ML063490295

License Renewal Project, Drywell Monitoring Program, Information for ACRS Subcommittee Reference Material, Volume 1

ML063490295
Person / Time
Site:	Oyster Creek
Issue date:	12/08/2006
From:	Wilson R GPU Nuclear Corp
To:	Zwolinski J Office of Nuclear Reactor Regulation
References
%dam200701, 5000-86-1116 NUDOCS 8612220272
Download: ML063490295 (632)

Text

Oyster Creek License Renewal Project Drywell Monitoring Program Information for ACRS Subcommittee Reference Material Volume I December 8, 2006 GPUHuclearr~tl Cpu0 Inuear 1r Parkway P;Rabppany.

New Jersey 07054-114.

4201 ) 263.6500 TELEX 136-482 Wfltei's Divec Dial Numit'e (201) 316-7246 December 18, 1986 S00O-80-1116 Mr. John A. Zvolinski, Director BUR Licensing Directorate 11 Division of BWR Licensing U.S Nuclear Regulatory .ommission Wasihington, D.C. 20555

Dear Mr. Zwolinski:

Oyster Creek Nuclear Generating Station Docket No. 50-219 Licensing No. DPR-16 Oyster Creek Drywell Containment On December I and December 10, 1986. the GPU Nuclear staff met with NRR to review certain facts, data and assessments related to measurements;showing, ...... , localized thinning of the Oyster Creek drywell. These measureieneis-.vee ofi" ,s.," , initiated by GPUN during the current refueling outage to'confirm'the" I. -- #' .. ....condition of the drywell containment vessel. This letter is a follow-up to the referenced two meetings and briefly summarizes the investigatiolis to date, the data obtained, our assessment of that data including a safety evaluation and future planned work.Background Data: Initial surveillance measurements, utilizing a UT probe, were made of the Oyster Creek drywell in the April/Hay time frame. The Initial measurements indicated containment plate condition and thickness consistent with the original desigin except for areas at the approximate elevation of the interior drywell floor directly opposite the exterior sand cushion and extending over.several bays. These early readings indicated apparent thinning due to loss of material on the exterior of the drywell down to thicknesses of about 0.95" compared to the as-fabricated thickness of 1.154". These early measurements led to an attempt to qualify the technique for painted surfaces and then to a much more extensive series of measurements.

The more extensive UT surveys confirmed the general corrosion wastage mentioned above and further indicated potentially highly localized Pitting with indicated shell thicknesses as small as .383". In order to confirm the adequacy ..d accuracy of the UTl measurements, to understand further the source of the highly localized UrF 8612220272 661218 00 PDR ADOCK 05000219 P PDR G11l'l IN%.( 'aL ~ kf(' k ,lht r .I .Ici ',. t -tq 1.r. I It q U.f"! *" -1 ? IN" ,"

Mr. John A. Zwolinski, Director December 18, 1986 Page Two readings, and to assess drywell containment below the level of the interior concrete floor, it was decided to take containment core samples in seven locations.

These samples were obtained early in December.

Based both on the UT measurements and on the examination of the containment shell samples, we have concluded:

A. The ultrasonic thickness probing of the drywell containment'has been ..confirmed to give accurate results with physical measurement of the'0lUg .thicknesses being consistent with UrT but, in general, about Ztgreatd-ti Therefore, the U1T measurements have been a conservdtive assessment of": thickness.

B. The highly localized UT measurements characterized as pitting are now believed to be Inclusions or laminations in the original plate. This Is based upon destructive metallurgizal examination of a containment core.C. The general areas characterized as exterior corrosion wastage have been verified.D. These broad areas of exterior corrosion seem to be localized at an elevation corresponding to the exterior sand cushion. Measurements of'drywell thickness below the level of the interior cnncrete floor (which were made by removal of the interior concrete at two locations down to a depth of about two feet, bay S and 17) show that wastage below the floor level is no greater than measured just above floor level. :lIhfactq"Tr'4'r measurements at the location where general wastage was Indicated abovez.the floor show the drywell below the floor to be about SO mils thicker than the immediately adjacent above floor area.As a result of removing core samples from the steel drywell, certain 6ther observations can be made. Where there was general corrosion, the sand cushion was wet. While metallurgical work on the corrosion films Is-still,'," ongoing, the films have a eliaracteristic of being magnetic, dark in color and exhibit chlorides throughout the oxide film and at the oxide to base metal interface.

The interface between the base metal and the oxide appears to be very sharply defined. In addition to the metallurgical work on the core samples and corrosion films, we have also removed samples of the backing sand and are subjecting those to chemical and other aualyses.

Results to date show high nitrates, chlorides and sulfates.

The source of the chemical species detected in the sand may be the insulating materials applied to the exterior of the drywell during construction, with contaminants carried by moisture to and in some manner concentrated in the .and bed or may be original sand contamination.

We have also attempted to culture samples of the sand and corrosion films to ascertain the possibility of microbiological activity.

Initial culturing shows an active presence of microbiological

..1.L -Mm=d Hr. John A. Zwolinski, Director December 18, 1986 Page Three species. These species have not been further defined and work is ongoing.We have concluded that the observed damage is not indicative of common forms of microbiologically induced corrosion.

Details of the UT measurements, metallurgical results, and chemical analyses are more fully sumarized in the attached GPU Nuclear Safety Evaluation.

Detailed backup is available.

Assessment:

Our ongoing assessment has concentrated on verifying the existing structural adequacy of the drywell, the source and form of the corrosive attack on the drywell, and source or sources of water In the sand cushion.With regard to the corrosion mechanism, our efforts have focused on either attack by aqueous films containing high levels of impurities or potential microbiological attack. We had separately made ground potential measurements which proved to be negative.

Drywell metallurgical samples from areas in which the underlying sand cushion was dry did not show unusual corrosive attack.With regard Lo the water source, we believe that the insulation materials and the gap between the drywell and concrete were wet during the construr.tion of the plant, anul we have confirmed that, in the time frame from 1980 to the present, we have seen periodically some moisture from the sand cushion drain---;'-

-pipes. This moisture was seen during times coincident'with the refoeling cavity being flooded. During the time frame from 1980 to the present, attempts were made to identify and repair potential leak paths. It is believed repairs to the connecting area between the upper drywell flange and the refueling cavity.in 1986 were successful.

It is possible water leakage could have been experienced during refuelings before 1980, but we have no record or any observations that would confirm or refute that postulate.

Based upon the observed wastage and experimental corrosion data, the corrosion rate that could be inferred would be IS mils per year with an upper limit corrosion rate of 50 mils per year. While our understanding of the corrosion mechanism is still not complete, our assessment of microbiological and concentrated chemical attack would not be inconsistent with the preceding inferences.

For safety review purposes, we have utilized what we believe is a conservative upper bound of 50 mils per year as future wastage allowance for the next operating cycle.Stnictural analysis of the capability of the Oyster Creek drywell containment was reviewed by Chicago Bridge and Iron, the original designer and installer of the vessel, with supplementary work performed by GPUN. This assessment shows that a shell thickness of 0.7" (actual averaged thickness equals 0.8"'), averaged over an area which could structurally respond to Mr. John A. Zwollnski, Director December 18, 1986 Page Four accident loads, results in stresses meting the original plant design bases, is consistent with the additional review conducted under the SEP program, and is consistent with applicable Section NE of the ASHE Section III code. This assessment was conducted assuming both the presence of the sand cushion as well as no residual structural support provided to the shell by the exterior sand cushion which is conservative.

To the 0.70" thickness would have to be added the appropriate corrosion allowance to ensure the structural analysis covers the future operational time frame of interest.

it iS 'Oiar conclusion the drywell meets the licensed structural Integrity requirements and that operation of the plant for the next cycle is consistent with License requirements.

The underlying Safety Evaluation again sumuarizes in more detail the basis for the above conclusions.

While the Safety Evaluation concludes that licensed safety margins are maintained for the plant during the next cycle, we intend to maintain an intensive effort to: A. Eliminate the source of any future water Incursions into the sand bed.B. Dry the moisture from the sand cushion and/or otherwise render corrosive"ack minimal.C. Continue the metallurgical and chemical investigations to determine, if possible, the exact cause of the attack.U. Further assess longer term corrective actions that may be appropriate.

E. Continue the 1IT shell thickness test program at future outages of opportunity including forced outages otherwise requiring drywell entry during the next cycle.If you should have any questions, please contact Mr. H. W. Laggart at (201)263-6205.V yours, it. F. Wilson Vice President Technical Functions!am M Nucer Technical Functions SafetylEnvimnmental Determination and 50.59 Review UNIT Oyster Creek 'luclear Generatinn 3tatton PAGE I OF , DOCUMENT NO. , SE No. Q0_02 4 3-1 7 el" S e(,fit applicable)

Rev. No. 0 ACT eIVT lTIi Tt'ýegSe RlecAyRe~t 9 .qion Type ofActrvwty Evaluation of Reduced Plate Thickness of the Drywell Steel Line (Modification.

procedure, test. experiment.

or document)1. Is his aclivitylocument listed in Section I or II of the matrices in Corporate Yes. -No Procedure 1OOO-ADM.1291 01?It the answer Ia question I is "no" stop here. (Section IV actnvitwti ocufffent, should be reviewed on a Caseby-cso basis to determine if thai procedure is applicable.)

Thais procedure is not applicable and no documentation is required It the answer a "yes" proceed to question 2.2. Is this a now activityfdocument or a substantare revwson to on activityklocu-Yes "No meno? (See Exhibit 3. paragraph

3. this procedure for examples of non-substantive changes)It Me answer to question 2 is *no" stop here. This procedure is not avolicable and no documentation is required.

if the answer is "yes' priceed to anserf all remaining questions.

These answers become the SaftylEnvironmental Deter-mination and 50S9 Review.3. Does this aclavdtylocument have the potential to adversely affect nuclear safety EYes ::No or sate plant operations?

4. Does the aWivity/document require revision of Me system/component descnp- EYes -No lIon in the FSAR or otherwise require revision of the Technical Specifications or any other Lcensing Basis Document?S. Does the activtlyidocurment require revision of any procedural or operating

-4, , descnphion in the FSAR or otherwise require revision of th Technical

.......*.Specifications or any other Licensing Basis Document?6. Are tests or experiments conducted which ate not described in the FSAR. the "Yes XNo Te :hnical Specifications or any other Licensing Basis Document?7 Does lhis document involve any potenlial Non-Nuclear environmental impact? --Yes ENo 8 Does the activty/ldocument require a review of criteria as outlined :m SDD. -Yes XNo TIO00. TMI.1 Division I Plant Level Criteria?It yes. ientty TRJTFWR.If any of the answers to questions

3. 4. 5. or 6 are yes. proceed to EXHIBIT 6 and prepare a written safety evaluation, If the answers So 3. 4. S. or 6 are no. this precludes the occurrence of an Unrevew*d Safety Ouestion or Technical Specifications change. t1 the answer to question 7 is yes. either redesign or provide supporting documentation which will permit Environmentat Licensing to determine a an a=verse environmental impact exists and of regulatn%

approval is required (Ref. LP-010). If in doubt, consult the Radflogical and Environmental Controls owis- -or Environmental Licensing for assistance in com-pleting the evaluation.

I)I ..',r.t.a *...*Signatures-See attached si'nn-nff sheet Oato Section Manager Responsible Technical Reviewer ¶Other Reviewwris) 8612?220289 866121 PDR ADOCK 05000219 P PDR 0 0 ]Nuclear Technical Functions_ Safety Evaluation UNIT Oyster Creek Nuclear 1 eneratinq Station PAGE 2 OF L .SE No. 000243-002 Drywell Steel Shell P1Ite Thickness Reduction at the Rev. N. 0 ACTMTYriYDOCIMENT TrriP gase San auushion Entrenchrient Document No.Region (if applicable)

Type of AcgoDoument Evaluation of Reduced Thickness of the Drywell Steel -(Modificaion.

procedure est. experiment, or document)

Liner..Th's Safty Evutim provides the basis for determining whether this actiMtykocument involves an Unreviewed Safety Ousetion or impacts on nuclear safety.Answe the following questions and provide meaon(s) for each answer per Exhibit 7. A simple satment of conclusion In bie Is not suffIcient The scope and depth of each reason should be commensurate with the say sIgnificance and complexity of the proposed change.1. Is the margin of safety as defined in Ucensing ausis Documents other than th Technical Specificalio reduced? C.s LNo 2. Will m of the ectlvitydocumnt adversely af~ nuclear safety or sato plant openrion =Yes ENo The following questions comprise the 60,59 consideratbons and evaluation 1o determine if an Unreviewed Safety Ouestmon exists: 3. Is the probability of oe ce or the consequences of an accident or malfunction of equipment Important to safety pmWvouW evaluated In '. , Mr the Saf"t Analysis Report increased?

LAV ~6 *~" 4. Is the possibility for an accident or malfunction of a different type tw a evaluated provkiuly in the Safety Analysis Report created? C-Y ENo I, Is the margin of safety as defined in the basis for any Technical Specification reduced? CYes ENo If any anwer above I "y an Impact on nuclear safety or an. UnreviewedlSafsty non 1 '." exists. If an adverse Impact on nuclor safety exists revise or redesign.

It an unreviewed safe.ty question with no Adverse impact on nuclar safety exists forward to Lcensing with any ad-ditional documentation to support a request for NRC approval prior to Implementing approval.6. Spe*ty whether or not any of the following are required, and if "yes" indicate how it was resolved Yes No a. Does the activityfiocument require an update of the FSAR?Explain: Yes; an analysis to suoport the new drvwell shell thickae~sbe included in the Final Safety Analysls Reoort Section 3.1.b. Does the activity/document require a Technical Specification Amendment?

Explain: 11o. the ,iinimum ý1ipll thickness found du,'inn :he insoe'tionn reets the des'in criteria snecified in the WCAIS Technical Snprif.ca:i ns WSW" 1104M SAFETY EVALUATION NO. 000243-002 Page 3 of 18 Preparers:

Siqnature M.D.P.L.S.R.Y.Laqqart Jerko Huebsch Garibian Giacobbe Greenwood N aqai Date L , -&Responsible Technical Reviewers:

S.G.M.Leshnoff VonNieda Sanford 12 -ii- R, Independent Safety Reviewer J. R. Thorpe SMFETY NO. 000243-002 Page 4 ot 163?ropfee I M. Lagsart D. Jerko P. Huebach L. Caribian 8. Ctacobbe 3. Creenvood 1. Nagai* tgU4tUtS Dato ljeuLusble Technical Rovieverst S. LtAhnoff C. Vonmieds M, Sanford 67 FIt a w I Independent Safety Raview r J, R. Thorpe SE No. OC0243-002 Rev. 0 Page 5 of 18 TABLE OF CONTENTS Purpose Systems Affected Effects on Safety Effects on the Environment Conclusion Attachments:

1. Description of Drywell Design 2. Extent of Damage 5. Causes of Corrosion and Corrosion Rate 4. Structural Analysis Section 1.0 2.0 3.0 4.0 5.0 SE No. Rev. 0 Page 6 of 18 1.0 PURPOSE The purpose of this safetj evaluation is to assess the structural integrity of the Drywell steel pressure vessel In light of a recent (Inspection) finding that sections of the drywell shell near the base sand entrenchment region have a thickness which Is below the thickness utilized In the original stress report prepared by Chicago Bridge & Iron Company ("Structural Design of the Pressure Suppression Containment Vessels", for JCPL/Burns

& Roe, Inc.. Contract No. 9-0971, by CB&I Co., 1965). In addition, this evaluation provides a justirtication for operation up to the end of the l1th operating cycle (18 months) for the Oyster Creek Nuclear Generating Station (OCNGS).2.0 SYSTEMS AFFECTED 2.1 System No. 243. Drywell and Suppression System, particularly the drywell sheil structure.

This structure Is directly affected by the localized thinning.2.2 Drawings showing original thickness

-Chicago Bridge and Iron Co..Contract Drawings 3-0971. Drawings #1, 2, 3, 4. 5. 6. 7. 8. 9. 10.11.2.3 Documents that Describe the Drywell Structure are listed below.2.3.1 Amendment

1. 15 to OCNGS FDSAR. Primary Containment Design Report.2.3.2 Updated FSAR. Paragraph 3.8.2.2.3.3 0CNUS Technical Soecification Section 5.2.2.3.4 CB&I Stress Report. "Structural Design of the Pressure Suppression Containment Vessels" for JCPL/Burns

& Roe. Inc..CB&I Company Contract No. 9-0971. 1965.

SE no. %C*243-0C2 Rev. 0 Page I of 18 3.0 EFFECTS ON SAFETY 3.1 Identificat'on of Documents 3.1.1 OCNGS Unit 1 racility Description and Safety Analysis Report,-Licensing Application, Amendment 3,Section V.-Licensing Application.

Amendment

11. Question 111-1S-Licensing Application.

Amendment 15-Licensing Application.

Amendment 68 3.1.2 Technical Spe:ification Documents 3.:.2.1 Technical Specification and Bases -OCNGS Unit, Appendix 4 to Fi'ility License DRP-16, JCP&L Docket No. 50-219. Sections 3.5. 4.5. 5.2 3.1.3 Regulatory Documents 3.1.3.1 IOCFRSQ. Appendix A, General Design Criterld for Nuclear Power ;:ants-Criterion 2 -Design Bases for Protection Against Natural Phenomena-Criterion 4 -Environmental and missile Design Bases-Criterion 16 -Containment Design-Criterion 50 -Containment Design Basis 3.1.4 :rdustry Codes and Standards 3.1.4.1 ASME Boiler and Pressure Vessel C:.,e.Section VIII, 1962 &itn Code cases 1270N-5, 1271N, and 1272N-S Ccae cases. Section II1. Div. 1. Subsection NE 3.2.4.2 See Attachment 1 for additioral codes and standards.

SE No. 000243-00Z Rev. 0 Page 8 of 18 3.2 Drywell Containment Structure 3.2.1 Attachment I provides a description of the Oyster Creek Oryweil Geometry, Design Bases, Materials.

Shop and Field Fabrication and Testing, and Concrete Interfaces.

3.2.2 Extent of Drywe1 1Thinning Background Information on the source of the sand cushion wetting, UT techniques, drywell thickness measurements, and core sample locations are I-:luded in Attachment

2. Based on Information contained In Attachment 2, the following conclusions can be stated: A. The ultrasonic thickness probing of the drywell containment has been confirmed to give accurate but conservative results. The physical measurements of the thicKnesses of the plugs were approximat:ly 0-4% greater than that determined by UT results.B. Destructive metallurgical examination of ore of the containment plugs verified that the highly localized UT indication was an inclusion and that pitting did not exist.

SE No. 000243-C02 Rev. 0 Page 9 of 18 C. The general areas chara:terized as broad exterior corrosion have been verified to be general wastage.D. Thes' broad areas of exterior corrosion are localized at an elevation c6rresponding to the exterior sand cushion.Measurements of drywell thickness below the level of the interior concrete floor (which were made by removal of the interior concrete at two locations) show that wastage below the floor level is no greater than that measured just above the floor level. Measurements at the two locations show the drywell below floor level to be slightly thicker than the immediately adjacent area above the floor area.E. The drain line gasket was found to be leaking and was replaced.

Leak tests were performed on the bellows. and":-'no leaks were detected.

Observations of the areas where leakage had previously been found Indicated that the leakage had been arrested.F. Based on the conservative methodology utilized in Attachment

2. the effective drywell thickness at the sand entrenchment region has a mean value of 0.87. This value exceeds the minimum required shell thickness calculated for structural stability and Integrity. (See Attachment

4)

SE No. 000243-002 Rev. 0 Page 10 of It 3.2.3 Drywell Corrosion Mechanism and Rate A review of the potential causes of corrosion and a conservative prediction of a future corrosion

!,-te is Included in Attachment

3. Based on Information contained in Actachment 3 the following conclusions can be stated: A. In all cases where general corrosion was present, the sand cushion appeared to be wet.B. No deep pitting was observed and no sulfide or substantial concentration of manganese was detected in the corrosion product. This indicates that microbiological influenced corrosion Is minimum..C. The corrosion observed c,n be explained by an aqueous corrosion mechanism assuming chloride contamination and oxygen depletion. " , ".D. A conservative corrosion allowance rate of 48 mils per year will account for any uncertainties in the assumptions of the corrosion mechanism.

3.2.4 Structural Attachment 4 provides an assessment of the Drywell structural capability assuming a reduced shell thickness of .7 Inches within the sand entrenchment area for two critical load combinations.

Conclusions which can be made from this assessment are:

SE No. 000243-002 Rev. 0 Page 11 of 18 A. The original 4llowable stress criteria of ASME Boiler and Pressure Vessel Code.Section VIII. 1962 with appropriate Code cases is met when credit is taken for the radially inward reaction due to the sand. Without the sand. (a beyond design basis condition) code allowable stresses'ire

exceeded by 2.7% with a reduced shell thickness of 0.7 Inches at the sand entrenchment region. However. ASME Sect. III. Oiv. 1. Subsection NE allowable stress criteria are met without exception using stress Intensities.

While peak local membrane stresses are less than the allowable.

the meridional extent of these is more than allowed by Section III (but 2X). The or'ginal Code placed no bounds on the extent of a local stress. It Is reasonabltni

.to neglect th;, departure from present Code guidance because the present situation is an In-service condition.

and not a design conL'tIon, and because the departure from present Code guidance Is ;mall.B. The load combinations selected for this analysis represent the design basis accident condition.

3.3 Effects o" Thickness Reduction on the Safety Function of Drywell Containmen, Structure (DCS)3.3.1 Structural Performance The reduction In thickness of the drywell shell at the sand entrenchiment region does not prevent the structure from performing its Intended safety function.mm=mid SE No. 000243-002 Rev. 0 Page 12 of 18 3.3.2 Quality Standards Repair of the core samples taken were made in accordance with the Quality standards of the plant.3.3.3 Natural Phenomena Protection Since the DCS is protected from the outside elements by-a safety class structire capable of withstanding a tornado'or hurricane, and since the plant elevation prevents natural flooding, these loadings do not contribute to the concerns posed by this activity.

However, in the evaluation of structural performance, seismic loads were Included and found that this event does not affect the integrity of the OCS when the event occurs singly or in combination with other design loads.3.3.4 Fire Protection

• , The thinning of the drywell shell does not affect the fire protection program for the plant, since the drywell was not considered as one of the fire protection measures.3.3.5 Environmental Qualifications The assumptions utilized in complying with IOCFRSO.49"Environmental QLalification of Electrical Equipment Important to Safety for Nuclear Power Plants" have not been altered.therefore there is no effect on Environmental Qualification.

SE No. 000243-002 Rev. 0 Page 13 of 18 3.3.6 Missile Protection The affected area Is protected by a concrete shield wall as described in Section 3.2.1 and by the Reactor Building which provides protection from external missiles.3.3.7 High Energy Line Break: Internal Flooding The maximum pressure Inside the DCS after a high energy line break has been conservatively assumed to be 62 psig.Subsequent evaluation of the affected area considering this pressure increase together with SSE and deadloac shows that DCS structural Integrity is stiil maintained.

3.3.8 Electrical Separation The reduction In thickness of the affected area does not Impact any electrical components.

3.3.9 Electrical Isolation

.1 **',... " The reduction In thickness of the affected area does not Impact any electrical components.

3.3.10 Electrical Loading Impact on Emergency Diesel Generators and Safety Buses.No effects per explanation 3.3.9.3.3.11 Single Failure Criteria No effects on single failure criteria since the structural Integrity and stability of DCS Is assured.

SE No. 000243-002 Rev. 0 Page 14 of 18 3.4 Licensing Basis Documents Margin of Safety Review of the FDSAR requirements as to the structural Integrity of the DCS during all modes of plant operation reveal that the minimum thickness of the affected regions still have ample margin of safety to satisfy Technical Specification 5.2 and the intended design as stated In the FOSAR. This was ascertained after reanalysis (see Attachment

4) of the structural response to the most severe load combinations considering the minimum thickness of the affected area.3.5 Nuclear Safety/Safe Plant Operation Since the structural integrity and stability of the DCS have not been affected by the thinning of the affected regions of the shell, and the corrosion rate determined will not degrade the structural integrity and stability of the DCS during cycle 11. nuclear safety and safe plant operation will not be affected.

The thlnningts-.

.limited to the area described In this evaluation; no'evidence of damage to other drywell areas or other safety related equipment was found.3.6 Probability of Occurrence or Consequences of an Accident Since the structural Integrity 6nd stability of the OCS Is still maintained, the minimum thickness of the affected shell region of the OCS will not affect the probability of occurrence of any accident when the plant Is In any mode of operation or plant condition.

Furthermore, since the containment isolation function of the DCS is Intact. the consequences of any postulated accident at O.C.N.G.S.

will not be affected.

SE No. 000243-002 Rev. 0 Page 15 of 18 3.7 Probability of Occurrence or Consequence of Malfunction of Safety Equipment The fact that the structural Integrity and stability of the DCS has not be~n affected by the condition, the probability of occurrence or consequence of a malfunction of safety equipmehtin the plant will not be affected.3.8 Possibility for an Accident or Malfunction of a Different Type Than Any Previously Identified In FDSAR.Since the DCS still meets design requirements no accident or malfunctions are different from what have been previously identified.

3.9 Margin of Safety on Basis of Technical Specification The thickness of the affected region of the shell has been ascertained to satisfy the original allowable stress criteria of-ASHE Boiler and Pressure Vessel Code,Section VIII, 1962' with apprepriate Code cases when credit Is taken for the radially Inward reaction due to the sand. Without the sand, (d oeyond design basis condition)

Code allowable stresses are exceeded by 2.7%. However. ASME Sect. III, Div. 1, Subsection NE allowable stress criteria are met without exception.

While peak local membrane stresses are less than the allowable.

SE No. 00024.,-002 Rev. 0 Page 16 of 18 the merldional extent of these Is more than allowed by Section III (but 2X). The original Code placed no bounds on the extent of a local stress. It Is reasonable to neglect this departure from present Code guidance because the present situation is an In-service condition and not a design condition, and because the departure from present Code guidance Is small.3.10 Violation of Plant Technical Specification The minimum thickness at the affected regions does not violate any section of the OCNGS Technical Specification.

As stated In Section 3.9, the allowable stress criteria Is satisfied.

3.11 Violation of Any Licensing Requirements or Regulations Review of OCNGS Licensing requirements and commitments reveal that the thinning of the drvwell shell does not violate any of Licensing requirements or regulations.

This Is primarily due to the fact that containment Isolation function and the structural integrity of the DCS have not been affected.3.12 Radiological Safety Concerns The reduction In thickness of the drywell shell will not affect any radiological safety concerns because the containment isolation safety function of the DCS Is still Intact. The drywell shell In the area of concern is within the biological shield, and adequate shielding of occupied plant areas will be maintained.

SE No. 000243-002 Rev. 0 Page 17 of 18 3.13 Change to FSAR This condition will require a change to the FSAR to reflect the change In the plate thickness, and the results of the analysis which support this evaluation.

3.14 Change to Established Practice or Procedure This condition will not require any change to an established practice or procedure.

4.0 EFFECTS ON THE ENVIRONMENT 4.1 Changes to Plant Environmental Interface The reductinn In thickness of the affected shell area will impose no changes to the OCNGS plant environmental Interfaces, because the structural integrity and stability of the DCS is still Intact.4.2 Potential Environmental Impact Since the activity does not affect the environment, It does not have any potential Impact to the following:

A. Environmental Technical Specification

6. Applicable Environmental Permit Requirements C. Final Environmental Statement D. Environmental Impact Statement Consequently, no additional evaluation Is required.5;0 CONCLUSION Recent findings revealed that sections of the drywell shell near the base sand entrenchment region have a mean thickness of 0.87 Inch.This Is less than the original thickness that was utilized SE No. 000243-002 Page 18 of 18 In the evaluation of structural stability and Integrity In support of Licensing the OCNGS. Extensive review of the original calculations.

load combinations and different plant conditions, and new calculations generated to evaluate the structural stability and integrity of DCS show that: 1. The structural performance of the DCS during the most severt'plant condition (DBA) will not be affected.

The arginsf 1 '6f safety found are more than enough to assure structural stability and Integrity of the DCS.2. The containment Isolation safety function of DCS Is still Intact.Consequently, no environmental or radiological concerns exist due to the reduced thickness.

3. FSAR and Technical Specification Commitments have not been violated.4. Plant Procedures and Safe Practices ire not affected.

"' ...5. The corrosion rate determined for will not degrade the structural int.grity and stability of the drywell during cycle I1.6. Based on Sections 3.6, 3.7. 3.8, and 3.9, there does not exist an unreviewed safety question as defined In IOCFRSO.S9.

SE No. 000243-002 Att. 1-1 Attachment I DESCRIPTION OF DRYIELL DESIGNI Primary Containment Geometry The primary containment consists of a pressure suppression system with two large chambers as shown in Figure 1. The drywell houses the reactor vessel.the reactor coolant recirculating loops, and other components associated with the reactor system. It Is a 70 ft. diameter spherical steel shell with a 33 ft diameter by 23 ft high cylindrical steel shell extending from the top.The pressure absorption chamber is a steel shell In the shape of a torus located below and around the base of the drywell.The two chambers are Interconnected through 10 vent pipes 6 ft. 6 in. In diameter equally spaced around the circumference of the prelsOre absorptlon chamber. The two chambers are structurally isolated by expansion bellows In the Interconnecting piping and analysis for each unit may be considered Independently.

The drywell interior is filled with concrete to elevation 10 ft. 3 in. to provide a level floor. Concrete curbs follow the contour of the vessel up to elevation 12 ft. 3 In. with cutouts around the vent lines.bn the exterior, the drywell is encapsulated in concrete of varying thickness from the base elevation up to the elevation of the top head. From there, the concrete continues vertically to the level of the top of the spent fuel pool.

SE No. 000243-002 Att. 1-2 The proximity of the concrete surface to the steel shell varies with elevation.

The concrete Is in full contact with the shell over the bottom of the sphere at Its Invert elevation 2 ft. 3 in. up to elevation 8 ft. 11 1/4 In. At that point, the concrete Is stepped back 15 Inches radially to form a pocket which continues up to elevation 12 ft. 3 In. That pocket Is filled with sand which forms a cushion to smooth the transition'of the shell plate from a condition of fully clamped between two concrete masses-t6 Aifrie'* ' .:'standing condition.

The sand pocket Is connected to drains provided to allow drainage of any water which might enter the sand.Above elevation 12 ft. 3 in. the concrete Is stepped back 3" measured radially from the steel shell. This gap was created during the construction by applying a compressible, Inelastic material to the outside of-the shell-prior.

to concrete placement.

The material was later permanently compressed by controlled vessel expansion In order to create a gap between'thO vesiel Undo',',";' 'the concrete.

.Drywell DesiLn Bases Design codes used for the original design are as follows with the effective dates at the time of design:

SE No. 000243-002 Att. 1-3* ASHE Boiler and Pressure Vessel Code. Sections VIII and IX with all applicable addenda In effect at the time of design.* Nuclear case interpretations 1270 N-S, 1271 N and 1272 N-5.* ASME Boiler and Pressure Vessel Code. Section I1 with all applicable addenda for the following material SA-212 High Tensile Strength Carbon -Silicon Steel Plates for Boilers and Other Pressure Vessels SA-300 Steel Plates for Pressure Vessels for Service at Low Temperatures SA-333 Seamless and Welded Steel Pipe for Low Temperature Service SA-350 Forged or Rolled Carbon and Alloy Steel Flanges. Forged Fittings, and Valves and Parts for Low Temperature Service ASTH A-36 Structural Steel* AISC Specification for the Design, Fabrication and Erection of Structural Steel for Buildings Pressure and temperature parameters in the original drywell design Include: o Drywell and connecting vent system tubes are designed for 62 psig Internal pressure at 175OF and/or 35 psig at 2814F. and an external pressure of 2 psig at 2050F.* In addition, the drywell Is designed to withstand a local hotspot temperature of 300OF with a surrounding shell temperature of 150F concurrent with the design pressure of 62 psig.* The lowest temperature to which the primary containment vessel pressure containing parts are subject to while the plant Is In service is 50"F.To provide an additional factor of safety. 30"F was actually used for the design basis.* During reactor operation, the vessel will be subjected to average temperatures up to 150*F at approximately atmospheric pressure.

SE No. 000243-002 Att. 1-4 Loadings considered in the design of the drywell Include* Loads caused by temperature and internal or external pressure conditions." Gravity loads from the vessels, appurtenances and equipment supports.* Horizontal and vertical seismic loads acting on the structures" Live loads* Vent thrusts J Jet forces on the downcomers W Water loadings under normal and flooded conditions W Weight of the contained gas In the vessels* The effect of unrelieved deflection under ,emporary concrete loads during construction.

• Restraint due to compressible material* Hind loads on the structures during erection SE 000243-002 Att. 1-5 Load combinations used the design of the drywell and vent system for accident conditions include:* Gravity load of vessel and appurtenances

• Gravity load from equipment supports* Gravity load of compressible material* Gravity load on welding pads* Seismic loads" Design pressure:

maximum positive pressure of 62 psig at 175'F decaying to 35 psig at maximum temperature of 2816F, to maximum negative pressure of 2 psig at 205"F.* Restraint due to compressible material* Vent thrusts J Jet forces Allowable stress levels used In the design of the drywell are based on Code Case 1272 N-5" General membrane (does not Include thermal) -19250 psi" Local membrane (does not include thermal) a 28875 psi" Surface stress -52500 psi SE 0002 3.-O2 Att. ,-'Drywetl Materials of Construction Steel plates are A-212-61T, Grade "B". made to ASTM A-300 requirements.

Minimum charpy vee notch Impact test values of 20 ft.-lbs. at OF were used instead of 13 ft.-lbs. at O*r as permitted by Code Case 1317. Test specimens were taken both parallel to and transverse to the direction of finat rolling of the plate.Forgings are A-350 Grade LFI. Minimum charpy vee notch Impact test values%ere 13 ft.-lbs. at OOF In addition to charpy keyhole impact test values required by the Burns and Roe specifications.

Pipe Is A-333, Grade "O" seamless.

Minimum charpy vee notch Impact test values were 13 ft.-lbs. at OF on full size test specimens In addition to charpy keyhole Impact test values required by the Burns andRoefpecifi-atlons.

Miscellaneous plate and structural steel (not within the scope of ASTH A-36): All permanent structural attachments and lugs, welded to the shells, were made of impact tested material for a distance of not less than 16 times the plate thickness.

The erection skirt supporting the drywell was also made of Impact tested material.

SE 000243-OC2 Att. 1-7 Drywell Shop Fabrication and Testing Components were shop welded, where possible, into large size shipping pieces, utilizing either submerged or metallic coated arc techniques.

In either case.low hydrogen electrodes were used, thus assuring the notch toughness requirements to meet the ASHE Code impact Tests.All seam welds In the shell of the containment were of the double bevel butt type. All butt welds in any accessories subject to the ASME Code were also of the double welded type or equivalent, and all the joints were full penetration welds. All welds subject to the Code were radiographed or otherwise examined in accordance with Code Case 1272 N-S. All mandatory provisions of this code were followed and all recommended provisions were also followed where practical.

Heavy weldments and penetration weldments were furnace stress relieved as follows: a. Any plate segment wholly containing a penetration, nozzle. r'olVm-connection was furnace stress relieved at the shop after Insertion of the penetration.

b. All large penetrations intersecting more than one shell plate were stress relieved as follows. Any portion of a penetration containing

"- 9s joining metal over I II/ In. thick at the joint was furnace stress relieved as a unit before welding Into a penetration assembly cr into the shell.

SE No. C00243-COZ Att. 1-o In keeping with the abo~e. the vent line penetrations were shop assembled to the reinforcing collar and the completed assemblies were stress relieved.

The weld between the collar and the shell plate was made in the field and was not stress relieved.All shop welds were radiographed in the shop. All welds In those parts of the work subject to the ASME Code were radiographed by methods complying with Paragraph UW-S of thi code.Prior to shipment, all materials were cleaned and painted. Surface preparation and painting 'was in accordance with the paint manufacturer's recommendations.

The Interior of the drywell above the concrete floor, Including jet deflectors and the exterior of the drywell above the water seal support bracket received one coat of Carboline Carbo-Zinc

11. The interior of the drywell below elevation 8 ft. 11 1/4 in. and the exterior surface of-the drywell adjacent to concrete surfaces at completion of construction were not coated. All other surfaces of the drywell were given one coat of Carboline primer.After erection and testing, all field welds and abraded places on the shop paint were cleaned by sandblasting and painted as noted above.Drywe~l Field Fabrication and Testing During field fabrication the drywell steel was supported on a steel skirt of approximately 39 ft. diameter with its base plate at elevation-Q ft. 1 in. and Invert of the sphere at elevation 2 ft. 3 In.

SE No. 000243-002 Att. 1-9 The 70 foot diameter spherical drywell and upper cylinder were field assembled and welded. The transition knuckle and top head flanges were field stress relieved iH accordance with the ASME Code.The heavy plate flanges for the 33 foot diameter cover and neck flanges of the drywell were subassembled in segments, welded, x-rayed and stress relieved as complete units.All completed shell plate assemblies, with penetrations installed, were stress relieved after fabrication.

All butt welds were 100% x-rayed. Other welds wh!ch could not be 100% x-rayed were magnafluxed before and after stress relieving.

Upon completion of fabrication of the drywell and pressure absorption chamber, acceptance testing was initiated.

This included soapsuds testing at S psig, a'.',: holding period at 40.25 psig and a second soapsuds test at the design pressure of 35 psig. This was followed by the overload test at a pressure of 71.3 peig which corresponds to a 115 percent overload.

The procedures for the overload test fulfilled the requirements of Section VIII of the ASME*Code and Code Case, 1272 N-5.At the time of the tests, the downcomers.

designed to pass the released steam and gascs from the drywell into the suppression

hamber were capped in order that a separate test could be conducted on each vessel. The drywell was tested with no pressure In the suppression chamber. The suppression chamber.however, was tested witt a balancing pressure in the drywell to avoid an excessive e,ternal pressure on the vent lines and header inside the suppression chamber.U, 1111 SE No. OWO'43-002 Att. 1-10 Drywell/Concrete Interfaces The drywell shell is designed as a free standing structure and, with the exception of concentrated jet forces. will resist all required loads without interaction with the surrounding concrete.

The function of the concrete Is to act as a radiation shield, provide a "back-up" to limit deformation due to concentrated jet forces and to form a support at the base of the tphere.At the base of the sphere, subsequent to completion of pneumatic testing, the volume Inside the skirt was filled with concrete while simultaneously pouring the concrete floor inside the bottom of the shell. The concrete pour outside the vessel proceeded in full contact with the vessel up to elevation 8 ft.11 1/4 in. where the concrete line was stepped back radially 15 Inches. This gap continues up to elevation 12 ft. 3 In. At points on the perimeter of the vessel where the vent lines penetrate the concrete, the forms were set back around the vent lines to provide clearance which would prevent Contact'betwekn the vent lines and the concrete surface during any design condition.

The 1S inch radial gap was filled with sand to provide a cushion for the shell plate during the transition from clamped between two concret.surfaces to'freo standing.At all elevations above the sand layer, the external concrete mass Is set back from the surface of the stee' vessel an amount calculated to allow unimpeded expansion of the steel shell during any design condition.

The gap was created by applying a compressible.

Inelastic material to the exterior surface of the I.mIuI.flEE U ~

SE No. 000243-OOZ Att. 1-11 vessel prior to pouring concrete.

The material properties were chosen to provide resistance to crushing by the pressure Induced by the head of concrete, but of low compressive strength to allow collapsing by Induced vessel expansion.

The criteria for maximum gap was established to limit the deflection of the vessel wall due to local Impa:t of jet forces. The criteria used was that the space between the steel drywell vessel and the concrete shield outsida must be sufficiently small that, although local yielding of the steel vessel may occur under concentrated forces, yielding to the extent causing rupture would be prevented.

Using this criteria, the formed gap was 2 inches from elevation 12 ft. 3 in. to elevation 23 ft. 6 In. Above 23 ft. 6 in. the formed gap was increased to 3 inches. This dimension allowed for Inelastic compression due to concrete pressure during the pour an,. residual thickness of gap material after compression by controlled vessel expansion.

........The criteria used for selection of t:e gap material was as follows: " It must adhere tightly to a curved, painted steel plate surface in flat.vertical and overhead positions.

• Could have relatively Insignificant deformation under fluid pressure cf wet concrete estimated at 3 psi.* W3uld be reduced In thickness inelastically hP !bout one inch from an initial thickness of 2 to 3 inches under a pressure of not more thin 10 psi.

SE No. *OO243-OO2 Att. 1-12" Hould remain dimensionally stable at the reduced thickness without significant flaking or powdering" Would be unaffected by long term exposure to radiation and heat* Should be susceptible to minimum damage which exposed on the vessel before concrete placement.

The 2 inch gap was formed using Owens-Corning Fiberglass SF Vapor-Seal Duct Insulation.

The material was supplied with a factory applied laminated asphalt kraft paper waterproof exterior face, and was attached to the vessel with mastic and Insulation pins. Joints between the boards, and edges and penetrations were sealed with glass fabric reinforced mastic.Tne gap material used above elevation 23 ft. 6 In. was Firebar-D, a proprietary asbestos fiber -magnesite cement product applied as a spray coat. The solid materials, asbestos fibers, magnesite 4ad magnesium sulphate (roughly 75% asbestos), were premixed and combined in a mortar mixing machine with water and. to control density, with foam to form a slurry suitable for spray application.

After application and curing, the material surface was faced with polyethylene sheets with all edges sealed by tape and held In place by Insulation pins. The posyethylene sheets formed the bond-breaker for the cr-rete pour.

SE No. 000243-002 Att. 1-13 Gap Formation and Results At the most critical location.

drywell expansion at 281°F and 35 psig was expected to be approximately 0.7 inches. Considering an allowance for material rebound. it was calculated that the required vessel expinsion could be achieved by raising its temperature 140"F above ambient. Concurrent with induced thermal loading, an internal pressure was created to balance the shell external compressive forces Induced by the crushing of the gap material; An internal pressure of 40 psig was calculated as appropriate for this function.and considering the expansion induced by internal pressure, the temperature differential was reduced from 140"F to 130"F.After placement of the gap material on the drywell shell, concrete placement continued in a staged schedule to complete encasement of the drywell. The vessel was then expanded to create the required air gap required for thermal and pressure expansion.

Expansion of the vessel was monitored via use of pairs of extensometers at 7 points around the exterior of the vessel at locations of penetrations.

The extensometers were read and recorded hourly and the reodlngs compared with calculated theoretical values. while the horizontai movements were In good agreement with calculated values, the upward accumulation of expansion expected due to the embedment of the lower region was at all points less than predicted.

Therefore, vessel discontinuity stresses at the embedment would have been less than calculated and the load on the concrete wall would havP been more uniformly distributed and with a lower maximum.

SE No. 000243-002 Att. 1-14 During the expansion.

It was noted that the gap material "ad entrapped moisture due to Incomplete curing and Introduction of water from external sources. This was evidenced by appearance of water at sleeves around several penetrations.

This was deemed to be of no practical significance since the moisture's effect on material compression Oiaracterist!'cs would .e a moderate Improvement through a slight reduction in strength and a lesser rebound.

T*t .* I C...

• n.e i Finure 1 Orywell Containrent Structure (tLl CIM (GO~ M UL v-1110 TCL zv, rt!ramss blij W r-womU. ttVw5 TO9040 F11IRR"R-0 PtitptPRoPf)r MW~L, EL 2'6(0~~-u 1 L Ifni '*476%IsTORUS 41 or cxf"TIMAA na .im V.Di SE No. *.XC'-24'3-5X2 Att. 2-1 Attachment 2 Extent of Daman EXPECTED SOURCE OF SAND CUSHION WETTING During the 1980 Oyster Creek plant outage, water was found leaking from various locations from the concrete surrounding the drywell. Containment penetration X-46 (Elev. 86'-0") on the scuth west, and penetration X-50 (Elev.47'-0") on the north east were reported to have water leaking from with'.i the concrete biological shield. These ide-tified areas correspond to Bays 7 and Bays 17 & 19, respectively.

In additicn it was reported that water was coming from the sand cushion drain lines in Bays 3, 11. and 15 into the torus room.Efforts were made to identify the source of the wate, and its leak path. The leakage was found to have the same range cf radioactivity as that within the reactor. The leak path for the water was believed to have been from the reactor cavity located immediately above the arywell. This cavity is filled with water during refueling operations.

It was believed that a leak from this cavity through the bellows seal at the bottcm drained to the space between the dry~tell anj the surrounding concrete (i.e.. the space filled with insulation).

The volume below the bellows was pressurized with service air and the bellows checked for bubbles. Another leak test was performed by injecting helium behind the bellows and the beitows sniffed. The results of these tests were negative.

The 2 inch reactor cavity drain line that includes a fletible pipe iecticn ads also tested with no significant leakage deteCEtd.Plans were made during the follcwing operating cycle to locat, and seal any potential leak path from the reactor retueling cavity.

SE No. 000243-002 Att. 2-2 During the 1983 outage the welds of the refueling cavity were leak tested.Some minor lepks were detected and repaired.

The bellows area between the containment and the refueling cavity was cleaned to remove contaminants.

The area was then Inspected and attempts were made to apoly various pressure tests to the bellows, however, no leaks were detected.

Also. during the 1983 outage the water lefel was dropped to the lowest reactor cavity shi-eld plug step.. At this time it was observed that leakage from penetrations X-46"4nd'X4S6-stopped. Furthermore.

leakage into the torus room had diminished.

Three of the four shield plug steps were inspected via liquid penetrant for the full circumference:

no indications were detected.

The single draln line used to detect leakage from the refucling cavity was suspected of being restricted.

A restriction In this line would cause any leakage to be directed Into the area between the containment and biological shield. This drain line'was purged with air and did not appear to have any flow restrictions.

When the refueling cavity was filled, similar leakage was found as previou'yde4vibed.4iowever;d' f v'.it had been reduced appreciably.

During the cycle 11 outage outage the drain line from the refueling cavity was Inspected.

Drain line gasket (30"x;") was found to have leaks. and It was replaced.

Leak tests were performed on the bellows, anJ no leaks were detected.

Observations of the areas where leakage had previously been found indicated that the leakage had been arrested.

SE No. 000243-002 Att. 2-3 DRYWELL THICKNESS HEASUREMENTS Because of these wetting conditions, there was concern that repeated exposure of the drywell steel to water could result In degradation of the drywell.measurements of the drywel. portion of the containment shell were made to verify its thickness during the 11R outage. These measurements were made using UT, a Non Destructive Examination (NDE) method, that Is able to accurately determine the thickness of material or presence of abnormalities.

i.e.. nonmetallic Inclusions.

• UT plate thickness measurements were made on the Oyster Creek drywell.Approximately 1.000 UT readings were eventually taken utilizing an ultrasonic thickness gauge device (D-meter).

Measurements were obtained by transmitting ultrasound through the plate and measuring the time it takes for the longitudinal wave mode to travel to a reflector (front'wa&l ttt~fattdr,'

mid-wall reflector or backwall) and back. Since the electronic measurement .of time results In the digital thickness measurement of the first significant sound reflector, the probability of a mid-wall reflector being measured verses the backwall Is dependent on the size of the reflector related to the surface area of the ultrasound transducer.

The larger the mid-wall reflector, the more likely the digital thickness reading will be the mid-wall number, and not the backwall value.

SE No. 000243-002 Att. 2-4 To further characterize the dry%,'ll and "A-Scan" UT technique was also employed. "A-Scan" is Imoortant for the expanded analysis of the character, location and amplitude of various reflectors.

The "A" scan is the ultrasonic Indication displayed on a cathode ray tube (CRT). The front surface pip or amplitude appears first, and the back surface p'r or amplitude appears sometime liter in the CRT sweep display. The space between the pips is a measure of the distance between the surfaces.

Pips In between the front and back surfaces may be mid-wall reflectors such &s laminations, inclusions or isolated holes and/or pits.Other characteristics of the reflector can be observed by a qualified technician when using an "A" scan that are not available with a D-teter.Profile of the amplitude, break pattern at the baseline, number of doublets following the amplitude pip. multiples of original reflectors, and amplitude height on the screen and other characteristics all give.'Informatlon that!may', be useful In analyzing orbins of ultrasound reflectors.

MEASUREMENT LOCATION Initial UT measurements were made from the inside of the drywell containment at elevations 51 feet and 10 feet. A digital UT system was used. The measurements opposite the sand cushion at the 10 ft. elevation in the Says corresponding to where water leaks were observed.

Indicated that the containment wall was thinner than expected.

Measurements above these areas in the same plate indicated thicknesses within the original plate thickness variability.

Additional UT readings in the same Say quadrants at elevation 51 SE No. 000243-002 Att. 2-5 indicated no abnormal thickness vi.-lations.

Although there are no specific requirements for surveillance of the containment wail thickness.

it was considered prudent to make these measurements due to the wetted conditions that had occurred.The initial measurements were made through the protective coatings on the Inside of the containment.

Since the effect of the protective'c6ating on the UT measurements was questioned, special test blocks were made that included the coating material to quantify the effects of the coating on the UT readings.

The accuracy of the UT system was established for the coating thickness of the upper portions of the drywel1. The effects of Carbollne Carbo-Zinc 11 coating on the accuracy of UT measurements was verified through an experiment conducted by GPUN. Two carbon steel plates 4pproximately I.IS-inch thick and six by six-inch square were coated with carbon :inc. One plate had five mils coating and the other plate had 10 l1is ceotifig-

'Both', plates had a half Inch wide strip on one edge left uncoated.

Both plates were laid out in a half inch grld pattern across the entire partially coated side including the uncoated strip. Similar equipment (D-meter of same make and model) transducers, and couplant as used in the field was utilized and measurements taken. Approximately 149 readings of thickness were taken for each plate. Additionally each gild (excluding the uncoated strip) was.measured by Dry Film Technique (OFT) gauge to determine the coating thickness.

The uncoated strip for each was measured by micrometer.

The three readings:

1) ultrasonic (coated and uncoated);

2) dry film technique:

and 3)micrometer (uncoated strip) were compiled, averaged and final factors SE No. 000243-002 Att. 2-6 developed.

The uncoated micrometer reading, plus the DFT reading was treated as the true reading of combined thickness.

The UT reading was found to overcall 0.3% for S mil coatings and 1.5% for 10 mil coatings after subtracting the DFT reading from the combined UT reading of steel and coating thickness.

It should be noted that the coating application on the test plates and the upper portion of the drywell were consistently uniform. The coating along the basement wall. however, was found to be considerably thicker at places effecting the UT readings.

For this reason the coating was removed and a new set of UT measurements were made. The new readings Indicated that the containment wall was thinner than exp':tee In several areas along the basement floor. The areas of Indicated tnln~ing was adjacent to the sand cushion.EXTENDED UT MEASUREMENTS As a result of the Initial UT readings adjacent to the sand cushion being considerably thinner than expected, a program was InItIated'to o6tal detailed measurements to determine the extent and characterization of the thinning.

UT neasurements were made In each Bay at the lowest accessible locatiors.

Where:hinnIng was detected, additional measurements were made In a cross pattern 4t the thinnest section to determine the extent and direction.

Measurements over a six by six Inch grV .ve-e then made, moving over the thinnest area to further quantify the wastage area.To determine the vertical profile of the thinning, a tiench was excavated into the floor in Bay 17 and Bay 5. The concrete floor and rebar was removed to expose a portion of the drywell wall about 18 inches wide and sufficiently deep to allow measurement to the bottom of the sand cushion area. Bay 17 was selected since the extent of thinning at the floor level was greatest in that area. It was measured that the thinning beow the Initial measurements were L I I SE No. 000243-002 Att. 2-7"5-' no more severe and became less severe at the lower portions of the sand cushion. Bay S was excavated to determine if the thinning line was lower than the floor level in areas where no thinning was detected although several Inclusions were found, there were no significant indications of thinning.

The Safety Evaluation (SE No. 328227-001) for the excavation and its treatment for continued plant operation is separate from this evaluation.

Heat Affected Zones & Reinforcement Structure Other areas of concern requiring additional UT Investigation were the plate to plate welds under the torus vents and the vent opening reinforcement plates.These areas were given extra consideration on the basis that material sensitized by welding nay have been attacked by a corrosion mechanism with greater damage or cracking occurring at those locations.

The extra UT investigation was conducted at three spots equal distahce alb iýtie'wtch toe"°of the vertical plate to plate weld and on either side of the bottom center gusset of the vent opening reinforcemmnt plate.D-meter thickness measurements were taken at all eight spots for Bay 5. 7 and 19. At these three PBy sites the six spot locations on each side of the plate to plate weld under the torus vent openings were also 45 shear tested to interrogate the weld Heat Affected Zone (HAZ). The 45" shear wave test was especially done to detect HAZ cracking.

The top two spots were alio the sites from which the plate to torus vent reinforcement plate weld was eramined for SE No. 000243-002 Att. 2-6 HAZ cracking.

No crack Indications were found and no wastage of the torus vent reinforcement plate was found. The plate to plate weld HA! as well as the weld when tested as part of a B or C lozation grid (6"x6") Indicated wastage similar to the surrounding plate wastage.Alternate UT Techniques and Verifications

£PRI NOE Center UT personnel were invited to Independently analyze the containment vessel plate. Their objective was to independently'analyze'the" condition of the drywel1 liner. They scanned two areas using a "Zero Degree Longitudinal Have Method". One area compared was just above the curb that we indicated had general wastage. Another area was where we had Indications of mid-wall deflections or laminar Inclusions.

Their observation and measurements Independently verified GPUN's results.Mapping of the wall profile Indicated a corrosion transition at seven to eight inches up from the concrete curb in Bay 19. This detaITfd~ma A% V f.A... -, ..... ^ -, -'..corroborated by the GE Ultra Image III "C" Scan topigraphical mapping system that will be used to obtain a baseline profile to track continued wastage.GPUN experimentally utilized the 1.0. Creeper or "30-70-70" technique (a UT integration method) to detect minor changes in back wall surface conditions.

This technique ccopared "A" scan presentations from one Inch thick corroded samples the results from Bay 13 locations "A" and "E". Reference standards were utilized representing light, moderate and heavy corrosion conditions.

This 30-70-70 technique defined surface roughness conditions by matching "A" Scan presentations from materials that have lignt. medium and heavy corrosion on their back surfaces.

It was able to verify the rouqhness condition of wastage and the light corrosion areas of the containment wall.

SE No. 000243-002 Att. 2-9 The "A" scan displays from the vessel plate were categorized by comparing them to the reference "A" scan displays.

Location A of Bay 13 (0"-6" up from concrete curb) showed typical "A" scan display of moderate corrosion on average and locdl sites of heavy corrosion.

Bay 13 locations "A" and "E" Indicated heavy corrosion between 0 to 6 inches above the curb, moderate corrosion 6 to 14 112 inches above the curb. and very low or no corrosion 14 1/2 to 17 Inches above the curb.LOGIC OF CORE SAMPLE LOCATION The selection of areas to obtain the core samples was made to evaluate If the UT measurements represented Indicated material wastage or If there .as localized "pitting".

Those measurement areas that Indicated thickness readings of less than half of tMe thickness expected.

I.e.. .4 to .7 inches.and had adjacent measuremonts of the expected thicknesses (nominally 1.154").were designated as "pitted" areas. Areas that had Indicated thinning at adjacent measurements were designed as wastage areas. A third area, above the uastage area. and within the sand cushion that appeared to have no thinning or"pits". was also selected as a sample site. The core sampling sequence and logic were to first obtain a sample of a suspected "pitted" area and two samples of a wastage areas but In different bays. Should the "pitted" sample turn out to be an Inclusion as suspected from the UT evaluation and tho adjacent areas were actually the thickness as measured by UT. additional samples of areas that were suspected as being "pitted" would not be required.

SE No. 000243-002 Att. 2-10 Core Samples Core samples of the Drywell wall were taken at seven locations.

The samples were 2 Inches In diameter.

This was considered the minimum diameter to produce an adequate sample of tht wastage area and provide an opening large enough to remove sand samples. The c,'enlng size also permitted Insertionq.

a miniature video camera. Larger openings ould have requlredai more complex plug design to restore the structure to Its ori'nal conditlion.-

The "pitted" sample #2 from bay 15 location "A" (GPUN 3E-SK-S-8S) was found to be an Inclusion In the plate with little to no indication of corrosion on the outside of the sample.Samples #1 and 83 were from bays 19 location "C" and bay 17 "D". respectively.

Both showed significant wastage with good correlation of actual micrometer measurement with the UT measurement (See Table 1). ..tt' ,,t, t "I*, 1, ., d The wastage samples (plug I & 3) were measured for thickness by ultrasonic (D-meter) and dimensional (micrometer)

In a four-point cross pattern and a center location.

The micrometer readings were taken with a ball micrometer to minimize the error obterved when a flat bottom micrometer measures a locally irregular surface. The micrometer measurements through the oxidized surface Indicated the UT measurements to be between 0 and 4% less than the micromvter measurements.

SE No. 000243-002 Att. 2-11 Two additional wastage locations were selected below the severely thinned locations (Samples 4&S) and two locations above the wastage areas (Samples 6&7) were selected to bound the conditions.

The wastage samples 4&5 were similar to samples 3&3 confirming the UT measurement accuracy.

Samples 6&7 did not have wastage and the sand behind them was found to be dry. also confirming UT measurement accuracy.

SE No. 000243-002 Att. 2-12 TABLE I CORE SAMPLE THICKNESS EVALUATION Sample No. Location Type of Sample Pre-removal Thick.Post-4Re0val Thick. (Ave.)19C -11'3 5/8" 2 ISA -11' S 114" Wastage Pitting.815" (avg.).825".490" 1.17 (min.)(avg.)1.170" center only 3 17D -11' 3 3/4" 4 19A -I1' 3 318" 6 I1A -11 3 I 1A -12' 2 314" Wastage Wastage Wastage.840" (avg.7.830" (avg.).860" (avg.)1.170" (avg):.8600.885" Above Wastage 1.19" center only 1.181" center only 7 19A -12' 1" Above Wastage 1.140" (avg.)

SE No. 000243-002 Att. 2-13 The openings in the Drywell wall were repaired and sealed with a special designed and fabricated steel plug. The final repair was accepted by the Authorized Nuclear Inspection (ANI) after successfully completing a magnetic partical examination of the welds on each plug. A final acceptance test for each plug was performed using a vacuum box bubble test. In addition a local leak rate test was conducted on each plug and met the Integrated leak rate requirements of the Code of Federal Regulations l')CFRSO Appendix 3. Actual leak rate measurement at each plug was 0.000 standard liters per minute at 3S psi. The repair left the Interior surface flush with the inside of the drywell wall. A separate Safety Evaluation (SE No. 328227-001) for the removal of the samples and for the repair of the Drywell openings has been conducted.

DATA

SUMMARY

The thickness measureme.,,ts obtained adjacent to the sand4cuht~i~ape tabulated on GPUN drawing number 3E-SK-S-8S.

Initial measurements were taken at four locations near the lower curb at each torus vent. These locations.

A-B-C-D.were selected to provide two thickness measurements of the left and right drywell plates that make up each Bay section. Eact tabulation heading defines the location of the tabulated matrix of measurements with respect to the top SE No. 000243-002 Att. 2-14 of the curb and to the weld between the two plates at the center of the vent line. The matrix of measurements are at one inch Increments both vertical and horizontal.

Those measurements around heat affected zones and on the vent line reinforcement were taken one inch on each side of the weld. No degradation or wastage was indicated on the reinforcement plate or around the reinforcement plate to the containment plate weld. Wall thinning Indications on the containment plate on each side of the containment plate weld was the same magnitude as surrounding areas Indicating that the weld heat effected zone did not cause or accelerate wastage.Data Reduction UT drywell thickness data was collected In each of the ten bays. The UT data Is presented on GPUN Drawing No. 3E-SK-5-8S Rev. I. The primary concentration of data was within , 6 inch wide circumferential banu above the drywell floor curb since data above this band Indicated minimal wastage of the drywell wall material.A new nominal wall thickness was sought for the affected lower portion. 6'wide band. of the drywell shell. Two approaches were taken."The first. was to establish the mean and standard deviation values nf all the UT data In the affected region of the drywell. The second approach was to establish the mean and standard deviation values of the UT data in the affected region which Is contained with!n a 60 Inch circumferential extent of the drywell. The second approach using six measurement locations in each bay yielded nine (9) 60 Inch combinations of mean and standard deviation values for eich of the ten (10)drywell bays. The significance of the 60 Inch spans Is that It represents a physical property of the shell.

SE No. 000243-002 Att. 2-15 This property is the deflection half wave length which defines the shell boundary relative to the location of applied primary and secondary loads beyond which the applied load does not cause shell deflection.

This property was calculated by Professor A. Kalnts of Lehigh University.

Although some of the low value UT Indications were Identified as Inclusions In some of tie areas measured.

they were used as thickness measurements for the statistical reduction of data.The first approach yielded a mean and standard deviation value of 0.96 Inches for all of the UT data In the affected region.The second approach yielded a value of 0.87 In. for the minimum mean wall thickness within the 60 Inch arc length criteria.

This segment Included 50 data points within a 26 x 6 Inch segment of the drywell shell In Bay #19. A wall combination In the same bay within the 60 Inch criteria included-148 data points with the data points extending over an area 57 x 6 inch and including the data within the former 26 Inch segment. This latter segment also yielded a value of 0.87 In. for the mean wall thickness.

For purposes of the engineering calculations regarding the structural Integrity of the shell, based on the above minimum mean values, a nominal wall thickness of 0.87 In. should be utilized.

SE No. 000243-002 Att. 3-1 Attachment 3 BACKGROUND Hater Intrusion Detection The first documented evidence of the Intrusion of water Into the annular space between the drywell shell and concrete shield wall came to light during the 1980 refueling outage when water was visible around penetrations X-46 at elevation 86' 0" and running down the wall to floor elevation 75' 3". Hater was also observed at penetration X-50 at elevation 47' 0' and running down the wall to floor elevation 23' 6". Hater collection was also observed on the torus room floor coming from the leak drains In bays 3. 11. and IS. Informal.

urdocumented communications.

however, also indicate water was observed on the torus room floor following construction.

Construction The primary containment pressure vessel Is contained within a concrete shield with a 3" annular space between the two structures.

The annulus Is filled with sand specified as ASTH: C33 from elevation 8' 11 1/4" to elevation 12' 3" and from the bottom of this sand bed are S drain lines.The sand appears to be a natural sand comcpsed of silica uith some alumina. An Owens-Corning Fiberglass SF vapor seal duct Insulation SE No. 000243-002 Att. 3-2 was applied to the vassel shell from elevation 12' 3" to 23' 6". The insulation was supplied as individual boards 2" thick with a factory applied laminated asphalt kraft paper waterproof exterior face. These boards were attached to the vessel shell with mastic and Inhulation pins.The Joints between the boards and edges and penetrations were then sealed with fabric reinforced mastic. The remaining annular region above elevation 23' 6" is filled with a Firebar-D material.

It was applied as a spray coat (approximately 2.75" thick) over the vessel shell. The material Is composed of asbestos fiber (approximately 75%), magnesite, magnesium sulphate and a foaming agent (Aerosol PK) to control density. Over the top of the Flrebar-D was placed a 4 mil thick polyethylene sheet.The primary vessel Is fabricated from ASTH-A 212 Grade B which Is equivalent to SA-516 Grade 70. The vessel was coated on;the-I.D.

with Carbo-zinc 11 and on the 0.0. with "Red Lead" primer Identified as'TT-P-86C Type I. Coating on th-k ,%terlor of the vessel extends from elevation 8'-11 1/4".4 SE. No. 000243-002 Att. 3-3 POTENTIAL SOURCES OF WATER INTRUSION Probable Sources Observations of leakage from the sand bed Irains during the 1980 and 1983 refueling outages Indicated that water had intruded Into the annular region between the drywell shell and the concrete shield wall. In addition, water samples withdrawn from the drains in 1980 were radiologically onalyzed and showed activity similar to primary water. From this Information It was concluded that the probable sources of water were (1) the equipment storage pool. (2) the reactor cavity, or (3) the fuel pool. It was further concluded that the leakage only occurred during refueling when the reactor cavity, the equipment storage pool.,and.the fuel pool are flooded. During the 1986 refueling outage, water samples were again taken from a drain line and analyzed.

In addition to tritium, these samples were also analyzed for contaminants.

The results-of these analyses are shown In Table I-1.

SE No. 000243-)02 Att. 3-4 TABLE I-M Orywell Drain Line Hater Analysis Parameter Na K Ca Mg Al NI Fe Cr Mn Pb NN,(N)Cl.-01 SO, P04 F TOC Organic Acid Total Sulfur Conductivity pH Alkalinity t.iCOI)Sample I (ppm)145 142 7.5 30.33 ( .01 ( .01' .01.01.06 3.6 32.5 8.7 153 5<(I S1 ( .1 153 1100 us/cm 8.9 Sample 1!96 6.4.02 ( .02.74 ( .02.02 C .02 25 6 60 N.D.23.3 814 us/cm 8.7 130 Samples taken SE No. 000243-002 Att. 3-5 UT Data lnterpreta:ion Prior to core sample removal possible causes of the low UT thickness readings were attributed to external corrosion.

laminations or a field of Inclusions within the plate. Because the very low readings were localized it was expected that they would be a result of laminations.

The general wastage, however, extended from plate to plate and the affected areas of the shell were within the sand bed only. Thus It was concluded that the plate thinning was most likely due to corrosion.

In addition, a qualitative assessment of the plate condition was made using an "A" scan presentation with a 5 mghz transducer.

This data was also indicative of corrosion on the outside.Numerous ultrasonic thicknems readings were taken In the drywell part Lularly at the elevation of 11' 3". Review of this ultrasonic test data showed that potential corrosion damage appeared to be confined to regions In Bays 11, 13, 17 and 19. Furthermore, the thinned parts of the drywell were limited to those areas which were in contact with the sand bed from elevation 10' to 11' 9". Numerical analysis of this data determined the minimum mean remaining wall thickness was .87".UT thickness readings below the concrete floor elevation showed the thickness to be greater than .87" and at the bottom of the sand bed to be nearly nominal design thickness.

..II 1 -E -=

SE NO. 000243-002 Att. 3-6 Sampi ing After the completion of the ultrasonic testing (UT) of each of the drywell bays above the concrete floor, the data was assembled and reviewed.

This data indicated that there were at least three regions which showed different characteristics.

One set of data showed regions of overall general wal reduction which we characterized as wastage. Another iet showed regions with little or no general wall reduction but localized areas with large wall reduction which we characterized as pitting/inclusions.

The last set of data showed regions of little or no wall reduction and no random large reductions, which we characterized as minor wastage. The characterization of each bay Is summarized In Table 2-M.

SE No. 000243-002 Att. 3-7 TABLE 2-M BUT Characterization I Minor wastage 3 Minor wastage 5 Pittinglincluston 7 Minor wastige 9 Pittinglinclusions Ii Wastage 13 Wastage IS Pittinglinclusions 17 Wastage 19 Wastage In Addition tu the above general characterizations.

It was also observed from the UT readings that above an elevation of approximately11'9" the wall thickness would return to the nominal value. This occurred even though the, readings were still within the sand bed and there was wastage below this elevation.

Likewise.

there were regions of the sand bed below the concrete.which heretofore had not been ultrasonic tested and hence no characterization could be made.

SE No. 000243-002 Att. 3-8 It was decided, therefore, that core samples should be removed from the drywell In each of these different regions in order to achieve the following goals: a) Vcrify UT thickness reading b) Characterize the form of corrosion c) Obtain sand samples and samples of other annulus materials d) If corrosion existed, characterize corrosion products and environment e) Provide access for visual examination of the outside surface of the drywell f) Allow for sampling of sand and/cl corrosion products for bacteria With these goals In mind, a first cut -as made at se'ecting"regions for sampling of the drywell steel. Twelve regions were sele:ted:

four from wastage regions, four from "pitted" regions, two from above the wastage region and two from below the concrete level. These initial selections were.however, modified slightly as the program progressed and additional Information became available from ultrasonic testing an. Initial core sample examinations.

SE No. 000243-002 Att. 3-9 Table 3-H identifies each of the seven core sample locations ultimately chosen ani the types of samples obtained.TABLE 3-H Core Samples Sample No.Bay/Location 1 2 3 4 5 6 7 19C Wastage 15A Pitting/Inclusion 17D Jastage 19A Wastage IA Wastage 1A Hinor wastage 19A Minor wastage Elevation U1'-3 518" l-5 1/4" 11'-3 314" l1'-3 318" 11'-3" 12'-2 314" 12'-1" Core, sand, bacteriological Core, sand. bacteriological Core, sand Core. sand. bacteriological Core. sand. bacteriological Core. sa:d Core, sand Samples Obtained SE No. 000243-002 Att. 3-10 Evaluation of Pitti,,4'!nci,,slon Sample Core sample #2 which was removed from bay 15 was taken to assess whether pitting or Inclusions were responsible for the low ultrasonic thickness readings observed in random locations.

In region C where the sample was removed, the general area had thickness readings on the average of about 1.17" with random low readings of .48". This particular plug had a region approximately 1/2" In diameter wnere the low readings resided.Upon removal of this plug it was immediately evident visually that no serious corrosion or pitting had occurred.

The outside surface of the plug was covered with a reddish brown oxide and the actual measured thickness of the plug was 1.17" (avg.). Figure I. Elemental analysis of this oxide by EOAX Indicated iron v the major constituent although in random location very high levels of lead were observed.

This lead is from reminants of the red lead primer originally applied to the shell. Other elements observed at trace levels were Al, SI, Mn, Ca. K, Cl. S.

SE ho. 000243-002 Att. 3-11 Metallographic specimen, were prepared from the core plug both parallel to the rolling direction and perpendicular to It. Examining the micro specimen at the outside surface of the core revealed ;ome minor pitting.These pits were filled with oxide which appeared normal for carbon steel corrosion.

At t4e mid-plane of the specimen, however, a band of aluminide stringers was found In the region where the l..w UT readings existed, Figures 24 -3M. These stringers were sufficiently dense as to form a lamination which could easily reflect ultrasound.

This observation validates the conclusion drawn by the GPUN NDE people via their "A" scan UT analysis of this region that the low "D" meter readings were a result of laminations.

In addition, the examination of this plug also validated the accuracy of the thickntss measurements.

It was concluded that UT could adequately define this type of condition and additional samples fiom pitting, inclusion regions were not required.i.U W SE N.o. C00243-002 Att. 3-12 r* *.A 2X Figure 1-H rPug #2 outside surface of drywell.Uniform red brown corrosion product.

SE No. 00243-002 Att. 3-13 4 K-/I S S---..50X'r.a a 0 500X Figure 2-H Plug #2 Aluminide stringer at mid-wall.Plane parallel to rolling direction.

SE No. 000243-002 Att. 3-14 0.'v.:.*sox t0 it 6. 0 6 .4 WM1 0 A *"a* ~ 0 S.1.* 0 a 0'. *dT/pow 500X Fiour" -j-i Plug 92 AluminIde Stringer at mid-wall.Plane perpendicular to rolling direction.

SE No. 000243-002 Att. 3-15 Examination of Wastage Samples As discussed previously.

four samples were removed from wastage regions.Three of these samples were sent to General Electric (Sample Nos. 3, 4 & 6)for analysis and one was analyzed by GPUN (Sample No. 1).When core samples (numbers 1, 3. 4 and 5) were cut. It was noted that a hard black crust remained in the hole on top of the sand. This crust was approximately 112" thick. It was quickly realized that this crust was the corrosion product from the iron and as such was collected along with the sand beneath It for later analysis.In general, all the wastage samples looked similar showing a relatively uniformly corroded surface-with some hills and valleys (Figure.4,.

Overal),..',.. .'...." ... ..... .,.*the surfaces were covered with a thin black adherant type deposit with some regions having a thicker more dense buildup of deposit (approx. .030" thick).Elemental analysis of this deposit showed Iron to be the major constituent with varying levels of chloride contamination.

Minor traces ofmangangse

.aluminum and silica were also noted and on occasion a trace of sulfur (Figure 5-H). On sample #1 a cross section was prepared through one of the valleys on the corroded surface (Figure 6-M). This valley had 6 layer of corrosion product on it approximately 30 mils thick. ELAX analysis of this deposit SE No. 000243-002 Att. 3-16 revealed a high chloride concentration In a 2 mil thick layer of deposit adjacent to the steel, while further into the oxide but adjacent to this region the chloride levels were very low (Figure 7-M). Although other samples did not show this dramatic variation In chloride, all did show that chlorine was a major contaminant.

In addition to EWAX analysis, x-ray diffraction was performed by GE on the black deposit. The results showed the material to be primarily FelO, 4 (magnetite).

This confirmed an Initial observation that the deposit was magnetic:

no other compounds were Identified.

Metallography on the core samples showed that there was no deep pitting and no signs of any type of cracking or Intirgranular attack. Manganese sulfides were observed within the microstructure which were typical ftrthis type material (Figure 8-N).

SE N4o. 000O243-002 Att. 3-17 Plug #1 9.Fiue4-M Plug #2 outaer wall surface microplane Is located AMt 3-18 I*LUG'I I ILAR J'1 OL II I-rr¶.i. -; ; F~.. .. .. .. .. ... ... .. .. ... .. .. .. .. .I-I K*1 I.r.I , : Figure 5-H Typical elemental analysts of sample 01 corrosion product.

A L L. P- L 7 56X Figure 6-H Thin layer of corrosion product remaining on sample 0I showing different layers and the presence of voids.

Att. 3-20 I LUG3 I SL&ALE Me cI t.' ~ ~ *:. .* .. -.*0 *....' .~.~.**~ ~A ~*.i.4* I ,. I .I ..Lt 4 F1gure 7-M EPAX scan through deposit shown in Figur, 6M.Deposit runs from 0 -49% full screen.Chloride peak Is at steel/oxide interface.

I -at -I Aitt-_=21 10OX O.D. surface; Plug 11 200X HnS inclusions below surface.Figure 8-11!

SE No. 000243-702 Att. 3-22 Analysts of Sand and Flrebar-D As was shown in Table 3-H sand samples were removed from behind each core plug. In addition, sand was removed from the Bay 11 drain line. EWAX analysis as well as leachate analysis was performed on representative samples of the sand. The results are shown In Figures 9-M and Table 4-M. The sand appears to be a natural sand composed of silica with some alumina present. As noted on the WDAX spectrum, some chloride Is present and this was confirmed by the leachate analysis which showed chloride In the range of 6.5 -9J ug/gm.Also noted In the leachage analysts was magnesium and sulfate which most probably came from the FIrebar-D.

Some organic carbon was also detected.These analysis indicate that a source of the chloride founJ In the corrosion product existed In tha sand which was probably lpached from the Firebar-D and that organic material as well as a source of sulfur exist which could provide nutrients for bacterial growth.A sample of the Firebar-O was obtained through one of the drywell penetrations and subjected to a leachate analysis.

As Might be expected..this materlai was high In Na. K. Ca. Mg and SO. as well as chlorine.

The results of these analyses are also shown In Table 4-M.

((Att. 3-23 TABLE 4-M Sand Lear-hate Analysis Analytical Parameeters Na K Ca Mg Al Ni Fe Cr un Pb NR 3 (N)Cl N03 SO4 P0 4 F TOC Organfc Acids Total Sulfur B Conductivity p11 Firebar-ri*

Leachate I Mr, 60 C (U-/g)777 784 176 1936 4 0.3-40.3< 0.3 4" 0.3-C 0.3 0.6 Sand Leachate Bay 11 Drain 24 Rrs, Room Temp (uNO)V Sand Leachate Bay 11 Drain I Hr, 90* C (ug!g) , Sand Leachate Plug #1 (19C)1 Hr, 60" C (us/g)Sand Leachate Plug 12 (15A)1 Hr, 60" C_ ,U,, 25 25 30 30 5.0 4 0.5 0.5 1.5 10.5 2.5< 25 N.D.N.D.39< 5 25 20 25 10 1.5 0.5 1.0<0.5< 0.5< 0.5 6.5 1.5 32 N.D.N.D.37 4C5 37 37 47 10 39 4 .33 82 4 .33 3.;< .33 47 23<.23<: 23 C 2.3 8.4<. 2.3 C 2.3 C 2.3 93.6 79 N.D.N.D.N.D.573 132 2850 N.D.14 1056-20" 50 45-C 17 28 N.D.N.D.46.6 588 8.46 7.43 7.58 7.02 5.99 Flrebar-D is a composite of foam, fibers- and concrete SE N~o. OOOZ4d-IYJ(Z Att. 3-24 ,.;AIJV-rL-I%,UUr-(311- S1,6141.1.

.*--.L.-, I.1.t I I 0 ° , ., Figure 9-H Photo shows distribution and type of sand particles.

Spectra shows basic elemental composition.

SE No. 000243-002 Att. 3-25 microbiological Assessment In order to assure a complete assessment of the corrosion damage It was decided early on that a microbiological analysts needed to be performed.

Because of limited in-house expertise In this area, an outsi'e consultant.

Dr. Carolyn Mathur from York College, was contracted to perform this analysis.

It was decided that four samples of material from the sand bed would be analyzed from the realons Indicated in Table 3-M. Two samples would be sand and two would be corroslor product. These samples were secured Immediately upon removal of the core plugs to assure minimal environmental effect on the bacteria.

Samples were treated for microscopic evaluation as well as for future culturing.

During core removal close attention was also paid to metal temperatures to assure temperatures did not exceed 150" which would kill the bacteria.Results of the cultures are not yet available; however. preliminary Indications are that there Is no strong presence of sulfate reducing bacteria (SRB). The microscopic evaluation results are shown In Table 5-H. Cell counts appear typical for levels of bacteria found In natural environments.

In addition, it was reported that the bacteria appear filamei,tous and In some cases bacteria was observed to be attached to the corrosion product.

SE No. 000243-0OZ Att. 3-26 Currently cultures are being grown aerobically and anaerobically to establish the type of bacteria present including the presence of sulfate reducing bacteria.Ground Potential Measurements The possibility of stray currents Influencing the corrosion rate was also considered.

In order to provide an assessment of this. external potential measurements were conducted to check for the presence of stray currents.

Potential measurements were taken between the ground and each of the five sand bed drain lines using a copper-copper sulfate reference cell. The measurements revealed no evidence of stray currents while the reactor Is shut down. however, these measurements will need to be repeated during power operation.

I SE No. 000243-002 Att. 3-27 TABLE 5-M Bacteriological Studies Preliminary Results Sample No.2-15A 1-1 9C 6-1IA 4-19A Type(dry).4acent to Drywell Corrosion Product Adjacent to Drywell Sand (moist)Away from Drywell Corrosion Product Adjacent to Sand Cell Count I lxlO' cells/gm 71% viable 5xlO6 cells/gm 50% viable 4x10' cells/gm 74% viable 6x10 6 cells/gm 40% viable SRS negative weak pos.weak pos.negative Stained with fluorescein Isothlocyanate SE No. 000243-002 Att. 3-28 Corrosion Assessment As discussed in the background section, wetting of the sand bed may have occurred as early as initial construction.

The only other documented evidence of leakage was during the 1980, 1983 and 1986 outages. Although the exact source of leakage during construction Is unknown. It was reported that during the application of the Firebar-D material that copi.ous quantities of water were observed coming from the Firebar and running down the'drywell presumably Into the sand bed. During outages water was most likely coming from a leaking gasket In the seal plate region. This gasket was replaced during the 1986 refueling outage and the leakage appears to be stopped. On the above basis and In view of the fact that there would not be other sources of water to enter the aniular region behind the drywell during operation.

It has been concluded that the introduction of water was an intermittent occurrence (i.e. during outages) which may have,,occ~uredd,.ur-jng, construction but definitely occurred in 1980. 1983 and 1986. Also. It can be concluded that the sand as a result of this water introduction is contaminated with chloride and sulfate along with numerous metal ions.

SE No. 000243-002 Att. 3-29 Sand Is generally ascribed with good drainage properties which would allow for the bulk of the water which entered the sand bed to flow out of the drain lines; however, because this region Is fairly enclosed with little dir circulation, high humidity Is believed to exist In the annular space which coull result In the sand remaining moist for Indefinite periods or time. This is partially substantiated by the fact that high hLmIdIty and sweating Is generally observed in the torus room here the sand bed drains exit. Above the sand bed, however, fiberglass boards and above that Firebar are applied to the exterior drywell steel which would help prevent moisture from coming directly in contact with It. In addition, during operation the average drywell air temperature is approximately 1400 F which again would prevent condensation from forming on any exposed steel surfaces.

The overall environment within the annular region can therefore best be described In the following manner: Hater was Introduced Into the. sand bcd possibly as early as in the late sixties and probably contained magnette and magnesium sulfate from the Firebar. The bulk of this water wotld have drained off leaving moist sdnd behind. He know that the exterior of the drywell was coated with red lead primer over which Firebar and fiberglass boards were applied which would afford general protection to these steel surfaces from corrosion.

Coating damaged areas and with time all areas within the sand bed would be expected to experience general corrosion as long as the sand remained moist or until a protective oxide film built up on the steel surface as a result of the corrosion process. It appears.

SE No. 000243-002 Att. 3-30 however, that a completely protective film did not result most probably because of the presence of chloride.

The actual metal loss which may have occurred during the time frame from initial startup until the next time water was reintroduced as a result of leakage Into the cavity Is unknown.The first documented incident of water Intrusion following startup which would definitely initiate corrosion was In 1980. Hater samples collected..

and analyzed at this time for radioactivity measurement, Indicated that It was refueling water and hence adds credence to the assessment that the source of the water was the leaking bellows gasket. Corrosion rates would therefore be properly based on the assumption that the corrodent was refueling quality water contaminated with chloride from the Firebar and that the corrosion process was aqueous genetal corrosion.

Some shallow pitting Is also occurring but It Is considered only in view of Its contribution to ovurall thinning.The possibility of stress corrosion cracking and hydrogen embrittlement were also considered.

However, these forms of corrosion are generally associated with high strength steels or high temperatures and not considered a damage mechanism for the environment or material associated with the drywell. Ultrasonic examination of the welds and heat-affected zone in the wastage regions also showed no Indication of cracking.

SE No. 000243-002 Att. 3-31 An upper bound general corrosion rate for carbon steel would be expected to be In the range of 10-20 mpy depending on the drywell plate temperature.

These corrosion rates, however. if applied generally to the drywell it.. on In contact with the sand bed are consistent with the average wall loss of .288" only If the corrosion Is assumed to have occurred since 1969 which was the first possible time water could enter the sand bed drains.In fact, however, close scrutiny of the UT thickness data Indicates that corrosion was extremely non-uniform as defined above In the section on UT measurements.

First. the region above the 11' 9" elevation shows little or no wall loss. Then the region from 10' 3" to 11' 9" shows the greatest wall loss followed by the region below 10' 3" which shows substantially less wall loss. Lastly, only two regions of the drywell encompassing four bays show any significant wall loss. A possible epplanation for.this is that due to channeling only these regions became wetted. This assumption Is potentially confirmed by the observation that the sand In the minor wastage regions was dry. Also, the Intimacy of the contact between the sand and the plate Is a factor. If the sand had been pushed away from the drywell in certain regions due to the preoperatlon pressure test causing the drywell to expand; this also can result In variations In corrosion rate. The protectiveness of the red lead primer will be a function of Its Integrity In the various regions and again may be leading to variable corrosion rates.Lastly, differential aeration may be playing a role In where corrosion Is occurring.

Clearly the presence of magnetite.

an oxygen defficient oxide, In some regions and hematite In other tegions suggests this Is occurring.

SE No. 000243-002 Att. 3-32 Conclusion Aqueous corrosion of the carbon steel drywell Is estimated to have Initiated In 1969 resulting from the first intrusion of water Into the sand bed region. This inventory of water may have been added to during subsequent outages but was definitely added during the 1980. 1983 and 1986 refueling outages. This latter water Is expected to flow down over the Insulation material in the annular space and pass through the sand and out through the drains. Depencing on flow rates. dn inveetco, of afater may te accumulated In the sand bed or channeling of the water may also occur leading to wetting In specific locations.

Irrespective of the water flow rate some sand will become wetted with oxygen saturated water and corrosion will result.This corrosion was most likely Influenced by the presence of chloride.leached from the Firebar-D.

as It was found to be Incorporated within the Fe,0 corrosion product. Becterid are not belier-1 to have been a major influence on the corrosion.

This latter conclusion Is based on the facts that no deep pitting was observed and no sulfide or substantial concentration of manganese was detected in the corrosion product. "11 of which are typical evidence of microbiological influenced corrosion.

In addition, the corrosion observed can be explained solely on the basis of chemical attack. However. because there Is viable bacteria present, any plans for Inhibiting future corrosion may also require destroying or rendering this bacteria harmless.

SE No. 000243-002 Att. 3-33 Review of the literature suggests corrosion rates can vary widely for carbon steel In aqueous environments.

Rates can be as low as 1-2 mpy in high pH aqueous environments or greater than 50 mpy In acid solutions.

Dr.Uhlig in his "Corrosion Handbook" lists corrosion rates for ambient temperature, seawater at approximately 1-7 mpy which at 140F.would conservatively equate to approximately 17 mpy. Uhlig further states that with the formation of corrosion products the rate of corrosion will be less than It would be if the steel were In direct contact with seawater and that the rate will stabilize and not change with time. In addition, heobserved that. "specimens of steel have been exposol to seawater where sulfate reducing bacterit were known to be prese- and in fact were found In the corrosion products which contained appreciabe percentages of iron sulfide.The observed rates of corrosion and pitting cf such steel fell within the normal range previously defined." ....If we then take the 17 mpy corrosion rate and project this over the 17 year life of the plant It correlates closely with the average corrosion loss of 288 mils. However, In order to insure conservatism In the structural analysis a factor of safety should be applied to this rate. To arrive at a defendable factor It has been assumed that all corrosion occurred over the past six years as a result of the water intrusion in 1980. This would equate to a corrosion rate of 48 mpy 4nd give a factor of safety of 2.8.

SE No. 000243-002 Att. 3-34 Conclusion Summary I. Wastage nf the drywell plate Is the result of an aqueous general corrosion process influenced by localized oxygen depletion, the degree to which moisture Is present, temperature and chloride contamination.

2. Although viable bacteria were identified In the sand and corrosion product, no substantive evidence exists as to its involvement In the corrc!Ion process, at oeait 'n terms zf currently publicized mechanisms.

However, because of the variable nature of microbial induced corrc.ion any attempts at mitigating corrosion should cons'der this mechanism.

3. 0-Meter thickness readings, which initially were thought .to.be eithe, pitting and later characterized by "A" scan UT as inclusions, were confirmed by metallography to be aluminide inclusions in the carbon%teel.4. The combination of using a 0-Meter for ultrasonic thickness measurements and an "A" scan for qualitative assessment of the ilatW condition are adequate for engineering evaluations.

5. Corrosion is limited to the steel in contact with the sand bed and Is present to a significant amount ti.e.. .25" -.35") only In bays lo.'3. 17 and l1 and only withtn elevations i0' 3" to 11 3".

6. The areas of observed ccrrosion oDDear to be those areas in which the sone has remained significantly wetted. This wetting most likely occurred during initial construction and then periodically during refueling outages as a result of leakage from the drywell bellows. Documented evidence of such leakage exists since 1980.7, Corrosion rates have been conservatively set at 4E mDY although more typically, through review of industry experience and corrosion literature, would be exoected to be opDroximately 17 may.

Attachment 4 SE No.

Att. 4-1 Structural Analysis Bases A reevaluation of the drywell containment structure has beer performed to ins$jre Structural integrity for the combined effects of local St1ell thinning.operating basis earthquake, pressure and temperature due to a postulated Design Basis Accident (OBA) and the mechanical loads. In performing this analysis the following desig: basL were used.Applicable Codes Establishing Allowable Stress Criteria. ASKE Boiler and Pressure Vessel Code,Section VIII, 1962 Edition.(2) Nuclear Code Case 1270N-5, 1271N and 1272N-5 (3) ASME Boiler and Pressure Vessel Code. Section 111. Division 1, appl'cable portions of Subsection NE-3000. namely. NE-3213.10.

NE-3221-2.

NE-3221.4 and Table NE 3217-1.Materials of :onstruction According to th; Chicago Bridge and Iron drawing No. 9-0971 sheet No. 1. Qe%.2. the material used in the fabrication of the drywell shell IS ASTH SA-214 grade B Firebo,. The evamination of the original mill certificates re-eals that all 1.154" plates used in fabrication of the drywell shell Iave a yield strer;:h of about 5 to 33% greater than the minimum 'Decified In the ASTM.

Attachment 4 SE No. 000243-002 Att. 4-2 DesjtnCondi tion The, drywell shell is analyZed for the malimum po~itive pressure 35 psig at 281"F and 62 psig at 175'F. The former condition represents the double end breaks of a recirculation loop. This Is the design basis accident.Other Loads All other !oads considered concomitant with accident conditions were taken from the Chicago Bridge and Iron origlral analysis.Load Combination Tne load combination representing a C-BA during normal operation.

as smecified in the original Bridge and Ircn Driginal report, was cbcsen fo, analysis Tnis load combination includes tme gravity load of vessel anC aDburtE,,ances.

gravity load from eauiomert Support%.

seismic loads OOBE). as we'l as accident conditions for temperature aid pressure.

Att4chment 4 SE No. 000243-001 Methods Att. 4-3 Structural Model The mathematical model used to evaluate effects of the reduced shell thickness within the entrenchment area consists of a lower region of the spherical shill between elevations 23'-f 7/8" and the point of complete fixity against translational movement and rotation at the foundation level at elevation 8'-11 114". This model Is developed to calculate the membrane and bending stresses at the point of flaity due to the accident internal press:" and thermal loads as well as loads associated vith normal operation.

The results of the structural analjsis will allow the determination of the minimum allowable oressure boundary thickness using'A!NE Code allowable stress criteria.EoceDt for the sand pocket zone. all other shell used in the analyses were those Shown in the Chicago Bridge and Iron Drawing No. g-)g71 Sheet No. 4 Rev. 1.The fincti-,n of the sand pocket is to provide a proportional reaction so that tne discontinuity stresses due tP the embedment will be gradual and lower rathe. than abrupt and high.In order to evaluate the sensit'tp

'*, if tne sand pocket. two separate structural models were c side ýd. :n tne first mode;, the sani is assmed to crovide an 4nward reaction '!near in p-'oortior to shell displacement.

T'emodel assumes :!e sai to cffee no starce against the Orvwell shell% e ie r Attachment 4 SE No. 000243-002 Att. 4-4 The thermal gradient in the sand entrenchment zone Is assumed to be linear.The attenuation of the thermal gradient in the meridlona!

direction is assumed to be complete within the sand pocket, that is. the temperature distribution is 175"FI281*F at elevation 12'-3" and 60'F (ambient temperature) at elevation Vl-1 114".The drywell shell membrane loads f-om the original Chicago Bridge and Iron analysis are introduced at t:.e tcp boundary of the structural models to simulate shell continuity.

Tlh structural model and the loading are assunted to be symme'scal1 the penetration and their effects are not considered.

T~ts is reatcrah)e since the reinfercement at the penetrations restores the shell to its original condition.

ne fibergldss insulation material withfn the annulu: Detween the drywell and the con:rete shield wal1 is assume! to have no structural stiffness.

Attachment 4 SE No. O0243-002 Att. 4-5 Stress Analysis The drywell containnent structure model described above is analyzed by the Chicago Bridge & Iron Corporation utilizing the Kalnins KSHEL Program for axisymmetric shells of revolut:on to evaluate the adequacy of the lower shell region within the sand entrenchment area.Each of the two models, with and withcut sand entrenchment.

is subjected to the mechanical loads, operating basis earthquake and the accident pressure and temperature conditions of 35 psig at 281"F and 62 pslg at 175*F.This analys's identifies meridional and circumferential membrane and merldional and circumferential bending stresses for the dead weight.earthquake.

pressure, and thermal loads.

Attachment 4 SE NO. 000243-002 Att. 4-6 The acceptance criteria used to establish structural acequacy of the drywell are taken from Section VIII. ASHE Boiler and Pressure Vessel Edition. Nuclear Code Case 1272N-5. and Section II. ASHE Boiler anj Pressure Vessel Code 1986 Edition, Division 1. Subsection NE. paragraphs NE-3213-10.

NE-13221-2, NE-3221-4 and Table NE 3217-1.For purposes of analysis, the shell thickness In the sand entrenchment zone is taken to be equal to 0.700".Mean of thickness readings as-representing structural response Structural loads will follow paths through the affected region having the largest stiffness (thickness).

Less stiff (thinner) sections will follow the strain of the stiffer sections such that the,-a will be a compatibility of strain through-out, as governed by the stiffer sections.

The condition of strain compatibility means that the stress In the thinner sections will be equal to the stress In the adjacent thicker sections.

It is reasonable to use the mean thickness, as opposed to the minimum thickness, because the mean represents the actual load reacting action of the shell.

Attachment 4 SE No. 000243-002 Att. 4-7 Potential for SucklinQ In addition, another analysis has been performed by Professor A. Kalnins of Lehigh University using the Kalnins shell of revolution computer program to evaluate the potential of buckling of the drywell shell in the sand entrpnchment zone.The mathematical model used to perform buckling analysis is basically similar to the model used for the stress analysis, except that credit was taken for the structural effect of the concrete that extends upward from the foundation arourd *he inside of the drywell and for the sand which povides an Inward reaction In direct proportion to shell expansion.

For purposes of analysis the shell thickness in the sand entrenchment zone is taken to be equal to 0.700".

Attachment 4 SE No. 000243-002 Att. 4-8 STRESS ANALYSIS RESULTS FOR SHELL THICKNESS TAKEN TO BE EQUAL TO 0.700" The allowable stress criteria are: 1) Local primary membrane stress (not Including thermal)

• P, 1.5.Smc -28.875 psi (No change since 1962).2) Surface stresses (local membrane and secondary stresses, both thermal and mechanic4l axial and bending) .Q 3 Sm a 52.500.(Q-3 S.,=57.900 from Sect. III, Div. I. Subsect NE)Table 1 shows the results of the stress analysis at the point of embedment taking credit for the radially Inward reaction because of the resistance of the sand and also for analytically removing It. Using stress *iptensity,-the stresses satisfy the former Code allowable stress criteria except for the condition of full sand removal when allowable stresse! are exceeded by 2.7%.This load combination considered the accident condition of 62 psig and 1750F.which Is not the same as the design basis accident (DBA) representing a double ended break of a recirculation line. Present Code Allowable stress criteria are satisfied in all cases.

Attachment 4 SE No. C-C'243-002 Att. 4-9 Except as mentioned above the stresses shown satisfy the allowable stress criteria of ASME Sect. VIII. 1962, with Nuclear Code Cases 1270 N-S. 1271 N and 1272 N-S, as well as those of ASME Sect. 111. Div. 1. Subsection NE. 1986.using stress intensities as directed in the latter code. Merldional extent.but not the peak value, of local primary membrane stress slightly exceeds (but< 2X) the guidance given in Sect. 111. It is reasonabie to neglect this small departure from present code guidance because the present situation is an In-sarvice condition and not a design condition, and because the departure Is small.

Attachment 4 SE No. 000243-002 Att. 4-10 Results of the analysis for buckling potential Stability margin Is identified in Table 2. Margin is defined as the rItio of the calculated buckling load to the actual applied load. The refererce Is the point of the embedment.

Normal and accident load combinations are considered with and without the radially inward resistance of the sand. The stiffness of the concrete on the Inside of the shell Is Included In both cases. The shell Is considered to be Imperfect.

The minimum margin to safety Is 3.80.Conclusion Structural Integrity of the primary pressure boundary Is maintained with a local shell thickness reduction limited to the sand entrenchment region. Code allowable stress criteria are met using a thickness equal to 0.10".A large margin to buckling exists such that buckling of a locally thinned shell is not a technical Issue.

C ((TABL% I M1ES3 IMr.I'IIES ALOI. in4EhII1Ar (r.Ul)IUAMD With Sand Pocket Without Sand Pocket

& l Idt MF1IBR AM & &IOdlJIG Cudculutcd Allce,.able' U.iiciu1.l-A

.1I1uwau l. %leu.atud lluw.blu Caulcultqjd Allovuuble P 1 psi( 32,1.7 8821 43,722 T 28I1" F LC .ju 1.5 Smc 3 Vi " 52,5 16# 28,875 52,500 6- 2 1 ' y 16,9;9 53,897&X )., ,Jc I,7.%jK)0 57 ,',,Vu C.TABL...(

DU1WE1L ANALYSIS SAND TRANSITION ZONE OYSTER CREEL CONTAIIWENT VESSEL CPU NjUCLEAR CORPORATION 7ARSMtANT, VIV JER.SEY Cz Sa!vices.

Inc.December 31,6 1966 contract 661172 Revision I December 30. 1986 1tvisi~on 2!Febriary 9, 1987 Table of Contents Introduction Aoplicable Codes Allowable Stresses Input Loading Description Table of Input Loads Mathematical Model of Reduced Thickness Zone Description of Kalnins Output Case I -Circumferential StressPlot Case 1 -Meridianal Stress Plot Case I -Stress Intensity Plot Case 2 -Circumferential Stress Plot Case 2 -Meridianal Stress Plot Case 2 -Stress Intensity Plot Analysts of Embedment Zone with Sand Considered to be Ineffective Case I -Circumferential Stress Plot Case I -Meridianal Stess Plot Case I -Stress Intensity Plot Case 2 -Circumferential Stress Plot Case 2 -Meridianal Stress Plot Case 2 -Stress Intensity Plot Conclusions Conclusions 1.0 1.1 1.2 1.3 1.4 1.15 1.6 1.7 1.8 1.9 1.10 1.11 1.12 2.0 2.1 2.2 2.3 2.4 2.S 2.6 3.0 3.1 0txJ-0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0i MA94 Y 'rJ a 0 Analysis of Embedment Zone with Sand Pocket Filled with Grout Mathematical Model -Grout to elev. 12'-3 Circumferential Stress -No Thermal Transition Meridianal Stress Plot -No Thermal Transition Stress Intensity Plot -No Thermal Transition Analysis of Embedment Zone with Sand Pocket Filled with Grout and Shell Insulated to Provide a Thermal Gradient Mathematical Model -Grout to elev. 121-3 Plus Insulation Case 1 -Circumferential Stress Plot Case I -Meridianal Stress Plot Case I -Stress Intensity Plot Case 2 -Circumferential Stress Plot Case 2 -Meridianal Stress Plot Case 2 -Stress Intensity Plot Conclusions Stability Analysis -(Task 4)4.0 4.1 4.2 4.3 4.4 4.5 Introduction Generation of Stress States Method of Analysis BOSOR Computer Program BOSOR Computer Program GOSOR Analysis BOSOR Model Suckling Criteria -Modification of Code Case N2e4 I 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 5.0 5. 1 5.2 5.3 5.4 5.5 5.6 5.7 5.6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 v 0 1 ' 5 'to c -e s 0 " i ,-1 yxac jj SAVt 1-3 1L1A7 I'1

• Modification to Capacity Reduction Factor Evaluation of Stability Table I -Case I BOSOR Input Table 2 -Case 2 BOSOR Input Table 3 -Case 3 BOSOR Input Table 4 -Case 4 BOSOR Input Table 5 -Case 5 BOSOR Input Table 6 -Capacity Margins Conclusions Appendix A -Kalnins Program Description Appendix B -Original Stress Summaries 5.9 5.10 5.11 5.12 5.13 5.14 5.15 5.16 5.17 At-A14 1A1-1A4w 1BI-184 Cl-C20 0 0 0 0 0 0 0 0 0 Appendix Appendix C 0-Stress Printouts-CBI Computer Program 778 Input and Output D1-038 Appendix E -Computer Program Documentation EI-E3 and Verification Information

• 9 0____ ____ ____ __ _ K A /47 OAU hi sw~.3 *P I Introdution The Oyster Creek uclear Power Plant Hark I Steel Containment Vessel was designed, fabricated and erected by Chicago Bridge and Iron Company in 1965. The configuration of the drywell portion is shown on page 1A1 of the attached Appendix B. The lower spherical portion of the drywell is embedded in concrete at elevation 8'-11 1/4. A sand pocket extends from the point of complete embedment upward 31-3 3/4 -to an elevation of 12'-3.This sand pocket performs two major functions:

a) Provides a transition from the completely embedded portion of the spherical shell to an unconfined portion. The sand "springs" help to ease this transition.

b) Provides a suitable means to dissipate the thermal gradient in the merldiqnal direction.

A recent inspection of the steel shell in the sand pocket region revealed that some degradation of the steel shell had taken place at some time during the twenty plus years since completion of construction.

Preliminary information indicates that the steel shell may have been reduced from the original 1.154- to as little as .80-.90" in thickness.

This report is an assessment of the stress levels which will exist if the shell is assumed to be reduced to .70 inches around the entire periphery in the sand pocket region. The analysis is performed for the following two cases: a) Pressure z 35psig Temperature

281lF b) Pressure : 62puig Temperature

175OF Other normal dead loads and earthquake lo~as for the operating basis accident are included.S ECT FFIE REFERENCE No.-MAOD BY CHKO BY MAOD aBY CIKO Sy IA^-PC flATC nAT I nLTU P~mo 0 VSA Ailileable Codes The Oyseter Creek Nuclear Plant Hark I Containment vessel was designed, fabricated and erected in accordance with the 1962 Edition of ASHE Code,Section VIII and Code Cases 12701-5, 1271N and 1272N-5. The allowable stresses used in this reduced thickness analysis are consistent with the original cod of record. Some symbols and clarification have been extracted from the 1986 ASME Section III. Subsection NE Code. The use of these references in no way changes the allowable stress levels intended for the original design. The references used merely reflect the current day interpretations of the stress state and tend to be more consistent with todaya analytical tools.Specific references to ASHE III, 1986 Edition are: I.2.3.4.HE -3221.2 Table HE -3217-1 NE 3213.10 HE 3221.4 WR ai SUBJECT OFFICE I REFERENCE NO.MADE BY CHKO BY MADE BY CHKO MY'4-llfr fdd A (-A Ajr t t DATE -UA-E " D0ATE DATE PnrAod in USA Go Of Ho. St44 Allowable Stnaepa *Primary Stresses (does not include thermal effects)Allowable Stresses General Membrane 1.1 x 17500 = 19250 pal 1272N~-5 5(s)(1)Local Hembrane**

1.5 x 1.1 x 17500 = 28875 Psi RE-3221.2 Local Membrane+ bending 1.5 x 1.1 x 17500 = 28875 psi 1272N-5 5(a)(2)Surface Stress 3.0 x 17500 = 52600 Table NE-3217-1 Secondary Stresses (includes thermal effects)Surface Stresses (PIPb+) =3.0x17600

52500pui 1272N-5 5. (f)and RE-3212.4 all actual stresses are either stress intensities 3000 or unidirectional stresses per ASHE VIII, Case 1272N-5, whichever is greater per NE and Code* a local primary membrane stress does not exceed 1.1.xl.lx17500

=greater than 1.0 / .Ref.in defined as one which 21175 psi for a distance HE 3213.10*** if bending moment at the edge is required to maintain the bending stress in the middle to acceptable limits, the edge moment is classified as Pb' Otherwise it is classified as Q.Note: For mrimarv stress evaluation " loads include (1) rnternal pressure (2) Dead weight of Steel (3) Dead weight of appurtenances (4) 11% Horizontal Earthquake

-OBE equivalent (5) 5 Vertical 1 Earthquake

-OBE Equivalent essentially service Level A in 1983 ASHE Code For Secondary Stress Evaluation

-loads include all of above plus meridicnal thermal gradient.SUBJECT Oyo-da,.Cb(~n.~ni Amlayfw -49Wdce4 P~sat~ h~ USA Q0.Im~ ~PSa P"MM W U" 0064 Ind SAP $4 I*In~put Leading Information The spherical portion of the containment vessel is assumed to be completely embedded at elevation 8'-11 1/4 (point G as shown on sheet lAl of Appendix B). This analysis of the sand pocket zone includes a segment of the spherical shell extending up to elevation 23'-6 7/8 (point F as shown on sheet IAl in Appendix B)The boundary conditions at point F are taken from the tables shown on sheet 1BI thru 104 as shown in Appendix B. This consists of the S values as described below: (S is the resultant load in tie meridianal direction In pounds pir inch)The table on the following page is a compilation of these input loads.Note; The earthquake stresses shown on pages 1B1 through 1B4 were originally calculated for a 22% horizontal earthquake and a 10% vertical earthquake.

This has been assumed to be the equivalent of today's description of a Safe Shutdown Earthquake.

Since the 1986 Code would permit higher allowable stresses for the SSE included earthquake, the SSE earthquake loads shown have been divided by 2; i.e. l1%horizontal and 5% vertical earthquake to simulate an equivalent Operating Basis Earthquake.

These levels are compatible with todays description of the Operating Basis Earthquake.

The allowable stresses for the loads in which the OBE is included are lower than those which include the SSE. An assessment of both earthquakes with their respective allowables indicates that the more critical is the OBE case.SUBJECT OFFICE 1 EFERENCE NO.Ma. G ~- oc REVISION NI/t4 o MA .DE&c d BY MADE BY CHKO BV T DAE D A DATE- DATE_ _ _ _ _ _ _ _ _ in US 0 RV 111 0 0 c Ar 7 Load Cam .rt t~z 24~v ~ P: '4 fit 7-z7 /7,,P'D ZfA~~I ~

• 730 *MI-271 , 77 *77 St -.1 4 *14 Mioor.- c 5.. -, "o -4C?'.,,oo; AO&Wav. +"a ,4vc,%U."Ztm "4-4-" 1232- /.O~fIr RIEF~C O SUJETREVW~ON FMG NC NO.MADE BY CKO BY MADE BY CHKO BY SRTL43...QATf OAT OT OATE P.,nt~ IRUA °a' UKCV"SS'" 11 IWomne4 In USA 40441tiv..

-

The CBI Hodel used in the analysis of the reduced thicknes Oyster C~tok embedment zone is described in the figure shown below. Complete fixity is assumed at elevation 8'-11 1/4.S Sand spring& are modeled as inward radial forces, the magnitude of which are dependent upon the magnitude of the shell displacements.

The sand spring constant is 274.3 psi/inch of radial displacement.

The attenuation of the thermal gradient in the meridianal direction is assumed to be completed within the sand_ embedment zone, that is. tke temperature distribution is 1750/2816F at elev. 1213 and 60bF at elev. 88-11 1/4.The embedment zone is analyzed by use of the Kalnins Shil of Revolution Computer Code. Complete begins at 36 from vertical axis. The model continues to 67 .See Appendix A for the program description.

Boundary loads at elev. 23'-6 design. *7/8 are taken from the original 4-0 Be-ftie'X eoqd.+Ov e%&. /fj -*M~.\%I J~I4.If.4~

4 AILSa0 Aoso 1 4 I 00l t *A9 '0*e Ioodr ~16S4~ )~dia~ deed /.4 J1&)(~ Mon Zfq* 4]J()X 44 &.9.Je~c IAI Mr 04( 7 I do OFFICE REFERENCE NO.~4r ,~'cd MADE SY ICNKO ST MADE BY CHKO ST wcaI '71A ftI SHTL5r3P_____________________

A J E OATE AV DATE_ _ _ _ _ _ _ _ _W__ _ _ _ _ I __ __ _in U&A QGo I61 M SIP 84 0 Du~m&/.!A -ia,9 Olt 4 e-d044.V.

Oh 075811cqll lost 13 131. CA113 11.SSTP slat.8 THZCK60.18 PACSGT l toD 'i; I "-s Tom& a~s 1'~ed [" 141416.04

• .1111.03 0.0001.00 footstool 4*1 1*0 34:8.3611003 1.1481.02 0.0001.000 ow0 1 01681042 a Slo .*0811.04 1.5110160) 0.M00010 1.4519008 195101003 1.Slt163481.0' 1.64,1.16 0.0001.000 1.8631.00

.11.44108

&27 1 04811.00 -3.39.102 000001000 16.1181906

.13*361.00 1 38.1 1.1831.0* -l.1011003 0.0001.00 1.1641*04

-*1.0?1903 1 S6.t 1.3891904

-1.0511003 0.0001.00 1.4941.04

-1.1*4511104 1 so.t 1087*9004

-1.4"90103 0.0001.00 1.42311.04

-1.4915103 1 z1474 Ile. 1.0231.08

-$.*3I5103 Ile. 161711404

-4.1041103 39.1 lS.01.04 *.1?191.03 40.8 I1*011.04

-1.1103 80.1 l.6491*0* -l.9181l03

• 0.8 1.1517004

-6.5161.03 0.0001.00 0.0001.00 0.0001.00 0.0001.00*

0.0006#00 0.0oo1.00 1.9821104-1.8311-.4

-4.1171*1.0

-46,l1.-03

-40S162043-I.1*11.03+

]0*0.8 1.80*10*e

-4...94101

• 0.8 1013911#04

-.40411402 60.8 1.IS81.0*

-6.4511.03 3 -1981611066 0.00001'0=

0.0002.00 0.001.000 0.0001.000 0.0001100 8.053110*1.0110640 a.301440%1.10110*6.47#11.03

-41*11,1.061

-1.31*1.04l 1 I 1 I, I.3101.0*,3Ill-03.01181.4.03311.0I.80116.0*,.88l1O04.8576I0.8'.':881.0 I:..,3*1.0* 1.831 i.1681l-41.1II051.0 I N.> 1fd6Wi$St7, (MSj

• tý24Ca i i XOGNr ^rfc" Jr.#*=,SE~im 4~cI~~ .V.~.4 7 S~d I!!.Ivpav I i I i I I I h I JE, OFF WV7 I-RE -VI SOI REFERENCE NO.0jds- dC~ a~nivt IRAEVSO w Y4 MADE B.Y CI41(0 BY .ii .iCMKDoBY Alw 1f r -Amdsocd-7%t,-nGJ7-rA )'6-T"OF-A E !OATEI OATE DATE r DAT ,,Vg , P-"wa to USA 6411 io$

0.4- Hombraii Stress Intensity 12 -Surface Stress Intensity (Ini~de)+ -Burf ace Stress Intensity (Outsilde)

I I-.4 2 Sc.ek I*"- r 1te%'r/I rj4 i qv. /Aosoid CIRCUMFERENTIAL STRESS ALONG MERIDIAN OYSTER CREEK EMBEDMENT.

CASE Is P=35 T(MRXl,28l MAXIMA 14102. 14951. 18321.THICKsO.70

.pJAGV~ofop n-wee em d-e~ t 77AC~ JJ*Z g~4~ a~~ I*~vi 4£AA CAY PN t /7op I

& -M~embrane Stwess Intensity 0 -Surfaei Stress Intensity (inside)+4- Surf ace Stress Inltensity (Outotde)0 h V. sr '-I a a Scoot 1":x W1/b:- ad" qp% /Assil MERIDIANAL STRESS ALONG MERIDIAN OYSTER CREEK EMBEDMENT, CASE 1; P=35 T(IMAX=281 MAXIMA 23615. 12719. 23475.TH ICK-0. 70 WaJaCT 99 ov coma aw --- no.10,AT C"" 7WA 'IT Aou "MM44a DATV 6rr.&r 48m, &.8 I 0.4 -Mosibrans Stress 1ctensityA*

S urf ace Stress In~tensaity (Insaide)* Surface Stress In~tensity (Outside)I~sv. ~t 1'~~4 A a ScsI. 6 t 'v aWg id I kew1444 STRESS INTENSITIES ALONG MERIDIAN OYSTER CREEK EMBEDMENT, CASE 1: Pw35 TIMRX)-281 MAXIMA 32147. 25251. 29914.THICK-O.?O wiWAC P C#-t~a VA COULD.Sie-A, a- a A- Pleabrane Stuess intensity 0 -Surface Stress Intensity (Inside)" Surftace Stress Intensity (Outside)43 1I A 1-Z1 1~i 4 A$Clt I" r al I. x JOIN om 10*144 CIRCUMFERENTIAL STRESS ALONG OYSTER CREEK EMBEOMENT.

CASE MAXIMA 12445. 12655.MER!0IAN 2: P-62 TIMAX)=175 13124.THICKsO.70" -1 9 M " fIl6o I, .I l*AU Membrane Stvoss Intensity 0" -uwtaco Stress Intunsity (Inside)+ -Surfao* Stress Intensity (Otzside I ft S$a 02 S U\1 A 4 3 2 I: 3" 2/ r@8.9 MERIOTANAL STRESS ALONG "ERIDIAN OYSTER CREEK EMBEOMENT.

CRSE 2: P-62 T(MAX)xl75 MAXIMA 34324. 20108. 33060.THICK=O. 70 U.J5~V ~ my[ , -a 'to o I u ,"0~y'Mo. 4fga&A 7r44~[~J 1 4 A- Hemabr Strno Intensity-Surface Stress Intensiity

[Iside)+- Surface Streso Intensty (Outside)a-I I A i I-y-dsz.J* Lj Is 3 0 Sce I " t~:o Kg%+r* so " k416/04: z d JI co 204 STRESS OYSTER MAX IMR INTENSITIES ALONG MERIDIAN CREEK EMBEDMENT.

CASE 2: P-62 TIMAXI-175 24027. 17889. 24935.THICKnO.70 ffUj9a DRU MR T;w*u I O71A 31 re.. /1d4Z S go 0 a mpý

'SI i I I Analysis of embedment zone with Rand convipnpd to be inaf..+4-....... The CB1 Hodel used in the analysis of the reduced thickness Oyster Creek embedment zone is described in the figure shown below. Complete fixity is assumed at elevation 81-11 1/4.Sand springs are assumed to be ineffective.

The attenuation of the thermal gradient in the meridional direction in assumed to be completed within the empty emb dment zone, that is, the temperature distribution in 1750/2818; at elev. 12'3 and 60'T at elev. 8'-I1 1/4.The embedment zone is analyzed by use of the Kalnins Sh p ll of Revolution Computer Code. Complete fixjty begins at 36w from vertical axis. The model continues to 67 .Boundary loads at elev. 23'-6 7/8 are taken from the original design.1.SUBJECT orri60. 644D4ad~ sd~rwit Afti4i -90*q@d -iMJ %AW zh,9.sqa*ta, -ýevd ad7,c~II IP^"e" ft YA 4- Hbmbrane Stress Intensit7-Surface Stress Intensity (Inside)+ -Surface Stress Intensity (Outside)Z&A 4 S*to II"Its IW -- o 04mANim144 CIRCUMFERENTIRL STRESS qLONG OYSTER CREEK EMBEDMENT.

CRSE MAXIMA 13620. 14265.MERIDIAN 1. P-35 TIMRX) -281 15245.THICK-O.70

-/" SAND U&J99T 6"09 ov mm fiv ov 91"OvAlke, 0 fl& r C,,eve ZMANWA104t-

%/Se 7v#q CkK*7%164,06&

ZJW a& I eava

.b- e""brazae stres ltseit* -.UrrtOe Ststre atem8ity (laside)+ -surtaoe stress (Outlsde)4 i S.e 4 1 jr x Kil/0 Z jo 814061"44 MERIOIANAL STRESS ALONG MERIOIRN OYSTER CREEK EMBEDMENT.

CASE I: P-35 T[MAX]=28l MAXIMA 43722. 8821. 26330.TH ICK-0. 70 w/O SAND 0 ,em Jt P- -- .. ...... E" Z.*.PAF CK I 01.4- Wrame, stress Iaqtaait* Surface stream In~tensityp

• -Surface Stress In~tensity Zaside)(Outside)A !q 3 0 1:34 04001. iD7 STRESS OYSTER MAXIMA INTENSITIES ALONG MERIDIAN CREEK EMIBEOMENT.

CASE 1: P-35 T(MAXI-281 30605. 7599. 18431.THICKx0. 70 Wi SAND U649CT i oftac of *V em, 6 saw ca I N4 1,14 Z 40 VA-agog, ct-wer 7V IAA*Ivrtr-oqcdvgcd 7,h,"44& 145leelli te j I I.4- Membrane Stress latemiaty* -Buztac* Stress Intensity (Inside)* -Sgrf moo Stress Inteus~tai (Ouatside)

A 0%fa us -A e/.v.4 Scohle ~ ow 10K$+rose he, P,: 54 o 9P ovah"4 CIRCUMFERENTIAL STRESS ALONG OYSTER CREEK EMBEDMENT, CASE MAXIMA 21828. 22635.MERIDIAN 2: P-62 TIMAX)-175 23721.THICKmO.70 WIO SAND 0.... r--wv em 011 4"Wa-a, 7-hmmt W air 'SA wN-14 447 QAT;

,4o- lesmbirae 8twes Inte.nsity 43 -Bu&ac. Stress Intensity (Inid.)S- -Surtace Stress Intensity (Outside)A S V@4 j'* .J2 !j 4 a A 4$.co I.to' l it:zJd4, IV 000 4"4 MERIDIANAL STRESS ALONG MERIDIAN OTSTER CREEK EMBEDMENT.

CASE 2: P-62 TIMAX)-175 MAXIMR 53897. 16944. 28734.THICK-O.7O WIC SAND dmwFCAw- jig ov WiA" B 169.joajp -1" 1AWmAv"1'*"r'tr--

4_ck"d 7hsawm& .4 JL I *ATE I Ion

.4- Miemrane Sties Intteasty N -Sueac. Stress Intensity (Inside)+ -urface Stress Intensity (Outside)I A i.3 0%/al,,. 41111 Q 4 3 i Seek /*/oSo I,, z o Qowl"44 STRESS OYSTER MAX IMA INTENSITIES ALONG MERIDIAN CREEK EMBEDMENT.

CASE 2: P-62 TtRAX)-I75 37728. 12257. 14006.THICK-O.70 k/O SAND I-Atq GAT! .Apýro- Rcdm", evQ @A vg L .*

• The preceeding analysis indicates that the reduced thickness section of the containment vessel shell located in the sand transition zone will meet the allowable stress criteria as prescribed in the original applicable code, i.e. ASME VIII, 1962 Edition and Code Cases 1270N-5, 1271N and 1272N-5. A review of the stress plots shown on sheets 1.7, 1.8, 1.1D and 1.1* for the case in which the sand is operative and pages 2.1, 2.2, 2.4 and 2.5 for the case in which the sand is inoperative shows that the local membrane stress and surface stresses are less than their respective allowablesQ The stress intensity plots shown on pages 1.9, 1.12, 2.3 and 2.6 show surface stress intensities less than the allowable as described in the 1986 Edition of ASME III.These same plots indicate that the local membrane stresses are less than the allowable of 28875. however, the length over which the local membrane stress intensity exceeds 21175 psi exceeds 1.0 VT.The local membrane stress as shown on page 1.9 Is less than the allowable stress of 28875 psi, however the Stress exceeds the 1.16-21175 for a distance greater than I.o0W- Since the amplitude of the local membrane stresi is significantly lower than the allowable stress, we can Justify the greater length of excess by comparing the area under the actual stress curve to the area under a similiar curve in which the height of the C¢Lfve reaches the maximum of 28975 psi over, a distance of 1.0 (RW. A calculatIon making this comparIson Is shown on the following sheet. This shows that the area under the actual curve is less than the area under the allcwablb stress curve.f0It is noteworthy that the surface stress shown on page 2.5 slightly exceeds the allowable of 5205 psi. The code of record indicates that the allowable surface stress is 3 times the value listed in Table UCS-23 of Section VIII. This results in a stress of Z5200 psi. Using the 1986 issue of ASME III would permit this same surface stressato be 57900*psi.

Based upon this comparison, it is reasonable to permit this surface stressa which is less than today's allowable while exceeding the original stress level by 2.7%.SUOJECT T OFFICE RE NCE M BY REVIS'ON 9 CHKRVSO MADE BY CHKD BY MADE BY CHKO BY A~~#/yr~v' -rJAig.. I JS~ %I V~A 4 I Sld 1 0F..U PVIWd Ia IBM 2 PWed Ift tMA C~moroo ci .Aoq~s L/~de 1 Aim~sm I apce i. 4rm V66/4k 0 5-.P/04- C --t*ftsr Z6kIf 4 7 Iv, a-r-y/m or j~~*dcl C:re J' 711 14 c 47 L.4 jo Actma I tASEI ,4IIII-one ji.tcsv o6mr I m I [ I Im Aff"Up44tAoft Z~dA' c". ai4"' V./7(m, r 47.6G wflei 30 t, ge.14 a%to AIlstwib aetw &wads~ dn .dJv6wq£# qo" ZI.1 7.(w, Amw Vn'/Je.- '7/../a.fI ffaa curm =di. 4#* win.m SUBJECT OFFI REVIIO REFERENCE NO.J gr MADESY CHKD BY MADE GY CHKD Y '!OF j ..'... I ..ftaI GO" REV up u ANALYSIS OF EMBEDMENT ZONE WITH SAND PO=ET FILLED WITH GR~OUT I- z -.n sna z;, razar.: ;.a iZ GE-7a:-a:nz:at*

izr .33Lne CXo 13.~ 014-tlZF 1=i 7*~~~~~~n .azprrzmr.;£res .ume.ce #ta. Az1~ c II Sr~i zr*~SE ZAre ;raet ar -al ;rsam~l :as t.-&~--t :m. ari aQ sss.-"rato zz=.e u an~*l .r-1 'tnc zz~m af~*zac.Tlarrr

t. : is

• mcalre cm -.. ;~ a r im tme F~a esl:.:.; 4.:.re ;egl v.;ara ta ;ra;E and t-eremanc:

f isr naiesar.nt;m ztfaro-.z sta!..aria =mi in::ta omeati-ire:s t'rQ lerv Za. *nzrt' :.i r.. '-asssarm

r.~eatips forn cme lcaze zacaa as.t ..1'c e temori-e afca

3ies Inse ý-zrze 3U C &*sea% Vi i amnoc. P~agesn .=. 4:. an 4.4:earc a .az a-.: .al a tec zsant~tlec ofcw~l the ci !ý.mfanisa.

h strss. :7.a-r aT-rz:a i'tr~assom.

ana tne strert~. intnsirte

cm Viatri zassr,7 I i iv,.m le q#' &t I U SA The COT Model used in the analysis of the f.1,d eabedment son* is described in the figure show below.Complete fizity is assumed at elevation 12'3 The attenuation of the thermal gradient in the meridicnal direction is assumed to be completed w.,th.ke earit' i Ijetj,.E, The embedment none Is analysed by use of the Kalnins Shell of Revolution Computer Code. Complete fixity begins at 44.41 frome vertical axis. The model continues to 87 .See Appendix A for the program description.

Boundary loads at elev. 23'-6 design.7/8 are taken from the original+ --S ~.pe°iL*1 I*b* , e , ",._+P I.qi.'Wa4 to elV.&# so& g .SUBJECT OFFIC. REFERENCE NO.MADE BY CHKD BY MADE BY CHKS BY 4-1 t..,A fo,, DATE- DATIE DATE____ ____ ____ ____I _IVA__01,ýW In USA GO0 64k Ai&lt UPJ 6 I~.I'-I A membrane.

Istges tateanlty o S surf ace stres lutenslty (Zaside)Surface Straps I tnltyf Ut 40d)6" I I I U I U I 0 0'4 a U a S is U 0 SD C a'is S S 0'A 6 S 0 4-S I C U I I I I d I I I a U U S 0 U is a U U 0 a'0 a U U.3 26-2 IF m._ wooffr--z 0.4 Uq.4 aq ,30-,o slow. 8-1 h a * -a ~ I-20 -~0 led i0 I-ke-4b Scale I" a .& kal 1I a 30" orc length tress., bat CIRCUMFERENTIAL STRESS RLONG MERIDIRN V'SIER CREEK MODIFIED EMBEDMENT.

CRSE It MRXIMR 27850. 21238. 28925.p ',;35 TimRXI-201 Auuw TWI T%~~d7^o~

y*er DATE 'MAUS'c0WT 1 CA I& -Uubmaae Stress Kutosolt7 o -Surface Stress taeanatty+ -Surface StreAs Inteualty F¸0 I C S S h 0 S S (Inside)Cuts do)v P4 lU a 66 All b-I. ____________

-- --Ii U a S a a C S 0 4 (4 4 a 0 w4.a 6m o 0l ,300 a olov. 4'-111 I a -& -' ~-, IC'O*to*Scale V" Is fb*Abt-2b*Streata kat I-20,-A0 *Z ho kla* 30 are leastb MERIOIANAL STRESS ALONG MERIDIAN OYSTER CREEK MODIFIED EMBEOMENT.

CASE 1: P-35 MAXIMA 92539. 7274. 77991.P1.6 sea& ~a~dS w#/6.p/~d kr /./lto T (MAX) u281 hCAJAd a AftPdS ,//~i~43 hMly a 4MA9YCO_ors,..; .--, m,,,,A.,, a3+ ,1 mIil w1

ANALYSISt Off EMBDEDMENT ZONE WITH SAND REI1OVEII AND FILED WITH 0 CONTAINMENT VESSEL The ;axat~ .mzcn -set ;or tl~zz :ize :z anxen?:.a.

-a a C=t...matmt .ass&! 'ýas a .)r zf ls,:Atlzm

= t?'* ;niza The :.'sL. fto:r -ha.-a to rL&ac.1% -lwrn i-.tz tne .mrt~~:-. er tz to e4ioct:,s..

T...o~: kn :aaai :: $amrv. P-zl:ee za an 7w- zagraea. 7~ -U. .~.a 4: *. asa Z. oa GF~J.sreascnaubla izr- tle =oz;rn mi an~t.: .SB ii l .r:aamic~ra th spcif. a.;& tmarma ;r r-. !.-2 SUBJECT O FFICE EVS REFERENC Ar~l ~MADE BY C*4KOBY MAEBY CHI BY I~r~~c~~~74 ,/Z.J.A PmtEw. in uS, The C31 Kodel used In the analysis Of the 41r,0 fr.lI4 embedadm t sona Is described la tM fthil. -hom below.Complete tixlt7 in assumed at elevation 1203 Te attenuatiom of the thermal giadlent in the airidicmal direction is assumed to be completed wua.nu J* 4& .rA.mu on th o & .a.The embedment sone is analyzed by use of the Kalnins $boll of Revolution Computer Code. Couplets fixiuy begins at 44.1* fzVm" vertical axis. The model continues to 67w. See Appendii " for the program description.

Boundary loads at elov. 239-6 design.7/8 are taken from the original 4-- -e iN~"&OP 44 Psatet Ate pose A%* 9 0 ISUBJECT OFFICE RVSOREFERENCE NO.0,YjREVSIO JL4@AC4! A QkI A/ I r MADE BY CI4KO BYAD my C14KO BY 4~~.*tq '* -C ~ l vj~'. dA ~ VcA SHT OF-~A wotei QA' DGOPOT OT I@"Mee In LO GO 6,6 4V SPG,4 o-.I -A -membras, Stress ZItenslty ao -Surface tress lotesIty (Maids)+ -Surface tress InteasLty (Iam1 de)I/i i ae 0*0 ft U" a, C a U k 0 U S U 0 at U U C A S 0 Wi 4 a'p.U Ul ft S8 S I I I I I I I U ii a a a U 34 ma£0 U p 0'a a I elev. 12--3 S30"'stoy. 3'..Ill I flev. 89-111t....- -I --- j

• 4jl 100 60-4,b Scale 10 to hal stress, kal 11 300 arc Ieuitb CIRCUMFERENTIAL STRESS ALONG MERIDIAN OYSTER CREEK MODIFIED EMBEDMENT.

CASE It P-35 T(MRX)-281 H/INSULATIC MRXIMA 15766. 8496. 12171.uiwinr 0"-=c T in~w.cg Wo.Ai. ~ ~ ~ m~a Ig/*~ 7%ewvo NA~ eo 1/~ 4_____7-.MA G CHOW WAM___ __v ween IIiv 0-..A -Voubtame Stress latessity a -urfacr'sltea ZstlaeaL (ltaside)-Surface Streus (ten Ity atu$de)I* *1 sibi gU as00 Ilo do A a -io -,b Scale 1 a L4 kat actress, kha V a 301 arc laostk MERIDIANAL STRESS RLONG. MERICIRN OYSTER CREEK MOOIFIED EMEEDMENT.

CASE Is Pz35 T(NRX)-21 W/INSULRTIC MAXIMA 52552. 5990. 4OS71.10 0 MUM, ,7 VA ByIi TIi MA.DAT m wew. CA vmm"ýomm"mm 0*

I.*.-6-Haabriae roeets stleot* y .... *-*-Surface:-

tress Zatuensty (Tasode)* -Surace Streus I tutenLty (utceotde) 6% 0 elev. 121-3 4#A ad 04 w0 oo ?a 6" ib "... -o-4b scale 1" -L kst stress, kst 11 a 301 arc loustb CIRCUMFERENTIAL STRESS ALONG MERIDIAN OTSTER CREEK MODIFI 'ED EMBEDMENT, CASE 2a P-62 TIMAX)-175 WIINSULATIO MAXIMA 13750. 12S26. 13627.YO IT. g I OATl 1 I 6

• I I esavewr.I mý, 1 0 I S N if a -yeabrase Stress Istsceht7.

(3 -lcW Stvass Istensity (16l0de)+ -Surface Sterme ThtouLty (QutsL4.)r.U I I I S&I 0'4 a'88!ao aq U1 1w 0 4,.a1 0 ad'4 a V II a'a Si a I I a'a ED 0 a'0 S U j iý O"'elev. 12'-3 300.,*LOT. V.oIlk E~I a 100 do Scale 1" 1t 4b0" ab-io a EO kat a 300 are lIaltk atre&&, kal STRESS OYSTER MAXIHA INTENSITIES ALONG MERIDIAN CREEK MODIFIED EMBEDMENT.

CRSE 2: 32084. S528. 16913.P.62 TiMAX ,l7S kt/INSULRTIPN

______DATE , , , I DL T I,__E____,.,___.,.,.

____, ___, __" __,, _,, I 2. __ __,,_\. 1 Conclusions for Section 4 -Sand Pocket Filled with Grout The filling of the sand pocket with grout up to elevation 12'-3 results in surface stresses In excess of 92 ksi. This is well beyond the AStE allowable of 52.5 kat.In order to reduce the surface stresses to the ASME allowable of 52.5 ksa, an insulation system is required which will ensure that the temperature of the shell is uniformly decreased from 2810F to SOOF within an arc length of 40 inches minimum., 0 I Ll TASK 4 STABILITY ANALYSIS OYSTER CREEK CONTAINMENT VESSEL GLOBAL FAILURE OF THE SPHERICAL PORTION OF THE DRYWELL DUE TO STRUCTURAL INSTABILITY CBI NACON FOR S GPUN FEBRUARY, 1987 0 This portion of the drywell analysis of the Oyster Creek Nuclear Power Plant is intended to assess the capability of the drywell, shell in the 'as found condition" to resist gross structural instability.

This analysis includes an investigation of five loading conditions as follows: 1. Conlainment internal presure at 35 psig and temperature at 281"T along with a .11g horizontal earthquake and a .05g vertical earthquake.

This load combination is being included because the previously presented stress information indicates that unusually high circumferential compressive stresses are present in the sand transition zone for the reduced thickness shell. The stress state used in the analysis at various points along the meridian are taken from the previously presented embedment tone analysis for points at elevations V"-11 and 120-3. Stress states for points at elevations above the embedment zone are taken from the stress output generated by CBI using CB2 computer program 778. A description of this prograw is included in Appendix D. A more thorough description of this reanalysis of the containment vessel drywell is included in the following section.2. Containment internal pressure at -2 psig and temperature at ambient along with a .11g. horizontal earthquake and a .06g vertical earthquake.

This load combination is included in order to investigate the most likely operating conditio which could result in gross instability.

The stress states used for the various points are taken from the original design report, page 1B9, except that the earthquake stresses have been adjusted to the llg and .056g.horizontal and vertical earthquakes, respectively.

See page 5.9 of this section for a tabulat4d listing.3. Containment internal pressure at -2 psig and temperature at ambient with a .lii horizontal earthquake and a .05a vertical earthquake.

This load combination is the same as that listed in 2 above, except thAt the stress states have been determined by use of CBI computer program 778. An explanation of the load inputs and justification of the use of the program is included in Appendix D. Analysis technique is the same as described in I above.SUBJECT OFFICE R REFERENCE NO.40 46ACK -r IVISION N6//4 7 MADE BY CHKD BY MADE BY CHKO BYA'"J'A _-- ATE. OATe '6PTE / 7/.67 4S UV W114

4. Containment flooded to elevation 74.6 feet and ambient temperature along with a .22a horizontal earthquake and a.10g vertical earthquake.

The stress states have been calculated by use of C91 program 776 and for this case, the flooded drywell is assumed to act as a cantilever beam.Although the drywell is coupled to the concrete shield building at the stabilizer elevation, the effect of the stay force is assumed to be zero. The cantilever condition presents the upper bound.5. Containment flooded to elevation 74.5ft. and ambient temperature with the same earthquake as listed in 4 above, however, in this case the drywell is examined-as a propped cantilever with a stay force imposed a the elevation of the stabilizer.

The concrete shield building is assumed to be stiff enough to exert a sufficiently large reaction force on the flooded drywell to reduce relative displacement to zero at that point. This is assumed to consitute a lower bound case. (See NURIG/CR-1981, pv 41 and 89 for a discussion on the effect of cantilevered vs. propped cantilever analysis)GENERATION OF 9TRESS 9TATEA The stability analysis is based on the use of the commercial computer code BOSOR which provides a means of determining the theoretical stress state at which structural instability will occur. CBI recently obtained an improved version of the BOSOR code which allows the analyst to incorporate the elastic spring effect of the sand in the sand pocket.SUGJECT OFFICE R REFERENCE NO.I za t 6EwIS Af, 114 7 MADEBY CHK Y B MATY CHKO BY"7-'JA IP14 s6"7,.1OF_.Z 7, E Pftft OR U" Q0GtWVW1A In order to use the BOSOR code, the stress states at approximately 10 points alone the meridian of the drywell must be available as input Information.

In the formula for theoretical buckling from code case N-284, par. 1712.1.3(a), the stress state it assumed to be equal biaxial compression of constant magnitude throughout the vessel. The actual stress states for the five loading cases described consists of varying meridional compression with varying circumferential tension for load cases 2-5. Case 1 is circumferential coupression with meridional tension.The original design report consists of a large volume of long hand generated calculations describing the stress states at 7 points along the meridian of the drywell. In the years following the design of the Oyster Creek containment vessel, CB! has developed a computer program 778, which performs the same type of analysis as the original design report$ but allows for more rapid consideration of various loading conditions, including earthquakes of various intensity.

In order to expedite the stability analysis, CBI generated an input data set for the Oyster Creek containment vessel. This data set includes a great deal of relatively non-essential input loads which are not readily available from the Oyster Creek original analysis.

In order to include some representative values, the input loads from the Fermi II containment analysis were used. All essential input loads unique to Oyster Creek are incorporated into the data set. Major components such as the-personnel air lock, equipment hatch and beam loads are correctly included.

The Oyster Creek earthquake accelerations are also included.Appendix D includes a copy of the input data as well as the printout for load cases 1, 3, 4 and 5 (Case 2 Is the original design report information)

METHOD OF ANALYSIS The analysis approach consists of an axisymetric shell of revolution analysis for linear bif'~cation buckling.

This approach produces a reasonable assessment of the buckling capacity of the structure.

Since the analysis is a linear elastic approach, it is valid only while the structure remains in antelastic state of stress.The analysis does not include any. beneficial effects of penetrations or attachments which would provide some support to the vessel. Any detrimental effects of penetrations or attachments which would cause a concentration of load over a local region of the shell are not included in the analysis.SUBJECT OFFICE REFERENCE NO.56,6 fitot'c REVISION MADE SBY C KO St MAD E BY CH KD Y I IVA N4 , ,_ , PATI DATE GATE ATE 2///aa7/87, ..I ftl mum44 MYGa Wi The analysis examines the structure as a free standing shell loaded by its dead load plus earthquake load using material properties at the temperature of the shell.This case is anlalyzed using an improved version of CB1 program 31443, "General Shell of Revolution Analysis with Stability and Eigenvalues CBOSOR4)." which allows the analyst to account for the elastic restraint in the sand transition zone. Linear bifurcation buckling elgenvalues are calculated considering axilsymmetric loading with buckling occuring In a range of potential harmonics.

No initial Imperfections are included.

The eigenvalue results represent a linear scale factor which when applied to the input loading produces a theoretical bifuracation buckling load. The results are then multiplied by an appropriately modified knockdown factor (reference ASHE Code Case N-284, Fig. 1512-1.)BOSOR4 Computer Program BOSOR4 is a comprehensive computer program for the stress, stability, and vibration analysis of segmented, ring stiffened, branched shells of revolution.

The program includes nonlinear prestress effects and is very general with respect to geometry of the meridian, shell wall design, edge conditions, and loading. However the wall must be thin enough so that thin shell theory is appicable and the materials must be elastic.A summary of some of the programs's features follows: 1. Analyses:

auisymmetric stresses and deflections using nonlinear theory for a stepwise increasing loading, vibration modes and frequencies, nonsymmetric buckling modes and load factors using an auisymmaetric prestress, either given directly or calculated from a praproblea (either symmetric or at a given azimuth of a nonaymmetric solution);

stresses and deflections due to nonaymmetric loadings.2. Geometry:

spheres, toroids, cones, cylinders, various types of rings, and general shell shapea using spline fits 3. Wall Construction:

layered construction with each layer of a different material; inner and outer surfaces can vary relative to the reference surface 4. Material Properties:

isotropic and orthotropic materials.

special cases include fiber-wound.

corrugated, monocoque, and semi-sandwich constructions SUBJECT ah OFFICE 1 REFERENCE N5.MADE BY CHKO BY MAEB MDB 1.4___ __ ___ __ ___ __ _A EZ9E ATE D ATE______ ______ ______ _ 2 ~ 7 Z 7/e nu VIA I I 0 5. Boundary Conditions:

displacements spocifled at any mesh point (including nonzero displacement), any point can connected to any other point.be 6. Loading: pressure and surface traction loading, line loading, thermal loadings including a gradient through the thickness (for linear stress analyses these all may be nonaxisymetric loadings in the form of a longitudinal distribution times a circumferential distribution), in buckling analyses there can be fixed and variable loads 7. The stiffness of the elastic springs in the sand transition zone may be input. The springs are ot equal magnitude both inside and outside the drywell.Program Limitations:

1. Thin shell theory must be applicable as well as thin curved beam theory if rings are used.2. Haterial must be elastic.0 3. Structure must be axisymmetric.

4. Prebuckled deflection, more than moderately meridinal rotation can while considered finite, must be no-large, i.e. the square of the be neglected compared with one.5. In the calculation of displacements and stresses for nonsyumetrically loaded shells, small deflection theory is used.SUBJECT OFFICE RREFERENCE NO.MAD B CMKO BY MADI BY CK BY PAE DA. DATE DATE....... 11 7/A P ý mum 0 q14 l14J 1 BUCKLfING ANALYSIS 5PMMICAL POiRTION O0I CTAUUHN? VESSEL Analysis A. Introduction The purpose of this analysis safetT against buckling of ti containment when subjected to program BO=OR4 (see proceeding B. Geometry and Modelling The dimensions of the model usi shown in Fig. 1.C. Haterial Properties Is to determine factor of hs spherical portion of the various loadings using the 2 pages).Dd in the 0OSOR analysis are The properties at ambient temperature for the SA -212B, FBX, steel are taken from the ASHI Sect. III code. The values used in the analysis are:..M Elastic Modulus = 29800000 psi Poisson's Ratio a 0.3 0 D.Since the eigenvalue is directly proportional toE, it can be scaled to account for other I's.Loading and Boundary Conditions The loading for each case is given in Tables I tblru 5.These stress resultants are given to the program as an initial axisymmetric prestress conditions.

The program uses linear interpolation for the values at points in between the given points.The shell is fixed at the base of t.he sphere and free at the top of the knuckle. Any mode shapes that involve significant movement at the top are ignored since the knuckle is in fact not free and modes in the lower regions are desired.The preceeding pages describe the sethod of determining the actual stress levels for the five loading cases and the DOSOR values which identify an amplification factor which will trigger gross buckling.

The followng section is a presentation of the determination of the buckling acceptance criteria.SUeJECT*Wmki U~ I*& 00 Gd MV UPSA P Wftd I% USA 606"rv"FIA O.C. CIONTR:NMENT, a8uCKL!N0 ANALYSIS.INITIAL UNDEFORMED STRUCTURE lN0tc o '4 a INCICATES EACH KISH POINT 400.300.200.100.z 0.4 7 I 0'N 0-100.-200.-300.-400.-100.0 0.0 100.0 200.0 300.0 400.0 S00.0 600.C 700.0 R , 84IUR498 ta/17/06 I.0 SUBJECT OFFICE REFERENCE NO.B060*.. ,,,,Y,,I" ab A/06 REVISION A./,21147 MAo BY A HKDBY MADETBY KAT gAy ATE AL OJ ATE DATE Go0i4 R[V ZP IA BUCKLING CRITERIA HODIIRD CODE CASE U-284 I. Modification of the Theoretical Buckling Value The buckling criteria described in ASHE III. Code Case N-284 is based upon determining the theoretical critical buckling stress (ref. 1712.1.3(a))

which may be taken for this case to be Or 0.605 tT/R.This theoretical buckling value as presented in the reference code case is based on a spherical shell of uniform thickness in which the unidirectional compressive stress is of a constant magnitude at all points both along the meridian and around the circumference.

The orthogonal stress is understood for this formula to be equal to zero.The loading conditions 2 thru 5 for the Orster Creek configuration consist of a meridional compressive stress which increases in magnitude at descending elevations along the meridian.

Corresponding respective circumferential tensile stresses also vary in a similar fashion. Recognizing that this loading condition cannot be realistically represented by use of the code case formula, we utilize the computer code BOSOR which is an equavalent formulation used to calculate the critical buckling value for the varying magnitude loading. The input for use in this solution consists of the calculated meridional and circumferential stresess at 10 randomly selected points along the drywell meridian (see figure on page D3). These values are shown on the following tables labeled "BOSOR Input Values".These tables also list the BOSOR ,output Kigenvalues.

The tigenvalue is the multiplier which,when multiplied by the actual stress provides the theoretical value for critical buckling of a shell not subject to any imperfections.

A 0 0 SUBJECT & OFFICE ' REFERENCE NO.SUBJET ~ ~~&a0c REVISIONI MADE BY CHKD BY MADE BY CHKD BY s/I o__ __,ATE DATE T DATE.. ... _ _ 7/67 1_1 I leffilftu" l 00 0 FAV W, 66 0 II. Hodification of the Capacity Reduction Factor The theoretical buckling value is aultiplied by a ocapacitY reduction factor' (ref. fig. 1612-1) herein referred to as a"knockdown factor" which accounts for the additional stability of spherical shells which are less subject to imperfections of construction, mainly through the use of stiffening rings. figure 1612-1 includes two curves; one for the case in which the stress state consists of equal biaxial compression and one in which the stress state is uniaxial compression.

Although the analysis presented in this report could conservativly utilise the uniaxial compression curve, further modification may be made to account for the case wherein first direction compression is accompanied by orthogonal tension. This orthogonal tensile stress has the effect of rounding the shell and reducing the effect of imperfections experienced duking the fabrication and construction phase. To arrive at a method of quantifying this , effect, the following technique is used.S-Q.l26"'E(T/R)

+&A' where h~r is the term which Accounts the stiffening effS~t Of the ciquivalent internal pressure and is found in accordance with "The Stability of Thin-Walled Unstiffened Circular Cylinders Under Axial Compression Including the Effects of Internal Pressure*

by Harris. Suer, Skene and Benjamin which appeared in journal of the Aeronautical Sciences-August 1957.Using this reference, the parameter X.may be determined to be x -JR12 The equivalent internal pressure, P. is the pressure which would result in a tensile stress equal to the calculated orthogonal I stress and may be found as follows Peq .2t x 6 (tensile)R using the Peq, determine the WaQr as follows: cr Y=.01983 + .7886 x -1.5272 x 2+ 1.5208 x 3-.73323 x 4 + .13398 x5 AG Et~cr R SUBJECT OFFICE REFE.ENCE NO.j!2.r REVISION A-14 MMADE BY1 OIKOmy BY ADI 4 ChKD GY M ADEIY SM ,1 0F P,4 1_____DAE ATE DATE DATE___ ___ ___ ___ ___ ___ ___ __ 7/67 1_ _ __ _ _POW" M W" Q0. ptWWWWW

• ,% a.. .IV%&B W 0 "rt 66%6n 4 ,,A~qMM 0mup. aft 6 vr~4 WA W" Wrof Wd1W "YKA d~IL "W M4 Abl dw m a 0 Fczmulas are a curve fit for charts included in paper above.listed Thus, the critical bucklinzg stress is the sun of the Code Case N-284 theoretical critical value multiplied by an initial knockdown factor of .207 plus the modified increment which accounts for internal pressure.

The result is Sx .125 1 t/r + AZ ocr or The ratio of the compressive stress So, to the theoretical compressive allowable V , is used to oetermine a modified"capacity reduction factSl". This will account for the enhanced ability of the shell to resist buckling.

Thus; the modified knockdown factor K~mod -V.,.S-r (theoretical compressive allowable (6c)IlI Evaluation of Stability E The capacity factor margin of the containment shell compared to the appropriate allowable stress is determined as follows%1. Determine the BOSOR Eiqenvalue from CBI computer code.2. Determine the modified knockdown factor as described above.3. Multiply the BOSOR Eiquevalue by the modified knockdown factor. This value is herein refered to as"Capacity Margin". These capacity margins may be compared to the ASHE required factor of safety -usually 2.0 minimum.The following table presents the "capacity margin" factors for the five cases analyzed.SUBJECT OFFICE NREFERENCE NO.MADE BY CHKO BY MADE BY" TiKosY B.DATE AT " OAT!E DATE..........__ _ _ _ _ 2~_ _ _ _ _ _^m in G006414II11

/A I S 0 C 7-o-, 616 ,,.- , Zo d,,v i /x'I P = lrprl 7-- 6 ;*F a ft.// P,7.. Pt.2 pr, f Pr. P5 Pr.7 Pr. 7 Pr.8 Pr. 9+PT,,o f £4 3293 45".3 6989 701C 7o4_9 64,80 e445" .,3 4404 46168/e /3z 4299 676C. ,7?8 4900o 1384 .847 4170 24o" 25.'oo f CeP 1331_6- 2te.6+' 20o12 7407 74/f. 8<Zo 8298 93.;7 1 //sz -"f5a-Eq /34. 2S6"9O /946 7a3Z 77Se 8314 6899 /10440-o4/

9.ZZ6 0"-'I lob-d41 i 0 2 M-691 44o4 E40o" S 91 9'2j P-eeAf 6200 Oo0504/4#0,= 17304i 1J,.40~___________

________ I I.....~..A I ________

0.rc7 6/u e P.ý,- ,Z 0 ad Ca re- e 8O$~'4 XsJI'L.'T VALUE$0"'170"/ A.,ftr5Xr j:6/u"v,.q4j II Pr.,or. OP F p f-/'r I- /Y -I i c0 1 AA: I L-Oas ea-l 3 VA L w C c PZ P-em 7-: 7o0F pr.7 Pr.,8 P ý.-9 orýo V Mexico-421-720-76k-Wo4a-745f-971t-7"6-.85r-/S45!-/361 2.4S$'41.-/320-4/17-24Z9-32O2~,CI4WA4-997-12/8-2374-3091-/74 49-"14-/9 f2 z 1124/550 2843 3~I6 2494-e~/f4 m-4 U,'8 p.-4 m~I £p.0 U-C-C U-C 0 I 0 0-C 0 I I z 0 t I I 1~~~~14*i mEqlo-4G645 All z- I ICIPXOAI 1-Cc?-Ifo S__________

+/- ~ -I -A £ ________ L ~1

/.ýM mw p U-I.744e A 4 Zood gve~ 4-1P454AP -lAd. OPW~ V/A I-~ v $.F4.S~J& e/8v._744-

/J7d*, 7/eVCe-r cbpd.61 -NO i 7 4 cc I I.212P)voP7:2 P3 P'nt 'pr.m5 P 0 7f, Pr77 Pr78 #Or. Prio-( 12,6 --4 /Ic 41Z -7989? -122Z3 -223-4t-Eq 425~7 -'-44627 79142 /'8/3' AVIS3 *g 7o74 N-q m a-4 Sm U.4 U.4 U.4 2 0h 0 17 1-Ae-/2?t.3 3 2 I ii JR p icloawU-te';, p Zwe =2A. 907 -"r-3____________

I I I I I .1 A £ I -A i... i........i

-_______________

I-id 0 I U.7a-6&.X -zea-d AZose S X~ Ao Pvr7 e473e ,;*F7oo~4 d~e/em 749-fr? Wot4 jý1,1vr Mw7*Pr. Z.p P. a oj&d.2~q I--, t 9-.-Pr.2 Pr. 3 p'rf Pr.', 1i')0, Pr7 Pr.78 Pr.. V/I Pr7,Io-I -£ 4 4 -I 4 4 a u-'E Alselo.I.COM-12e-4267-/70 -a331 -/14o-42S3jI-d5aoj-1"473 411=44~o It7G/3 fr.~4l~G fr/~z4t ~~'fl9~o 4-0 0 in'U K U S-C K 0 fq S 0 U-C I 0-m444g.4 4 4 4 4 4 4 I-eq vAl IA,,, foe PY y1reJ~.- &5"~'44 i. 4 S f Irl I gt -I -9 1 U W 0 I a I W c Ad~lasemeo 4 adei Miw/234 A:e.* PaYo" acq~~7coP- ov-t/~Aoapo

/ tb'pqm3 4jIaV 77-ft' ' ~c~~d I s-,h E17660oooO*pi E 28 3¶0ooo Art.4v 7'I'-p, , .t. #, -s .p -d834 ÷Z9to.97 gjl /6aS3 .449 7. s .Z6 K -,a. -.* oao ..7-3 3765 /9?37 '333 /Z.2.+ 4.07 71 1* A.__ I__ _ _ _ _ _ __ _ _ __ _ _4"I iR I a*1~$? 721f., , -s / ./ .-/AiJ I 8-7/4339 L"~ '~ I ~ 4 4 4-4 ~ I -t a-4 C U.K 1 a m U C)I U-C'IU-0 2&, cle., .,-,I*./ I+1,6297 95-3'rk ~UP 74.*5 fI 247 ," ."'1. '31; 9 4 j.$i/ 4.I 2.3?i _____iii_____

I 8 U m'%m A z r.., 4to Ye4,v4rA'?0 v /gprf, a -/ X0.-

• 1 dr, 4 74 -.. A-' ,, .z,/'-I-I-I-'

-0 f -t I The column labeled "Capacity Margin" shown on the preceeding page is a tabulation of the capability of the containment vessel to resist meridional or circumferential buckling in the sand transition zone for a minimum shell thickness of .70 inches.These capacity margins may be compared to the ASHE required"factor of safety" against buckling.

The table shows that the capacity margins for load case 1, 2, and 3 are 3.26, 4.07 and 3.25, respectively.

All 3 capacity margins are greater than the ASME required factor of safety of 2.0 and are considered to be acceptable.

Load cases 4 and 5 are included to provide a range of capacity margins for the "flooded to elevation 74.5 feet" cases. Case 4 shows a capacity margin of 1.64 for the containment vessel acting as a free standing cantilever. (The original stress report allowed a capacity margin of 1.0 for the flooded plus.22g earthquake condition).

Case 5 shows a tapacity margin of 2.39 for the containment vessel acting as a propped cantilever. (This is the condition utilized in the original stress report). The true condition of support probably lies between the cantilever and propped cantilever condition, i.e; capacity margin of 1.64 and 2.39.Either capacity margin meets the requirements of the original design specification.

Further refinements of the earthquake loading, the coupling of adjacent structures and the fluid structure interaction is recommended.

These refinements were beyond the state of the art used at the time the original design was performed.

It is highly probable that additonal refined analysis will improve the capacity margins for cases 4 and 5.SUBJECT OFFICE REFERENCE NO.tf .QL~gCREVISION NG4 114 7 MADE BY CHKO BY MADE BY CHKO BY 77A 0001f-./7o

_DATE DATE DATE DATE 114 A S/ /E Pom" in UIA 00 4W OW UP at 0 Appendix A CBI Computer Code Description 0 Kalnins Shells of Revolution Program S J-.gA InS .ON 112- 6 l'fto.raa1 O : te she ls. Of Revolution Program is the Chicago Bridge. &'iron Coqpany Pivgraz 7-81. The program calculates

.-the stresses and displacements in thin valled elastic -shells of revolu'tion when subjectea to stitic edge," surface and/or temperature loadi with arbitrary distribution over the surface of the shall. The geometry of* the shell must be.jsy. tric, -but the shape of the median is arbitrary.

It Is.possible to include up to three branch shells with the main shell in a single model.* In addifion, the'shellSwall may cu..tst of four layers of different orthotropic materials, and the thickness of each layer and the elastic proiprties

"f each liyer "lcni`thi "iidai " The 7-81 program nuserically integrates the eight ordinary first order differential equations of thin shell theory derived by H. Reissner.

The equations are derived such that the eight variables are chosen which appear on the boundaries of the axially symmetric shell so that the entire problem can be expressed in these fundamental variables.

Chicago Bridge & Iron Company has extensively revised the Kalnins Program. The program has been altered such that a 4 x 4 force-dUsplacement relation can be used as a boundary condition as an alternative to the usual procedure of specifying forces or displacements.

This force-displace-teat relation can be used to describq the forces at the boundary in terms of displacements at7 the boundary, or the displacements at the boundary in terms of forces or some compatible combination of the two. In this manner, it is possible to study the behavior of a larse complex structure.

SUBJECT OFFIC= EFERE ff./ , ý P EVISION I / 14 Aafyr tsal-- 7v&s7 MADE BY CHKO BY MAEBY CI4KO mYSH TA::-A,1ev4 A -P-7~.. ba 6 ATE DATE DATEjATE-__"Yo __/f _____I pft" $A U" it Is also possible to introduce a "spring matrix" at the end of any part of the stress model. This matrix =st be expressed in the form, force -spring matrix x displacement.

In this manner it is possible to model the restraint of the sand cushion in the transition zone at the point of embedment.

In addition to the above changes, the Kalnins?rogram has been modified to increase the size of the problem that can be considered and to improve the accuracy of the solution.

a I-I A'tw SUBJECT OFFICE REFERENCE NO.o"7 CHK0BY MADEBY CeNKD 6 SNTAIo47 Ae,,,.w,-x A -0ow ,w- & I" ATE DATE- DATE DATE , '//c~/IN", Mu " U 00 to REV lip U I LdMJ I C L. KALNINS As"**t~l PFfstw of obibwo w g A~pggd kigiwe.'Few tVagty*Nowv KOewe Cesa Ahow AwM Analysis of Shells of Revolution Subjected to Symmetrical and Nonsymm.nttrical Loads'I rhe Iendarroalue probhr of deforn..aiox of a roaiana Bly symmawiricskel i steated in wwffof a new Osyna eff Au-at eder ordinary ia~eren&)sguai"Lqe

UJJeA carn be &riai~e Jor any s.uontria liamo hendix&IAwry of cAid. lb dependent ecriabhtu eomimedeni A&I syamea of ogsaMlsensa re okere guasi %Akfz th appeal in Mhe naftsri lioawivy uuaditiesu na roweiieuaify slns,.egrio edge qf a s&eW af ro&exeuir.

A iramerdnl MeAWd. ofl"IS&% whieh emneiss the edwuagess of Wet Ake direet 4iseratieu end ase Ania .~j'reae aprachis netpedforSh anlyi .1reltiefy tymn.metric AdaIt ThM .tA~d elimiaate the k"as .1~ w scu ogecenauxred in LUSial europpficadsie of &U 4-dir isugmtkn eppreack to s ae .adyis of Shells. Fa4r ate tarpone .1 Z~udssiu.stresses gad 6isplacasenak of e pressariaed briar ow mkskktd and d&Mtaild ammerica resUl eff pau-vnu Ii I in. shell of revolution is an imnpartian structural elmen sand he liturature devoted to its saalyeie Is extensive.

With regrd to axisyimmeric deformation.

various metd have been employed to obtain etiutions of the banding theory of shells of revolution by moans of the H. 6Aima-Maimer equations.

For example, Nagldi and DeEva 111' we symptotis tapu&e.tion; Lotnan 121. MOna II, Ml&e 141. employ a direct meu,.ical intaopagon, p"ra"c; CaGlely, at &L 181 And the solt-* , Nailoeal Ioeam Grant No. 3 Raport So. L.July. 1093.INumbem in brak.ets damlgtale Raefrmam at wAd o1 paper.Peamstasd at the S Immo Confirence of the Avlpld MeAa'.Boulder. Colo. June 0-IL 1064. at Tim Aleuuox locta or Mrmxic:A.

Earorrnma.

Discusion tof s paper thoAld ba addresed to the Wditoral Do.ptartmens.

AiXM. United F, nagnwl Center. 34 1r ma47th Strmet.New York. N. Y. 10017. and will be semoled unto OCelaer I. 19084.Disumion soeosved after th gaming dat will be oftmL Usaw.scrivt sceived by AMlE'Appld Memesllm Divisi. Jtr It. 19I.Pmer No. 6--A- 43.L%ion for an ellipsoidal e of revolution by both the InitJodiffwr ence and the Riange-Kut a method; &ad Penny [61. Radkowskl.

et &1. (71, mand Stpetoshi at uL. [1 utilize the Wntaodifrarnce technique.

A aumber of additional refernsm which " with the *Dlution of the Hf. aibmmaer-Uelammr equations can be found in the Ppuae citad..For problema it bending In the abosncef axala eymmetz. a reducton of the goerning equations of arbitrry sIesla eo nvol-0"o to.& eystae of four escond-dr differenti.l equations in-volving four unknowns hes boa cmied out by Budiansky mad Radkoweld (91. A method for *taining the mslution of thvme equations is ,i in lei which Is an extnsion of that employed in and M181. Fwubrterors.

VLumente of sonaymmstrio dfortation, of shells of revolution an found In paper by Gold-beg and So~l 1101. whe meytaem 1 c( drsordmr differential equation for emial &be& b deived. sad by 8Mel e111 and labile 1121. where solutlos 4t emltai "p are eonsdered by mnas of Asymptotie integration.

Among the papers whih employ auamoural aslyule two df--NamlnfIltIlra C -eoordinatm of aPint ef o a dst s mesuaed from an abitrasy origin along meridias in positivt dinection of*to. 0e. a -unit vectors manget to cooreinate eurVes (M Fig.11).-pricipal radii ot aurva-lure of middle eurfamc r "distance of a point go, middle e7surfae fo*axisaof syrnnaty Z = Young's mocdubus V -Poison's ratio A = tlackneesofsh&W a -coefficient of tbermal ex-9 410/12(1 -KC -ft 5 (I- NI)a*. Uh v MComponents ofd Aplsto.ment of middle mudace M~e agen of rttionof Nor-mel-components of mechua-cal surface loads ,* ata m components of m 1mat of sirface lods", T, -tmperature incrment and temperature rw-sultanta No. e, NaN -membrane guess riemml -ants MV. Art Meo -Moment resultant Qe e-tranaeverseebear result-.)., A(s)derivative rith rempect t4 any Coordialto order of ssteum of eqaua-sumber of seamzats independent variable, witMer* of end point of gment (in, 1) matrix. fundamen-tl vailabim (N ,) matrix, Coc&-cents of dwuegratual o~luatiool (w, 1) mauix. nonho-mogeeowus coefficients (a,. w)maturt, ho-ogene-oa solution$(a. t) onho.motmaeous solutions (%. 11) matri&, arbitrary unit Mataix 5(s) -0 1 ofl~ctive0hesw resultants r(s)J iI/R. + Vn 0/r S- /R e-#in /O a i-ntegr. designating t C -Fourier esmposoet$lkngth factor r A-3 V.

I fatent metboda of frtea d Qe prolsm of deformatieoa of sbells moul be noopized; La., db direct lttsgrs, tics [2461 ad the MAnta diftwe. approleh [$41. Whe the diec iAte&r a io proahbaa u sain important Mvataug, it o ha balse crious disadvantag; L.e.w*ken the lmA of the shell" -. -isiner~uea alm y iniviably nrdt- Tkis pheaome-son wus clearly pointed out in [81. Tba lose of acc-,ur dones not atault from accumulative srors in but It Is caused by the subtraction a( elmost equal numbers in thb procea of deter.mination of the unknown boundary vahmL It follows that for every at of geometric and material paramees" of the s"ell ihee is a critical length beyond which the solution losm .s aOcuracy.The advantage of the cute-dicronce approach eve direct iatw gration io that it Cataa void such al.oess c uWIaq. It iseo-eluded from 19I that it the eolution of the of algebrais equations.

which rmeult from the Iits-dU.4ierss squatlena.

is obtained by mean of Gausian eliminati.

than so lose of an.curacy ie experienced if the length of the shell 4 hiaesd.This paper is concerned with the goneral problem of defonan-tion of thin. elastic ehalla of wevolution, at non-symmetrcally loaded, end with the development eof ,- -A method of i olution wheck "M the direct intot Lion tecb.niqu*, but eliina Ote Ovin jacue C. t h. hn ~or p~i~unaeyvu~e poblm whic go!e ed.-WMhaneain-tervi- bya syutem of at fltre iered~ differential eqaaatwnUý-

j32unayooii rrwenh at Cub and of the intrva, It is shown that te boeund7valu-problem of a oa symmetric s "a be swtad in this form for say conuitaent nawr beding theory of shells in trn;s of quantities which appear in the natural boundary condi dlons an a rotatioully symmetric edge The method of tis papr offe defite advntag. er the wniu-diffsrence approach.

The main advantages as: (e4LL* " t- can be aE edconveniently to a bIn s te a6eord dif--.ent~ meqatinsand C6 iemst , nQ-ttomatie sl a Of Lan oPimum stepansa of ifptioat dxiseach se 1j-h daeilred accuracy of temsouti T'L bra Fkatme-Ansta

.~uaion ofthetha~i-or hg1 jeourlon chrsctezissd in te so f &at..rsrdar diferetia evatina. &MnM he iteatedfuth8 reuctiont attM eqatfons t- smsller sum.bir of unknowns is act aseosary.

.s t mu t be of preat importore ifa truly melm d i d w noes-n soon ith the Anite-diftermas approsakh a meaning-l d PP ut mataU' oft4he step " m " often dis ilt. if not 1m,-possible, epecially when mpid thap. sad discontnultio in t"e shall parameteus am encountered.

IV a direct Integration approach is employed wiQ the method ofth paw, tbn the stop ass an be seleted automAtically at each step which ensuru a procnibed o the solution and optumom e aiency in the clulaiona.

Tho method given in this paper sa be divided into two peart: (e) Direct integration of .t + I ilnitval siue problems over pro-eclected segments af the totw intervl, mad (6) the un of Goue-san el, tination, for the solution at the Irsulting sstem of matrix equations.

The Amnt peat at thi method is a Ituarslization at that which Is employed over the whole interval in [M-41. ere, however, the initial value problem us d4flned ve sgmients of the total interval, the lengtha of which we within the rang the applicability of the direct integratio approach.

After the initial value problems am integrated ever these smemts. eontiAulty eonditions on all vsriablme ae written at the andpointa of t se nts.mand they a simultanouos system oG law mnauix equations.

This system at matrix equations Is tha, solvedby mea&s of Gaussian elimination.

The rmlit Is that the Sdirect Integration method is employed and at the eama time there is to loss of accuracy because the longths of %te asgmente ar selected in such a way that the solutios of th initial valus problems us kept suAcifttly small. A eTavaniunt Parameter I gives furom wh" go pprpritee agowte te t am be mum" "May.tn the appiatsU ft ti "metod 114 th aalY1le Of wetaMUtlairl termas o liseeder eda emelrod equa t s Fos t hr i purpose. sktart with __ the o h Dscauia e agm eth~ thei d nl v w ich thOe 2 1e,21a e l MtL kerstain cirwpuatjv eiJJujjgt~t dii-furential natio nvulS eigt !ou i hs m1anman thatthe sotem the &o!id veSTV f the a$, teral parameters, thicUSM, or principa radi of cMurvtue The absence of the derivatives in the Coefficiants Of th differential equations peTmtstho coleulatle of the oomelenht" a fta 'L pot without reard to the values at h shell parameters at preceding of following Panlts. t e.ki7 lestv o tK, Wmet As independeft%

variables.

the desried iSAets V, ".'~ ~ ~ ~ a ta uffle tdi twh-piUiont

-

derived YsrIAM Of A. symmetric shell with bir meridoal valratlowm (including discootinuitim) in yoiung~a mod"le Poinons satia, rAdi of cuvat&ure thickfk.and soefficans of thema openaiow Wkhl such a System Of equatione is derived in thi Paper onl SWt eOW verlio Of th daedcsal thoeoy 04 e1e1%, It coo be deried in the cMe way for all other onment loa banding theorigeso indludInt than which accunt for the dynamic sea trAnve shear d tim, sonhoinogessity.

and e , soýtr o.Finally, with the we d the method sand the equations Omivn I this papea, str-- and displamerutsanu calculated is a thin-Walled WMna subjet t IQ tns 11WAprenUM The soldtio shows that the meridional membrane Main, bs almost ideica to tha-predictod by membrane theory. but tha Wh bending srm eve for a melaUT*v thi toru may wotbe negligible.

Goomotry and iasc Equadsu The postim Of a point at a s&h o revolution is given by the toordinat" e #. r~ Measured aims g th tipletaofunit vactrorate, .a. repecivelY, u4 shown ft 1. shape th Wse "b de-by secifying the two Principal nod f cO Orte A*s otfm idd surfaceuaetiomof.

Inted at As. it is canvenlant to me the dstac r foes a point w Whe middle oin&ustothoa4zI

bSox&N 1oe tIIsfollws r fJda#(1)It the genrati s the middA e surfn given byr -s thOM X%. scshmeoeoile A-4

%+" +[ 1(- ) (2)The followinganalysis nquirs fnquent eis(e *)with repect t4. and it is eMawsient to CprM thi deivative by& WA rh od a ft e~o dr The diaplacanent bm4oNents at the middle vuAdce of the shell and the otatou of th normal we by te exprsiA of the displacemet tecto. U of the turi U -N + r4 4 + (W + A.* + m (4s)The shell is mubjected to the vechunc load vecto p. which Is mesuud " form per unit am of te middle surface and witten end tie momnt vector m, which is moeeured " nmoment per unit am &and gieon by.ve --ot -(1 -w)Kgea (96)(JOE)*1 -A +nWO- R +V OTS (100)(10.Are, ft -r (I -v)Oes Uu~dispiacsinant relatIons:

i M 4 -I -I4 + U (44) -J ,. * ',p (ord I i iiliwT di dli4 9'"'Q tiffi Pq~OS~IS41IUM

• LI 40M G ,Th temrpor÷ e da =2 eusd by th1r.ma beds is acaut for in the byar 6f te Integrated teepa4tUWe effect, of the foris-) -(1 + .(0)r.- ((, a ). -(t(.÷ .,)'rT (a)The dorivation it a neweam of equatdons esured out ID the sot*moto Is based 4% a lnu ust lsdemi teoy 4d shells Ome by Rsimew 1121. When Wmted In arbtrsy ehello a( revolution, the governing av's~ of equaton of 1131 a& be dmits In the followingtiorml Eqatkeonsitequillbzmz 4 # *w -N o+r (7)MW M e+ ' U4 +(Mo MI) a e d Q+ M40 (9b)-stfs4tusin relatlone so- Xes + P40) 0+ F)aK?. (6s)+0£e*.-C+r~E, (Sim (0j -*w soo 0 + W4n ()e)Ce (" 0 o# 024l)24# , 0 -,A so 0)w + A" 024 1 *1 --(4a+Dtea .4.(1-5)-1 1! I llelnl£ Io wlh vl e t S 0.0jU +on#. (e 134)T,, postive directions of Wte Mn o ultants in the tamping eqution an the sum as the oormpoding swm on the edge of the shell. The definitions of the sueis mwultants am found in 1131.The order of l bs sy m of equationi (6)o(13) is eight with m-oped to % and tonsequently it is posble to reduce (#(I13) to sigh firh-order astial equatios which involve eight a-..kniownL f the eight unknowns quantitie which enter Into the satut boundeaY eodition, at the edge # -am t. then the bounda-lh ie problem al aotaowiolly tom etris s, & .an he eDmpkty sae JA trm4 Of tMh ese nWOns. T ar th-.sason. th eight diffset equataonk derived. in the Iflwing ectlns sad th oight unknowns am allsd Mw fud-metal at of eqlutlos and the fundameta vuarishim ,mpectvely.

Darvasua93Tn au tal Sit of Equ39Atos Awaording to She "iadea theory of shells, the quantities which appear In the matUusi boundary oonditl=ona a ratationally sym.metri edge ofs aW shlf revolution include the ofective shea win sutat N end 0 endb iT an te ou gnallodfam , ý%N. sad Mo Th@s she fundeamena dgiv4xsls th ihe Wobem am toat wit thel brulb. derivation of l bs fudM ena eOf ios itd imoe on t- ,;I A A Aj a pollmis m sp. seM. qtsaOm M#, M..a Is iWINS Ot the t tmmaZ vsh.in. Ifkm (9S) it 1tW N. -.e 4. £-(uda*+ wJ+ ,,a')-dUCl- ,,)y. (11)I I&we -u a..-t" * , , 0.)sad fmm (10a) dat its,- #A* 0 w vie OU + AO dLAA-amC -Pr)T Cie)U]anoMnaao se.'. sad w.& rfta equatios (11.) hs& (am 3ep idft &hMe ias farm÷ j- 1 2 4 .++ a e 4+ Rd N 14j + m 17 L t 1+. "60 ... I- " " Is the dvatim of tbe fot equatlon of the tuadmintal an which livolve the derivatives of %be item mAlthata with derpen In by ZUL U. Of (14) bhe Iatlel.LmNIt itSo ,am(6s) sa (b)bymemsftIOaikwdt Nda.. r M N ...i -*5 0_ d Vf F&WIiiuYS sllminlago Ot Ot fift (7) MAd (S.) OIvel M(11).-- N.- + .. , + # so) --201 I+ No. + act + Me (22b--LD £l a-I + 9dl , ÷ 1 JA-L++ + 0 + O, M D .. -swe at# .8"4, It ) ..,., ,. ,+ 0+ ", +.-.dima -t LDS metL..1 1-a r-' 14u)Z-~ <,..- ,-

F. W'.i-z-a-f( --TD) (2I2+.( Iv+[OrDwe 4 -Cl 9+ am JILE + 1 + +J AD _ O eeQ-l-v Q Is. AL Gi C r 0 si 9 o o(W sad is (QO"lw trom (a1) ta" Whaereu ans.A , N,# sad Qj. won liminated wt the aun o The fundammntal est of equations iondstes e (lSM L whemt N, Us. ha be pilU ad in trm it theA fw stdl vLWauih by m~ams at (10-07L, and four additional equations involvicS the dedoivauvs of m, fA#. v4 j% with romp" toes. wbish aebtained from(13L (ll1 C1i),L (101), epuWtiU'y.

FIDsllyY thes Vat= it Sight dfMUMal equations WSa umma the dietervAtkLa of ea 1 nrvolutio us be exmmd is towm of t& eighsiA tal wmdahl6 send ni t as 0. +r 1110+U +ptid"d-(1-u) i~+, iw-( -) RlAWE + 0I + ,,.,> So l,-,, + ),( + -ID) .-+ , , -[,+, --,,...I 9..~(,_~~.r )... 01--. eve.aobt.ained kas form.act I4V9.QluIN+.M...i.J)

} (2110)+D(1-) +) .0+__ML.e+ D +600 K~ega aml{ }(2k.)The ..dependsnt coslDakata with subscripts a on the sighe.he"d side of1(2) anm Crovrs" by a system *I equatkon which ks Obtalaed from (2) a" after O"Si the mzamptla thaS the SWei Is "Ian'a be writsesam (~)p p Equations (22), (14), an (15) to (17) dete.-mie all unno wo variablem exept Qj which can be found from (Is) end writen in the form 1-so ..(CA)Sqebim -Ow* -p p Q-ij+Me+-

Me* + w4a (23)+(25b)Bly alculatinl 14... from (17) Lad mali- 1 useo ((16, dt I poi-ble to oenpiOse directly in terms #1 the fimdamental variables.

This expression is lengthy and contains derivatives with reepeat to*eelthe ebellparameters.

SizvosQedo.

not inter into any bound-ar ooditions on the edgie @- co-st. it bs prefirable to calculate Os" the lastunknown dirsctlyfrorn(23 Thederisivatrniif.

an be usily obtained by aumerical duifntlation.

The procedure for She derivation.o an *qWvaknt st at equa.ti.- for otbar How caseical theodu of Isotropic edWhu In identi-ca] to that giv en fre. Tor guwerl aninotro),c and/cr son-hmogeneous s of revolution with rotAtloaly symmetria properties, the fundamental atod equations is derived in she Sway as (22) ecept that (9) d (10) imust be replaced by the appropriat einrswetrain relations given. for etakple, by Am-.bar..umyan 114). OtherwIs.

the derivation straightforward.

For tdo iprond theory of shells reuch as the oe given by Nahd.(IS). in which the afeceef atranmeN.ehea deformation an accounted for, the following ten ftndamental vaiablu ean re-CS appear in (13), the ellinlna-tioa 4C Go from (64) (M) ("k) is done by means it (13a). The required equations flothW derva-slve of the genaralised faore. an obtained directly fromn t& Ave equations o equilibrium. (A) (7M (M The rse&mann A" eque-tions an derived by foWowing a procedure similar to that of tho foregoing.

tAgOdn4 4 #+ I 914 .(I -N (26 00 Ins .IF MML I+ B Af#% + a(I + Ore,. (2U~)D Mie S h IO o Q.,,. I- ' ((I + u)PD+ h as*(1 (3 + .)Kr ia' it" (I., ( +Wea +, +.)L,.,]u..

+ G(1 -wX, + ,), O.. -Q.. + UK..Y ~ ~NO + 2 .ý .'+/-.-ow' to'Fundam atW Equatacsu for Seprars Solutions For shels of revolution which of cplete atLtude circles. surfaft lod anre nerIo r: with rep 1to I with a-dK1- P(La dn 0 sia+ D~T. 2 A at 1w end th&v can be swuned to be of "TUS Form K.A ip..,,m.I

-(S")NOa1 -w 1(I + N)Jrn' as lJO in..W4 'a I~~a ~tl I an9 (2a)I , "I.4 -[Pl,. i f " oil (249)10. -(1 wheg the variables with subscripte a depend =ly_* a., and jach 'to615 deriat"al of the cum& of eQUIUMon ($)-(IS) the AsMUM.e.JntZreiv~p*lue of a in (24) "an begar.ded use rora eaneisi FW a" I" "h Sis Nmadiestl' ALLn so "haI + WIMP8t ponent Me~ 1. wben A caotew Wh misimem vrlaripe isdii. it eurwvAum.___________swim_

Thisears apem~a&UsadbIto ebesa the flwlag qiastge purarf Oloas hemn 023.A-7

-410 -SAW ' (:M/amCIu)r Jul. 34 .') +TI~~~~M" P"bh a(0 ouo oxtpE~f tW~o*f. -aDI- iX +P) ve.(1-')1--M'D vib ADnI apm #t b), e )lU ! E-,.o sk I to D L JI+ K +P)a* 0+2PM*0+Q0 U to gp 4.D- (Catlmwhc

+.an mt+included Q --- iMeun*uena variaobles ugn e exrese (2y eoneepd ste tparatio bofo vaimbas-TU double dXMI agia esmupd Wo the Uop of bottom triP.-ewuettia fAWCtlo Imployed Imn24). (25). #ad =3)The fne"aar paoer " sonearza with the Solution of the satam of equation (25. swblect. to the boundmar moditions ontwo edg. , ca It shouldbe oatd after sion of the load in Fourier riam. the solution to (26) ie obtained for sadk integral) value of a separately, and &ben the solutions anm superimposed to form.a Fourier set= expansion for the u~nkar Riductla ta to uWas alus PWabws Mhs ection, Is concerned with the reduction of a two-pInt ho~unduyvslu pmoblrm. p'mred by-A -,(&)y(a) + 40(s) (ofm)to a ai of aitialvalu.

problems.

le (29al v) i -an (m, 1)sa.uix which rerweents a unknown functions; s is the inde-pendent variable; A(s) damot. th (m, a) moemcknt matriz;aud(x)b ths(m 1)ma of the monbomog o trms TMe eleoments df AWs and Bit) a*m given pecemwiee mcatinmuog tune& biaibe toG a boundary gondltlom setated kn termas of lUnseal comabnastion.

of V(a) end NO() Is the form wbrm f..p e u -(m.m)m trics =d G s an (% 1) matrix, wich ane knovaIfrom the etatamant of the boundu eonditions of the problem It should be clear that the goverming artsem of equa-aos (26) drived In the preceding section in eted in form of (29a), md at the approprialta bmodar7 CoIDitioUS tO t aL"Of revolution sa be expressed In Qh form of (295).Let the eomplata solu6tcion (20a) be eoruse ua I[eA s, A.Q#) * (NIPW Me.. "eI MIS 27)INAA, Me.s Q) INse-. Mso" 1 0ft (7{a e )orhaza the sdeendnt coefficlents with subecriptena mast satiy a set dof quation obtained froem equataion (10-01) and (23) b theta..by, OR + C-a)C.d4+

.* ma)-m 0-0rome (29a)+t.r- sA4%, + 0 1+3,:#V#*A me. K- Mod.'D (284)Q0.m Q. W Ms.. (23f)V(z) -TCSzC +.Z(S)(30)wher the (., 1) mutrix C reprante me wublumy &ad Yjs) b man (m, a) mand X(s) an (ms, 1) mastrix whlchmu detnd as'.(s)-+ B(s)(31.)(311)(32a)(32b)TM WW conditions tot dtermix"g Fr(s) ad 4,(s) Lm Y(e) -I Z(a) -, whee I isths unit matrix.Evaluation d1 (30) atz -4 Wadsat amoc in view a( (32,a. b%, to C -(sa)vdmndb(30)&ts*m&csnbewntanenm y(b) -YIbrM(s) + Z~b)(33)Together with (290)X equation (33) conelituta a eymzsm of 29S Har almbraac equatiom from which the 3* unknmw. y()and VC), am detamind.

Once ~iK) is known, the solution at any value of, is butad from (30) provided that the Values of r(s) end X(s) as that partcula r an stond. Thie complpt. the rnduction ofa two-pont boundur-value problem by (29)to a + I tdsm-value pmblems giv by (31,32).As stuted In the introductio, the solution for sheila obtined by m4as of this procedure sot a eOmpl ete be Qo eccu.a. at mm ential lngt of the intvtarl.

The roem for this pha..beaaame acn be mn dearly from (33). Wbe the initil-vaius problht dSn- d by (11, 32) en solved with th e of theo equa., A-6

~vUwsuu wwrý1ZJ44 -0 mosdtilly of all elsmesnt of y(s) a" the plate S,-2.3,.. e U + I. On follwing i-matrix equations a(eob-Wooed from (38):-----4Z.IVZJ4-j+~

(37)i I&AX&MMu4noeoaaMM X1 Xi X3 3tu 0 fto. 2 Meigite hr dh4Ie.SO of M461 hI.,eI 1.5. t.g.Mu tions (26) for aahlb of revolution.

it is observed that e eUments of Y(s) &Ad Z(s) isersu in maeituda in Auch a way that if the length ia incroeasd by any factor a. then thass enltione increase in mapiUgd a pprXiMLtey 4XiOeniAlly With a.Consider, for the aiymmeatric case when the defor-matson in the shen is caused by some prmecibed e*d conditions as z.a.sn~Y~byM#Co)w3&zedN#(e) mQ(6)-&0 Ilhuuesonable to expect that the csrmpoodingaolutiom ats -i bec=ome ller and smaller when the interval (4, 6) is incrnesed In length. The connection between Y) end Y(4) i given by the matrix equation (33) with t following magnitudes of the elements:

yvb(-smafl.

Y(b).large, r(s).usity.

Clearly. the only way that the matix product of (33) can give small values of 0) is that a of dits w e nof (S)subtatot.

When the klng of the interval i increased, M() inzsse while ,it() doe , and invaribly eal accurac is lost a some critical length because all signilcant digit& of M() in (33) em lost. This simple example sa en iutration fr the los of accuracy encoU"t*d in %Me awayoia of eswe I the fregoing redudon, technique ie employed.__ A convenient length fato, defined by whoet ia1.l%...t. -Equations (37) involveM + Iunknown (K, 1) matrice: Y(xi. 1, X.... 0At +1 ENowever, it the quantities prescribed as the es of the hll " am the (undamestal wariables tbhe the total number of unknowne is reduced by a% be-craus a/2 elaemat of Y(z.) and utf7 elements of t(z~e) ame known. The same a true it the boundary conditioa etated is tum's of linar eoabinations of Whe fundamental variables in the form of (294). ik saa. I(,z) sad aI(x..) should be premulti.pied by wasningul:

04 a') trnaformation matrices & Lad Fe.4,.mrcpectlyv, so tha t he elements of the products contain the quantities prescrbed at each edge. After eliminatng g(z,) and I(xw.j from (37) by moans of thes products, it Is conduded"t.l (37) will swni its formnt. &f.utr integration sand, before sub.sustion into (3n Yj(z#) it poetuoltiplted by Ps*1. while Y7x(sm.) and Z(zg..) an premultiplied by 7At.. 1n the followmng, it will be ruegrded that this tranaformation is carrid out and that tKs) sad contain among their elements tiscos quantitie which at a -zx and * -xau. repeo-TIus for eal boundwt in the form of (29b). the eye.tam of At matrix equations (37) Involves exactly M times a va-knowns., and formally t n be Iolved by any method which in applicable so a lerge number of equations.

'Howmovr.

the success of the Procedure given in this paper life in the application of Gt,,mia elimination directly an the matrix equations (37).irst a earrangment of elements is performed.

Since those=/2 elements of g(s.) end y(sK.I) whi euw known through the boundary sonditons an be say af/2 of the s-.elments, it is secessary to mairang the rows of V(x,) and V(zu..) so that the known elements an maat*d from the unknown elerenta." It is asumed her th the Ist m x/ elemsnta of V(z, denoted by vdJ. arn kenow and that the laet a/d2 elements, dezoted by Wsd, eas unknown. 0n the other hand. ji,(aw.) ame she un-known and ,zir..,) msthe known elements of V(s.r.). Since the order of the variables in the column is arbitrary, It should be emphseid that thia sparation 4f eements doe not involve any muatctloou en the boundary eanditlon, Land that sany natural boundlay codition in the form of (2b) can be prescribed aS edge The sparation is achieved by a simple rarrange.ment of the eoluma at Y)'A) and the sows of ,(zm.a) and 4(zxx4&) efter integating the initIal-vAlue problems defined by (36) to the "d' of the segments A and So a"d multiplying by , -1 and Ps.& as stated the (orsgcing.

Once ik is etabliobod which parts af &6) and v(uir)d ) amr known. the econnuity onditioca (37) are rewritten as a patti.gaoud Mali= product o the (om 0 J D -11301 -PR-P~whe.I in the length of t& meridian of the Obl and Albo Irn-mum radiua of nvatusme, sea be eed. for an appro.mnate estl-mat o th citiallenthof heshel.If s olutions (s) end Z(s) are obtained with a six-digit aacurscy, then the foregoing procedure giv tod results in thb ang O 3 -L However, the lose ofcuracy o the solutitos be avoided and shes of revolution with much lmrgre vah Oef c ran be analyse by osoe of the direct integration Ischniqus i th m6 ultUleat method gIVeA n lthe eSt sWains III SIMPICYOd Mustiastgmt Meth odf Itapztiaa Let the shell be divided into M4@sgmetx (denoted by 8d. where i- 1, 2..., M)at abitrary lngth In .each of wkich i a3.Denoe the of tsh ends of the egments by a -.A, where the left-band edge of the shell in at a -A~ and the right-hand ge isat -.,,uasshownsIn.ft2.

Inanalogyto(30).

the aolutiodnia the total Interval I s , s.. win be writtn 48-rr--- -- JLq.& -jS. .(38)Ax NXsleOs. + 1A4)(3u) so that, each of the equatione (37) turns juto a pair of equations, given by where r.(s) and Z,() denote t Matrices to V(s)and ZIs) in each atzment St(z, S a S and a'sm gven by Yxz(&.,))s,)

+ Y,'a,.) -g(&..) -I¶Oi Y, 5 (s..dpdzJ

+ F 4'(si)VO-s)

-ba(c-.) -Zz-!Us A(S)r'4s)daq Yrfj) W I The result is a simultawn orus ytm of 2M linar Matrix equa-tkons In which the known oefficitnts T.'(i"a) and Z.#(&..,)

are (366). (*&A *A/) and (me%. 1) matrices, respectively, sand the urn-knownavA)rJas (etAI) matices Sincam J n ~sa. r (35c) knwn hare amexactly unAr wtL s.M rt(s Wilk' .213-..IZ48)~s + 645)A-9 Nj' SeaM of Caau~aa elimlasalse., the system etequatioee (39)e hiuabeughttstheloem where IW dot, Indicate b -,4.... M-- I. Ty I& -I a 0 0 0a X A&*o C 0oC, 0 -0 0 ra L 0 -N,) a, a 0 A -g a 0 fins) A S m I 1 0 iC:) As the OqUtUo. (39) Vt 0 on the sha Such Iodr introduce discontisuitie in the sole-1e (mu. m/2) matrices i. C, a defined tion for the orrepoanding sire resulants, sad they can be repro: amted at e&-7 a by an (i. 1) discontinuity mitrix which is A a -it (41&) simpyaddedtothematrxZ.(z.j}ontherighthandsidsef(3n).

c, (411) Thb fature hot greWat vale if shell aw odefed. Azy discntinuiy.

sther ingometry or in loade. Is as~ily handled by that he end point of a segment coincides with the lact-4-, + YrC.- (41) tion of the discntinuity.

Since integration is restarted at theof each oeent. th p eect of the discontiuity is (M,' + Y .'-,)Ud-'

(41d) obtained.

The progrum outputs -&l fundamental varibls"W at a A. 8, are gven by number o( dared points within sch segment, and it also com.put" the value of V,() twice; onc from (43) and than from d(42') (33). f a osel in aumber of egnicant figyure of these vunes cat. -mac than the continuity conditions anr known So be eati to theee tm offinberef In, thsway, a convenient snoatim maw afte soluton is o taie o every ease.and fri -%3,.., A'4'ft C, f 7U (mA J)mstriome A, --Z, as --ZS and for, -2.3&.. .A, --l T-~~~5~(429)ad --z,6 -ri,,-B -(r,4' + YsV"-')&i'A4 OM2d Fnanity, far the 11th sagrneat Aar --4'i -.r S -g(gx.) -Sl- YulartASC.ad5.a WO(at + rzeCM 4',)8,M'Adr (42n)For bitvity. in placsh Tbt x.4) end 5,i('). the symbols Y.0 ad Z.0 have bean used.By imam eo(41)enad(42}.

the siknowset(N})aerebteindd by Cx3MPI: PrsUvilzad Torus In this section the strtsse and dplacamenta san determined is a complete torus subjected to a condtant internal prinure. It ij well known that the solution of this problem, when obtained Ly moans of the linear membrane thery of shells, bees a discoatinsuzt in .O displacement ULd. It has ben shown by Jordan 111) nd by Sanduenr ad Iepine 117t tha a satisfactory solution with p&-gard to thesplaement fMd for a sfficientdy thin shell can be obtsaned i the membrane theory of shels is employed.Subsequently, Raisener 1131 etablished bounds on ertain Parameters which show when the wonlinar membrane and when the linr bendin theory I applicable.

It eseme to give hem the oltion for a prasuised tona a prdkited by the Tb geomr7 Of the torus is shownI in ft 3. With regud to the quentibtie Mpkyd in equation (2), the two necsary Parameters for a to ean given as TOO)-Co-a yis)- le-r"Itsdros)

+ Ae,)and (or i- l.2.. .,M -I 11srd)- C,.-'(p 1 SMe60) + 5asr-d1 (439)Vics,.4.)

-Zvau.a.,)+

AWAti (4Md 4 ao&(44e)It should be noted that (41)-(43) miut be In succemio.because each equation involves the resaut obtained by the pmeed-in equation Once all the uknowas &J am found the fundamental variabla am determined from (3) at any vale of sat which she solutios Y,(x) Lad Z5(s) en stored during the integration of the initil.value problems of (U8) The integration of (38) can be accomplshed by mans of any of the etandard direct ittegcmtonOn the basis of the rstem o (26) given in an sarler sectio and the method of eolution developed in the lat two ao-tions, thd author bee pMpUred a computer program" which bae been applied to many shell ronfigurstlon havaing l value of 0 and sueccsfully toted sgpinet known resulht. One ,ample of a pressurized torus with 0 -57 is presnted in the neot section.The proram admits rbitary meoddiozal variations, inudia g diacontinu:'iak in all ll parameters.

It OWs admita ring lds i inth ormof precribed valume of Ne, #.M. or Q stan vulus ot-T~s progrem was wernn e all ealculatioas were esaned eut by ase em the hIB 709 em, puter at th Yae Computer Censor. The diret integration 13 is petomed br sem of teM Adam s vmdlktov-mwcowr shed. whish selects as optikmsas oW"e at ever tp seer" t a presubred seomay.r M a + a B (440)Because of symmetry with respect the plane XX. Fig. 3. the x 0 me. a O.eeey e ~ue eeealduwd am eseeple A -1,:

leble I Swesemsad dkoewma o at Vpeweu~qd twi. 01&a g- f U. ~- i.s0 1.001 -0.003 -0.031 -0.016 1.249 1.234 1.2M 10 1.213 -0.IM -0.003 -0,01. 1.201 1.313 1.312 120 1.504 -0.13" -0.123 -0.030 1.329 1.1=7 1.417 144 13.20 -1.014 --0.20 -0.0=0 1.21 1.,07 1.6n in1 1.53 ILr -4,505 3"73 0.90 2.8= LAWiS 2.1h30 171 LNG00 1.002 0-168 --0.30 3.467 -3.0 ISO L. 3,3 "3-.717- 4 -.482 0 .07 4.334 1 4? .M14 1S4.3 2.042 3,.0W 3.03,5 l.SrA 4.150 4.376 3.248 ISO 2.104f 4.770 3.119 I.3":0 4.208 -,..637 3.143 10M.S 2.17S 4.1T78 2. W40 .="i 4.160 4.3W0 4.t1WM 10,5 2.2,54 3.610 1 .360 -0.274 3.9M8 4.221 4.102 210 -^-&3 0.7 0.3,7 0.0V I.L6n 2.S2 2.431=34 3.168 --1.t43 --0.201 .-0. Ow i.200 I .2 31.m3 -0.717 -0k --0." 0.416 0.411 0.414 270 3 9?7 ,-0824 -4.331 -0.481 '0.103 0.101 0.100 I h/bsO.b 00 002 3 2* 4 Is e ed e U0"to* 27", ndthebund~yitlmaexthe**

-,dpon 11 0.05 0.02 0.005 Z89 1502 A--.4i arem.D~Q 0.Fo te 1U9OeO1maereo

~ (e.I) 2' .053 2.082 2.042 rult1e1l61 anud 171, theloe, dpaaametm~ke~hosmee$l/E (e~.e/5) x i@' 0.427; 0.$22 0.197-0.002aa e/t lb 1.,5. 100 (tde,,) 20.8 14.0 9.4 The muuartcisalwuus

.1 the aeruzil diapaesnn~t.

meridional membrne, euwi it. -11,/A, cad meridional bending s,.ns Vol- IWO' N 5 -PJ2 foe a pWorm0 ed torus er shown in*Tabl I and an Tags. Oe 3.l Thee result wer take (rom the It Is of eigzutano to set that eve for the thi*kuse rciio output of ths cornputwr propram prepared for en erbtnaJ shl o1 8,/3 -@0.00, which tor many esppliatione would be ,eprde ae iegol non afte precin the eometc hr u gin by teha1. ee e.-wam bedg et I aout 10 ipsrnet of the (44). th d m ba atre dtrbtint agree ve'y memb2W.e adhbun at the dcme poast. Sute aenedt of be.ding3 .a .welwt that obtaind n (1?) by measa o the mminabe theory, to, us wer pvoul Rated by asek 1191, an they are else in of ehefl e-nd it ehoe ony csnail variation with &/. The de- epeemn~t with the etatmot made by Gokidevie 120] that aormd eh0#Wp Q = the %rme pectuon .v the toni ehowa I Fag. 8 for when the 1iddle eur"cc m a closed-ple ee, whc Inci threevlul of 11 M 117qualtative eoad entwiththose gven " m o4uoa -1.03 2thn 8th 2v.cnit thiseee in U1M1 and &iTA hut thei quantitative agrement annot be a- bendin suei sho"ld be eLpLcte.d ed the memh .a. theory s peTd becaue the values of %/b need a this c et not rothea where the bending ec ar eIbl This Is con TheA bounda lae show I. ag. 4 r -lAo fo apreement wath Tabrmed bI tLa ein nr o of the bendn strumes bmin from. 4 the es ofso gmached mm 1181 to thte eve tha thicness rt The maximum vaueof at4, 189'.forkI, -0.05 nd givec by otpt o 184.80 for m /t -0.005P which ar aleb the apoAta wf (4a16mTm normidal diracmbrwgen s a diswvtributioen a Jg. S. ver mem(12tem tesamo- t aSuchegactotanin Thev omvalueo At the mebuane asnd t he m thiosm e bein U aThemtvioum valueY of too/bme " ninT bhL p -AA -- wSpa/nd g,-n)b A- 1 le ant k46 emparend to Valt7. *0 a £bouandeay hISM hk 1% Gish-borbeod of # Me 180 eMW be 6111dpe1`

For the Proeas-- ample. p ramp bbm44tO4@ "fD Wpf*Ina£S0 ic Ho~wevr osaca V is Ut. eal had panwwe df the peoblem. te Solutions sbmaws Fin MC 4 and & am pcsportloatu to p. and tde botodary layer rozmnaii ansf~ctad if p plea. s bvaried. Of enwue for veq Laiga vaiuuoflV tU deformatlon

.1 the twot way a-end the Iianita eta& Uooar theoy which acooa~iat to LIS) re"fc p to tOe rana~p 0 CPI TW@ meauch baa. been. supported by- the N&Uoaa2 &Uses originated from. %Ma oneaaliha work perfoime by the author for Owe United Tocknoloa Cauter. Sunnyvale.

CaliornWa.

The author wiahaw to thank the M9a of the A~ppie seabaft, Do-panonout of UTC hor many Wuminasdag diecumlom ooncerning this e'jbjtc UR~urelae by Fiche OUNNIOd MAWhd. Jeournl of Maewoe SnOwastio'i

&ioke. veL L 1061.g~ w 047 7 P.MF. RAdhuueL A. M. Davie, and U. ft. bldode "tNames"n AA9Mg~ of Cquadvas at Tha Sbel of Rgvela11e." Ameri-Aakdd~oirflereat." IOU3.162 pM.3641.S W.3L usaeiee Q&i.3Porboa.

1. W. Dincoo~l ed A. W.MUMna -A Olaise Ceanput Paaeum tea IMe Clene AaWIar SY-m meetS. Thia4het Psoblow.-

JouwoAs Of APPURS 34ecUAnCe VOL.21, Tauui. ASM&~ voL 8SC Soral. IL14OM pp. W3-1401.1 S. Budianukr ead F. T. RadkowW~. "MNurerlea Aalyaiai ft V~amm*Wuiea Bondlgi" Sfhalls of Revolution" AIAA Ju*-v&L. vol.I.M96. pp. 51-82 10 J. L. Goldberg and J. L. Socanaol. "Stalls eand Orn-is Analysise at 114sunaiferi Can"ea Shells under 8Y~matatcal end Us-olWmaet*Wa C&o"Uo. Preuadi. of tho .siat sympeosim so Re~tgis& jkttest and Aoe'svpwco ftoioelow.

Academia PIo'N. ow Verb. W. V.. Wo. S. MI~. po. )1-628.ll C. ft. simle. "lib"ofl Rettvolution Wthk WOU Ledso atltpi CIkeustnirnulal VtAMAWe. Jouawaz er AppU" 34SCRAM1M, vOL.29. Tux&s £8313 wat. K4 Isail IL 1902. pp.702-707.

12 R. D. labile "Asyunp11 1 Solutioen of Nonshola-Mwi Solte Revolution Subleaetd So Xeosmmtaule load." J.urel q/ IAs Asne.ape Sclooms.." Stl. 91942. Mp Im137f-t.13 L. flimsoer. "A New Dowivulif of the Equation(at toShe 14 L A. AsaSIzM~MYoa. "Theso

• r-imoUopi Shells" 69". Momoow. VUSS 19411. V. St.is PT .36egbd1ML "001 the Ihene of Thin Molts. Sbal." QUePUrV ofAPPli~ditionia v*l. 14.1ISO. pm 36"M80 16 P. T. jordso. Museses, and fotrmeiouse of the ThIn-Welled Pimuaalsed Tefarm"." weel 4f As AwWPec Jtdwm6a VOL. 29. 1062.pp. Ila-62S 17 3. U. Seadeuo Jr, &&4 A. Liepbma "Tosoidal 1Zambusne Under latami ynm j."AojIAAr.a .e wneavl. 1.13. pp 210-4110L 1S L. Retwer. 'On Stramee a"d Detawmatift, In Toe"ada Shlle ot Obuler Ctom Salim WMAo Ane Aotd Upon by Uniom Normal?meaou%." qabwiwtw q4 A4pWe AMamotinoe

.vol. St. 1I63 pp. 11?-19 3.L Alrc .Oathe 1 TheWof ti IWOeetla ToColdel ShAIIe.v Aledre it MOAM...it Ow ?Aaitea voL 2. Ism0 wp.14W170 20 A. L. GAeoIAVIA.

T&WI 4f Iletaof Mie &Wd.. ?ergaui PaW. ew Tak. E. 16.M p.4W0 I P. M. Ne"bd end C. N. Deaumo Dftaametoaoef Elail El~ipsojdal SMell of .ovoluti~oe

?weawdiwW M 445 ,a- -V. X Xasi~enw Cooneme *f A pptWe M~aeul. IN4. pm 833-3,11 2 W. Loeanon. -asitagme outegmaw~soa dot Rainer-Momoor-.cM. Schalenelicua ftr DOWANTo eater bo="atMU zanertraruk tngsco~rAft~3.

VOL C ML fV. 339-364L0' 3 I. 3S~aM. -'Eln l0"teawoavsiteohro ftO die Beehouenag der Sicvapenaunspr eceiomatie e ks woeanyw achawneym-aetrieche S>letwtu&L" 1vgai'uf-Are.

vol. 19. 2961. pp 103-117.4 r. Daghell. "ur Thoorio der Romelacachaiea van SteaL.punkg num.het Rah nugaJdniwa~th vo 7, 1919. pp.60.9D.0.3.117.

W. T., Kner. and C. IL Matter. *Nwhaeuhel Usthods end %he bsadiat Gf IllipokWe SMaOW. .Ime',a U 30itaadg Vjnda~~~albt 12n1 1M.~a 4uAB4eO.v-63t. Pennyca. "lymssu eadn ofd G r Sbell of'>5 Reprinted froiziSep1tambe 196 isme of th~e Jotanal1 of Applied Xaomnics.

I I___________

V ..I )ri oI P Forcal ibfatlou f bllinflua sizctric ba;:a Rob A. IKAUaNS'.~ , s el ~~a h It, " to cat a a* MAbW of a'wysui d utic sldll gem if Iru " ea.t ppemtis W e i l who sh- shel wal of any umbev of ,.1 ot W mak eJwnitv',k mtiaL lb *OWSn ehi Sin& be m:L-A tt etmo e lG -VA pn&IprtkS are VMbwit ti l. atm,-sul- .hi-. in turn, affcct wily the M-dkkenits oe ib 1 jM r-order fmINvntal 4iffwrntkn rtejtians dwbmvu for the mubtlonn of attaibntialt-Ivalt prulnAA .The SW~mnle~nst methd OMe ba *4hVe iwUIMta n eu threb siaist d erI or &adts) eatitou a s$stalk &eforuautitn muins emscl* the mod 9 On deitvtiva ot "bs ekwientAi deivd ib ithi lKte, an e4plcyl ka *A iilal-at bv wtcgimttimn.

In the pIeset irmnabtik

& AAl b deryd by aaI"n s( mny mva"Mlt, wistmo*y guyMettr..

seetinuma rftfifrtm.

vAud f& Aqamreu boanding mas a thd se "ayea of t Ohn 1an suoted with alope t tbs uerfemw wfacs IR any axblreQ inna. Uhi ktown Ut uutficlay umful Iat sactL-a qqpkiwkub beemn dw uftienn saurtabe an be cest a the simplest oawso liba shllt and aw Iots a .t&td to the "mWidle""adac w toa speei am *Ed bI det euined from the dustie betc IMt d t sItalNk dutotas a the do x, ar-mahlants sad te etlia etssurm rm a lyayent arthoetwlt 9.%t!!an o t from Anbsritmfl.*

and. after ad4ing the tern.erstmt tems, the ma be wuitten b re km"' +Wgnnc, IIJMACCgI*~~~~g~~~~tT ofn~. aawa esf~ lh. ar bbw"" "~Isi-a' (va -4")ro 8,, (t-)I ("6.' -S).'a)' ,'F l(,~t..,'

-,,* ..(2,00 -1,,)I I 56 -ZOAt -vti%)W~CS. .+ Cat.+ rug#. + .ro.+HUS+A

+ Et.". + 8.7.e+ Me + +)*+ Ute+.,, *g. + + ajr..-r.r.

-V.. -,. bw + Di&. + 4.4.- + Xea.+ slfe. + Far..MNt O * + D *.+ k0/0 + af,...+,..,)

whers tno SA.AS a im."" the .th Fo ,,iE b.ir,,cie.

se, she &,dI"ase s aaluw rte of tbs taeerevia a'ot IA" shd lbo eletue psautmt V,. 1,,.. is OMuI.g Itk (1) aftl (2) an AkEr..d lot a stall #Nwi.tlng?IA. ft1t. hI &l , form IZ" Titetalem.

Depsaftowues al Mweknls. LshI#ftraity. se~be.l.Veedatut

.Ue~'n rom 1Cilletrw APPWa rieawe Yalek lvt~ mlia" as Im A I L .&kW&J 'Amnb.&6 of Atweta of Ueutunai 1kIwtot 4n" GO("ta &Md WVoasntIRAQW01 Lft&" JOCMAL WT AMPICts UaIowud eel. 21.d ISiW#so JAbtrei %el. 3ba .1144. pp.Usmsý Moilead by ASIML Appihi Moletiled Di Awb aORK& 4. 1143'. "IIIM Sat.lay U. MIS, 5.6 uui.I -vft.) -v*MsI --aMI -'s Va Idex i dmoato the ith hysbruiuded the WwenoIIltm and , fit. I; b 4f"a,0.rml,,anthe acfi,,nts ottfI+en epatision and modi ib tW # and 0 dlUctk4, OW ively; # f ar W corupadleg Paissoa's ntif; SAd Go k ohms mbodulus.

Tor on ibotmo*l kyn of this shOW&VOs a-ý ý4 w # w;p or,@mw f P nd Gw -.E f ~+ P)tewpraturs twuhata To. savd IN. are ddied by (3d1) ,wmo ?gj#) "nd ?s.(*) arm Prwled bempw to"e dun 1kw of t aA nih anuwak Si the %pw (a -A.") end IV (a a a,) sudifaro Ow lsUAbe, wsediiely The oymbos %m.(at AWW..#,. u~deipsat SWe amh barsonscs of the ntuenaewf strins and %tesDng strans, .aeqedlivey.

and th e Sprai (2 eaftn &has suae Msaftu ft 11w diaplaftmait sMOpMo sheW t referencesrfm mat s be found in %Me prerilsa papt.'.C onsvanvRM we Urs defined 61t s, I toald be veaus that 6 qumantint de~las VMAIAes I~WE M prt i s w w Whe t *$do"' bmaws (so ad *ft Ow 0e.~aa S.p~atmes awawluan.~

ee:eftwi.1 eerq~sM 4 IN.. gimeat.e aa auft an ww deami te s- WVr With ft~m% ..nigi 1 aeava" status. ?T. gwwjmeuhunes 06" tw OWl of wh Mnfa SsrAt suffvine .ut denote byWO a.w 1O A. Aas~tusuu"Pjos.

?&&e #I Ad' k5~.u iA.U W?, s) a' Tft*tinu TItuiebiadm

.11*.WAiretw-D. C.. Way. I10% or Wi. VA 44.Aait k,,usalf I t,'tld .p., 44.vube in s t tall her A adShlty mm ermia6,ed e'rpa.hou lee"d7 To. ,"A'ls Iithe lemneutu','

1' a T,* +4 ST&.' ki s8ewnl the elk -la-e. &*a te taetslta am dserIsd 1 r aeqVIIAt 0 h .1 ofnI"biaten "d IM out of b*., atft ,, bmam11 .taml of~ b) M A-8?

It drui be ws oalmd G jat , eo .l as utalt8seu (mis di-neaethmm) 41 bet th 1W awl ,,,tawut ,k% f.-q4 % 0h dmatr@MIWkl iWeelWAt &. CaOW-fstuy, the ihik.s"h.

eml a te owk mpMto.ri dqtndeeist ha adswomwkr.

bt " " ,omu- drk, o ,.* 0Ia Ithe tupth of tip balkl-vh w dt4ned prt I t It m.vem9 ts Wefleut at 9 gi's p' t Ow valom of in urivatha of Ga.e' I metskuelA iwbiat fe Ie.o ,.. p... Os. U , "M:1 I thi l vanmlhe ammd.knows- tlUbg (l c,'.d (a) m Sl rsir, ,.,.equdibriam eq"atloen (isiti mir .hl. turn* W) fam e the Pmv"a tram .@ Ome mca k tho e of the &A-r, m's be amaz'g'd I the %flwing eado: v.b + V4. an 4 + W .z " #It?k sok*ub or edonhooh.

4t 6i rapboftW ,. --W owu wit, I. w"llm, 4t ml tw Onoi mi,' ow, SAl "I* ,a. LI%..... jet.all -Ak s'h, .v i"6 1 i u ,mswi hp of P, Rd tho tosdwnw ,whbimb (em sbutw k emYMih., Iatus iit l (2i s i'mdoklkg witi (ft Ztesthes (a) mna I* tw dk asn4w~ j FORM., ew amts k ampl pum umsme rmwy 9meatlty ewwulu~ hue o qu&W s.he hr. &4kd by a Fr#w 1 ThaftftLh(N) ansrasW&k 6e She Aaajiehof tie sody.StaeW Rmv.~o of &a aubitriri mailiely epaftzwej&

Ad?msOOReIl1Y euuiAWla rmiaee vil and/e hw ma 16 To End the srwi, defeuamaIu, we shj apyit Is (S) a I'rt WNshofe(,,qlatgie, 416ftIlug slauut Its exis ot 4imuelacy with*gl~iLf weelyf 0. eAD ane Inedg W w an sot eqI 1l te.ut w for Ift-vUfearm proma.m ant beds one shes~t (.a vs. w to. -T. a 16 &W* e )e rudaih " eo pftvm wlye 169 oetshiatl st the eAtwwa fmqtwneiw sad M&d alsnp" of &by kymnd SheW at r.VOhutkw We~ smasas hu bme "upwted by Owe K&Ath &mI de Crtas ICAi. !Zfl I 4m. -mJ + 0,. as of (--)(al)46 -b% ON0/4 a+OVO.jAfRo

+ (Omt AXik -2 im $ -.(39)41. -(0/lJ1(K*-

-- Nor.I -RANT -emo -£CU4.)Du-(Mo. -sNue. -ferm- irN, -DW-4.Kl (a)* 8g. X MA#Ltj-.-

s -to' -sM. -- bua3C-(P*. -NJ 0.-SIuts -Cti. -ro'.o-el (a)X& ' C04. + *'n& + 4, J'e + + liOT6--Dog. C. + Ceta. + £wej. 4 slo.'e+ iro...C. ++eT,., (+-* +IV -$/r + Ke 0/41 a ~ ~ ~~W +u,, +s minr. +/T (an~L'e+R Mw. aet. + '.) + d(. 4 Va.. aa (N)I a + m(l/to +d Se/F)*Ja + IV*. anA/?SV,- an -'./-p-OW. (so)I ff rm ief if,, #It* +l 0h lk .Tl clsii N...-mok COO t/0 IN+Md#+ sslitoeh -P. -Wýu. (36)wher OIhe dee- gin~~' D ha#. heft. d.Ine amo Is erm dof he paftx drlt it w~ e ,Th bys.Tudikae 0o"he symbols ftna b6 Swind ebcwhmeo i A-14

.4 ski* dC~s..Appendix B Containm~ent Vessel Drywall Configuration Ohl and Stress Summariea from Original Design Report Oyster Creek Nuclear Plant I

qn 4 .1"'-V ..... l cWcAGoIRIDG&

I~ *covuir uI&~~'V.'GR WIVIMLLE EVINHEERING DEPT.0.111'-*z-- E. rv'I I-Ev " so'-I t 2-0 ('C vj7lTr (~L___-_t -(.. -SiAI, U OMCAGO BDem MM IRN WANY GREENVLLE EI4WHEERING 00 L&~r YA&. L Am SA1a ~ --MIU.%. Q E.i& & ~ -£ &4 **--. --"a's. týLz--v-.f 0$Also by OACACGO SEDGE ION CWaANY GRWl"LLE &QMEEraxc otP CAJWý2) eu A Qpe utfb-c1 4 7 Z 144, .T4 IT !u CPUIL0 4) L~j~nt.r CON-r,&4.JV 44fQ --Co e. A 0/,. 4 A. mj v' C C"k,&" ATI.:. z P0M ...-,' -,. Cm..2L,* .L .L~Lyt. Li .2` .0%Q.IA.t CINCAGO BRIDGE & ION COMPANY REImLLE ENGINEERING DPt.Q --.L=,ca FQ~.V,, C MA,,t,.IA I, U o..r bA< .-..",) L uoea. A.-'E'idA m.Gý le .r tr,, .A- %-I,, Q, u A ,w,-ims 0 ..~~,c, _ 4 ,_'0 -rl LU 5 ftI-. 10-4 5 I-oQ C )TV' 464%....*D V 9..ops~h ~uI i 6, 2-*--I ~-~1w*(~N. -...~ r-T-I %S'YeU.ist.

j~7.I4mO.Ui.

1107. V&..~.i C -. --.-'-- I -'~r~q.blab L0~I wbi.-IfrV4 S" Lrh'm."T wa~^",~rk~YU ~ I ~ t..t&W-no I f-a's IT-TV- I a- Im, : 0 a 16 3Z S'lls :F dA Z S.-'81 4.ii .o, i;)L4RO Lji4_LIZO L Q N-O Zo 42o1 4zo-Zqio j Io15 " "za 8Ib" 11001 Z.4001 55 Z; z3q 1101 14 siý I BSI :; VIO 1- M + ZS. -91,01*clio

19171,, -1 1,-i .i~w 4. -I 444444 -I ----If*. %'*1 4 8i%4 1 4 1 091t (91 4. 144 4 4*4 I -M-1be , Is Z- -.4 1~1-'C-a a.0 0 U-I I-Z~l ac i "i 1,e4~ .qo0 -C 3I "15SF S 1 4. 6-4-4-4w

.~ -4. -I -w--q 8a S1 955 4 9591468'41agl 8I I ~4 4 4~4-4-I

-*" Adt -101t -4 S84fS8~4 1*%e'C)&8bI%

1: Nl-..-* ---a -S -S -* -S -* a ~u ~0)

I ...... 1 ..... W;A ,- , 1 b,t,,4, ,- -M _ "--- .T, A. ... .." A ;.q-9 .-4: ... .--_ ---I -! -_ -'-- 'lms- I -- !- ------ t -I_- /-4-I, .L, SI -' ---W=. F r. F17 3 9--(6-5 _ l -4z 4 0 0 a I -a -W--II.O * -s.---- --- -)

i SU 4 , -""- -t1 , --- .._ ._ _I.a I-- , ,- o. ,, -.,,- .,o -, --,i -,_. 1 -._ _ .-._A' ,,-',- -~ -, -" -, --1; a0...C 4 ..,.... .. .-._.. .- " : ---... -"- -' -, "1": 3. ... ,:._ L 0 :, -ZL---,'~ m!,=-- -.J ...B-.'.. 8.... .L N :F T C LI~L ZS '93. z8 9~ -.,: 'eit I ..I 0 0 3 0 0 i I I/

(lc.,,.; j v~... Loa~b~bh VeQ~ aaU leq-'1 lll 11 --.T C- 4 4 %k A 10 t---'4b , 1/2-.,,'I ,)- "- ----4S 4 V. 4.q 4 4-t9 if.- as-I----'- -' --" .1'4.1 4.1S %7t .'j.,_ -- -- ,, -.- .. -I@..e *C i I ,I/

0 I.Appendix C CEI Kalnins Computer Frogram Printout of Stresses in Embedment Zone 0 I

• T *..* -"-1* 0 *.1 0-.. ~ 4 -* ..~ .4-.4 .4 .4 4.4 -.4.4-4.a.4~ .4 -4*~~.L. ... 4.1. .-4 -.4 -. -.4 .-* 4 -. --4~-4 4 4* S --* 4.-4.-4... *45.. ** *7..4..4 O 444 ~ -.4.; .-4-.~*4. 4~4* -* 4 .4..-..4.

• 4 4 4...4.-4.-.4..~4~4~~* -4.. 4.4.-, .4-.~...4 .4.. 4* .4 -4 *.4 --4*4.4 .. .4 ~*4 .~.*..4,.-.4..4.-1 .~ .-.*4.4-' I *~. * -.4. .4 4 4..4 4-. *4 ---.e--.4 ..~... -.4.1.4 -. --. 4~ S~~44 ;..4 -.- 4..4 ...4 --.4..4---.4 -*4~.4 .4 44 -.4.4-.......4.-. -4.-4.4 ..-4 .4 .-I h~g I ~AAe~NG.,!O ,I MV JL .ca "v In o.IVA A p~, ~©r' ..

• 9**.. .4-..~... .9 a.-.4e.9* ---£4 .~ --~.9. .-..4 -.4. -* .1.9. -4..* 4 a -.4.. -.9 ~4'. -4 4 -94.*. 4...4* --41/4.4.o o ..44 .1-4 4...4 .--..4 -4 ..4..~ ...9.4..9-.4- .. 4.4. -4..4 4.-.4. ---4.4. -4 .-4.. 4 4.*.4* i*4.....:y- 4.-I.4 ...- 4*.e : -~.4 4 4.eq-4.-~ .44.4.4..7, ~4.*~ .4.9- 4**~-- :~4 -. -... 4.4 4..4. 4.'. *-4 *44* 4 .4.a..~4 4*9-4.4 -. -4 .-..4.[4 -. 4-p.4.1~. 4--.4-44.4.-..9*.* .4 9 4.. -9 9 .-S li~~s.&:Au STf l.I

, .73°. .0 ~0 0 -* -,p-0 ..0 .*-a -.0 -b-* a-a.~0 10.* ..a.40 a-* a.0 -.3.....a..0~*.-b 0*~U.-. a.4.. a.0.0*i Mags *3 anaw SW F ",s%I I I I a/ 411 PAVU SIcf GOP m I ItEil I

.v.. -* .* * .4-4. ~.-4. .. --* 4. .4 -J. ....-* .4 ..-..* 4 -..s.. .. -.* *..

• 4 * .4. ..4 .. -.9... 44-4* 4..* 4.4.. 4 4 -4.-

• 4.4. .4 -444. -.4 4 --. 4... a.4... -* .4* .. .. --..4 -4 .4 .- ..4.. 4 .4.. 4 ..4 --.-4 .4 1~ ** .. 4*.* *4..* ** .--.4 .4. .-..-.4 .- .. .4* r ...a-- .- -. *~4~ .4 ...* ..4 .a* .4 ...4 .-* .* ....**.4 .-** ... .4-.-.4. .4 -.. 4..44. 4 .* .. -- 4 .-.4... 4 .4 4**4~~ ~*4". .* ...-.4** .4.~ .4**. 4 ..4.4 .4 *::.- ...&4 I* -4 4 .-* ..4 4 4. -.9

• 4.~ .. ~4 .. 4 -.4 4 .. ~..* ..4. ....4- 4 I..... SY 1 l -~ ~&-Avg a t.I 5 4 ... -. .-O.. .4 -...4 --k4.-4.4;..-. 4.4 .- .4 4 4B"r/ ** 4 .1* .. .4* -.-4-.4 .4* 4 .4.-4 .4......4.4..4.4 B.4.4. ..4..4 .4 -.4 S *~ ;* p ...4* .4 .4.4.. .4..5 .4.4...IS.v.4.~.4.... ...4** .4* *..£4.. .4* --.4.4..4 -.4. -...4 .4. 4.4.4,.4. 4-. ..~4.4. -4.4. ..4 ..4* * :-..4. 4.4 --4.4. .4-.4 -.4 .4.4....4. -* ...4..4 ..4.44-......4 1.. -* .4 *1.4. -.4.* 4-. .4-.4* .4.4* **-.4 .4.*9.5.4.4 ..4 4* S .* ...4 -..4..4. --4 .4~ ..4 .4* S -.4.. -.44* 4.4.*444~ 4.* 4.4 .4..4 --.4,. -I -.4.* -.4. 4 *5.4. -4 m I I n I Ilse N U U lA1 1L.-7 I .*, ° ..I ."'-**- a a *t. -.-J.. .-40 .I -4* a * ** ...~ -* *.* ** ,..S.* **4 -* a.S. S S S* .9 S. *j a a* ~. *.*~e L.~ *..S*.* St ** .S.5 .S..'. *J. a 5*~.* 5~.'- .5* .* S d 4* .5 a *a-, ¶**s* ~S**. S* 5.0 0 a Sfl 3 INK SI I I I.9'C9R 16VhAO eo OA ~RV o460*Can 5AAC ZA am C6 O I

• 41 l'~.'/; , I 1"-4* *S .*.. S 4 .* °g 1.4 4 -4 4-I 1* 4.4* 4-*.-0.4.~ .4 0.4*.4.'4..----4.4.*4 --0 -4.4..0.4 1~. U ..-*..'-*.0.I*'.0 -4.0 .*0* I.K*.4 .4.4~4 *0~ -.* 4. .0. * -.4.4. -4* a .1* ~ 1-. .S I -.0 0*' ~4*~. -.---.4.0 .~.0.-.4.4**.0 4*~. 4 4.--. 4.* 1** *~*-0~.~ .~ * ---4. .-4*44* *4 4.444 -.4.44.- -.**- 4-.*~0~~--4* ~~~: 4*44 4.0'.* 0 ~**4*0.4 -4* 4..~~1 ii i i i i-.I *4^4aI6 too.I~~ *T7 'or-

-~ .4 ..V.-k J 'a.9 -* 9 ...4 -. -...1. ..*.~* .1' *.* -.9 I -* -'a-9 ....-...-.-... S.9 9 ---4 -9*. 1 9 9*.* ...._ _ L.. .*-~ -9..1* 9*. ** ... * .9.~... .. .2 ** .--S.* -.9...9 -.~.9.... .9 9 9-*. ~ ----.~ -.-.... .... .**.. 99.* *** ..9.- .-.L. **- .*. .-9. 9... ..9 ..9. .9.* 9 *,4 ..-9-.. ,* 9* .9 .-.. 9.9*7.

• 1* -. S 94.. -.9 9.4 -9..* *;. 'a* .. -*. S-4 9--- .'a.. -I...9 -.9**~-.4~'* *..9 -.4* '9 9-IS o.9. -9 9.. -.9.9 a 9. ..9.9 -* p9.4* *...* .SW. -9* -* -.. .9'I.* ;$ .-.4 a ., ,,.I ~SE "A7'7 I it.,r i i CAI L I* .~ .. --..9~~4qj~IIUI~i

~ -~- ..~*

-4* D9**

1~?*;.:-Z.. **4* .9..i A. -¶a --.a.9 *.9 9.9 .--* *1.9 9 ..9.a -*99 9.a. ~*

• a* S ~-a, *9 -* -. 9-9 99-* 9.9 --.'*.IK.a *~. *-a. -.9. ----9.9 9-: 0 a i J JI* IADESVICSKO*Wl IjsgIrvnI

~~j7hAae~w~

I t L...A Tt9 I C A SI'" 7P i1 --.. .. ...-- -'

6° .+

• 6...: -.- .6" * -."* ",.6. .. ."-6. 64-- -6.* o. ....* -.9 * *.

• 46

• 6 T 1: ' .a a'"./. ,3 ',"-..p* 9 -.9 .-4* .* -.9 4.4.1*6..6 .~* S*6* .6 ,99.6499 *46~.~**--.9. 6* * :.6 .-.6.6-6.4~6.6**96~.66.4..6 6 -.6. .--.9 6*6~* I 6*.*6 ~.9* 6. .6-669* .6 ~ .~ -.6 -6. -.6.9 * .* 6.6. ..6--~ *~* -I.. .6 -4 .6 .6-6. .6. .6. -- C* 4'- -*.6 -...*'.6. .6 --.9 6*96. 4. ...~ --*- f..... -. -6 *6.~ --.6* 6~ *~ p..-6*

• 4***6 ..6* -.. .6. * -.6.6 ~ ..6* *~. .6. .* 6* 6~ ** 6.49. 96 -* 6:.. .961 .6* -. *. 6.~* .4 -9 .* 6.6 .* .6 6 6 * -S* 9-..- 9-...~ ... -* -.:- -4?~ *.*~ **~ *6-... .6.6 -~ 96 * *~ 9 -6~ 6 S -~.* ~6 6. .*~ 6*.* Se-9 p .9* .**. ..6* -*6~

• 6 .~ 4 I Wý -a r MADJIT .H CHAR6K NO.%Ise c.cm tI 6 A1/47.c *lr .. Znfld/0 op.

-. ..Jz.*L *I *.~g~1* * ... -S 0* .. .4-.4. .. *a -* .5 5*-a--.a.5. a a.4-.,* 44*5..4* 5 *** .S-5 5 S-S .5 4-S a.-* S-.~ 1..- *s* *5 **. I p.4... .S. *. ..4.-.* 4.5..4--.4-I. ** a-...5.~. ..-.-.5. *S*45.4. .5.4. .4.p -S.* *~ -p..-.5 .-.4 .. .S.5 -..4.4. -.5 -I -*~ *.5 --5 5* *-5 .*.~ ** 4. .* *.4..- .5...5. 5 5...41:.- ** **~ .5.-S a.....~. ;a~.5 -.. A. **...5*- -a 55 **~*4*5 --4 .5. 5.5* 5 .4. .a.44 -4.* *.* 4* S a* 4..4..*55*5 -a.5 .S* .5* .5. -* p* ..5..4 .4~1.4 -.5-y

.1~...I 0 I *0.. ..-~. ...-** 1, -S --*5*4 *..~* ~.C... *... ~ -.5 -* *.. -..5.-.5. **5-5 .5.* S *t.~.* .7. -;C.

• a L S i-r I OV3a~ CREEK ROMuPSEi WOIG Can Ix~u P=333 aa FJD I 78 d COORD~~~~~

S PH I. t SS S-0 PART I.~d ~ .oil. -ek~m.L~wL .L-O &d~m4~.I /if** 4&FACESO--.

t. 2.44.4 9.25*04 2.785E1404 o0.00E010 6.469E+04 2.751*04 9.2541404 44.4 -.7l99f*04 -2.3311404 0.O0001400

-7.799E104

-2.331E.04 46.7 -Z.3421E03

-1.s869E04 O.0001+00 1.635E*04

-1.669E.04

-2*.342E103

".M86 lOZ"-- !00. '2- 296E.04 1--ti!O4" "S.786E103 46.7 ..3911.04 -1.S651E04 D.001E.00 2.95"1104

-1*.565E04 46.9 -9.e616403

-4o198E+02 0.000E400 9.39U8103

-9,8116E03

-4.198E*02

... 49- '.*3m9 03 -3UTE03- OOOOEl00 .t;16tE.03-48.9 2o.0601E04 .969E+03 OoO.E*00 1.2631E#04 7.9691E03 2.060E404 M9. -t0&tFF (.pIME9U02 0.Ut9U W.MEW03 -&*UUIE*tU

-(.*9FWOZ9U lg we 7eAuvcw9.L AOMCC1.Wr UQ.VUIUVW L.OUEWJ &*71LLWu1 U WV-0oo0e001# -t- I"3EWO*,~

1. iaF3v03-2*019 40 --Vw~ 0--S.Z- S.4lE* Or 9;4S0E03--

0.0001400=

-?-.a E03- -.4- EO?1E .4501403*03 51.2 5.4511403 9.040E.03 0.0001.00 36569E*03 5.4511.0O3 9.0401403-... 53 -- 5;38 2E-03* 3 4""03 "OOOE400 O 53.5 S.551.*03 8.4301403

_OO000E100

--13.S 33E4O03--" q.-" a I.ui,@E00-

2. 9630 - 3G2103--1-.345E#03 2.872E*03 5.S551.03 8.430E403 2. 7114E03 -..733EV03-&

8 W,*03 S3.S 5.4231*03 8.397.103 0.000.*00 2.974E403 S.423E*03 8.397E*03 9307 "OSS6.0-#03 V-*fOfvw0-2e-nt1*603 3*9 .r~6 O 'm... --53.5 S.693E403 8.539E403

.000ECOE Z.845E103 5.6931E03 6.5391+03 55.7 6.293E103 7,6001+03 0,0001.00 1.307E103 6.293E+03 7.6001*03 9904 .6.~u 5083.e elt&9wu uOOOWWW telI'ZEwu 9*523eYWJ IS.tSEYv3 55.7 7.230E*03 O.000E#00 2.2771.03 4.9531+03 T.2301.03 56l.O 5.9331403 T.066L+03 O.OOO00E+

1.135E+03 5o9334E*03 7.06.E*03-.----5 i t0- 5-"62.t03--- -i9"@- (-.O00! .U -M900 3,29t14+ Or 5.620- -.99 1.0t 58.0 5.391E*03 6.913m403 0.0001*00 1.SZE.03 S3.3911E03 6.913E103 M-- t7-6 .,. i. Let-eVi3 OsOO0!0,-"-OO?,1".

--58.0 5.662E103 T0O26E+03 OO001O00 1.360E103 5.662E*03

1. 026403.... S8. 3. 0QE "--69fl #3 --O.o001DE00 1-;53?E#03

$O.*390!uw 6.9314*03 60,2 5.693E103 6.961E03 O.000E100 1.*21E40.

9.6931E4f0 6.9g1E+03 6u4a goe 16903 66916wV03as "0uO40 t*293 083 SMSuE'uz soiw40-lSl -1 S 444 7 -..... .~~16 ' ciUlmm ~

01a3em hU19 iF&rI., TAS"g5Ip*lPU3-Tt Will Sp USIA a 5Wamm alm i I SPART I elj~wmrn4 at I~mAy)3~A

~u~Aftt~~akyL.

FACES-t 2.62e5 5.645E*03 6.942E* 0. oOOOE o Z*297E+0)62.S S.T96E*G3 6.9616f03 0.000100 1.392E*03 5.6451+03 S.*T *03 6e9421*03 606SW0 6.@96&E*03

-6t.lS.610[,0-7.003(.03 00OEOOD t. 306ME* S.-Sir0V0 s1.OO*03 62.5 5*7201.03 7*003E.3 0.0001.00 t*262E*03 5.720E+03 7.003E403.... 64.--r --5.-6991Io3-

-. 6-791f0-~0 0O-T -$ I.-79E+03-64.? 5&745E+03 6.993E403 OeO0E00 1,248E+03 6.9&3E03~S-~.e 9.EW4i. g.&uuutuw

-00 01'00 IZZIM03 $?9tE*03 -1001E*03--

6t0 5.766?.03 6.940E.03 O.0OOE000 1.1?3f*03, S?.671*03 6.*0E*3 6100 c*3 StO0OM 0~4E05iUw363t0 I S. ______* 14 q r ~ --. --* 'j~4gL..v~g.~

A.~k.rc IL!~'V2 ~w" ! IAw;ll W__ -A_L_ -M= -,. --.

--oys~fteCrEEK3-0WIFt EXS.KENT, CASE I, P10z T &XI-11, .....-- - SPHI-S THETA--S-SHEAR

-Sr-SZ- -Sr-S3 SR -* PART t ....-...o.e. .FACES-l. to 2.44.4 6.655E+04 2o00 t*04 O. 000600 4.654E*04 2.00 10EO 6.6ssE604.

+1 &LOLUCOVl%

V*VUIUt:4O w

Se'%Vpt#03O Le IL3It#04 -444 -4.356E*04 -l.302E+04 O.0000E*0 3.053E+04

-4.356E+04

-1.3026*04 46.7 .5.36SE.03

-4.5346E03 O0000E.00 9.8986E03 5.363E*03.... 41--.S6E*O4P-J.-166Y0-3 00E0 0 "'l.'13E.G4----3."SIUBE0-

] 04 46.7 1.576E*04

-2.602E603 0.00EFO00 1.637E,04

-2.602E*03 1.5T6E.04 46.9 50.676E+02 To.0+,03 O.000640o 64.133*63 "5.'+67662 7.301+.03----4.-9-'-r;033-+04

-9 99VOE6403-O O00oE+O " 3;44S'E+OZ--9VO9'90Ef--

r.-031@334r.-..

48.9 2.OLOE404 1:260E+04 O.000EO0 1.422E*03 t.Z68E604 2.0106E04 48.9 .O334E*04 9.762E*03 0.000E#00 5.698E*02 0a Iuw U. Ua*&FW 947626.03 1.0336404 la.Z5[E#4f

2. 02mE-04-" 9to- &- % "1"17 71 rlr [ --n rT--- &t r' --0 a 0-51.e "s90OlEfO3-1.2Z2S

--53?6E.0-6.V0 3

51.2 1.039E404 1.330E*04 O.O000E00 Z.oIZ4E03 1.039E*04 1.3306*04 Ldmk I S L.SWat9U4 l.4S.5E9u%

U.V*VWIO %e4ut*uz &oJuut*M &* ssJttgi--53j. +TI0366404

-"l284E'04--.O-00k'0'-Z.

47 EV03-- r.036E-- 'T Z84E404 -53.5 1.O48E*04 1.2894E04 0.000EO0 Z.413E603 1.046E*04 1.289E604 5 -T19E 4 eZ. 0*0

-" 53.5 1.039E*04 1.267E+04 O.0O0000 2*479E603 1.0396O40#

1.267E*04 93-3I .BE &.ut. le9E0 u.UW0E*0u 24S.EIpttu*

loMEWt04 L~Z 53.5 1.057E*04 1.296E6.04 0.000E+00 2.397E.03 1.051*+04 1.296Et04 55.1 1.097E*4 1.2l341*04 O.O000OO0 1.3?1lE03 1.O9E*04 1.234E*04 55.?_ 1-.0OE604 1.2106.04 OO006E*00 2.003E.03 1.1OOE.04 1.t2OE0D4 58.0 1.OT66E04 .e996E*04 O.O006*00 1.Z3.6E*03 1.OTE604 1.199E*04.. .u p iO ik ~..4-1;'9i

'--.ooo6 .00"- 1; 386. 0I--- -$56.,0 1.039E#04 1.18+9604 0.0004*00 1.494E*03 l.039E+04 1.l89E604.10... 04 0 1.3*4037 1057i6.04 1&1962--ZE6.04 5l.0 1.196E*04 O.OO E00 O0 1*364E*03 1.O3tE*04

l. L96E*04 5004160 IUft94 &.i'gi.Tu iiuu~I00 1035EISUj L.U5,&9U, LotvzEtvm-- *u~ .opg .v~~u .--60.2 1.016E*04 1.192E6004 0.000E600 1.312E6003 1.061E404 l.1926E04"-*' to 9E901%'. 9V it *u i.000E6T---.-3ZUE"I 0"-- 1x)39' -I"911-04 1;4-----i ~I~ c~lLnZ~1AIA I UT..--.--..--.

£ ~i~JSNYCt54hVI I I* , I

...rTaCrW PoP JIgtIMENT.-

CASE it 1-U Is ,11---7-wCOmR-"- V- IxI-T -.-1 S" ARX"-- 'S-sL .sr-w--- = ..* ,PART' -_4,,.J,,,.

@a e/,c,,,-2

-./-4,. J2,.. /l.

FACES-I. t. 2.62.5 1.059E104 toLS9E#04 O.O00E*O0 t.3OOE*03 toO9fE,04 I.1O9E*04 62.5 1.068E#04 1.192E*04 0.000EO0 Le.24ZE403 1.068f+.4 1.192E04-" 274i.5" taGIE*04 t7¶21*0* Oi00tE4rO tro304E*0" ts0GIE4- .e I92E*04 62.S t.0631E*0 t.1921E04 0.OOO08* t.26eE103 1.O63)E04 1L.92E*04-64.? +' 31[.04 1* r0e0.o4--VOo1e.o0

"-1". 2691.03- t-o063fV2 T.'T90'404

-64.7 1.066E.04 1@191+.04 O,.O00E00 1.2418.03 I.0661.04_

_1191E04..... 64' --- I-o'+ 69lwQ0W-"I"'",to

.'2O41 "Oo-"OOOa "26V+Ol'03""r7-69'V-, * ......-- 61*0 toUS9ET04 112E904 0*000tM N2261.03~W v 'U00i551V;V0 4 -e55W 67.0 1.068E+04 1.186E+04 O.0000+O0 L.1a8E403 t.061[*04 1.186E.04--'.04-- 1+;? '0- "I-t162103"-10-~t[-4-"W;r871A 11 1-._ _ --...............j ... ..:....

0 wow5P I..aIII Pi/iTua 5flU[WI211g bI Li r57 II M*SWL U5I4.EIAUR O' PART I FACES--I.

1. 2.44.4 5o255E04 1.5177E04 O.O000E#0 3.6?9E904 .a5TTE*04 5.2SSE#04 44.4 -4.051r1.04 -l.o21Ee04 O.000E*O0 2.640E,04

-4C.571l04

-1.21?E704 46.7 .9o91LSE03 I.300E#03 O000oO1 0 6.614E*03 1.300E+03 9o.9lE503--4* f7V6?6E03 -T.1. U UEZ- 0CrOEf00 'G4-W7103--7i 7ZTUEW02 '"096E+03 46.7 1.4771.03

-2.742E103 0.000E#00

1. .o2t9E*03 -Z.142E03 1.4771E03 48.9 -S.6LZE*03 -S.ZOE+02 O.OOOEO0 5.091E.03 -S.6t2E*03

-5.201E02-- -4.;9--5.'S36EF-2J-Z.-I0X)-0-.O0.0EO0 3.233E+03 -Z.-3VM03L-5S-.536E-03-..

46.9 1.46661404 5v.12803 O.O00EE00 I. 16E+04 5.1ZOE,03 1.666E804 ,48.9 -5*275EVu3

-4ovv,,tVtic :)uL:u =, .,9."t:nuc 48.9 5.536+03 2,316E1O3 O.O000E+00 3.216E*03 2.311E403 5.5361*03 5 3.2 .46E'03 6.9?EV03 0OO00EI0*06 i1o' 069E*03 -3ow.. .t. u lY9M03 51.*2 S.461E+03 6.4931.03 0.0001.00 1.90321403 5.4611.03 6.493E403 O .. S3- ?09EO 3"-Tit4f "3 3;2091f03

--T9/41.03-53.S S.5391E03 6.496E*03 0.000E1.0 2.9579403

$.539E#03 8.496E103.... S3"--,fl.t Y ,

0;"O01 .3291Y 0--T;S86E*0r-1981.303 53.5 3o.23SE* 03 To635*03 O.O000E100 4.600E+30 3.235E103 T7.35E*03 93*3 243E0 as 3flE"u 000001E -2;V8W*V3-.S*53EW0 52 E 443 53.S T7.43E10 go9.22163 O.oGOOE*W0 1.3TTE*103

.843E,03 9.ZZE,03 55.7 5.854E*03 7.8501E03 0.o00E.00 1.9TE710 5o85 41.03 .SOE.03 5S.T 5.375E+03 T*7.4E543

• .OOOE'e00 2.3010E03

$.37E503 7.7451403 58.0 6.054E+03 7.255103 O.O000E*00 I.o26t*03 6.054E*03

?.ZSSE03.0 5,,.*03 -7.-T SEV3 00 t-#'481E*03$.

G8-S-o 611tE03 T. t4SE,3-" -58.0 5.268*03 To.036[*03 O.0001400 t.T68E*03 5.2661*03 7.036E+03 S as *.UVDt903 iIMSttUJ U.UI1IE'w E-L Itt3 GoOIS0 -7;306V+93 58.0 5o661E*03 T.183E+03 OO000t*00 1*5*oE*03 5.661E*03 7*153E*03 8 S -" -.< 0 ,, .. ...Ift T V i mimrtm n x anicw, v, is tw)nr I. ....4" Wi N I--a--60.6002b S*6941103 6-9931.03$513t~EM 5-99flF03 O.O000E00 L.300E+03 5.6941E*03 6.931E.03 0o 000F4Q0 -T551r*r-6;" 584 T03 U U £~0-i,+o,,*++

+.+_+..,_C.

a, .Ill l ...

--- aiTmLcarut-uoozu mmNeat-cast-ir-NPon mr 1"Zfl-vnsuL rtoor--* -S_ -- S .. Sr-' ----s -- 7-- .SEX ' ..O PART It " .FACESa-I.

to Z.62.5 5.676E#03 6*9351403 O.OODE0O0 1.259E03 5.6761E03 6.935E+03 62.5 5.765E+03 6.96.030 O.O00E0*0 1.1971103 Se651E03 6.962EO03 62695 5.716t07 -A*869 !VO3 ~M000 Kt26GE,03 SeULBE90I 6elt6E403" 62e$ SeTZUE*03 6*9866E03 0000E00 i,266E*03 5,*71.E03 6*986E*03 64,? -5;695E. 03-- "6.tOE.3"---0OOOEfO0 tt7M 64.? 5.745E+03 6.9SE*03 0-0001.00 1.240E103 S. 45E+03 6.985E103 Goi. SoI9995& V 9ZEV a.L~~ U.UUIE9IN1 LocUO9EUj SeNIET0J *.92ut've 67.0 5.761E*03 6.937.03 O.0001.00 1,17l1-103 ST611.03 6.9371E03 a .i ISO 7 ~O 6;93E 03 000 W*00 M3ug Sue 6*~g23f3 9S31.03--- S -m9

--OtSoi-tCtt-Xo WtO f EECIMWO TI- CA .21 1'T ",Xv,,IlNS1AATbw--

... COR---S-PHI-" S-SWT1l -Sw5Z ----Sl-a *.* PART I*llý- ^.A. I -SJIJ FACES-t. 1* 2*44.4 4.563E,04 1.375E*04 O.000E140 3.208E*04 o3SE4*04 4.S83E104 44.4 -Z.416E*04

-7.146E*03 O.*OOE.Oo 1.69E104 -2.416E÷04 -To246E4,03 46.7 1.I53E104 5.6911403 O.0004O00 So.341+03 5.69E1O0 1.153]E04 46- ?- t-09U*1404 4Te882lvc 0, 000qQ 'O 5636E+O346UUa".M03 -WO52E404' 46.7 9.S11/*03 4.0731.03 O.OOOE0" S. 436E*03 4.0,T3E#03 9.StIE103 46.9 2.622E*03 7.103E403 O.O00E+00 4,4681E03 2.622E103 7,103.E03 48a9"- Z"O4t+..04--

9,23!'3O0001;O00

". 2 *0-- ' -to-UM -1;0411t.04 48.9 1.OS19*04 o11515*04 O.001E*00 1*044E'*03 1.1151E04 le 1819104 46.9 1.041E*04 9.l39E*03 O.OOOE*000 1267?E03 9.139E*03 1.O411[04 4009- -tr-- 11 4---1.l0lt4 1..@ 000wOo r.O5E503 !sit"04 --t-I-Tt#04'

-- "-it---- 61*03 --t;OSO0704-- S-o 36E&03--s

.-enlw 03-- re 0 50E*04 ..51.2 1.040E*04 IoOIE*04 0.000E*00 1.60*9f03..

o.040*04 l.201E*0G4 2t* tobS00 LeUYY *J5L2WU4 0.UIJULUU sl59EO3'i"l.J*35lcYO tm358&t9 SM -9;256E,03

--t-2S6E10 --0 400- 3;300E*03-" '9S236103- "t-Z5&6E404 S3-S L.O47E-04 1.293E*04 O.OOOE00 2.456E*03 1.04TE#0*

1.293E#04-

.

.-3301[,04" 53.S 9.276E.03 1,259E*04 O.001E*00 3,309E903 9.276E*03 1.ZS9E*04 53.5 1.166E*04 1,331E*04 O.000O100 1.6496*03 L.L661.04 1,3311*04 S5.T 1.070SE#4 1.247E*04 0.000.*00 .aT2E203]

1o075.E04 1.241E*04 M I L.USJCYU' lo2*22YOU9 II.U0JYJCYU

• ..UOO$c*0.

twUSIEWOS teSSt#Ui 55.e 1.031E.04 ls.Z34.04" 0.000.*00

.2.050E.03 .O311.04 1.236E.04 8.0 1.08e2.04 1.t2091404 0.0004*00 1.201E+03 1.o0,2104 i.209E+04.... S6O--I-SE4---04 2O2Y04

1. 44E*03 -;'tOS'Ef04 1.;20E204-56.0 1.053E*04 1.195E104 O.001E*00 1.6191.03 1.033E÷04 1.195E.04 s--o-a-

• t '74E*03 , I.2"[ IE204 58.0 lOSTE.04 1.Z04E*04 0.00E1400 L.464E*03 1.S0r1104 1.2041E04....5 T0 3.IEVOGtT' 0.OOOE'0r" 1.....IIIII6S)Ef0)

-31LE0- 1.1961404

""--.u .JUUIW 5.-8I 003*I U~mi*60.2 1.06E104 L.9ZE1.04 o0.* l.lvufO4 O.000E*00 1.36E103 l.061E#04 00OO E#--1*0t.7637

---34E*3O-o. 192E104"o190*E04---~*-V i.~J ~Eh .~ ..~....*

---

a~c0ou 3 "'a- 7SHEA SEA S3S PART I FACESaO-L.

to I*1,061E.04 000(400 I.ZI1E*03 1.061E#04

o. 189E+04 6Z.5 1.O66E(04.

l.0l9E0404 O.OOOE00 1*245E03 1.066E404 t90OE*404.-.. 6Z- -17,O6E*04-"It-*-O 1 *1EW040--00

+00 --"1"-2&E4f lE.4 -62.5 L.O63E*04 Omo.00(0 I.2g9E*03 GZ* LsUO*ttUi

&*KVZL*9U U.UUCVU Lo~g4tWg.5 K*UFtttU4 toIVlt*0U--4ir- .0631*04 .ZS~qw~0-0W-OadE*0

.1;Z6TE. 03- 1 3EV047 -T1 flewfO4 -44.?T 1.0664104 9l(OE*O4 O.000E400

,1,244E+03 1.066E404

1. loggE 04 GoO. U LeUOVF04 1*1522T000oOOE*09fl=2017,003 1*06W*04 taa.mrPU~W

-67.0 1.068E404 1.L864E04 O.000E400

.1otZ403 1*.O46SE04 1.1866404 4 ...... 1.0 412Ec' 04 18TEV040 .OO..;tE4 03 1. OIZE 04 1 7E*04- -I I r " /I. -..A I ,Mo- DV CK .no.I 0 APPENDIX D CBI COMPUTER PROGRAM 778 INPUT AND OUTPUT FOR STABILITY ANALYSIS CEI PROGRAM 778 DRYWELL PRIMARY MEMBRANE STRESS ANALYSIS This Program performs a primary membrane stress analysis of a containment vessel drywell. The drywell shell can be analyzed for any combination of 14 loading conditions, including earthquake.

The drywell shell is analyzed for stresses due to the customer specified loading combinations.

Primary membrane stresses are computed for each of the loading combinations.

The resulting stresses are compared to ASME Code allowables.

In addition, the compressive stresses are compared to an allowable buckling stress.The drywell primary membrane stresses are found using the general equations for an axisymmetrically loaded shell of revolution.

The derivation of the general equations can be found in Chapter 14 of Theory of Plates and Shell by Timoshenko.

The equations are as follows: General Equation #1: 14+1 General Equation 02: 2IrroNSBIR

+ az = 0 of0 where NA pe ro meridional membrane stress resultant.

circumferential membrane stress resultant.

radius of curvature tn meridional plane.radius of curvature in circumferential plane.pressure.angle between pole of revolution and point.resultant of total load on shell.R 0 SINaF.0 HORIZONTAL EARTHQUAKE The effect of the horizontal earthquake is to produce a shear load acting on the shell at the elevation of the load. This shear is found by multiplying the load by the horizontal earthquake factor for the elevation of the load. This factor is taken from curves for horizontal earthquake given in the customer specifications.

From statics the shear load can be considered to produce a moment at a lower elevation.

This moment tends to rotate the drywell shell about the plane under consideration.

Kj@0 In the earthquake analysis the drywall Is analyzed as a free standing, cantilevered column. However, the drywall can be supported by the surrounding building at the stabilizer elevation.

This support is separated from the stabilizer of the drywell by a 10 mil gap. Thus, during the incidence of an earthquake, the vessel may generate a shear in the opposite direction to the shear of the applied loads. This shear is the reaction at the stabilizer elevation, which is treated in the same manner as the other shear loads. The reaction is found using a combination of Castigliano's First Theorem and the unit load method using the following equations:

ISM oMr. Jx4-q I-L f TIapoed = Ahor. EaqrthquaJ.K +o Rc ke + ni load ~Rato SUBJECT I.. J C. OFFICE R REFERENCE NO.MADE SY CNKC BY MACE BY MDY A C#4K0j BY IsH, oF_E_ _ __ DATE..ATE

_Pv~~m.tbrUjA 00 7t IP I ftmo0 in kA44 GO 04 Rev UP 54 I~ IA aPT.. S~.-~ N I° 0 I INPUT FOR ANALYSIS OF NUCLEAR CONTAINMENT ORYWELL w ROGRAM 778 -REVISIO% I DATED JANUARY 1974, IS IN SFFECT.NUMBER OF POINTS TO NUMBER OF POINTS TO CODE FOR ANG OR EL-RADIUS OF RADIUS OF RADIUS OF RADIUS OF EPBEOMENTS 10 SKIRT

• 12 1 IzELEVATIONS INPUT 2:ANGLES INPUT 420.0000 IN 198.0000 IN 72.0000 IN 197.5000 IN SPHERES CYLINDERc KNUCKLE £HEADS ELEVATION ELEVATION ELEVATION ELEVATION ELEVATION OF OF OF OF OF EQUATOQx FLOODING&STAY FORCES TOP OF mEADx FLANGEs 3702500 7?45000 B6216TO ITC.Tsoo 94.7S00 FT FT FT F T FT INTERNAL DESIGN PRESSURES 35.00 PST INTERNAL OPERATING PRESSURES 0.00 PSI EXTERNAL OPERATING PRESSURE.

2.00 PSI ALLOWABLE PRIMARY MEMBRANE STRESS* L9250.00 PSI WEIGHT OF STEELx '0.80 LBS/SO.FT/IN.TmK OVERAGEIIN PERCENTI.

0.00 WEIGHT OF COMPRESSIBLE "ATERIALSIO.00 LBS/SO. FT MODULUS OF ELASTICITY%

29600000.

PSI*SHEAR MODULUS- 11500000.

PSI THICKNESS OF CYLINDERs 0.6400 IN THICKNESS OF KNUCKLEs 2.7SO0 IN THICKNESS OF HE&A a 1018Ts IN THICKNESS OF CYLINDER ON HEAD z 2.2500 IN THICKNESS OF CYLINDER ON BOTTOM FLANGE s 2.2500 IN LENGTH OF CYLINDER O0 BCTTOM FLANGE x 40.00 IN THICKNESS OF CONE

• 1.5000 IN ANGLE OF CONE a 30.0000 DEGREES IMPOSED DRYWELL DEFLECTION a 0.03 IN JET LOADINS ON HEAD YIELD STRENGTH OF STEEL a 33700o PSI JET LOAD a 34000. LBS ON AN AREA x PERCENT OF YIELD FOR ALLOWABLE STRESS AT 300.0 DEGREES 0.9110 St.FT a 90.0NICAGO BRIDGE AND IRON CO.INPUT OAK 6ROOK ENGINEERINIG CONTRACT GPU/O.C. DATE 01/15/67 BY1IA REV I I O INPUT FOR ANALYSIS OF NUCLEAR CONTAINMENT DRYWELL PROGRAM 778 -REVISION I DATED JANUARY 19749 IS IN EFFECT.POINT ELEV 1 71.5230 z 67.6130 3 66.5100 4 65.2030 S 5069250 6 37o2500? 23.5730 8 15.6040 9 12.2500 10 8*9350 11 8$5000 12 6.1850 CODE FOR USE OF ROUTINE PENETRA- DEAD LIVE TION LOAD LOAD* EARTHQUAKE CURVE DESCRIPTION NO.CUrVES 3 7.00 16.00 24.0(CURVE %O. SEISMIC COD 1 0.2200 0*2200 0.2200 2 0.1100 O.100 0.1100 3 0.2200 0.2200 0.2200 SHELL THK 2*7500 2.7500 270SO0 0*7220 0.7220 0.7700 0O7700 1.1540 191540 1.1540 1.1540 1.1540 WATER LOADS I CONE ON CYLINDER I SKIRT AIR t I ELEVATION 0 37.00 EFFICIENT 0.2200 0 0.1100 0 022OO Oo IN FEET S0o.O 7000 90.00 10000 ,2200.1100 2200 0.2200 0.1100 0.2200 0.22OO 0.1100 0.2200 0.2200 0.1100 0.220c VERTICAL SEISMIC COEFFICIENT

0.0500 CHICAGO BRIDGE AND IRON CO.INPUT OAK BROOK ENGI'EERIG CONTRACT GPU/OoC. DATE 01/15/87 BYVA SHTDS REV I C45 A 7

• LOADS TO BE CONSIDERED IN

SUMMARY

0 MEANS NOT CONSIDERED I MEANS CONSIDERED ACCIDEtIT PRESSURE.e..oeL MEANS OESIGN PRESSURE 2 MEANS 1-25 X OESIGN PRESSURE EARTHQUAKE CURVEe..o.vel MEANS CANTILEVER 2 MEANS STAYED 3 MEANS FLOODED CBI CASE NUMBER I ACCTI35 PST)ACC PRESS 1 OUTS BE LL 0 OPER PRESS 0 INS BELL 0 EX T ER PRESS 0 EARQU CURVE 2 STEEL wT I C0OMP MATL I P ENE wT I DEAD LOAD I LIVE LOAD I REFUEL WAT LD 0 AIR 0T 0 STAY FORCE 1 HORIZ EARQU VERT FLOOD EARQU I 1 0 CBI CASE NUMBER ; OPER(EX-2 PSi1 ACC PRESS 0 OUTS BELL I OPER PRESS 0 INS BELL I EXTER PRESS I EAROU STEEL COMP PENE CURVE WT MATL WT 2 1 I 1 DEAD LOAD L LIVE LOAD I REFUEL WAY LO 0 AIR WT STAY NORIZ VERT FLOOD FORCE EAROU EARQU 0 L 1 1 0 CBI CASE NUMbER 4/1 FLO ELEV 74.50 ACC PRESS 0 OUTS BELL 0 OPER PRESS 0 INS BELL 0 EXTER PRESS 0 EARCU CURVE 3 STEEL WT I COMP%ATL I PENE wT 1 DEAD LOAD 1 LIVE LOAD 0 REFUEL WAY LO 0 AliR WT 0 STAY FORCE I HORIZ EAROU I VERT EARQU I FLOOD 2 S i CHICAGO BRIDGE AND IRIN CO*INPUT OAK BROOK ENýINEEPSICG CONTRACT GPU/0.C. DATE 01/1S/Si BYVA SHYOf*REV I 4.dkAO AMf 216/57 PENETRATIONS ITOTAL NU1BER a 461 I(TOTAL NUMBER x All MARK ELEVATION wEIGHT IN L$SIEST)X -54A 87.00 1000.00 X -5 A THRII M 16.00 'in0000.00 X -6 16.00 6000.00 X- 7 A THRU 0 30.00 46AO.0f0 .x -8 26.00 2450.00-9A.qB .4-n) ...X -10.11 26.00 6650,00 x 12.4 31.00 16500.00 X -13Ao13B 33.00 15450.00-14,19,12P 10,00 s7sa.00 x -43,44 54.00 7850.00 y- OAA. 71.tn0 RA4fl.Of x -17 90.00 ?S0.00 x -20.21.22 40.00 so0.00 A >12,4, 20.00 6,000.0n x -25 90.00 3750.00 X -ZOA-G 34.00 5450.00 x -3flAiA.1A IA.-nn I7lan-An x -31AB*53 16.00 3150.00 X. 620.00i jqnfl.0 x -35A THRU G 16.00 -0000 y -lt&~nInn X -37 A THaU 0 40.00 8100.00 r- %jtA T"ALi n &.In.0 an A no-x -42 20.00 400.00 X- 19A0 %A ,%es6n X -40AB946A 30.00 2400.00 X- 4C.P 'q2, A t)O I ASO an X -49.50 35.00 1500.00 x- St A2.0) 7s0.o0 -_X -100AS91046 40.00 2500.co x US~ A~nflA17A 40.00n --2tfl 5 0ln ~X -100C00,G*104 40.00 i.15000 x -100E9103A*10 40.00 2500.00 x- A2A &A.80 An.nn0 .X -lOIA-F 40.00 5100.00 X -104CO 40.00 -105O000 X -54B 90.00 1000.00-55 A'B -90.0 2000,IOQ X -102A#1041.10 40.00 z250.00 X IOOF91038-40.00 1850.00 X -29AtB,47,v8 90.00 4030.00 rI4tAGn ARInp t ItON mC.MPAN¥Y K_V ENETRATIOnS CaNTRACT GPU/O.C. D&TE 01/15/87 BY thJA SHt C3A7 V I a P.NFTRAT IWNKt MARK FLEVATIO~g w~l;,4T IN LR X -32!'q33A*338 16.00 3?50.00 I iincfl ISSL3pp X 4- iO000 500.000\I i&NiAATONSCOgTRACT GPU/0.Co IATE 01/15/87 BYV~A SmiTD6 R.Sv I-IW" A^ P41 I'l DEAD LCOAS IMAID UPPER 14P*DFA i nlor- tHPAor-;LlfPFPP UPLO FAQJS MDD~LF wPL0 PADS I nw~Ft w1:1 r PAD'TOP FLAN1C.F 413TTnm ELA44GF TUNFIZ W1TPP q.hI STARTL T7Crs UPEFR REA1 SFATL I lwrPA CIXAM C.ATq FLFE~tLLLN 60*00fl-4n, 0 6S.00-Aon 32000.00 216SO.00 ii 02000 .0a ii>A2AfA..mm 12 El flTA~l Faf ani A-25 4A0fl0.n.E PFAlONEL iLfLL. -10f.1)n 64 Q v~jjj 1 4 .d t" nnrnaina IA FT nTA FP1n '4024 q7fl00.0n IAPPFQ wifln PAO% lbs.00 t2000,00 MyrD.-AF WLFln PAQ% 60.n0 19202.O0 UldWF WELD PA35 860 400.00 DRYWELL SHELL. IN CANTILEVER CONDITION4 CHICAGO 9RIOGE E IQON COMPANY O)AK 3A0OOK ENIEE1'IF A iI IAn 4 r ,p.,TgAtT tno#'i.rpc. .iA vr ni SI/TR7 VTA qmT 09-1v t 641pco PfAI Zh/5I7 I II LIVE LOADS L OAO ELEVATION WEIGHT IN LBS UPPER sEADER 60.00 4200.00 LOWFR AR 40. 0 t71so.o0 uPPeF Witt0 PArnq AS. 20000.00 MIDOLE wELD PADS 60.00 20000000 LOWER wELO PADS 56.000 2400000.UPPER OEAM SEATS 50.00 0.00 EQUIP DOOR 30.25 O00000.O0 PFRc.ONNUL LirK 10.00 .sooo.0a LOWER EFAM SEATS 22.00 0.00 DRYWELL SHELL IN STAYED CONDITION CHICAGO BRIOGE E IRON COMPANY OAK ,RO0K EN'a4EERI';G t fv

• I"At' rnNTR.CT CaPU/.c. o'lIs/8"i AY¶JA /HT13 E.v.L_OOt p /,~

SuJMATtON OF STRESSES, C!t CASE I ACCT(35 PSI$LOAD POINT POINT POINT 3 MERID CIRCUM MERIO CIRCUM MERID CIRCUM ip/FN LA/f!N tq/TN lI BT4 f A/ TN I R/lo npFGN ITNIET AL PPEESURP 146S. _7S2. 2749S. 73SO. _14 OPERATING PRESS 0. 0 0. 0O 0. 0.;XTFRNAL PRESS 0. 0. no p. O. .mORZ SEISMIC ON SMELL 49. 67. 63. 237e 620 197.V,;Rt flmtc nN tmFLL R .. 11- 2 S .I?,. ..... _. COMPRESSIBLE MATERIAL -17. -23. -24. -121e. -34. -148.VERT SEISMIC ON C.M. 1 1* 19 60 2s 7a LflAn -270 -27. -27, -110 _ -38. -03, HORZ SEISMIC ON PENE 3. S. S. " be, &. 16.VERT SgTsmfc nN PFNF 1. 1, ..... se 20. ... _DEAD LOADS -SS -7?b -650 -245. -864 -207... nRz sc Mt smi 0% .L. 14, ... q 19,IS 69, 24d .._ ... VERT SEISMIC ON OL. 3a 4. 3% 12. 4. 10.HORZ SEISMIC ON L1L1 0. 0. 0. 0. 0. 0.VERT CuTsmff ION 1 .. 0n. 0. fl- 0. n,--. .t-V vERTICAL AIR LOAD oe 0. 0. 0. 0. .o INSIDE SELLOWS 0. 0. 0. 0. 0. 0.REFUELING WATER Oe 0. 0O 0. 0. 0.mnnY~ tFiSPTr ON wATF& M. i). n_. O_. m_VERT SEISMIC ON WATER o0 0& 00 0. 0. 0.STAY FORCE 0. 0* 0. -1* 00 "..SUMMATION 4.#EýILBS/INI 3293. 13315. 4513. 26412. o989. 20212.StIMf!AYTflnp i.r0IiDqft 12.7%.. ..A.S5.5...-

I1A9- --.91t~g*.-2AOaea 7s52.1g SUMMATION I-EOIIFPSI 1150e 48?Z. 1600. 9522. 2495. 72t4.bUCKLP4G RATIO (#EQI 0.0lQ 0a0O CHICAGO BRIDGE & IRON COMPANY OAK BROOK ENGINEEqING AffT~l~r DsyI rnN.TRArl Cn01Iin.c..

nAtp milt%/t57 PYfJZA SHTft 1 REVIy.CeJ. Pf 2~R, SUMMATION OF STRESSES, CBI CASE I ACCT(3S PSI)LOAD POINT POtNT P3INT 4 56 MERIO CIRCUM 14ERto CIRCUM IEoR IO CIaCUM 1 vu f Lft/1N ta/iN tp/l'. LF/TN I/'DflSIN TNTERNAL PRg.SURV 73AS. 73sOm 73S0. 7350. 7350. ..i m s.OPERATING PRESS 00 0. 0. 0. 0., 0.EXTEROjAL PRESS 0. O. .0 0 Smail t d T T 'l' .. At!. -Ij n 14SO -1in?. Is?.HORZ SEISMIC ON SMELL 66. -660 34. -34v 41e -.1,....VER.T SEISMIC ON SMELL 16, -1l3, q,. -7, 9. .=9 COMPRESSIBLE MATERIAL -31.4 .5 -27. 190 -35. 35.HmR7 SFISMITC 134 r.M. 4, -3. -.... -3. .5*._VERT SEISMIC ON C,., 2. -1 1t. -1. 1. -20 r%& -12,.",. -I I, .Ia.. -31. 1Jl,*4ORZ SEISMIC ON PENE So -5. 3e -3. 4.DEAD LJADS -71. 71. -122. 122. -536. 536..4m0RZ -F TMC. Jay n O .t. 19, -IQ, I14. _-14., %4o .liL.VERT SEISMIC ON DoL.. 4. -4. 6. -6. 2?. -27.t vE I nA04; .-In n O ... -34f. in,. -29, 2'.a...HDRZ SEISMIC ON L.Lo 0. 0. 2. -2. 4. -4, VFRT SFT9r-1C ON L.L. 0. o. 2. -27 w.VERTICAL AIR LOAD 0. 0a 0, 0. 0.INSIDE BELLOWS 0. 0. 0. 0. 0. 0.nuT~tnF Atiinwq 0. 0. D_ 0. 1. 0.REFUELING 4ATER 0. 0. 0. 0. 3. 0.MnqZ AN WATPR 0. 0. 0. 0. 0.VERT SEISMIC ON WATER 0. 0. 0. 0. 0, 0.STAY FORCE 0. O, 0. 0. 0, 0.SUMMATIO4N I.EC)IL5S/IN) 7012. ?bO?, 704,9, 7611. b6b6o sO2c, SUMATION c TB32. 19OO. 0782. a900..._ 5 6 .SUMMATION

(,FElfPSI) 9712. 10 5O3 .. 9263. 10542, .p IT_.g ._ r 0_Lj*SUMiMATION 0 -EO)IPSI1 93866 10848. 9557. 10743. 8291. 10799.BUCKLING RATIO t.EQ) 000 .0COOC... 1-9-011 -.eme" ....aea~~aaL CHICAGO BRIDGE C IRON COMPANY OAK 6ROOK ENG1NEEVI".CD ACETI3S PSI) CO.TRACT GPULO.C. DATE 01115/.?7 BYIIA S,4TlIQEV L._.ca,~ PM4 Ak/7--

/SUMMATION OF STRESSES*

C8I CASE I ACCT(35 PSI$LOAD POINT POINT POINT 7 a ,ER8D CI CUM MER O CIRCUM MERKO CIRCUM!FSTnN ~INTPRNAL ERCq~tIRF 715M. ?SO 711s5fls 71ASO 23,50, OPERATING PRESS D. 0 .0 0. 0. 0.PXTFRNAC P8Pg% n. 0. 0. 0. .., HORZ SEISMIC ON SHELL 750 -75. 145. -145. 2210 -221.VPRtT qPTsmTl rit SHPLL 13, -IS, 21. -25, 27, COMPRESSISLE MATERIAL -540 63o -85o 106. -113. 137.H.y7 KFtmitC ON C.M. It. -it. 2P., -22. 35. ..3i,_VERT SEISMIC ON C*M& 3. -3. 4. -So 6. -7.4ORZ SEISMIC ON PENE 110 -11f 27. -270 47. -47.WF R T ý Uf H T Q -- P FN F S .-s .1 2 , .-1 2 , ts .-1 5 , I DEAD LOADS -708. 706. -1677. 1b77? -2113. 2113.VERT SEISMIC ON O0L, 35. -35. 84. -84. 106l -10b.HORZ SEISMIC ON LeL. 10. -10. 24. -24e 38. -38.VE61 ¶~SE~ I l 1-. 4-. -4 A-. -fk. 7.VERTICAL AIR LOAD 0. 0. 0. 0. 0. 00 INSIDE BELLO,S 0. 0. 00 0. 0. 0.IkTYfl imnw% A. 0. 0,. 0. 0. 0.REFUELING kATER 0. 0. 0D 00 0. 30 HrR2 SEISM1IC ON~. wATER 100 0. as0.o VERT SEISMIC ON WATER 0. 0. 0. 0. 3. 0o STAY FORCE 0 00 O. 0. -1". 1.SUMMATION t*EQIILBS/INI 6445. 8298. 5473. 03ZT. 5132. Q684.%cIMPjATIIN I-P0Q(i.RS/11N

!58470 ag~g. !kilo. 10A40., I I V. 1lebQ5... UMMATInN 8170. 107tT. 4743. 8083. SUMMATION

(-EQIEPSII 7594. 11558. 36"14. 9220. 2715. 10134.BUCKLING RATIO t*EQI -eee e.3060 eee gur-u1%I0 RATIO I-EQ) -- AaAA efe Z CHICAGO ERIDGE I IRON COMPANY .AK BROOK EGPEERIG AECT1IS pt11 CONTRACT CPU/I.t. OTAF 01IlS/f? AYXIA SeY ILLEV_. ..em 7 SUMMATION OF STRESSES, CbI CASE I ACCTI3S PSI)LOAD POINT POINT POINT IQ MERID CIRCUM IAJIN lIlTN DESIGN INTERNAL PRESSURE 7350. 7350e... ..EXTERNAL PRESS 0. 00 SHELL VERTICAL LOAD -808. 919.NGRZ SFTSMIC ON SbiELL 1.01. -406.VERT SEISMIC ON SHELL 40@ -46o COMPRESSIBLE 4ATERIAL -168. 19, s HORZ SEISMIC ON COMO 65. -65.VERT SEISMIC ON C.oM a. -10.PENETRATION LOAD -'24. 424*HORZ SFISMTC ON PFNF 924, -94.VERT SEISMIC ON PENE 21. -21.OfAD LOAOS -2991o 299 ..NORZ SEISMIC ON DeL. 944. -944o VERT SEISMIC ON O.L. 150, -150.LIVE LOADS -208. 208.moR7 SFPTMTC ON toL. Tsk. -72.VERT SEISMIC ON LsL. 10. -10.VERTTCAL AIR L3I4 .0 0.0 TNrlOP %ELlWn n. n.OUTSIDE 3SLLOWS 0. 0.HORZ SEISMIC ON WATER O. 0.VERT SISMICT ON WATFR fl. 0.STAY FOR9 -1. ,.ttimm&T~nN t.EQIIIRS/YN3 4SS4. 10277, SUMMATION A-EZIILBS/INI 9'46 13899.SUMMATION I.EO3IPSII 3946. 8906.SUMMATION_4-fClteStI 82(j. 12L04'. ________AUCKI TNrG RATIO l#FO)BUCKLING RATIO (-EQI .e1--ee*CHICGOAIE L IRON COMPANY ..AK BROOK ENGtNFRIRNQ ACCT13S PSI) CONTRACT GPU/OoC. DATE 01/15/4? BY72A SHT014REV I AI I r SUMMATION OF STRESSES, CBI CASE ; OPERIEX-Z PSI)LOAD POINT POINT POINT 2 3 MERID CIRCUM MERID CIRCUM hEERID CIRCUM I )TNi I A/TNJ I 4k/TNi I AITN I c4ITk~ I A/TNIk FrlflN TNTFRNAI n_ n- n. 0. n. 0.OPERATING PRESS 0. 0* 0. 30. 0 0.F:YTFRNIt Ice-?& -272. _1%1?1. -42n.

mgRZ SEISMIC ON SHELL '9* 67. 63. 237. d2, 107T uAY& SETt.M1jC ON~. tk.4j A 1- 17 SI 17.. 55...COMPRESSIBLE MATERIAL -17, -23o -24. -121, -34. -148o mnQ7 S~f¶Mrr ~n cM. -i -It VERT SEISMIC ON C.m. I* Le to* b 2- 7.P~fiCTRA r?n?fN I *'lArl _;r%_ _27- -21, -1A -i. -'Ak--l MORZ SEISMIC ON PENE 3o 5. so 18. 6. 160 vRTT Sgtt*4tr IN PPNF 1- 11._..DEAD LOADS -SS, -'*. -65. -245o -86, -2070 M.'IR? rrT~m1c nN n.,. 14. tq. 1q. .24. ST.5 .VERT SEISMIC ON D.L. 3. 4. 3, 12o 4. 10.iiip iflAntn n~. A.0 OXRZ SEISMIC ON L.L. 0. O 0* 00 0. 00* v-mr S;Tt~mTr Dly _ n_ n_ 0. 0 _VERTICAL AIq LOAD 0. O 0. Do 0. 3.INSIDE SELLOWS 42. S8. 50. 189. 66. 16o.REFUELING WATER 0. 0. 0. 0. 0. 0-wnD7 S !$cr n WATER n_. n_. n-. n..VERT SEISMIC ON WATER 0. O 0. 0. 0. 0.STAY FORCE 0. 0. 0* -1, J. -9.SUMMATION (oE^)4LBS/IN)

-360. -997g -507. -2610 -'66. -2374.ruimm&TION

/fpj -1%21. --t21 fl -720, -11.12. -in4 , .._.-_=.109 ILL SUMMATIONt 44EQIIPSI1

-134. -3L3.f5. 971. -ZV --0?SUMMATION

(-EQIIPSIS

-194. -453. -268. -127T. -390. -IL50.BUCKLING RATIO I*EQ)-e ...., D.C." 13 -i. .i..RuCKLI%G RATIO I-F01 CHICAGO BRIDGE & IRON COMPANY OAK BROOK ENGIk4EERING-GPI X-2 PS.I. CONTRACT GPUln.C. DATE 01/1/87 BYWA SHT~AL5ýj...

____ ____ ____ ____ ____ ____ ____ _ ~2/ h, qmplM pL SUMMATION OF STAESSES9 CBI CASE 3 OPERIEX-2 PSI)LOAD POINT POINT POINT 45 6 MERID CIRCUM MERID CIRCUM, MEqIO CIRCUM LE/IN La/I14 Le/14 LE/IN Lb/IN Ls/IN D!STSN INTERNAL PRESSURE 0. 0 Of. 00 0. Go OPERATING PRESS 0. 0. 0* 0. 0. 0.EXTERNAL PRESS -420a -420. -420. -420. -420. -42.@%oFti yFRTTCAL LCAD -320, 252, -17T1. I&S, -t,3T let H3RZ SEISMIC 0 SHELL @6. -6t. 34. -34. 41. -,1*VERT SEISMIC ON SmELL 16* -13. 9. -7T 9. -9.COMPRESSIBLE MATERIAL -31. Is -2?. 19. -35. 35.HOR7 SEISMIC ON C.-. 4a -4. 3@ -3. 5- -5.VERT SEISMIC ON C.m. 2. -1, 1. -1, 2. -2.PENFTRATfO' LOAO -32. 12, -19. Ia. -31. 31.HORZ SEISMIC ON VENE 5. -5. 3. -3. 4- -4a vFRT rN PINE .2 -2. 1. -1. 2. a2 DEAD LOADS -71. 71. -122. 122. -536. 536.mORZ SEISMIC ON D*L. 19. -19. 14. -14. 54. -S.VERT SEIS4IC ON O.L. 4. -4. 6. -6. 27. -27.lTVF L O1 S 0D 0. -30. 30. -29. 29.#IORZ SEISMIC ON L.L* 00 0. 2. -20 4. -4.VERT SCISMIC ON J&.L. 0. 0. 2. -2. 10 __J1.2* VERTICAL AIR LOAD 0. 0. 0. 0. 0 .3*INSIDE BELLOWS s5. -5S. 24. -24. 20. -23.intISTMED RELIW; -47. 42, -IA, la, -is, ISO REFUELING WATER 0. 0. 0. 0. 0. 0.HCR7 ON WATER 0. 0. 0 .0. Q .VERT SEISMIC ON WATER 0. 0. O. 0. 0. 3.STAY FORCE 0. 0. 0. 0. 0. C.SUMMATION

.*EQ)ILBS/INt

-745. -076. -716. -164. a 10e5. 2'5.%tjMmATTnN 1-EOILES/1NI

-979. 49. -55&. -19. -13al. f.S4" SUMMATICtN I.FQIIPS!

-1)oi. -244., a-Q91. -2zl. -14.Q9s SUMMATION

(-EQ1tPStl

-1356. 68. -1198. -27. -17930 7S3.SUCKLING RATIO I.EI) -039 'a CHICAGO FRIDGE G IRON COMPANY :AK BROOK ENGINEEPING 1 Q-ATE mt /Is/a?c-.je 0 I ASUPMATION OF STRESSES, C91 CASE j OPEq[Ex-2 PSIT LOAD POINT POINT POINT MERI C IRCUP MERID CIRCUM ME ID CIRC'JM Lp/IN LA/IN LA/IN LI/IN L/TN LAI/l!OEnIC'J TNTFRNAL PRESSURF O. 0. 0. a0 0. 0.OPERATING PRESS 0. 0. 0. 0. 0. 0.EXTFRNAL PRFS. -420. -420. -420. -420. -421, -.HORZ SEISMIC ON SmELL 75. -75. 145, -145. 221. -221.VFRT %FfS"Ir nN SHFLL 13= -IS. 21. -2S. 27. .COMPRESSIBLE MATERIAL -54. 63* -859 106. -113. 137, HnRl SFTSMIC ON C.M. 11. -11. 22. -22. 35. 11_...VERT SEISMIC ON CoMe 3a "3 -5. 60 -7.* PFNFTRATtn%

tnAn ,. -91, 0.1- -21114 I F. --A3oO. 100, HORZ SEISMIC ON PENE 11 -11. 27. -27. 47. -47.VERT SFPSMIC ON PFNF S. -s, 12. -12, 15. i.sm DEAD LOADS -70S. 708. -1677. 1677. -2113. 2113a HORZ SPISMIC ON D.L. 131. -l3t. 307. -307. 493. -4c8.VERT SEISMIC ON DeL. 35. -35. 84. -84. 106. -1C6.I fvr- I nA-I -A!L. Al. -it?., 117, -147 , 147, HORZ SEISMIC ON L.L. 10 -100 24o -24* 39. -38.VFRT CsF;e-Ml QN L-L. 4. -4. 6. -6. Im VERTICAL AIR LOAD 0. 0. 0. 0. C.INSIDE BELLOoS 24. -24. 32. -32. .1. -41.REFUELING oATER 0. 0O 0 30. 0. o.0R8 SEFISMIC ON WATYR 0. 0. 0. 0.VERT SEISMIC ON MATER 0 0. 0. 0. 0. O.STAY FORCE O. 0. 0. 0. -1. 1.SUMMATI0N f#EOIILBS/IN)

-1320. 522. -2Z89. 1550. -2629. 1905.SUMMATt0Q I-FO)ILBS/INI

-1917. 1124. -3592, 2853. -62S. 3916.SUMMATION t#;QllP5II

-1714a -U 34 ... 1343, -.A!A.. v_SUMMATION

(-EOI(PSII

-2490. 1460. -3tL3. 2481. -'010. 3393., -T ----t.- ..... ..-- -. 5 .1 CHICAGO ERIDGE E IRON COMPAvY OAK dROOK ENGTERING nPfRtgl-7 PTI CONTRACT nPu/".C. DATF glIS/l/7 gyIA SN"T Jni Lg .__ýjqjpi CM .LLý_

SUMMATION OF STRESSES9 C81 CASE 3 OPER(EX-Z PS11 LOAD POINT POINT 10 MERIO CC I, iCUm 0/ iT.6 I pI/tN DESIGN INTERNAL PRESSURE 0. 00 OPERAT1NG.

P;FS n n -a.EXTERNAL PRESS -420. -420.SMELL VER T ICAL LOAD -608.kd~o ,rtr,.*gmTt nk rmi¢, i .4ni- VERT SEISMIC ON SHELL 40. -46..fR[F MATfRTAL -- 1k2. lq 4 HORZ SEISMIC ON CoMo 5.5 -65.V9-RT qglMTC;£ nN r.M. 0, -in)PENETRATION LOAD -424. 424.HmR7 f.,_tNMTr ,N P-Pr- 94 -911: VERT SEISMIC ON PENE 21. -21*HORI SEISM.IC ON OL, 944. -944.VFRT SFIMICT ON D.L. 1so. -1sO.LIVE LOADS -208. 206.tj,IR2 QN I L 72. .7;..VERT SEISMIC ON L.L* 10. -10.VFRTIrAL ATR 1h"ll .. 0.OuTSI'VE BELLOWS -4.4. 44o , PrItt~i Twc .&TFR a..- -'-HOARZ SEIS.MIC ON WATER o0 00N ArIT A-. -..STAY FORCE -1. II ,UMMATION I.dO(LRS/TNI

.-2;02. 24Q4, SUMMATION i-EQ) LBS/IN) -0811.. 6116.SUMMATION 4.ECI(PSII

-2775. 216L.SUMMATtON I-FO1(PS1 -S902. 5300. a PERIEX-N ROATIO i TTI UstrAtGn ARIntP f- IRAnk rnMPANX aLK ARgbIC F~artI 1 cccr-, OPERIEX-2 PST) CONTRACT GPU/O.C. DATE 01/15/8? BY-= SIHT 0I8qEV I-PiM RAIZ~

//SUMMATION OF STRESSES AT EMBEDMENT, CBI CASE 4 FLOODED TO ELEVATION NOTE STRESSES DUE TO HORIZONTAL C VERTICAL EARTHQUAKE ARE

• OR -*.*1 EARTHQUAKE STRESSES YIELD A I() 'ERIOIONAL STRESS RESULTANT P6.4It I MER ICKONAL 74o5O Fl-ý 4* SEISMIC -SEISMIC ILS/191 ILS/INI OVERHANGING WATER 0o 0O VERT SEISMIC ON WATER 0. 0O WATER ABOVE FLANGE 0. 0.VERT SEISMIC ON WATER 0 0.NOR SEISMIC ON ALL WATER 5s -5.SHELL VERTICAL LOAD -161. -161.NOR SEISMIC ON SMELL 97, -970 VERT SEISMIC ON SMELL 16. -16.COMPRESSIBLE MATERIAL -170 -1?.N3R SEISMIC ON C*Me 4s -4.VERT SEISMIC ON C.Mo 2. -2o PENETRATIONS

-20. -20.NOR SEISMIC ON PENE 7. -7.VERT SEIS4IC ON PENE 2. -2, DEAO LOADS -559 -5S.64OR SEISMIC ON O.e1 ZBe -28.VERT SEISMIC ON D.Le be -6.LIVE LOADS 0 0.HOR SEISMIC ON LeL. co a.VERT SEISMIC ON LoL° O 00 BUOYANCY 2949 294.STAY F3RCE 0. 0.TOTAL 168 -1 MAX CIRCUMFERENTIAL 2.57 LB/IN Att8W eldGKLItIC Lehr, ; 2429. ielit;.f *TOIg -OF Skf-+V---

a--ae-OO CHICAGG BRIDGE t IqON COMPANY OAK BROOK FLO ELEV 74.S CONTRACT GPU/OoC. DATE 01/15/87 BY-& SHTY 09REV 1 C4xv pm. 2/-;/07 0 74.50 Fl SUMMATION OF STRESSES AT EMBEEOMENT CbS CASE 4 FLOODED TO ELEVATION NOTE STRESSES DUE TO HORIZONTAL

£ VERTICAL EARTHQUAKE ARE

• OR -.* EARTHQUAKE STRESSES YIELD A I.) IAERIOIAL STRESS RESULTANT F6,%t t MERIDIONAL OVERHANGING WATER VERT SEISMIC ON WATER WATER ABOVE FLANiGE VERT SEISMIC ON WATER NOR SEISMIC ON ALL WATER SHELL VERTICAL LOAO NOR SEISIC ON SHELL VERT SEISMIC ON SHELL COMPRESSIBLE

%ATERIAL NOR SEISMIC ON C.M.VERT SEISMIC ON CoM.PENETRATIONS NOR SEISMIC ON PENE VERT SEISMIC ON PENE* DEAD LOADS NOR SEISMIC ON DoL.VERT SEISMIC 3N D.L.LIVE LOADS NOR SEISVIC ON LeL.VERT SEISMIC ON L.L, BUOYANCY STAY FORCE TOTAL* SEISMIC ILB/ 14 0.0.O.0.27.-236, 125.24.-24.7.2.-29.10.3.-65.37.6.be 0.0.0.309o 0.-- LB/IN-SEISMIC ILB/IN)0.0.0.0.-27e-236.-125.-24.-24.-7,-2.-29.-10.-3s-65.-37.-6.0.0.0.309e 0.MAX CIRCUMFERENTIAL AtLk0w etpegLtiN6 tCA CHICAGO ERID-E E IMON COMPANY FLO ELEV 7'.. CONTRACT OAK dROOK E4GI:vEEQtNG GPU/I.C. DATE 01/15/8?7 BYJASATbIOEV 1

SUMMATION OF STRESSES AT EMBEDMENT, Cal CASE 4 FLOODED TO ELEVATION NOTE STRESSES DUE TO HORIZONTAL E VERTICAL EARTHQUAKE ARE

• OR -, (*1 EARTHQUAKE STRESSES YIELD A (*) WERIOIONAL STRESS RESULTANT Polet 3 MERIOIONAL

?4eSO FT E1* SEISMIC -SEISMIC 4LB/INI ILB/IN OVERHANGING WATER o. 0.VERT SEISMIC ON WATER O. 0.WATER ABOVE FLANGE 0. 0.VERT SEISMIC ON WATER o. 00 NOR SEISMIC ON ALL WATER 48. -48.SHELL VERTICAL LOAD -348. -348o NOR SEISMIC ON SHELL Les. -1650 VERT SEISMIC ON SHELL 35. -35.COMPRESSIBLE MATERIAL -34. -34*NOR SEISMIC ON CoMe 9o -9, VERT SEISMIC ON CoMe 3e -3.PENETRATI ONS -399 -39*NOR SEISMIC 01 PENE 13o -13.VERT SEISMIC ON PENE 4. -4.DEAD LOADS -870 -86.NOR SEISMIC ON DoLe 48. -48.VERT SEISMIC ON OoL. 9. -9.LIVE LOADS O0 0.NOR SEISMIC ON LeLe 0. 0.VERT SEISMIC ON LeLo 0. 0 BUOYANCY 350. 350.STAY FORCE 0. Os TOTAL 17G --4qf MAX CIRCUMFERENTIAL z -L 3/IN AiLLW 61JCKLIPC G-LAl-s43 LBO- ! 641 t gaeT4-t4W.SAFET.'

6 GOB0 CHICAGO BRIDGE E IRON COMPANY OAK OROOK EGIEiEPING FLO ELEV 74.S CONTRACT GPU/O.C. DATE 01/15/97 BY-JA S'lT01REV I ciir. M 2h/

SUMMATION CF STRESSES AT EMBEOMENT9 CE! CASE 4 FLOODED TO ELEVATI2N NOTE STRESSES DUE TO HORIZONTAL L VERTICAL SART40UAKE ARE

• 04 -, 1.) EARTHQUAKE STRESSES YIELD A to) MERIOIO';AL STRESS RESULTANT Pot-* 4 MERIDIONAL 7T.50 Ft OVERHANGING WATER VERT SEISMIC ON WATER WATER ABOVE FLANGE VERT SEISMIC ON WATEF NOR ESMIC ON ALL WATER SkELL VERTICAL LOAD NOR SEISMIC ON SHELL* VERT SEISMIC ON SHELL COMPRESSIBLE MATERIAL NOR SEISYIC 0N Co.l VERT SEISPIC ONJ C.o.PENETRATIONS 4OR SEISMIC ON PEVE VERT SEISMIC ON PENE.OSLO LOADS MDR SEISNiC ON D.L.VERT SEISMIC ON D.L.LIVE LOAOS NOR SEISPIC ON L@L*VERT SEISMIC ON L.L.BUOYANCY STAY FORCE TOTAL* SEISMIC I LB/IN)0.0.Do 0.490-322.132o 32.-32.8.3.-3Z°3.-71.38s 7.0.0.COo 336.0 141 Lo/w-SEISMIC ILB/IN)0.0.0.0.-322.-'132.-3?.-32.-3.-3?.-11.-3o-71.-38.-7.0.0.O.336.0-t4c4.U MAX CIRCUMFERENTIAL x t CHICAGO BRIDGE & IRON CýOMPANY FLO ELEV 74T5 CONTRACT GPU/O.C OAK BROOK ENGINEERINwG

. DkTE 01/15/87 8Y'7rA SHT*b2ZREV I CjMA~c PolA0

~wq I SUMMATION OF STRESSES AT E48ECMENT, CBI CASE 4 FLOOOEO TO ELEVATION NOTE STRESSES DUE TO ORIZONTAL 9 VERTICAL EARTMQUAKE ARE # OR -a I*) EARTHQUAKE STRESSES YIELD A *i* "ERIDIONAL STRESS RESULTANT'I8.DL ME RID!ON AL 7T9SO FT OVERHANGING w VERT SEISMIC dATER ABOVE F VERT SEISOIC MOR SEISMIC 0 Sb,4ELL VEqTICA NOR SEISMIC VERT SEISMIC COMPRESSIBLE NOR SEISMIC VERT SEISMIC PENETPATIONS NOR SEISMIC VERT SEISMIC CEAD LOADS NOR SEISMIC VERT SEISMIC LIVE LOADS HOR SEISMIC VERT SEISMIC BUOYANCY STAY FORCE TOTAL S SEISMIC (L8/I4 ATER 0.ON WATER O0 LANGE 0.ON WATER 0.N ALL WATER 150.L LOAJ -179.ON SMELL 68.ON SHELL 18.MATERIAL -28e ON Como 6.ON CoN. 3e ON PENE 6.ON PENE 2.-1220 ON DoLe 28*ON O.Lo 12.0.ON L*Lo 0.ON L.L* 0.299.Z45 MAX CIPCUMFERENTIAL

• 4677 LB/IN 96"t* .....biN. LOD 7:O .'^ t .-SEISMIC fLS/INI 0.0.0.-150.-179.-68.-ISO-23.-5.-18.-6.-122.-29.-12.30 0.99.0*2.34..... Il_. A--

i CHICAGO BRIDGE E IRON COMPANY 3AK BROOK ENGINEERIG FLO ELEV 74e5 CONTRACT GPU/OC. DATE 01/15/97 BYrJA SmTb*3REV I CA"& Pm c;/6~/07 0 74.50 PT I SUM4ATION OF STRESSES AT EMBEOMENT, CE5 CASE 4 FLOOCED To ELEVATION N0TE STRESSES DUE TO HORZONTAL E VERTICAL EARTHQUAKE ARE

• OR -o* I#) EARTHQUAKE STRESSES VIELO A Lto %ERIOIONAL STRESS RESULTANT 4ERIDIONAL OVERHANGING WATER VERT SEISMIC ON WATER*ATER ABOVE FLANGE VERT SE!S."IC ON WATER'OR SEIS4IC ON ALL WATER SHELL VERTICAL LGOA 43R SEISMIC ON SHELL VERT SEISMIC C% SMELL CO3PRESSIBLE MATERIAL NOR SEISMIC ON CeMe VERT SEISMIC ON C.M.PENETRATIONS NOR SEISMIC ON PENE VERT SEISMIC O PENE* Z-AD L-OADS-43 SEIS6MIC ON O.Lo VERT SEISMIC 3N 3,L, LIVE LIAOS'4QR SEISMIC ON LoL*VERT SEISMIC ON L*L.BUOYANCY STAY VORCS TOTAL* SEISMIC ILB/14I 0.0.O, 0.390a ez.19.-35.10.30-SEISMIC IL/tINI 0.0.00 0o-390.-185.-a?.-19°-35.-10.-30-31,-S.-3.-53a."04 0./14.9? o4 O°IIo-31.3.-536.IC7, S4.g 00 00 ERE%TIAL ' 7942. LV/IN NA~ A-k =-. r -ýMAX CIRCUMFI 41I0-A AHAIr-*i*e4OR-:r3 S*FErT- 137 CAICAGO BRIDGE E IRON COMPANY OAX 3qOOK ENGI.'.EPI'IG FLD ELEV 7?4.5 CCNTRACT -PU/O°C° 04TE 01/1S/87 BYfrA SmToL4qEV I 0 SUMnATION-Or STRESSES'AT " 4 DNE-NCT-CalCASE-4--FEWD. -T-EfrATI0--

14.50 FT" NOTE STRESSES DUE TO HORIZONTAL" i VERTICAL--EARTHQUAKE'AltE'-

t -.1.3 EARTHQUAKE STRESSES YIELD A 1.1 qERIOIrONAL STRESS &ESIULTANlr PNt 7* SEIS1IC -SEISMIC N AIR -LIT. -all*VERT SEISMIC ON WATER 16. -ISO .MATER ABOVE FLANGE 0. 0.VERT SEISCNIC0N WATER- -0oW- .-HUR 5E1HIC UO ALL WATER 934. -934.SHELL VERTICAL LOAD -260. -260.NOR SEISMIC ON SHELL 134. -134.VERT SEISMIC ON SHELL Z6. .--26."UMPKR.SI5LL MAILKIAL -50. -.No*NOR SEISMIC ON C.M. 19. -i9.-VERT- SEISMIC ON C.M. -."- ". .*-5.PENETRATIONS -Sg, -a0. -HOR SEISMIC ON PENE 19. -19.VERT SEISMIC ON P'NE 9. -9.DEAD LOADS -T7:..OR SEISMIC ON D.L. 226. -226.r ' T SEISMIC ON D.L. .68o -66.I/LlVt LUAU3J U- U*HOR SEiS1IC ON LeL. O. O.VERT SEISMIC ON L.L. -"...0.. .0.-NUOYANCY-

.. ...... 2- --- -' -sIAy FunK, 0 0 M CrCA- -.TO. * " ".... .*MAX CtROMPERIEIl~r'L-" /1/99a L'BF'IN- : ......L*t:t:um ---- -&.Wd a.. ,m Ib.raplq67

-- All CRICAGO-" DGE"C-TROW COMPANIY U,-IOU ENGINEERING

--FLO ELEV 74.5 CONTRACT GPU/O.C. DATE 01/15/07 BY VA SNT DO*REV I C9ktoAMl-v7

.+..7 9W _______

SU&AMAIOi EhetliNEMT, C4 CAsE 4"1E , 1o-LFU*TL( 7IrW3 f NOTE STRESSES DUE TO WORIZONrAl.&

VtATICA& 'EAATHQUAKE ARE4 OR -, 6.1l EARTHQUAKE STRESSES YIELD A 1*l MERIDIONAL STRESS RESULTANT (0 EARTHQUAKE STRESSES YIELD A 1+1 MERIDIONAL STRESS RESULTANT Not+ SEISMIC -SEISMIC OVERHANGING WATER -16210 -d162t VERT SEISMIC 0N WATER _62. -, 2.WATER ABOVE FLANGE _0. O VERT SEISMIC"bk"'_ATlk"" -*-; .NOR SEISMIC ON ALL WATER 2523. -2623.SHELL VERTICAL LOAD -437. -*37m O.OR SEISMIC ON SHELL 291. -12970 VERT SEISMIC ON SHELL 44, -449 CUMPRESSIBLE MATERIAL -85. -855 NOR SEISMIC ON C.M. 45e -45.VERTSEISMIt-ON C.N,, --- -g.PENETRATIONSi

-238- 23......-238.*

NOR SEISMIC ON PENE 54. -54.VERY SEISMIC ON PENE 240 -24o DEAD LOADS ....... -16"T. .-16]*r R SEISMIC ON D.L. 614, -614.RI SEISMIC ON D.L. 168.o ;1680 LIVE LOADS 0. 0.HOR SEISMIC ON LoL. 0. 0.VERT SEISMIC ON L.L. 0;. -0.STAY FORCE"TOT AL - 75.-MX-CTkCO#FEl EWRrl -ie-- .L 1B7IN " °ALLOW-. LC. L, -2. 0 -05,00. Le;ft.CHmI-C-',G TD-"U-, IRO'NI "AY CON- t' OK AEMIRNCER1N-FLO ELEV 74.S CONTRACT GPU/OeC. DATE 011S/8T BY7rJA SHT 02GREV 1 4Ka Pm7 I I[]I I 1 IUkMlTIo%

Of STIESSS°ATrE.

XPENT-SCVCAE-74- ETF OOEL fbELEVATION'

-14.S0 FT -NOTE STRESSES DUE TO HORIZONTAL

-'EKTI CAL: EARTHQUUAKE ARE

• Oft -.L.a EARTHQUAKE STRESSES YIELD A t5l MERIDIONAL STRESS RESULTANT*l 'I NERiDIUNAL

• SEISMIC -SEISMIC tLL5IIM LL17&NJ OVERHANGING WAlYER -3146. -3146.VERT SEISMIC ON WATER 315. -315.WATER ABOVE FLANGE 0. 0.VERT SEISMIC ON WATER 0.- 0., HOR UN ALL WATtR 4055. -4055.SHELL VERTICAL LOAD -578. -ST8.NOR SEISMIC ON SHELL 453. -453.VERT SEISMIC ON SHELL se. -56.oUMPRESS'BLE HAITiIAL -113. -.13.NOR SEISMIC ON CoM. TOo -70.VERT SEISMIC ON C... 11; .. -110 PENETRATIONS

-300. 3000 NOR SEISMIC ON PENE 940 -94o VERT :kL:HIL ON PFtNE See -.30LOADS -Z1130 ... -2113.O SEISMIC ON D.L. q9s5 -99s.VERT SEISMIC ON DoL- 211. -o1k.M-TVE LUAD$ go 0.NOR SEISMIC ON LeL. O. 0.VERT SEISMIC ON L.L. -" 0. 0.BUOYAMCY ... "L9 -.3rg.5TA&Y FMIGO 0 0 MAX CIRCUNFREINT1?I-r 34qa"-[B/IN" HICfGO iO- GL- OITLONFAMY

..""--If'7UJGINEEAZNG

"-'LO ELEV 74..5 CONTRACT GPU/O.C. DATE 01/15/87 6',' SKI 17REV I0/167 SUMMATlON*

C G -rBi- AlE FLOO-[EL 0 ELEvfArlGI0W

-.'5 fl -" NOTE STRESSES DUE TO HORIZONTAL 6 VERTICAL EARTHQUAKr ARE * #1 -.1+1 EARTHQUAKE STRESSES YIELD A M.l MERIDIONAL STRESS RESULTANT_ SEISMIC -SEISMIC TLEBIIN) LSIANJ OVERHANG 0G WATER -6470. -6410.VERT SEISMIC ON WATER_ __ _ -64T-WATER ABOVE FLANGE 0. 0.VERT SESMIC-N-iATEI

... ... .-. 0-;NOR SEISMIC ON ALL WATER 7664. -1664.SHELL VERTICAL LOAD -ass. -Iss.NOR SEISMIC ON SHELL 823. -823.VERT SEISMIC ON SHELL $6. -96.COMPRESSIBLE MATERIAL -168. -168."OR SEISMIC ON C.R. 129. -L29.VERT SEI'SM[C 0R C.M. fT.- -1i.PENETRATIONS

-424;'" -424.NOR SEISMIC ON PENE 188. -L88.VERT SEISRMI ON PERE 42. -42.qiID LOADS -z I"T. -29'1.SEISMIC ON D.L. 1886. -to8o.VERT SEISMIC ON D.L. 299. -299.LIVE LOADS 0. 0.HOR SEISMIC ON L.L. _0. 0.VERT SEISMIC ON L-L. 0. 0.STAY FORCE a...AX-1I 4FEI T,(U" ,,-.,-,_C 67 IN CfIIC'AG6o--0"E -CrT N COMPANY 'AXT-&DK ENOG"tEERING FLO ELEV 74.S CONTRACT GPUIO.C. DATE 01/15/81 By6¥/4 SKTD2bIREV I c.tAxO pm 21&167 SUmmATION OF STRESSES AT EMBEDMENT, CSI CASE f FLOODED TO ELEVATION NOTE STRESSES DUE TO HORIZONTAL

& VERTICAL EARTHQUAKE ARE

• OR -** (*) EARTHQUAKE STRESSES YIELD A (*I MERIOIONAL STRESS RESULTANT NERI0IONAL 74.50 F OVERHANGING WATER VERT SEISMIC ON WATER WATER ABOVE FLANGE VERT SEISMIC ON WATER NOR SEISMIC ON ALL WATER SHELL VERTICAL LOAD HOR SEISMIC ON SHELL VERT SEISMIC ON SHELL COMPRESSIBLE MATERIAL"OR SEISMIC ON Come VERT SEISMIC ON Com.PENETRATIONS HOR SEISMIC ON PENE VERT SEISMIC ON PENE DEAD LOADS HOR SEISMIC ON D*L, VERT SEISMIC ON O.L.LIVE LOADS HOR SEISMIC ON LeL, VERT SEISMIC ON L.L.*SEISMIC (LB/IN I 0.0.0.0.so-SEISMIC ILS/INI 0.0.0.0.-.5.-161.'7.16o-17.4.2.-206 7.2.-55.26.be-161.-97.-16.-17.-4.-20-20.-7o-2.-55.-28o-6.0.294a 00 0.0.BUOYANCY STAY FORCE TOTAL 294.00 2.57 LB/IN 2 ,6 7 R JqL e-Va 4 Z 9 -w .w 0.MAX CIRCUMFERENTIAL a at:-,,--~1

--rAG-TOR SF-SAFE*V e, we-CHICAGO BRIOGE E IRON COMPANY FLO ELEV 74.S CONTRACT OAK BROOK ENGINEERING GPU/0,C* DATE 01/15/67 BYrJA SHTOt1REV I cgLO.J9f//6/e7 SUMMATION OF STRESSES AT EMBEOMENT.

CBI CASES FLOOOED TO ELEVATION NOTE STRESSES DUE TO HORIZONTAL C VERTICAL EARTHQUAKE ARE

• 01 -O I*) EARTHQUAKE STRESSES YIELO A 1t. MERIDIONAL STRESS RESULTANT MERIDIONAL 74.50 F?OVERHANGING WATER VERT SEISMIC ON WATER WATER ABOVE FLANGE VERT SEISMIC ON WATER NOR SEISMIC ON ALL WATER SHELL VERTICAL LOAD NOR SEISMIC ON SHELL VERT SEISMIC ON SNELL C04PRESSIBLE MATERIAL NOR SEISMIC ON C.M*VERT SEISMIC ON C.Me PENETRATIONS NOR SEISMIC ON PENE* VERT SEISMIC ON PENE DEAD LOADS NOR SEISMIC ON DoL.VERT SEISMIC ON OL, LIVE LOADS NOR SEISMIC ON L.L*VERT SEISMIC ON LeL-* SEISMIC ILS/IIN 0.0.00 0.2?.-236.125.24a-24.7.-29.10.3.-65*37.be-SEISMIC ILB/INI 0.0.0.0.-27.-236.-125.-24.-24.-7.-2.-29.-10.-3o-6se-37.-6*0.0.09 309.Oo 0.0.309.BUOYANCY STAY FORCE TOTAL 0.19, ' " Oe-28(.MAX CIRCUMFPERENTIAL

-1.5/IN aLLo. CLefICNLt LOAID :;2956W l~L/IN C'FO nr v 4rduu CHICAGO BRIDGE E IMON COMPANY OAK BROOK ENGINEERING FLO ELEV 74.S CONTRACT GPU/O@C. DATE 01/15/87 BY'IIA SHTDSOREV 1 C.imw PM EZ/-4'7 SUMMATION OF STRESSES AT EMBEDMENT, CBI CASE( .FLOOOED TO ELEVATION NOTE STRESSES DUE TO HORIZONTAL

& VERTICAL EARTHQUAKE ARE

• OR -, f.1 EARTHQUAKE STRESSES YIELD A I.) MERIDIONAL STRESS RESULTANT peat I MERIDIONAL 74o50 F OVERHANGING WATER VERT SEISMIC ON WATER WATER ABOVE FLANGE VERT SEISMIC ON OATER NOR SEISMIC ON ALL WATER SHELL VERTICAL LOAD NOR SEISMIC ON SHELL VERT SEISMIC ON SHELL COMPRESSIBLE MATERIAL NOR SEISMIC ON Com.VERT SEISMIC ON CoMe PENETRATIONS NOR SEISMIC ON PENE VERT SEISMIC ON PENE* DEAD LOADS HOR SEISMIC ON DeL.VERT SEISMIC ON O*L.LIVE LOADS MOR SEISMIC ON L.L.VERT SEISMIC ON L*L.BUOYANCY STAY FORCE TOTAL* SEISMIC ILB/I 4 0.0.0.0.418.-346o 35.-34.90 3.-39.13.4.-87.10.90 Ole 0O 350 3°~0.-LB/IN v 60-SEISMIC ILB/INI 0.0.0.Do-48.-348.-165°-35.-9,-3*-39o-13.-'to-87.-4.e*-9.0.0.3Oo 350.0.-49Z MAX CIRCUMFERENTIAL a t~* ?. fl -CHICAGO BRIDGE & IRON COMPANY FLD ELEV 7495 CONTRACT GPU/OC. DATE OAK 8ROOK ENGINEERING 01/IS/87 By rT, SHTO5) REV I

\SUMMATION CF STRESSES AT EMBEDMENT, CBI CASES FLOOOED TO ELEVATION 7?.SO FT 0 NOTE STRESSES DUE TO H3RIZONTAL

& VERTICAL EART64OUAKE ARE

• OR -.I1) EARTHQUAKE STRESSES YIELD A I.) MERIOIOMAL STRESS RESULTANT POet44'%ER IDIONAL* SEISMIC ILB/INI-SEISMIC ILB/INI OVERHANGING WATER VERY SEISMIC ON WATER Oe 0.0.0.WATER ABOVE FLANGE VERT SEISMIC 04 WATEP 0.0.Do 0.mOR SEISMIC ON ALL WATER 49a SMELL VERTICAL LOAD NOR SEISMIC ON SOELL VERT SEISMIC ON SNELL-322o 132o 32.-322.-1329-32.COMPRESSIBLE MATERIAL NOR SEtSVIC O0. Come VERT SEISPIC 0N C-Me-32.to 3o-329-S*-3*PENETRATIONS MOR SEISMIC ON PENE VERT SEISMIC ON FENE-32o 110 3.-32.-3.DEAD LOJADS AOR SEISMIC ON. O*k.VERT SEISM4IC ON- O*L*-71.38.7.-?I*-38.-7.LIVE LOADS NOR SEISvIC.ON L.L.VERT SEISMIC ON LLe 00 0.0.O 0o BUOYANCY 336.336o STAY FORCE 0 3 a-4o4 TOTAL mAX CIRCUMFERENTIAL

• -- LB/INý&ý -9%ur-A-. m AA 6 i T .-t: Of !to[In~v' flu ---qro v CHICAGO BRIDGE C IRON CZMPANY FLO ELEV 74.5 CONTRACT GPU/0.C* 04TE OAK BROOK ENGINEERING 01/1S/67 BSV JA SHT13tREV 1

SUMMATION OF STRESSES AT EMBEDMENT, C31 CASE5 FLOODEO TO ELEVATI 7 0 74,50 FT* NOTE STRESSSS DUE TO HORIZONTAL C VERTICAL EART1QUAKE ARE

• OR -.14) EARTHQUAKE STRESSES YIELD A (.) "ERIOIONAL STRESS RESULTANT 4ERIDIONAL

• SEISMIC (LB/141-SEISMIC I L 3/ I '4 OVERHANGING 4ATER VERT SEISMIC ON WATER 0.0.0.Oe OATER ABOVE FLANGE VERT SEISMIC ON WATER G.0.00 O.HOR SEISMIC ON ALL WATER 150.-150.SHELL VERTICAL LOAD NOR SEISMIC ON SHELL VERT SEISMIC ON SHELL-179.68.18.-179,-68.-L.COMPRESSIBLE MATERIAL HOR SEISMIC ON CaMo VERT SEISMIC ON -28.6.3.-23.-be-3.PENETPATIOS NOR SEISMIC ON PENE* VERT SEISMIC ON PENE-180 ble 6.2.-S+3/'DEAD LOADS;OR SEISMIC ON ODLe VERT SEISMIC ON DoLo-122.28.12.-"122.-12.LIVE LOADS'OR SEISMIC ON L.L.VERT SEISMIC ON L.L.0O 0.0.3.0o BUOYANCY zqq.29q.STAY FORCE-167.167.a TOTAL 75.-170*.MAX CIPCUMFERENTIAL I 4460. LB/IN.AkLLO. 34Cr~Li?~

L3.I;70w' LoJ'Z Il-I- --$CHICAGO BRIDGE E IRON COMPANY FLD ELEV 74,5 CO NTRACT GPU/O.Co DATE 3AK BROOK EN'EERI'4 01/15/87 I C4"~ RA4 Z4/~

k I ski J SUMMATION OF STRESSES AT EMBEOMENT, CBI CASEf' FLOODED TO 6LEVATION NOTE STRESSES OUE TO HORIZONTAL E VERTICAL EARTHQUAKE ARE # OR to) EARTHQUAKE STRESSES VIELD A (.) MERIOIONAL STRESS qESULTANT MERI31ONAL 7'.5O FT OVERHANGING WATER VERT SEISMIC ON WATER wATER ABOVE FLANGE VERT SEISMIC ON WATER HOR SEISMIC ON ALL WATER SHELL VERTICAL LOAO HOR SEISMIC ON SHELL VERT SEISMIC O SHELL C3MPRESSIBLE MATERIAL HOR SEISMIC ON C,.M VERT SEISMIC ON CvM.PENETRATIONS HOR SEISMIC 10' PENE VERT SEISMIC 0% PENE OEAO LOADS F43R SEISMIC ON D.L.VERT SEISMIC ON D.L.LIVE L3AOS"OR SEISMIC ON L.L.VERT SEISMIC ON L.L.BUOYANCY STAY VORCE TOTAL* SEISMIC ILS/INI O.0.0.0.390.-165, 2.0 190-35.10.3.-31.d.3.-536.107.54a.Go Oo 00 0Q.-3299-153.7613. LB/IN SEISMIC 1L5/tN1 0.0.0.-390.-lase-82.-190-35.-3.-31o*-So-3.-536.-107.09.0.0.294a 321.-833.MAX CIRCUMFEREnTIAL a AL-.: ..... LA Mw-1 S q'llP 1~ 7L~CHICAGO eRIGGE L IRON COMPANY OAK 3ROOK E'GI'.SE:IG FLO ELEV 74.5 CGNTRACT GPU/OC* O.TE OL/15/87 BY'rJA SmtlD.4Ev t cF't.o PA 1 SUMMATION OF STRESSES AT C CASE-5" FC E TO ELEbiVA7rON' 74.50 FT NOTE TRESSES oU"10 HURILONTAL V&MrICAL E..ATiHQUAKE ARE * -, I.1 EARTHQUAKE STRESSES YIELD A 1,1 MERIDIONAL STRESS RESULTANT Pl*mt 7 NERIDIONAL

• SEISMIC -SEISMIC ILEIIUNS ILul.LNJ OVERHANGIlNG WATER -LIT, -177.VERT SEISMIC ON WATER 18 -180 WATER ABOVE FLANGE 0. O.l VERY SEISMI'. ON WAITR 0. 0.HNOR SISMIC UN ALL WAIEK 934. -934.SHELL VERTICAL LOAD -Z60,. ..6.NOR SEISMIC ON SHELL 134. -134.'VERT 5tISMI. L UN SHELL 26. -Z60 LUMPRIE51BLE MA'IRIAL '-0. * -.0.HOR SEISMIC ON C.M. No. -190 VERT SEIUM5L ON C.M, so ENTRAITIUNS HOR SEISMIC ON PENE 19. -190 VERT SEISMIC ON PENE 9.ORSEISMIC ON D.L. 226. -226oIt O -- 68. LIVE LOADS .. o.;HOR SEISMIC ON L.L. 0. 0.VERT SEISMI'. ON L.L. O. 0Go .....sU.YANc Y '297. '9T..siAY FUORE -TVZ. 2;1.TUOAL 307'. ate MAX IDlS eR L S/~4-d.1LL.

91 t -1 .-1 10a 'I CRHIA BRKIDGE 1"IRUN LUUMANY OAK BROOK TESINEEAING FLD ELEV 74.5 CONTRACT GPU/O.C. DATE 0115/St SHT" .5LV 1 cm9c pm 7 I m t .. ...SUMMATION OF STRESSES At EMUROJENT, U,, CASE, FLOODED TO ELEVATI-ON 74.50 FT NOTE STRESSES DUE TO HORIIONTAt" & VERTICAL EARTHQUAKE ARE

• OR -.* I.) EARTHQUAKE STRESSES YIELD A 143 MERIDIONAL STRESS RESULTANT MERID1 ONAL* SEISMIC -SEISMIC OVERHANGING WATER -1621. -1621.VERT SEISMIC ON WATER 162. -162.WATER ABOVE FLANGE 0. 0.VERT SEISAIC ON WATER O. 00 HOR SEISMIC ON ALL WATER 2523. -2523.-SHELL VERTICAL LOAD -437. -437o HOR SEISMIC ON SHELL 297. -2gT.VERT SElSMIC ON SHELL 44.COJMPRESSIBLE MATERIAL 8 MOR SEISMIC ON C.I. 45. -45.VERT SEISMIC ON C.R. go --.-I PENETRATIONS

-238. -238a HOR SEISMIC ON PENE 54. -54*VERT SEISNIG ON PENE Z4. -24.OSSMIC ON D*L, 614. -614.VERT SEISHICO' DLI-. T. -1668 LIVE LOADS 00 a.HOR SEISMIC ON L.L, 0 0.VIVEKT TEISMIC ON L.L. O.BUOYANCY 309. o09.STAY FORCE -Z211. 21.TOTAL -1977. -5523.MAX CIKCUKFERENTIAL

-"162-4Z.

LB/IN...O...... Q Kit.C b..JO -.CO-U'. LC./'.T C.g 4 rTUR 6f 5.,AUFFV -3. 6 ft CHICAGO BRIDGE --fRON COMPANY OAK AOouK--MG-RT-EERING.

FLO ELEV 74.5 CONTRACT GPU/O.C. DATE 01/IS/87 8YAI SHTD3£REV I.,cv PM Vf.l/67 Iz'I SU&MATION"OF 0 TRESSE5 AT E.- .1MENTt CKU CASES- U °O -A EL.EVA5'N-1T.50 Fl U-TO R A L VERTlIcAL EARTHQUAKE ARE " UK -.It) EARTHQUAKE STRESSES VIELD A I.1 MERIDIONAL srEss REsuLrTAr--MKIULURAL 4 SEISMIC -SEISMIC I L5/1I4J ILU/IINJ OVER NANGINZG WTER -3L46. -31466 VERT SEISMIC ON WATER 315. -3LS.WATER ABOVE FLANGE 0. 0.VER"T"$EI54 IC; ON wATER 0'MDR SEISUMIC U ALL WATlER 40i5. -4055e SHELL VERTTrAL LOAD ----. ,7U7 HOR SEISMIC ON SHELL 453. -453*VERT SEISMIC ON SHELL 58. "587Lt M L"-J.3 --113.HOi SEISMIC ON C.M. 70. -70.VWRKI SEISMIL UN L.m* 110 -Lao P

-"OO.HOR SEISMIC ON PENE 94. -g4*VtR1 SLIAIiIL UN PENE 30. 30 ESMIC ON Do.L. g995. -9So"LIVE .0.HOR SEISMIC ON L.Lo 0. 0.VER" 515MlG U LLo 4 .....BUOTYANLY

--319;-----iAY 1IUKLE -3583. 3583.' T]T'1'1-2Z "8580.-W ArCTMWRFEREN1IAL L9910. TW.?-~lLOku vU LftiKJ LiuA v LE' r CHICAGO B'--ME £ IRON COMPANY -OAK SI ENGINEERING FLO ELEV ?4.5 CONTRACT GPUIO.C. DATE 01/15187 SHT " REV I... .K JD ~0~~~

N I f SUKMATION OF STRESSES AT EMBEOMENTs CBI E5TFLODED TO ELESATION 74.50 FT NOTE STRESSES DUE TO HORIZONTAL

& VERTICAL EARTHAQUKE ARE

• O1 -.1+1 EARTHQUAKE StRESSES YIELD A 1*1 MERIDIONAL STRESS RESULTANT No11t I0 MERIDIONAL

-UVLRHANUINb VERT SEISMi WATER ABOVE VERT SEIS!NOR SEISMIC-SHELL VERTII HOR SEISMIC COMPRESSI BLI HOR SEISMI4 VERT SEISMI PENET RAT IONS HOR SEISMIC VERT SEISMI N A LOADS OR SE1SMIC VERT SEISMl+ SEISMIC ILULKNI WATER -6470.IC ON WATER 64.1 FLANGE 0.IC ON WATER 0.ON ALL WATER 7664a:AL LOAD -855.: ON SHELL 823.IC ON SHELL 66.-MATE'IAL -168.* ON C.M. 129.IC ON C." .17.-424s.ON P'ENE 188.It ON PENE z4.'-299L.ON O.L. leD8 I 4-oo -OL. a--SEISMIC ILB/INS-6 4 7if.-647.0.0o-1664.-823.

---

129.-LOS.-1868.-Zvq'2°I LIVE LOADS 0, 0.NOR SEISMIC ON L.L. 0. 0.VERT SEISMIC ON L.L. 0. 0.BUOYANCY 339. 339*3TAY FORCE -7394. 7394.TOTAL -6266. -141173.MAX CIRCUMFERENTIAL 26806. AR9 UUilEf-INGL LO1AD w- 21061-15 t'.jiN CHICIGO-BiDGE

& IRON EOMPANY HAK 8-00K ENc-hEERING FLD ELEV 74.S CONTRACT GPU/O.C. DATE 01/15/8? BY'ASHNTbadREV I.........

z/:,,,1 J Appeadtx R Comuter h03tTa Documetattio and VerifIcatioc lbdormation The following CBI Computer Programs have been used in the preparation of this report: 0 1. Program Mmber and Namet E0778A Analysis of Containment Vessels Description%

Program performs a primary membrane stress analysis of a containment vessel drywell. The drywall shell can be analyzed for any combination of 10 loading conditions, including earthquake.

Revision Identification:

Rev. I Dated January, 1974 Computer Type: IBM 4381 Nainframe Verifications Documentation data is on file at Oak Brook office.Inputs/Outputs:

Inputs and outputs are shown in Appendix 0 of this report and are on file at CBI's Oak Brook office.I-Uj 0 a i 0 Ok5.- 4C.,7 ....a"M "r' A' 14 s/ 1. 7 JJM e j A & -1 CO"K Ii ga s M.Onas%'s' 7ni~ 3~e~ Z' OATS 17U7/Z7S

2. Progran Number and Narme: 90781A General Shell of Revolution Stress Analysis Descriptiont Program calculates the stresses and displacements in thin-walled elastic shell of revolution, when subjected to static edge, surface, and/or temperature loads with arbitrary distribution over the surface of the shell. The geometry of the shell aust be symmetric but the shape of the meridian is arbitrary.

Since the program is based on classical shell theory, it has the same limitations.

Some of the features of this program are: The shell thickness, physical properties of the materials, and loading may all vary arbitrarily along the meridian.

The loading, including temperatures, may vary arbitrarily around the circumference (by using Fourier Series). There may be junctions or branches (maximum or three parts meeting at one point).0 A variety of forms are available in order to describe the shape of the shell (cylinders.

cones, spheres.torroids, ellipses, and parabolas).

Revision Identification:

Last Rev. June, 1982 Computer Type: IBM 4381 Klainframn Verification:

Documentation data is on file at Oak Brook office.Inputs/Outputal Inputs and outputs are partially shown in Appendices A, B and C and complete listings are'bn file at CBI's Oak Brook office.'" l Ay : C"Wo l ,,_ _ _ _ _ _ _ _ _ _ _ _ _ 7 A peI L ---4,

• 4.-3. Progrm Number awl KRam: 11443D 50SOZ4BPLL olZn BOSOR4 is a comprehensive computer program for the stress, stability, and vibration analysis of segmented, ring.stiffened branched shells of revolution.

The program includes nonlinear prestress effect* and is very general with respect to geometry of the meridian, shell wall design, edge conditions, and loading.However the vail must be thin enough so that thin shell theory is applicable and the materials must be elastic.Revision Identification:

Rev. 3/XX/85 and revised version Purchased 1/12/87 Computer Typet IIBM '381 Hainframe Verification:

Documentation data Oak Brook office.is on file at Inputs/outputs:

Inputs and outputs Oak Brook office.are an file at avs' 0-NUCLEAR ENERGY BUSINESS oPERATIONS GENERAL 0 ELECTRIC L F&PHT Transmittal No.87-178,003 I I I I I I I 1 FUEPLANWT hAUEIAIS TECMIOWG CORROSION EVALUATION 07 TIM Date: Ifarch Go 1987 9 Prepared Ayg Approved Byt P Plant Hater$.s' Performnce Fuel & tatials Tecbsolosy GENIERA. ELECTPIO-CWGIJI LE'aL WFOMATION NOTCE The only undertakings of General Electric Company (CE) respecting Information in this document are contained In the agreement between CMU Nuclear and 0E and nothing contained In this document shall be construed u changing the -agreement.

Mhe use of this Information for any purpose other than that for which It is Intended.

Is not authorized and with respect to any unauthorized use 92 makes no representation or warranty, (express or implied) with respect to the comnpleteness, accuracy.

or usefulness of the Information contained In this document s or that the use of such Information may not Infringe privately owned rights, nor does G1 assume any responsiblity for liability or damage of any kind which say result fro the use of of the Information contained In this document.31108706 .11"go 60? IneV 10/11)

FC&T Transmittal No. 87-178-003 lay. I

1.0 INTRODUCTION

CPU Nucleates Oyster Creek BVR is characterized by a Katk I containment as shown in Figure 1. During the 1980 refueling outage water was noted around penetrations at elevation 86'0s and running down the wall to floor elevation 75'3". Water was also observed at a penetration at elevation 47"0 and running down the wall to floor elevation 23h6h.1 The presence of water at these locations indicated that an intrusion of water into the annular space between the drywell &hell and concrete shield vall had occurred.

Water collection was also observed during this outage on the torus room floor which originated from the leak drafts in bays 3. 11 a" 13, as shown in Figure 2.Water on the torus room floor was also noted following construction.

Whem water saxples were witbdraw from the drafts In 1980 end were subsequently radiololically analyzed, the results indicated an activity level similar to primary a 1 I Itw" concluded at this time that the iobeble.sources o -rater were the- () equipment storage .poa, (2) reactor cavity, or (3) tuel pool. It was further concluded, that the leakage occurred only during refuelnlg when the reactor cavity, the equipmnt storage poo.l and the fuel pool- are floodedd When water was again'qoted leakin from the sand bed drains during the refueling outage in 1983, It ase decided that corrosion of the drywell shell could be a concern end an inspection would be performed during the next outage (1986). However, prior to discussing the details of the program, it is critical to examine the geometrical configuration and construction of the Oyster Creek drywall.2.*0 OYTER CREEK PlRAkY CONTADWNT GUO*ThRY The Oyster Creek KMrk I containment consists of a pressure suppression system with two large chambers as illustrated In Figure 3. The main chamber is 70' diameter spherical shall with a 33' diameter by 23' high cylindrical F&PNT Transmittal No. 87-178-003 Rev. 1 shell extending froe the top. The pressure absorption chamber is a shell in the shape of a 301 dt:ueter torus located below and around the base of the drywell. The two chambers are Interconnected by 10 vent pipes 6'46 In diameter equally spaced around the circumference of the pressure absorption chamber, figure 4.The drywell Interior to filled vith the concrete to an elevation 1003" to provide a floor. Concrete curbs follow the contour of the vessel up to elevation 12"3" with cutouts around the vent lines, Figure 5.The drywall exterior is encapsulated in concrete of varying thickness from the 'asi elevation up to the elevation of the top head FigurSe 3. From this juncture, the concrete continues vertically to the Iolw of the top of the spent fuel pool. The proximity of the concrete eurfaca to the shellwith elevation.

The concrete is in full contact with the shell over the bottom of the sphere at its invert elevation

.3" up to elevation 811.25". At that transition, the concrete is radially steapped back 15" to create a pocket which continues up to elevation 12"3 Figure 4. This pocket is filled with sand which creates a cushion to smooth the transition of the shell plate from a fully clasped condition between two concrete masses to a free standing condition.

The sand pocket is connected to drains to allowof any water which might enter the sand. It is within this sand cushion contact area&with the dryvell shell where corrosion was identifled.

Above elevation 12"3* the concrete Is radially stepped back 3 from the shell. This gap is created during constructIon by applying a compressible, Inelastic materiel to the outside of the shell prior to concrete Installation.

This material was later permanently compressed by controlled vessel expansion to create a gap between the vessel and the concrete F&W T ranauittia so. 87-178-003 Rev.2.1 Drywel Materials 2.1.1 !. .ry l Stee The drywoll shell Is fabricated of ASMf 22 L3 made to AST A-300 requirements.

This material ts basically a hIgh tensle (70 jel)carbon-s.iicon steel with a basic composition listed Ln Table I end to equivalent to SA-516 Grade 70 steel. gb mt .Wiside surface with an Inorganic ,inc (Carboline

'ft l 1iOS) and with -(Pb 3 0 4) 4 1 K.le4#tug vest 1 euti e ter.p of the us4frod~.t, t~lt'2.1.2 Sand Cushion The sand cushion was filled with *&an specified asASU 633 from.elevation 6'1l.25* to elevation 12h3". The "and was a natural sand, composed pri 311117 of ailica (810,2) With some' alumina, (A1203). Since the send was stored at a local dealer uncovered and *ezoed to the environaiest during storage and instsalltio 8.It to a safe asusmption that the saWno u placed into the sand cushion region In a moist or vet condition.

There is so Information concerning the method or condition of the backfilling of the sand into the sand cushion gap.Both GMV~ and 62 performed leachate studies on thea sand. Table 2 presents the GMU leachate results on various sand samples plus an insulstion sample which will be discussed in Section 1.1.3. The results of *the GPU2I analysis of the vat Say It anid indcates measurable quantities of Was Xt Cat Pbs, Hg end Cl which are naturally occurring Inu.and. The KS and Cl are also present in the insulation as wIll be discussed In Section 2...The Fe present is probably from any corrosion product incorporated Into the sand; the Pb to from the red lead primer.~1~~

FO&NT Transmittal No. 87-178-003 Rev. 1 The sand samples from core samples 19C and 13A (to be discussed in detall in Section 5.0) also indicate nominal values of contaminants with the exception of plug 19C which has a sigulficantly higher Fe content. Since this plug is characterlzed by considerable corrosion.

this result ts not unexpected.

Plug ISA, which Lnitilly was considered "pitted" but actually had inclusions with essentially no corrosion, has a significantly lover Fe content in the sand behind It.CMUN also performed an energy-diepersive x-ray analysis (CED or EDAX)with a scanning electron microscope (SIM) on sme sand and small pebbles as shown in figure 6. The EDAX spectrum confirms the presence of silica and alumina plus chloride.

The presence of C1 is consistent with the leachate analysls.The CE leachate ehemical analysis of the sand cushion specmlen (plus other samples to be discussed later) is presented In Table 3.2 Table 4 Ulsts the iconc constituents of leachate samples in units of milliequivalents per liter. Those values are calculated from the chomfcal analysis results using some assumptions of metal species. Note that this charge balance calculation sum of the anions differs from the sum of the cations by no more than 14Z for any of the test solutions.

This degree of agreement tends to verify the quality of the chemical analysis.

Flnally, Table 5 provides the checal analysis results of Table 3 expressed relative to the original samples by multiplying the concentrations reported for the leschates (in mg/L) by the leachate volume L) then dividing by the weight of the leached material (a).A comparison of the GE sand leachate analysis "torus sand", with the Table 5, GPUW analysis Bay 11 sand, Table 2, reveals similar results, (note ppm vs. ppb units). 'The major difference, albelt of little technical consequence, Is the level of Fe. The hay 11 sand contains 1.0-5.0 ppb of Fe while the GE analysis of the torus sand contained (0.04 ppb Fe. Since the sand sample may be from different locations the results are not significant.

The lead content In the Day 11 sand sample also appears to be higher.

F&NT Transmittal No. 87-178-003 Rev. I 2.1.3 Insulation At all elevations above the sand layer, the external concrete "ass Is set back from the surface at the Steel shell an amount calculated to allow unimpeded expansion of the shell during any design condition.

As noted In Section 2.0, this Sap was created by applying a compressible, Inelastic material to the exterior surface of the vessel prior to pouring concrete.

The material properties were selected to provide resistance to crushing by the pressure induced by the head of concrete but of low compressive strength to allow ccllapsfng by induced vessel expansion.

Design considerations.

necessitated that a Sap of 2' was required from elevation 12'3" to elevation 23'6" andta sap of 3' above 231.6.The criteria used to select the gap material "a as folloes: 1. TiSht adherence to curved, painted steel plate surfaces in horizontal end vertical positions.

1. Znsignificant deformation under fluid pressure of vet concrete estimated at 3 pal.3. Would be reduced In thickness inelastically by approximately one Inch from an initial thickness of 2 to 3 inches under a pressure of got more than 10 psi.4. Dimensionally stable at the reduced thickness without 8sinificant flaking or powderIng.

5. Unaffected by long term exposure to radiation and heat.6. Unaffected when exposed on the vessel prior to concrete Installation.

FUM Transmittal No. 87-178-003 Rev. 1 2.1.3.1 Duct Insulation The 2" Sap discussed above was formed by using 0vens-Corning Fiberglass SY Vapor Seal Duct Insulation and was applied to the vessel &hall from elevation 12'S" to 23'6". The material was applied as individual boards 2" thick with a factory applied laminated asphalt kraft paper (blih in sulfur and chlorine) water proof exterior face and was attached to the vessel with mstic and Insulation plns. The fiberglass strands were embeddd In phenol foraaldehyde.

Joints between the boards, edges and penetrations were sealed with glass fabric reinforced mastic.2.1.3.2 .irehba-D Insulation 2.1.3.2.1 ahckground The gap material used above elevation 23'6."was PIrabar-D, a proprietary asbestos fibar-magnesits ceant product applied as a spray coat.The manufacturer of this material was All Purpose Fireproofing Corporation.

The material was subsequently modified by Certified Industrial Products, Inc.to achieve a reduced density. The solid saterials, asbestos fibers, magnesite and maSneslum sulfate were premixed and combined In a mortar mixing machine with water and., to control density, with foan (aerosol P1, a protein, as a faming agent) tO fom a slurry suitable for spray application.

The first coat (W t.*Ahick) was standard density while the second and third coat (one inch thick each) was at a reduced density. After application and drying, the surface was faced with GrIffolyn (chloride content not known) 4 .l. clear polyethylene sheet with all edges sealed by tape and held in by Insulation pins. The polyethylene sheets formed the bond-breaker for the concrete pour.It is Isportant to note that the Flrebar-D is said to be 751 asbestos in magnesite.

To a geologist, aguemlite is the 1mieral. foa of sagnesium carbonate, KgC0 3.3 Complete calcination or dead burning of magnemite produces magnesium oxide, MSO. Commercially however, magnesite refers to "dead burnt" F&AP)O Transmittal No. 87-178-003 Rev. I maguesium oxychloride, also known as Sorel'& tenan This material ts produced by the exothermic action of a 20% solution of magnesium chloride*HgCl 2 , and a blend of magnesias, Mg, by calcining Ragnesits and magnesia obtained from brines 3 MgOWl 2+l1 520 -- 3 gO.Hgal 2-11 20.The resulting crystalline oxychaoride contributes the cementing action to the commercial cements. The product in hard and stron& but Is diaensionally unstable, lacks resistance to weathering, and most importantly in readily attacked by water which leaches out the )aCl 2 and thua Is highly corrosive.

2.1.3.2.2 Firabar-D Analysis Leachate analysis of the Flrebar-D was performe by both GMlN and GE.As shown In Table 2 for the 01WN analysis of this Insulation.

Flrabar-)consisted of a mixture of fiber, foam and concrete had high levels Xa, K, Ca, XSg, Cl, N03s 504 and total organic carbon (TOC). The R.a X and Ce are contained in asbestos.

The MS is also present In asbestos and if course the 71rebar-D.

The C1 e&a 804 are major compositional factors of the ?irebar-D.

The TOC of 1056 ppm Is mot likely due to the fosamg agent Aerosol PR which Is a protein. (This material along with the sulfate could serve as a food for any aicrobiologically Influenced corrosion (HIC). However,, this subject is beyond the scope of this report.] Although the source of the nitrate Is not obvious, It ts prevent in small quantitie In mesavaters.

The C& analysis of a 14 &ram Insulation leachate is shown In Tables 3-5. A comparison between the GPUM and GE results, Table 2 end 5o respectively, revealed similar resul. For example, 573 vs 610 ppb Clo 1936 vs. 1400 ppb Kg, 2850 vs. 2500 SO 4 , 132 vs. 130 N03', 1056 vs. 900 ppb total organic carbons, etc. vere observed for the GlUN and GE analysis, respectively.

Transmittal No. 87-178-003 Rev. I 3.0 POTENTIAL SOURCES OF VATER INTEUSIoN 3.1 Leakage Paths and History As noted in Section 1.0, leakage of water from the eand bed drain$were observed during the. 1980 and 1983 refueling outages. A series of investigations were performed by GP1UN to identify the source of the water and its leak path. Since the same range of radioactivity was found In this leakage water as Is found within the reactor, the leak path was believed to have been from the reactor cavity located lsmediately above the drywell. This cavity is filled with water during tefueling operations. "It was believed that a lak from this cavity through the bellows seal, Figures 7 and 8, at the bottom drained into the space between the drywall and the space filled with -Firebar-D.

Etanslve leak tests finally revealed that the most probable source of the water was the drain line gasket, Fig=* 8. This gasket was replaced and subsequent leak tests performed on the bellows revealed no additional leaks. Inspections of the areas previously characterized by.leakage indicated that the leakage had been arrested.Nowever. this history of leakage, which may have initilted at the first refueling outage plus any condensation in the gap between the FIrebar-D and the drywall shall, means that the Firebar-D could have been intermittently wetted and leachad of corrosive K'Cl 2 which collected In the sand cushion. .-As will be discussed later, the establishment of this electrolyte in the sand is considered the key factor in the drywe;l corrosion phenomenon.

3.2 Leakage Utter Analysis During the 1986 Oyster Creek refueling outage, water samples were obtained from a drain line and analyzed by GPWI and GE. In addition to tritium, these samples were analyzed for contaminants.

The results of the GTUN analysis is shown in Table 6. Significant amounts of Ns, K, MS, Cl and SO 4 are present. The sources of these substances F&M Transmittal No. 87-178-003 ROev. 1 Include the natural substances found In sand, a marine environment and the Firebar-D.

The conductivity is high (680-1100 uS/c. as cempared to 0.2, 1. 70 and 1000 vSlca for good reactor water, 0ood quality distilled water, excellent quality rev water, and 0.03Z MaCi soluti4o, respectively) and thus clearly Indicates that the drain line water would serve as a suitable electrolyte for corrosion.

The results of the GE analysis, Tabl 3p of the leksI p water sain reveals isiasr results for elements K, Na. Co, Use Alp Clg, 80,4. ToC adm conductivity.

The only measurable differences between the two analyses Was in the ?a and Sr.3.3 Deposit.Analysis Various scraplngs from horizontcal and vertical surfaces were obtained between the toru* and drywall at Oyster Creek. Those deposits wee analyzed by couple plasma by GFPr. Table 7, and by EDAX and leachate by GZ, Tables 8 end 3, respectively, Table 7 reveals the presence of various metal oxides with 702035 heutite (rust-) being the dominant material for material removed by Bay 7 and 11. The only unusual oxide is the 3103 which Suggests the presence of reactor water. The GE analysis of a scraping from Bay 7 only as Investigated by EDA]2 Table 8, revealed high percentages of Fes C19 K, S and Pb. The results of the 4E analysis is fairly consistent with GW's investigation slthough different analytical methods were utLlized.

The quantity of Ie Is consistent and anticipated.

The presence of 1ed, Pb, is consistent with the red ead prime* coat. Manganess say be due to the presence of manganese sulfides in the steel. Although the existence of Be, Cl, end K are consistent with the presence of Firabar-D, the presence of these three elements plus bromine is suggestle of a marine environment sice Ir is also found in savater.The results of the leachate analysis of the Bay 7 deposit is presented in Table 5. The results are consistent with the other analytical results from F&T No. 87-178-003 Rev. I this sample in that significant amounts of S0 4 , Cl, Co. K and Ka leached out of the specimen.

Pb, 8, Sr, Ba and Al were also Identified.

9ith the exception of Pb. all of thesn elements are present In *e&water.4.0 DYWELL THICKNS Motivated by the presence o wer fn the draln Ifes. ar oathar penetrationso, MP performd extensive ulttesoni thikueks me"e'sements of the zywel'-0e Aetri if 4egradatIm "a occurrus.

4ppiimiately lop0 Ulrasomld testIns (UT) measurements wers obtaned, throaghl use of An ultrasonle thickness gun. device C(D-etsr).

The 0-.etat uqsures the Wlba for a longitudinal ultrasound wave to travel to a refloction Cbackwall or aidvall reflector) and back. Zxpanded UT examinations were accomplhed through the use of a "A-Scan" UT technique where the character, location and amplitude of various ultrasound reflectors are displayed an a cathode ray tube.The initial UT measuresents (D-ueter) vere made from the Inside of the 64wiam at eleVations.

of wl and the Ml!, sand,_cushion, fipft 9. 21wk.4 cushiom asesurate obte~4.a the. baya to known water eAki ieated, t "hatwtinnb hadW occurN. Measurements just above these area in the same plate and at the 51' elevation Indicate nomisal plate thicknesses.

Measurements were obtained with both the Inside surface coating of Carbo zinc 11 In place and removed.As a result of the Initial low thickness readings, additional thickness measurements were obtained as described In detail elsewhere.I To determine the vertical profile of the tbinning, a trench was excavated Into the concrete floor In Bay 17 and Bay S. Bay 17 was chosen since the extent of thinning at the floor level was the most severe. The additional thickness measurements indicated that Whaning below the inItial measurements were no more severe and become less severe at the lover portions of the sand cushion.Bay 5 was selected to determine if the thinning line was lover than the floor-I ~

F&PMT Transmittal87-178-003 Rev. I level in areas where no thinning was identified.

No significant Indication of thinning was found In the sub floor region of Bay S. Aside from UT thickness measurements performed by the CPUN staff, independent analysis wes performed by the EPRI NDZ Center and the GE Ultra Image III *CO scan topographical mpping system. the EMI! Investigation verified Gfl's thickness and mid-wall reflector 7 results and the GI mapping confirmed a corrosion transition at seven to eIght inches up from the concrete curb in Bay 19. The Ultra Image results will be used an a Easeline profile to track continued wstas..GPON also used a UT Integration method (30-70-70 technique) to detect minor changes In back vall surface conditions.

This technique ma able to verify the roughness condition of wastagS and the Light corrosion areas of the containment well as compared to reference standards.

Finally$, UT Investigations of various plate to plat* welds and heat affected Zones revealed no Indications of wastage or cracking.1 5.0 CORE S#JIPLINO To evaluate if the UT measurements were valid, characterize the form of damage, and determine the cause (i.e., due to the presence of contaminants, microbloloSical species, or both), *It was considered prudent to obtain core manplos from various bay locations.

Areas that were characterlsed by sharp deviations In thickness of less than half the 1.154' nominal wall were designated "pitted/incluslon" areas. Legions that had UT indications of thinning were designated as "wastage" areas. Regions above the wastage area.and within the sand cushion region that appeared to have no thinning or"pLttin" were also selected as candidate core sample sites. Table 9 snumarizes the UT characterizations by bay nmber.Core samples of the dryvell. wall rer obtained at seven locations.

To produce an adequate sample size, an opening large enough to allow removal of sand samples and insertion of a miniature video camera and allow a simple plug design, the sample diameter was optimized at 2" In diameter.

Table 10'I Transnittal No. 87-178-003 Rlv. I sunmarizes the seven core sample locations, the type of samples obtained and the organization who performed the subsequent analysis.The core uamples were cut in such a manner to eliminate any possible contamination from the cuttilg operation.

Distilled water was used during the Initial cutting operation as a coolant. The final cut through the wall was performed vithout coolant and the shell temperature was maintained below I450!F to prevent the premature death of any viable amcrorganula.

Biological samples ware taken from four plugs and analyzed by another party for GPUX.The next five sections present the results of the core sample analysis.5.1 Core SaMle 15A -KI.nmus Thickness Speimen -GPWI Core sample 15A which was removed from Bay 13 %as the key specimen for detailed analysis.

Thls particular ea was characterized by the lowest through-wall thickness (0.490") as observed randomly by UT examination, surrounded by adjacent areas with nominal thickness of 1.17". Therefore, the question was whether this area was suffering from some sort of locallzed tpitting" attack ot did the plate in this location contain Inclusions.

The removal of plug IS Immediately revealed that there was no pitting or in fact any serious corrosion attack. The sample measured 1.17" average thickness and was covered with a uniform dark brown (magnetic) .Elemental analysis of thi oxide by EM indicated that Fe was the major (.10 W/o) constltuent, followed by Pb (3,1 V/o) from the red lead primer and traces (01 w/o) of Al, Si, Ca, Cl, K, S and Mn, Table 11. Pigures 10 and 11 present overall crpss-sectIon viev of plug ISA and detail region where ElAX was performed, respectively.

Figure 12 presents the energy dispersion line profile of plug 13A which clearly reveals a constant low level distribution of Cl and a high level concentration of Fe In the scale. ElMX analysis of a sand sample from plug ISA revealed that Sl vas the =ajor constituent (210 w/o) with minor amounts of Al and Fe (21 w/o) and trace amounts of Cl, K, Ib and Ti (e1 V/a), Table 12.~,

FU&ri Transnittal No. 87-178-003 Rev. 1 GPUN also prepared metallographic speclmens from this core plus in both the rolling direction and perpendicular to the rolling direction, Figure 13. As shown In Figure 10 and 11 , minor pitting (C5 ilU) varn observed on the surface. The amd-plane of the specimen vas characterized by a band of alu-Inide stringers, Figures 14 and 15. These Inclusions are sufficiently dense to produce a reflection for ultrasound.

In fact, the measured depth of these Inclusions correlated with the depth determined by the initial UT examination.

The validity of the overall UT thickness measurements vns also confirmed by actual thickness measurements.

The ability of theWA-scan to Identify areas of inclusions, as opposed to pits, was also confirmed.

5.2 Core Sample 19C -Wastage -GPU(Vhan core sample 19C was cut, GlPU noticed that a hard black crust remained in the hole at the sand Interface.

The crust was approximately 0.3" thick and was subsequently removed for analysis.

Other wastage sample@ were also characterized by this corrosion product crust.Figure 16 presents an overall view of plus 19C. The surface has the classic appearance of general corrosion.

The measured thickness was-approximately 0.825" which corresponds with the UT determination of O.815".The surface was covered vith a thick black powder deposit vhich varied In thickness up to %30 mle. EDAX analysis of the surfaces Table 11, revealed that again Fe was the major constituent (310 V/o) as was the came of plug 15A.However. for this wastage sample the minor element 01 W.o) was C1 and not Pb.Trace amount of Al. S1 and Wn were also identified, Figure 17. A cross-section of plug 19C, Figure 18, prepared through one of the. valleys on the corroded surface reveals the corrosion prodict. An EDAX analysis along the indicated profile location, Figure 19, reveals a chloride peak in a 2 anl thick region adjacent to the steel surface. ilAX analysis of the magnetic crust/flake deposit removed with plug 19C revealed that the primary con-atituent was Fe (010 W/o) with only trace amounts of Si and C1 (<1 W/o), Table 13. The pH, as determined by litmus paper, of the scale was measured at 4.

F&PHT Transmittal No. 87-178-003 Rev. 1 Metallographic ezamination of the plug sample 19C also revealed that scale contained manganese-sulflda inclusions, Figuro 20, Manganese Inclusions were also found beneath the surface of the plug, F1guro 21. These two figureu clearly Indicate that the wastage Is proceeding through the vall and io capable of retaining Inert materials In the original position and orientation.

This result also explains the presence of Mn In the EfAX analysis presented in Figure 17.5.3 Core S1aple 17D -Wastaue -GE This plus sample was also ciiracterized by general corrosion/wastage.

The pre-removal UT thickness was determined to be approximately 0.440%. Upon removal actual thickness measureuent revealed an average thickness of 1%.860': SEN ezamination of the surface of plug 17b, Figure 12, revealed a fairly uniform distribution of oxide particles, Ana EX analysis of this surface revealed a high concentration of C2 (3.71 -4.92 V'1) and ft (92.73 -94.60 V/o). Table 14. Similar results were obtained for an analysis.

Figure 23, of the cross-section of the oxide, Table 15, where 3.455 W/a Cla 94.40 WIo Fe was Identified.

This analysis confirms the GPVN studies on wastage asmple 19C where a high chloride peak was associated with the significant general corrosion attack.The corrosion product crust removed from plug 17D was analyzed by both EDAX and z-ray diffraction (XID). The EDAX analysis of the crust reveals that Fe is present In the highest concentrations (88.32 -98.26 7/o), Table 16, followed by Mn (1.54 -10.42 V/0.) Si (0.00 -0.63) and Cl (200 -3800 ppm).XMW) analysis of this dark brown to black crust was performed on magnetically separated material ai discussed balbv.The results of the KID revealed that the non-magnetic aliquot was composed of major amounts of alpha quartz (a -Si02) With sall amounts of face-centered-cublc (FCC) 11304 spinel type phase plus trace amounts ('2 W/o)of an unidentified phase.5 The magnetic aliquot composition was essentially F&PWT Transmittal No. 87-178-003 Rev. 1 Just the opposite of the non-maguetlc sample, that is, the magnetic aliquot consisted of a major phase (390 V/a) of FCC 1304 spinal with smail to trace amounts of 4 SI0 2.The lattice parameter value for the spinal phases was determined to be so a 8.387

• 0.004 A. This value Is clome to the lattice parameter of stolchiometric Fe304 at ao " 8.3963 A. The slightly sall.er measured lattice parameter of this oagnatic phses can be moat likely attributed to small amounts of other transition elements in substitutional solid solution with the major element Fe. It should also be noted that the error on the lattice parameter could Indicate no cag in coarmition has occurred and the spinel phase could be pure h 3 0 4.Other compounds such "s FeCd 2 , Fe*C 3 , a Fe 2 0 3 and y F1*03 were specifically analyed for in the samplep but none were identified with the possible exception of a weak trace of a Feah3. The detection limit for theac types of phase In this type of material is estimated to be we to two weight percent.Metalloppahic euamination of plug 17D revealed similar corrosion product buildup as seen on wastage plug 19C. an seen by comparing Figure 18 vith Figures 24 And 25. The microstructure of the steal$ Figure 24& and the hardness values N 81-84) were typical for this type of. steel.The leachate analysis of the sand behind plug 17D revealed significantly less contaminants than observed for plugs 19C ad ISA. The only contaminant present in significant quantities Is 19 pMa K, 9 ppm WA and 4 ppm Ca. The chloride content In this sand, 1.8 ppm, Is significantly less than observed in the sand behind plugs 19C and ISA at 45 and 93 ppm, respectively.

It should be noted, hoverer, that the plug core sand samples were received in plastic jars and not as a core sample per se. Therefore.

any higher concentration of contaminate adjacent at the plug/sand interface could have been diluted by ixaing.

FtPUf Transmittal No. 87-178-003 Rev. I 5.4 Core Sample 19A -uastage -GE This sample was the second wastage sample received by Gl for analysis.The pre-removal.

UT thickness measured by GPUN averaged 0.830w. The post-removal average thickness vas 0.847%.SEX examination of the surface of plug IM Figure 26, revasled a surface which is quite different that that observed an the previous wastage sample plug 17D. Only a very fine powder deposlt 1. observed op this surface.An ZWAX analysis of this surface revealed primarily to (96.07 -97.45 V 1 a)with lower amounts of Cl (0.34 -1.25 '%.) than plug 170, Table 14. The lower chloride content could explain the difference In surface sorphology.

The cross-section anslysais Figure 27, of plug 19AI Table 15p revealed the absence of many of the alements observed In plug 17D. Again Fe dowmntes the analysis (98.37 V/a) followed by Mn (1.24 W/o). The source of 3a Is moat likely the nanganese-sulfida inclusions in the steel.The corrosion product crust removed from plug 19A was also analyzed by EDAX and 3DD, Table 16. In this case, plug 19A crust was characterlsed by Fe (64.69 -93.36 V/)), $i 3.81 -30.34 V/), Mk (up to 1.5O w/o), Ti Cup to 2.98 w/o) and CM (3300 to 19,300 ppm). The M30 analysis revealed a non-magnetic and magnetic phase compositions that are nearly identical to that obtained on plug Me .3 The ony difference found was that the lattice parameter for the M 3 0 4 spinal was %° a 8.396 a 0.003 for plug 19A which is exactly the value for stoichlometri 1e 3 0 4.Am was also the case of the crust from plug 17p, no FelC 2 , rFed 3 , a-F*203 or y-Fe 2 0 3 were identified In any measurable amounts.)etallographic ex-amiation of wastage plug 19A, Figures 28 and 29, revealed'similar results as observed on plug 17M, that Is, thick corrosion product on the surface, normal elcrostructure and hardness values (% 80-81),

FUNT Transmittal No. 87-178-003 Rev. I The leachate analysis of the sand behind plug 19A reveals aimltar results to that obtained behind plug 17D, Table 5, but again different results as compared to the two CPUN analyzed pluSe.5.5 Core Sample lA, -i -Above Vast&e& -G2 This core sample was removed from above the wastage region of the drywall but still in a region In contact with the sand cushion. The thickness measured by UT was 1.170". After removal of the plug, the thiekess measuroxent measured at the center of the plus was 1.19'. Thus there was essentially no corrosion on this specimen.SIr eamination of the surface of-plug IA-3 (E -high. I.e.. above-sample, il). Figure 30, revealed a surface with Isolated islands of deposits.Higher magnification examinstion (1500X) revealed the presence of a uno-uniform powdery scale. The EDAX analysis Indicated the ?b (32.61 -59.77 w/o) and not Is (21.72 -28.87) dominate the surface, Table 14. indlcates that the red lead paint (Pb 3 0 4) was still present on the surface.This result is anticipated since this plug suffered no corrosion.

It Is important to note that the Pb Is present because no corrosion occurred at this area and not that the red lead Inhibited the. corrosion.

The high presence of S in both the surface analysis and cross-sectlon analysis (Figure 31 and Table 15) may be due to the affinity of sulfate to combine with the red lead paint to produce 1Pb0 4.The sulfate may be a leachate from the Firebar-D or from the marine environment.

Significant amounts of chloride are also present.Since plug lUA-R had essentially no corrosion, there was no crust to analyze by EDAX oi XRD.Metalo$raphlc examination of plus 11A-H, Figures 32, 33 and 34.revealed the absence of severe corrosion.

There was only mild attack observed at higher agnifications (125X) on the cross-section of the plug. Figure 33.Hardness measurements again revealed nominal values (R, 80-81).

F&MPH Transmittal No. 87-178-003 Rav. I The iachate analysis for the sand behind low-corrosion plug 11A-H reveals some Interesting differences as compared to plugs 17D and 19A. For example, despite the fact that this sand was characterized by an order of magnitude higher chloride (26 vs. 2 ppm), sulfate (40 vs. 4 ppm) and magnesium (16 vs. &3 ppa) content, this plug had essentially no corrosion.

This result is consistent with the GPUN results for plug@ 19C and 15A said, where no-corrosion plug 15A was characterized by higher chloride, sulfate and nagaesium In the sand. The key difference In corrosion response lies not with the relative contamiatlon levels in the sand, but rather the piesture content. As Is shown In Table 5, the sand sample behind plug 11k-I[ ws dry as opposed to 1.1 -2.61 moisture for plugs 17D and 12A, respectively.

the absolute difference in the contanination levels of the sand are significant an a percentage basis, but not on a corrosion basis. The key here is the absence of an electrolyte.

6.0 DISCUSSION The results presented In the previous sections on the analysis of.various sand, plug, deposit and water samples Indicate that a .ustable environment for the corrosion of carbon steel Is present In the sand cushion ares, In other words# the corrosion of the drywall as exposed to this particular environment could not be considered unexpected.

The question is whether the amount of corrosion is particularly high and what role did other factors such as the lirobar-D Insulatlon, contsaliants, differential aeration, red lead prlmer, or concrete play in the corrosion phenomeno.

6.1 Cenetral Factors Afecting the Corrosivi!t of the Send Cushmen There are numerous factors which would affect the corrosivity of the sand cushion relative to the carbon steel drywall. These factors include the sand porosity, electrolyte conductivity, contaminant level, moisture level, acldlty/alkalinity and the presence of bacteria.

FUNT Transmittal No. 87-178-003 Rev. I The relative porosity of the sand cushion would be affected by the method of beck-filling the sand Into the sand cushion region construction, the settlinS'of the sand, the Initial moisture content of the sand, whether it was subsequently vetted after Installation, etc. The porosity of the sand vould affect the moisture that could be retained In the sand cushion and the establishment of local areas of high aeration.

The more porous the sand the more moisture would be present over An extended period of time and the more optimm the degree of aeration.

Both of these factors would tend to increase the Initial corrosion rate. The degree of aeration of the sand would also affect the type of corrosion products formed on the eteel surface.For example, studies by Roonm-ff6 have indicated that in ell2-aerated soils the rate of pitting/corrouion, although high, falls off rapidly with time because in the presence of an abundant supply of oxygen.oxidation and precipitation of iron as ferric hydroxide

[fe(CH)3]

occurs close to the metal surface, and the protective membrane formed in this manner decreases the subsequent corrosion rats. A noted on the plug specimens from the Oyster Creek dryvell only shallov pitting was observed on eaom of the specimens.

In poorly arated regions, tomanoff noted that the Initial rate of corrosion decreases slowly, if at all, with tine. Under such conditions the corrosion products, remainW In the deovidized state, tend to diffuse outward into the soil, offering lttle or no protection to the corroding metal. (Te actual corrosion products observed on the drywall will be discussed in nore detail below.)The role of conductivity of the sand cushion is more straight forward.The hiSher the conductivity, the greater amount of corrosion would be anticipated.

The conductivity of water samples removed from various drain lines at Oyster Creek ranged from 680 to 1100 uS/cm. The conductivity of pure*water at similar temperatures is three orders of magnitude lower than these values. B ence, the sand/water environment was sufficiently conductive to establish a viable electrolyte for corrosion. ________

F&IMT Transmittal No. 87-178-003 ley. I As noted in Tables 2 through 6, the sand, scrapings and drain water had high levels of contaminants which would be expected to Increase the corrosion rate of carbon steel. In particular.

high levels of detrimental chloride and sulfate were noted in virtually all the analyzed.The mare fact that corrosion occurred at Oyster Creek indicates that moisture was present In the sand cushion. As discussed In Section 3.0, the sources of moilture include a known leakage of water from' fuel pool which most likely occurred through a drain lne gasket, Installation of moist sand during construction, water *squeezed" out of the FLrebar-D slurry durlng pressure testing of the drywll and 'condensation.

The moisture content of the sand samples as measured by GE ranged from 1.1 to 12.61. The only dry sand sample was from plug I1A1-, which did not suffer any seigicant corrosion.

High pR Is beneficial for the resistance of iron base'-alloys. The p1 observed from sand end drain wator samples ranged from 5.99 to 8.90, Table 2, 3.and 6. Most of the pM values were somewhat greater than neutral pH 7. However# average pH values alons ca be misleading*.

A will be discussed later, the establishtent of local anode and cathodic altos due to differential aeration will affect the local p1 values. Deserated anodic regions will have a lower pl while the cathodic regions will have a higher pH.Also sections of the drywall adjacent to the concrete would benefit from the high alkalinity of concrete.Corrosion induced by alcrobes is a widely recoSnized phenomenon in a number of systems such as oil wells, pipe lines and municipal savage.Hicrobiologlcally.influenced corrosion (CIC) has also been identified in nuclear power plants. However, the role of IdC in this particular system is being Independently investigated and Is therefore beyond the scope of this report. It should be noted that prelimliary evidence presented during discussions of the drywall corrosion at Oyster Creek have indicated that the role of MIC, if any, is not considered to be signlflcatt. _____

_

F&PUT Transmittal No. 87-178-003 Rev. 1 6.2 Specific Influences on Oyster Creek Corrosion 6.2.1 Firebar-D Due to the known high corroslvity of Firebar-D on steel,34 one of the primary motivations In the Investigation of the corrosion of the OCyter Creek dz2vell was the detearafttion of the role of Firebar-D on the corrosion echanisam.

An noted in Section 2.1.3.2. Firebar-D Is composed pf MSO, NSC12 and vater. StudLes by Biliuski, et al? have revealed that swuf (E)1* .C 2.8H 2 0 Is the predominant reaction product In m cally-sound hardened magnesium ozychlorlde cement. This material Is extremely sensitive toto water @Ince there is an extremely narrow concentration range of magnesium end uhlorLde ions In solution In which 5Mg (OH)2 20 3jgl

• 8U 2 0 1e stable. It Is the presence of leachable 2 which can produce severe corrosion problems.The specific corrosIvIty of magnesuim oxychlorlde cements has been investigated by Revaller.

Observation@

of steel exposed to direct contact with damp manesiumn ozychlorlde reveals a distinctive dark black rust (y-Fe 2 0 3)1 typical of *corrosion which occurs In either a law oxygen or a caustic envIroment.

D Minvestigations by CZ specifically designed to determine the presence of f-I*203 were negative.An analysis of the chemical structure by avalUer revealed that when magnesium ozychlorlde Is exposed to 100Z humidity, leaching of surplus magnesium chloride resulta In the formation of magnesium hydroxide.

Carbon dioxide extracted from the atmosphere ecmbines .with this material to form a surface layer of magnesium chlorocarbonate

[N(COf)i MdC12 -2gC03 @ 6'H 2 01.This surface layer slow& the leaching process. As addltlonal MrC1 2 'a leached, a surface crust of hydromagneuite C5MgC0 2

• 4WC 2 ' S2O) to foamed.These insoluble carbonates and hydroanguesites help to Improve the weathering stability of magnesium oxychloride materials.

F&M Transmittal No. 87-178-003 Rev. 1 The lack of Y-Fe 2 0 3 In the oxide an the core plus surface/crust, the relatively low amount of NS in the sand samples and the absence of corrosion at the 51' elevation level suggests that the role of Firebar-D In the degradation of the Oyster Creek drywell corrosion phenosena In not significant.

The levels of chloride and .aguaitu identifLed In the various laboratory samples may be as such the result of the marine environment as the leaching of the Firebar-D.

The formation of the Insoluble carbonates and hydrooagnesites discussed above may have reduced any potential FontributIon of Firebar-D to the corrosion reaction.6.2.2 -Contaminants Table 17 presents the typical constituents of seawater.9 A tomparison of Table 17 and the results of leachate analyses, Tables 2 and 3, the drain water analysie.

Table 6, ond the deposit analyses.

Tables 7 and 8. reveal that many of the contaminants observed in these analyses could be from the Oyster Creck marina environment.

In particular, the presence of gal, Al, Xr, B,- Ca and Sr can be ezplaIned.

ffowevor, the boron and strontita may be from the fuel pool as discussed in Section 3.0.The primary role of any of the tons in the corrosion of the Oyster Creek drywall would be the enhancement of the electrolyte, that is, an Increase in the conductivity.

A secondary role for these Lons. and In particular, chloride and sulfate, would be the breakdown of any passive film established on the carbon steel surface. As seen In Figure 19, hiher concentrations of chloride are observed at the drywefl va.ll-ozide layer Interface.

The presence of the higher chloride at this Interface may be the result of the alternate vettinS and drying of the sand cushion.

of the source of the conttanxtiaon, that is, the marina enviroment and/or the Firebar-D.

the presence of such known "bad actors" as chloride and sulfate will Increase the corrosion rate of the drywall.~, 1~

F6P)T Transmittal No. 87-178-003 Rev. 1 6.2.3 Differential Aeration In most systems which are In contact with atmospheric oZygen, geometrical situations arise where transport of oxygen through the solution by convection (natural or forced) and diffusion to one part of the metal occurs rapidly, whereas it is slow or even negligible at another. The areas characterized by high oxygen will serve as cathodes where the reduction of oxygen to hydroxyl occurst o2 + 2o +4e g--4On Areas depleted In oxygen w= become anodic with the corrosion of the carbon steel: S-e re 2÷ + 2e" Therefore, areas of the sand cushion adjacent to ready oxygen access such as lower regions near the drain line and upper regions near the Insulation Sap would become cathodic while areas In the middle of the sand cushion would become anodic. UT measurements appear to verify this topographical evaluation.

Also, differences in Icca concentrations of XtCI may result In differences in oxygen concentration a tssgested by Schaschl.

and Harsh.1 0 The higher the salt concentration the lower the solubility of oxygen so that these depleted zones become the anodic zones of the differential aeration cell.6.2.4 Role of the Red Lead Primer The outside of the dryell was painted with red lead which is lead oxide, Pb 3 0 4 , or more accurately Pb 2 PbO., in linseed oil* Water reaching the surface dissolves a certain amount of p1pent and makes the water less*corrosive." In general, corrosion Inhibiting pigments must be soluble enough to supply the minimum concentration of Inhibiting Ions necessary to reduce the corrosion rate, yet not so soluble that the are soon leached out of the paint.

FISfPT Transmittal No. 87-178-003 Rev. I The inhibiting Ion for red lead is probably PbOJ'-4 vhich can passivate steel. Noverer, In the presence of SO 4 or C02 the passivatiug effects of red lead can rapidly disappear.

Sulfate was Identified in many of the analyses and carbon dioxide is readily available In the atmosphere.

It was noted throughout the analysis of the removed core plugs that lead was found on the surfaces of the plugs that suffered m alm or essentially no corrosion.

It is strongly believed that lead was found on such samples because no corrosion occurred due to the lack of *oIeture (dry sand)and not due to corrosion inhibition of the red lead paint. led lead paint alone simplydoes not provide long tars corrosion protection.

6.2.5 Role of Adjacent Concrete Concrete provides an alkaline environment and, under moist conditions, passlvates Iron and steel, lesions of the sand cushion/dryvell adjacent to the concrete could be benefited by this local alkaline environmont.

This factor can explain why the lower regions of the drywall below the 8'11.25'elevation which are In direct contact with the concrete did not suffer any Smeasurable corrosion.

Since part of the drywell is In contact with the passivsting concrete and part of the drywoll in contact with the moist-high conductivity sand l.macro-salvanic cell is established.

This will result in the acceleration of the corrosion of the dryvell in contact with the vet sand cushion. As will be discussed In Section 6.3. the presence of chloride in the stnd will only amplify this effect.6.3 Relevant Corrosion Reactions It Is considered prudent to briefly ezaalne the possible corrosion reactions which are occurring on the surface of the dryvell embedded in moist sand. Zron (steel) ions will go into solution at anodic areas in an amount electrochemically equivalent to the reaction at the cathodic areas. As noted-2.4-F&PU Transmittal No. 87-178-003 Rev. 1 earlier, the anodic areas of the dryvell are characterized by the following basic oxidation reaction: Fe --) rf" + 2e-The relevant cathodic reaction In aerated solutions is the reduction of oxygen to hydroxyl ionsa 02+ + R2o + 4 -4or However$ the corrosion of iron or steel is not as straight forward as illustrated above. An shown In figure 33, numerous corrosion ractions can occur depending on the local oxygen concentration, intat slil. As vil1 be -discussed belov, the presence of chloride =nd sulfate as observed In the Oyster Creak sand cushion also affects the corrosion reactions.

In the absence of chloride and sulfate, Figure 335 hydrous ferrous oxide (FeO a u M aO) or ferrous hydroxide Z1.(COB)2 composes the diffusion barrier adjacent to the Iron surface through which oxygen must then diffuse.The pH of saturated 1e(08)2 is approximately 9.5. fure Fa(0H)2 I. typically white In color but rapidly oxidizes In air to green to greendsh black. At the outer surface of the oxide film, access to dissolved oxygen allows the ferrous oxide to react to forn hydrous ferric oxide or ferilc hydroxidet Fe (0)2 + 1/2 120 + 1/4 02 Fe (08)3 Hydrous ferric oxide ls orange to tcd brown in color and in the main constituent of "rust." It primarily exist nounmaugntic atF 2 0 3 (hematite) or magnetic yFe 2 0 3 where hematite is more stable. Saturated Fe(ON)3 hUs a nearly neutral pH. A magnetic hydrous ferrous ferrite, F304- nf2O. often forms a black. lnteuediste layer between the hydrous 1z203 and FeO. Figure 35.Therefore, as observed on same of the core samples from the drywell, rust films of various colors (states of oxidation) can exist s8inutaneously.

FUNTZ Transmittal No. 87-178-003 Rev. 1 Motivated by the denting of carbon steel support plates In PWR stea generators, sore sophisticated studies had been performed an the "ruating" of carbon steel. It Is believed that this work performed by Pourbalz, et al 9 is particularly relevant to Oyster Creek. In particular, Pontbaiz, at &I vere looking for conditions wbich would produce acid chloride attack of the carbon steel. The mechanism proposed to explain this formation o acid Is the hydrolysis of soluble corrosion products with formtion, inter ella, of no-protective magnetite which Is found In large quantities where denting his occurred.

If no contaminsnts are present (contmilnant are ions other than le, O" and te'. ), no acid hydrolysis would occur. Wihen contaminants (such as CC. Bi% SO) are present, no acid hydrolysis.

will occur provided there are no oxidants (such as dissolved oxygen) and no concentration by evaporation.

Problems may result froe the presence of contaminants shen concentration-by " evaporation occurs even without oxidanta and from the presence of contaminants when oxidants are present, even without evaporation.

Since the Oyster Creek sand cushion Is most Likely characterized by all three factors (contamnants, ozidants mad alternate vetting and drying concentration mechanisms), acid formation Is expected.The hydrolysis of ferrousions In the presence of chloride or sulfate leads to acid and concentrated ferrous chloride or ferrous sulfate solutinas:

Fe -4 re2+ + 2e" Fe2÷ 2 O -eO÷ + U0+Feoe+ + 2 + i 4 Io + HCo + ueo FeORi + 49+ + 250,4 Fas eS 4 + 8 2 S0 4 +Ri The corrosion rates of iron in these solutions are higher than in neutral or alkaline solutions.

For example, Instantaneous corrosion rates were measured by Pourbaix.et at 212"F-(higher than the drywell) in 4 solar ?eCl 2 solution In closed system without an oxidant was 1.6 ails per year (apy). When In contact with magnetite, the instantaneous corrosion rate of Iron increased to 8 upy. and F&PIT Transmittal No. 87-178-003 Rev. 1 was well over 120 upy when ferric contaminants vera present. Magnetite, as was identified in the plug crusts at Oyster Creek, i1 considered by Fourba1s as such an oxidizer end not a stable form of Iron In mildly oxidizing environsents.

The oxidizing power of magnetite is Illustrated in Figure 36.The stable form of iron is a ferric oaide or & ferric hydroxide.

HaSuetite was considered as the normal and stable corrosion product of iron in boiler conditions because mcat boilers operate aatisfaotorily.

Hovever, the opinion that protection is due to f rric oxide, end not to*aSgntita, now receives nore and more support.9 At room temperature it has also been more and more generally accepted that magnatite Is not protective in the presence of aqueous solutions.

The passive films on iron In aqueous -solutions at room temperatures appears to consist of to304 at the matal-oxide interface of of yFe 2 0 3 Sarheuite at the oxide-solution interface.

Although yre 2 0 3 Is difficult to distinguish from maguetite since magheadta is also black and magnetic, and has the sm XlD lines as magnetite, It was not identified In the GE analysis.

Hovever, this similarity between magnetite and maghemite could be responsible for the long accepted opinion that magnetite Is the protective oxide in boiler waters.When magnetite part Wcas are removed from the steel surface, they can be oxidized to hematite (aft2 03s), maheuate (01620$) or goethite (-1e00H), in the presence of water containing as little as I ppb dissolved oxygen.'6.4 Corrosion Rate of Oyster Creek Drrve!ll Zt Is mandatory that the corrosion rate of the Oyster Creek drywall be.estimated so that the present design life can be calculated.

A comparison of this value with corrosion rate data available In the open literature will also be useful in determnling the relative corrosion performance of the drywvell.An estimation of the Oyster Creek dryvell corrosion rate is straight forward since the reduced shell thickness as measured on the removed core plugs, Table 18, was approximately 0.85" and the initial thickness was P&F Transmittal No. 87-178-003 Rav. I approximately 1.13". the typical loss In thickness lis %0.3". If It is assuted that the corrosion initiated with the installatlon of the sand 17 years ago, the average corrosion rate Is approximately 20 spy. The assumed Initiation date Is considered realistic since the sand was Installed in at least a=uolst" condition, was vatted/rewetted during the expansion of the dzyewll which squeezed out the water from the FLrebar-D slurry, and exposed to numerous condensation cycles. f It is assumed that corrosion only Initiated 6 years ago Whe the first fuel pool leak wag noted, then the dEtlinted corrosion rate increases to approximately 50 spy.A review of the open literature on investigations concerning the corrosion of carbon steel In air saturated Is sumarized in Table 19 and FiLpu 37. Data was selected for only toots with reasonable epofsure perlods, that In. corrosion test data based on 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> exposure were not used. In some cases, the exposure period was not provided.

A more recent. Literature review performed for GPII/ESI an this subject by Pednekar of lattelle Colusbus Division also reveals similar corrosion rate.1 6 It is Interesting to note that the 20 mpy corrosion rate estisate for the Oyster Creek drywall falls among the data for carbon steel exposed to water ranging In quality from distilled to aebient seawater to a mixture of soils. If the 6 year-based average of 50 spy Is used, the results are compa~cable to Warsaw tap water or warm seawatar.The results of this comparison may, at first, appear somewhat surprising.

The discussion and results from Section 6.3 suggests that the sand cushion environment with high chlorLde and sulfate, oxygen* high conductivity, etc. would create an environment wbich would produce higher corrosion rates than speclmens immersed In air 'saturated hish quallty wster.However, there are a few factors which may be reducing the overall corrosion rate of the drywell. First, when a metal corrodes in a substance like sand, the sand tends to retain the corrosion products in place which physically stifles further corrosion.

For a specimen imersed In water, the corrosion products can be transported away from the surface allowing corrosion to F&PST Transmittal No. 87-178-003 Rev. I continue physically uninhibited.

Second, during operation the sand cushion adjacent to the drywell may dry out and thus tenporarl.y terminate any corrosion reaction.

When the sand is rectted die to condensation andlor leaks, corrosion reinItietes.

This last scenario would evolve an overall lover average corrosion rate, that is, a conbiuation of high corrosion followed by long pariods of dormancy.Pednakar16 notes that the corrosion rates, corrosion pfoducts, and p1 changes observed in the Oyster Creek drywel coroTscion are those that are observed for corrosion of carbon steel* in aeratede chordle solutions.

6.3 Possbile Corrosion Scenario of Oystert'relk

-As Illustrated in Figure 30, a of factors/events most lIkely affected the corrosion of the Oyster Creek drywell. Such a corrosion scenario is listed below: 1. Backfilling of moist sand into the transition tona creates an initial electrolyte.

Sand is contamuinted by open exposure to marine environment during storage and installation.

Backfilliug also affects porosity of sand which affects moisture retention quality and creates random air pockets.2. Upansion of dryvell during pressure testing "squeezes" water out of the Pirebar-D slurry which flows down into the sand bed. This water contains Initial high quantities of chloride and sulfate.3. Corrosion of the steel drywell initiates.

Red lead primer provides some initial protection due to the formation of Pb0 4.Kfowever, carbon dioxide from the air and sulfate from the sand/or Firebar-D accelerate the breakdown of the lI/iting inhibitive qualities of the red lead primer.

F&PXT Transmittal No. 87-178-003 Rev. I 4. Areas with more ready access to oxygen such as the insulation gap and drain become local cathodes.5. Areas adjacent to concrete are provided soms corrosion protection due to local alkalinity.

A macro-galvanic cell is established between the steal adjacent to the concrete and the steal adjacent to the sand cushion.6. Condensation cycles and leaks from fuel pool bellows gasket contribute aft saturated water to. maintain moist sand cushion.Additional chloride and sulfate may leach out of Firebar-D and be carried into the sand cushion.7. Soam regions of the sand cushion see alternate vetting and drying during utartup/shutdown'cycles,.

This results in a concentration of chloride at the metal/send interface.

S. Sand maintains corrosion products close to metal surface and thus physically stifles corrosion rate.9. Corrosion proceeds interalttently during "ettaings periods (condensation, leaks) or on a continuous basis.7.0 RECCIIE NS The loss of containment Integrity at Oyster Creek is an obvious concern. The corrosion mechanism is fairly vell defined and an estimated overall corrosion rate of 120 spy has been established.

It is nov time to address this problem and identify potential mitigation steps for this phenomenon.

At the specific request of CPUN, three areas of mitigation have been analyzed by GE; 1) polymer replacement/addition to the sand cushions 2) corrosion Inhibitors#

and, albeit superficially, 3) cathodic protection.

F&PNT Transmittal No. 87-178-003 Rev. I 7.1 Polymer Leplacement/Addition to Sand Cushion 1 7 It has been suggested that the removal of the sand cushion could be accompllshed by sluicinS.

If the sand cushion via removed and if the subsequent void was desslcated, corrosion of the drywell would essentially stop. However, due to structural requiremints on the contalmeut, which are beyond the scope of this paper, It say be necessary to refill the sand cushion void with an alternate material which would have suitable mechaidcal properties.

Therefore, GPWF has requested that a bzlef review be performed on candidate cushion materials with particular emphbasi on polymer..The first concern for a polymer replacement would be suitable mans of Injecting the material into the void. It would be possible to spray pellets of through numerous core holes cut through the containment.

Although there would be some procedural difficulties, it should be possible to obtain a relatively uniform distribution of plastic pellets.Scrap material such as polycarbonate resin (e.. LeUan) and thermoplastic resin (e.g. Noryl) are available.

Lexan and Noryl can withstand doses of approximately 81O6 MA MX10 reds, respectively, before any structural damage occurs. Above these total dose levels, the materials would experience degradation by cracking.

However, this cracking and eventual fracturing wxoud probably have little effect on its structural qualities to serve as'a transition cushion. Slice, In this particular application, the sand and plastic would obtain their respective spring constants more fro& the voids in the cushion rather than any Intrinsic material property, both materials should have ilnilar spring constraints.

However, it Is recomended that this assumption be confirmed by a soll geologist.

If It is desired to have a cushion with more support strength, then any candidate polymer should be able to be applied in a sufficiently fluid state so that It could be poured into place. This material could then completely cover the steel surface and fill the sand cushion void. Since.ql .......__ __ __ __ __ _

FUNT Transmittal No. 87-178-003 Rev. 1 there will probably be no opportunity for beat curing, then the candidate material should be characterized by an ambient temperature If a particular polymer in not completely wettable, It may fern a crevice against the steel surface which can promote localized corrosion if any electrolyte Is allowed access. Therefore.

assurance that any poured-In-place polymer adheres well to the steel must be obtained.

Good adhesion will also depend on the skill with which the monomer or partially resin is installed.

Epoxies would probably be the be at candidates.

for an iztrawaal resin injectlon since any remaining sand would behave as a filler. The epoxies would also be likely to adhere to .steal surfaces.

Th short "pot lite" and the viscosity of the epoxy resins would sake application troublesomes addition epoxies are relatively costly.. Coal tar epoxies would probably be the best candidates.

Presumably a "Nuclear Grade* material (i.e., one eapecially low In halosens, sulfur, and embrittlIng Petals) would not be needed.An epoxy spray paint could be used if the main concern Is to prottct the steel surface. If the sand can be removed* possibly a coal tat epoxy paint could be sprayed or poured into the intrawall region (i.e,, Napko 538 alne Coal Tar Epoxy). Then dry sand might be tr-installed into the intrawall area for mechanical support, If necessary.

Xapka 539 Aluminum Kastic Epoy is an aluminun powder-filled polyaside epoxy paint that is good for application to "minimally cleaned" surfaces.

It satisfactorily penetrates residual rust on steel surfaces and generally vets steel surfaces thereby assuring more thoroush coverage.

Napko 682 Splash Zone Batrrier Coating is an epoxy saide capable of application under water, if required.If the sand is not removed, a paint may still be used since spray paint versions of epoxies or other resins would be more fluid than the corresponding resin and would be more likely to penetrate the sand and reach the steel surface. There is no obvious way to assure that complete steel F&T Transmittal No. 87-178-003 Ray. 1 surface coverage In the sand aru Is obtained.

The most that sight be accomplished would be assuring that excess paint ti introduced to the intravall region I.e., there is enouSh paint present for the sand to be saturated and coat al1 thi entire drywall wlvl.For fillers that say provide support as a substitute for or an addition to the sands materials generally used as temporary sealants for valves# flanges, and pipes might be suitable.

These materials

• ould be the elastosers (fiber-reinforced, the fiber usually being glass) used for leak sealing by LUsk Repairs Inc. (Division of Teas I=c.# Houston, TX) at by Furtanite Inc. (Virginia Beach, VW). It would not be possible to use these materials with fiberglass If the sand was not removed. HObweverl, it say still-.be desirable to paint the steel surfaces first.If a polymer with good mechanical strength Is desired, then materials might be used that are applied like polymers used for electrical insulation of moters (i.e., pouring of the prspolymerited material Into place in a large holding container).

However, the highest strength material, (20 hal M) would be 202 glass-reinforced polyaryletherethezketot (i.e., PEW*The polymer is castable at 700"? but to currently rather costly. It is available from ZCI Americas Inc., (Wilmington, DR) or- from a licensee (Greene,Engineered Plastics, Harleysville, Ph) under the trademark "Arlon.Arlon is injectiou-moldable.

Arlon 1260 (carbon fibet-reinforced PEWK) has a 30 ksi UTS. Addition of polymer reasi to existing sand precludes the use of fiber,-cotainiLng resins. There is no assurance that sand as a polymer filler would add to the strength of a polymer; such a In fact, usually results In a weaker product. A high hardness polyetberurethans polymer would provide up to 7 k.i MTS and a 152 carbon fiber-reinforced aromatic polyetherurethan*

would provide approximately TO kal UTS. A styrene-maleic anhydride

opolymer with 20% glass fiber reinforcement and proper processing may have a 10 to 14 kha VTS..The only other high strength materials approaching that of PEEK, are the fiber-reinforced epoxies. Injection-grade, 202 glass fiber-reinforced AUS-ii1-

\ /F?&P1T Tranamittal No. 87-178-003 ev. I have a 10 to 13 kal UTS. Sii1cone/nylon 6, 6 pseudo interpenetrating networks (l.a., Petrach Systems, Bristol, PA), made by mizing the two components into a povder or granular fora, may achieve 10 to 12 kal UTS.7.2 Corrosion Iuhlbitors 1 7 The primary problem with corrosion inhibitors involves obtaining a uniform distribution over the entire surface of the steel or, ap with the paint discussed above. corrosion may become focused at unprotected areas. At the same time, some prohibited Inhibitors (i.e., chromates) may have to be avoided. Limited lUfe or short-term Inhibitors are not useful unless tha sand cushion area Is made virtually airtight.

Therefore, Inhibitors that operate by scavenugng oxygen may not be usable. However, those that promote protective oxide formation an steel surfacea appear to be the ecat praofming.

The difficulty Is in identifying all of the required properties for this imnbi tor In one Inhibitor.

Volatile Inhibitors are usually of the type that oxygen thereby makiug them limied-life Inhibitore.

Yet water-soluble but volatile corrosion Inhibitors would be most likely to provide complete coverage of the steel surface of the sand cushlon area.Molybdate.

could be used as a teplacement for chromate to provide an Inhibitor that promotes protective ozidt. fila formation on steel; but molybdate is not available In a volatile form. Sodium solybdate is available from Noah Chemical, Framingdale, NT. -olybdate corrosion Inhibitor@, but only for cooling water, have been studied by Nouseman (Burnham)

Ltd. of the fortals Water Treatmezt Group In Geat Britain. Molybdates are also available from Clla Kolybdenum Co.. Calgon Corp., Exxon Chemical Co., Magau Corp., Nalco Chemical Co., Nevage Industries Inc., and I.T. Vanderbullt Co.A vater-soluble (in case of the presence of any liquid-phase moisture), volatile corrosion Inhibitor such as one that might be used for packaging or in long-term storage is the only type feasible for sand cushion use to inhibit further steel wall corrosion.

Cortac Corporation (St. Paul#MNI contact Baorl Ifkuic) Is outstanding Ln this area. They have produced a F&PHT Transmittal Nfo. 87-178-003 Rev. 1 volatile.

water-soluble Inhibitor dicyclohuxylannoulum chromate (U.S. Patent 4,275,835; June 30, 1981). They may also have the molybdate version of this inhibitor or the chrmate say possibly be acceptable for Oyster. Creek since it is not likely to escape the sand cushion region).Just as vas the case for coatings, incomplete coverage by an Inhibitor can concentrate corrosion in unprotected areas. Howeverg some corrosion inhibitors pose another problem. Nitrites, for example, shoul4 be avolded since there are certain moderate concentration ranges (depending an other environmental parameters) wbich would proote corrosion Instead of Lnhibiting it.If the presently existing snud to not removed, volatile "corrosion inhibitors may not work. The sand will readily absorb this type of corrosion inhibitor.

In fact, this would also be true of any Inhibitor applied as a solution, aqueous or otherwise.

The same problem exists hers an It did for considering the use of resinous filers or paints Ln the presence of the exiatein send: a sufficient excess of Inhibitor, as a solution or as a vapor, must be used so as to assure that the Inhibitor reaches the steel vll and costs It completely.

Otherwise localted corrsion say occur. Since liqulds wil be absorbed throughout the sand more readily than vapor, an oil-soluble version of an Inhibitor may be suitable for application In this case.Cortec Corporation has oll-soluble versions of its Inhibitors.

Using one that is oil-soluble and volatile may be suitable mince it would help ensure coating of the steel wal with the inhibitor In one manner or another.Cortec VCX-320 wold be one such product. Preservative petroleum lubricating oils could also be suitable.

Examples are Oilcoat VT and Cl.coat A. (formerly Gulf products but nov Chavron products), )Mobil VaporTech Light Oil, SACI100 (Vitco Chemical Corp.) and Tover 6401 (Tower Chmical Corp., FA).Similar materials may be available from other sources, but It to best to use a product containing a volatile corrosion inhibitor. (Note that these materials are bound to very flammable.)

F&PT Transmittal No. 87-178-M03 Rev. 1 Pednekar also provides & ]let of organic And inorganic Inhibitors for carbon steel in aerated chloride solutions16 7.3 Cathodic Protection Bically, cathodic protection is & means of reducing the corrosion of a component by making the motel a cathode by seanm of anLIMpre8sed current or attachment to a sacrificial anode (such as magagsLumit 4zInum br zinc).Since the cathodic protection (CI) syeste forces electrons into the metal creating this cathode, the basic principle of applying CP Is quite simple. In general, the practical application of this corrosion control method is much sore difficult.

For the specific came of the Oyster Creek drywall, It may be-extremely difficult.

For ezamples CP systens are designed to protect coated structures, that Is, provide protection against any defects (holidays) in the coating.This inlaftres the required applied current for protection.

For Oyster Creeks the dryvell Is presently uncoated end therefore significant and perhaps prohibitive currents may be required.

Other concerns include ;hat source of direct current should be use; can a suitable anode be designed and, in fact, Installed around the entire sand cushicon and how can It be ascertained, on completely buried structure, whether or not the entire surface has, in fact, been made a cathode and ali corrosion mitigated.

Information which can answer such questions are beyond the scope of this report.7.4 Ltigation Recommendation It appears that a multiple approach should be used for the mitigation of corrosion of the Oyster Creek dryvell. Since It appears that the main source of the corrosion problem is the vet chemically contaminated sand, the mest suitable mitigation stop would be the removal of the send and dryin of-36L F1& Transmittal No. 87-178-003 Rev. I the cavity. This, by itself, would reduce the corrosion rate of the dryvell to a vanishingly small level.If so structural support is required, a further corrosive iltigatoln Improvement would be a back spray paintin of the drywll to provide crating protection vith an alumim powder-filled polyaside epozy paint followed by application of a volatile corrosion inhibitor to mitigate any holidays In the coatinge.If the sand cannot be reamoed, than the application of an excess quantity of an oal-soluble vapor phase inhibitor say be the best approach.

If excess water is a problem, then an application of an excess of a sufficlentl.

diluted epoxy paint such as Napko 682 Splash tane Barrier epoxy aside may be the best choice. This paint application could then be followed by the excess application of an oil-soluble volatile corrosion inhibitor.

As noted above. the use of cathodic protection as a suitable corrosion mitigation step Is considered beyond the scope of this review and therefore will not be discussed.

8.0 The results of metallurgical analysis by both GPU= and GC, data from the open literature and the above discussions have indicated the following conclusions concernliS the corrosion of the Oyster Creek drywells 1. The corrosion/wastage of the dryweUl ls due to the presence of oxygenated moist/vet sand and exacerbated by the presence of chloride and sulfate in the sand cushion. A contesinste concentrating mechanisn due to altenate wetting and dryln of the sand cushion may have lso contributed to the corrosion phenomenon.

917-

?6iT Transiuttal No. 87-178-003 Rev. I 2. Althwgh Firobem-9 Is a known corrosive saint to steels its role in phenommnon is probably secoUdary.

The source of courants In the sand cushion say have been from the local martae onvivoumint.

3. Since the wal thithness measured by UT are strmeeli claoss to those tessred on actual removed specmens.

OT appears to be an accurate non-4estructive ethod of woultorfti wall 4hickness.

.4. bte estimated corroslon rate of the Oyster Creek drywll is 420 a". This rate reflects the verage corrosion o"er 17 years of servics regardless of the relative continuity of the corrosion reactions i.e. there say be periods of high corrosion rate actIvIty , uring vatting cycles followed by dormancy durin "dry" periods.3. Excuding cathodic protection which ts beyond the scope o. this report, the optiaum metbod of miti4gtion of the corrosion of the Oyster Creek dryall. to be the cOmbluationo*

saa Ce@ val back spraoytg of a protective paint aWd appication of a volattle corrosion Inhibitor.

90.. ,

F&WP Transmittal No. 87-178-003 REMMCES 1. GPUN Safety Evaluation No. 000243-002, December 12, 1986.2. C.R. Judd, "Leaching Tests and Chemlcal Analysis Results for Oyster Creak Drywall Samples," FIN Transmittal 87-212-0004, January 23, 1987.3. G.T. Austin, Shrews's Chemical Process Industries, Fifth ad., llcGrov-Hull.

New York, 1984.4. C.A. Sorrell and C.L. Armstrong, "Reactors and Equilibria In Masnesun Ozycblorlde Ceaent, Jour. of ACS, Vol. 39, No. 1-1, Jan-Feb., 1976.5. J.2. Levis Letter to S.M. Gordon, "x-lay Diffraction Analyses of Corrosion Crust from Inner Surface of Oyster Creek Drywall,*

December 19, 1986.6. K. omanoff, Cnrerotand Corrosion, National Bureau of Standards, 1957.7. BUllnski, at &L. 'Me Formation of Magnesium Ozychlorlde Phases in the Systems KgO-KfSCl -K 0 ind &aOE-NaCl-80

• , Journal of the American Cermaic Socelty, 2 Vol. 64, No. 4, pSi;619184.

8. 5.1. Kauarler, *Update on Mfagnesium Otychlorlde Fireproofing," Fire Technology, Val. 13, Kay 1977.9.. K. at a&I "Chsmical Aspects of Denting in Steru Genarators

," NP-2177, BPS.!, Palo Alto* CA. December 1981.I1. E. Schaschl and G.A. Marah, Corrosion, Val. 16, 1960.11. 3.11. fblig, Corrosion and Corrosion Control, John VLiey and Sons, Nov York, 1971.12. R.N. Ublig, editor, Corrosion Handbook, John Wiley and Sons, New York, 1948.13. 1.1Z. Berry, at al, "Survey Report on Corrosion Behavior of Carbon Steel in Pure Water at Ambient Conditions," June 30, 1975.14. L.L. Shzefr, editor, Corrosion, Nevaeus-Butterworths, London 1976.15. N.E. Banner, Corrosion Date Survey, ACE, Houston, Toeas, March 1974.16. S.F. Padneckar, *Corrosion of Carbon Steel in Aqueous Environments at Temperactures Below Boiling -A Literature Review," Battelle Columbus Division, February 10, 1987.17. R.S. Tunder letter to I.K. Gotdon, "Informatiou Relating to "ryeall Containment WHal Corrosion at Oyster Creek," January 8, 1987.

TABLE 1.DRMLL SIM~l

(ffSIl)(flzCA)ASS: A-212-61T Or I FIKOK FIRW -NNORALIZED 1-1.06 P -.010 S- .023 Si -..21 mu, TENSILE -75,0001 I YIL -50,000 PSI ELK -35 NOE: NATEIAL WA NoA~ TEmE

M Nuclear TELECOPIED TO J. CHARTERINA 1/28/88 Memorandum FEB 01 i988

Subject:

OYSTER CREEK REACTOR CAVITY LEAKAGE Date: January 28, 1988 From: Manager Mechanical Components

-R. H. Greenwood Locationmorris Corp. Center 5310-88-018 To: Director Engineering Projects -D. K. Croneberger Engineer -A. Collado Engineer -J. A. Martin Engineer -J. Charterina H/X & Pressure Vessels Manager -J. D. Abramovici Manager Plant Engineering

-J. DeBlasio Manager Special Projects -B. .D. Elam Materials Engineering Manager -R. L. Miller Plant Engineering Director -A. Rone Project Engineer -A. Spivak Supv. Mechanical Engineering

-C. Schilling The following are January 27, 1988.review the cavity the repairs.resolutions and action items from our meeting Tuesday, The next meeting will be Thursday, February 4 In FlA to coating options and to develope an integrated schedule for CAVITY STEPS The reactor cavity steps were inspected per Procedure 6130 Q AP 7209 that included visual examination with PT of suspect areas. Defects were not identified from inspections conducted 1983 nor 1987. The step plate thickness was verified to be 1/4 inch for the top three steps and 1 inch for the bottom step landing. It was agreed that this inspection should be repeated.

The top three steps can be inspected as soon as the top three concrete shield plugs are removed using the fourth shield plug as a platform.

Provisions need to be made for access to inspect the last step after the last shield plug is removed. (ACTION -3. Charterina) 9073d/343 N70648-06.86)

Distribution January 28, 1988 5310-88-018 Page 2 CHANNEL The transfer channel from the cavity to the fuel pool and to the equipment pool is lined with 1/4 inch plate. There is still a question as to the existence of leak tracing on the weld attachment to the pool liners. There is no record of any inspections of these channel areas or of its attachment to the vertical 12 gage liner. (ACTION -J. Martin)Since the channel area is subjected to loading similar to the steps when the shield gates are put in place, it was decided that these areas also need to be inspected.

This will require that the fuel be moved away from the gate area prior to removal of the shield plugs to prevent shine through the fuel pool dam. (ACTION -J. Charterina)

WELD REPAIRS Should there be any defects found in the step area or the channel area a weld repair procedure will be needed. Although there is a procedure in place it needs to be reviewed and updated. A separate procedure is needed for repairing the liner in the areas where samples were removed or where there is unacceptable liner damage. (ACTION -R. Killer)RELINER Two coatings are being considered as a permanent fix for sealing the cavity liner. (1) Ceilcote 222HT, a vinyl ester with flake glass fill-that would require grit blasting and two coats; (2) A Dow silicone requiring two coats but may not need grit blasting.

The Ceilcote coating meets the coating selection criteria.

It is unclear at this time as to whether the Silicone coating meets all GPUN requirements.

Additional details are being sought for the Silicone coating. The option of using Hypolon sheets was discounted since it degrades at high temperatures. (ACTION -R. Miller)A strippable coating option, Isotron, is being pursued. This material is supposedly suitable for underwater service but would have to be removed after each use. Its advantages are that it can be sprayed rolled or brushed without any surface preparation and it provides an additional benefit by removing radiological contamination.

More information is needed on its strippability to assure it remains in place until it is intended to be removed. Another concern is the leach ability of impurities and its teflon content. Should this option be selected, the inspection or repair of the steps or channel area would not be needed. An economic evaluation needs to be prepared to compare these three options. (ACTION -R. Miller)9073d/343 Distribution January 28, 1988 5310-88-018 Page 3 TROUGH Access to the concrete trough will be obtained through a 6 inch hole drilled through the steel trough. The concrete will require repair at two locations.

One being at the drain and the other about 60 degrees clockwise from the drain. These areas were identified from the camera inspections and indicated that the lip of the trough was not sufficient to assure that water would not enter the area between the concrete and the containment.

Video enhancements of these suspect areas were made to determine if cracked concrete existed. No evidence of cracking was observed.

A repair procedure is needed. (ACTION -J. Martin)Additional video camera inspection will be needed to precisely locate the areas to be repaired.

A locating method will be needed to assure that the repair hole is drilled at the correct locations. (ACTION -A Spivak)CONCRETE REPAIR The concrete may be repaired with Ceilcote 665 or an available Belzona Magna Quarts. A wood mock-up of the steel trough area is to be assembled so that the technician can practice the repair since he will have to watch his activities through a video monitor. If this can not be done, it may be necessary to drill two holes for each repair. Drawings of a wood mock-up model of the work area are complete. (ACTION -J. Charterina)

DRAIN REPAIR Since it was reported that the 2 inch drain pipe appeared to be clear of obstructions no pre-outage borascope inspection of the pipe up to the concrete trough was advised. However, when the access hole is drilled to repair the edge of the concrete trough, the drain will be examined.

A sump hole about 4 inches in diameter and a few inches deep will then be drilled adjacent and intersecting the drain. The objective will be to provide a drain pot to collect then direct the flow from the trough. Removal of the concrete plug and section of the drain pipe needs to be reviewed for feasibility. (ACTION -J. Martin)TROUGH SEALING The 6 inch access holes in the steel trough will be plugged and seal welded shut. The bottom of the steel trough will be hydrolazed and coated with Ceilcote 665 or promotec lose applied in a liquid state to cover all of the carbon steel welds. It was reported that the bellows assembly area had been factory welded and leak tested after installation including the SS/CS weld on the lower part of the bellows. There is a video record of this lower weld, and it was reported to be in good condition but should be reviewed.(ACTION -R. Miller)9073d/343 Distribution January 28, 1988 5310-88-018 Page 4 Video inspection of the underside of the bellows area was discounted since the air pressure test gave confidence of the bellows integrity.

A plan was presented that the air pressure test would be repeated should leakage be detected in the drain line. An extension plate would be installed over the notch at the drain bellows access plate cover so that the air bubble behind the bellows cover would include the SS/CS weld. Introduction of air in this area would reduce the detected leakage should it be coming from the bellows or the weld. If it was determined that leakage existed, a leak rate monitor would be used to supply a constant supply of air to limit water leak through. Provisions needs to be made so that air can be supplied to this area. (ACTION -J. Charterina)

FUNDING The distribution of funding between capitol and Operation

& Maintenance would depend on the repair method for the cavity. The repair of the trench and inspections will be Plant O&M. (ACTION -A. Spivak)RHG:am R. e wood Extension 7404 cc: Director Engineering

& Design -G. R. Capodanno Mgr. Mat. Engrng./Chem Support -F. S. Giacobbe Plant Systems Director -D. Slear CARIRS 9073d/343 10110106 -9:-54:39 Nuclear Technical Functions SafetylEnvironmental Determination and 50.59 Review (EP-016)UNIT_ Oyster Creek _AG_ 1 OF 13 DOCUMENT/ACT1VITY TITLE Tenporary Repair of Rx Cavity SE Re. No. 4 DOCUMENT NO. OCIS-328257-002 oOC REV. NQ 0 SE No. 328257-002 Type of Activity Repair (Modification, procedure, teSt, experiment, or document)1. is this actklvty1document listed In Section I or IH of the matrices in Corporate Procedure

-_Yes ONo 1000-ADM-1291.D1?

If the answer to question 1 Is "no" stop here. (Section IV Wctivitiesldocuments should be reviewed on a case-by-case basis to determine if this procedure is applicable.)

This pro-cedure Is not applicable and no documentation Is required.

It the answer is "yes" proceed to question 2.2. Is this a new activItyfdocument or a substantive revision to an actMty/document? (See l&Yes CNo Exhibit 3, paragraph 3, this procedure for memmples of non-substantive changes)If the answer to question 2 is "no" stop here. This procedure Is not applicable and no documentation is required.

If the answer is "yes" proceed to answer all remaining questions.

These answers become the Safety/Environmental Determination and 5059 Review.Does this activity/document have the potential to adversely affect nuclear safety or safe LKYes LNo plant operation?

4. Does the activity/document require revision of the systemlcomponent description in the FSAR lJYes l[No or otherwise require revision of the Technical Specifications or any other part of the SAR?5. Does the activity/document require revision of any procedural or operating description in I-Yes b)No the FSAR or otherwise require revision of the Technical Specifications or any other part of the SAR?6. Are tests or experiments conducted which are not described in the FSAR, the Technical liYes 0No Specifications or any other pert of the SAR?No because Installation/Application of the waterproof liner is a temporRry repair which will be removed after refueling and draining of the cavity.Documents checked, If any of the answers to questions 3, 4, 5 or 6 are yes, prepare a written safety evaluation on a Safety Evaluation form.If the answers to 3, 4. 5. and 6 are no, this precludes the occurrence of an Unreviewed Safety Question or Technical Specifications change. Provide a written statement in the space provided above (attach additional sheet if necessary) to support the determination, and list the documents you checked.7. Does this document involve any potential Non-Nuclear environmental impact? eyes NJNo 8. Are the design criteria as outlined in TM-l1 SDD-TI-OO Oiv. I or OC-SOD-000 Div. I Plant 0Yes lt]No Level Criteria affected by, or do they affect the activity/document?

If yes, Indicate how resolved It the answer to question 7 is yes, either redesign or provide supporting documentation which will permit Environ-mental Licensing to determine if an adverse environmental Impact exists and if regulatory approval is required (Rel. 1000-ADM-1216-03).

If in doubt, consult the Radiological and Environmental Controls Division or Environmental Licensing for assistance in cormipleting the evaluation.

Signatures D4 -Date EngineerOriginator S. K. SaM ...Section Manage ! F. S. Giacobbej Responsible Teclinical Reviewer _ _____ .-______Other Reviewvers) N 5047 (02-88) 1O10O10 09:54:39 Nuclear Technical Functions Safety Evaluation (EP-016)UNIT Oyster Creek PAGE 2 of 13 ACTIVITYIDOCUMENT TITLE Temporary Repair of Rx Cavity SE No. 328257-002 LinerRev.

No.OCIS-328257-002 0 DOCUMENT NO. (if applicable)

-Rev. No.____Type of Aotivity/Document Repair (Modification.

procedure, test, experiment, or document)This Safety Evaluation provides the basis for determining whether this activityfdocument involves an Unrevewewd Safety Question or impacts on nuclear safety.Answer the following questions and provide reason(s) for each answer per Exhibit 7. A simple statement of conclusion In itself Is not sufficien.

The scope and depth of each reason should be commensurate with the safety significance and complexity of the proposed change&1. Will implementation of the activityldocument adversely affect nuclear safety or safe plant operationa?

The following questions comprise the 50.5 considerations and evaluation to determine If an Unreviewed Safety Question exists: 2. Is the probability of occurrence or the consequences of an accident or malfunction of equipment Important to safety previously evaluated In the Safety Analysis Report increased?,Is the possibility for an accident or malfunction of a different type than any evaluated previously In the Safety Analysis Report created?4. Is the margin of safety as defined in the basi for any Technical Specification reduced?r3Yes 1No QYos ftNo ClYes 12No ClYes liNo If any answer above is "yes" an Impact on nuclear safety or an Unreviewed Safety Question exists. If an adverse Impact on nuclear safety exists revise or redesign.

If an unreviewed safe-ty question with no adverse Impact on nuclear safety exists forward to Ucensing with any ad-ditional documentation to support a request for NRC approval prior to implementing approval.L. Specify whether or not any of the following ae required, and It "yes" indicate how it was resolved Yes TRITFWR/Other No a. Does the actlvlt:ocument require X an update of the FSAR?Explain: Application of the water proof barrler(s) are temporary and they will be removed after refueling.

b. Does the activity/document require a Technical Specification Amendment?

X Explain: Same as Item 5 (a).N 04 (0248) 1310/106 09:54:39 PAGE 3O0 13 C. Does the activity/document require a Euality Classification List (CL) Amendment?

Yes TR/TFWR/Other No X Explain, Same as Item 5 (a).d. Other: (If none, use NA)X/IA This form with the reasons for the answers, together with all applicable continuation sheets constitutes a written Safety Evaluation.

Ust of Effective Pages Page No. Rev.No. Page No. Rev.No. Page No. Re_1 4 11 1 2 4 12 1 3 4 13 1 r. No.4 5 6 7 8 9 10 2 2 2 0*2 0 2 t Sinatures C7I 10Date EnglneedOrgtinao

s. K. saha bO tI0l Reeponsible Technical Reviewer _ g -f-er Independent Safety Revie -r Other Reviewer(u)

_______I N 5046B (02-M) 10110106 69:54:39 SE 328257-002 Rev. 2 Page 4 of 13 1.0 PURPOSE 1.1 The purpose of this safety evaluation is to address the adequacy of design and safety impact of Installation of a temporary barrier on the carbon steel trough and the stainless steel liner, of OC-Reactor Cavity Pool to prevent leakage of water during refueling operation.

2.0 SYSTEMS AFFECTED 2.1 During the proposed activities, the following systems will be affected: Reactor vessel & recirculating system° Main steam system o Condensate and feedwater systems* Reactor Core Components

• Control rod drive system* Standby liquid control system* Reactor cleanup system* Reactor shutdown cooling system* Fuel storage and handling* Spent fuel pool loading" Radioactive waste system* Stand by gas treatment system 2.2 GPUN Drawings: 2.2.1 GPUN 3E-153-02-001 through 009, "General Arrangement Reactor Building".

2.3 General Electric Drawings: 2.3.1 Dwg. No. 237E516 Sh. 1 -- Fuel Storage Pool Arrangement 2.3.2 Dwg. No. 237E547 Sh. 1 & 4 -- Arrangement of Fuel Storage Pool 2.3.3 Dwg. No. 237E975 Sh. 1 & 2 -- Study Refueling Equipment Storage Arrangement 2.3.4 Dwg. No. 3E-153-88-014

-- "Reactor Cavity Cross Section" 4410H/339H/4 10/10/06 09:54:3-)SE 328257-002 Rev. 2 Page S of 13 2.4 Burns & Roe Drawings: 2.4.1 Dwg. No. 4056 -- R.B. 4th Floor el 95'-3 Plan & Section 2.4.2 Dwg. No. 4057 -- R.B. 5th Floor el 119'-3 Plan & Section 2.4.3 Dwg. No. 4068 -- Rx Bldg. Storage Pool Section & Details 2.5 OC Final Safety Analysis Report (FSAR) -Section 9 -"Auxiliary Systems".3.0 EFFECTS ON SAFETY 3.1 Safety Functions Documents that define the safety functions of the system are: 3.1.1 OCNGS FSAR -Chapter 9 3.2 Description and Function of the Systems Affected The reactor refueling cavity at OC is a SS lined concrete cavity which Is located between elevations 91V91 and 11913". It is approximately 37' in diameter and it completely surrounds the drywell head. The Rx refueling cavity Is connected through gates and channels to Equipment Storage Pool and Spent Fuel Storage Pool.During refueling, the cavity Is flooded with demineralized water from Condensate Storage Tank which is at ambient temperature.

The water from CS tank flows through the reactor into the cavity and the level of water in the cavity is maintained to an elevation of 114' maximum. The temperature of water is maintained below 125°F during refueling operation which lasts approximately 10 weeks. The transfer of new and spent fuels is carried out under water to reduce radiation level. Upon completion of refueling, the reactor cavity and equipment storage cavity are drained, after installing the refueling gates, through lines at the bottom of these pools to the suction of the fuel pool pumps and hence to the Main Condenser hotwell or to radwaste.

Supplementary drains from these cavities are directed to the Reactor Building Equipment Drain Tank. There is a curb around the cavities to direct any overflow to drains.3.3 Statement of the Problem During a recent Inspection of a portion of the stainless steel liner on wall of the Rx Cavity by Liquid penetrant test, numerous unacceptable defects were found. A large number of such defects were found to be through wall defects by vacuum test. Two samples containing defects were removed from liner wall for investigation of the failure. The failure mode was determined to be fatigue (Ref. 1). No evidence of stress corrosion cracking was found on these samples. Although no such tests/examinations were carried out on the CS trough, it was perceived that the trough floor can also be a contributor to the water leakage due to the presence of about 600 linear feet of fillet welds and 117 plug welds all of which have experienced some deterioration over time.4410H/339H/5 10110/6ý6 09:54:39 SE 328257-002 Rev. 2 Page 6 of 13 Based upon the finding, it was postulated that numerous through wall defects will allow leakage of demineralized water from the cavity into the concrete wall.If this leakage of water (when the cavity is flooded during refueling) is not rectified, the demineralized water may deteriorate the concrete wall and will corrode the drywell shell from OD. To prevent water leakage through the cavity liner two major options were considered I.e. (a) -weld repair of the defects and (b) temporary barrier over the entire cavity liner. The weld repair option has following drawbacks

-(a) too many defects (b)weld repair of so many defects will produce large residual stresses and warping of the liner and (c) the repair areas will eventually fail by the same fatigue mechanism in the future. Therefore the latter option (i.e. temporary barrier) was selected to prevent water leakage. To prevent water leakage through trough floor the options considered were (a) weld repair and (b) temporary repair.The temporary barrier option was chosen since weld repair will involve large manhours and manrem exposure.3.4 Proposed Rectification 3.4.1 Cavity-Liner Repair The proposed cavity liner repair consists of a combination of welding of larger defects, application of ss adhesive tape over certain size defects followed by application of a temporary coating barrier. The coating barrier(s) and ss tape will be qualified to Ref. 3 and 4 respectively.

During application of the coating barriers, the reactor head will be kept on to prevent introduction of foreign materials.

After completion of the refueling, the polymer barrier and ss tapes will be stripped off. The removable temporary barriers have been selected because no proven permanent barrier material could be found which could successfully withstand both operating and refueling environment detailed In Para. 3.6.3.4.2 Cavity Trough Repair: The trough will be hydrolased to remove rust, oil, grease or other debris followed by drying and solvent wiping to remove any trace of oil or grease. It shall then be coated with the same temporary polymer barrier as qualified for the stainless steel liner removed later after refueling is complete.3.5 Materials of Construction 3.5.1 The existing Rx cavity liner is fabricated from ASTM-A240 Type 304 stainless steel in the following thicknesses:

Walls: 0.109 inch Floors: 0.250 Inch Shield Plug Steps: 1.00 inch lowest step base.500 inch remaining steps 3.5.2 The cavity trough is fabricated from ASTM-A212 Grade B material in the following thicknesses:

Bottom Plate: 2 -3/4" Bottom Plate: 1" Side Plate Expansion Joint 7/8" 441OH/339H/6 10/1U/06 09:54:39 SE 328257-002 Rev. 0 Page 7 of 13 3.6 Environments 3.6.1 Refueling Environment (a) The medium is demineralized water of the following quality: Parameter Admin. Limit Chloride i 50 ug/L Conductivity

@ 25°C < 2.0 timho/cm pH 5.3 -7.5 Silica < 100 jig/L Total Organic Carbon < 500 ttglL (b) The temperature of the water during refueling is less than 125*F. The temperature of the liner before flooding can be as high as 140"F.(c) The pressure on the liner will be hydrostatic pressure of the water, i.e., maximum of 10 psig.(d) The radiation level of the cavity prior to flooding is estimated to be 20-100 mR/Hr -General area and 50-3000 mR/Hr contact with drywell head on.3.6.2 Operation Environment (a) The medium is dry air (enclosed by shield plug).(b) The temperature on the liner can vary from 225*F to 280*F (Ref. 8).(c) The radiation level at the liner location can be as high as 100 Rem/Hr of gamma radiation during reactor operation.

3.7 Technical and Safety Concerns on the Proposed Repairs Concerns which must be addressed and dispositioned include 1) Hill the coating(s) be able to seal various types and sizes of defects, 2) will the coating(s) be able to withstand refueling environment without delaminating or allowing water leakage, 3) will an explosion hazard be created by the application process, 4) will the Standby Gas Treatment System (SGTS) charcoal filters become fouled with solvent vapors during coating application, 5) will the leachates from the coating material adversely affect the reactor water/fuel pool water chemistry, 6) will loss of cooling system during refueling affect the coating adhesion and/or water chemistry, 7) does the coating application and removal produce any health or safety hazards, 8) Is there any adverse reaction between coating and the substrate, 9) what are the impacts of residual coatings left inadvertently on the liner during operation, and 10)what are the impacts of inadvertent introduction of liner pieces into Reactor or Spent fuel pool cleaning system.4410H/339H/7 10130106 019:54:39 SE 328257-002 Rev. 2 Page 8 of 13 The evaluation of the above items of concerns are summarized below: 3.7.1 Reference 6 is a GPUN evaluation of test results on the performance of stainless adhesive foil. The subject tests were carried out as per Reference

3. The test results and the evaluation indicate that a) the SS foil seals defects up to 1/4" width as evidenced by vacuum box testing of the foil repaired areas In air. The force/pressure generated by the vacuum test is greater than the expected maximum hydrostatic pressure on the repaired areas, b) the SS foil maintains adequate peel strength In contact with water as evidenced by the Immersion test results. Since peel strength did not show any appreciable loss of adhesion during 5-10 week immersion test, it can be deducted that the water sealing capability of the SS foil will be maintained over areas containing defects up to 1/4" width, c) the test results show that the foil by Itself can provide an acceptable barrier on the liner under anticipated refueling environments against water leakage.Reference 7 is a GPUN evaluation on the performance of sprayable coatings.

The subject tests were carried out to the requirements of Reference

4. The test results indicate that the subject coating a) can seal defects up to 40 mils In width under simulated refueling environment, b) maintain adhesion to the substrate under water as evidenced by immersion test results, and c) can provide an acceptable barrier against water leakage under anticipated refueling environment.

Based upon review of test results of the sprayable temporary polymer coating on the smooth stainless steel surface, it has been concluded that the same coating will perform equally well on the carbon steel surfaces.

Since carbon steel surfaces will be rougher in texture thereby providing a greater adhesion of the subject coating.3.7.2 Reference 9 is a GPUN Fire Hazard Analysis Report. It shows total amount of solvent that will be released by the application processes.

Explosion can occur upon ignition of concentrated solvent vapors in a confined area. The lowest concentration (M) at which this can occur is defined as the lower explosion limit (EL). The FHA report concludes that the solvent vapor concentration in and around Reactor Cavity area will be significantly less than the LEL anytime during coating application process when manufacturer's recommended coating application procedure is followed.3.7.3 Reference 10 is a Nucon Evaluation report on the effect of solvent releases on the SBGT system and resulting contamination of the charcoal filter. The report evaluated solvent release from spraying the Rx cavity with approved strippable coating system. The report concluded that the impact on SBGT system from Rx cavity coating application is not expected to be significant considering the low solvent content of the coating. In addition to the above, Plant Surveillance procedures require testing of the SBGT system 1 0/'010/6 09:5j4:39 SE 328257-002 Rev. 0 Page 9 of 13 3.7.4 In order to ensure that the proposed coating work (or its removal) will have no adverse effect on safety, plant personnel or public health, the following precautions have been required: a. The use of welding, grinding equipment or open flame is forbidden in or around the Rx cavity area during coating.b. The amount of solvents and other materials utilized in the Reactor Building which could be absorbed by the carbon filters has been limited. Use of solvents to be utilized in the Reactor Building has been restricted.

No solvents other than those associated with the coating will be utilized in the reactor building at the time of coating application.

The Reactor Building normal ventilation system will operate during coating and during the drying period. This will allow any solvents generated to be exhausted and discharged to outside.c. Use of solvents prior to coating for degreasing will be restricted.

d. All coating operations shall be terminated upon loss of the normal exhaust ventilation.

e. There shall be complete compliance to the following plant procedures:

119 Housekeeping 119.4 Consumable Materials Chemistry Control 120 Fire Hazard 120.4 Fire 120.5 Control of Combustibles All of the above controls will also reduce fire hazards and fouling of charcoal filters as discussed previously under Para. 3.7.2 and 3.7.3.3.7.5 Reference 11 is a calculation for total leachates expected In the Rx cavity water due to use of GPUN approved strippable coatings.

The results indicate that water chemistry of the affected systems will remain within acceptable limits.3.7.6 The loss of shutdown and fuel pool cooling system has been reviewed.

It has been estimated that a temperature rise to 212OF can be expected due to total loss of cooling systems.The test results (Ref. 6 and 7) on approved strippable coating and stainless foil products reveal that all of the barrier materials can withstand 212*F boiling water environment for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> without significant loss of adhesion.441 OHj /310Y -'

10/10/06 09:54:39 SE 328257-002 Rev. 2 Page 10 of 13 3.7.7 Any chemical or metallurgical reaction between the coating materials and the stainless substrate was evaluated.

It has been concluded that coating materials (i.e., latex, or the foil adhesive) have no chemical reactivity with the SS substrate up to 212OF in presence or absence of water.Similarly, no metallurgical reactions are expected since the temperature during application or service is too low.Electrochemical/corrosion reactions are not expected since 1) the coatings/adhesives are electrochemically inert and 2)demineralized water is a poor conductor.

3.7.8 Some coating may be tightly trapped into cracks and crevices of the liner and trough floor such that 100% removal may not be feasible.The operation environment (Para. 3.6.2) and its prolonged exposure is expected to embrittle the trapped coating/adhesive material.

The anticipated movements (e.g., crack propagation or differential thermals expansion or contraction etc.) within the cracks/crevices are expected to dislodge the trapped materials in future. Since 1) the amount of such trapped material will be very small, 2) the leachate/bulk analysis showed no harmful effect on the fuel pool water chemistry, no adverse effect is expected.3.7.9 The question of inadvertent Introduction of the coating barrier materials Into the reactor was evaluated.

It was determined by immersion and boiling tests (Ref. 6 and 7)that the approved coating materials will maintain its adhesion and will neither spall off nor fragment into small pieces. In addition, during coating application either the reactor head will be kept on or the Rx head opening will be securely covered up. All of the above should preclude introduction of coating materials into the reactor.3.7.10 The quality standards of the cavity are not affected by the proposed coating work since the objective of this work is to prevent water leakage through liner and trough during refueling.

The first barrier to prevent water leak is the strippable polymer Coating. The SS foil tape over the known areas of leakage on the liner is the second barrier.3.7.11 Natural Phenomena Not affected.

The use of temporary barrier material will not change seismic classification or tornado/hurricane flood protection.

3.7.12 Fire Protection Not affected.

The existing fire protection systems are not affected by this activity.4410H/339H/l0 Vh/10/0 i 0 9:14:39 SE 328257-002 Rev. I Page 11 of 13 3.7.13 Environmental Qualification Not affected.

The Rx cavity coating does neither affect existing EQ components nor the criteria used to qualify the components.

3.7.14 Missile Protection Not applicable.

The Rx cavity is located inside RB.3.7.15 High Energy Line Break Not applicable.

No high energy line is affected by this activity.3.7.16 Electrical Separation/Isolation Not applicable.

No electrical system is involved in this activity.3.7.17 Single Failure Criteria Not applicable.

No electrical system is involved in this activity.3.7.18 Containment Isolation Not applicable.

The strippable coatings will neither be applied at any containment isolation boundary nor is required to perform any containment isolation function.3.7.19 Material Compatibility The compatibility of the coatings with the existing materials has been tested and evaluated under Ref. 6 and 7.Para. 3.7.7 summnarizes the test conclusions that the coating materials have no adverse effects on the existing materials.

4.0 Effects on Licensing Basis 4.1 After completion, the proposed coating work will not increase the probability of occurrence of the consequence of an accident since there will be no changes to the physical configuration or the operating parameters of the effected systems because all of the coatings are temporary which will be removed after refueling.

4.2 The proposed coating work will not increase the probability of occurrence or the consequences of a malfunction of ITS equipment for the same reasons discussed above.4.3 The proposed coating work will not adversely affect nuclear safety or safe plant operations since all coatings will be removed after refueling and plant startup, along with the additional considerations addressed In para. 3.7.8.44108/339H/l I

I I)/ I L) 106 09 : ýA : ?9 SE 328257-002 Rev. 1 Page 12 of 13 4.4 The proposed coating work does not create a possibility for an accident or malfunction of a different type than previously analyzed.

The configuration and function of each affected system is unchanged.

4.5 The proposed coating work does not reduce the margin of safety as defined In the SAR or any Technical Specification for the reasons discussed above. The coating will prevent water leakage through liner and therefore reduce potential for degradation of concrete and/or drywell.4.6 No Oyster Creek Technical Specification violations are produced by the proposed coating work.4.7 The proposed coating work does not violate any license requirements or regulations.

4.8 Coating work does not produce a radiological concern. The impact of airborne solvent in exhaust air passing through the SGTS in the unlikely event of emergency SGTS activation.

simultaneous with coating, will not significantly affect Iodine releases from the plant.4.9 No changes to the FOSAR are required.4.10 Plant Procedures do not need to be changed.5.0 Effects on Environment 5.1 None. The secondary containment integrity will be maintained during coating application and removal. Therefore no release of coating material is expected during application or removal processes to the outside environment.

Since solvent contents of the coating material is very low, the release of the solvent vapor to the outside environment (via RB ventilation) will not pose any environmental concerns.6.0 Conclusion The proposed coating work will not reduce the performance of the affected systems, affect the safety functions of these systems, increase the probability of occurrence or consequence of an accident, create a possibility for an accident, decrease the margin of safety as defined in the bases of Oyster Creek Technical Specifications, violate any licensing requirements, cause a radiological concern, and will not affect the environmental permits.4410N/339H/112 SE 328257-002 Rev. I Page 13 of 13 7.0 References

1. GE -F&PMT Transmittal

1. 88-178-005

-"Corrosion Evaluation of the OC Cavity Seal" dated March 17, 1988.2. GPUN Tech. Spec. SP-1302-22-006, Rev. 4, "OCNGS -Repair of Reactor Cavity and Storage Pool Lining." 3. GPUN Tech. Spec. SP-1302-56-707, Rev. 0, "OCNGS -Selection Criteria for Temporary Hater Tight Metal Foil for Rx Cavity Liner Repair." 4. GPUN Tech. Spec. SP-1302-56-106 Rev. 0 "OCNGS -Selection Criteria for Temporary Hater Tight Coating for Rx Cavity Liner".5. GPUN Tech. Spec. IS-323505-002, Rev. 0, "OCNGS -Application of Temporary Polymer Coating for Rx Cavity Liner." 6. GPUN TDR #937 Rev. 0, "Test Data on Qualification of Coating Science's SS Metal Foil".7. GPUN TDR #938 Rev. 1, "Test Data on Qualification of Isotron Products".

IR 8. GPUN TDR #713 Rev. 0 -OCNGS Upper Drywell Shield Wall Thermal Analysis.9. Fire Hazard Analysis.

-FPE No. 328257-001 Rev. 0.10. NuCon evaluation of Coating Materials solvent release on SGTS charcoal filter DRF #067072.11. Calculation for leachateslbulk analysis.

-Calculation No.C-1302-243-5340-046.

4410H/339H/13

'0/1019t 09:ý4:J9[ Nucear DOCUMENT NO.SE 328257-002 TrtI Temporary Repair of Rx Cavity Liner REV

SUMMARY

OF CHANGE 2 Substantial changes to include application of temporary polymer coating on carbon steel trough floor in the Rx Cavity. Paragraphs 1.1, 2.3, 3.3, 3.4, 3.5, 3.7.1, 3.7.8, 3.7.9 and 3.7.10 expanded to include the above change.3 Para. 7.0 -Ref. 7 -GPUN TDR 938, Re'. 0 revised to extend the test immersion time from 10 weeks to 12 weeks at 125OF for Isolock 300 modified Latex coating system.I 4 Pages 1 and 2 -Non Substantial Change-Correct Design Docment No. and Rev. No.provided.0 U A0000036 12.83 4 0 E Nuclear MR No 851 Revision Technical Data Report Activit No. page _ of 21 Project: ASSESSM ENT OF OYSTER CREEK Departm ent/Section r _ _ _ _/_ _ er__ _.__ _ _ _ _ _ _ _DRYW.LL S.HELL .~oOe. o ~DRYWELL SHELLRelease Date. ý ýRevision Date______

Document Tile: A"s tJ -F 0 ,/ C /-" --4 Originator Signature Date (s) Date John A. Marti~L L4~Y-LU~-

i Approval for External Distribution Date Does this TDR Include recommendation(s)? -Yes M/o If yes, TFWR/TR#* Distribution Abstract:

STATEMENT OF PROBLEM This TDR covers the taking of the U.T., preparation and removal of core samples gathered in the evaluation from 1986 to November 1987 of the drywell containment pressure vessel shell. The concern for this assessment of the vessel shell thickness arose out of the continued observation of leakage of water from around drywell penetrations at reactor building floor elev. 75'-0" and 23'-0" and the drains from the torus rooms (sand cushion entrenchment).

This observation generated concern regarding possible corrosion of the shell.KEY RESULTS The U.T. data indicated reductions in drywell shell thickness in a band at approximately elevation 10 feet. This band corresponds to the sand entrenchment region of the drywell.Based on data collected in November, 1986 the lowest wall thickness in this area is 0.87 inches. The "as specified" nominal thickness is 1.154 inches.U.T. data taken at elevation 50'-2" indicated some areas of local thinning.

The lowest average thickness at elevation 50'-2" is 0.757 inches in plate that has nominal thickness of 0.770 inches. This is based on data collected in November, 1987.The lowest average thickness at elevation 871-5" is 0.619 inches in plate specified as 0.640 inch plate. This is based on data collected in November, 1987.__(For Additional Space Use Side 2)This Is a report of work conducted by an individual(s) for use by GPU Nuclear Corporation.

Neither GPU Nuclear Corporation nor the authors of the report warrant that the report Is complete or accurate.

Nothing contained In the report establishes company policy or constitutes a commitment by GPU Nuclear Corporation.

• ibt r c O n y Oi iO A 1* Abstract Only A00X)OOMM 1-8 TDR 851 Rev. 0 Page 2 of 21 CONCLUSION At the present time a specification (Ref. 6) has been initiated for continued monitoring of the drywell containment vessel wall thickness as required during unscheduled "outage of opportunites and refueling outages".

This data will be evaluated as it is generated and captured in TDR 948, Statistical Analysis of Drywell Thickness Data.9223d/347 TDR 851 Rev. 0 Page 3 of 21 TABLE OF CONTENTS SECTION DESCRIPTION PAGE 1.0 PURPOSE 3 2.0 METHODS 4 3.0 RESULTS 19

4.0 CONCLUSION

32 5.0 RECOMMENDATIONS 33

6.0 REFERENCES

36 7.0 APPENDICES

1. PHOTOGRAPHS CORE DRILLING 9223d/347 TDR 851 Rev. 0 Page 4 of 21 1.0 PURPOSES This TDR captures the technical information gathered for the evaluation of the drywell containment pressure vessel shell thickness.

The concern to make this assessment of the vessel shell thickness was born out of the continued observation of leakage of water from around drywell penetration observed from reactor building floor elev. 86'-0", 23'-0" and the drains from the torus room (sand cushion entrenchment).

A program was undertaken to accomplish a sampling of thickness readings using ultrasonics at various elevations.

This sampling of data was taken and evaluated at each outage of opportunity (10 foot elevation only). For the purpose of this TDR and as a guide, the data collected will be referred to in the following nominal elevations.(a) Elevation 10 foot (b) Elevation 51 foot (c) Elevation 87 foot Because of these wetting conditions, there was concern that repeated exposure of the drywell steel to water could result in degradation of the drywell in the sand cushion region.2.0 METHODS 2.1 DRYWELL THICKNESS MEASUREMENTS Measurements of the drywell portion of the containment shell were made to verify its thickness during the 11R outage. These measurements were made using UT, a Non Destructive Examination (NDE)method, that is able to accurately determine the thickness of material or presence of abnormalities, i.e., nonmetallic inclusions.

UT plate thickness measurements were made on the Oyster Creek drywell. Approximately 1,000 UT readings were eventually taken utilizing an ultrasonic thickness gauge device (D-meter) (Attachment 1). Measurements were obtained by transmitting ultrasound through the plate and measuring the time it takes for the longitudinal wave mode to travel to a reflector (front wall interface of mid-wall reflector or backwall) and back. Since the electronic measurement of time results in the digital thickness measurement of the first significant sound reflector, the probability of mid-wall reflector being measured versed the backwall is dependent on the size of the reflector relative to the surface area of the ultrasonic transducer.

The larger the mid-wall reflector, the more likely the digital thickness reading will be the mid-wall number, and not the backwall value.9223d/347 TDR 851 Rev. 0 Page 5 of 21 To further characterize the drywell an "A-Scan" UT technique was also employed. "A-Scan" is important for the expanded analysis of the character, location and amplitude of various ultrasound reflectors.

The "A" scan is the ultrasonic indication displayed on cathode ray tube (CRT). The front surface pip or amplitude appears first, and the back surface pip or amplitude appears sometime later in the CRT sweep display. The space between the pips is a measure of the distance between the surfaces.

Pips in between the front and back surfaces may be mid-wall reflectors such as laminations, inclusions or isolated holes and/or pits.Other characteristics of the reflector can be observed by a qualified technician when using an "A" scan that are not available with a D-Meter. Profile of the amplitude, break pattern at the baseline, number of doublets following the amplitude pip, multiples of original reflectors, and amplitude height on the screen and other characteristics all give information that may be useful in analyzing the origins of ultrasound reflectors.

The "D" meter was chosen for the continued surveillance of thickness readings because of its o Accuracy o Ease of reading o Repeatability 2.2 MEASUREMENT LOCATION Initial UT measurements In 1983 were made from the inside of the drywell containment at elevations 51 feet and 10 feet. A digital UT system was used. The measurements opposite the sand cushion at the 10 ft. elevation in the Bays corresponding to where water leaks were observed, indicated that the containment wall was thinner than expected.

Measurements above these areas in the same plate indicated thicknesses within the original plate thickness variability.

Additional UT readings in the same Bay quadrants at elevation 51 indicated no abnormal thickness variations.

Although there are no specific requirements for surveillance of the containment wall thickness, it was considered prudent to make these measurements due to the wetted conditions that had occurred.The above initial measurements were made through the protective coatings on the inside of the containment.

Since the effect of the protective coating on the UT measurements was questioned, special test blocks were made that included the coating material to quantify the effects of the coating on the UT readings.

The accuracy of the UT system was established for the coating thickness of the upper portions of the drywell. The effects of Carboline Carbo-Zinc 11 coating on the accuracy of UT measurements was verified through an experiment conducted by GPUN. Two carbon steel plates approximately 1.15-inch thick and six by six-inch square were coated with Carbo Zinc. One plate had five mils of coating and the other plate had 10 9223d/347 TDR 851 Rev. 0 Page 6 of 21 mils ofcoating.

Both plates had a half inch wide strip on one edge left uncoated.

Both plates were laid out in a half inch grid pattern across the entire partially coated side including the uncoated strip. Similar equipment (D-meter of same make and model)transducers, and couplant as used in the field was utilized and measurements taken. Approximately 149 readings of thickness were taken for each plate. Additionally each grid (excluding the uncoated strip) was measured by Dry Film Technique (DFT) gauge to determine the coating thickness.

The uncoated strip for each was measured by micrometer.

The three readings:

1) ultrasonic (coated and uncoated);

2) dry film technique; and 3) micrometer (uncoated strip) were compiled, averaged and final factors developed.

The uncoated micrometer reading, plus the DFT reading was treated as the true reading of combined thickness.

The UT reading was found to overcall 0.37 for 5 mil coatings and 1.5% for 10 mil coatings after subtracting the DFT reading from the combined UT reading of steel and coating thickness.

It should be noted that the coating application on the test plates and the upper portion of the drywell were consistently uniform. The coating along the basement floor(elevation 10), however, was found to be considerably thicker at locations where UT readings were taken.For this reason the coating was removed and a new set of UT measurements were made in 1986. The new readings continued to indicate that the containment wall was thinner than expected in several areas along the basement floor as with prior measurements, the areas of indicated thinning were adjacent to the sand cushion.2.3 EXTENDED UT MEASUREMENTS As a result of the UT readings taken in 1986 adjacent to the sand cushion being considerably thinner than expected, a program was initiated to obtain detailed measurements to determine the extent and characterization of the thinning.

UT measurements were made in each Bay at the lowest accessible locations.

Where thinning was detected, additional measurements were made in a cross pattern determine the extent. The cross pattern had the lowest reading as the center and was a 1" center with a 5x5 pattern after the cross pattern was completed the lowest reading was then used to expand the UT to a 6x6 grid on V" center with the lowest reading as its center.9223d/347 TDR 851 Rev. 0 Page 7 of 21 To determine the vertical profile of the thinning, trenches were excavated into the floor in Bay 17 and Bay 5. The concrete floor and rebar was removed to expose a portion of the drywell wall about 18 inches wide and sufficiently deep to allow measurement to the bottom of the sand cushion area. Bay 17 was selected since the extent of thinning at the floor level was greatest in that area. It was measured that the thinning at elevations below the initial measurements were no more severe and became less severe at the lower portions of the sand cushion. Bay 5 was excavated to determine if the thinning line was lower than the floor level In areas where no thinning was detected.

Although several inclusions were found, there were no significant indications of thinning.

As a result of the above, the area above the concrete was considered to be conservatively representative of the trenches and no further readngs were taken. A repair specification Ref. 1 was initiated to provide instructions for the repair of the concrete floor after the readings were taken: specifically the repair consisted of filling up the cavity with silicone elastomers in order to restore the insulation properties of the removed concrete.Additional U.T. measurements for the continuous monitoring program will be obtained during future outages to ensure that: 1. Cathodic protection is being properly implemented in the sand bed region. In addition, we cannot monitor C.P.effectiveness without U.T. in the frame area because a reference cell cannot be installed in the frame area from the torus room side of the sand bed.2. Previously uncorroded bays remain that way.3. Finding standing water in the core hole of bay 11 during the C.P. implementation would be properly assessed.2.4 HEAT AFFECTED ZONES & REINFORCEMENT STRUCTURE Other areas of concern requiring additional UT investigation were the plate to plate welds under the torus vents and the vent opening reinforcement plates. These areas were given extra consideration on the basis that material sensitized by welding may have been attacked by a corrosion mechanism with greater damage or cracking occurring at those locations.

The extra UT investigation was conducted at three spots equal distance along side each toe of the vertical plate to plate weld and on either side of the bottom center gusset of the vent opening reinforcement plate.9223d/347 TDR 851 Rev. 0 Page 8 of 21 Eight D-meter thickness measurements were taken at each bay, or Bays 5, 7 and 19. These readings were on each bay of the welds.(4 each hole). At these Bay sites the six locations were also 450 shear tested to interrogate the weld Heat Affected Zone (HAZ). The 450 shear wave test was especially done to detect HAZ cracking.

The top two locations were also the sites from which the plate to torus vent reinforcement plate weld was examined for HAZ cracking.

No crack indications were found and no wastage of the torus vent reinforcement plate was found. The plate to plate weld HAZ as well as the weld when measured using a 6x6 grid indicated wastage similar to the surrounding plate wastage.2.5 ALTERNATE UT TECHNIQUES AND VERIFICATIONS EPRI NDE Center UT personnel were invited to independently analyze the containment vessel plate. Their objective was to independently analyze the conditon of the drywell liner. They scanned two areas using a "Zero Degree Longitudinal Wave Method". One area compared was just above the curb that we indicated had general wastage.Another area was where we had indications of mid-wall deflections or laminar inclusions.

Their observation and measurements independently verified GPUN's results.Mapping of the wall profile indicated a corrosion transition at seven to eight inches up from the concrete curb in Bay 19. This detailed map was generally corroborated by the GE Ultra Image III"C" Scan top graphical mapping system.GPUN experimentally utilized the I.D. Creeper of "30-70-70" technique (a UT integration method) to detect minor changes in back wall surface conditions.

This technique compared "A" scan presentations from one inch thick corroded samples to the results from Bay 13 locations "A" and "E". Reference standards were utilized representing light, moderate and heavy corrosion conditions.

This 30-70-70 technique defined surface roughness conditions by matching "A" Scan presentations from materials that have light, medium and heavy corrosion on their back surfaces.

It was able to verify the roughness condition of wastage and the light corrosion areas of the containment wall.The "A" scan displays from the vessel plate were categorized by comparing them to the reference "A" scan displays.

Location A of Bay 13 (0"-6" up from concrete curb) showed typical "A" scan display of moderate corrosion on average. Local sites of heavy corrosion also were identified.

Bay 13 locations "A" and "E" indicated heavy corrosion between 0 to 6 inches above the curb, moderate corrosion 6 to 14 1/2 inches above the curb, and very low or no corrosion 14 112 to 17 inches above the curb.9223d/347 TDR 851 Rev. 0 Page 9 of 21 2.6 METHODOLOGY OF CORE SAMPLE LOCATION The selection of areas to obtain the core samples was made to evaluate if the UT measurements represented indicated material wastage of if there was localized "pitting".

Those measurement areas that indicated thickness readings of less than half of the thickness expected, i.e., .4 to .7 inches, and had adjacent measurements of the expected thicknesses (nominally 1.154"), were designated as "pitted" areas. Area that had indicated thinning at adjacent measurements were designated as wastage areas. A third area, above the wastage area, and within the sand cushion that appeared to have a thinning or "pits", was also selected as a sample site. The core sampling sequence and logic were to first obtain a sample of a suspected "pitted" area and two samples of a wastage areas but in different bays. Should the "pitted" sample turn out to be an inclusion as suspected from the UT, additional samples of areas that were suspected as being "pitted" would not be required.It was decided, therefore, that core samples should be removed (Ref.2)from the drywell in each of these different regions in order to achieve the following goals: a) Verify UT thickness reading b) Characterize the form of corrosion c) Obtain sand samples and samples or other annulus materials d) If corrosion existed, characterize corrosion products and environment e) Provide access for visual examination of the outside surface of the drywell d) Allow for sampling of sand and/or corrosion products for bacteria With these goals in mind, a first cut was made at selecting regions for sampling of the drywell steel. Twelve regions were selected: four from wastage regions, four from "pitted" regions, two from above the wastage region and two from below the concrete level.These initial selections were, however, modified slightly as the program progressed and additional information became available from ultrasonic testing and initial core sample examinations.

Table 1 identifies each of the seven core sample locations ultimately chosen and the types of samples obtained.9223d/347 TDR 851 Rev. 0 Page 10 of 21 TABLE 1 Core Samples Sample No.1 2 3 4 5 6 7 Bay/ Reason for Location Sample Removal 19C Wastage 15A Pitting/Inclusion 17D Wastage 19A Wastage 11A Wastage llA Minor wastage 19A Minor wastage Elevation 11'-3 5/8" 11'-5 1/4°°11'-3 3/4" 11'-3 3/8" 11 '-31 12"- 2 3/4"1 12'-1" 9223d/347 TDR 851 Rev. 0 Page 11 of 21 2.7 METHODOLOGY FOR MEASUREMENTS OF WALL THICKNESS ABOVE 23' ELEVATION (NOVEMBER 1987)Wall thickness measurements using "D" meter equipment, were taken at elevation 50'2" approximately eleven inches below the seam weld on the joint to the next highest plate. Readings were taken in a one Inch wide circumferential band extending around the drywell.Readings were taken in all accessible areas (areas that could be accessed from existing floors or gratings without scaffolds or equipment removal).

UT readings were obtained on six inch centers. If four consecutive readings (on six inch centers)yielded readings more than 25 mils lower than nominal thickness, the interval between readings was shortened to one inch centers.In addition to this band, "D" meter readings, on six inch centers, were taken in a two foot long one inch wide circumferential band above accessible drywell penetrations between elevations 46'6" and 49'At elevation 87'5" a one inch wide circumferential band was scanned with an "A" scan in all accessible areas to characterize the outside surface of the drywell wall. Readings were also taken with a "D" meter on six inch centers. As done on elevation 50'2", the reading interval was shortened to one inch on center if four consecutive readings were more than 25 mils less than nominal wall.If a drywell penetration intersected the inspected band at elevation 87'5" then an additional two foot band (centered one foot on each side of the penetration) located six inches below the penetration was inspected.

In three areas on the 50'2" elevation and four areas of the 87'5" elevation, a six inch grid of 49 UT measurements was taken to provide additional data on the extent of wall thinning.2.8 METHODOLOGY FOR ASSESSMENT OF CORROSION Assessment of corrosion was performed by removing two core samples from elevation 50'2" (Table 2). A metallurgical assessment of the plugs was performed to characterize the form of corrosion, obtain Firebar samples, characterize corrosion products and environment and provide access for visual examination of the outside surface of the drywell if a gap exists between the Firebar and the drywell (Ref. 5 & 6).9223d/347 TDR 851 Rev. 0 Page 12 of 21 TABLE 2 50'2" ELEVATION THICKNESS EVALUATION Sample No.8 9 Location Bay 5 Bay 7 Type of Sample Uniform Thinning Uniformly at or above nominal with low spots 2.9 METHODOLOGY FOR SELECTING THE SIZE OF THE DRILLED (CORED) HOLE The selection of a 2" dia. steel core from the drywell containment wall was chosen to facilitate, I .2.3.4.Surfaces meaningful to evaluate the corrosion mechanism.

A hole large enough to facilitate examining the backside of the drilled hole with a miniature video camera.A hole large enough to extract sand samples.Routine test and repair of drilled hole. A larger opening would have required a more complex plug design to restore the structure to its original condition.

9223d/347 TDR 851 Rev. 0 Page 13 of 21 2.10 METHODOLOGY FOR DRILLING OF CORE SAMPLE It was agreed to drill the carbon steel area with a 2" Milwaukee"STEEL HANG" with carbide teeth(Ref.

2). This drill bit was combined with an electric drill motor attached to a magnetic base for positioning on the drywell wall. A drilling sequence was developed to keep the temperature of the plug sample during the drilling operation below 150°F. so that the sample could be evaluated for Microbiologically Induced Corrosion.

This was accomplished by using two (2) drill bit assemblies.

The first drill assembly was used with a self contained distilled water spray. The second drill assembly was specially designed and rigged with a magnet slotted to fit inside of the drill bit to attach to the plug so that it would not fall through the newly drilled hole.This second phase of the drilling operation was done slowly and without coolant to keep from contaminating the plug core backside so that it could be evaluated in its pristine state by the laboratory.

This design combined with operator skill allowed a clean even cut of the plug sample. In addition, at no time did the sample temperature exceed 150°F.Prior to doing the actual drilling, the drill operation was qualified with identical equipment.

This training session further guaranteed a successful operation.

2.11 METHODOLOGY FOR REPAIRING DRILLED (CORED) HOLE Repair of 2" dia. hole in drywell wall was accomplished using specification (Ref. 3) which outlined the plug manufacture requirements and instructions for replacement of the drywell core samples. Specifically, the plug replacement covered the following sequence: o Filling sample core hole (optional) o Surface preparation of the hole o Manufacturing of the plugs o Welding o Post Weld NDE o Leak Testing o Painting and corrosion protection o Preparation of final document package In addition the design of the drywell plug and groove weld stresses was shown by calculation and verified (Ref. 4).9223d/347 TDR 851 Rev. 0 Page 14 of 21 3.0 RESULTS 3.1 DATA

SUMMARY

(Below 23' Elevation)

The thickness measurements obtained adjacent to the sand cushion are tabulated on GPUN drawing number 3E-SK-S-85 (Ref. 7). Initial measurements were taken at four locations near the lower curb at each torus vent. Four locations, A-B-C-D, were selected to provide two thickness measurements of the left and right drywell plates that make up each Bay section. Each tabulation heading defines the location of the tabulated matrix of measurements with respect to the top of the curb and to the weld between the two plates at the center of the vent line. The matrix of measurements are at one inch increments both vertical and horizontal.

Those measurements around heat affected zones and on the vent line reinforcement were taken one inch on each side of the weld. No degradation or wastage was indicated on the reinforcement plate or around the reinforcement plate to the containment plate weld. Wall thinning indications on the containment plate on each side of the containment plate weld were the same magnitude as surrounding areas indicating that the weld heat effected zone did not cause or accelerate wastage.3.1.1 U.T. Data Reduction UT drywell thickness data was collected in each of the ten bays. The UT data is presented on GPUN Drawing NO.3E-SK-5-85 (Ref. 7). The primary concentration of data was within a 6 inch wide circumferential band above the drywell floor curb since data above this band indicated minimal wastage of the drywell wall material due to lack of sand bed region.3.2 UT DATA INTERPRETATION (Below 23' Elevation)

Prior to core sample removal possible causes of the low UT thickness readings were attributed to external corrosion, laminations or a field of inclusions within the plate. Because the very low readings were localized it was expected that they would be a result of laminations.

The general wastage, however, extended from plate to plate and the affected areas of the shell were within the sand bed only. Thus it was concluded that the plate thinning was most likely due to corrosion.

In addition, a qualitative assessment of the plate condition was made using an "A" scan presentation with a 5 mghz transducer.

This data was also indicative of corrosion on the outside.9223d/347 TDR 851 Rev. 0 Page 15 of 21 Numerous ultrasonic thickness readings were taken in all bays in the drywell particularly at the elevation of 11' 3". Review of this ultrasonic test data showed that significant corrosion damage appeared to be confined to regions in Bays 11, 13, 17 and 19.Furthermore, the thinned parts of the drywell were limited to those areas which were in contact with the sand bed from elevation 10' to 11'9". Numerical analysis of this data determined the minimum mean remaining wall thickness was 0.87".UT thickness readings below the concrete floor elevation showed the thickness to be greater than 0.87" and at the bottom of the sand bed to be nearly nominal design thickness.

After the completion of the ultrasonic testing (UT) of each of the drywell bays above the concrete floor, the data was assembled and reviewed.

This data indicated that there were at least three regions which showed different characteristics.

One set of data showed regions of overall general wall reduction which we characterized as wastage. Another set showed regions with little or no general wall reduction but localized areas with large wall reduction which we characterized as pitting/inclusions.

The last set of data showed regions of little or no wall reduction and no random large reductions, which we characterized as minor wastage.The characterization of each bay is summarized in Table 3.9223d/347 TDR 851 Rev. 0 Page 16 of 21 TABLE 3 1 Minor wastage 3 Minor wastage 5 Pitting/Inclusion 7 Minor wastage 9 Plttlng/inclusion 11 Wastage 13 Wastage 15 Pitting/inclusion 17 Wastage 19 Wastage 9223d/347 TDR 851 Rev. 0 Page 17 of 21 In addition to the above general characterizations, it was also observed from the UT readings that above an elevation of approximately 11'9" the wall thickness would return to the nominal value. This occurred been though the readings were still within the sand bed and there was wastage below this elevation.

Likewise, there were regions of the sand bed below the concrete which heretofore had not been ultrasonic tested and hence no characterization could be made.3.3 UT MEASUREMENTS (Above 23' Elevation)

UT data obtained at the upper elevations of the drywell is presented on GPUN Sketch Drawing No. 3E-SK-S-89 (Ref. 8).3.3.1 MEASUREMENTS AT ELEVATION 501-2" Data was taken on a one inch wide circumferential band at elevation 50'2" covered over half of the drywell's circumference.

Approximately 230 readings, on six inch centers, were taken in plate specified as 0.770 inch. 90%of the readings were within 25 mils of specified wall thickness.

Approximately 30 readings were taken on plate specified as 1.063 inch. Three readings were less than nominal. Of these two isolated readings were more than 25 mils below nominal. The wall thickness for the 0.770 inch specified wall plates ranged from 0.705 to 0.800 inch. The wall thickness for the 1.063 inch plates ranged from 1.04 to 1.11 inches. There were two areas where the reading interval was shortened to one inch due to consecutive low readings (as outlined in Section 2). The two areas were at approximate azimuth 188 to 194° (area 1) and between approximate azimuth 63 to 66° (area 2) (see GPUN Dwg.3E-SK-S-89 Ref. 8).Area 1 included several of the lowest readings (.705 inch)and is directly above Bay 11 and below the refueling bellows drain cover plate. To surround the lowest reading, five additional measurements one inch above and five measurements one inch below the site were taken on one inch centers. The thickness readings in this area ranged from .705 to .770 indicating that the thickness of the wall was not uniform.The area 2 readings indicated an approximately uniform wall thickness ranging from .730 inch to .755 inch.All of the 49 points in a grid were averaged.9223d/347 TDR 851 Rev. 0 Page 18 of 21 3.3.2 MEASUREMENTS AT ELEVATION 8715"1 All of the plates at this elevation are specified as .640 inch thick. Data was taken on a one inch band at elevation 87'5" covering approximately 75% of the drywell circumference. "A" scan presentation was relatively smooth with occasional depressions.

Approximately 150 "D" meter readings, on six inch centers were taken. About 90% of the readings were within 25 mils of nominal wall. All of the low readings were isolated with the single lowest reading being 0.540 inch. There were no instances where consecutive low readings required the interval between readings to be shortened.

Of the grids at this elevation the lowest average thickness was .619 inch. Incremental averaging of the data in a circumferential band has yielded a minimum average within three mils (lower) than the 0.619 inchminimum grid average.3.4 ASSESSMENT OF WALL THICKNESS MEASUREMENTS AT ELEVATION 50'2" The "DV meter measurements indicated some thinning of the drywell shell.3.5 ASSESSMENT OF WALL THICKNESS MEASUREMENTS AT ELEVATION 8751" A-scan of the evaluated band indicated a smooth outside surface with occasional depressions.

'D" meter readings indicated some wall thinning.3.6 CORE PLUG SELECTION AND VISUAL ASSESSMENT The basis for selection of the core plug locations is as follows: Plug 8 was removed from an area of apparent general thinning.

Plug 9 was removed from an area where the UT indicated that nominal thickness or above existed with isolated low readings.

This plug was centered between a reading of 0.798" and 0.710'. See Table 2 for details.Both plugs removed from the elevation had surface corrosion.

The Firebar in the region of the drywell surface for both plugs was"chunky" and denser than the Firebar toward the concrete which fell apart rather easily. There was no visible gap between the Firebar and drywell. There was no visible evidence of water or moisture on either plug or Firebar sample.9223d/347 TDR 851 Rev. 0 Page 19 of 21 3.7 REPAIR OF DRILLED (CORED) HOLE The openings in the Drywell wall were repaired and sealed with a special design and fabricated steel plug. The final repairs were accepted by the Authorized Nuclear Inspector (ANI) after successfully completing a visual magnetic particle examination and a vacuum box bubble test on each plug weld. In addition a local leak rate test was conducted on each plug and met the integrated leak rate requirements of the Code of Federal Regulations 1OCFRSO Appendix J. Actual leak rate measurement at each plug was standard liters per minute at 35 psi.3.7.1 The repair of all seven (7) core sample holes below the 23'level (Dec. '86) was accomplished with no undue problems.Two holes were found with Indications using magnaflux and weld repaired and one hole was found with a pin hole leak using vacuum box test and weld repaired.

A successful local leak rate test was performed using a special test cup held in place with the same magnetic drill used for the drilling operations.

3.7.2 The repair of the two (2) core sample holes at 51' level (Nov. '87) was again accomplished with no undue problems.The repaired area was magnafluxed, vacuum box tested and given a successful leak rate tested.3.8 In December 1988 the scope of UT was expanded to: 1) Provide basis for verification of CP effectiveness.

2) Provide baseline data to monitor these sand bed areas not protected by CP.3) Verify that the presence of standby water in the sand beds did not adversely affect previous results.

4.0 CONCLUSION

4.1 The ultrasonic thickness probing of the drywell containment has been confirmed to give accurate results with physical measurement of the plug thicknesses being consistent with UT but, in general, about 2%greater. Therefore, the UT measurements have been a conservative assessment of thickness.

4.2 In the sand entrenchment region, broad areas of exterior corrosion seem to be localized at an elevation corresponding to the exterior sand cushion.9223d/347 TDR 851 Rev. 0 Page 20 of 21 Measurements of drywell thickness below the level of the interior concrete floor (which were made by removal of the interior concrete at two locations down to a depth of about two feet, bay 5 and 17)show that wastage below the floor level is no greater than measured just above floor level. In fact, measurements at the location where general wastage was indicated above the floor show the drywell below the floor to be about 50 mils thicker than the immediately adjacent above floor area.4.3 At elevations 50-2" and 87'5" the wall loss is 33 mils and 46 mils respectively.

This is estimated from the "average" drywell wall thickness in areas of general wall thinning compared to the maximum encountered wall thickness.

Use of an "average" wall thickness is appropriate in evaluating the shell strength; individualized localized pits will not alter the structural integrity of the drywell.4.4 Details of the UT measurements, metallurgical results, and chemical analyses are more fully summarized in TDR 854 (Ref. 5).5.0 RECOMMENDATIONS 5.1 Eliminate or re-direct water intrusions.

This effort is ongoing via BA's 328257 and 323505.5.2 A cathodic protection system has been selected to avert the corrosion in the sandbed. This is being installed via BA 402873.5.3 Monitor (UT) the drywell containment vessel wall as required during an unscheduled "Outage of Opportunity".

This is established by means of Ref. 6. & Ref. 9

6.0 REFERENCES

1 GPUN OCIS-328227-003, Repair of Concrete Floor Removed for U.T.Readings 2 GPUN IS-328227-001, Drywell shell Vessel Sample 3 GPUN IS-328227-002, Replacement or Drywell Vessel Core Sample Plugs 4 GPUN C-1302-243-5310-030, Calculation of Drywell Plug and Groove Weld Stress 5 TDR 854, Drywell Corrosion Assessment 6 GPUN IS-325227-004, Functional Requirements for Drywell Thickness Evaluations 7 GPUN DRG 3E-SK-S-85, Ultrasonic Testing Level ll'-6" 9223d/347 TDR 851 Rev. 0 Page 21 of 21 8 GPUN DRG 3E-SK-S-89, Ultrasonic Testing Level 50'-2" -87'-5" 9 GPUN Memo 5360-88-304 Rev. 1 dated 11/22/88 Expanded UT & Thickness Inspection of Drywell.9223d/347

,.Alp I -W IR91"ll ((pzý.ý dmw C..

((ANA4Y t"9 C tab

""tw (C W.V et C VM4

('40jo r I (

(4

7AL SiAY(IVA N A AOL pu ((V*ýft (

7JL 5j47(I6A (e (

11.ILWI, K MSA (

I a -MAY (1"/C IN q (.o %.( A (..A,-I 9*'ý4s"""tot (wS

'I.'(DAY 1,4 4 C Cd*47Y WA (AN (VA.C 4-

'Ul (r A if (A-. ah, aik, XWC4 (-tit-(, ISA C (7 A ow- I 4¢fvkai 41C r My~114 Wit,

(.(i C

()(C (i (i (.

C ( (

((-C.

(C (

((J (

((.(

C, C (

((, C/

(-((

((C S Nuclear Calculation Sheet at CIC NoRvN" 5 -tN SS~c.ct STATISTICAL ANALYSIS OF DRYWELL .T.DAATHRU 12-31-88_______

1.0 PROBLE. STATEMENT 1.1 Background The design of the carbon steel drywell includes a sand bed which is located around the outside circumference between elevations 8'-11-1/4" and 12'-3". Leakage was observed from the sand bed drains during the 1980, 1983 and 1986 refueling outages indicating that water had intruded into the annular region between the drywell shell and the concrete shield wall.The drywell shell was inspected in 1986 during the 1OR outage to determine if corrosion was occurring.

The inspection methods, results and conclusions are documented in Ref. 3.1, 3.2, and 3.3.As a result of these inspections it was concluded that a long term monitoring program would be established.

This program includes repetitive Ultrasonic Thickness (UT) measurements in the sand bed region at a nominal elevation of 11'-3I1 in bays 11A, 1iC, 17D, 19A, 19B. and 19C.The continued presence of water in the sand bed raised concerns of potential corrosion at higher elevations.

Therefore, UT measurements were taken at the 51' and 87' elevations in November 1987 during the 11R outage. As a result of these inspections, repetitive measurements in Bay 5 at elevation 51' and in Bays 9, 13 and 15 at the 811 elevation were added to the long term monitoring program to confirm that corrosion is not occurring at these higher elevations.

A cathodic protection system is being installed in selected regions of the sand bed during the 12R outage to minimize corrosion of the drywell. The long term monitoring program was also expanded during the 12R outage to include measurements in the sand bed region of Bays 1D, 3D, 5D, 7D, 9A, 13A, 13C, 13D, 15A, 15D and 17A which are not covered by the cathodic protection system. It also includes measurements in the sand bed region between Bays 17 and 19 which is covered by the cathodic protection system, but does not have a reference electrode to monitor its effectiveness in this region.Some measurements in the long term monitoring program are to be taken at each outage of opportunity, while others are taken during each refueling outage. The functional requirements for these inspections are documented in Ref. 3.4. The primary purpose of the UT measurements in the sand bed region is to determine the corrosion rate and monitor it over time. When the cathodic protection system is installed and operating, these data will be used to monitor its effectiveness.

The purpose of the measurements at other locations is to confirm that corrosion is not occurring in those regions.N 0016 (06-86)

Caic. No. C-1302-187-530C-005 Rev. No. 0 Page A of 1.2 Purpose The purpose of this calculation is to: (1) Statistically analyze the thickness measurements for Bays IIA, IIC, 17D, 19A, 19B and 19C in the sand bed region to determine the mean thickness and corrosion rate.(2) Statistically analyze the thickness measurements for Bay 5 at elevation 51' and Bays 9, 13 and 15 at elevation 87' to determine the mean thickness corrosion rate.(3) To the extent possible, statistically analyze the limited data for the 6" x 6" grids in the sand bed region of Bays 9D, 13A, 15D and 17A to calculate the mean thickness and determine if there is ongoing corrosion.

(4) To the extent possible, statistically analyze the limited data for the 6" x I" horizontal strips in the sand bed region of Bays ID, 3D, 5D, 7D, SA, 13C and 15A to calculate the mean thickness and determine if there is ongoing corrosion.

Statistically compare the thickness data from December 1986 and December 1988 for the trench in Bay 17D to calculate the mean thickness at various elevations in the trench and determine if there is ongoing corrosion.

(5) Statistically analyze the thickness data from December 1988 for the Frame Cutout between Bays 17 and 19 to calculate the mean thickness.

10/251D 1 C;7: Caic. No. C-1302-187-5300-005 Rev. No. 0 Page 5 of j'2.0 SUMMkRY OF RESULTS S Bay & Area Location Corrosion Rate**Mean Thickness"*

2.1 6"x6" Grids in Sand Bed Region at Original Locations 11A I1C 17D 19A 19B 19C Sand Sand Sand Sand Sand Sand Bed Bed Bed Bed Bed Bed Not significant Indeterminable

-27.6 +6.1 mpy-23.7 +4.3 mpy-29.2 +0.5 mpy-25.9 +4.1 mpy 2.2 6"x6"' Grids in Sand Bed Recion at New Locations 9D 13A 15D 17A Sand Bed Sand Bed Sand Bed Sand Bed Indeterminable*

Not significant*

Possible*Indeterminable*

908.6 916.6 864.8 837.9 856.5 860.9 1021.4 905.3 1056.0 957.4 750.0 620.3 635.6 634.8*5.0 mils+10.4 mils+6.8 mils+4.8 mils+0.5 mile+4.0 mile+9.7+10.1+9.1+9.2 mile mils mile mile 2.3 6"x6" Grids at Upper Elevations 5 9 13 15 51' Elev.87' Elev.87' Elev.87' Elev.-4.3 +0.03 mpy Not significant Not significant Not significant

+0.02 mile+1.0 mils+0.7 mile 70.7 mile 2.4 Multiple 6"x6" Grids in Trench 17D 171/9 Trench Frame Cutout Not significant*

Indeterminable' 981.2 +6.7 mils 981.7 +4.4 mile 2.5 6" Strips in Sand Bed Region ID 3D 50 7D 9A 13C 130 15A Sand Sand Sand Sand Sand Sand Sand Sand Bed Bed Bed Bed Bed Bed Bed Bed Indeterminable*

Not significant*

Not significant*

Possible*Indeterminable*

Not significant*

Not significant*

Not significant*

1114.7 1177.7 1174.0 1135.1 1154.6 1147.4 962.1 1120.0+30.6 mils+5.6 mile+2.2 mils+4.9 mile+4.8 mile+3.7 mile+22.3 mile+12.6 mils 2.6 Evaluation of Individual Measurements Below 800 Mils One data point in Bay 19A and one data point in Bay 5 Elev. 51' fell outside the 99% confidence interval and thus are statistically different from the mean thickness.

• Based on limited data. See text for interpretation.

"*Mean corrosion rate in mils per year + standard error of the mean***Current mean thickness in mils + standard error of the mean Caic. No. C-1302-187-5300-005 Rev. No. 0 Page 4

3.0 REFERENCES

3.1 GPUN Safety Evaluation SE-000243-002, Rev. 0, "Drywell Steel Shell Plate Thickness Reduction at the Base Sana Cushion Entrenchment Region" 3.2 GPUN TDR 854, Rev. 0, "Drywell Corrosion Assessment" 3.3 GPUN TDR 851, Rev. 0, "Assessment of Oyster Creek Drywell Shell" 3.4 GPUN Installation Specification IS-328227-004, Rev. 3, "Functional Requirements for Drywell Containment Vessel Thickness Examination" 3.5 Applied Regression Analysis, 2nd Edition, N.R. Draper & H. Smith, John Wiley & Sons, 1981 3.6 Statistical Concepts and Methods G.K. Bhattacharyya

& R.A. Johnson, John Wiley & sons, 1977 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page~of 5o 4.0 ASSYUMPTIONS

& BASIC DATA 4.1 Selection of Areas to be Monitored A program was initiated during the 1IR outage to characterize the corrosion and to determine its extent. The details of this inspection program are documented in Ref. 3.3. The greatest corrosion was found via UT measurements in the sand bed region at the lowest accessible locations.

Where thinning was detected, additional measurements were made in a cross pattern at the thinnest section to determine the extent in the vertical and horizontal directions.

Having found the thinnest locations, measurements were made over a 6"-6" grid.To determine the vertical profile of the thinning, a trench was excavated into the floor in Bay 17 and Bay 5. Bay 17 was selected since the extent of thinning at the floor level was greatest in that area. It was determined that the thinning below the top of the curb was no more severe than above the curb, and became less severe at the lower portions of the sand cushion. Bay 5 was excavated to determine if the thinning line was lower than the floor level in areas where no thinning was detected above the floor. There were no significant indications of thinning in Bay 5.It was on the basis of these findings that the 6"x6" grids in Bays 1IA, l1C, 17D, 19A, 19B and 19C were selected as representative locations for longer term monitoring.

The initial measurements at these locations were taken in December 1986 without a template or markings to identify the location of each measurement.

Subsequently, the location of the 6"x6" grids were permanently marked on the drywell shell and a template is used in conjunction with these markings to locate the UTprobe for successive measurements.

Analyses have shown that including the non-template data in the data base creates a significant variability in the thickness data. Therefore, to minimize the effects of probe location, only those data sets taken with the template are included in the analyses.The presence of water in the sand bed also raised concern of potential corrosion at higher elevations.

Therefore, UT measurements were taken at the 51' and 87' elevations in 1987 during the IlM outage. The measurements were taken in a band on 6-inch centers at all accessible regions at these elevations.

Where these measurements indicated potential corrosion, the measurements spacing was reduced to 1-inch on centers. If these additional readings indicated potential corrosion, measurements were taken on a 6"x6" grid using the template.

It was on the basis of these inspections that the 6"x6" grids in Bay 5 at elevation 51'and in bays 9, 13 and 15 at the 87' elevation were selected as representative locations for long term monitoring.

Calc. No. C-1302-187-5300-005 Rev. No. 0 Page & of The long term monitoring program was expanded as follows during the 12R outage: (1) Measurements on 6"x6" grids in the sand bed region of Bays 9D, 13A, 15D and 37A. The basis for selecting these locations is that they were originally considered for cathodic protection but are not included in the system being installed.

(2) Measurements on 1-inch centers along a 6-inch horizontal strip in the sand bed region of Bays ID, 3D, 5D, 7D, 9A, 13C, and 15A. These locations were selected on the basis that they are representative of regions which have experienced nominal corrosion and are not within the scope of the cathodic protection system.(3) A 6"x6" grid in the curb cutout between Bays 17 and 19. The purpose of these measurements is to monitor corrosion in this region which is covered by the cathodic protection system but does not have a reference electrode to monitor its performance.

4.2 UT Measurements The UT measurements within the scope of the long term monitoring program are performed in accordance with Ref. 3.4. This involves taking UT measurements using a template with 49 holes laid out on a 6"x6" grid with I" between centers on both axes. The center row is used in those bays where only 7 measurements are made along a 6-inch horizontal strip.The first set of measurements were made in December 1986 without the use of a template.

Ref. 3.4 specifies that for all subsequent readings, QA shall verify that locations of UT measurements performed are within +1/4" of the location of the 1986 UT measurements.

It also specifies that all subsequent measurements are to be within +1/8" of the designated locations.

4.3 Data at Plug Locations Seven core samples, each approximately two inches in diameter were removed from the drywell vessel shell. These samples were evaluated in Ref. 3.2. Five of these samples were removed within the 6"x6" grids for Bays 11A, 17D, 19A, 19C and Bay 5 at elevation 51'. These locations were repaired by welding a plug in each hole. Since these plugs are not representative of the drywell shell, UT measurements at these locations on the 6"x6" grid must be dropped from each data set.

1 0125/0D6 '.4 :21 9: 0 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 7of .The following specific grid points have been deleted: Bay Area Points 11A 23, 24, 30, 31 17D 15, 16, 22, 23 19A 24, 25, 31, 32 19C 20, 26, 27, 33, 5 20, 26, 27, 28, 33, 34, 35 4.4 Bases for Statistical Analysis of 6"x6" Grid Data 4.4.1 Assumptions The statistical evaluation of the UT measurement data to determine the corrosion rate at each location is based on the following assumptions:

(1) Characterization of the scattering of data over each 6"x6" grid is such that the thickness measurements are normally distributed.

(2) Once the distribution of data for each 6"x6" grid is found to be normal, then the mean value of the thickness is the appropriate representation of the average condition.

(3) A decrease in the mean value of the thickness with time is representative of the corrosion occurring within the 6"x6" grid.(4) If corrosion has ceased, the mean value of the thickness will not vary with time except for random errors in the UT measurements.

(5) If corrosion is continuing at a constant rate, the mean thickness will decrease linearly with time. In this case, linear regression analysis can be used to fit the mean thickness values for a given zone to a straight line as a function of time. The corrosion rate is equal to the slope of the line.The validity of these assumptions is assured by: (a) Using more than 30 data points per 6"x6" grid (b) Testing the data for normality at each 6"x6" grid location.(c) Testing the regression equation as an appropriate model to describe the corrosion rate.

IOIZS /06 74:28 ! 7 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page g of, These tests are discussed in the following section. In cases where one or more of these assumptions proves to be invalid, non-parametric analytical techniques can be used to evaluate the data.4.4.2 Statistical Approach The following steps are performed to test and evaluate the UT measurement data for those locations where 6"x6" grid data has been taken at least three times: (I) Edit each 49 point data set by setting all invalid points to zero. Invalid points are those which are declared invalid by the UT operator or are at a plug location. (The computer programs used in the following steps ignore all zero thickness data points.)(2) Perform a chi-squared goodness of fit test of each 49 point data set to ensure that the assumption of normality is valid at the 95% and 99% confidence levels.(3) Calculate the mean thickness of each 49 point data set.(4) Using the mean thickness values for each 6"x6" grid, perform linear regression analysis over time at each location.(a) Perform F-test for significance of regression at the 95% confidence level. The result of this test indicates whether or not the regression model is more appropriate than the mean model. In other words, it tests to see if the variation of the regression model is statistically significant over that of a mean model.(b) Calculate the co-efficient of determination (R') to assess how well the regression model explains the percentage of total error and thus how useful the regression line will be as a predictor.(c) Determine if the residual values for the regression equations are normally distributed.(d) If the regression model is found to be appropriate, calculate the y-intercept, the slope and their respective standard errors.The y-intercept represents the fitted mean thickness at time zero, the slope represents

` W 2; 5/06 2 4 : 20 : 07 Calc. No. C-13C2-187-5300-005 Rev. No. 0 Page I of the corrosion rate, and the standard errors represent the uncertainty or random error of these two parameters.

(5) Use a z score of 2.58 and the standard deviation to establish a 99% confidence interval about the mean thickness values for each 6"x6" grid location to determine whether low thickness measurements or"outliers" are statistically significant.

If the data points are greater than the 99% lower confidence limit, then the difference between the value and the mean is deemed to be due to expected random error.However, if the data point is less than the lower 99%confidence limit, this implies that the difference is statistically significant and is probably not due to chance.4.5 Analysis of Two 6"x6" Grid Data Sets Regression analysis is inappropriate when data is available at only two points in time. However, the t-Test can be used to determine if the means of the two data sets are statistically different.

4.5.1 Assumptions This analysis is based upon the following assumptions:

(1) The data in each data set is normally distributed.

(2) The variances of the two data sets are equal.4.5.2 Statistical ApProach The evaluation takes place in three steps: (1) Perform a chi-squared test of each data set to ensure that the assumption of normality is valid at the 95%and 99% confidence levels.(2) Perform an F-test of the two data sets being compared to ensure that the assumption of equal variances is valid at the 95% and 99% confidence levels.(3) Perform a two-tailed t-Test for two independent samples to determine if the means of the two data sets are statistically different at the 0.05 and 0.01 levels of significance.

A conclusion that the means are not statistically different is interpreted to mean that significant corrosion did not occur over the time period represented by the data.However, if equality of the means is rejected, this implies that the difference is statistically significant and could be due to corrosion.

I102510- 14:28:07 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 10 of 4.6 Analysis of Single 6"x6" Grid Data Set In those cases where a 6"x6" data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken in 1986 to identify the thinnest regions of the drywell shell in the sand bed region. For the most part, these are single point measurements which were taken in the vicinity of the 49-point data set, but not at the exact location.

Therefore, rigorous statistical analysis of these single data sets is impossible.

However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 49-point data set, the t-test can be used to compare the means of the two data sets as described in paragraph 2.5.When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed.4.6.1 AssMptions The comparison of a single 49-point data sets with previous data from the same vicinity is based on the following assumptions:

(1) Characterization of the scattering of data over the 6"x6" grid is such that the thickness measurements are normally distributed.

(2) Once the distribution of data for the 6"x6" grid is found to be normal, then the mean value of the thickness is the appropriate representation of the average condition.

(3) The prior data is representative of the condition at this location in 1986.4.6.2 Statistical Approach The evaluation takes place in four steps: (1) Perform a chi-squared test of each data set to ensure that the assumption of normality is valid at the 95%and 99% confidence levels.(2) Calculate the mean and the standard error of the mean of the 49-point data set.(3) Determine the two-tailed t value from a t distribution table at levels of significance of 0.05 and 0.01 for n-I degrees of freedom.

14:28:07 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page jf of (4) Use the t value and the standard error of the mean to calculate the 95% and 99% confidence intervals about the mean of the 49-point data set.(5) Compare the prior data point(s) with these confidence intervals about the mean of the 49-point data sets.If the prior data falls within the 95% confidence intervals, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If it falls within the 99%confidence limits but not within the 95% confidence limits, this implication is not as strong. In either case, the corrosion rate will be interpreted to be "Not Significant".

If the prior data falls above the upper 99% confidence limit, it could mean either of two things: (1) significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 49-point data set in 1986. There is no way to differentiate between the two.In this case, the corrosion rate will be interpreted to be"Possible".

If the prior data falls below the lower 99% confidence limit, it means that it is not representative of the condition at this location in 1986. In this case, the corrosion rate will be interpreted to be "Indeterminable".

4.7 Analysis of Single 7-Point Data Set In those cases where a 7-point data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken in 1986 to identify the thinnest regions of the drywell shell in the sand bed region. For the most part, these are single point measurements which were taken in the vicinity of the 7-point data sets, but not at the exact locations.

However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 7-point data set, the t-test can be used to compare the means of the two data sets as described in paragraph 2.5.When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed.4.7.1 Assumptions The comparison of a single 7-point data sets with previous data from the same vicinity is based on the following assumptions:

(1) The corrosion in the region of each 7-point data set is normally distributed.

'0125/Of-14 : 28 : G-, Calc. No. C-1302-187-5300-005 Rev. No. C Page 2-of f2) The prior data is representative of the condition at this location in 1986.The validity of these assumptions cannot be verified.4.7.2. Statistical Approach The evaluation takes place in four steps: (1) Calculate the mean and the standard error of the mean of the 7-point data set.(2) Determine the two-tailed t value using the t distribution tables at levels of significance of 0.05 and 0.01 for n-i degrees of freedom.(3) Use the t value and the standard error of the mean to calculate the 95% and 99% confidence intervals about the mean of the 7-point data set.(4) Compare the prior data point(s) with these confidence intervals about the mean of the 7-point data sets.If the prior data falls within the 95% confidence intervals, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If it falls within the 99%confidence limits but not within the 95% confidence limits, this implication Is not as strong. In either case, the corrosion rate will be interpreted to be "Not Significant".

If the prior data falls above the upper 99% confidence interval, it could mean either of two thingsi (1)significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 7-point data set in 1986. There is no way to differentiate between the two. In this case, the corrosion rate will be interpreted to be "Possible".

If the prior data falls below the lower 99% confidence limit, it means that it is not representative of the condition at this location in 1986. In this case, the corrosion rate will be interpreted to be "Indeterminable".

4.8 Evaluation of Drywell Mean Thickness This section defines the methods used to evaluate the drywell thickness at each location within the scope of the long term monitoring program.4.8.1 Evaluation of Mean Thickness Using Regression Analysis The following procedure is used to evaluate the drywell mean thickness at those locations where regression analysis has been deemed to be more appropriate than the mean model.

10125/!C 14:2S:01 Calc. No. C-1302-157-5300-005 Rev. No. 0 Page j3of (I) The best estimate of the mean thickness at these locations is the point on the regression line corresponding to the time when the most recent set of measurements was taken. In the SAS Regression Analysis output (Ref. 3.7), this is the last value in the column labeled "PREDICT VALUE".(2) The best estimate of the standard error of the mean thickness is the standard error of the predicted value used above. In the SAS Regression Analysis output, this is the last value in the column labeled"STD ERR PREDICT".(3) The two-sided 95% confidence interval about the mean thickness is equal to the mean thickness plus or minus t times the estimated standard error of the mean. This is the interval for which we have 95%confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for equal tails at n-2 degrees of freedom and 0.05 level of significance, where n is the number of sets of measurements used in the regression analysis.

The degrees of freedom is equal to n-2 because two parameters (the y-intercept and the slope) are calculated in the regression analysis with n mean thicknesses as input.(4) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95%confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for one tail at n-2 degrees of freedom and 0.05 level of significance.

4.8.2 Evaluation of Mean Thickness Using Mean Model The following procedure is used to evaluate the drywell mean thickness at those locations where the mean model is deemed to be more appropriate than the linear regression model. This method is consistent with that used to evaluate the mean thickness using the regression model.(1) Calculate the mean of each set of UT thickness measurements.

(2) Sum the means of the sets and divide by the number of sets to calculate the grand mean. This is the best estimate of the mean thickness.

In the SAS Regression Analysis output (Ref. 3.7), this is the___ value labelled "DEP MEAN".

I '0/2ý1/06

'4:Z8:07 Caic. No. C-1302-187-5300-005 Rev. No. 0 Page 14 of (3) Using the means of the sets from (1) as input, calculate the standard error. This is the best estimate of the standard error of the mean thickness.

(4) The two-sided 95% confidence interval about the mean thickness is equal to the mean thickness plus or minus t times the estimated standard error of the mean. This is the interval for which we have 95%confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for equal tails at n-i degrees of freedom and 0.05 level of significance.

(5) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95%confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for one tail at n-1 degrees of freedom and 0.05 level of significance.

4.8.3 Evaluation of Mean Thickness Using Single Data Set The following procedure is used to evaluate the drywell thickness at those locations where only one set of measurements is available.

(1) Calculate the mean of the set of UT thickness measurements.

This is the best estimate of the mean thickness.

(2) Calculate the standard error of the mean for the set of UT measurements.

This is the best estimate of the standard error of the mean thickness.

Confidence intervals about the mean thickness cannot be calculated with only one data set available.

10/25106 14:26:01 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 15 of 5.0 CALCULATIONS 5.1 6"x6" Grids in Sand Bed Region at Original Locations 5.1.1 Bay 1lA: 5/1/87 to 10/8/88 Six 49-point data sets were available for this bay covering the time period from May 1, 1987 to October 8, 1988. Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is more appropriate then the regression model.(3) The current mean thickness

+ standard error is 908.6+5.0 mils.(4) There was no significant corrosion from May 1, 1987 to October 8, 1988.

jalcNo F No elN jC. 7O2-/ Ssj/6o,--PROGRAM: DWCHISO ENTER NAME OF DATA LIST eila ENTER FPT NUMBER LIST ýnt:(! .4,9'ENTERF NAME OF DATE LIST dlte.7iia N 3 4 EPiA E' 1A612 Ei 1 A704 El 1A705 EI i A708 El i A~e?Eli A307 Ei Asi C ENTER NO. OF DESIRED DATA 2."3iA,5,6.7 E 1i A704 E I i A70'ElII A7OB9 El i A709 El I A807 Ei I AB1O 09881 3 1 .5175 2.6733 6.7947..438 DA TESI i A 4/29/87 5/1/87 8/187 7/1 '/388 1 0108/88 MEANTHK.91866.90464.9052.91297.3138 2.2 SD 0"163 S040982 023 7 247.049865.045901.038926 STDERR X*4*4***.0081466.)061783.0056152.0075174.0079902* 0058027 n' r mi Ix * *2 CH1952°ý.5.99 5.99 5.99 5.99 5.99 I F CHI992 9.21 9.21 9.2i 9.21? .21 EXP 8.6863 9.3218 9.3218 9.3218 6.9914 9.5337 7.767 8.3354 8.3354 3.3354 6.25115 8.5248 8.0934 8.68`6 8.6856 8.6856 6.5142 8.883 7.76'7 8.3354 8.3354 8.3354 6.2515 8.5248 C.6863 9.3218 9.32i3 9.3218 6.'9914 *'2.3I7 OBS 7 9 9 10 11 12 li 8 10 8 2 4 8 9 6 12 4 10 9 l8te 5 3 8 9 10 9 9 8 11 GRAND MEAN THICKNESS

= .90863 STANDARD ERROR OF THE GRAND MEAN = .0049825 January 18, 1989 12:54 PM No________

A~N heN 1 II A704 l14705.991 .995 1.05t .975 1.009 .951 .903 .967 .962 .923 .919 .964 .147 .873.909 .944 .966 .788 .938 0 .383 .904 .914 .893 .891 .903 .896 .922.846 .86 .9 .896 .9 .861 .909 .849 .893 .932 .854 .875 .836 .869.884 0 0 .87 .892 .961 .91? .854 0 .8 2 .9 .883 .909.9 0 0 0 .949 .906 .887 .907 0 0 ..926 .89 .851.838 .931 889 .868 1.006 0 .?01 .812 .941 .851 .895 .948 .902 .915.941 .922 .959 0 .968 .871 91 .989 .958 .97 .932 .932 .833 .938 EliA706 EliA709.955 .941 .964 .929 .966 .943 .868 .955 .959 1.058 .915 .992 .942 .974.942 .934 .905 .91 .927 .904 .897 .898 .916 .896 .88? .899 .903 .884.872 .84 .904 0 .867 .899 .913 .847 .822 .9 .846 .863 .848 .872.859 0 0 : .899 .874 .8? 5 .845 0 0 .845 .893 .86 .905.945 0 0 .8Y4 .89649 938 .902 .936 9 0 .91 .929 .915 .937.96 .999 .S81 .978 .968 .945 .936 .821 .941 .877 .895 .946 .861 .882.926 .914 .949 1.011 .928 .891 .934 0 .885 .978 1.011 .953 .A8 .942 E81A80? EilAsio.983 .965 .903 0 .954 .929 .857 .937 .944 .832 .89 .947 .911 .85.91s .943 .917 .873 .932 .927 .864 .881 .897 .905 .94 .876 .885 .857.87 .86 .895 .874 .857 .86 0 .833 .81 .889 .831 .824 .829 .857.885 0 0 0 .858 .947 .945 .874 0 0 .881 .898 .853 .998 0 0 0 944 0 809 92 0 .906 .922 .872 .884.97 0 04 .982 0 0 :eS2 .935 .876 .879 .944 .881 .924.96 .922 0 .947 .964 .856 .956 .915 .894 .952 .943 .913 .841 .913 O3 K/lIjNCAR RtEGRESSION PhO.T FOR DU WALL THINNING A ALYSIS OF~ PAY IIA V ABOVE CURB UPS YEARS MILT 0 .00 flo.?0.00 ?4.6 4 0.37 9R.i .B 57 ,ED"USDAY, JANUARY 4. 1909 62 ii.

r (4.FOI'l~ E 43ANF9MA410 LYSIS OF DAY 11A 3" ABOVE CURB.M17AN 2 V. R8N .9 3333333 12.22565608 4.9 0.57 WEDNESDAY, JANUARY 4, 1989 63 VARIABLE H ERROR HEAN 9116329 7 PR)ITI 182.05 0.0001 HILS 6 909.6 Mean 'qoer6 t :5.6/vqý" IL K LlHNAR REGRESION P4ýOT FUrOR WALL H RING t.YXIS OF BAY i1A 3" ASOVE CURB 8!37 WEDNESDAY.

JANUARY 4, 1909 64 l: VARIAB.LE Mi.S SOURCE DF C TOTAL 5 DEPC C.V.ANALYSI 30U1 OF S IJA$ S 248.59965 498.73469 741.33333 1U61,1 i .6e3W s or VARIANCE 248.59865 Q24.6R367 F VALUE 1.994 PROB)F 0,7308 R-J R-VAR I AK E V RrEp PARAMETER ESTIMAIE 9ý4 9 11 STANDAR ERROR 6. 9jj40 PARAMETER ESTIMATES T FOp 10: PARANiErER-0 144.935-1.4iz PROR ) ITI 0.0001 0.2309 TYFK I Ss 4953687.21 248 ,9865 STANDARDIZLD ESrIMArF 0-0.5761561t COLLINEAfrTY DIAGNOSTICS ores 4 5 6 ACTUAL 911.7 9G4.6 922.1 905.2 913.0 098.2 NUMBER E I G:NVA.Ut: 944.8 6. 11I 914.8 iis 9t11, 5,1039 910.6 4.7746 901.2 6.9664 R98.5 9.5037 CONDSTlOZ MEAR 897.874.9 VAR PROP'INTERCEF UPPER95%MEAN 9 2.3 926.0 923.9 920.5 92.1.VAR PROI'YEAkS LPtWER9ý1 879.2 079.2 871.a 876.9 964.7 859. 5 950 4 950.4 946.0 944.4 931.7 937.5 REXIDIJAI.

3.9024-10.1976 10.2253-5.4381 11.9050-10. 2969 Nf UIt Or STID X IJAI.$2 STUM Or RF$I )IJALS F'P:F'Dlrl'f,,D RESID .$5 (1"RES.S)1 .1741.1443.319ý- qS((, .2) .2. 774 all t (9'57 WUDRE$DAY.

JANUARY 4, 1989 65 LIhFAR !ErRE$$ION PIF'L FOR DALL T7HZHNGN ANALYSIS Of BAY t1A 3' ABIOVE CURB DEP VARIAPiLE.

MILS SOUR moot RRV ANALYSIS~CE DF S~'~PAL L f 248.59965 2 R 4 498.73469 t: 9TAL 5 747.33333 C.V. 1.128899 OF VARIANCE 4q.59865 24,60367 9-SQUAR'ER-$Q F VALUE 1.994 PROR)F 0.2388 VARIAPLE DF INIb.CE." i PAýAMEILR 9t4 79762-11 .24485554 PARAMETER ESTIMATES STANRARD F ROR 1 177910.96147274 T FOR Ho: f'ARAMHFIER=O 144.3) II lips AAcI UAI 3 922.1t 4 905..5 913.0 6 88R. 2 STUDf NH 08 RE$ I DUJAL-0.4231-1.1071 3 1.0'9A 4 -1.4229 F'REDII:tED E.SID SS (PRFSlX)PREDICt ý(DR .ta VACOIFCT O N 911.9 5.1039 897.7 910.6 4.7746 897.4 901.2 6.9664 881.9 898.5 9.5037 874.9 926.0 923.9 920.5 922.t LUWER9?f 879.2 H19.2 877.8 0t6.9 864.7 859.5 tJIF'ERY5%

950.4 950.4 946.0 944.4 937.7 937.5 RESIDUAL 3 902'4 10:1t976 10.2253--5.4381 11.9050--10. 969 RTffLURR 9.111 9.9314 10.0939 8. 1166 7.2368-2"'l"O 1 2 1443.379 COOK

• S D 0.140 0: i i.39A i'ao 08 0 Llt4IEAR:

UCCRE. 0SIDN PRtOI fIW DU UALI. IH!NNINi ANALYUSw'Of P4AY 11A 3" ABOIVE: CURP VM01 OF .tt!LS*YEARS SYMB!4OL tIfSED iS X FLO1 OF PRE.DY SYMBOL. USlD IS F P'.01 OF U950YEAR$

SYMBOL I1$IS IS U FtOLT OF SYMtUL USED IS L tl Wt Wf.DtL.DAY, JANIIAkY 4. i9H9 66 1*I,'I)C V A 1.U L 1 .100 1050 t I000 95,0 900 050 n00 I+I÷4..o ,+;7 ,h ,.'I x F.L U IJ 1, U I+..I>L L 1.fA IA'N 0,~... .D ..... .. .......~ 4 .....,+..1 )2~ 0..1 0.4 0.1 4 IWhH 10P14I.N 0, 6 , ft A .', i .. .4. .. 4... *0 .1 f .? 1)

• I*().4 ....... +.. * .- .... ......+ 4 ..S. 1 1.2 $ .. 1 .4 NI) I

--C t.INEAR If(L.GRlI..SSIN F101 FOR DU WAI.L THINNING ANALYSIS OF BAY IiA 3' ABOVE CURB Pt.Ot OF kEL.IDIJALKYEAR " SYMBOL. IUS'ED Is RWE.DNfSDAY, ,JAMIJARY 4, 19fI9 67 40 35 35 25 15 1 0 o)ft, I, x I I I I 4.I 4 I I J,)o kt 1.0-20* 25-35 40\0t 0 o,1 o ,-, 0.3 0,4 0.5 O.A 0.1 0.0 yI f. Ak s 0.9 1o *0 1 1.2 1.3 1.4 I C,/ Z5/ n6 ' 4 : 2; : t-, Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 9 of 5.1.2 Bay 11C: 5/1/87 to 10/8/88 Five 49-point data sets were available for this bay covering the time period from May 1, 1987 to October 8, 1988. These data were analyzed as described in paragraphs 2.4 and 2.8.2. The initial analysis of this data indicated that the data are not normally distributed.

The lack of normality was tentatively attributed to minimal corrosion in the upper half of the 6"x6" grid with more extensive corrosion in the lower half of the grid. To test this hypothesis, each data set was divided into two subsets, with one containing the top three rows and the other containing the bottom four rows.The top subset was normally distributed but the bottom subset was not. For both subsets, the mean model is more appropriate than the regression model.Since there is an observable decrease in the mean thickness with time, there appears to be some on-going corrosion at this location.

Further analysis is required.The current mean thickness

+ standard error is 916.6 +10.4 mils for the lower subset and 1057.6 +16.9 mils for the upper subset.J 2C/2 5 14: 2 Lcd/3ReNO.

s a see L0 PROGRAM: DWCHISQ ENTER NAME OF DATA LIST dlic ENTER PT NUMBER LIST intz(¶,49)

ENTER NAME OF DATE LIST d34567 N DuOI 1 Di i 612 2 I I C1705 3 D110C708 4 D1i1C?09 I Dl i cbo 6 LNI I C1, I EWTER NO. OF' DESIRED DATA 2,3,4,5,4 D1110C70.5 Di I C708 Dli i C709 Di 10807 DI i c~i 0 CHISQ 40.419 i4. 689 60.493-)23 27 .104 D34567 VI /87 5/1 /87 ,1117 8/1/1378 9/ 10/8 7 7/i 2/SE I 108/8 CH?1952 5.99 5.99 5.99 5.99 5.99?8 MEANTHK SD.96735 .092336 1.0132 .10'737.97744 .10944.9`79 .0998195.94133 .094979 XTODRR.1 3614.015498.016315.016885.013568 DFM2'2 2 CHI?92 9.21 9.21 9.21 9.21 9.21 9.7456 10.i69 9.5337 7.4!51 10.381 2.7142 9.0931 8.5248 6.6304 9.'826 9.004 9.475-2 8.883 6.90? 9.6726 8.7142 9.0931 8.5248 6.6304 9.2826 9.7456 10.169 9.5337 7.4151 10.381 CBS 4 9 i 3 6 25 .3 2.9 17 23 4 8 5 6 6 3 -" 1 2 3 i() 11 10 7 11 GRAND MEAN THICKNESS

= .97243 STANDARD ERROR OF THE GRAND MEAN = .0129 ,January 1, 1989 121 -55 PM lCjlc H6O 1 C.I3O2/0 S314 0 j V7 O1iC70 DlIC799 0 0 .90958 .92 .998 1.117 958 0 1.017 *j 9 1.4.151.109 1.06 5 11 1, 1. 1136 116 i. tes IA&6 if1 a~ : .2 1'.191 1.276 1.244 .. .213 1:117 Dl.,lO? I -"C"UG?" "'1.,, :92 4 .901 ., .1.01 .94 ..91 .9493 .9 286 .91t .. 96:0 :91I :992 :? .,0 , Sol :i .°94 : 1 1:9 ., DlIC719 IIGO.98 91 .1:7 .39.6 91 :,310 :1 :914 15 4 .892 a.884.9 9 41:ZJ4 44:903 .90 .92 9011 .91 449 89 0* 101 a :84 :8 .91 7 nilcels.,, J:!' 4i ::1 Mt": O I4 LINEAR REGRESSION PLOT FOR DU MALL. THINNING ANALYSIS Of SECTION tiC DEP VARIABLE.

MEASURE SOURCE DF MODEL I ERROR 3 C TOTAL 4 ROOT MSE PEP MEAN C.V.ANALYI SUM OF SQUARES 1490.70825 1832.09175 3322.800o0 24.71229 972.2 2.541094ýIS OF VARIANCE MEAN SQUARE 1490.76825 610.69725 R-SQUARE ADJ R-SQ F VALUE 2.441 0.4486 0.264a PROW)V 0.2161 VARIABLE DF INTERCEP I YEAR i PARAMETER ESTIMATE 992.11 07-30.59444577 STANDARD ERROR t6.B6860259 i9.502099M2 PARAMETER ESTIMATES T FOR HO:

38.814-1.562 PROD ) ITI 0.000i 0.2161 TYFE I SS 4725864.20 1490. 70825 STANDARDIZED ESTIMATE 0-0.66979859 L'., 0BS I 2 3 4 5 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)ACTUAL 967.0 iota.e 977.0 958.0 941.8 5.11591 E 1832.4858.NUMBER EIGENVAL 1 1.755 2 PREDICT STi VALUE PRI 992.1 16 984.4 13 981.1 12 955.4 15 948.0 19 COLLINEARITY DIAGNOSTICS CONDITION VAR PROP LUE NUMBER INTERCEP 488 1.8000080 0.1223 512 2.679471 0.8777 D ERR LOWERYSS UPPER975 EDICT MEAN MEAN.8666 938.4 1045.8.5324 941.3 1027.5.4235 941.5 1020.6.4207 906.3 1004.3.0152 887.5 1008.$VAR PROP YEAR 0.1223 0.8777 LOUER9SZ PREDICT 896.9 894.?893.0 862.7 648.8 UPPER95Z PREDICT 1087.3 1074.1 1069.1 1048.1 1047.3 RESIDUAL-25.1109 33.5989-4.0663 2.6025-7.0243-13 092 022 tq-(Is-2 (S-X 3./8 z VA CA

/'e LINEAR fREGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION tiC DEP VARIAKLE; MEASURE SOURCE DF MODEL i ERROR 3 C TOTAL 4 ROOT MSE DEP MEAN C.V.ANALY!SUM OF SQUARES 1490.70925 1832.09175 3322.80000 24.71229 9?2.2 2.541894;IS OF VARIANCE MEAN SQUARE 1490.70825 610.69725 R-SQUIARE ADJ R-SQ F VALUE 2.441 0.4496 0.2648 PRON IF 6.2161 VARIAI.E INTERCEP YEAR OF I 1 PARAMETER ESTIMATES PARAMETER STANDARD ESTIMATE ERROR 992.118e?

16.06860259

-30.59444577 19.sB209912 t FOR HO: PARAMjETERse 58.814-1.562 PRORf I T I 0.2161 ofs 2 3 4 5 ACTUAL 967. 0 t015.0 977 .0 958.0 941.6I S PREDICT VALUE 992.1 984.4 98t .t 955.4 748.0 SID ERR PREDICT 16.58696 13.5324 12.4235 15. 4207 19.0152 LOWER95Z MEAN 938.4 941.3 941.5 906.3 887.5 UPPER952 MEAN 1045.6 1027.5 1020.6¶004.5 1008.5 LOWER952 PREDICT 896.9 894.7 893.0 862.?8419.8 UPPER95%PREDICT 1087.3 1074.1 1069.1 1048.1 1047.3 RESIDUAL-25.4109 33.5989-4.0663 2.6025-7.0243 TrD ERR RESIDUAL 18.096 20.6718 21.3624 19.3106 15. 7935 STUDENT RESIDUAL-t.3904 1.6249-0.1903 G.t348-0.4450-2-1-0 1 2 I 'I I I I I 1 0.843 0.565 3 0.006 4 0.006 5 0.144 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID $S (PRESS)t 0'5.11591E-13 1832.092 4858.022 (LINEAR REGRESSION PLOT FUR DU WALL THINNING ANALYSIS OF SECTION 11C PLOT OF MEASURE*YEAR SYMBOL USED IS X PLOT or PRED*YEAR SYMBOL USED IN P PLOT OF U750YEAR SYMBOL USED is u PLOT OF LR95YEAR SYMBOL USED IS L$000+I IP IP I P I x I IX I x 950 P x P R E D T E 900 D IL L L V I A L I U L EI 850 + L$00.o..0. -,.2- 0. 0.4 ..6 -.- + 1.---.------------

---

4 .0.0 0.1 0 .2 0.3 0.4 0.5 0.6 0.7 0.8 0.? 1.0 1. 2 .1.3 1.YEAR NO IU 6 Ob$ HAD MiSSIOG VALUES OR WERE OU1 OF RANGE 0 LINEAR REGRESSION PLOT FOR OU WALL THINNING ANALYSIS OF SECTION t1C PLOT OF RESID*YEAR SYP¶OL USED IS R I R E U A L S 4,++++-4 +-S ÷-1!R R IA flu'&--*.. ..... ...-.. ...-... --.... .... --.. -... .... .. .. .--.. .. .. ..,-. ..... .. ---... .. .4- ... .... 4-.. .. ... .... ... .. ,-.... ...-.....

......--4.. ......---oo 0.,1 0.2 0.3 0.4 0.4 0., 0.7 0.8 0.9 1.0 t.f 9.2 1.3 1.4 YEAR NOIL 2 0VE HAD MISS(NG VALUES OR WERE. OUT OF RANGE LINEAR RE.GR$SSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 1iC UNIVARIATE I, , VARIABLE-RESID RESIDUALS N MEAN STD D£V SKEWNESS USS CV T !MEAN=0 SGN RANK NUM "m 0 W:NORMAL STEM LEAF 3 4 2 i MOMENTS S SUN WGTS 1.023E-13 SUn 21.4015 VARIANCE 0.922373 KURTOSI$1832.09 CSS 99999 STD MEAN 1.069E-14 PROB)ITI-1.5 PROB)ISI S 0.932876 PROD(W S 5.116E-t3 458.023 1.97891 1832.09 9.57103 0.787406 8.545 100% MAX 75% Q3 sox MED 252 Q1 02 NIN RANCE Q3-QI"ODE QUANTILES(DEF24) 33.5989 99t 18.100? 95X-4.06627 90Z-16.0676 102-25.1109 52 1z 58.?798 34.1683-25.1109 33.5989 33.5989 33.5989-25.1109-25.1109-25.1109 EXTREMES LOWEST HIGHEST-25.1109 -25.1109-7.02427 -7.02427-4.06627 -4.06627 2.60247 2.60247 3S.5989 33.5989 41 03 1-0 74 2-i-2 51 MULIIPLY STEM.LEAF BY 10**,01 BOXPLOT I I*- !* -.. .354 54 NM P AL 4)NORMAL PROBABILITY PLOT'61 44+446--25+-2----- -1 ------ -----01 84 I

~O/2$IC~ 14:2~:CP C ac No Rev No Sh.eet ,o PROGRAM: OCDWCONF BAY: IIC TOP3 DI I CIOFP3 1 046 Ii103. 6 1079..1 10,8.9 M¶EAN THICKNESS "i057.STANDARD E6RRR OF" TEMEA 16.9094 )- 2.7763 r(.efi2, 4 = 4.6041 CONFUEHNCE INTERVALS F'OR:h THE MEAN 95 , UPPER BOONE = i104.5 9.Z LOWER BOUN 1 i ) .7'40r% JF'E2R P= 0 27 19 9-'j' nulary 20, 1989 12: 56 -.

ENTER "A"E OF DATE LIST 34167 6 D O1 ENTER NO. OF DESIRED DATA 2,3.4,5.6 D34567 MEANTHE SD DFM2 CHISj CH1952 NUACUSW* U*%U*U** U*U**"* **4*** ewe, ;**No* wewewe D1IC?7O 5/l/07 1.846 1.344 2 .4172 5.99 D11C708 8/1/87 1.1096 .18947 2 3.563 5.90 D 1; I a 4 iAIII 99 Eila:? ioiib9/ 1.008 .1 0? Z IM60 3:;;OS EXP 4 7 6 3.8135 4.J372 4.6811 2.?542 4.4491 4.,1 + 3+5 :gII3 J:1711~ ?~A~ :"42 3.908 10 4 4 6 3 4 384099 3. ?8 3.2205 2.4627 ? .9702 5 5 3 4 6 3.0135 4.2372 3.6016 2.7542 4.4491 C, M .Is I- /f LINEAR KGRVý$$iS 6LUI FOR DU UALL TI1ZNNI, ANAL.Y.IT OF BAY I IC UPPER 3 ROws Obs YEARS MlLS:% 8:3R 408 3 0.36 1079..1 4 1 20 1045.4 5 1.44 1000.9 H 5i7 WEDNESDAY.

JANUARY 4. IV98 53 47 tA~

/1 f VARIKBLE N LINEAR RE JSSON PLOT FOR Did WALL 114 NN 06N AN4ALYSIS OF DAY 11C UPPER s3 KRowd MEAN 110MAHM W ARO 0000000 37-90995321 16.90909040 8.51 WEDNESDAY, JANUARY 4, tve9 T PR)ITI 62.55 0.0001 54 MILS 5 10' .64/V,CW THMCAI'.sSc 1 657.e. +/- 5. ,

1" 0 L NEAk ON PlOT roa DW WALL HINlINC AIALYSIS OF DAY tiC UiEER 3 RDI DEF VARIABLE-Mlt.S SOURCE oF C TOTAL 4 C.v.ANAL. YýxUHt OF SQUARES 5718.34000 3.051722 IS or VARIANCE"EAN SQUARE 8,? UEDNESDAY.

JANUARY 4, 1989 ROBW )F 9.212?7 55 F VALUE 2. 490 R-SQUARE ADJ RU-SQ VARIABLE OF PAR A%EX T 15AT PARAM!ETER ESTIMA1ES$TANRAR9 1 FOR HO: PARA'ETER=e F'ROB ) I I I IN cp I 53.30987 -0.67342963 OB$ AC S 10 4 te 5 go~mF ES IDAL$Um a U R IDUof.ED 1.OI'E Rto1 (PREksS)IUAL 79.1 NU"PER UILENV PRU!IR W VALUE 1069.3 164 ;0 COLLILEARITY DIAGNOSTICS CONAI TON ~prO ALUE M RB.I IqTE £C ERR tPPER1S2 EDICT MEAN MEAN='72 1017.7 1120.9&24 971.3 1099.5 272 946.7 t104.7 VAR f*01'Y(EARS LOWER95D PREVJCr 719.5 9 6.6 954.3 W14.3 a96.1 PREDICT 1208 -.1190.9 1184.3 RESIDUIAL 64B19 9.7999 9.9899-16.8273 i.13687"-l 312.3 9624. O12 IA VL 0 [,

-- K I JNEAR REcRfsT $ N rLOT FODN WALL. THY H! NG ANAL.YsIS Or BAY tIC UPPER 3 ROWS B hI I.DNI.DAY, JANIJARy 4, 1989 56 DEP VARIAOL.E:

MILS SOURCE OF FORD I C TOTAL 4 C.V.VARIABLE OF[HTERCEP I I YEARS t 4 ANALYSIS H$125. alt 5718.34000 3.05t 722 OF VARIANCE 9 .09al 49?.2767t 20Q-ty-F VALUE 2.490 8:11 T rop Ho F'ARANEITER-O PROB)F 0.2127 PARAMETER EsT "ATE 01308B2439

).34MM?1 PARAMYETER ESrItIATES S TAWRD F'ROo ) I11 o0 oot 0.2127 Obs ACtUAL 1 1046.0 2 1108.6 3 1079.1 4 1045.4 5 1008.9 lifts 0HRULbI~5 0.3513 0.3963 5 -0.8160 PREDICfED RE$ID SS (PRESS)PREDICT VALUE 1083.8 t073./S069.1035.4 1025.7-2-I..0 1 2 9624.1112 f~pogq O~rg~gA UPPER952 HEAN 22.0131 1153.9 17.6903 1017.4 1130.0 S6.2272 101?7.7 20.1524 971.3 1 24.8272 946.7 114.1 COOK ' S 0 959.5 956.6 954.3 914.3 896. 1 1200.1190.7 4184.3 11 56.5 1 155.3 NE SI DUAL-37.6244 34.8619 9.7999 9.9899.16.8273$Ij ?ORR 2A,. 60.40 26.9950 27.0990 25.2103 20.4225 I 0 021 0. o8o I 00 tA 4 zi 0 LINE~AR PEGRE.SSjg VLON 1 rtJR DW WALL 71 INN 1 ANALYSIS or bAY ItC IUPPFR 3 ROWS~18I 8~P~SWYAI xY f~Lip PLOT or 1.9 YEARS SYMBOL UiSED IS L 13 5? WCONtSDAY, JANUJARY 4. iV9W 57[A L.U 1250 I ?00 1150-1050 I 1000 950 850 BC0 I'U IL U u 11 P u K P x p F, L L I.xý--n qA...3.. ..4 D.) 6.t, ... ...... ý .. --........

-, -. .. ... ... -0.7 0. R 0.9 1.0 V I At($-4 ... 4 ... + ....1.. 1 3 1 .4

-4,0 of DAiY i c UP FR .5 POUs fi 0! Of' W.S IDIIaI 'rAras SYMP01a USED IS R-.'it WI!NIN.DAY, JANUARY 4. 1909 58 40.35.30 20 15 10 0 1).10f-15 Ft 1,) ,o¢D R I.D I, A L$,30-35-40 0.0 I ft N N 0. q 0.4 5.' 0.6 01 1 Yr fAS 0.11 S ...4...1.0 1.1 1,. .4.

PROGRAM : OCDWCONF lI:Ay: liC POT4 D11ICB) T"4 ji * ** *9i6.8 953.6 915.7 906.1I 890 .7 MEAN THICKNESS

= '16.!-STANDARD ERROR OF THE MEAN 10.37 T(.5/2. 4 )= 2.7763 T'/2.4 )= 4.6041 CONFIDENCE INTE! FULFR THE MEANUF'f'EF' i.NI) B 945.37 9i% LOWER BOUND B87. 79 99% UPPER 1'OUND = 964.33 9Y% LOWER POUND 268J33 January 20. 1989 12'~

• P, 'K)0

1. I" ENTER NAME O DATE LIS d'351 N DlIC80 6 DlI11C10 ENTER NO. OF DESIRED DATA 2,3,4,5.4 D34561. SD fl9 .INU MUiCGS 5/1/B? 5k 679 .0264"9 2 4.035 5*99 DiIC368 8/1/97 .9364 .037M589?

.781 1:;oilC91e 10/09/98 .806 :81146 16:399 ass EXP 4~ 4:2 2539321 4.6609 J.32 96 ~12 :1 Mal 4~~c 4 567 :0 4 6 3 11112? 4:69 5932 (A L.

I-ON V'IU1 FOR U WALL. -HNIN(. ANALYSIS OF PAY SIC LOWER 4 ROWS 014S YEARS MILS 8'5? ULDNFXDAf.

JANUARY 4. i9BV 35 4 0.36 915.7 1.20 906.1 1 .44 8f90.7 In K.VARIABLE LI fFAFR REGRESSION

~o FaR Dt UALL TJNNING ANAIY$STS OF PAY LOWER 4 ROWS MEAN DA R0o0000 23.18894133 10.3 9.57 /EDNE..SDAY, JAMUAkY 4. i91? 3A N 7040993 1 PR)ITI 60.38 0.0001 MrL$5 9f6.58 16.[A I 0N4AR RL(,RESSIUN K'L.U FOR 01 WALL THINNINI ANALYSIS OF DAY IlC LOWER 4 ROWS NUE' VARIAL..r:

MILS SOURCE oF C TOTAL 4 C.V.ANAL.YI SUH or SRUARES 2150.90800 19 15448 4t6. 8 2.089778 SI OF VARIANCE MEAN SQUARE 1950.22535

&66:9422 8 57 ULDNESDAY.

JANUARY 4. 1989 37 PRO8)r 0.1992 F VALUE 2.862 R-SQUARE ADJ R-SQ F'ARAM51IER VARIABLE or ESTI ATE!UCEI9 58 YE ' I S ai&PARAMETER ESTIMATES STA2RA~ ~ T S T N D R D H O O : ERO ARFARETER=0 3064277g3 71:137 0.75230 0 -1.692 PROP ) I1I II209594.48E

'TYPE SS s-A 6976E A,229 -.97A lift AC'3 9 4 94 5 8 IUM OF RESIDUALS UM OF SQU RED ESIDUALS PREDICIED RESID SS (IPRESS)NUIM1ER 1 2 PREDICT TUAL ALUE i .7 924.0 e6.1 902.5 90.7 896.3 6.25278E-13 150:683 A2158.051 F 161 NVA 1.751)0. 244 SID PRE 9.CrILIN:ARIIY DIAGNOSTICS CONDTI, O4 AR PRNUWRITERCE-ERR LOWER?5% UPPER95%DICT MEAN MEAN.6303 893.4 954.7 9600 864.4 940.5 7344 849.4 943.2 VAN FVROP YFARS 8: S LOWER91Z PREDICT 355.9 830.6 819.4UPPER95%PREDICT"76: 1 992.3 974.3 973.2 RESIDUAL-8.3257 3.6411-5.5970%A (A , S i L.0 iN kk RE:rjE,'ON P'LO)I-OR ALt "I'NNTG ANAL,YSIS Of RAY lI1 LOW.ER 4 ROWS H'57 WEI).IXPAYo .IANIIARY

4. 1989 3i PWF" VARIAItE MIL.X SOURCE DF C TOTAL 4 C.V.AHA Ys SUM (IF SQUARES N150.90800 2.089778'IS Of VARIANCE MEAN SQUARE F VALUE 2. 962 PROB)r 0. 1892 VARIAFI E DF INTERCEP t YEARS t PARAMETER ESTIMATE 933 '6858-25.67A47266 PARAMETER ESTIMATES STANDARD 13.06421493 15.17523 00 T FOR HO: PARAETER-O 71 4j7-1.6 2 p'ot ) ITI.0 o01 0. 1897 nNIS 3 4 5 AClUJAL 916.8 953.6 915.7 906.1 190.7 S FIIDF NT opS RE X I DUIAL 4 0.2434 5 --0.4573 t'RLDWtED ReESI SS (PRESS)F'REDICT VALUE 933.3 926.8 924.0 902.5 R96.3*. 21. t- 1 2258.051 STD VRR 13.0643 10.4988 9.6305 11.9600 14.7344 LOIER957 MEAN 891.7 993.4 893.4 064.4 849.4 COOK'S 809 0.043 0.019 0.152 IJPPrR95%MEAN 974.8 960.3 954.7 940.1 943.2 LOWER95X PRED I T 859.5 a5s .3 855.a 830.6 819.4 PREDICT t007.1 996.4 992.3 974.3 9?73.2 RESIDUAL-16.4606 26. l!01 3.6411-5.5970 RES DUAL 14.0078 16.0209 16.5574 1 4.9617 1 2.2390 I (A (Al I 0*>I INEAR REG4ESSION FPt.OI FOR DU WALL THINNINGf ANAI.YLYS OF PAY tic LOWE.R 4 ROWS ,:I1 I P A IS X ,..M PRgY ARS S is r F'.)0 1 OF iJ951YEAR$

SYMItOL IISF*D IS LI NL0 OF' l.?5WYAItS SYMOI. USED I.":%? WFI)NE.SODY, ,JfNlJAkY 4, 1997 39 E D I c v 1)V A L 115)0 1 050 900 4.O0A I, 4"'0 U 0,0 u U u 1)x P'x x I.x L L tA 10 1..4..I)..2 .. .....4 (-). 2 0.5, 0.4-..4ý ......1 4....0.5 0 .A 0,7 0. 0.9 1 .0 Yt. ARN..1 1.2 1.3 1.4 NoI I l, l 014S H illI)EN

-4: LINEAR REGRESSION FT.O1 F.O IOk D A.L A tPINNING ANAIYNJS OF PAY iC I.OIJER 4 ROWS PNOT OF RESIDrIJAINYFAR SYMBOL USDP IS R H-t'/ KDRE.fSDAY, .IANUARY 4. 1?89 40 I.n U A L 40 20 i 10.0!*t!, I 20.10.0 ri R R R'I Vt i,.,.g (~. ') 0 s 0.4 4. 4 0)5 0J~ " 0.14'Vt ANS 0.9*...I 1.0 1.1 1.2 1.1 1.4 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page q4of 5.1.3 Bay 17D: 2/17/87 to 10/B/88 Six 49-point data sets were available for this bay covering the time period from February 17, 1987 to October 8, 1988.Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 84% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 864.8+6.8 mils.(6) The corrosion rate + standard error is -27.6 +6.1 mils per year.(7) The measurements below 800 mils were tested and determined not to be statistically different from the mean thickness.

&6t25/0* Czac No 1 Rec NO 1 Shet No c.:302-1es-Sl ' I',L1 PROGRAM: DWCHISQ ENTER NAME OF DATA LIST e17d ENTER PT NUMBER LIST ints7(1.49)

ENTER NAME OF DATE LIST d234567 N I 4 51 El 71612 E17D~e Ei D705 El H1708 E It7D? 09 El 7D807 Ei t7D810 ENTER NO. OF DESIRED DATA 2.3,4,5.6.7

• • T0T**Ei 7D705 El 7D708 El 7D709 El 7D807 El 7D810 CHISQ 7.5153 2.8389 2.1086.43573 2.0383 1 .2028 D234567 2/17/87 5/1/87 8/1/87 9/1i/87 7/i!2/88 S0/08/88 CH1952 5.99 5.99 a.99 5.99 5.99 5.99 MEANTHK.92217.81507.P8969.89528.87793.86222 SD.061283.051215.054341.061832.061168.055095 STDERR.0094561.0076346.0081006.0094294.0094384.0082131 DFM2 2 CH1992 9.21 9.21 9.21 9.21 9.21 9.21 EXF 13.8981 9.5337 9.5337 9.11 8.B981 9.5337 7.9565 8.5248 8.5248 8.1459 7.9565 8.5248 8.2908 8.883 8.883 8.4882 8.2908 8.883 7.9565 8.5248 8.5248 8.1459 7.9565 8.5248 8.8981 9.5337 9.5337 9.11 8.8981 9.5337 OBS 9 9 11 9 8 9 4 8 5 8 8 8 13 12 10 10 10 7 11 5 iO 7 5 11 5 11 9 ? 11 10 GRAND MEAN THICKNESS

= .89056 STANDARD ERROR OF THE GRAND MEAN := .0081735 January 18, 1989 12: 56 P M S

" l N E1 7D?02 1 3 -lJ $360 uB~luOaltisN.~U0u*wUult5ueNumgw**a*~Ulu 93 923 .913 .943 .'J7 82 .9,12 1. ..943 .96 :997 :999 .801.909 :91o .869 .964 .909 .874 .89 ..94 .83 .871 .856 .843 .83 0": .1864 .063 .99, .284 .9 Y: 07 .916 0% ..9 .828 .es .8 8 905.89 .991 .954 .8 .897 .939 .942 .923 .922 .89.922 .945 .909 .955 1.018 .987 t161 .896 .963 .8 .893 Y2S .977 .91 El70708 ElD7709.924 882 .912 .945 .967 .997 .784 .914 .eet .92 1.063 1.025 .979 .784.908 .88 6 797 .i1 M 90 .904 .87..82 si : 95Sl~ 799 0 0 A1 .1i 6 .889 .843 .804 0 0 :881 :6 0 .97ý :786 0 0 .915 .862 .829 -925 s888 0 0 .835 .866 .829 .931 883.842 .869 .689 .968 .888 .04 , 93 .836 .869 .AV .971 .o05 :9 .908.948 .9:94 .O.87 8 93 921 .925 9 1 :91 .867 .921 .9%. .94 .927 ? .9 0.964 .918 .91 .954 .979 1.012 .904 .965 .915 .955 .95 9 E17D0807 £E17DO.935 .941 .898 0 .97 .962 .764 .891 .883 .915 .917 .954 .95 .7?1.892 .863 .775 .846 .77 .3 .7A7 .883 .47. ..762 80 752 0 0 .86 .914 .918 0 757 0 0 96 .5 .878 .825 .731 0.3 0.7 .794 .892 .821 .858 .8 t 0.5 .787 .842 .802 .982.8 S74 .849 .874 .74 .8 .7 9 .967 .841 .A76 .905 .88-a, 964, 85,,4 .93-:8 .-J59 :e+.121103,134:91o 11.971 7 .95 0 .945 .9 .1 ..9 7 9 5 944 L 4 I q 4 LINEAR REGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 17D DEP VARIABLE:

MEASURE SOURCE OF MODEL I ERROR 4 C TOTAL 5 ROOT MSE DEP MIEAN C.V.ANALYSIS OF VARIANCE SUN Of MEAN SQUARES SQUARE 1673.42584 16?3.42584 328.07416 A 2.01954045 200f.5000 9.056409 R-SQJARE 890.5 ADJ R-SQ 1.et0T73 F VALUE 20.403 0.936t 6.7951 PROD)r 0.0187 VARIABLE DF INTERCEP I YEAR i PARAMETER ESTIMATE 910.07272-27.60669793 STANDARD ERROR 5.69613550 6.1i163939 PARAMETER ESTIMATES T FOR H0: PARAWETERwO 459.770-4.517 PROB ) ITI 0.0001 8.010?Q 41 TYPE I SS 4757941.50' 1673.42584 STANDARDIZED ESTIMATE 0-0.91437730 2 3 4 5 6 SUM OF RESIDUALS SUN OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)ACTUAL 922.0 893.0 891.0 895.0 878.0 062.0 7.39964E-328.0?789.43 NUMBER EIGENVAI 1 1.760 2 0.239: PREDICT STJ VALUE PRI 910.1 5 9"4.6 4.897.6 4 B94.6 3.871.4 5 064.8 6 COLLINEARITY DIAGNOSTICS CONDITION VAR PROP LUE NUMBER INTERCEP 71 0.1196 202 2.712624 0.8904 D ERR LOWE R951 UPPER95Z EDICT MEAN KEAN.6961 894.3 925.9.8319 891.1 918.0.0171 886.4 909.7.8063 934.0 905.2.6129 855.9 897.0.7"8 045.9 8r83.6 VAR PROP YEAR 8.1196 0.9804 LOWER95Z PREDICT 980.4 076.t 878.1 967.3 841.8 833.3 UPPER95%PREDICT 939.9 933.1 925.1 92t.9 901.0 896.2 RESIDUAL 11.9273=9. 55t5-6.5948 0.4143 6. 5?58-2. 7711 I t4 .'(A.i 13 42 397 t 1 56( -2)z 7 7 74 All-ii'r " I 0.,40.Ai = 8ý111, a 16.9 LINEAR REGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 17D I 4 DEP VARIABLE:

MEASURE SOURCE DF RODEL I ERROR 4 C TOTAL S ROOT MSE DEP MEAN C.V.ANALYSIS OF VARIANCE SUM OF MEAN SQUARES SQUARE 1673.42584 1673.42584 329.0t416 92.01854845 2601.5"000 9.056409 R-SQUARE 890.5 ADJ R-SQ 1.017003 F VALUE 20.483.O8361 0.7951 PROD)F 6.e107 VARIABLE INTERCEP YEAR OF I PARAMETER ESTIMATES PARAMETER STANDARD ESTIMATE ERROR 916.07272 5.69613550

-27.60608793 6.11163039 4 T FOR HO: PARAIETER-0 159.770-4.517 PROD ) ITi 0.0001 0.0107 1 2 3 4 5 6 ACTUAL 922.0 095.0 891 .0 895.0 879.0 862.0 COOK'S PREDICT VALUE 910.1 904.6 997.6"94.6 8?1.4 064.8 STO ERR PREDICT 5.6961 4.8319 4.0171 3.9063 5.6129 6.7908 LOWER95Z MEAAN 894.3 891.1 886.4 884.0 845.9 UPPER9SZ MEAN 9215.19 919.0 9GO .7 908.7 905.2 B987.0 883.6 LOWIER952 PREDICT 80..4 8?6. 1 870.1 867.3 841.8 833.3 UPPER95Z PREDICT 939.8 933.1 925.1 921.9 901.0 896.2 RESIDUAL 11.9273-9.5513-6.5948 0.4143&.5759-2.7711 LTD ERR RESIDUAL 7.0408 7.6597 8. 1168 8.2177 7.1073 5.9920 STUDENT RESIDUAL 1.6940-1.2470-0.8125 0.0504 0.9252-0.4625-2-1-0 I 2 I.II Obs D t 0.939 2 0.309 3 0.001 4 0.000 5 0.267 6 0.137 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID 9S (PRESS)(A a%. -J 7. 38964E-13 328.0742 789.4397 I q LINEAR REGRESSION PLOT FOR D60 WALL THINNING ANALYSIS OF SECTION 17D PLOT PLOT PLOT PLOT OF MEASURENYEAR OF PRED*YEAR OF' UtWUYEAR OF L"*YEAR SYMBOL SYMBOL SYHBOL SYMBOL USED IS X USED iS P USED IS U USED IS L)P R E D I C T v D V A L U E ieee.'U I I'P I I0 Il 900 +d I I 6 U P x L U P x U x U X P U L L P x 850 +L L eOB-4-------.---------------4-------

---

4--- ----------

4---------

---

4----- --------4-------------4---------------------------4-------------4-----------

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.9 2.0 2.2 2 YEAR I. I OBS HIDDEN IN J I01)./HN 6 6 LINEAR REGRESSION PLOT FOR DV WALL THINNING ANALYSIS OF SECTION 17D 6 7 PLOT OF RESDUYEAR SYMBOL USED IS R 10 B +6 +4+-2 ft R D U A L$0 R[(vie~i~1 o I.-4-6-e RR-4-------4-------4--------.----------4------------+-------------4----------------

---

---

+---- ---------e-----------------------

0.0 0.2 0.4 0.6 0.5 1.0 1.2 1.4 1.6 1.5 2.0 2.2 2.4 YEAR I OBS HAD MIISSING VALUES OR WERE OUT OF RANGE-10 tA1 (A 3.3.1 NOTE: j, L 4-q 4 LINEAR REGRESSION PLOT FOR OW WALL THINNING ANALYSIS OF SECTION 17b UNIVARIATE VARIABLE-RESID RESIDUALS MOMENTS QUANTILES(DEF=4)

EXTREMES MEAN STD DEV Us$CV T:MEAN-O SGN RANK HUM "* a U: NORMAL 6 I.232E-13 8.t1"3 0.452516 329.874 9"9,, 3.724E-t4-0.5 6 0.964399 AM WGTS VARIANCE KURTOSIS CS$STO ME"N PROS) IT I PRMO) I S 6 7.39"E-13 65.6148-0.9Bt5121 32B.74 3.30693 1 1 0.918 sez 252 ex MAX 43"ED"IN 11.9273 7.91368-1.1.841-7.33395-9.5515 21.47"9 15.2476-9.5515 992[951 9e2 51 Ii 1f.9273 11.9273 11.9273-9.5515-9.5515-9.5515 LOWEST-9.5515-6.59476-2.77113 8.414299 6.57581 HIGHEST-6.59476-2.77113 0.414299 6.57581 11.9273 Q3-Ql Moog STEM LEAF 12 2 a? 1 0-0 30 2-0 7 2 MULTIPLY STEM.LEAF BY 10Ww+.1 SOXPLOT÷- --*-!.'I a I I I NORMAL PRORAVILITY PLOT 2.5*-2 -1 0 +1 +2.1 ItA vi'[iy~:l i*;'ti L~i

0/25/CE :4E:c~Calc. No. C-1302-187-5300-D05 Rev. No. 0 Page_;ýof 5.1.4 Bay 19A: 2/17/87 to 10/8/88 Six 49-point data sets were available for this bay covering the time period from February 17, 1987 to October 8, 1988.Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) The data are nearly normally distributed.

(2) The regression model is appropriate (3) The regression model explains 88% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 837.9+4.8 mils.(6) The corrosion rate + standard error is -23.7 +4.3 mpy.(7) One data point that was below 800 mile at two different times was tested and determined to be statistically different from the mean thickness.

The probability of this occurring is less than 1% at each specific time.is 1C/25/0E 14:2&.:07 ca.c No --Rev No iShectN7oI 1.302/Jle7-S36O0 I$I5o i t S PROGRAM: DWCHISQ ENTER NAME OF DATA LIST ei9a ENTER PT NUMBER LIST tnts(l.49)

ENTEr NAME OF DATEF LIST d234567 N E-IA I E19A6i1ýS19A( )02 3 E19A705 A E19A708 E: i ?;A 7!39-E19A799 ENTER NO. OF I)ESIRED DATA 2.3.4,55,6,7 El 9A702 E I9A70-5 El 9A708 El 9A709 EI9A807 Ei 9A310 CH I S70 4.8162 9. 13015 8.6057 8.9579"39'79.113"02 D234567 2/1 7/87 5/1 /87 8/1 /87 9/10/8?7/12/88 10/08/88 CH1952 5. 9?5.99 5.99 5.99 5.99 5.99 MEANTHK.88364.87293.85829.341357.133691 SD.050725.056352.05677.053896.061395.063663 STDERR.0076472-0084004.0084628.0080344.0092557.0094903 01: r NitWit r)I 4~-~1 I, CH1992 9.21 9.21 9.21 9.21 9.21 EXP 9.3218 9.5337 9.5337 9.5337 9.3218 9.5337 8.3354 8.5248 8.5248 8.5248 8.3354 8.5248 8.6856 8.883 8.883 8.883 8.68:6 8.883 8.3354 8.5248 8.5248 9.5248 8.3354 8...242 9.3218 9.5337 9.5337 9.5337 9.3218 9.5337 0B$12 10 9 8 7 9 8 14 15 i5 ii 9 7 2 4 5 9 9 4 6 6 5 6 8 13 11 Ii 12 11 10 GRAND MEAN THICKNESS

.85982 STANDARD ERROR OF THE GRAND MEAN = .0068177 January i8, 1989 12:57 Ph S C&IC N40 El9A702 E19A705.776 .91 .861 .837 .862 .854 .868 .768 .845 .857 .737 .846 .804 .81S.826 .852 .81 .817 .835 .842 .837 .853 .849 .904 .813 .827 .805 .921.809 .929 A72 .86 .844 .82 .872 .857 8J7 .944 .822 .858 .847 .918 8 .9 .o7 0 e84 .909 .929 .923 9 7 0 .871 .815 .826 141 .87- 0 .843 .875 .953 .969 .904 0 .834 .838 1.011.939 .872 .948 .902 .945 .921 .956 .93 .837 .853 .89 .918 .919 .919.967 .884 .951 .965 .942 .894 0 .942 .864 .846 .897 .968 .915 .913 819A708 El9A709

• U~UtmlN.W*NN N U NNN **WNW*N U *U.766 .843 .808 .729 .827 .785 .791 .801 .838 .814 .712 .826 .772 .788.841 .822 .899 .792 .807 .781 .839 .87 .821 .893 .799 .818 .786 .847.868 .844 .9 .83 .822 .835 .836 .834 .844 .915 .844 .82 .834 .836.917 .925 0 0 .83 .813 .827 91 .921 0 0 .836 .811 .82 1.007 .948 0 0 .828 839 .894 :98 .979 0 0 .825 .839 .9.934 .835 .854 .881 .925 .898 .916 .925 .834 .854 .884 .923 .893 .926.912 .86 .847 .916 .958 .906 .902 .92 .861 .856 .932 .977 .905 .918 El 9A807 E19ASIO.729 .841 .831 .714 .804 .74 .767 .724 .806 .793 .668 .76 .73 .753.858 .843 .911 .769 .794 .757 .84 .842 .807 .874 .78 .78 .747 .819.852 .823 .896 .859 .e39 .824 .815 .854 .828 .906 .783 .865 .852 .831.912 .913 0 0 .821 .804 .804 .901 .891 0 0 .829 .793 .814.946 .897 0 0 .814 .818 .898 .947 .889 0 0 .818 .81 .88.938 .812 .839 .884 .903 .882 .925 .883 .799 .845 .869 .907 .904 .903.927 .856 0 .899 .999 .866 .874 .884 .842 .825 .921 1 .877 .828

//q LINEAR REGRESSION PLOT FOk DW WALL THINMING ANALYSIS OF SECTION 19A 4 DEP VARIAKiE:

MEASURE I SOURCE DF MODEL I ERROR 4 C TOTAL 5 ROOT MlSE DEP MEAN C.V.ANALYSIS OF VARIANCE SUM OF MEAN SQUARES SQUARE 1236.97830 1236.97830 163.02170 40.75542519 t400.e0000 6.383998 R-SQUARE 860 ADJ R-SQ 0.7423253 F VALUE 30.33$0.8036 0.8544 PROB)F 0.0053 VARIAKLE DF INTERCEP I YEAR I PARAMETER ESTIMATE 876.92796-23.73464127 STANDARD ERROR 4.01529094 4.306919802 PARAMETER ESTIMATES T FOR HO: PARAMETER*S 218.372-5.509 PROD ) ITI 0.0001 0.4533.3 TYPE I SS 4437600.00 1236.97830 STANDARDIZED ESTIMATE 0-0.93997656 Des I 2 3 4 5 6 SUM Of RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)ACTUAL 804.0 873.0 959.0 858.0 849.0 937.0 7.95808E-163.02 346.34 NUMBER EIGENVAL 1 1.760?2 0.2392 PREDICT STO VALUE PRE 876.8 4.872.1 3.9466.1 2.863.5 2.843.6 3.837.9 4.COLLINEARITY DIAGNOSTICS CONDITION VAR PROP IUE NUMBER INTERCEP Is 0.11962.712624 8.8804 I ERR LOtKR93Z UPPER95Z DICT MEAN MEAN 0153 865.7 888.0 4061 862.6 891.5 8317 958.2 874.0 6831 956.1 071.0 9566 932.6 854.6 7069 824.6 051.2 VAR PROP YEAR 0.1196 0.8804 LOUER95X PREDICT 855.9 852.0 846.7 944.3 922.7 015.7 UPPER9SX PREDICT 097.9 892.2 885. 5 882.7 864.5 860.0 RESIDUAL 7. 1721 0.9491-7.0998-5.5127 5. 4006-0.8793 a ol 13 17.43 Z-- q 56- 1)- P p7 7 ,S .7,Z q2 .'f, 4 LINEAR REGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION i9A DEP VARIABLE; MEASURE SOURCE DF MODEL I ERROR 4 C TOTAL 5 ROOT MSE DEP MEAN C.V.ANALYSIS OF VARIANCE SUM OF MEAN SQUARES SQUARE 4236.97030 1236.97830 163.82170 48.75542519 1400.00000 6.393998 R-SQUARE 860 ADJ R-SQ 0.7423253 F VALUE 30.351 9.8036 0.8544 PPOB)F 0.0053 I1 VARIABLE INTERCEP YEAR Of It I PARAMETER ESTIMATE 876.92706-23.73464127 PARAMETER ESTIMATES STANDARD ERROR 4.01529094 4.3"19502 T FOR HO: PARAMETER-O 218.372-5.509 PROD ) ITI 9.0001 0.0053 ailS 1 2 3 4 5 6 ACTUAL 084.0 873. 0 859.0 858.0 849.0 837.0 COOK IS PREDICT VALUE 876.8 872. t 866.1I 863.5 843.6 837.9 STD ERR PREDICT 4.0153 3.4061 2.8317 2.6R31 3.9566 4.7969 LOWER95%MEAW 865.7 962.6 959.2 856.1 832.6 924.6 UPPER95Z MIEAN 888.0 901 .5 874.0 971.e 954.6 051.2 LOUER95Z PREDICT 955.9 852.0 646.7 844.3 822.7 15 .7 UPPER95Z PREDICT 897.8 892.2 885.5 8R2.7 864.5 R96.0 RESIDUAL 7.1721 0.9191-7.0995-5.ý127 5.4006-0.6793 STD ERR RESIDUAL 4.9632 5.3995 5. 7216 5.7928 5.e*00 4.2238 STUDENT RESIDUAL 1.4451 0.1702-1.2409-0.95 17 1.0780-0.2082-2-1-0 1 2*1lN, II Ots D 1 0.603 2 0.006 3 0.189 4 0.097 5 0.362 6 0.028 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID St (PRESS)LI tou (A, 6 6 Il 4 J-J 7.95808E-1 3 163.0217 346.3443 r S Ul LINEAR REGRESSION PLOT FOR DU WALL THINNING ANALYSIS OF SECTION 19A 0 PLOT OF PLOT OF PLOT OF PLOT OF MEASUREUYEAR PRED*YEAR U95*YEAR L93UYEAR SYMBOL USED IS X SYMBOL USED 1S P SYMBOL USED IS U SYMBOL USED iS L P R E D I C T E A L U£teee ÷I?50 +Ix IL I I 850 800).J u K L U U P P K x L L 0 U x P L U P K tA zJ 0, I L 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.YEAR LU

---

I 4 LINEAR REGRESSION PLOT FOR DW UALL THINNING ANALYSIS OF SECTION 19A PLOT OF RESIDeYEAR SYMBOL USED IS R R E S I 0 U A L S 20 ++-4-6-6-1 +ft (3 4 I I I R R R I- ---1 IE-~I~i~1 I0I%~fl I St.iJ R---- --------- -------4.-.........---

---. .+-.. .---+ --. ------ --..-4 -------- -------0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.B 2.0 2.2 2.YEAR v-I--F-i-~qQ 7 zC

/1" LINEAR REGRESSION PLOT FOR Dd WALL THINNING ANALYSIS OF SECTION 19A UNIVARIATE I, 4 4 a VARIABLEaRESID RESIDUALS I1 MOMENTS QUANTILES(DEF-4)

EXTREMES N MEAN STD DEV SKEWNESS USS CV T:MEANMw SCN RANK NUW' "1 0 U:NORMAL STEM LEAF aI 6 1 .326E- 13 5.71002 0.00192?7 163.022 99999 5.690E-14 8.5 6 0).9421973 SUM WGTS SUM VARIANCE KURTOSIS CS$STO MEAN PROSB)ITI PR)D)ISI 6 7.95BE-13 32.6643-1.66529 163.022 2.33111 I I 100% MAX 75Z Q3 set MED 252 Q1 RANGE Q3-Qt MODE 7.17214 5.84351 0.0198766-5.9095-7.6"9 14.2719 11.753-7.6999 992 95!95Z Is%7.17214 7.17214 7.17214-7.998e-7.0998-7.0998 LOWEST-7.0998-5.51273-0.979314 0.919068 5.40064 HIGHEST-5.51273-0..879314 0.919068 5.40064 7.17214 0 .635 9 2!UtIXPLOT U--,--.I INORMAL PROBABILITY PLOT-01 1-0 76 2 MULTIPLY STEPI.LEAF BY t0*04OI-7.54-2 -1 a +1 +2 iA Calc. No. C-1302-187-5300-005 Rev. Ho. 0 Page &,týf-t 5.1.5 Bay 19B: 5/1/87 to 10/8/88 Five 49-point data sets were available for this bay covering the time period from May 1, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(I) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 99% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

standard error is 856.5+0.5 mils.(6) The corrosion rate + standard error is -29.2 +0.5 mpy.(7) The measurements below 800 mils were tested and determined not to be statistically different from the mean thickness.

I N-S i T!Ca No Rt I'"% .S- e s,-jio2- 1J7-S3ao-0,S[vVo

$ ýl"D PROGRAM: DWCHISQ ENTER NAME OF DATA LIST d19b ENTER PT NUMBER LIST ints(1 .49)ENTER NAME OF DATE LIST d345.-67 N D19B I D19B612 DI?19705 3 DI 9B708 4 D19B709 5 Di9r.807 6 D19819B0 ENTER NO. OF DESIRED DATA 2,7,4.5,6 Dl 9R705 D 198708 D t 9P7708 DI 98?709 Di 9B807 DI 9.481 0 CHISQ 3. 2344 2. 3594.74185 2.3425 2. 8577 D3456 5/1 /87 8/1 /87?/10/8 7/1 2/8f I 0/08/1 CH1952 5ý. 99 5.99 5 .99 5.99 37 38 MEANT1IK**f*. f****.a 87 V3.89221.8876.86398.85641 SD STDERR 0='"1 " 6 ) 00.5-76,6 .0082294

.0086491.057 59 .0088864 05,6871 .0088817.053922 .0077031 DF M21 ft N. * *(2 C'H199 2 9.21 9.21 9.21 9.21 EXP 10.381 10.169 8.8981 13.6663 10.381 9.2826 9.0931 7.9565 7.767 9.2B26 9.6726 9.47., 8.2908 8.0934 9.6726 9.2826 9.0931 7.9565 7.767 9.2826 10.381 10.169 8.8981 8.6863 10.381 BPS 13 11 8 9 12io 10 3 5 10 8 7 12 9 7 11 10 12 12 9 6 10 GRAND MEAN THICKNESS

= .87956 STANDARD ERROR OF THE GRAND MEAN = .0081549 Jauarits18, 1989 954 .9 92 .975 911 :827 .92j .7 .95 .98 .999 .817:a91 46 :91v :9 6 .36:19 :tj:14 91 :.02 9.602:,; .396 9 Ole -i lI :4 89 I, & .91 .9o 6 .94 Dl 93769 D191807:134 sg .96, t .111 .13~ :I~ :14 :1 5 *.957 SJ.:85::7

1176.924 64 74 .99 .89 .882 .9 .8 97.77 :885 4 4 .9 .9 ,6 9 at. .05 j, if 1. :l4.911 .8892 .917 0 .99i4 .9 .9. 912A .9 I .94191 .11 I I G 1154.916 H169 .It6 .689 .~ ad 5.79 :13 71 9 37 .~.79 '1 0 1f r LrNEAR REGRESSION PLOT FOR DW4 WALL THINNING ANALYSIS OF SECTION 19R DEP VARIABLE:

MEASURE SOURCE DF MODEL I ERROR 3 C TOTAL 4 ROOT NSE DEP MEAN C.V.ANALYS SUM Of SQUARES 1361.79728 1.4e272314 1363.20000 0.6837941 579.6 0.07773921OF VARIANCE MEAN SQUARE 1361.79728 0.4675?438 R-SQUARE ADJ R-SQ F VALUE 2912.4?2 PROR)F 0.0001 0-.990 0.9986 0 VARIABLE DOF INTERCEP I YEAR i PARAMETER ESTIMATE 898.63049-29.24169165 STANDARD ERROR 0.466T5769 0.54184669 PARAMETER ESTIMATES T FOR HS: PARANETERwS 1925.261-53.967 PROB ) ITI 0.*0001 0.0001 I TYPE I SS 30684B8.80 1361.7972R STANDARDIZED ESTIMATE 0-0.99948537 OBS 2 3 4 5 SUM OF RESIDUALS SO" OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)ACTUAL 898.0 892.0 899.0 864.0 856.0 3.97904E 1.402 4.544 NUNBER EIGENVAI f 1.755i 2 0.244: PREDICT STt VALUE PRI 899.6 0.991.3 0.998.1 0.863.5 a.656.5 0.COLLINEARITY DIAGNOSTICS CONDITION VAR PROP IUE NUMBER INTERCEP 408 i.000000 8.1223 312 2.679471 0.87??D ERR LOIMR9SZ UPPER9SX EDICT MEAN MEAN 4668 897.1 900.1.3744 890.1 892.5.343e 887.0 S99.2.4267 862.2 864.9.5262 B54.8 858.2 VAR PROP YEAR 0.1223 0.8777 LOWER95%PREDICT 896.0 888.8 885.6 861.0 853.7 UPPERO5Z PREDICT 90i.3 593.7 890.5 866.1 859.2-0.6305 0.7384-0.0742 0.4595-0.4932 4.°I1,;4-13 723 196 t95(/4-2)r= 3.J)2 41/4~~ //,,ck.qlD),o

-&:~;~-~. S .t 0.5. ~

(f 0 a LINEAR REGRESSION PLOT FOR Did WALL THINNING ANALYSIS OF SECTION 19B 0 DEP VARIABLE:

MEASURE I SOURCE DF MODEL I ERROR 3 C TOTAL 4 ROOT MSE DEP MEAN C.v.ANALYS SUM OF StUARES 1361.79728 1.40272314 1363.20000 6.6937941 879.6 6.07773921 3IS OF VARIANCE MEAN SQUARE 1361.79728 0.46757438 R-SQUARE ADJ R-SQ F VALUE 2912.472 PROD)F 0.000t 0.9990 0.9996 VARIABLE INTERCEP YEAR DF I I PARAMETER ESTIMATE 898.63049-29.24069163 PARAMETER ESTIMATES STANDARD ERROR 0.46675769 8.54184069 O T FOR HO: PARAMETERmO 1925.261-53.967 2 3 4 5 ACTUAL 8998.0 892.0 999.0 864.0 856.0 CooXI $D PREDICT VALUE 698.6 891.3 899.1 863.5 056.5 STD ERR PREDICT 0.4668 0.3744 0.3438 0.4267 0.3262 LOWER95Z MEAN 897.1 890.*1 097.0 862.2 954.0 UPPER95Z MEAN 900.1 892.5 09 .2 864.9 858.2 LOWER95Z PREDICT 996.0 98G.B 885.6 861.0 853.7 UPPER95X PREDICT 901.3 893.7 890.5 866.1 959.2 RESIDUAL-0.6305 0.7384-0.0742 0.4595-e. 4932 PR0V ) Ill 0.0001 0. 008 1 STD ERR RESIDUAL 0.4997 0.5722 0.5911 0.5343 0.4367 STUDENT RESIDUAL 1-0 1 2-1.2617 I 0*l 1.2906 I In"-0.1256 I 1 o.9600 I I1-1.1293 1 *~I ObS 1 0.694 2 0.357 3 0.003 4 0.236 5 0.926 SUM OF RESIDUALS SUM O0 SQUARED RESIDUALS PREDICTED RESID SS (PRESS)S ,41 I 1 LLJ 3.97904E-i 3 1.402723 4.544196 4 LINEAR REGRESSION PLOT FOR DU VALL THINNING ANALYSIS OF SECTION 193 PLOT OF MEASURE*YEAR SYMDOL USED IS X PLOT OF PREDUYEAR SYMBOL USED IS P PLOT OF U95%YEAR SYNMOL USED IS U PLOT OF LD3WYEAR SYMBOL USED IS L 1000.4 4 Y50*4 P E 0 IL I I v I E YO *3 IL tA L o 1A BO----------------------------------------------------------------

---

------- ---------------

---

~0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 YEAR -NOIE. 10 ObS HIDDEN C.(0 4 LINEAR REGRESSION PLOT FOR DII WALL THINNING ANALYSIS OF SECTION tMB e \41 PLOT or RESIDwYEAR SYIIOL USED IS R I a 0))4 +0 IR..2 +4 i a'I, R E S U A L S R R ft R-4 4,-6 .-8 *-o 0*I~Ij'-- -------------------------.......

---

.-

--..---------------------

4 -----------------------------------------------------

.0 0.1 0.2 0.3 0.4 0.5 0.6 0. 0.8 0.9 1.0 1.1 1.2 1.3 1.4 YEAR (K LINEAR REGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 19b UNIVARIATE VARIABLE-RESID RESIDUALS MOMENTS N MEAN STD DEV SKEWNE$S Us$CV T:NEANO0$GN RANK NHU Is 9 W:NORMAL 5 7. 958E-1 4 0.592183 0.259109 1 .40272 99999 3.005E-13-0.5 5 0.925954 SUW WGTS SUM VARIANCE KURTOSIS CUS STD MEAN PROB)ITI PROS) I$1<5 3.979E-i3 6.350681-2.31676 1.40272 0.264832 0.497 100e 752 502 252 01 MIAX 03 NED Q1"I N QUAWTILES(DEF=4) 0.738413 992 0.599975 952-0.0742422 90!-0.561854 1e2--.636493 52 1.36991 1.16063-0.636493 EXTREMES 0.738413 0. 738413 0.739413-e0638493-0.630493-0.630493 LOWEST-0.630493-0.493215-6.0742422 0.459537 0.7384t3 HIGHEST-0.630493-0.493215-0.0742422 0.45Y537 0.738413 RANGE Q3-Q1 MODE a STEM LEAF 9 64 1 46 6 2 0-07 7-2-49 9-63 1 MULTIPLY STEM.LEAF BY te*4-01 D0XPLOT I S4. I I I 4.-. ..4.0.74-0.74.NORMAL PROEABILITY PLOT N4 ++4.4+ N---. .--. .-- ..-- -----" ---- ---------.-

---

2 -1 + ÷2 U'w~~1 ci 1~I I'j~Li I0/25/-0f 14:2e!O" Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 44f 5.1.6 Bay 19C: 5/1/87 to 10/8/86 Five 49-point data sets were available for this bay covering the time period from May 1, 1987 to Cctober 8, 1988. Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 91% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 860.9+4.0 mils.(6) The corrosion rate + standard error is -25.9 +4.1 mpy.(7) The measurements below 800 mile were tested and determined not to be statistically different from the mean thickness.

)

1C/Z5/C6 ;4:ZE:U7 NO Re. No ,Sh-eeNo tC. 13o2-/& 7- s3 o -,=,s[PROGRAM: DWCHISQ ENTER NAME OF DATA LIST elc ENTER PT NUMBER LIST int'(1,49)

ENTER NAME OF DATE LIST cI34567 N 3 4 6 Et9C El 9C62 E190705 El9C708 El 9C709 El ?C817 El 9C81 0 ENTER NO. OF DESIRED DATA 2.3.4,5,6 El 9C(05 El 9C710§E1 9C709 E19C8{}7 El 9C807 CHISQ 2.793 3.2861 1.2392 1 .2081 1.3084 D.3456 5/1"87 3/1/87 9/1C/-s 7/12/OE CH1952 5.99 5.99 5.99 5.99 5.99 7 1*1 iS MEANTHK.90051.88816.88831.87346.85627 ED 08 1 25 (-. 91 1154-,')63771-728128.072399 Y:TDERR 01)2112.012234 9098401.013016.(10915 1) FM21 2 2 9 CH1992 9.21 9.21 9.21 9.21 9.2i EXP?.`337 9.3218 8.8981 8.2625 9.3218 8.5`248 8.3354 7.9565 7.3882 8.3354 8.883 8.6856 8.2908 7.6986 8.6856 8.5248 8.3354 7.9565 7.3882 8.3354 9.5337 9.3218 8.8981 8.2625 9.3218 OBS 10 12 11 10 7 6 4 6 5 10 12 10 8 8 8 Ic' 8 7 8 10 7 10 10 8 9 GRAND MEAN THIVTCKNES

= .98134 STANI)ARD ERROR OF THE GRAND MEAN = .0075929 January i8, 1989 t21: 5 ?PrM E19C705 r19C ?08.969 .927 .92 1.067 .996 t.112 1.104 0 .818 .864 1.06 .952 1.016 .961.775 8 ,839 .929 .894 1.03 .891 .767 .804 .97 .f74 .807 .933 .895.882 A0 .936 .965 .908 0 .912 .83 10. .O7 9 .947 .868 0 .998 S.761 .61 .75 0 .71 .7.0 :0 :.833 .823 .843 .864 0 834 ~9 81.64 .823- :87908 .01.M 8P4.926 .978 .92 .796 .894 .813 .856 .914 .874 .915 .789 .898 .775 .91.897 .921 .936 .899 .92& .955 .989 .918 .864 .894 .97 .973 1.039 .956 EI9C709 E19C807 WW**WWNNWWUNUW*NMNMNNMNN*WWNW*WUUWWWUW UWNWPWW*WNWU*,,*M~gWW*

.92? .802 .891 1.02 .961 9 .944 .871 .781 .874 t.066 1.005 1.056 .947 0 .792 .785 .059 .812 I3 .874 70 .78 .757 .847 .786 .959 .866:33 ' :2 3 .9 ".887 903 : A 0  : -.85 :774 .78 0 0 :93.875 .2t .827 .85 0 .837 .893 0 .804 0 .87 0 0 .879.942 .884 .926 .839 .898 .791 .958 0 .967 .917 .803 .879 .809 .9B7.866 .942 .944 .926 .958 .982 .993 0 .853 .916 .902 .944 .932 .949 0 .777 .965 1.002 .907 .975 .935.705 .727 .2 .848 .803 .931 .864.856 .IM :4.11 .905 .85 0 .897.834 .91 .735 .768 0 0 .934.82 .7M .965 W82 0 .900 .86.87..74 .3 .81 .84 ."13.R62 .878 .91N 121T 905 :19412.93J.rI (A]a f./I a V LIER IEf-d:SSIO jaw~Fil? U WALL. 1HTNUE NG ANALYXSS Ov PAY I 9C V AIIOVF CURPt)OHN4 YCARS hit$3 0.36 888.3 4 1.20 073.5 5 1,44 956.3 13:0O3 UFt)NEXIAY.

JAWIARY 4. 1989 7o, a Li V LIWfAR KGtRE SIUM 1.-L01 FOR DW WlALL 114j HUH. ANALYSIS OF PAY VYC 3 APOVE CORP9 M~AN SI1RR STD ERROR D(.Vi11 N OF MFAN Bei.36000000 fh.96372601 1.58640890 13'03 WEDR'.¶DAY, JANUARY 4. 1989 8 VAR IAIALE MILS: N r FIR) IT I 116.19 0.000i..tA w (A 6 LINEAR RELGPILSSION 11.011 fIUR tDW AUAI 114IINNINt ANALYSIV OF~ PAY 19C 3' AI4OVE CIJRP JS 03 UW.DWSPAY, IANjAr~y 4, t989 9 DI-" VARiAIAII till S SOURCE Dr C IlIrAI. 4 ROOT IISL~EP MEIAN C.V.AlNALYSIS OF VARIANCE AUAR S SQIZARtI.ti0st j48 11i 48 F WALKE 40.1 ?4 I'R0.0 ) F O.0019l~.162:1 ?2 s813A 3 05857 166 R -.SQUAR AbDJ R--sQ VAR 1A14I.IwI ERCEP Y A 'S DF i i PAiAMF Tkh rSTIMATE S. 926i"Wi ERRFOR s-~ g FARAPIEtrR USTIhIArEs I FOR ti.CflAME)E lI0!'ROB ) I I I iY pri I s.0..OA4641119 90.0U7 01 .24 V007 M .24 COIL IffARI1Y DlAL.NDSTTCS

¢:, ovis 4 5 AC rthAl¢ 00 5 138.2 B838.3 ti'e. 3 0*13.5 13',A .3 NUM13I. I. I lEIVALIUE'lay. 3.5;09 860.9 3.9719 cnH Ii f ION LOuERY 880.856.82 VAR PICOP UPPER 5 89r.1I orp.4 8,13.5 V..AR poS o ~: 43 3 9130.5 847.?1 1440. 1 lITPE kySz 907.3 881.6-V1.1162-0.5794 6. 4009-4.'%?61 pL n1 81Sr D r.. S I DUALS PRI.DICErD RFXID SS (RPK'SS)i .19371E -i2 79.94 716.?O.0524 t 1:5 (") -2) = B. /,? 2.e, L Aleilp cf4o,9 t 4,ol;pý(A-Vc

/xv I INEAR REtklf:SXION PtO1 mrn Dw WALt ItHINNING ANALYSIS Of PAY I9C 3' A1OVI: CrOPP iA 03 WLI0N.SDAY.

JANIOARY 4. I(Dt.I" VoldA"Lil.

MlT S SOURCE DF C; TOTAL 4 VANIAwL.F DFl NAfRIEP I Yt Rs t 2'ANAIYSIS OF VARIANCF..1)1* OF MEL _XiJARES' 5051AR 10t; .tfl4j4 107¶1 '414 9.9416 92 .6.6 4 0" 1 ilsi.072(00 S1.162?72 R. SOUAkS.1 .36 ADJ R-$1I 0.5.?O7t 6A F* VAt OIr 40. 194 0.9305 0.9074 T FOR ffO;PARAMtFILr'O FPFROP : F 0.0019 FARAMILTER tSFIHATC 5.920.94103

'AcAMrTER ESI IMIATIS SIANDARD ERkflF 3 5.529l6 4.069's'.442

• ROl ) I I I 0.000i 0.0019 4014y 4 S 900.5 OUR0. ?531113. A4 173.5!8056.3&S I 01545 VI',.1 Ill-T 091 .7 06?. 1 A60.9-- 2I.. I 2 79.94716:980.05:'4 XFD lICI 3.5ý.09'.3955 3.2233 3 .?110 t 1flF R957.ME.AN 856.8O R48.2 COOKIJ ' S D 0. 15?0.144 0. 003 0. 805 I. 395 thII~ll.'95Z MEAN 909.4 900.?897.1 871.4 873.5[I WLR91 z PRLD1el 813.0 810.5 847.1 840.1 918.1901.3 806.5 881 .A r-.*F IA DIJAL 2. 1162 3.531 A"0 .ti 194 6.4009-4.576t s ID lk.Rw P.E S T DUAI 35. 71!12 4.31/1 4. 4623 4.032.3.5.-101935 1 -.C)8171 3 "0. 12W8 4 1.5874 S ..58/.3 SlIp IO 1% SIDUAL. S S Uh InI s IJAN10D RLS IJ.AL "!N I'l lt) II: D s5 (frcrsX )I 0 i I ! NI.A1R IFG3RL1SUIN PLOT f0li Di IJWALL III.INNIN AHALYSIS OF DAY iyv 3" AT4VF_. curl[4'loi Flh.SmYEARS SYmpIY .UIiE11) J S X P-L"iOT UF U9RED*YE k.V SYMu1 ,l LISE s 11 f'tlT OF Lg."9Y.ARS syvt.tso 1JS1t Is 11 P'LO'T [IF L95%YE.ARS SYMIBOL liJsC Is L 1.3 03 t.DNI.SDIAY, JANIIPA1Y

4. 49119 it I:.D v I)I E V A L E 1200 11%0 1 100 1050 9.~3 v 700 R00 4.1 I it I I1 4.jl J*.0.,.e.U I.x 0.1 O.. 0.3 II x I.IJ x 1,..,4 0.4 0.5 0., A ./ 0.10 7.9 1 0 V EARS.4. ~. --9...t.~ t..~3 MI:S 141vDoN t I I,-----I I IHF.Ak RE.&St..LON PLKOI FiO rTU WIALL THINNIMN.

ANALYSIS OF' DAY i9t .3' ABOVE I.Ukd 11.111 (If RIXLSTDUAL wYVAI." sy.tMll IISFI) Is k<t$303 JAMIARIcY 4, 19g?I..L I..4 x 404 L4 +-O4o 20 .16 +14 +4 4 10 4 7 4 14 4 10 4 30 ,4 4.4'2?-30 3.4-36 40 0. 0 1" HC5 I (A.. t. .. ............

...-. --. + ..-0.1 ).: 0.1 0.4 0. S 0.6-.--*-.4.4. .4.4.4..-1 1.4 .0 (A

.NCaic. No. C-1302-187-5300-005 Rev. No. 0 Page of 5.2 W"x6" Grids in Sand Bed Region at New Locations 5.2.1 Bay 9D: 11/25/86 to 12/19/88 The 6'x6" grid data was taken in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection system being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, previous measurements were taken in November 1986 in a 10-point 6"x6" cruciform pattern.Measurements were also taken in a 6"x6" grid in December 1986. The new data were compared with both of the previous data sets. These comparisons were made using the chi-squared test, F-test and two-tailed t-test as described in paragraph 2.5. The mean thickness was determined as described in paragraph 2.8,3.(1) The data are normally distributed.

(2) The variances are equal in both comparisons.

(3) It is appropriate to use the two-tailed t-test in both comparisons.

(4) The difference between the means of the 1988 49-point data set and the 1986 10-point data set is not significant.

However, there is a significant difference between the means of the 1988 49-point data set and the 1986 49-point data set. Therefore, significance of the corrosion rate is classified as"Indeterminable".

(5) The current mean thickness

+ standard error is 1021.4+9.7 mils.S Nuclear Calculation Sheet I Cawc NO Re. No I S#e¶ No-o -'o -,- -`-_ __ o..5..?./1&9Y 9/) (aIAJ7-Iu OCDWCONF /,2-/If PAY' ?P D8702659 ftf tf t* * **f * *t *1 *t ** * *t *t f * *t ** ft *t ; ;** t * *** ft f It ftft4tf

• 9***-t itftt 1.0?7,? 1.104 1.034 1.144 1.09 1.14 1.'-;__'7.949 .964 1.089 .993 1.084 .1I9...7 .953 1.054 1.077 1.031 1.117 1 vi54.919 .9519 .976 1 .038 1 -038 1 .0-31-S 1..056.971 1.046 .922 .991 .947 .983 1.031.987 .947 1.021 1.026 1.041 1.015 959 1.02 1 .975 1.021 1.022 .947 .976 MEAN 1HICKNESS

= 1.0214.S"TANDARD ERROR OF THE MEAN = .0097164 J(.05/2. 48 ) 2.0106 T( .131/2. 48 )2 26822 CONFIDENCE INTERVALS FOR THE MEAN 95% UFPPER ROUND = I .0409 95% LOWER BOUND = 1.0019 Janr 16ý, 1989 12:3 10.m!'RfGRAM OCDWCOI1F" /.; --6!)860)494i 1.175 1.162 1 i?4 1.13 1.182 1.i62 1.12 1.197 1.219 1.168 1.143 1.119 1.045 1.':25 1.145 1.107 1.085 11.i26 1.132 1.085 1 .e2 1.119 1.119 1.031 1.038 1.048 1.061 !.074 1.07 .963 1.03 1.051 1.007 .?91 .983 1.063 1.059 1.059 .968 .977 1. '2. ..2.,, 7 1.079 .987 1.049 .926 1.018 .984 .9,28 MEAN THICKNESS

= 1.071'STANDARD ERROR OF THE MEAN = .010397 T(.06/2. 48 = 2.10106 T(.01/2. 48 = 2.1822 CONFIDENCE INTERVALS FOR THE MEAN 95% UFPER BOUND = 1.0924 95% LOWER BOUND = 1.05"-01?9% UPFER BOUNT, =" .0994 t9% LOWER iBOUNI 1 0 ,436 juaAary 6. 198?5:, .MP* 0016 (06-10/25/0( 14:2ý:C7]Nuclear Calculation Sheet 4.x- 7 -S7 30 -60 )ý'R:'IGRAM " OCDWCONF IATY: 9D Drq 6 49 19 1 .114 1 .054.997.732.985 1.06.981.9?4 MEAN THICKNESS 999 STANDARD ERROR OF THE MEAN = .032583 T(.O5/2, 9 )= .2622 T(.01/2. 9 ) 3.2498 CONFIDENCE INTERVALS FOR THE MEAN 99% UPPER BOUND = 1 .1049 99% LOWER BOUND .89311 January.16, 1989 5:3e -w &.5C 'OF Ti mi r O / 5, lE..9 -- o. cA In./.2 C4 9(- /. 6715 OJ " N 0086 (06-84 lC~sc NO Shet N PROGRAM: DWCHISQI BAY: 9D D870265¶CHISG 4.4488 nfl **, 9 9.13 9.4 9.12 io.DATADATE 12119/88 MEANT'HK 1.0214 SD.068015.0097164 DFM2 DE: C HI 9952 5.9Q CH1992 9.21 LXP.381 2826.6726 2826 ,38 PTNOS PINOS 1 11 313 4 14 5 15 6 16 7 17 8 18 9 19 10 20 JanuArv 18. 1989 1:17 PM PTNOS 21 23 24 25 26 27 28 29 30 PFTNOS 3 1 32 33:Z4 35 26 37 38 39 40 PTNWOS 41 42 43 44 45 46 47 48 49 I-Nuclear Calculation Sheet 0o O,.Y.P~Tor Oaleeew.bvDl 6,.ýAWY 90 6 MJTI PROGRAM: 2.AY: 9D DATASET D8702659 DATADATE 121t9/88 MEANTHK 1 .-0214 SD.068015 1FM2 CHISte CHIT.4.4488 5!.99 OB$9 13 4 12 PT N O 2 3 4 6 7 8 9 10 EXP 10.381 9.2826 9.Z726 9.2826 10. 381 PTNOS 13 14 is 16 17 18 19 20 PTNOS 21 2-2 23 24 25 26 27 28 29 30 PTNOS 31 32 33 34 35 36 37 38 39 40 PTNGS 41 42 43 44 45 46 47 48 49 Jantjr6. 1989 S N 0016 (064 M:i2~,'OE~

L4:2E:C?o 52-eo o SeeNo 0 PROGRAM: DWCHISQI tAY: 9D D8604941 CHISQ 1.5i67 OFS E 10 9.10 91 12 io.: DATADATE 12/04/86 MEANTHK SD.072782 S1T9E R R* DFM2 C1.952 5.9 CH1992 9.21 XF 381 2826 V726 2826 381 PTNOS PRNOS PTNOS¶ 11 21 212 22)3 13 23 4 14 24 5 15 2 6 16 26 7 17 2 ?9 19 29 10 20 .30 January 18, 1989 I1-18 PmI 31 32 33 34 35 36 3?38 39 40 PTNOS 41 42 43 44 45 46 47 48 49 01' 42 : r 2 Rev -o 'S01~o cait No P.N h tN-V PROGRAM: DWCHISQI BAY: 9D DATASET D8604919 CHIXO 8.3595 OPS E)1 2.11 DATADATE 11/25/86 MEANTHK?99 SD.1 0304 S TDERR DFM2 2 H1952 5.99 CH1992 9.21 (P 186?44 74'44'86 ETNOS'1 3 4 6 8 10 Ja Pay's 1989 42"Is PM 0 4r/51e6 14:2&:Oi[ fNuclear Calculation Sheet C -1 so --1 COMPARISON OF MEANS USING TWO-T-AILED -T-EIE BAY DATASHTS DATASETS DATADATE MEANTHK 9D 3702659 D8702659 12/I 9/88 1.0214 8604919 0860491? 11/25/86 .999 D8702659 f****f****f**f*f*ft*****ft***ft*ftftftf**ftf***tf****

1.077 1.104 1.034 1.144 1.09 1.14 1.157.949 .964 1.029 1.089 .993 1.084 1.109.827 .953 1.054 1.077 1.031 1.117 1.154.919 .959 .976 1.038 1.038 1.038 1.056 SQ 1.046 .922 .991 .947 .983 1.031 9 I7 .947 1.0-21 1.026 1.041 1.015 .959 1.02 1 .975 1.021 1.022 .947 .976 D8604919 1.114 1.054.732.985 1.015 1.058 1 .06.981.994 F TEST FOR EOUAL POPULATION VARIANCES VARA VARB DFA DFB*4*4*** i*4*4** *4* *44.010616 .004626 9 48 F = 2.2949 F(.05/2, 9 , 48 ) 2.3925 F(.01/2, 9 , 48 ) 3.1133 TWO-TAILED T--TEST DF = 57 ALPHA 8 9551 T(.LO/2. 57 ) = 2.0025 T(.01/2, 57 ) = 2.6649 Janar13, 1989 5 :52 (.M*~o tctJs) .4 oO1m (w6e

[ lfNuclear:subject Calculation Sheet Calculation Sheet CAl¢ NO Ri ev No No ,id.-/3o2- -60S LA j£o,..rDae Rev.eweo by _ _ e S.2,/!!7; A~ (i, e, , J COMPARISON OF MEANS USING TWO-TAILED T--TEST BAY DATASHTS 9D 8702659 8604941 1)A!TASETS D8702659 D8604941 DATADATE 12/09/86 12/04/86 MEANTHK 1 .0214 I .0715)b D8702659 1.077 1.104 1.034 1.144 1.09 1.14 1.157.949 .964 1.029 1.089 .993 1.084 1.109.827 .953 1.054 1.077 1.031 1.117 1.154.919 .959 .976 1.038 1.038 1.038 1.056.971 1.046 .922 .991 .947 .983 1.031.987 .947 1.021 1.026 1.041 1.015 .959 1.02 1 .975 1.021 1.022 .947 .976 D8604941 1.175 1.162 1.174 1.13 1.182 1.162 1.12 1.197 1.219 1.168 1.143 1.119 1.045 1.025 1.145 1.107 1.085 1.126 1.132 1.085 1.082 1.119 1.119 1.031 1.038 1.048 1.061 1.074 1.07 .963 1.03 1.051 1.007 .991 .9B3 1.063 1.059 1.059 .968 .977 1.052 .987 1.079 .987 1.049 .926 1.018 .984 .928 F TEST FOR EQUAL POPULATION VARIANCES VARA.0052973 VARB.004626 DFA DFB 48 K4 48 48 F = 1.1451 F(.05/2. 48 48 ) = 1.7728 F(.0t/2, 48 , 48 ) = 2.13 TUO-TAILED T-TEST DF = 96 ALPHA = 3.2857E-4 T = 3.5221 T(.05/2. 96 ) = 1.935 T(.01/2, 96 ) = 2.628 Janupar 13, 1989 5:49 PM 6. _>1(55):. /~'.r~c awL./7 9/ er 7,--Ny 0016 J06-f

[f]Nuclear Calculation Sheet/ 69Y 9AD if'- c tv 4 eE7 etVashifLC A N 0016 (06-Calc. No. C-1302-187-5300-005 Rev. No. 0 Page ¶Iof 5.2.2 Bay 13A: 11/25/86 to 12/17/88 The 6"x6" grid data was taken for the first time in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, previous measurements were taken in November 1986 in abutting 6"x6" cruciform patterns across the entire bay. As a best approximation, 13 of these data points are at the same location as the new 6"x6" grid data set. Therefore, the new data were first compared with these 13 data points, and then with 21 data points which include the 13 plus 8 additional points within one inch on either side. These comparisons were made using the chi-squared test, F-test and two-tailed t-test as described in paragraph 2.5. The mean thickness was determined as described in paragraph 2.8.3.(1) The data are normally distributed.

(2) The variances are equal in both comparisons.

(3) It is appropriate to use the two-tailed t-test in both comparisons.

(4) The difference between the means of the data sets is not signficant.

Therefore, the corrosion is classified as "Not Significant".

(5) The current mean thickness

+ standard error is 905.3+10.1 mils.

4 11/2 ý,AIE 14 : 2 E ; L'J[f]Nuclear Calculation Sheet/,3/?y Z3-'q 3"i* 9460r- dLW-dS 7~~'~A~ 4~~2E T~~-VO £T~r- ' ~~' tS~7o,//-/5 -SC..87.- 5 2' so. 4kj )77~ U-5-4 A97 t4ýLo 7yetA /4) 7-wd .srnes We6.P WH/,qC IS 4 5PD9WAJ 10A- 41 /Z< ". 7-7,ý'I 6WI~~S~)S Iffo AAAO 13,6. 77-,--5 C- P*9r T-H1'ICAIC.SS 0-J /3q, 6/o9 A'tD SNe'67-/(1v,=1w-/nEA/rs Seerw!A45j 7' ~~SjAIL/?~'

Ad W-V,,CG 7;9/<ýA.)

4ed&-~ ,.9AID> 7/-*CCL e Rcz c 7-H6 6*edA4 R64DiA&5- ,Ogr /- 11r&C Amvr:* 4 0*

• 0** ~....~* *

• 0 S& S* -A4ck. QE.~P.CIt'.I~

a 0 P.)y i(A IA"-, I~ WE r.i. " 106-M)

4:2e:(,?[ ~Nuclear Calculation Sheet F-IAy 33A, Pic ,,EL-0 E " " O e J$ kL / 7"9- T.I &,E/D CCA/ /p',,".L=, 415 I" A. 5 -A? 06ýS7-,9~4,7O, 727'd6 1.2-1 7-86 &AU (,Pr AM.A7- ,4.Dpl A~r.*3~'o 4 a=.2S)tEW-6ir&e cv-l.7* 22& a a 12?a A j Appaox. LoCA T7,01J OF 12 8t GlZiO.<2 A U-C- D -A -Se t~e3A At6I.CALLC-D ,'>T-S % ) Jw6 06-86)

Nuclear Calculation Sheet Ca j eNo R h~ Nt Dl Rev.ewoa by Dale W/" n..-,.S7501 ZXI Zvv-O6 77?z .-J ,6.6 LU.&,D 7,b e m,'c$/ACF,, 7- .." 7. 7- , C ZW /. S&7:S"l-HE- 12-17-< 2,,-E Wr (,P/3,9e-/Z ) Z.&8g f1,ocZs J.eZ6,Iu J,,r,/ /oIC //---Z^.9w49 eT (D /7'6 2747,LZ).7,-) SU 7,- OFc 496qz0 .4-4 7,VS 7- 7Z C/c le 1P~0b7-6 CHRC,,t Pef/ AOU.L4 6~6A. 6r 7o -1Es 7-A4 PZO9/?0 W7H,3% 0016 (06-861 10,125/06 14:2iý:0 MJNuclear Calculation Sheet SubjectNo No Stmet N, The ,Z'° 7°r .. _ ,"- eoZ7li °sv'Oug-ato Dae Reewe bVDate 4./.s7/iAJ~a O9AJ 4Po6ýCO:Z1 ,oM7'A (= 7ft11 4C e -q SF-T 7es ~4A/zt TrHCr 7-5 re7* 7- T ,,7,;V'o",'

6 , .2 ae ,,9 -,,:A /0 @,//v_. 717"A: ,Oý44610<*/FI'f ,Tyl JI )7- XTfsr -gzz- , 1CA) 4r~e51- 771,q£ 7,y-,-- AQyor 7#-19?~r j"~ /~/1AC2r~

e~ c,5-9rde IS A4 7- 1672FC7V'" I it 1,zS7 -0,00c /1 ~ OICA4 (0 Th'40-0/.(-5) ________,o1, Alj~- // 7;/I w v, 2o/,o T_9. -:, 19 C A49W9, IS A/6 r ,'IT11 15/F ~7-A 42 0.~~N70016 1O6-86)

PROGRAM: OCDWCONF BAY: 13A.941 .862 .88 .963 1.016 1.046 .9.85 .9 1.141 .94 .892 .884 .802.837 .959 .849 .861 .809 .784 .855.907 .936 .908 .911 .843 .1397 .951.962 .834 .902 .927 .357 .925 .935 1.108 .863 .829 .998 .845 .876 .881.908 .998 .839 .879 .899 .967 .902 MEAN THICKNESS

= .90527 STANDARD ERROR OF THE MEAN = .010109 T(.05/2. 48 )= 2.0106 T(.01/2. 48 )= 2.6822 CONFIDENCE INTERVALS FOR THE MEAN 95% UJPPER BOUND = .92559?57 LOCWE:R BOUND := .88494 99% UF:'iEF; BOUND =. ?9% LOWER BOUND =.3781516, 1989 12:42 FPM 0 CA&N No Shee PROGRAM: DWCHISQI BAY: 13A DATASET DATADATE MEANTHK SD STDE'RR DFM2 D13A812 12/17/88 .9052? .07076 .010109 2 CHIS00 CH1952 CH1992 UBS EXP 9 1?.381 12 9.2826 12 9.6726 8 7.2826 8 10.301 PTNOS FPTNOS PTNOS PTNOS PTNOS 1 11 21 31 41 12 22 32 42 13 23 33 43 4 14 24 34 44 5 15 25 35 45 6 16 26 36 46 7 17 27 37 47 8 18 28 38 48 9 19 29 39 49 t227 S) 30 40 I113. 1989"':14 FPM 10/'25/06 14:2tý:01 Ca-t No R. No No~ So PROGRAM: DWCHIEOI BAY: 13A DATASET D13A8601 CHISQ* 58781 OBS E£3 2.4*3 2.4t 3 2.74 3 2.7~DATADATE tf/I 5/86 MEANTHK SD.91908 .041422 S'I'DERR* i14****.011488 DFM2 2 CH1952 5.99 CGH 992 9~ '24 (F'542 627 S&2 62 7 542 PTNOS 3 4 6 7 8 9 10 Januar" 18, 1989 2:15 ,Eer. PnrO S~tr e&-o.,'-oy 1Q/25b/v6 NO Ae'. No ISheet Wo$9 ioflt I" PFROGRAM:

DWCHISQI BAY: 13A DATASET DI3A8602 DATADATE 11/15/86 MEANTHK.9361 SD.045934',')DERR.)10024 DFM2 CHIs3 3.7292 CH1952 5.99 CH1992 9.21 0 4S 4 2 EXP 4.4491 3.9782 4.1454 3.9787 4.4491 PTNOS 'TNOS 3 14 4 15 Cý 16 6 '17 7 18 o 19 22 9 2 1o 21 Ii January 18, 1989 2:16 FP 94r-: wmn cnr 8o&-Ovf-oy 0

its-COMeAO7-N3 F 6sN IN T .-IS COMPARISON OF MEANS USING TWO-TAILED T-TEST EAY DATASHTS aIt* ****N***13A 8604909 8702658 DATASETS DI 3A8601 Di 3A81 2 DATADATE 12./ 1/88 MEANTHK.9?190.90527 Di 3A8601.903.987.934.937.862.839 ,919.887.926.932ý897 ,963.962 D)13A812.941 .1362 .188 .963 1.016 1.046 .9.85 .9 1.141 .94 .392 .884 .902.837 .959 .849 .861 .809 .784 .955.907 .936 .908 .911 .843 .897 .951.962 .834 .902 .927 .857 .925 .935 1.108 .863 .e29 .?98 .845 .876 .881.908 .898 .839 .879 .899 .967 .902 F TEST FOR EQUAL POPULATION VARIANCES VARA.0050069 VARB.0017157 DFA 48 DFB 12 F = 2.9182 F(.05/2, 48 , 12 = 2.8771 F;(.0l/2, 48' 12 ) = 4.1754 TWO-TAILED T-TEST DF = 60 ALPHA = .25229 T = .67134 T(V05/2, 60 ) = 2.0003 T(.01/2, 60 ) = 2.6603 January 13, 1989 5::58 PM L z~0 I 10/fl/Of.al 14240 TPN c /1o2Ce7- S36o -0 sIt2 t COMPARISON OF MEANS USING TWO-TAILED I-TEST PAY DATASHTS 13A 8604909 8702658 DATASETS DI3A8602 D13ABf12 DATADATE ii/15/86 12/17/88 MEANTHK.9361.90527 Di 3A8602.903.987.934.937.862.839.919.887.926.9 3t2 D13A8602.963.962.943.932.98 1 .057.956.93.954.958 D1 3A81 2.941 .862 .88 .963 1.016 1.046 .9.85 .9 1.141 .94 .892 .-84 .802.837 .959 .849 .861 .809 .784 .855.907 .936 .908 .911 .843 .897 .951.962 .834 .902 .92?7 .857 .925 .935 1.108 .93 829 .998 .845 .376 .881.908. .T a 839 .879 .899 .967 U902 F TEST FOR EQUAL POPULATION VARIANCES)V VARA.0050069 VARB.0021099 DFA DFB*41* *20 48 2.0 F = 2.373 F 2, 4 20 ) = 2.2557 Ft,:01/2.

48 20 ) = 21.9692 TWO-TAILED T-TEST DF = 68 ALPHA = .035529 T = 1.8338 T(.05/2, 68 ) = i.9955 T(.01/2, 68 ) = 2.6501 January 13, 1989 L-4 00t-C ac. No. C-1302-187-5300-005 Rlev. No. 0 Page 1, 0 7.. of 5.2.3 Bay 15D: 11/25/86 to 12/17/88 The 6"x6" grid data was taken for the first time in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, a previous 1-point measurement was taken in November 1986. The location of this point may have been somewhat removed from the location of the new 6"x6" grid data set. The previous measurement was compared with the new data set using the methods described in paragraph 2.6. The mean thickness was determined as described in paragraph 2.8.3.(1) The new data are normally distributed.

(2) The previous measurement falls above the 99% upper bound of the new data.(3) This implies that the corrosion may have occurred in the time period covered by this data. Therefore, the corrosion is classified as "Possible".

(4) The current mean thickness

+ standard error is 1056.0+9.1 mils.S C atc ka4ot NeON Sheet No 4f-SO le 7- 530O0 X'S.CA3.%PROGRAM: DWCHISQI BAY: 1 5D DATASET D155M12 CH 1 170 1.8429 OBS 8 9 13 40 DATADATE 12/1 7/88 CH1952 CI*.99 MEANTHI I 056 ZD.0636 STI00 8RR 3*

• 3*3*'1 HI792 9.21 EXF'.381.2826.6726 2826.381 PT N OS 4 5 6 10 F'TN OS'13 14 15 16 17 18 19 20 PFTNDJS 21 23 24 26 27 28 29 30 PTNI)O 31 32 33 34 36 37 38 39 40 PT NO S 41 42 43 44 45 46 47 48 49 9r*

1989.:18 FM

.4-s-/e I-Re.-0 seet Ni1 PROGRAM: OCDWCONF BAY: 15D D15D812 1.127 1.131 1.127 1.136 1.143 1.125 1.139 1.091 1.11 1.088 1.142 1.127 1.128 .133 1.033 1.035 1.03 1.064 1.1,05 1.097 1.091.989 1.023 .995 1.036 1.036 1.09 1.066.996 1.022 .842 1.053 1.113 1.063 1.047.944 .994 1.035 1.047 1.026 1.054 1.038.955 .968 .96 .99 1.016 1.071 1.074 MEAN THICKNESS

= 1.056 STANDARD ERROR OF THE MEAN = .0090857 T<.05/2. 48 )" .0106"(.01/2, 48 )= 2.6822 CONFIDENCE INTERVALS FOR THE MEAN** ** * **** *** * ***~*****

• • • • • **** **95% UPPER POUND = 1.0743 95% LOWER POUND = .0370 99v UPPE Z. EOIJND = O4, 49% LOWER BOUND -1:87 January Mt6. 1989 12:48 1,89 vC I ,-%- ,t4 CcQ.0t ~ LOn'Q~e Co o~' P5U I3/2EiOF 34:28:CýCalc. No. C-1302-187-5300-005 Rev. No. 0 Page jo5of 5.2.4 Bay 17A: 11/25/86 to The 6"x6" grid data was taken for the first time in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, a previous 1-point measurement was taken in November 1986. The location of this point may have been somewhat removed from the location of the new 6WIx6 grid data set. The previous measurement was compared with the new data set using the methods described in paragraph 2.6. The mean thickness was determined as described in paragraph 2.8.3.(1) The new data are not normally distributed.

However, the top three rows and the bottom four rows are each normally distributed.

(2) The previous measurement falls below the 99%confidence interval for the top three rows, and above the 99% confidence interval for the bottom four rows.(3) The corrosion is classified as "Indeterminable".

(4) The current mean thickness

+ standard error is 1133.1+6.9 milsfor the top three rows and 957.4 +9.2 mils for the bottom four rows.S t l q IC/2S/10 14:28:V7 a NO4.---!Rev-ko -heeP401/4PROGRAM: DWCHISQi BAY: 17A DATASET DATADATE MEANTHK D -S"TDERR DFM2 1****.** 1****** ******* ******0 ****W** ****D17A812 12/ 17/88 1 .0327 .01?7129--

.013899 2)CHIS6 CHI992 11.601 5.99 CHI 1?92 9.2 CBS 4 4 17 PTNOS 2 3 4 5 7 8 9 10 EXF'10.381 9.2826 9.6726 9.2826 10.381 F:TNOS Ii 12 13 14 15 16 17 i 8 19 20 PTNOS 21 22 23 24 26 27 28 29 30 PTNOS 31 32 34 35 36 37 38 39 40 F'TNDS 41 42 43 44 45 46 47 48 49"I Januarv IS. 1989 2:19 Ph IOI2'LiOC 24:2R,:O (alc No ee eNIo FPOGRAM; DWCHISQI BAY, 1?A DATASET DJl TA81 2 CI IS 12 5.4566 4M* Eft 4- 4.4 2 3.);3 4.1 4 4.4 DATADATE 12/1 7/88 CH1952 CI**f.** %1 IT5. ?4 MEANrHK 1.1331 I'D TE RR****** ********.0 1 .46 Z C .0 60 6.57 DFM2 H1992 9.2t XF'1491 454 4781F'TNPS F'TNOS 112 2 13 3 14 4 15 6 17 8 19 9 20 iG 21 1 i 2 : Pr F'i i .8 1989 cfl.c No Rev No Sheet '%o Ic-,3o2-le 7- 5 -3, L__PROGRAM: DWCHESOI BAY: 17A DATASET D 7A81- 2 6 52 4 5.6 5.C*7 DATADATE 12/17/88 CH1952 C*4*4*K*4*

• 5z QC ME.NT HK.95736 ED 04f!67*S01DERR.0091 919 DF ý2 H1992 9.21 EXP?321 3043 52"72 3043 93"2123 24 26 27 28 29 30 31 P TNUW 32 33 34 36 37 38 39 40 PT N05 4 1 42 43 44 45 46 47 48 49-'I Janj~ary 18, 1989 P~ROGRAM¶:

OCDWCONF B~v': h A\ "i Dl 7A81 2 01.168 1.157 1.16 1.142 1.141 1.1731t.172 1.129 1.153 i1.35 1.48 t.134 1.14,81.142 1.063 1.t46 1.113 t.1tl

.069¶ .993 1.00I 1.011 1.0-35 1.006& .968.976 .925 .934 .965 .89 .969 1.0323.3-79 .983; .9i,6 .873 .846 .95:2 1.012.9?92 .97 .951 .924 ,929 .912 .971 MEAN T

= 1.0327 STANDAR!)

ERROR OF THE MEAN =.0'13899 T(.0..5/2, 48,)= 2.010 g'6 1 (. e1/2. =

CONFtDEN4CE INTERVALS FOR THE MIEAN 95%/ UPPE'E: BOUND 1 q.0606 95%. LOWER f.QUND = 1.0047 99% JF'PER E'OUJNP 1 .07 99%. LOWER" ROUJN .995396 1989 K l~alc Noe. No ' Sheet ,N.PR~OGRAM : OCDWCON.F RAY: i1?A TOP3 1-168 1.1R7 1.16 1.142. 1.141 .71.2 j .i J l71. i 3 1A3.13` 48 1 .34 1 .148 1 .1 4i2 M~EAN THIC~KNES

-1.1331 STANL)AFA)

ERROR* OF THE MEANz ,)0iJf, T(.)52.20

)~2086 T(.0i/2:.

210 1= 8A453 CONFIDENCE INTERVALS FOR THE MEAN U~ 'PPER P~OUND 1 l. 1474 9%LOWER~ YOUI4D = 1 li88 99% UPPER P~OUNtD = '. 1 26 997. LOWER SOUND) = Ii i36 January 20. 1989 1:36 PM Ti-~fl-2s-e4 tjirckA4CS.

OF 4.1fl Catc No 0 setn PROGRAM' OCDWCONF BAY: 17A BOT4 Di 7ABOT4 1 .993 1 1j 1 .011 1.035 i-."06 .968.976 .934 .96` .09 .969 1 .0123.879 .983 .916 .373 .346 .95.2 1.'12.992 .97 .951 .924 .92? .912 .971 MEAN THICKNESS

.95736 STANDAR:D ERROR OF THE MEAN .OQ?1,979 T" .05/2-, 27 )2 2.015s18 T(.01/2. 27 ,= 2.7707 CONFIDENCE INTERVALS FOR THE MEAN'95%T UPF'PER BOUND = .623 95% LOWER BOUND .93848 (1,9% UPPER BOUND = .92284 99% LOWER BOUND .93187 January 20. 1989 S: 36 ('m2 7-H R 1.2-5-C& r-jc-,AcCeS o0 57 Potts h5vts r lwc f% QfeAo

G/2~AE i428~C~Calc. No. C-1302-187-5300-005 Rev. No. 0 Page //Iof 5.3 6"x6" Grids at Upper Elevations 5.3.1 Bay 5 51' Elevation:

11/01/87 to 10/8/88 Three 49-point data sets were available for this bay covering the time period from November 1, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) Except for the first data set, the data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 99% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 750.0+0.02 mils.(6) The corrosion rate + standard error is -4.3_0.03 mpy.(7) One data point was determined to be statistically different from the mean thickness.

The probability of this occurring due to expected random error is less than 1% at each specific time.S 10/2.!/0C

'4:2e:V7 Faic No Re T- N o e PROGRAM: DWCHISQ ENTER NAME OF DATA LIST udin2 ENTER FT NUMBER LIST iAn1t-!(1,49)

ENTER NAME OF DATE LIST dte'i N tJDi 2 I E8702626 2 £8702640 3 E8702650 ENTER NO. OF DESRED DATA i .2.3 BAY i5ELCV 5)E-8702626 E8702640 E8702650 CHISO 25.367 1 .608'?. i733 4 7 7 4 10 3 11 9 15 19 6 7 3 10 10 DATESI MEANTHK 11/01/87 .75385 7/12/88 .75095 I1/08/88 .75019 SD.024144.e086446.'01716 ST DEF: F.0037706.00133339.3026478 DF'M2 2 21 CH 1952 5.99 5.99 5.99 C1H1992 9.21 9.21 9.21 EXP 8.6863 8.8981 8.8981 7.767 7.9565 7.9565 8.0934 8.2908 8.2908 7.767 7.9565 7.9565 8.6863 8.8981 8.8981 fRAND MEAN THICKNESS

= .75!67 STANDARD ERROR OF THE: GRAND MEAN = .001116 January. 18, 1989 J:eo IM 0

'Gf2VU 14:28:C)3 Jux~y s -sl E8702626.765 .735 .764 .735 .754 .772 .?4., 6 .706 .761 .748 .7? .74" .'t..765 .763 .766 .73 .777 ," .767.736 .76i .759 .716 0 C 0.77 .768 .761 .751 ( 0 c.776 '. 3 .75 .73? .766 .75S .743 L638 0 .751 .76 .75 ' .758 .761 E8702640"'!5 .728 .765 .746 .749 .765 743.'756 .746 .754 .745 .761 7.47 .?48.7; .761 .76 .747 .75 0, .753.759 .748 .752 .75 0 1Q.759 "74'.3 .753 .751 $ C 0.768 .762 .744 .7.' .754 .748 .737 S- ..739 .741 .754 .! 5t2 .75 T -S4 E8702650.772 .73 .765 .748 .747 .764 .7%.752 .696 .753 .743 .764 .723 .778.756 .756 .758 .721 .747 ? .745.759 .746 .756 .706 0 0 0.763 .764 .755 .751 0 0 0.769 .75 .742 .73 .752 .'?A6 .734.746 .77 .774 .752 .772 .749 .754 (K 0 LINEAR kCEGRCSION IPLOT FOR DU WALL THINNING ANALYSIS OF SECTION 51&.4 DEP VARIABLE; MEASURE SOURCE DF MODEL I ERROR i C TOTAL 2 ROOT MSE DEP MEAN C.V.ANALYSIS SUm OF SQUARES 8.66615345 0.0.000513212 0.0 0.66666667 0.02265419 751.6667 8.e03013861 OF VARIANCE MEAN SQUARE AM&&5345 00513212 R-SQUARE ADJ R-SQ F VALUE 16886.103 PRob)r 0.0049 0.9999 0.9999 VARIABLE OF INTERCEP f YEAR I PARAMETER ESTIMATE 753.99542-4.27817350 PARAMETER ESTIMATES STANDARD T FOR HO: ERROR PARAMETER=0 0.02219620 33984.979 0.03292257

-129.947 PROB ) ITI 0.0001 0.0049 TYFPE I S$i695008.33 8.66615345 STANDARDIZED ESTIMA1E 0-0.999Y7039 NUMBER EIGENVAI I i.070 2 0.192: PREDICT STI VALUE PRI 754.0 00 750.0 0 COLLINEARITY DIAGNOSTICS CONDITION VAR PROP LUE NUMBER INTERCEP 748 1.eo6ee0 0.0961 252 3.066429 0.9039 0 ERR LOWER95Z UPPER95Z EDICT MEAN MEAN.0222 753.7 754.3.0140 750.8 751.2.0184 749.8 750.2 VAR PROP YEAR 0.09A1 0.9039 LOWER95Z PREDICT 753.6 750.7 749.6 085 I 3 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)ACTUAL 754.0 751.0 750.0 2.27374E-0.00651321 0.014886 UPPER95Z PREDICT 754.4 751 .4 750.4 RESIDUAL.0045809-0.0179 0.0132 13 19 667ý (n-2-) = /.?, '70ep'.S ti!. *, 7 ,6 f .2,n,

./l LINEAR REGRESSION PLOT FOR DW WALL T14IMNING ANALYSIS OF SECTION 51 I DEP VARIABLE:

MEASURE SOURCE DF MODEL I ERROR i C TOTAL 2 ROOT ISE DIEP MEAN C.V.ANALYSIS SUN OF SQUARES 9.66605343 a.0,000513212 0.04 8.66666667 0.02265419 751.6667 0.093013861 OF VARIANCE MEAN SQUARE 66615345 00513212 R-SQUARE%DJ R-SZ F VALUE 16886.103 PROD) F 0.0049 0.9999 0.9999 VARIABLE INTERCEP YEAR Dr 1 PARAMETER ESTIMATE 753.99542-4.27817350 PARAMETER ESTIMATES STANDARD ERROR 0.02218620 0.03292257 T FOR HO: PARAMETER=O 33984.979-129.947 PROD ) ITI 0.00,01 0.0049 ORS 3 2 3 ACTUAL.754.0 751.0 750.0 COol( 1S D PREDICT VALUE 754.0 751.0 750.0 STD ERR PREDICT 0.0222 0.0140 0.0104 LOWFR?5X MEAN 753.?750.9 749.8 UPPER?5%MEAN 754.3 751.2 750.2 LOWER?5Z PREDICT 753.6 750.7 749.6 UPPER95% STD ERR PREDICT RESIDUAL RESIDUAL 754.4 .0045809 .0045809 751.4 -0.0178 0.0$78 750.4 0.0132 0.0132 STUDENT RESIDUAL 1.0000-1.0000 1.0000~2~t-O 1 2 I MI ju 08S I 11.728 2 0.309 3 0.966 SUM OF RESIDUALS 2.27374E-13 SUM OF SQUARED RESIDUALS 0.0005132119 PREDICTED RE$ID SS (PRESS) 9.01488667 id tU w 1L;

~~~1'4 LINEAR REGRESSION PLO0 FOR DU WALL THINNING ANALYSIS OF SECTION 51 a PLOT OF MEASURE*YEAR PLOT OF PREDwYEAR PLOT OF U95*YEAR PLOT OF L95*YEAR SYMBOL USED IS X SYMBOL USED IS P SYMBOL USED IS u SYMBOL USED IS L P R E D C T E 0 V A L U E 1100 4 1000*900 Ix 700 +IX I I 600.0 f, x x P~J.4..!.+ -.. .-.. ... .... ..--..............

-... ... -.. ... ---. -- -.-----+--..----.----..........-......+....4.

... ..--..+..-- -... -..0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 YEAR N0llt 1) tOBS HIDDEN

/0 R D U A L S 6tG 4 ,+-2-4-6 4 48 4-1 I1 LINEAR RGRE$SION PLOT FOR DU UALL THINNING ANALYSIS OF SECTION 51 PLOT or RESIDWYEAR SYMBOL USED IS R R f-4 ...... ...... --...... --. --. ..-- -. .- ---. ... .... ..- ,....... ... ..-........

.... ,. ..... ... ....,-..... .... ....÷ ........ ..-0.1 0.- 0.3 0.4 0.5 0.6 0.? 0.8 0.9 1.0 YEAR f4 tU'V 4-7 z 0 II i/0 LINEAk REGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 51 UNIVARIATE

'I tit 1 VARIABLE-RESID RESIDUALS N MEAN STO DEV SKEWNESS USS CV T:MEAN=8 SGN RANK Plum 'a 0 W:NORMAL STEM LEAF' 3 e 5 0-o-Q-1 MOMENTS 3 SUM WGTS 7.579E-14 SUM 0.0160f19 VARIANCE-1.18162 KURTOSIS.0005t3212 CSS 99999 STD MEAN B.M9SE-12 PROB>ITI 0 PROB)ISI 3 0.938649 PROB(M i i 1 3 2.274E-13.000256606

.08e513212 6.00924853 0.452 t002 MAX 75% 03 50Z NlED 251 Ql RANGE Q3-Ut MODE QUANTILES(DEF24) 0.0132295 99%0.0132295 95%0.00458099 9ez-6.0178104 102-0.0178104 5Z 12 0.9310398 0.0316398-0.0179104 0.01 25.-0.0025+I 0.0l32295 0.0132295 0.0132295-0.0178104

-0.0178104

-0.0178104 EXTREMES LOWEST HIGHEST-0.0178104 0.0045099 0.0132295

-0.0178104 8.0t32295 8OXPLOT I I I,'I I I .. .I NORMAL PROBABILITY PLOT%.++-1 MULIIPLY STEh.LEAF BY iOn*-02-0.0175S++ 1*4-tm-2 -1 0 .1 +2 LW 0 (A 0,_N I /2146 4:26 :07 Ca.c No..vNo IC-1302- le7-5300 -5 5 ..7355 .764 ... .. ... ..Al? -76 7'.3 7.-6 ..71 7 .-. .777 z .76; .Tf'L" .761 .:"r9 7.7"5 +.768 .761 .7.'.fr U 0\ .763 .75 .73' .766 .7513 .743 01 .751 -76 .7.> ... .. .56 E,8 +-'"7 +' Z" .65i .746 .749 7 .7"4.756 ~T7 +.754 -745 .761 .747 .748*759 .761 .76 .747 .75 0 .75.?59 .748 .752 .75 0 0 0.7413 .753 .751 0 0 0.7,8 .762 .744 .731 .1.4 .?48 -737'745 .739 .741 .75:4 .753 .75 .754 E8702650.77n .76" .748 .747 .764 .75..752 ..3 .743 .764 .."723 .778.,56 r7"6 .758 ." .747 0 .745 759 '7446 .75 I 0 0.763 .764 .75:5 ('i 0.76 .75 .'742 .73 .7'2 .746 .73.4.74 .77 .774 .752 .772 .749 .754---S , po,5 9/7 0 1 14 :26 : 07 CaLc. No. C-1302-187-5300-005 Rev. No. 0 Page) Iof 5.3.2 Bay 9 87' Elevation:

11/6/87 to 10/8/88 Three 49-point data sets were available for this bay covering the time period from November 6, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is appropriate than the regression model.(3) There was no significant corrosion from November 6, 1967 to October 8, 1988.(4) The current mean thickness

+ standard error is 620.3+1.0 mils.

" 012 5"c-c '. 4 : , ý : Q lCac Noa '-ev 1 IC0 12 PROGRAM: DWCHISQ ENTER NAME OF DATA L.I:T u2O ENTER PT NUMBER LIST int.v(1,49ý ENTER NAME OF DATE LIST date20 U20 1 D8702630 2 DB702641 3 D8702651 ENTER NO. OF DESIRED DATA 1.2.3 D8702630 D8702641 D8702651 7.4024 2.8986 i .2047 OBS 9 8 10 8 a 7 13 14 12 12 10 10 7 9 10 DAIE20 11/06/87 7/20/88 10/8/88 MEANTHK.61892.6"233.6 i157 SD STLIERR******* ~**.g***.014675 .0020965.014447 .0020639.013885 .0019136 DF M2 2 *:H1952 1.99 SH'.191 2 9.2-9.21 9.'21 i-* v 10.381 10.381 10.381 9.2826 9.2826 9.2826 9.6726 9.6726 9.6726 9.2826 ?.2826 9.2826 10.381 10.381 10.381 GRAND MEAN THICKNESS

= .62027 STANDARD ERR'OR OF THE GRAND MEAN = .0010444 JanturK.18:

1989 0 1 ;:: 2510;6 4 :2F:C', No

• Ae, I No She., D871,2630.,8 .,604 .6 .601 .34 O.N0, .615 .618 .617 .6-1 ..,5- .639.618 .614 .615 .6"8 .628 .604 .631.6:6 .604 .62 .5.65 .627 .626 .623.624 .607 .666 .641 .61S .3641 .61.624 .618 .617 .622 .616 .6",29 .641.608 .60? .5?3 .598 .622 626 .611 1)8702.)64

'.633 .625 .625 .S2"' .62', .601 .. 6 )`: .6'14 .619 .617 .6313 .638 .638.612 .6218 .615 .623 .628 .627 .622 56Y- .66 .628 .585 .63z .627 .619 i'.603 .647 .6 39 (37 .99"2 .617 662 "638 .614 62-5 637 15 .60)3 5 92 597 1522 6 431.62)D 37 026 f1* ~ ~ ~~~~I It******44441

• • **ftt 4** * *4*** I t 44444.6'6 .629 .629 .60"7 .633 .601 .634.606 .616 .618 .6i7 .623 .587 .63?.60? .&2 .619 .626 .627 .61 .6.23 62 .6 .623 .5864 .6T .."37 .624 t626 .iS5 .644 .64 .61? .635 *6.61"7 .617 .624 .615 .628 .639 14 .608 .593 .598 .622. .634 .616 0 0 LINEAR REGRESSION PLOT Fnk DIJ WALL ININNING ANALYSIS OF SECTION 20 DEP VARIABLE:

MEASURE SOURCE OF MODEL i ERROR I C TOTAL 2 ROOT MSE DEP MEAN C.V.ANALYSIS OF VARIANCE SUM OF MEAN SQUARES SQUARE 0.50567679 0.50567679 5.49432321 5.49432321 6.00060000 2.34399? R-SQUARE 620 ADJ R-SQ 0.3780641 F VALUE 0.092 0.9843-8.8314 PROW)F 0.8125 VARIABLE OF INTERCEP I YEAR i PARAMETER ESTIMATE 619.43455 1.04263.156 PARAMETER ESTIMATES STANDARD T FOR WO.ERROR PARAMETER-O 2.30336565 268.926 3.43679113 0.303 PROD ) I11 0.0024 0.9125 TYPE I TS 1153200.00 0.50567679 STANDARDIZED ESTIMATE 0 0.29030926 09S 2 3 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)ACTUAL 619.0 622.0 619.0 2.27374E-5.4943 1B4.33 NUMBER EIGENVAL 1 1.8091 2 0.190B PREDICT STD VALUE PRE 619.4 2.620.2 1.620.4 1."13 123 369 COLLINEARITY DIAGNOSTICS CONDITION VAR PROP tUE NUIMER INTERCEP 199 1.009000 0.0954 aet 3.079305 0.9046 ERR LOVIER93Z UPPER93Z DICT MEAN MEAN 3034 590.2 649.7 4629 601.6 638.8 8923 596.5 644.3 VAR PROP YEAR 0.0954 0.9046 LOWER95Z PREDICT 577.7 585.1 582.2 UPPER95Z PREDICT 661.2 655.3 658.6 RESIDUAL-0.4345 1.8314-1 .3969 4A K.t LINEAR REGRESSION PLOT FOR DW WALL THINNIIG ANALYSIS OF SECTION 20 DEP VARIABLE:

MEASUtE SOURCE Df MODEL I ERROR I C TOTAL 2 ROOT MSE VIP MEAN C.V.ANALYJ SUM OF SQUARE$0.50567679 5.49432321 6.00900000 2.343997 620 0.3780641 ,IS OF VARIANCE MEAN SQUARE 0.50567679 5.49432321 R-SQUARE ADJ R-SQ.3~qJ 4.)F VALUE 0.e92 0.0843-0.8314 PROB)F 0.8125 VARIABLE INTERCEP YEAR DF I PARAMETER ESTIMATES PARAMETER STANDARD ESTIMATE ERROR 619.43415

' 2.30336565 1.04263256 3.43678113 T FOR HO;PARAMETER-O 268.926 I.303 PROD ) II1 0.0024 0.8125 SID ERR RESIDUAL 0.4345 1.9314 1.3969 Os 2 3 ACTUAL 619.8 622.0 619.0 COOX' $PREDICT VALUE 619.4 620.2 A26.4 S1D ERR PREDICT 2.3034 1.4629 1.8923 LOUER95Z MEAN 590.2 601.6 596.5 UPPER95Z MEAN 648.7 638.0 644.3 LOWER952 PREDICT 577.'7 595.1 592.2 UPPER95Z PREDICT 661.2 655.3 658.6 RESIDUAL-0.4345 1.8314-1.3969 STUDENT RESIDUAL-1.0000 t.0000-1 .0000-2-1-0 t 2 I I I I*I '1 OBS 1 14.048 2 8.319 3 0.9"s SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)2.27374E-13 5.494323 184.3369 clot (A 0 5 LINEAR REGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 20 I-6 9 PLOT PLOT PLOT PLOT OF OF OF OF MEASURE*YEAR PRED*YEAR U95*YEAR L95*YEAR SYMBOL SYMBOL SYMBOL SYMBOL USED IS x USED IS P USED IS U USED IS L ttee 900 800 P I E D I C E D V A L U£ 700 +~~j4~0I I -I I'U Ix 600 +I IL 500+U x P L U L IA IA -. -- ...... -- ---,- ----- --- --- 4-- --......

• 4--- ------.-0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.4 YEAR NOTE 2 UPS HIDDEN

/1.LINEAR REGRESSION PLOT FOR DU WALL THINNING ANALYSIS OF SECTION 20 (1 I PLOT OF RESIDVYEAR SYMBOL USED IS R E S I 0 U A L S 10 +2 +0-2 +-4 +-*-1 I S.ft hr4 4"I tA f..-. ....--.---..

...- --.........-

.........

+ -- -- -.-.-.-.4- -- +- --.... -..--4--..-----..

-........ 4--0.00 0.06 0.12 0.18 0.24 0.30 0.36 0.42 0.48 0.54 0.60 0.66 0.72 0.78 0.64 YEAR

(°K\LINEAR REGRESSION PLOT FOR DU WALL THINNING ANALYSIS OF SECTION 20 UNIVARIATE I I VARIABLE-RESID RESIDUALS N MEAN STD DEV SKEW4NESS USs CV T:IIEAN-6 SGN RANK W: NORIMAL STEM LEAF 18B t MOMENTS 3 SUN IVGTS 7.579E-14 SUM 1.65746 VARIANCE 1.0987 KURTOSIS 5.49432 CSS 99999 STD "EAN 7.920E-14 PROB)ITI 0 PROB)ISI 3 0.948429 PROM(U 3 2.274E-13 2.74716 5.49432 0.956933 I I 0.475 0o8z MAX 752 03 50% MED 25Z Qi 0% MIN RANGE Q3-Qt mODE QUANTILES(DEF-4) 1.83144 99z 1.83144 95%-0.434546 902-1.3969 10o-1.3969 5z 12 3.22834 3.22834-1.3969 1.75+0 1 1 .83144 t.fl3144 1.83144-1.3969-1.3969-1.3969 EXTREMES LOWEST HIGHEST-1.3969-0.434546 1,83144 -1.3969-0.434546 1.83144 0 0-e 4-0-1 4.... ... ÷ .. .÷ ... --ROXPLOT I -I I I I.) -- -4 NORMAL PROBABILITY FLOT+4,+1 1 644$-2 -t 0 *1 +2'1i IA-z I 0125106E .'4 : 28: 0 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page )2of 5.3.3 Bay 13 87' Elevation:

11/10/87 to 10/8/88 Three 49-point data sets were available for this bay covering the time period from November 10, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) There was no significant corrosion from November 10, 1987 to October 8, 1988.(4) The current mean thickness

+ standard error is 635.6÷0.7 mils.0 Calc No Re. No I Sheet No ,C 130ý2- le7- 530 -0 PROGRAM: DWCHISQ ENTER NAME OF DATA LIST u28 ENTER FT NUMBER LIST int7(i,49)

ENTER NAME OF DATE LIST date'28 660, 13 JE LC-v rc?, N U28 D8702637 1)8702652 ENTER NO. OF DESIRED DATA 1.2.'1 D8"0..2637 D8702642 8 70 265.CHISf (1. 24.r1, 7.7853477 OEI$9 7 8 8 9 7 9 6t12 12 16 I "o .II 9 10 DATE2B!i/10/87 7/20/88 10/8/93 MEANTHK.637.63453.63 53.3 3D.01 9096.0 95 A.018936 STDEPR.002"728..0028532.002 ". 05 4 DFr2 2 2:H19!52 5.99 5.9?5.99 CH1992 9.21 9.21?. 21 EXF'10.381 9.9574 10.381 9.2826 8.9037 9.2826 9.6726 9.2778 9.6726 9.2826 8.9037 9.2826 10.381 9.9574 10.381 GRAND MEAN THICKNESS

= .63562 STANDARD ERROR OF THE GRAND MEAN = 7.2`133E-4 January 18. 1989 1: 05 FM t

!Caic N) Rev No 5S,-.iNa ac I3o V-/ 7-34-S -7~fh/De702637.602 .613 .645 .643 .643 .633..639 .652 .653 6 642 ....623 .64 .649 .621 .646 .651 .-65.602 .63"7 -3 .575 61 .".629 .627 .652 .629 649 .55 .66.59 .639 .638 .662 .651 .64i .661.641 .639 .628 .66 .653 ..38 .6"5 D 8-70 26 4 2.5,6 .608 .643 .64 .643 .645 .627.626 .638 .t48 .648 0 ( .621.624 .641 .647 .617 .646 .649 .652.596 .635 .633 .572 .602 .65 .6/i.62? .624 .647 .626 646 .652 .6 5.5S? .643 .64 .657 .654 ..S36 .652.636 .632 .t2'i .6.53 ..b .62"? .646 D8702652 9t" ,61 i .64? .644? 646 .65e-,, -.623.,32 .643 .65 .651 .63 .64 .626.628 .645 .651 .62 .646 .649 .656.6 .636 .638 .579 .01 .657 .654.631 , 626 .642 .627 .649 .657 .659.M85. :659 .639 .657 .642 .639 .652.638 .633 .628 .637 .634 .612 .648

-Kf 0 I LINEAR REGRESSION PLOT FOR OW WALL THINNING ANALYSIS OF SECTION 29 4 9 C, i DEP VARIABLE:

MEASURE SOURCE OF MODEL I ERROR I C TOTAL 2 ROOT ISE DEP MEAN C.V.ANALYS SUM OF SQUARE$3.32654781 1.34011886 4.6666666?

1.157635 635.3333 6.1822egi;IS OF VARIANCE MEAN SQUARE 3.32654781 1.34011886 R-SQUARE ADJ R-SQ F e F VALUE 2.4R2 0.7128 0.4257 PROB)F 0.3600 VARIABLE OF INTERCEP I YEAR I PARAMETER ESTIMATE 636.78260-2.70991515 PARAMETER ESTIMATES STANDARD T FOR HO: ERR9q PARAMETER-0 1.13703086 560.036 1.71937336

-1.576 PROD ) ITI 0.0011 0.3600 TYPE I SS 1210945.33 3.32654781 STANDARDIZED ESTIMATE 0-e.84429359 NUMBER EIGENVAL 1 1.5094 2 6.191'PREDICT STI VALUE PRi 636.8 1 634.9 0 634.3 0 COLLINEARIIY DIAGNOSTICS CONDITION VAR PROP LUE NUMBER INTERCEP 0OO 1.900000 0.0955 see 3.077534 0.9045 D ERR LOWER95X UPPER95%EDICT MEAN MEAN.1370 622.3 651.2.7215 625.7 644.1.9311 622.5 646.1 VAR PROP YEAR 0.0955 0.9045 LOWER95%PREDICT 616.2 617.6 615.4 1 2 3 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID SS (PRESS)ACTUAL 637.0 634.0 635.0 I.13687E-1.3401 43.981 UPPER952 PREDICT 657.4 652.2 653.2 RESIDUAL 0.2174-0.9053 0.6879-13 19 171 2,,76 (.r)..4 (A I L.INEAR REGRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 28 I DEP VARIABLE:

MEASURE SOURCE DF MODEL 1 ERROR I C TOTAL 2 ROOT MSE DEP MEAN C.V.AHALYS SUM Of SQUARES 3.32654781 i.340ttBB6 4.66666667 1.157635 635.3333 0.1822091'IS OF VARIANCE MEAN SQUARE 3.32654781 1.34011886 R-SQUARE ADJ R-SQ F VALUE 2.482 0.7129 0.4257 PROB)F 0.3600 VARIABLE DF INTERCEP I YEAR I PARAMETER ESTIMATE 636.7T260-2.70991515 PARAMETER ESTIMATES STANDARD ERROR t.13703896 1.71937330 T FOR HO: PARAMETER=0 560.036-1.576 PROD ) ITI 0.0011 0.3600 STD ERR RESIDUAL 0.2174 0.9053 0.6879 OB$t 2 3 ACTUAL 637.0 634.0 635.0 COOKDIS 0 OBS PREDICT STD E VALUE PREb)636.8 l.i1 634.9 0.7;634.3 6.92 1.13687E-t3 1.340119$) 43.99571:RR rCT tI15 tll LOIIER95Z MEAN 622.3 625.7 622.3 UPPER93X MEAN 651.2 644.1 646.1 LOWER95Z PREDICT 616.2 617.6 615.4 UPPER95Z PREDICI 657.4 652.2 653.2 RESIDUAL 0.2174-0.9053 0.6879 siUDENT RESIDUAL 1. 0000-1.0000 1.0000-2-1-0 1 2 I '1 I I'1 13.679 2 0.318 3 0.916 SUM OF RESIDUALS SUtM OF SQUARED RESIDUALS PREDICTED PESID SS (PRES$'z Vt IA I LINEAR REGRESSION PLOT FOR DW UALL THINNING ANALYSIS OF SECTION 28 I 0 Ii PLOT OF "EASURENYEAR PLOT OF PREDOYEAR PLOT OF U95"YEAR PLOT OF L930YEAR SYMBOL USED IS X SYMBOL USED IS P SYMBOL USED IS U SYMBOL USED 1 L p R E D I c T E V A L U E 9100 +I 1000 IuU Ix Ix IL L 600 900..4.....-*.-----------------

-- ----------

---

.

~0. QO 0.06 0. 12 0.18s 0.24 0.30 0-36 0.42 0.48 0.54 0.60 0.66 (1.12 0.10 0.84 YEAR 0 a4)I 0.NATF! 3 CBS HIDDEN

-a 0 0'LINEAR REGRESSION PLOT FOR DU UALL T7INNING ANALYSIS OF SECTION 20 PLOT OF RESIDUYEAR SYMBOL USED IS R k E S I U A L S 10 4,t-6 *-I-1 I 9 4 u ,..$, 91 0 R R IM Lii 0.1U 0.24 O.JO 0.36 0.42 0.48 0.54 0.60 0.66 0.7 o 0.78 0.4 0.1 YEAR V 0 LINEAR REGRESSION PLOT roR OW WALL THINNING ANALYSIS OF SECTION 28 UNJVARIATE 4~s-I VARIABLEwRESID RESIDUALS'OMEMTS QUANTILES(DEF=4)

EXIRE"ES N MEAN STO DEV SKEWNESS U$$CV T :MEAN=G SGW RANK wjul -" 0 W:;HORMAL TTEM LEAF 07-0-o 9 3 3.790E-14 0.815"72-1.11882 1.34012 9999" 8.OI:-14 0 3 0.947082 S WGT VARIANCE KURTOSIS 1TO MEAN PROD') I T I PROD') I1I PROD(W 3 1 .137E-t3 8.670059 1 .34612 0.472663 0.472 100z MAX 735 03 502 MED 25X as 0% MIN RANGE: 03-Ri MODE 0.687928 0.897929 0.217397-0.905325-0.9'5325 t.59325 1.59325-0.965325 99X 9Y5 90%19%stT 55 0.697928 0.687928 0.69792;-0.905325-0.9"5325-0.905325 LOWEST-0.905325 0.217397 0.607929 HIGIIEST-0.905325 0.217397 0.607928 t 1!'DXPLOI S-0--s--p 1 I S-S-0.75+NORMAL PRORAhILITY PLOT+$0+4**4++4+++..... + ... + ... + -....-2 -1 0 +1 +2 (A ma~-1i IzI Calc. No. C-1302-187-5300-005 Rev. No. 0 Page I3ýf 5.3.4 Bay 15 87' Elevation:

11/10/87 to 1018/88 Three 49-point data sets were available for this bay covering the time period from November 10, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) There was no significant corrosion from November 10, 1987 to October 8, 1988.(4) The current mean thickness

+ standard error is 634.8+0.7 mils.

I lý12!,1iU, :24:j le = e'. No 5netNo 1 PROGRAM: DWCHISQ ENTER NAME OF DATA LIST u31 FNTE:R PT NUMBER LIST i ntc(i.49)ENTER NAME OF DATE LtXT c:hte3l N U31 if ~*****~**1 D8702638 D P8702643 3 D8702653 ENTER NO. OF DESIRFED DATA 1.2,3* %t* i* ***D 8702638 D8702643 D8702653 C Lij:IS.82927 1.5885 2.58.36 OBS 8 9 8 9 7 7 I1 10 13 10 12 It 11 11 1 ()DATE31 11 0/87 7/20/88 10/8/88 CH1952 Ci 5.99 5.99 MEANTHK.6361.63402.63422 SD STDERR.(;17368 .0024811.01673 .00239.Gi6704 .0024149 WFM2 2 2-11992 9.21 7.21 9.2i EXP 10.381 10.381 10.381 9.".8,6 9. 2826 9.2"826 9.6726 9.6726 9.6726 9.2826 9.2826 9.2826 1(.331 10.381 10.381)aRAND MEAN THICKNESS

= .63478 STANDARD ERROR OF' THE GRAND MEAN = 6.6249E-.4 January i, 1989 1:07 (M 1Of~~~ it/C ::C NoR/D8702638.655 .648 .639 ..62 6 .641 ,559 .643 ,646 ,64 .634 .65i .641.622 .657 .673 .639 .654 .62i .632.65 .652 .646 .638 .61? .433 .634.656 .633 .637 .623 .634 .568 .62.65 .63 .625 .607 .6215 .606 .614.64? .648 .615 .649 .628 .628 .647 D070"'.1643

.65i .645 .633 .643 .&65 .S26 .634.651 .642 .643 .641 .651 .644 .638.627 .654 .654 .633 .65 .6S2 .634.644 .6w2 .654 .635 .616 .634 .632.652 .63 .64 .622 .635 .566 .623.645 .627 .619 .6C4 .624 .6f5 .617.648 .646 .6i3 .639 .622 .610 .543 1"8702653.651 .645 .632 .642 .618 .622 .636.655 .641 .644 .638 .63 .643 .637.629 .654 .645 .635 .649 .649 .643.65i .65 .619 .636 .616 .632 .636.664 .63 .635 .619 .634 .562 .626.65 .646 .622 .605 .63 .608 .622.654 .645 .6T2 .642 .628 .622 .643 K LINEAR REGRESSION PLOT FOR DU WALL THINNING ANALYSIS Of SECTION 31 4, DEP VARIABLE:

MEASURE SOURCE DF MODEL I ERROR I C TOTAL 2 ROOT MSE PEP MEAN C.v.ANALYS SUm Of SQUARES 2.52560013 0.14106654 2.66666667 0.3?%883 634.6667 0.05917882 3IS OF VARIANCE MEAN SQUARE 2.52560013 0.14te6654 R-SQUARE Abi R-SQ F VALUE 17.904 0.944 PROI)F 0. 1477 VARIABLE OF INTERCEP t YEAR t PARAMETER ESTIMATE 635.92947-2.36037375 PARAMETER ESTIMATES STANDARD T FOR HO: ERROR PARAMETER-e 0.36890594 1723.825 0.557a4lil

-4.231 PROD ) ITI 0.0004 0.1477 TYPE I SS 1208405.33 2.52560013 STANDARDIZED ESTIMATE 0-0.97319065

.-..NUMBER EIGENVAL 1 1.8099 2 0.1911 PREDICT ST$VALUE PRI 635.9 0 634.3 0 633.8 0.COLLINEARITY DIAGNOSTICS CONDITION VAR PROP VAR PROP liE NUMBER INTERCEP YEAR wee 1.000000 0.0955 0.0955 G0ft 3.077534 0.9045 0.9045 D ERR UPPER95Z LOUER95Z EDICT MEAN MEAN PREDICT.3689 631.2 640.6 629.2.7341 631.3 637.3 628.7.3621 629.9 637.6 627.7 CBS 1 2 3 SUM OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICTED RESID 55 (PRESS)ACTUAL 636.0 634.0 634.0 2.27374E 0.1410 4.63 UPPERY5%PREDICT 642.6 639.9 639.9 RESIDUAL 0.0705-0.2937 0.223Z-13 665 012 633.8 0 *iu IA LI

-C LINEAR REGRESSION PLOT FOR DW UALL THINNING ANALYSIS OF SECTION 31 A 0 0 0 DEP VARIABLE:

MEASURE SOURCE OF MODEL 1 ERROR I C TOTAL 2 ROOT MSE DEP MEAN C.V.ANALY.SUM OF SQUARES 2.52560013 0.14106654 2.66666667 0.3755883 634.6667 0.05917882 HIS OF VARIANCE MEAN SQUARE 2.52560013 0.14106654 R-SQUARE ADJ R-SQ F VALUE 17.9e4 0.9471 0.8942 PRob)F 0. 1477 VARIABLE INTERCEP YEAR OF t I PARAMETER ESTIMATES PARAMETER STANDARD ESTIMATE ERROR 635.92947 0.36890594

-2.36037395 0.55784M11 T FOR HO;PARAtETERao 1723.925-4.231 PRO? ) I1I 0.0004 0. 1477 DBS 1 2 3 ACTUAL 636.0 634.0 634.0 COOKI S D PREDICT VALUE 635.Y 634.3 633.9 STD ERR PREDICT 0.3689 0.2341 0.3021 LOWDER95%HEAN 631.2 631.3 629.9 UPPER951 MEAN 640.6 637.3 637.6 LOWER95Z PREDICT 629.2 628.7 627.7 UPPER95%PREDICT 642.6 639.9 639.9 RESIDUAL 0.0705-0.2937 0.2232 srD ERR RESIDUAL 0.0705 0.2937 0.2232 STUDENT RESIDUAL 1.0000-i.0000 1.0000-2-1-0 1 2 I I*I Wj1 OBS 1 13.678 2 0.318 3 0.916 SUt OF RESIDUALS SUM OF SQUARED RESIDUALS PREDICIED RESID SS (PRESS)Li Iz 0.2.27374E-i3

0. 1410665 4.63012 0 'K 4 LIHEAR REGRESSION PL.OT FOR DW VALL THINNING AHALYSIS OF SECTION 31 Pl01 OF MEASUREOYEAR PLOF Or PREDWYEAR PLOT OF U95?YEAR PLOT OF L95%YEAR SYMBOL. USED IS X SYMBOL USED IS P SYMBOL USF0 tS 0 SYMDOL USDO IS L P R E-D C v V A L U r 1100 ÷1000 900 +800 *700 4 Ix IL I 600 +I 500:', ti L~A U x LA.-.. .. 4 ... .-- 4.. -..* ..+ ... .,...,-..-.

-.. +- -- p--- -..------ + --..-- --+..-..- .+..-...--+.

..- .--k. fit) Q.(,' 0.06 0.18 0.24 0.30 0.36 0.42 0.413 0.54 0.60 0.66 0.72 0.7?8 0.84 0.9(YEAR HOLE, 6 CBS 14IDDEN I a 6 LINEAR REGRESSION PLOT FOk DU WALL THINNING ANALYSIS OF SECTION 31 UNWIVARIATE VARJABLEORESID RESIDUALS N HEAN STO DEV SKEWNESS USS CV T : EANBG SGN RANK MUM 'a a. STEM LEAF 2 2-1 0 7-0 MONENTS 3 SUM WGTS 7.579E-14 SUM 0.265581 VARIANCE-1.11082 KURTOSIS 0.141067 CS 99999 STO MEAN 4.943E-i3 PROC)ITI 0 PROB)ISI 3 0.147082 PROB(W I QUANT!LES(DEF'&4) 3 2.274E-13 0.0705333 0.i41e67 0.153333 0 9 .472 iooz MAX 75Z Q3 502 NEp 25Z Q1 o0 hIN 0.223194 0.223194 0.0705332-0.293728-0.293728 992 95%902 10%52 i%RANGE 0.516922 Q3"-Qi 0.516922 MODE -0.293728 0.223194 LOWEST 0.223194 -0.293728 0.223194 0.0705333-0.293728 0.223194-0.29372B-0.293729 NORMAL PROBABILITY PLOT+4@4 EXIREMES HICHE ST-0.29372;0.07e5333 0.223494 DOXPLOT I I I I* -t 0.25.-9.254.4,-9T NtILTIPLY STEI.LEAF DY 10e0-0.25*" 4+w+----- -- --- .-----2 -1 0 41 +--4 0 fINEAR kELRESSION PLOT FOR DW WALL THINNING ANALYSIS OF SECTION 31 0 In 19 PLOT OF RESlDOYEAR SYMBOL USED IS R R E I D U A L S 10 4-2-4 ++-6 4-8I-I I'...6.. ..1..."....

.. ..... 4 ....0... .... ---- .. 0... ... 4 -.6-0 0.. .. .. 8 0-- .. 4 o.06 0.111 0.2,' 0.30 0.A6 0.42 0.48 0.54 0.60 0.66 0.72 0.76 0.84 0 YEAR Calc. No. C-2302-187-5300-005 Rev. No. 0 PagepOf 5.4 multiple 6"x6" Grids in Trench 5.4.1 Bay 17D Trench: 12/9/86 to 22/23/88 The two sets of measurements in the Bay 17D Trench were taken on December 9, 1986 and December 23, 1988. The 1986 data is a 7 column by 36 row array. The 1988 data is a 7 column by 42 row array. The 1986 data is at the same elevation as the lower 36 rows of the 1988 data, but is centered about 3-/12 inches to the left of the 1988 data.To compare these two data sets, the 1986 data set and the lower 36 rows of the 1988 data set were each subdivided into six 7 column by 6 row subsets. Each pair of subsets was compared as described in paragraphs 2.5 and 2.8.3.Fourth Subset From The Top: The chi-squared statistic for the fourth subset from the top from the 1986 data set slightly exceeded the critical value for level of significance of 0.05, but was within the critical value for level of significance of 0.01. Also, the F statistic exceeded the critical value for levels of significance of 0.05 and 0.01. Therefore, it is inappropriate to apply the two-tailed t-test based on equal variances.

However, the approximate t-test based on unequal variances can be applied. From the results of this test, it is concluded that the difference between the mean thicknesses is not significant.

This implies that corrosion at this location was not significant.

All Other Subsets: (1) The data are normally distributed.

(2) The variances are equal.(3) Comparison of the means using the two-tailed t-test is appropriate.

(4) The difference between the means of the subsets was not significant.

This implies that there was no significant corrosion in the period from December 9, 1986 to December 23, 1988.(5) The current mean thickness

+ standard error of the top subset is 981.2 +6.7 mils. This is the thinnest area in the trench.S iRev No s-t No c.1302-/e7-S5360~

sZ 0 4t-itabulLate(d17d6l2t:nosplit)

DI7D612T.93" SUM- .943" .958 .927 .889 .91 1,1.014 -.~ .8 987 .973 .939 .956.99V 1.005 .951 .968 .939 .945 .956.995 .995-1.038 1.031 .992 1.003 1.011 I. 25 1.011 .96Er 1-024 1.004 1.002 1.055 p pq 714 I .0i7 1 036 1 .0219 1 .03 1 1.084 1 -1 M..45 1.009 1.024 1.026 1 GOB.991. 1.012 1.041 1.031 1.017 1.076 1.076 1.031 1.101 1.081 1.077 1.04 1.076 1.072 1,087 1.059 1.069 1.05? 1.102 1.088 1.047i.4 1.019 .98' 1.024 1.0t 1.014.961- .I i t1.083 1.011 1.047 1.016 1.028 1.063 1.012 1.029 1.047 1.056 .972 .907 1.021 1.097 1.071 1.068 1.033 .91t .952 1,0U4 3.921 i.00 1.9063 .055 I-= 5 .990 2;I .l2ý 1.0.57 1.044 1 .078 ItQ:5 1.054 1.OtO 1.037 1.015 1.026 1.064 1.07 1.056 1.044 1.065 1.059 1.026 1.058 1.047 1.067 1.095 I. .)8 i.046 1. W9 1.103 .9931 1.86 1.041 1.056 1.045 .995 1.044 1.042 1_026 1.116 ,V6 I.V() 1.,044 .ifI~i.069 .9,65e VBE8 1.122 1.034 1-032 1.07 1.097 1.02e 1.051 RI? 1.059 1.015 1.005 1.135 1.022 1.076 1.0e .957 .981, 1.0213 1.0'23 1.049 .987 1.085 1.048 1.072 .99.0S-i74 i .'I I .- 7 .9 5 f? 1 -04 4 .992 1! 5 t46 1 -7 1.053 1.0 1.025 ,98 t.0311.5 ?;1.005 1.049 1.006 1.058 1.058 1.01l.97 98 1.012 1,009 1.067 1.017 .97..?2 .97~ 974X .9 6 .9 S.;54r1 .061 .94.2 .984t, .64' :99. t.004_journaL off 0 WrJ Nuclear Calculation Sheet Suboject Cac No C\30 -J t'Revwevec by Dale *15qy/7P7&C'

' tP-9G rT -VJ OFLA Z>04ETA _____________~~ri 12-23-&a I?- zI- -Jzr aFof 4tz L- 64 T'* -G5Qn .J-uD, I&Ti"Ie 12 2- 3 -g a -ATr A w ?% ~ voEr.~ #4- 1 3G-rtte Ztjc N ID.r F(~vrs m urn- r--se tz 4e o F~o r Thre ,z G-T c3 m A~s jwr 9 )iTt A1~2 TAISOLATED C~t-, ~-I'r.T5 911 rj is, 1.).2) 4 2',V 1)19I'A LST1NC%5 Gt ZrtrS 84 OF 2 ----U-5* r.y GT P4ý,WS AN D CeOLUWP3.1 N\ X-4 UP 5 0 Foa Trt* L-oe-iTtorQ-oF PL cw S-~ 12.'-5-84 DATAI 15 Ak 7 CQLU.WNC A U. IZOW 4PP(A,/-THC i2--ZS-6 DATfA IS A 7 co.urmiI 4 4*2- Row ARReA/y H-Fti- iZ-q-6,6 DFt\A is AT 1't-k ZPAM ELevAT7o*

A-' 7i-i 56*c~iV -!G. <t cF -%TAs 12.- Z 3-8 8~A'4 0016 io 6 8 6 S) 10,12 14 : 2F. :T" M~Nuclear Calculation Sheet SvtbleCt 0d~ NO R No Sheel Org~niinr

~ ii~v~wedt by a i;ý,Ay 17 D' -r42~c(7tht r2-9G ViTIA WcA5 CEWTQeRED AT 5S./z" fRPcvrA-n~f (0V l(21 &HC-tt,WftJ T)HC LOCA~TION.

Or- PLQC%1#'OM-/f & ~ /k G2 -2~ CA~dn 9, S~ 1//5 Cr-" eP-G0 I l S~r / sJ,- 94ýPAl ;P' c5.ATvs- 711. 12~u Jd 22w7-, Is Foor,)D 4&,cr /y lwH e 8-6d9v-4ejT TH'V5 /;r /.T C&NT&264Z 58' Fý44I& 7C LZ.00 279HC i-f/CD.T,4eE 7-eelor.2-64T .49At.e T 17?'kq (4Al- 7sM5 4n- b.p9#9 .'rv )P-ur- ,Gc~2,FFCI I'dAJ7% 0016 (0646) 101251;0,, 14:26:1ý1[o Nuclear Calculation Sheet 7 4) -rq6ýA.CH~se 7~$/Ep~ va~a'I li /Mo 7 COLLc.-Vw A 6o~eO&i/iDaDP,6V2r (AlOr USED)., V7 Dc u i2.r CNL.C'1 N1,WI aLD2./Vt-4I z.L$D 3 6L1_P 4-At .W3 f vew 4-AIS M6 / e* ,,0!6 (06-86) d Nuclear Calculation Sheet iSub,,,c" Rev N .h, No Date Rev~ewe(S Oy Ote3t£ ,j_,',j. 0,, ;/o,,/ 72$ Az.. (-. A40 7,t,,,e, 13 , /7 / ,7,p Z ,oeoV .AO ,,(L),og,,v 7 77vo -Y7 '/E" 7E:4 T'xrS. ,,t .z 0,,

S -.F_- 44* ./ ,,9 ~ ?ZAS,17 ,,V , AO 7744 F W A'D 2-11.5 7=7-sr- AV.Eec 4~7I~VS///~77tpE-i SR,6-&q-Z 1/e6~i60e0 ZI,16w 5 4A/Lb 61-'- 7--1-6~e A~% 006 (06-86)

C 'Rev No Sheol NO i~ ~ Catt sNO7-thkiaeanc- I 7 sds-E TOWMEAN ýAe 9.? POMPiUENT ARR:AY.)9 2;'.? 6 2 .9Q8i21 r ., 1. 74o tdI nip. wI t Pi r.- j 'an dd v ( tht: ýi 1. L'.tderi-standere r tr ( h k a NI u t95=abs1( 7pd-ob i nvert'( 1 .6.'e..an,:thk kI I J t.9hceanrte(-oanihk--td'51.dnewr-ta. t t-3 a.t"e meanlthk ., ,.- t? 9.fne:1.7.m uy.f.imean rutMr4N 'rK2IDER LC1MEP': U?'REAN re-4 4 1 0 t 4"702. 1 .I2 0 6 '9467 1 .442 ejdvtitde(tI~L

'4 :2*f,,. : r![ Nuclear Calculation Sheet Subtect Calc No Re.. No Snee- No~jf.-to2-J-53605cDWS 19 6 -O~gnDr ate lev,ewved by -,Ate 199- -.6 *~' IE ____/A___

_____ ____Di 7D6l 2T.93 .932 .943 .958 .927 .889 .913 I.!D14 .953 .984 .987 .973 .939 .95,-.991 1.005 .951 .968 .939 .945 .956.995 .995 1.038 1.031 .992 1.003 1.011 1.025 1.011 .968 1.024 1.004 1.002 1.055 1.017 1.036 1.029 1.031 1.084 1.026 1.05 1.041 1.055 1.044 1.047 1.043 0 0 1.045 1.009 1.024 1.026 1.008 1.07 1.07.991 1.012 1.041 1.031 1.017 1.076 1.076 1.031 1.101 1.081 1.077 1.04 1.076 1.072 1.087 1.059 1.069 1.057 1.102 1.088 1.047.998 1.065 1.048 1.004 1.014 1.016 1.016.964 1.019 .987 1.055 1.045 1.022 1.061.906 1.04 1.01 .98 1.024 1.01 1.014.964 1.105 108 1t 01i 1.047 1.016 1.028 1.063 1.012 1.029 1.047 1.056 .972 .907 1.021 1.097 1.071 1.068 1.033 .911 .952 1.066 1.023 1.006 1.063 1.045 1.035 .992 1.052 1.037 1.044 1.078 1.05 1.054 1.051 1.037 1.015 1.026 1.064 1.07 1.056 1.044 1.065 1.059 1.026 1.058 1.047 1.067 1.095 1.088 1.046 1.019 1.103 .993 1.086 1.041 1.056 1.045 .995 1.044 1 04i 1.026 1.116 1.102 1.001 1.044 1.082 1.020 1 1.08 1.106 1.05 1.002 1.017 1.042 1.034 1.037 1.069 .965 .9e8 1.122 1.034 1.032 1.07 1.097 1.028 1.051 .951 1.059 1.015 1.005 1.135 1.022 1.076 1.058 .952 .981 1.023 1.023 1.049 .987 1.085 1.04e 1.072 .98 1.1 1.017 .958 1.044 .991 J.056 1.074 1.053 1.03 1.025 .987 1.031 1.059 1.087 1.005 1.049 1.006 1.058 1.058 1.011 .992.972 .985 1.012 1.009 1.067 1.017 .975.985 .979 .974 .961 1.0t7 1.008 .982.999 .987 1.021 .958 .954 1 e$4 .942.923 .981 .976 .97 .964 :9 1.004:_oLdi = dt7d6i*2t(ints(l,6),)

_oLd2 = d17d612t(ints(7,12),)

_oLd3 = d17d6i2t(jnfs(I3,i8),)

_oLd4 = d17d6i2t(intg(19,24),)

_oLd5 = d17d612t(ints(25,30),)

_oLd6 = d17d612t(ints(31,36),)

N 001i 0 JNuclear Calculation Sheet ,.C-Juoi-ft-53C.-c OLD¶.93 .932 .943 .958 .927 .889 .913 1.014 .953 .984 .987 .973 .939 .956.991 1.005 .951 .968 .939 .945 .956.995 .995 1.038 1.031 .992 1.003 1.0111.011 .968 1.024 1.004 1.002 1.055 1.017 1.036 1.029 1.031 1.084 1.026 1.05 OLD2 1.041 1.055 1.044 1.047 1.043 0 0 1.045 1.009 1.024 1.026 1.008 1.07 1.07.991 1.012 1.041 1.031 1.017 1.076 1.076 1.031 1.101 1.081 1.077 1.04 1.076 1.072 1.087 1.059 1.069 1.057 1.102 1.088 1.047.998 1.065 1.048 1.004 1.014 1.016 1.016 OLD3.964 1.019 .987 1.055 1.045 1.022 1.061.906 1.04 1.019 .98 1.024 1.01 1.014.964 1.105 1.083 1.011 1.047 1.016 1,028 1.063 1.012 1.029 1.047 1.056 972 .907 1.021 1.097 1.071 1.068 1.033 .911 .952 1.066 1.023 1.006 1.063 1.045 1.035 .992 DL D4 1.052 1.037 1.044 1.078 1.05 1.054 1.051 1.037 1.015 1.026 1.064 1.07 1.056 1.044 1.065 1.059 1.026 1.058 1.047 1.087 1.095 1.088 1.046 1.019 1.103 .993 1.0 6 1.041 1.056 1.045 .995 1.044 1.042 1.026 1.116 1.102 1.001 1.044 1.082 1.028 1 1.08 OLDS 1.106 1.05 1.002 1.017 1.042 1.034 1.037.98 1.122 1.:04 1:031 1.07 11:051 .951 1.059 1.01 1.005 1.135 1.022 1.076 1.058 .952 .981 1.023 1.023 1.049 .987 1.085 1.048 1.072 .98 1.1 1.017 .958 1.044 .991 1.056 1.074 OLD6 1.053 1.03 1.025 .987 1.031 1.059 1.087 1.005 1.049 1.006 1.058 1.058 1.011 .992.972 .985 1.012 1.009 1.067 1.017 .975.985 .979 .974 .961 1.017 1.008 .982.999 .987 1.021 .95e .954 1.064 .942.923 .981 .976 .97 .964 .99 1.004 N 0016 6O&

10 121 5.10 f, 14 : " is : ý -, E2]Nuclear Calculation Sheet S ubjeci Calc No Aev No Soei Nw Di7D812" 1.003 .94 .887 .843 .864 .845 .78.885 .801 .734 .811 .858 .806 .77.793 1.072 .807 .841 .824 .839 .859.815 1.105 1.123 .807 .851 .891 1.031.823 .836 .852 .893 .859 .898 .927 A842 .85 .875 .928 .899 .917 .9'41.914 .861 .907 .889 .948 .979 .977.983 .945 .969 1.034 1.003 .974 1.004.957 1.053 .9,66 .949 .95 .985 1.001.915 .981 .949 .963 1.012 1.026 1.009.962 .99 1.008 .964 1.013 1.009 1.039.984 1.005 1.06 1.01 1.03 1.013 1.065 1.039 1.014 1.022 1.027 1.104 1.012 1.033 1.013 1.015 1.005 1.025 1.056 1.056 1.095 1.051 1.023 1.021 1. 06 1.085 1.054 1.134 1.057 1.048 1-048 1:194 1.056 1.055 1.017 1-071 1.016 1.044 1.062' 1.052 1.077 1.08 1.037 1.043 1.014 1.037 1.028 1.069 1.024 1.01 .9g8 1.13 1.047 1.071 1.007 1.062 1.017 1.024 .974 1.024 1.075 1.016 1.063 1.012l 1.057 1.04 1.068 1.02"' 1.01 .986.976 1.03 1.039 1.057 .969 .927 .949 1.06 1.041 1.06 1.043 .974 .953 .938 1.006 .983 1.07 1.054 1.03 .976 .956 1.044 1.O11 1.03 1.019 1.105 1.068 1.002 1.078 1.045 1.029 1.107 1.016 1.036 1.065 1.02 .968 1d024 .964 .962 1.035 1.041 1._e5 1.089 1.065 1.097 1.025 1.089 1.027 1.042 1.049 1.115 .985 1.11 1.051 1.07 1.088 .925 1.038 1.095 .971 1.032 1.058 1.05 1.0213 .974 1.002 1.035 .993 1.059 1.049 1.015 1.052 1.083 1.029 1.001 1.015 1.G73 1.021 1.003 .975 1.079 .947 1.03.981, .989 1.11 1.055 .951 .95 1.056 1:039 1.074 1.025 1.124 1.045 1.026 1.087 1.026 1.003 1.045 .983 1.056 1.112 1.065 1 1:0i2 .999 .993 1:027 1:05 1.033.986 1.011 1.051 .995 1.041 1.0D02 .991.98 .964 1.00? .953 1.025 .975 .972.949 .986 1.016 .977 .949 1.083 .932.965 .942 .941 .936 .961 .977 .9712:-neW2 = d17dei2t(inltS(1 3 ,ie),).-new3 =d17d~i2t(jnlts(19,24),)

.:jnew5 = di7d812t(iflts(3i,36>).

_nreW6 = d17d8I2t(inlt5(37,4A 2),)", ,06-8!

I.C!2ý,10t

4:2P:r7 Nuclear Calculation Sheet Sub,.ecl, C/aleNo e-- No ' Sneet .r+Orngwllot Daie Re v.e weeby BAY /7D T,.'we'NEWI.914 .861 .907 .889 .948 .979 .977.983 .945 .969 1.034 1.003 .974 1.004.957 1.053 .966 .949 .95 .985 1.001.915 .981 .949 .963 1.012 1.026 1.009.962 .99 1.008 .964 1.013 1.009 1.039.?94 1.005 1.016 1.01 1.03 1.013 1.065 NEW2 1.039 1.014 1.022 1.027 1.104 1.012 1.033 1.013 1.015 1.005 1.025 1.056 1.058 1.095 1.051 1.023 1.021 1.06 1.085 1.054 1.134 1.057 1.048 1.048 1.194 1.056 1.055 1.017 1.071 1.016 1.044 1.062 1.052 1.077 1.08 1.037 1.043 1.014 1.037 1.028 1.069 1.054 NEW3 1.031 .988 1.033 1.047 1.071 1.007 1.062 1.017 1.024 .974 1.024 1.075 1.016 1.063 1.012 1.057 1.04 1.068 1.022 1.01 .986.976 1.03 1.039 1.057 .969 .927 .949 1.06 1.041 1.06 1.043 .974 .953 .938 1.006 .983 1.07 1.054 1.03 .976 .956 NEW4 1.044 1.011 1.03 1.019 1.105 1.068 1.002 1 $1I,45 1.029 1.107 1.016 1.036 1.065 1:831 :-68 1.024 .964 .962 1.035 1.041 1.085 1.089 1.065 1.097 1.025 1.089 1.027 1.042 1.049 1.115 .985 1.11 1.051 1.07 1.088 .925 1.038 1.095 .971 1.032 1.058 NEW5 1.05 1.023 .974 1.002 1.035 .993 1.059 1.049 1.015 1.052 1.083 1.029 1.001 1.015 1.073 1.021 1.003 .975 1.078 .947 1.03.982 .989 1.11 1.055 .951 .95 1.056 1.039 1.074 1.025 1.124 1.045 1.026 1.087 1.028 1.003 1.045 .983 1.056 1.112 1.065 NEW6 1.078 1.02 .999 .993 1.027 1.05 1.033 1.032 .992 1.035 1.052 .989 .952 .989.986 1.011 1.051 .995 1.041 1.002 .991.98 .964 1.007 .953 1.025 .975 .972.949 .986 1.016 .977 .949 1.083 .932.965 .942 .941 .936 .961 .977 .972 N" 00 16 41064 Ca:c No PYý o he PROGRAM: DWCHISQ ENTER NAME OF DATA LIST oLd ENTER PT NUMBER LIST hnt.(1.42)

ENTER NAME OF DATE LIST oLddatl:-N 2 3 4 OLD OLDI t( L. Ill.OLD3 0LD. D)4 OLD')ENTER NO. OF DESIRED DATA 1,2.3.4,5,6 OLDI OLD2 OLD3 OL.D4 OLD5 OLD6 CHISQ*2*.* 7*2.1073 2. 368 5. 991 6.1309.3404 2.72 75 OLDDATES*2/9/86 12/9/86 12/9/86 12/9/86 1 2/9/86 12/9/86 CH1 9 52 5.99 5.99 5.99 5.99 5.99 MEANTHK.98762 1.0468 1.0207 1.0508 i.0358 1.003 SD.043i69.029404.046986.029182.045087.037224 S I)E RR.0066611.046492.e 072501.00450*9.0069571-0057438 DFM2 2 2)2 CH1992 9.21 9.21 9.21 9.2t 9.21 EXF 8.8981 8.4"744 8.8981 8.8981 8.8981 8.8981 7.9565 7.5776 7.9565 7.9565 7.9-65 7.956-8.2908 7.896 S.2908 8.2908 8.2908 8.2908 7.9565 7.5776 7.9565 7.9565 7.9565 7.9565 8.8981 8.4744 8.8981 8.8981 8.898! 8.8981i1 8 9 9 a 6 4 3 5 8 12 6 9 13 14 7 E 8 7 9 59 6 ij 10 9 9 a B CRA.ND MEAN THICKNESS

= I .0ý241 STANDARD ERROR OF THE GRAND MEAN = .010251 JanuaTy 18, 1989 1 :09 FM MW, A-e,,=W 98 7. & z +/- 4D. , hb2d, ,( )

ITac.-No No 7VfiO' elt Noj jc.3o21e-530-0's 0 0`1 0 PROGRAM: DWCHISQ ENTER NAME OF DATA LIST new ENTER FP NUMBER LIST irit-(1 ,4')ENTER NAME OF DATE LIST ýiwd-ates N NEW I NEWt 2 NEW2 3 NEW3 4 NEW4 5 C EU5 6NEU6 ENTER NO. OF* DESIRED DATA 1,2.3,4,5,6 NEWJDA IES NEW2 1212/3/88 NEW3 12/23/88 NEW4 1"'/23/88 NEW5 12/23/88 NEW6 1./.3/88 CH ISQ CHmI952 3.0481 5.99 4.8554 5.99 4.2277 5.99 1,.5117 5.?9 1.1584 5.99 1.5888 5.99 (I8S 6 9 11 7 9 10 9 3 9 6 8 13 8 10 9 1 11 5 9 6 10 7 6 1l 10 8 1 MEANTHK 981 21 1.0501 1 .0171 1.0423 1.0312 99476 SD.043301.038922.040598.044753.A43294.038463 STDERR.0066315.0055275.0062644.0069055.0066804.005935 D'FM2 2 2 2 2 2 2 CH1992)9.21 9.21 9.21 9.21 9.1,7 EXPF 9 8.8981 8.8981 8.8981 8.8981 8.8981 8.8981 8 7.9565 7.9565 7.9565 7.9565 7.9565 7.9565 0 8.2908 e.2908 .?'908 8.2908 8.2908 8.2908 5 7.9565 7.9565 7.9565 7.9565 7.9565 7.9565 0 8.8981 8.8981 8.8981 8.1,981 8.B981 8.898!GRAND MEAN THICKNESS

= 1.0194 STANDARD ERROR OF THE GRAND MEAN = .011071 January 18, 1989 1 10 PM aRe NN R Sheet No COMPARISON OF' MEANS USING iWO-TAILED T-TEST BAY DATASHTS DATASETS DATADATE MEANTHK*** *ft*N****

• ttf*******

ftt**f**ft

• • • • • • • ~f 17D 8604956 OLDI 12/09/86 .98762 8702664 NEWi 12/23/89 .981 21 OLDI.93 .932 .943 .958 .92' .889 .913 1.0i4 .953 .984 .987 .973 .939 .956.991 1.005 .9 1 .968 .939 .945 .9'6.995 .995 1.038 1.031 .992 1.003 1.011 1 ....*968I 5'1:.0119 1:8 1 2 96§18'41 1.004 1.002 1 ol!I1.064 1.02& i05 NEWi.914 .861 .9Q07 .889 .948 .979 .977.983 .945 .t269 1.034 1.003 .974 1.i004.957 1.053 .966 .949 .95 .985 1.001.915 .981 .949 .963 1o012 1..026 1.009.962 .99 1.008 .964 1.013 1.009 1.039.984 1.005 1.016 1.01 1.03 1.013 1.065 F TEST FOR EQUAL POPULATION VARIANCES VARA VARB DFA DFE.00187.5 .0018636 41 41 F= 1.0061 F.O$/. 41 , 41 ) = 1.8604 F(.0i/12.

41 41 ) = 2.2716 TWO-TAILED T-TEST DF = 82 ALPHA = .24957' = .67885 T(.05/21 62 ) =1.9893 T(.01/2, 82 ) = 2.6371 January 13, 1989 6:08 PM c e No P.it No I s hee, %a COMPARISON OF MEANS USING TWO-TAILED T--TEEST BAY DATASHTS 17D 8604956 8702664 DATASETS OLD2 NEW2 DATADATE 12/09/86 12/23/88 MEANTIHK-.0468 I .0501 OLD2 t.-41 1.055 1.044 1 .047 1.043 0 0 1.045 1.009 1.024 1.026 1,08 1.07 1.,0?.991 1.012 1.041 1.031 1.017 1.076 1.076 1. 03 1 1,10 1 .081 1.077 1.04 1.076 1.072 1.087 1.059 1.069 1.057 1.102 1.088 1.047.9913 1.065 1.048 1.004 1.014 1.016 1.016 NEW2 1.039 1.014 1.022 1.027 1.104 1.012 1.033 1.013 1.015 1.005 1.025 1.056 1.058 1.095 1.051 1.023 1.021 1.06 1.085 1.054 1.134 1.057 1.048 1.048 1.194 1.056 1.055 1.017 1.071 1.016 1.044 1.062 1.052 1.077 1.08 1.037 1.043 1.014 1.037 1.028 1.069 1.054 F TEST FOR EQUAL POPULATION VARIANCES VARA.0012832 VARB 8.6459E-4 DFA 41 DFB 39 F -1.4842 F(.05/2, 4,1 39 ) 1 .3803 F,.01/2, 41 39 ) = 2.305 TWO-TAILED TEST DF = 80 ALPHA = .32681 T = .45043 T(.O5/2, 80 ) = 1.990i T.91/221 00 ".6387 January 13. 1989 6:10 Pf:M TW4Yo- ez7-s 3 oooci" IT COMPARISON OF MEANS USING TWO-TAILED T-TEST PAY DATASHTS DATASETS DATADATE MEANTNK 17D 8604956 OLD3 12/09/86 1.0207 8702664 NEW3 I 12/3/S'8 1. 0 171 OL.D3.?64 1.019 .987 1.055 1.045 1.022 1.061.906 1.04 1.019 .98 1.024 1.,.'1 1.014.964 1.105 1.083 1.011 1.04"7 1.016 1.028 1.063 1.012 1.-29 1.047 1.056 .972 .9'37 1.021 1.097 1.071 1:068 1.033 .911 .952 1 .0,66 1.023 1 .06 1.063 1.,045 1.035 .992 NEW3 1.031 .988 1.033 1.047 1.071 1.007 1.062 1.017 1.024 .974 1.024 1.075 1.016 1.063 1.012 1.057 1.04 1.068 1.0.22 1.01 .986.976 1.03 1.039 1.057 .969 .927 .949 1.06 1.041 1..06 1.043 .974 .153 .938 1.006 .983 1.07 1.054 1.03 .976 .956 F TEST FOR EQUAL POPULATION VARIANCES VARA VARB DFA DFB.0022077 .0016482 41 41 F 1 1 .339541 , 41 ý = 1.3604 1 .7 1/2, 41 , 41 ) = 2.2716 TWO-TAILED T-TEST DF = 82 ALPHA = .35423 T = .37522 T(.05/2, 82 ) = 1.9893 TQI1/2, 821 4 2.6371 JAnuary 13, 1989 6: 11 PM CuIc NO No IN- SheolNo COMPARISON OF MEANS USING TWO-TAILED T-TES1 BAY DATASHTS DATASETS DATADATE MEANTHK 17D 8604956 OLD4 12/09/36 1.0508 8702664 NEW4 12/23/88 1.0423 (3L.D4 1.052 1.037 1.944 1.078 1.05 1.054 1.051 1.037 1.01i5 1.026 1.064 1.07 1.056 1.044 1.065 1.059 1.026 1.O)R 1.047 1.067 1.095 1.088 1.046 1.019 1.103 .993 1.086 1.041 1.056 1.045, .995 1.044 1.042 1.026 1.116 1.102 1.001 1.044 1.082 1.028 1 1.08 NEW4 1.044 1.011 1.03 1.019 1.105 1.068 1.002 1.078 1.045 1.029 1.107 1.016 1.036 1.065 1.0' .968 1.024 .964 .962 1.035 1.041 1.685 1.089 1.065 i.097 1t0'5 1.089 1.027 1.042 1.049 1.115 .985 1.1l 1.051 1.07 1.088 .925 1.038 1.095 .971 1.032 1.058 F TEST FOR EQUAL POPULATION VARIANCES VARA VARB DFA DFB.0020028 8.516E-4 41 41 F = 21.3518 F>F(.05/2, 41 , 41 ) = 1.8604 F(.9/2.41

,41 ) = 2.,2716 :lecT 4 vrfJsrnt TWO-TAILED T-TEST /Aqairs DF = 82 ALPHA = .15277 T = 1.0311 T(.05/2, 62 ) = 1.9893 T(.OI/2, 82 ) = 2.6371 January 13, 1989 6:12 rM

Calculation Sheet/ 7D 7-WEtcA4 --47#V SUser-, L4 7- A4~ 5.ýe, OC' A 1 EAMS WMm- 44V 4L p 5~p'ys ~c'i.. 'e-~e ~jVTh 7e 4* ,.^cr~Cz is OT-_ ,- ---+ -E /S ,, /s.lAJh"pre :/7 CRCS oFOLD 4 SOF ~5 fXCC-O 0-P~A AWA'4-I -9 r~bAAAUV.

A4..=e. '-/Z 4A,,,&Nce C =r, C.= ~~-5-t 0 N,70016 (06-86)

[IMNuclear

! 4 0 2 .: .7 Calculation Sheet 0./ 6-,), (",,,Lz'C 6,0. 72?/7 0 z 4- ( -C jC ;, ~ ~ FQG-Cbam O F -,, (-o. 707)770.53 7/"-& (6S&o, 7/)e(o-o /a, w')... A8 r ,fC?.EjEpCr

/IýyPOTHC-sl 7-,-/Rrn'~J 7T-tis 1,vpej,&.s 7-H&~7- 7WI9A C~o 5oslo.)/;V IVMor61A/P

, S/71]N',"016 (06-86)

IOIZ5ICC 14:26:$' ,. .. ____._.,___

] C.1* NO [Rev N0 1 Sheol NO !'Caic No le'eN COMPARISON OF MEANS USING TWO-TAILED T-TEST BAY DATASHTS DATATETS DATADATE MEANTHK l7D 8604956 OL.D5 ¶2/09/36 ¶.0358 B702664 NEW5 12/23/88 1.0312 OLD5 1.106 1.05 1.002 1.0J7 1.042 1.034 1.037 1.069 .965 .988 1.122 1.034 1.032 1.07 1.097 1.028 1.051 .951 1.059 1.015 1.005 1.135 1.022 1.076 1.058 .952 .981 1.023 1.023 1.049 .987 1.085 1.048 1.072 .98 1.1 1.017 .9.58 1.044 .99i 1.056 1.074 NEWS 1.05 1.023 .974 1.002 1.035 .993 1.059 1.049 1.015 1.052 1.083 1.029 1.001 1.015 1.073 1.021 1.003 .975 1.078 .947 1.03.982 .989 1.11 1.055 .=51 .95 1.056 1.039 1.074 1.025 1.124 1.045 1.026 1.987 1.028 1.003 1.045 .983 1.056 1.112 1.065 F TEST FOR EQUAL. POPULATION VARIANCES VARA VARP, DFA DF'B.0020328 .0018744 41 41* F = 1.084541 41 ý' 1.3604 F('.91/2, 41 , 41 ) 2.2716 TWO-TAILED T-TEST DF = 82 ALPHA = .31752 T = .47643 T(.05/2, 82 7 1.9893 (.0112, 82 ) =,2.6371 ,Januar: 13, 1989 6:i3 H 1Catc NO Pg.. No iSheet 940 COMPARISON OF MEANS USING TWO-TAILED T-TEST BAY DATASHTS DATASETS DATADATE MEANTHK*** ******** ***W**** ******** ****W**1 17D 8604956 OLD6 12/09/86 1.003 8702664 NEW6 12/23/98 .99476 OLD6 1 .053 1 .03 1.0'15 .9087 1 .03'1 1 .059 1.087 1.005 1.049 1.006 1.058 1.058 1.011 .992,.972 .985 1.012 1.009 1.067 1.01? .975.985 .979 .974 .961 1.017 1.008 .982 999 .987 1 0'11 958 .954 1 .064 .942?923 .981 :976 :97 .964 .99 1.004 NE Yb 1.078 1.02 .999 .993 1.027 1.05 1.033 1.032 .992 1.035 1.052 .989 .952 .989.986 1.011 1.051 .995 1.041 1.002 .991.98 .964 1.007 .953 1.025 .975 .972.949 .986 1.016 .977 .949 1.083 .932.965 .942 .941 .936 .961 .977 .972 F TEST FOR EQUAL POPULATION VARIANCES VARA VARD DFA DF'B.0014794 .0013856 41 41 F = 1.0677 F(O5/2, 41 , 41 ) -1.8604 F(.(.1/2,, 41 , 41 ) = 2.2716 TWO-TAILED T-TEST DF= 82 ALPHA = .16005 T = 1.0003 T(.05/2, 82 ) 1.9893 (.01/2, 82 ) = 2.6371 January 13, 1989 6:14 10/25,10C 14 .2'ý: ', CaWc. No. C-1302-187-5300-005 Rev. No. 0 Page/ of 5.4.2 Bays 17/19 Frame Cutout. December 1988 Two sets of 6"x6" grid measurements were taken in December 1988. The upper one is located 25" below the top of the high curb and the other below the floor. There is no previous data. The upper location has been added to the long term monitoring program. With no prior data, the only possible analysis was to check the data sets for normality using the chi-squared test.The data at the upper location are not normally distributed.

The lack of normality was tentatively attributed to minimal corrosion in the lower half of the 6"x6" grid with more extensive corrosion in the upper half of the grid. To test this hypothesis, each data set was divided into two subsets, with one containing the top three rows and the other containing the bottom four rows. These subsets proved to be normally distributed, thus confirming the hypothesis.

The current mean thickness

+ standard error is 981.7 +4.4 mils for the top three rows and 1003.8+6.6 mils for the bottom four rows.The data at the location below the floor is normally distributed.

Also, the mean thickness is higher than at the upper location.

The mean thickness

+ standard error is 1034.1 +6.8 mils.

Cj~c No C. 130 7-: I 3 y-s.0~D8702666.986 .98 .97 .975 .975 .97 .928.982 .986 .97 1.01 .93-1 .973 .959.986 .983 1.00i .993 .975 1.032 1.001 1.122 1.1.' .983 .997 1.015 1.01 .978 1.005 1.003 .975 .986 .979 .997 .96 1.082 1.038 .985 .978 .939 .97 1.017.976 1.012 1.052 1.011 1.049 1.01 1.019 D8702663 1.027 1.057 .993 .958 1.062 1.025 .897.988 .973 1.011 .99 1.04B 1.141 1.101 1.079 1.113 1.033 1.017 1.076 1.064 1.04 1.017 1.007 1.051 1.021 1.028 .97 1.039 1.064 1.005 1.052 .983 .96 .991 1.042 1.087 1.014 1.054 1.049 1.039 1.017 1.044 1.142 1.017 1.019 1.001 1.059 1.109 1.095 S" Caic No 1 Rev ~ SP t o c. 1i3o2-I6e7-S~o PROGRAM: OCDWCONF!i;AY : 17/19FR D87:32666.986 .98 .97 .975 .?75 .97 .92c.982 .986 -.97 1 .Q 1 .981 .9'73 .959.986 .983 1.001 .993 .97-' 1.032 1.001 1.122 1. 0i7'2 .983 .997 1.Ji 5 1.01 .978 1.00.5 1.063 .975 .986 .979 .997 .96 4.028 1.038 .985 .978 .939 .9? 1 017.976 1'. 012 i.052 1. Q)1 t.'1 49 1.01 1o.19-MEAN THICKNESS

..99433 STANDARD ERROR OF THE MEAN = .0044713 T(..05-,/2.

48 ) 2.0106 T.01/2, 48 ) 2.6822 COWNFIDENCE INTERVALS fOR THE MiEAN fft*

• f* ft*** ********1I fttf***ft*

fttfttft*

f* K*t***9`--! UPPER' BOUND = 1 0033 95t. LGWEt: FOUND = .98534 99% UPPER BOUND = 1.0063 99% LOWER BOUND = .98233 Janu.n-'r, 20, 1989 10:44 AM 0 Calc No "- .No Shect No PROGRAM: DWCHISQI BAY: 17/19 D8702666 CHTSQ 28.617 OB$ E: 4 10.23 9.5 9.ii 9.6 1 (.: DATADATE 12/30/88 MEANTHK.99433 A312??.0044713 DFM2 2 CHII52 5.99 CH1992 9.21 381 2826 6726 02826 381 PTNOS I 4 5 6 8 9 10 PTNO]S0 11 12 13 14 15 16 17 18 19 20 PTNOS 21 22 23 24 26 27 28 29 30 PTNOS 31 32 33 34 3-r.J .36 37 38 39 40 PTNOS 41 42 43 44 45 46 47 48 49 0 Janua'y 18, 1 989 j :13 41 0 c acLN Rev No Sheet NO~PR:OGRAM:

DW~CHISQi B~AY: 17/19 D8702666 CHI Sri c 7.884-, 4.4'7 4.1-19.4 4.4~DATADATE 12/30/88 MEA NTH K f t** ft at** f.98171 SD ft ft* ft* ft ft T TD E R.0~04441 2 DFM2 5.99 CH 1992 9.21 491S 49 1 P'TNOS F"TNOS i 12 2 13 14 4 1.: J 16 6 '7 7 18 6 1921.January 12, 1989 1L14 PM 1012510C 14:2f:C'C -lc ---ev No ! Sheet No C~0 PROGRAM: DWCHISQI PAY: 17/19 D8702666 CHISQ {A ****4.3034 OUS E)*** ***I 4 5.9?7 5.3(9 55: 4 5.3(4 5.9: FTNOS 23 24 25 26 27 29 30 31 DATADATE 12/30/88 MEANTHK 1.0038 SD.034903 STDERR.*0065961 D FM 21 2 CHI952 5.99 CHI992 9.21 (F 321)43 272)43 S21 PTNOS 32 33 34 35 36 37 38 39 40 F'TNOS 41 42 43 44 45 46 47 48 49 Januarv 18, 1989 I :15 PM 0

!4:2e.e' C "W PROGRAM~:

DWCHISQi BJAY: 17/19 D 8702666 CHISO 11 .812 OB$ E;***t *M*11 5.3(6 FPTNOS i 3 4 5 6 7 9 19 10~Q DATADATE 12/30/8e MEANTHK.99046 SD.032714 STDERR.0061023 DFM2 CHI 9952 5.99 CH 192 9.21 XF'321 043 272)43 321 PTNOS 11 12 13 14 is 16 17 18 19 F:TNOS 2 1 22 23 24 26 27 28 Januari i8, 1989 0 24V254~ !4:2:f ra No N c./302-/d'7a53Co~95f34j PROGRAM: DWCHISQ1 BAY: 17/19 D8702666 CHISL I*6 9258 OBIS E4 3.92 3 4.1'5 3.92 4 4.44 PT NOS 29 30 31 ,32-333 34 35 36 37 38 39 DATADATE 12/30/88 MEAN 1'.K.t***94.99948 rD.029286 0T6DE:RR.0063907 DFM2 2tf A 1I952 5.99 491 F82 454 182 9 1 FTNOS 40 41 42 43 44 4tr 46 47 48 49 CH1992 9.21 January 18, 1989 1 1s PM 0 N 10Q,'2-=/06 14:26.:(-Calc No 40. l,,et ---1 PROGRAM: OCDWCONF PAY: 17/19FR D8702663 1 027 1 5057 .993 .9"SI .062 21 .0. .,.9?8, .973 1.011 .99 1.048 1.i41 1. 10,;1.079 1.113 1.033 1.017 1.076 l.0J64 1.04 i.017 1.007 ¶.051 1.021 1 .C2- .91 1..039 1.064 1.005 1.052 .983 .96 .991 1.042 1.087 1.014 1,.i ,t4 1.049 1.039 1.0.7 1.044 1.142 1 .0i7 1.019 1.0( 1 1 ..59 1 .09 I .095 MEAN THICKNES$

= 1.0341 STANDARD ERROR OF NHE MEAN .0067931 T(.0../2., 48 2. 0106 T(.01/2. 48 )= 2.6822 CONFIDENCE INTERVALS FOR THE mEAN 95% UPPER BOUND = 1.047?95% LOWER IBOUND '0204 99% UFPPER BOUND 1 .0523 99% LOWER BOUND = 1.0158 January 20. 1989 i0:43 AM 0 Tcalc N~O ~i ,c.i o2-/e7-S 3 0 PROGRAM: DWCHISQi BAY: 17/19 (4ffI.AAJ Pzo oe)D8702662 CHI:nQ.61772 OBS E 10 10.11 P.9 9.10 9.9 10.DATADATE 12/23/88 MEANrHK t .0341 SD)(475* * *t4.047551.0T0ERF3.0067931 DFM2 CH1952 5.99 CH I992 9.21"XF'381 2826 6726 2826 381 PTNOS I 2 3 4 6 7 8 9 10 PTNOS 11 12 13 14 15 16 17 18 19 2)PTNOS 2!23 24 25 26 27 28 29 30 P'TNOS 31 32 33 34 35 36 37 38 39 40 PTNOS 41 42 43 44 45 46 47 48 49 Januarv ¶8. 1989 1:12 FPM Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 174of-5.5 6" Strips in Sand Bed Region 5.5.1 Bay ID: 11/25/86 to 12117/8 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls below the 99% lower bound of the new 7-point data set. Thus, the corrosion rate is class-ified as indeterminable.

The current mean thickness

+standard error is 1114.7 +30.6 inils.ke.I A.LL C T=.-W 01-2.5-6G6

&t-ir.Le pa,Tjr Tj:Ztw4s AQE bo .&ie tcmMC O#%3-. o/q %

14 : 2 6: 07 i subject Calculation Sheet...... i -j/.Date Ri-v ewed tv Datef Oti kulor PROG RAM G E) 14C 0, Nr f'AY: iD D8702654.932 i.i4ýI .... 43 141 4, ,"IEAN THICKNES'S 1.1147 STANDARD ERROR OF THE MEAN .03058 S.,, 2.4469 T'..i/2, 6 ?= 3.7074 CONFIDENCE INTERVALS FOR THE IEAN 15% UPPER POUND = 1 .1895 95% LOWER BOUND = 1:0399 9ý9% LE UN 1 M38 99% LPPER EI:U.D'= 1: i January 16, 1989 12:36 PM~~eeýýIew&vs 79.e 0. 79e$ ,6~0* 0016 (064 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 179of 5.5.2 Bay 3D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls within the 99% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 1177.7 +5.6 mils, 0~ ~ ~~J /21/CC1 I~Jw'i~uguar Calculation Sheet Or 9nator Date Revteved by Date0; 1 v/FROGR'AM:

0CDW(CONF BAY -31)D8726 55 1 .194 I ,192! .192 iI ?ii 4. .15"4 1 .i8 I 1 66 MEAN THICKNESS' 1.1777 STANDARD ERROR OF THE MEAN = .0055751 T(.05/2. 6 )= 2.4469 T(.I/2. 6 )= 3.7074 CONFIDENCE INTERVALS FOR THE MEAN 95% UPPER BOJUND = I .I1?14 95X LOWER BOUND = 1.1641?9% UPPER BOUND = 1.i984 99;- LOWER BOUND = 1.157 January i6, 1989 12-37 PM VV/TH M /"rtE 5# 77 v,,L_..5 ,,,AF <,:>F 1/.F/CV' 7"L.0 14 Xie (0" Calc. No. C-1302-187-5300-005 Rev. No. 0 Page 1'of 5.5.3 Bay 5D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls within the 95% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness rate + standard error is 1174.0 +2.2 mils.

'2 5 / G fi 14 : 2 E : 0 7]Nuclear Calculation Sheet Subject CIec No 4W _~ ~ te/aRev-e~d by "we~55*3',r 4o5* 27-;>FROGRAM; OCDWCONF*F.:AY: 5D D8702656 1 412 I .177!

• 1 '79 1 .1t77 I .174 1 171 1 .178 MEAN THICKNESS 1.174 STANDARD ERROR OF THE MEAN = .0022467 T(.05/2. 6 )= 2.4469 T(.01i/2, 6 ) 3.7074 CONFIDENCE INTERVALS FOR THE MEAN 95% UPPER BOUND = 1.1795 95% LOWER BOUND = 1.1685 January 16, 1989 12:39 PM"TiEy: //- .2 -&1 A: r ýC* SS c /177 " V17-4rq/,A 7W.C 95. X n ,T -dqs* ~v~ .5 4 sýJj.4Cc0CAIi 7.VM- V4 N 0016 (06,4 1(42510ý-

14.'.'E,:V Calc. No. C-1302-187-5300-005 Rev. No. 0 Page ý2of 5.5.4 Bay 7D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data was compared as described in paragraph 2.7. The previous measurement falls just above the 99% upper bound of the new 7-point data set. This implies that corrosion has possibly occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 1135.1 +4.9 mils.4 0

[ ]Nuclear Calculation Sheet 0 Subject=.Otngltoi~

"3 D le pevvewed h, Date/-/06 ýd't 5X4464e9V 7b~ (2coJio)FPROGRAM-h:DWCONF BAY": 7D D870.26., I .146 1 .146 1 .147 1.141 I.129 1.121 1.116 MEAN THICKNESS 1.1351 ETANDARD ERROR OF THE MEAN .0049156 T(.05/2, 6 )= 2.4469 T(.OI/2. 6 )3 2.7074 CONFIDENCE INTERVALS FOR THE MEAN 95% UPPER BOUND = 1.1472 95% LOWER BOUND = 1.1231 99% UPPER BOUND = 1.1534 99% LOWER BOUND i 1169 January 16. 1989 1 2 .8 PM 7)-I/-c oF' I /NO"" Pcss 1 4-3L". 4)'6 406-Calc. No. C-1302-187-5300-005 Rev. No. 0 Page jIjof 5.5.5 Bay 9A: 11/25/86 to 12/17188 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls below the 99% lower bound of the new 7-point data set. Thus, the corrosion rate is class-ified as indeterminable.

The current mean thickness

+standard error is 1154.6 +4.8 mils.0 0ElNuclear Calculation Sheet Suobject Cac No iRevNo Shee t___. __ _ _ __ _ _-__ _" 5____ _s ,/0c 1-.O na'or Date Revievved by oee.5 5 .z 9y FROGRAM: 0CDWCONF BAY: 9A D8702660 1 .161 1 i6l 1 163 1.161 1 .t57¶ .152 t o 12?MEAN THICKNESS 1.1546 STANDARD ERROR OF THE MEAN = .0048001 T..0,/2. 6 )= 2.4469 T(.01/2, 6 )= 3.7074 CONFIDENCE INTERVALS FOR THE MEAN 95% UPPER BOUND = 1.1663 95% LOWER BOUND = 1.1428 SL BUND =I. 1368', / #-, 7-e .797, eO-,o,, 0.'3UA .4,707/C(f0~S4J1 N 0016 (06-1 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page jiof 5.5.6 Bay 13C: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls within the 95% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 1147.4 +3.7 mils.0

,FcarC .o -74.. No She., k0 PROGRAM: OCDWCONF BAY: 13C D9702661 1.154 1 .156 1.15 I .152 1.15 1 .t27* 1.43 MEA14 THICKNESS

= t.i474 STANDARD ERFROFR OF 'HE MEAN .0037407 T<,.11/2, 6 )= .4469" )= 3.7074 CONFIDENCE INTERVALS FOR THE MEAN 95% UPPER BOUND = 1.1566 95% LOWER BOUND = 1.1383 99% UPPER BOUND = 1:1613 994" LOWER BOUND = 1:1336 Jansuar 16, 1989 12:46 rmN IVI FH ~74.? fS7- Co"F1Z>SIOJI.C lAzr6.4t*/9L' IS Al A)o 51 ,FJ 6FICA)J7-0 Calc. No. C-1302-187-5300-005 Rev. No. 0 Page / of 5.5.7 Bay 13D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls within the 95% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 962.1 +22.3 mils.

t... .40 SieN F'FROGRAM:

OCDWCONF BAY: 13D D8702667 1.03.985.898.871.949 S.007 HEAN THICKNESS 96:)14 STANDARD ERROR OF THE MEAN .022'261 T(.05/", 6 )= 2.4469[(.01,2, 6 ) 3.70,74 CONFIDENCE INTERVALS FOR THE MEAN* **** *** *** * **** * * * * ** ** * * ** ****95"! UPPER BOUND = 1.0166'I:% LOWER BOUND = .90767 79% UPPER BOUND = .0447??% LOWER POUJND = .87961 January 16, 1989 12:47 Phol-LtS TrHG RS .I 0o-:4tt(-)4 Cale. No. C-1302-187-5300-005 Rev. No. 0 Page/7 f 5.5.8 Bay ISA: 11/25/86 to 12/19/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. Also, a 6"x6" grid data set was taken on December 2, 19B6 at this location.

As a best approximation, the first 5 points in the 7-point data set are at the same location as points 38 to 42 of the 6"x6" grid. These five points all fall within the 99% confidence interval of the new 7-point data set.The single measurement falls below the 99% lower bound.This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 1120.0 +12.6 mils.

/2sE : ... .R, No 1 hew No PROGRAM: OCDWCONF IAY: 15A D9702662 1.1I38 1 .147 i .1"14 i .123 1.14 1.-io 1 .05'EAN THICKNES i.12"TANDA,'"R ERFFGR OF' THE MEAN 1 ."i 2632 T(.05/2. 6 )= 2.4469 1k./! ./2, 6 ) -3.i074 CONFILDENCE INTERVALS FOR THE MEAN 95% UPPER POUND = 1.1509?5Z. LOWER BOUND 1 .'M91 99% UPPER :OUNB = 1 ,!668 99% LOWER 8OUND 1 .0732 Januarl1 6, 1989 12:47 WM MEp%&3 OF~ 5 pl-. rf:ofln IASOG TdItZ FALLS \,JtT1-,3 TrHC 917-i ClfoWtc FRKA-LS 13e"IV~ 77,06 rf % Lwee&`:>p- 4.'), TftC- Coar-.o ,uQ Is /: r""l'oT 0 10 12^ E/ lmý 14 -.2 P, -. 01 1'Eic %0 Re~ 140 St'eel No PROGRAM: OCDWCONF BAY: I5A IJ;)ATA Zfir: Z&-04i-2S-D I 5A61 2.6989 1.163 1.16" 1 1i69 1.. 6 7 1 .1': 1.t6-2.759 4.168 1.62 .66 1 ! 1 .1,.55.404 .754 1.167 '.16?' 1..69 1.164 1.166.64 i.L15 .736" 4..1.56 1.15 i.156 i.1.54.645 t 1 3 1 _ 1 j .1 43 :.5J3 1 126 1",.1,11 1ý .14Al 1 1:; .1 .11.:.)M.EAN THICKNESS

= 0 c33'"-OT,,N ARD E:RRfIR OF THE; MEAN .032-' -T'(.5/2, 48 ) 2.01'6 T(.(.1 /2. 48 )= 2.6822 CONFIDENCE INTERVALS FOR THE MEaN~f t* * * **t* ftft*** Y***********t****K*f***

'95% UPPER BOUND = i.095Y 95Z LOWER 6OUJND = .9678 9?.UPPE PR =Q UM.LO(W R Ef U 4w ,Jal(LarV 20, 1989 10:44 AM APPIb,. LO-T" CF QE6^01IIS I-S 0

!0/25/06( '4:2t:C07 Calc. No. C-1302-187-5300-005 Rev. No. 0 Paged/5of 6.0 APPENDICES 6.1 SPEAMEZ Programs 6.2 SAS Program 0 C Ic Nio Acev No 1Sheet No 1 ,C-13o2-/1f7-530p-0aS 1 Aeecom( (.. I 1 :2$ B~Y J .P. MOORE, PE'VU7U:P 1H1-: I .,Jt)RN(AL.

OM'PROGRAM DU~CH: S"U* .!4'" "NTER NJAME~ OF DATA L.IST' @PNCEFflR trL.IST Is-R 1-1 Y ETf r PFE iT; -T0S F:.rN01",NTS( 1 .49) 1`A EY 'EM 7 ' ý:' N4AME OF DAT E S. sTH -I, r-- H~cF-orkrN-D ATF I- I*-' 0 $fGEt DA~TAL 1"' Or 1GL 'DCA T S0TA.-i'1 ilL N. PATALI'T A E"T FNT-If.' 0 U EDr !E S CR EL D D TA NF. Ic LEC T-1-0 tNOEL.'k .ELEC I-r 4EI,- *~ 44--iD , 2.^~: 1.8944E-1, i.9740E~i I.4E- '3 4~90 R 10,1,0 = A ih .OS. -o:.', .25. f)."0 0 N = ^iP)(JýIi .~ 6 D 24IDJ Ii .'TDERR pfib'.'12.6ACARTH A 4 7 (J: 16.0) VFM AlD(J1 17.0 CHISG' AlDt"J:)18.0 F: O.*, I i *J 19.0 1< SLLECT(I)20.0 Al AiD(OBJECT(DATAL[VT(V)))

'210 2 A AI(LI25NAAO T0)~j ) NOELS ( 1*1"3.0 St)(i1 ITANDDEV(AI) 23 .K STDEFFR(I)

=STANDERR(AI)

'1.( MEANTHK(1)::-

MEAN(AI)B~INS =- D(*..*EqNSE)

+ IME:ANrF'1(T)

.., B EIN( .1) 1- 1 0 0 E-tS( .w' It NN ED( A 28.0 E XF'C. N" ~( Iiff CONF (I ) = H I SQUAFELV<OBS(,IEXP jI) DFr! F. CHfI :-CH I 30 0 DFM? ( T) DF -2 31j 0 c 14 1 aT d, ( T. = I-l 32 .NEXT I ME9NEAN (ME:AN TH<323 17E-MEAN .9TANDF;RF:'(ME~h1THK)-J. $ OUTHJT 22~~ C92 1( J: + .99 ,3.4 CH1992 AiD(J:) + 9.211"74.0 TAIBIJLATE( DATAI.isi(SEL ECT) ~DATELIST.

MEANTHK..

Fr- --fDERR 1)f7M2 74.!:, ABULATE(CHISQ, CHIV5~2, rH09.2 3¶0 TAE4ULATE(OBS, E YP'I FACE(I1I TYPE 'GRAND MEAN THiSS'GMAN

3 TYE

• TANDARD EPRO O.GADMA EMA H".4 SPACE(!)35.5 DATE;T IME~6 0 NEWPAGE 37.0 JOURNAL OFF 38.0 END o$ic ND Rea oSn,,N 2 -e 7 5- .0 0& fr L.ISTING OF PROGRAM DWCHIt-Qi.0" PROGRAM 4'.3 $ BY J.P.MOORE

-!8-.13 4.0 ASK(*ENTER NAME OF DA[rA :.ru, ",HENCEFORTH NAMEX 12 " 4£.!. DTASET = NAMELITT(NAMEXý 4.7 AS7KNAME( 'BAY NUMEFP','BAY= "i 5.0 ASKW(ENTER FT NUMBER LIST4, "F'TNOS= "TNOSINTS(1,49) 6.0 ASKNAME-(F.

'NTER E.ATE'.. TAISTE..1 "., FP = D, 1186E--., .4-sE-- I .-;: .-1 ; 2 .Iil3F -:'Y ')BINSQt-

= AiI'* : -h ..t. *0 .( 9. c*_:, ,,.', $,"7.0 PIN = AiD(' 16 j!,() O' .CX -Ajfl(':'Df,: BXF' =: AID(5: E:~~~~ X " A.,D A' A 14) ( A 15 JECT (DAIA , A -AI(FTNOX)..,0 AI A(LOCS(AI.GTU0).

23/4') N =hNOELS(Al)

n! SIANDI)EV(AI) 28.5 STDERP

= VTANDERPVAI) 9Q 1Q ¶r4.IANI'D

-:: MEtAN(AfI 3C'. : SINS ,*PINkSO + MEANTHK 31.0 BIN BIN" 32.0 0'T' = N I'NEDIr)3Z.0 EX N*F 34.0 ,ROB 2 HISQUARED(0BY.EXF-DF'F CHI=CHI)7':? flFM2HI -- 2'36. 0 HIr. I z S UT HI 38. 'S 0UTEU T 38. 2 JJURNAP4 ON:L.8.3 TI-E 'PROGRAM:

DWCHIXQ¶']8. 4 T'YPE "r'PAY : " BAY 39.0 CH1952 = 5.99 CH1992 = 9.21 46 .... TABUL4TE (DATASET.

I)ATADATE, MEANTHK .2T')ERR DFM2)40. 5 T'AVULA TE (CHI QCHi 52, CH1 992);-41 " TABULATFI(OBS.

EXf:')41.5 TABULATE PTNUS 4> d- DATE]T"I ME 43.0 NEtW AGE 44.0 JOURNAL OFF 45. 0 ENt' i-/ 3o-/o7- 53oo- .1 LISTING OF PROGRAM OCDWCONF 1.0 PROGRAM 1,5 t BY J.P. MOORE, REVI. E: i--1&--89 2.0 AS"K DATASET NAME'?' .

DATArv'i-.

3.0 fsKNAME(*BAY NUMBER?',f:A',-

4.) MEANDATA-=MEAN(DAT'ASE.T 5.0 STVERR=S'TANDERR (DATASIET?

.6 ':F NOELS(DArSET)T9F -ABS"TPROI"NVERSE'

.975.DF)3.0 T99

,1NVER.E(.'? .F?.0 UB95 = MEANDATA + T?!5STERRF IA.. 0 ' L, o = MEANDATA -T9.9 T'nF R. R i1 v. U1?9Q = MEACIl)ATA

+ T-99*Srl'EhR 12 .0 :::? =MEANDATA

-T99.,1lDERR 1 3.(- JOURNAL ON 14.0 TY-PE "PROGRAM:

1t(:DW(CONF" i5 .0 TYFi-E '"PAY: 'BAY 16 .0 TABULATE )"ATASET 17.0 TrYF:', -tEAN THi:CKNESS

=' MEANI)AITA I S. .Q YPE 01-TANDARD ERROR OF' THE M AN '" 'IPERR 19.0 rYF'E T.05/2,D LF .=" T95 20.0 TYPE !T(01/20 F ").-'99 21I. 0 SPACE (I 22.0 TYPE "CONFIDENCE INTERVALS FiOR rHE MEAN'23.0 TYPE ****************.********, 24.0 EPACE( I 25.0 T'Y'E "95% UFFER BOUND = IP95 2U.0 TYPE 005% LOWER BOUND =LB9 27.0 SPACE (I)28.0 TYPE "99% UPPER PO(JUND = B99 29.,G TYPE "o9"- LOWER BOUND =" ?.99 30.0 SPACE(1)50.0 DATE- TIME 51 .0 NE:WFPACE:

52.0 JOURNAL OFF 53.0 END C"Itc No R o Sheet No LISTING OF F!ROG~RAM T2TAIL 1 .0 PROGRAM 1.5 S BY J.P. MOORE REVISED 2.0 ASKNAME(ENTER AAY NUMBER E. "BAYz- '.3.0 ASK('ENTER NAME OF i.T DATA 5"E." "HENCEFORTH I)ATAt IS ";4.0 ASKNAMEENTER DATA SHEET NO.'. -.'4YASHTi=

P)5.0 ACKNAME('.

TER DATE OF ¶T I.T-TA ..... "1)ATE1: E 6,0 SK'ENTER NAME OF '?NO DATA SET, "HENCEFORTH DATA2 1S 7,0 .lEKNAME ("ENTER DATA OHEE, NU., ,)ATASHT2:.) ASKMAME("ENTER DATIE 'OF "'IND D"Tr , " 1- A- T .'?.0 T" = !q NAMEL.Elf"il)AT"'A

!TTA72)0',.0 DA'T '-tS 1AIA IT2 i .(-) DATADATES DATE. ,I)AT2 12" DI AID(DATAI)

13. 0 V' = AMP'DATA2)

14. A DZOI0K = DI (l..OCS.'(Di .NE E ) .)0 DICK =D(LOCS'I2.NE.., 5' 0 C.,.'Vr: i 7. M.iEANTHK " ME AN (! 1 F 0 MEA N -02`3K 1.9.0 FrE.J TI-N9.0 NI J NOELS(DlOe.)

N? q.OEL.S(D2r3Kt 2i.') DFI Ni-I 22' DF"2 W2-" 320 -1 VARIANCE(DIOK) 215 .1 F s;H^W- .v. ANCERI D2OK.) = D20K N.0 NA = N1 27.0 NB = N2 29.0 DFA = DFI.9..().0 PF? = DF2-3.'.', VARA VA~i 3(1 VARD VAR2"32. .OTO FTEST3 33.63 FTE'T2.2 DA D20K 7,,4 DB = liOK 34.03 NA =N2 3ftz:'- NB ="N!'.w" ":' DFA = rjF2= DFi 3e." VARA :VAR2 39.0 VA -. VARI 40. C, FITEST3 41 .0 F = VARA/VARP 42. 0 F-'9; FPROB INVERSE ('."C125, DFADFP , 42.5 F99 = FPROBINVERSE(.0O5'.DFADF'B) 43.0 44.0 ALFHA = TINDEPT(DA.001, T,DF)45.0 1"95 ='F'ROINVERSE(0.v:..D.)

46.0 T99 = TPROEBINVERSE(0.0()05,PF) 47.0 OUTPUT: 49.G JOURNAL ON 49.0 PRINT 'COMPARISON OF MEAN;2 USING r!'W,]-TAIL.ED T-'!E:.,T 50.0 PRINT , 1l.0 TABULATE B4AYDAITASHI'SDAIASE.

I'S, lATADATE:S, MEAN T HK 5'2.0 TABULATE DATAI*3.c TABULATE DATA2 54 .0 PRINT *F TEST FOR EQUAL FPOPULAT'ION VARIANCES"'55.0 P'PINT *****************~V 56.0() TABULATE VARA;VARWI)FA.DF7B "46 'g-A. TYPE *F=.05/2." DFA ,'2 1)FP .'y " n .. r , 2, 15F 7'," DF'- I F-9' 0 -, I'Ti" *TWO-TAILED V-'TEST':'.Q0 PRINT .****, ***, 69.0DF, ALPHA T3'3 "YPE " (.5 1: .D F "" .-" 0'T4.E I) A .I 1 0 4 4 0 F.JUI1/2: L. ,)F' C alc N o v N o S N' 7- S*oo- o° I

• data dwdata;i li'p tit 05 dates rmniddyvS.(M15 siddev 3.3;retain dayO.;if -n.. = 1 then da'vO = dates;= in-t:k('day',dav0.date.3).= ay s/365;.ards;5/1/37 92.336 8/1/87 107.370 9/10/87 109.444 7/12/8 9 9 .ISt5 10/)3/8'8

?4.979 proc print data=dwdata; titlel 'LINEAFR REGRESSION PLOT';titte2 'FOR DW WALL THINNING ANALYSIS';

titte3 'OF TIAY IIC 3" ABOVE CURP';f'ori mat dates mnmddyy8., format day{0 ismddyvv.

i proc meanr data=dwdata n mean std stderr;var stddev;proc reg data=dwdata; model. stddev =

stb ctm cti cothn: proc reu9 data=dwdata; model s+ddev = years/ p r cLi ct.m;output out=*a p=pred t95=1?5 u95=u95 r=residuat; clear;proc plot data=a;plot stddev*years='x' pred*years='p' u95*years='u' t95*year5='L'

/ overlay vaxis=800 to 1250 by 50;Proc plot data=a, plot residual*years='r'/

vaxis = -40 to 40 by 2; Nuclear R No. 948 Revision No. ...Budget Technical Data Report Activity No. 315302 Page 1 of 26 Project: Department/Section 5300 OYSTER CREEK Revision Date 2-1-87 Document Title: STATISTICAL ANALYSIS OF DRYWELL THICKNESS DATA Originator Signature Date Approval(s)

Signature Date___ _Approval for External Distribution Date Does this TDR include recommendation(s)? -Yes X No If yes, TFWR/TR# -* Distribution Abstract: J. D. Abramovici Statement of Problem F. P. Barbieri G. R. Capodanno The design of the carbon steel drywell includes a sand D. W. Covill bed which is located around the outside circumference D. G. Jerko between elevations 8' 1/4" and 12'-3". Leakage X. W. Laggart was observed from the sand bed drains during the 1980, L. C. Lanese 1983 and 1986 refueling outages indicating that water S. D. Leshnoff had intruded into the annular region between the J. A. Martin drywall shell and the concrete shield wall.J. P. Moore M. A. Orski A long term monitoring program was established in 1986 S. C. Tumminelli to take Ultrasonic Thickness (UT) measurements at rep-M. 0. Sanford resentative locations on the drywell shell to determine D. G. Slear the corrosion rate and monitor it over time. The R. W. Keaten initial program included six locations in the sand bed region. The program was expanded in 1987 to include measurements at higher elevations.(For Additional Space Use Side 2)This is a report of work conducted by an individual(s) for use by GPU Nuclear Corporation.

Neither GPU Nuclear Corporation nor the authors of the report warrant that the report is complete or accurate.

Nothing contained in the report establishes company policy or constitutes a commitment by GPU Nuclear Corporation.

Abstract Only Abstract Continuation TDR No. 948 Revision No. I A cathodic protection system is being installed in selected regions of the sand bed to minimize corrosion of the drywell. The long term monitoring program was further expanded in 1988 to monitor the effectiveness of the cathodic protection system and to monitor additional sand bed regions not covered by cathodic protection.

A critical part of the long term program is the statistical analysis of the UT measurements to determine the corrosion rate at each location.

This report documents the assumptions, methods, and results of the statistical analyses of UT measurements taken through December 31, 1988.Summary of Key Results Bay Area Location 11A 11C 17D 19A 19B 19C 9D 13A 15D 1?A Sand Sand Sand Sand Sand Sand Sand Sand Sand Sand Bed Bed Bed Bed Bed Bed Bed Bea Bed Bed 5 9 13 i5 51' Elev.87' Elev.87' Elev.87' Elev.Corrosion Rate*" Not significant Indeterminable

-27.6 +/-6.1 mpy-23.7 74.3 mpy-29.2 ;0.5 mpy-25.9 ;4.1 mpy Indeterminable*

Not significant*

Possible*Indeterminable*

-4.3 +0.03 mpy Not significant Not significant Not significant Not significant*

Indeterminable*

Indeterminable*

Not significant*

Not significant*

Possible*Indeterminable*

Not significant*

Not significant*

Not significant*

908.6 916.6 864.8 837.9 856.5 860.9 1021.4 905.3 1056.0 957.4 750.0 620.3 635.6 634.8+5.0 milo+1-0.4 mils+6.8 mile+4.8 mile;0.5 mils;4.0 mile+9.7 mile+10.I mils+9.1 mile 79.2 mils+0.02 mile+1.0 mile+0.7 mile+0.7 mils Mean Thickness*"*

17D 17/19 Trench Frame Cutout 981.2 +6.7 mile 981.7 +4.4 mile ID 3D 5D 7D 9A 13C 13D 15A Sand Sand Sand Sand Sand Sand Sand Sand Bed Bed Bed Bed Bed Bed Bed Bed 1114.7 1177.7 1174.0 3135.1 1154.6 1147.4 962.1 1120.0+30.6+5.6+2.2+4.9+4.8+3.7+22.3+12.6 mils mile mils mile milo mile mils mils One data point in Bay 19A and one data point in Bay 5 Elev. 51' fell outside the 99% confidence interval and thus are statistically different from the mean thickness.

• Based on limited data."Mean corrosion rate in***Current mean thickness See text for interpretation.

mile per year + standard error of the mean in mile + standard error of the mean Page la N 00308 (02.88) 16.26:33Nuclear DM NO, TDR 948 TITLE STATISTICAL ANALYSIS OF DRYWELL THICKNESS DATA REV

SUMMARY

OF CHANGE DATE (two places).Deleted redundant discussion of Bay 15D on pages 12, 19, 25 and 26. 1-. 4

-16:2f;-33 TDR 948 Rev. 0 Page 2 of 26 TABLE OF CONTENTS Sections Page

1.0 INTRODUCTION

3 1.1 Background 3 1.2 Statistical Inferences 4 2.0 METHODS 5 2.1 Selection of Areas to be Monitored 5 2.2 UT Measurements 6 2.3 Data at Plug Locations 6 2.4 Bases for Statistical Analysis of 6"x6" Grid Data 7 2.5 Analysis of Two 6"x6" Grid Data Sets 9 2.6 Analysis of Single 6"x6" Grid Data Set 10 2.7 Analysis of Single 7-Point Data Set 11 2.8 Evaluation of Drywell Mean Thickness 13

3.0 REFERENCES

15 4.0 EVALUATION OF DATA THROUGH 12/31/88 15 4.1 Results for 6"x6" Grids in Sand Bed Region at Original Locations 15 4.2 Results for 6"x6" Grids in Sand Bed Region at New Locations 18 4.3 Results for 6"x6" Grids at Upper Elevations 20 4.4 Results for Multiple 6"x6" Grids in Trench 22 4.5 Results for 6" Strips in Sand Bed Region 23 4.6 Sunnmary of Conclusions 25 iCI191(]6 TDR 948 Rev. 1 Page 3 of 26

1.0 INTRODUCTION

1.1 Background

The design of the carbon steel drywell includes a sand bed which is located around the outside circumference between elevations 8'-11-1/4" and 12'-3". Leakage was observed from the sand bed drains during the 1980, 1983 and 1986 refueling outages indicating that water had intruded into the annular region between the drywell shell and the concrete shield wall.The drywell shell was inspected in 1986 during the 1hR outage to determine if corrosion was occurring.

The inspection methods, results and conclusions are documented in Ref. 3.1, 3.2, and 3.3.As a result of these inspections it was concluded that a long term monitoring program would be established.

This program includes repetitive Ultrasonic Thickness (UT) measurements in the sand bed region at a nominal elevation of 11-3' in bays 11A, 11C, 17D, 19A, 19B, and 19C.The continued presence of water in the sand bed raised concerns of potential corrosion at higher elevations.

Therefore, UT measurements were taken at the 51' and 87' elevations in November 1987 during the hIM outage. As a result of these inspections, repetitive measurements in Bay 5 at elevation 51' and in Bays 9, 13 and 15 at the 87' elevation were added to the long term monitoring program to confirm that corrosion is not occurring at these higher elevations.

A cathodic protection system is being installed in selected regions of the sand bed during the 12R outage to minimize corrosion of the drywell. The long term monitoring program was also expanded during the 12R outage to include measurements in the sand bed region of Bays ID, 3D, 5D, 7D, 9A, 13A, 13C, 13D, 15A, 15D and 17A which are not covered by the cathodic protection system. It also includes measurements in the sand bed region between Bays 17 and 19 which is covered by the cathodic protection system, but does not have a reference electrode to monitor its effectiveness in this region.Some measurements in the long term monitoring program are to be taken at each outage of opportunity, while others are taken during each refueling outage. The functional requirements for these inspections are documented in Ref. 3.4. The primary purpose of the UT measurements in the sand bed region is to determine the corrosion rate and monitor it over time. When the cathodic protection system is installed and operating, these data will be used to monitor its effectiveness.

The purpose of the measurements at other locations is to confirm that corrosion is not occurring in those regions.This report documents the assumptions, methods, and results of the statistical analyses used to evaluate the corrosion rate in each of these regions. The complete analyses are documented in Ref. 3.7.

C (, I C .2 C : 3.4 TOR 948 Rev. 0 Page 4 of 26 1.2 Statistical Inferences 1.2.1 Statistical Hypotheses The objective of these statistical analyses is to make statistical decisions or inferences about populations on the basis of sample information.

In attempting to reach these decisions, it is useful to make assumptions or guesses about the populations involved.

Such assumptions, which may or may not be true, are called statistical hypotheses and in general are statements about the probability distributions of the populations.

In many instances we formulate a statistical hypothesis for the sole purpose of rejecting or nullifying it. For example, in performing a t-test to test the difference between the means of two samples we first hypothesize that there is no difference between the two means. This iB referred to as a null hypothesis.

Any hypothesis which differs from the null hypothesis is referred to as an alternative hypothesis, eg., the means are not equal, one mean is greater than the other, aet.1.2.2 Tests of Hypotheses and significance If on the supposition that a particular null hypothesis is true we find that result. observed in a random sample differ markedly from those expected under the hypothesis on the basis of pure chance, we would say that the observed differences are significant and we would be inclined to reject the hypothesis (or at least not accept it on the basis of the evidence obtained).

Procedures which enable us to decide whether to reject or not reject hypotheses are called tests of hypotheses.

1.2.3 Type I and Type 11 Errors If we reject a hypothesis when it should not have been rejected, we say that a Type I error has been made. If, on the other hand, we fail to reject a hypothesis when it should have been rejected, we say a Type II error has been made. in either case a wrong decision or error in judgement has occurred.1.2.4 Level of Significance In testing a given hypothesis, the maximum probability with which we would be willing to risk a Type I error is called the level of significance of the test. This probability is usually denoted by the Greek letter alpha. In practice a level of significance of 0.05 (5%) or 0.01 (1%) is customary.

If 0.05 has been selected, we say that the hypothesis is rejected (or not rejected) at a level of significance of 0.05.

16:2g:35 TDR 948 Rev. 1 Page 5 of 26 2.0 METHODS 2.1 Selection of Areas to be Monitored A program was initiated during the 11R outage to characterize the corrosion and to determine its extent. The details of this inspection program are documented in Ref. 3.3. The greatest corrosion was found via UT measurements in the sand bed region at the lowest accessible locations.

Where thinning was detected, additional measurements were made in a cross pattern at the thinnest section to determine the extent in the vertical and horizontal directions.

Having found the thinnest locations, measurements were made over a 6V-6" grid.To determine the vertical profile of the thinning, a trench was excavated into the floor in Bay 17 and Bay 5. Bay 17 was selected since the extent of thinning at the floor level was greatest in that area. It was determined that the thinning below the top of the curb was no more severe than above the curb, and became less severe at the lower portions of the sand cushion. Bay 5 was excavated to determine if the thinning line was lower than the floor level in areas where no thinning was detected above the floor. There were no significant indications of thinning in Bay 5.It was on the basis of these findings that the 6"x6" grids in Bays 11A, 11C, 11D, 19A, 19B and 19C were selected as representative locations for longer term monitoring.

The initial measurements at these locations were taken in December 1986 without a template or markings to identify the location of each measurement.

Subsequently, the location of the 6"x6" grids were permanently marked on the drywell shell and a template is used in conjunction with these markings to locate the UT probe for successive measurements.

Analyses have shown that including the non-template data in the data base creates a significant variability in the thickness data. Therefore, to minimize the effects of probe location, only those data sets taken with the template are included in the analyses.The presence of water in the sand bed also raised concern of potential corrosion at higher elevations.

Therefore, UT measurements were taken at the 51' and 87' elevations in 1987 during the IlM outage. The measurements were taken in a band on 6-inch centers at all accessible regions at these elevations.

Where these measurements indicated potential corrosion, the measurements spacing was reduced to 1-inch on centers. If these additional readings indicated potential corrosion, measurements were taken on a 6"x6" grid using the template.

It was on the basis of these inspections that the 6"x6" grids in Bay 5 at elevation 51'and in bays 9, 13 and 15 at the 87' elevation were selected as representative locations for long term monitoring.

TDR 948 Rev. 0 Page 6 of 26 The long term monitoring program was expanded as follows during the 12R outage: (1) m easurements on 6"x6" grids in the sand bed region of Bays 9D, 13A, 15D and 17A. The basis for selecting these locations is that they were originally considered for cathodic protection but are not included in the system being installed.

(2) Measurements on 1-inch centers along a 6-inch horizontal strip in the sand bed region of Bays ID, 3D, 5D, 7D, 9K, 13C, and 15A. These locations were selected on the basis that they are representative of regions which have experienced nominal corrosion and are not within the scope of the cathodic protection system.(3) A 6"x6" grid in the curb cutout between Bays 17 and 19. The purpose of these measurements is to monitor corrosion in this region which is covered by the cathodic protection system but does not have a reference electrode to monitor its performance.

2.2 UT Measurements The UT measurements within the scope of the long term monitoring program are performed in accordance with Ref. 3.4. This involves taking UT measurements using a template with 49 holes laid out on a 6"x6" grid with I" between centers on both axes. The center row is used in those bays where only 7 measurements are made along a 6-inch horizontal strip.The first set of measurements were made in December 1986 without the use of a template.

Ref. 3.4 specifies that for all subsequent readings, QA shall verify that locations of UT measurements performed are within +1/41 of the location of the 1986 UT measurements.

It also specifies that all subsequent measurements are to be within +1/8" of the designated locations.

2.3 Data at Plug Locations Seven core samples, each approximately two inches in diameter were removed from the drywell vessel shell. These samples were evaluated in Ref. 3.2. Five of these samples were removed within the 6"x6" grids for Bays 1IA, 17D, 19A, 19C and Bay 5 at elevation 51'. These locations were repaired by welding a plug in each hole. Since these plugs are not representative of the drywell shell, UT measurements at these locations on the 6"xW" grid must be dropped from each data set.

10/ý 9P56 I f: 26 : "';TDR 948 Rev. 0 Page 7 of 26 The following specific grid points have been deleted: Bay-Area Points 11A 23. 24, 30, 31 17D 15, 16, 22, 23 19A 24, 25, 31, 32 19C 20, 26, 27, 33, 5 20, 26, 27, 28, 33, 34, 35 2.4 Bases for Statistical Analysis of 6"x6" Grid Data 2.4.1 AssuMtions The statistical evaluation of the UT measurement data to determine the corrosion rate at each location is based on the following assumptions:

(1) Characterization of the scattering of data over each 6"x6" grid is such that the thickness measurements are normally distributed.

(2) Once the distribution of data for each 6"x6" grid is found to be normal, then the mean value of the thickness is the appropriate representation of the average condition.

(3) A decrease in the mean value of the thickness with time is representative of the corrosion occurring within the 6"x6" grid.(4) If corrosion has ceased, the mean value of the thickness will not vary with time except for random errors in the UT measurements.

(5) If corrosion is continuing at a constant rate, the mean thickness will decrease linearly with time. In this case, linear regression analysis can be used to fit the mean thickness values for a given zone to a straight line as a function of time. The corrosion rate is equal to the slope of the line.The validity of these assumptions is assured by: (a) Using more than 30 data points per 6Rx6" grid (b) Testing the data for normality at each 6"x6" grid location.(c) Testing the regression equation as an appropriate model to describe the corrosion rate.

TDR 948 Rev. 0 Page 8 of 26 These tests are discussed in the following section. In cases where one or more of these assumptions proves to be invalid, non-parametric analytical techniques can be used to evaluate the data.2.4.2 Statistical

%pproach The following steps are performed to test and evaluate the UT measurement data for those locations where 6"x6" grid data has been taken at least three times: (1) Edit each 49 point data set by setting all invalid points to zero. Invalid points are those which are declared invalid by the UT operator or are at a plug location. (The computer programs used in the following steps ignore all zero thickness data points.)(2) Perform a chi-squared goodness of fit test of each 49 point data set to ensure that the assumption of normality is valid at the 95% and 99% confidence levels.(3) Calculate the mean thickness of each 49 point data set.(4) Using the mean thickness values for each 6"x6" grid, perform linear regression analysis over time at each location.(a) Perform F-test for significance of regression at the 95% confidence level. The result of this test indicates whether or not the regression model is more appropriate than the mean model. In other words, it tests to see if the variation of the regression model is statistically significant over that of a mean model.(b) Calculate the co-efficient of determination (R 2) to assess how well the regression model explains the percentage of total error and thus how useful the regression line will be as a predictor.(c) Determine if the residual values for the regression equations are normally distributed.(d) If the regression model is found to be appropriate, calculate the y-intercept, the slope and their respective standard errors.The y-intercept represents the fitted mean thickness at time zero, the slope represents TDR 948 Rev. 0 Page 9 of 26 the corrosion rate, and the standard errors represent the uncertainty or random error of these two parameters.

(5) Use a z score of 2.58 and the standard deviation to establish a 99% confidence interval about the mean thickness values for each 6"x6" grid location to determine whether low thickness measurements or"outliers" are statistically significant.

If the data points are greater than the 99% lower confidence limit, then the difference between the value and the mean is deemed to be due to expected random error.However, if the data point is less than the lower 99%confidence limit, this implies that the difference is statistically significant and is probably not due to chance.2.5 Analysis of Two 6"1x6" Grid Data Sets Regression analysis is inappropriate when data is available at only two points in time. However, the t-Test can be used to determine if the means of the two data sets are statistically different.

2.5.1 Assumptions This analysis is based upon the following assumptionsi (1) The data in each data set is normally distributed.

(2) The variances of the two data sets are equal.2.5.2 Statistical Approach The evaluation takes place in three steps: (1) Perform a chi-squared test of each data set to ensure that the assumption of normality is valid at the 95%and 99% confidence levels.(2) Perform an F-test of the two data sets being compared to ensure that the assumption of equal variances is valid at the 95% and 99% confidence levels.(3) Perform a two-tailed t-Test for two independent samples to determine if the means of the two data sets are statistically different at the 0.05 and 0.01 levels of significance.

A conclusion that the means are not statistically different is interpreted to mean that significant corrosion did not occur over the time period represented by the data.However, if equality of the means is rejected, this implies that the difference is statistically significant and could be due to corrosion.

Ci I 9/C6 26 TDR 948 Rev. 0 Page 10 of 26 2.6 Analysis of Single 6"x6" Grid Data Set In those cases where a 6"x6" data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken in 1986 to identify the thinnest regions of the drywell shell in the sand bed region. For the most part, these are single point measurements which were taken in the vicinity of the 49-point data set, but not at the exact location.

Therefore, rigorous statistical analysis of these single data sets is impossible.

However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 49-point data set, the t-test can be used to compare the means of the two data sets as described in paragraph 2.5.When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed.2.6.1 Assumptions The comparison of a single 49-point data sets with previous data from the same vicinity is based on the following assumptions:

(1) Characterization of the scattering of data over the 6"x6" grid is such that the thickness measurements are normally distributed.

(2) Once the distribution of data for the 6"x6" grid is found to be normal, then the mean value of the thickness is the appropriate representation of the average condition.

(3) The prior data is representative of the condition at this location in 1986.2.6.2 Statistical Approach The evaluation takes place in four steps: (1) Perform a chi-squared test of each data set to ensure that the assumption of normality is valid at the 95%and 99% confidence levels.(2) Calculate the mean and the standard error of the mean of the 49-point data set.(3) Determine the two-tailed t value from a t distribution table at levels of significance of 0.05 and 0.01 for n-1 degrees of freedom.

TDR 948 Rev. 0 Page 11 of 26 (4) Use the t value and the standard error of the mean to calculate the 95% and 99% confidence intervals about the mean of the 49-point data set.(5) Compare the prior data point(s) with these confidence intervals about the mean of the 49-point data sets.If the prior data falls within the 95% confidence intervals, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If it falls within the 99%confidence limits but not within the 95% confidence limits, this implication is not as 'trong. In either case, the corrosion rate will be interpreted to be "Not Significant".

If the prior data falls above the upper 99% confidence limit, it could mean either of two things: (1) significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 49-point data set in 1986. There is no way to differentiate between the two.In this case, the corrosion rate will be interpreted to be"Possible".

If the prior data falls below the lower 99% confidence limit, it means that it is not representative of the condition at this location in 1986. In this case, the corrosion rate will be interpreted to be "Indeterminable".

2.7 Analysis of Single 7-Point Data Set In those cases where a 7-point data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken in 1986 to identify the thinnest regions of the drywell shell in the sand bed region. For the most part, these are single point measurements which were taken in the vicinity of the 7-point data sets, but not at the exact locations.

However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 7-point data set, the t-test can be used to compare the means of the two data sets as described in paragraph 2.5.When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed.2.7.1 Assumptions The comparison of a single 7-point data sets with previous data from the same vicinity is based on the following assumptions:

(1) The corrosion in the region of each 7-point data set is normally distributed.

TDR 948 Rev. 0 Page 12 of 26 (2) The prior data is representative of the condition at this location in 1986.The validity of these assumptions cannot be verified.2.7.2. Statistical Approach The evaluation takes place in four steps: (I) Calculate the mean and the standard error of the mean of the 7-point data set.(2) Determine the two-tailed t value using the t distribution tables at levels of significance of 0.05 and 0.01 for n-i degrees of freedom.(3) Use the t value and the standard error of the mean to calculate the 95% and 99% confidence intervals about the mean of the 7-point data set.(4) Compare the prior data point(s) with these confidence intervals about the mean of the 7-point data sets.If the prior data falls within the 95% confidence intervals, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If it falls within the 99%confidence limits but not within the 95% confidence limits, this implication Is not as strong. In either case, the corrosion rate will be interpreted to be "Not Significant".

If the prior data falls above the upper 99% confidence interval, it could mean either of two things: (1)significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 7-point data set in 1986. There is no way to differentiate between the two. In this case, the corrosion rate will be interpreted to be "Possible".

If the prior data falls below the lower 99% confidence limit, it means that it is not representative of the condition at this location in 1986. In this case, the corrosion rate will be interpreted to be "Indeterminable".

2.8 Evaluation of Drywell Mean Thickness This section defines the methods used to evaluate the drywell thickness at each location within the scope of the long term monitoring program.2.8.1 Evaluation of Mean Thickness Using Regression Analysis The following procedure is used to evaluate the drywell mean thickness at those locations where regression analysis has been deemed to be more appropriate than the mean model.

TDR 948 Rev. 0 Page 13 of 26 (1) The best estimate of the mean thickness at these locations is the point on the regression line corresponding to the time when the most recent set of measurements was taken. In the SAS Regression Analysis output (Ref. 3.7), this is the last value in the column labeled "PREDICT VALUE".(2) The best estimate of the standard error of the mean thickness is the standard error of the predicted value used above. In the SAS Regression Analysis output, this is the last value In the column labeled"STD ERR PREDICT".(3) The two-sided 95% confidence interval about the mean thickness is equal to the mean thickness plus or minus t times the estimated standard error of the mean. This is the interval for which we have 95%confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for equal tails at n-2 degrees of freedom and 0.05 level of significance, where n is the number of sets of measurements used in the regression analysis.

The degrees of freedom is equal to n-2 because two parameters (the y-intercept and the slope) are calculated in the regression analysis with n mean thicknesses as input.(4) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95%confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for one tail at n-2 degrees of freedom and 0.05 level of significance.

2.8.2 Evaluation of Mean Thickness Using Mean Model The following procedure is used to evaluate the drywall mean thickness at those locations where the mean model is deemed to be more appropriate than the linear regression model. This method is consistent with that used to evaluate the mean thickness using the regression model.(1) Calculate the mean of each set of UT thickness measurements.

(2) Sum the means of the sets and divide by the number of sets to calculate the grand mean. This is the best estimate of the mean thickness.

In the SAS Regression Analysis output (Ref. 3.7), this is the value labelled "DEP YMX"A.

9/GE 1c, : -1 E, : 3,A TDR 948 Rev. 0 Page 14 of 26 (3) Using the means of the sets from (1) as input, calculate the standard error. This is the best estimate of the standard error of the mean thickness.

(4) The two-sided 95% confidence interval about the mean thickness is equal to the mean thickness plus or minus t times the estimated standard error of the mean. This is the interval for which we have 95%confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for equal tails at n-1 degrees of freedom and 0.05 level of significance.

(5) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95%confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for one tail at n-I degrees of freedom and 0.05 level of significance.

2.8.3 Evaluation of Mean Thickness Using Single Data Set The following procedure is used to evaluate the drywell thickness at those locations where only one set of measurements is available.

(1) Calculate the mean of the set of UT thickness measurements.

This is the best estimate of the mean thickness.

(2) Calculate the standard error of the mean for the set of UT measurements.

This is the best estimate of the standard error of the mean thickness.

Confidence intervals about the mean thickness cannot be calculated with only one data set available.

3.0 REFERENCES

3.1 GPUW Safety Evaluation SE-000243-002, Rev. 0, "Drywell Steel Shell Plate Thickness Reduction at the Base Sand Cushion Entrenchment R'egion" 3.2 GPUN TDR 854, Rev. 0, "Drywall Corrosion Assessment" 3.3 GPUN TDR 851, Rev. 0, "Assessment of Oyster Creek Drywall Shell" 3.4 GPUN Installation Specification IS-328227-004, Rev. 3, "Functional Requirements for Drywall Containment Vessel Thickness Examination" TDR 94B Rev. 0 Page 15 of 26 3.5 Applied Regression Analysis, 2nd Edition, N.R. Draper & H. Smith, John Wiley & Sons, 1961 3.6 Statistical Concepts and Methods G.K. Bhattacharyya

& R.A. Johnson, John Wiley & sons, 1977 3.7 GPUN Calculation C-1302-187-5300-005, Statistical Analysis of Drywell Thickness Data Thru 12/31/86.4.0 EVALUATION OF DATA THROUGH 12/31/88 4.1 Results for 6"x6" Grids in Sand Bed Region at Original Locations 4.1.1 Bay 11A: 5/1187 to 10/6/88 Six 49-point data sets were available for this bay covering the time period from May 1, 1987 to October 8, 1988. Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) The current mean thickness

+ standard error is 908.6+5.0 mils.(4) There was no significant corrosion from May 1, 1987 to October 8, 1988.4.1.2 Bay l1C: 511/87 to 10/o/88 Five 49-point data sets were available for this bay covering the time period from May 1, 1987 to October 6, 1988. These data were analyzed as described in paragraphs 2.4 and 2.8.2. The Initial analysis of this data indicated that the data are not normally distributed.

The lack of normality was tentatively attributed to minimal corrosion in the upper half of the 6"x6" grid with more extensive corrosion in the lower half of the grid. To test this hypothesis, each data set was divided into two subsets, with one containing the top three rows and the other containing the bottom four rows.The top subset was normally distributed but the bottom subset was not. For both subsets, the mean model is more appropriate than the regression model.Since there is an observable decrease in the mean thickness with time, there appears to be some on-going corrosion at this location.

Further analysis is required.

TDR 948 Rev. 0 Page 16 of 26 The current mean thickness

+ standard error is 916.6 +10.4 mils for the lower subset and 1057.6 +16.9 mils for the upper subset.4.1.3 Bay 17D: 2/17/67 to 10/8/88 Six 49-point data sets were available for this bay covering the time period from February 17, 1987 to October 8, 1988.Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 84% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 864.8+6.8 mils.(6) The corrosion rate + standard error is -27.6 +6.1 mile per year.(7) The measurements below 800 mils were tested and determined not to be statistically different from the mean thickness.

4.1.4 Bay 19A: 2/17/87 to 10/8/s8 Six 49-point data sets were available for this bay covering the time period from February 17, 1987 to October 8, 1988.Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) The data are nearly normally distributed.

(2) The regression model is appropriate (3) The regression model explains B8% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 837.9+4.8 mils.(6) The corrosion rate + standard error is -23.7 +4.3 mpy.

I0/191C( TDR 948 Rev. 0 Page 17 of 26 (7) One data point that was below 800 mils at two different times was tested and determined to be statistically different from the mean thickness.

The probability of this occurring is less than 1% at each specific time.4.1.5 Bay 19B: 5/1/87 to 1018188 Five 49-point data sets were available for this bay covering the time period from May 1, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 99% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 856.5+0.5 mils.(6) The corrosion rate + standard error is -29.2 +0.5 mpy.(7) The measurements below 800 miils were tested and determined not to be statistically different from the mean thickness.

4.1.6 Bay 19C: 5/1187 to 20/8/88 Five 49-point data sets were available for this bay covering the time period from May 1, 1987 to October 8, 1988. Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 91% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 860.9+4.0 mils.(6) The corrosion rate + standard error is -25.9 +4.1 Mpy.--I-. ---- -. -

TDR 948 Rev. 0 Page 18 of 26 (7) The measurements below 800 mils were tested and determined not to be statistically different from the mean thickness.

4.2 Results for 6"x6" Grids in Sand Bed Region at New Locations 4.2.1 Bay 9D: 11/25/86 to 12119/88 The 6"x6" grid data was taken in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection system being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, previous measurements were taken in November 1986 in a 10-point 6"x6" cruciform pattern.Measurements were also taken in a 6"x5" grid in December 1986. The new data were compared with both of the previous data sets. These comparisons were made using the chi-squared test, F-test and two-tailed t-test as described in paragraph 2.5. The mean thickness was determined as described in paragraph 2.8.3.(1) The data are normally distributed.

(2) The variances are equal in both comparisons.

(3) It is appropriate to use the two-tailed t-test in both comparisons.

(4) The difference between the means of the 1988 49-point data set and the 1986 10-point data set is not significant.

However, there is a significant difference between the means of the 1988 49-point data set and the 1986 49-point data set. Therefore, significance of the corrosion rate is classified as"Indeterminable".

(5) The current mean thickness

+ standard error is 1021.4+9.7 mils.4.2.2 Bay 13Ai 11/25/86 to 12/17188 The 6"x6" grid data was taken for the first time in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, previous measurements were taken in November 1986 in abutting 6"x6" cruciform patterns across the entire bay. As a best approximation, 13 of these data points are at the same location as the new 6"x6" grid data set. Therefore, the new data were first compared with these 13 date points, and then with 21 data points which include the 13 plus 8 I I C/ !I*,/ -6. 1c;2E:33 TDR 948 Rev. I Page 19 of 26 additional points within one inch on either side. These comparisons were made using the chi-squared test, F-test and two-tailed t-test as described in paragraph 2.5. The mean thickness was determined as described in paragraph 2.8.3.(1) The data are normally distributed.

(2) The variances are equal in both comparisons.

(3) It is appropriate to use the two-tailed t-test in both comparisons.

(4) The difference between the means of the data sets is not signficant.

Therefore, the corrosion is classified as "Not Significant".

(5) The current mean thickness

+ standard error Is 905.3+10.1 mils.4.2.3 Bay 15D: 11/25/86 to 12/17/88 The 6"x6" grid data was taken for the first time in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, a previous 1-point measurement was taken In November 1986. The location of this point may have been somewhat removed from the location of the new 6"x6" grid data set. The previous measurement was compared with the new data set using the methods described in paragraph 2.6. The mean thickness was determined as described in paragraph 2.8.3.(1) The new data are normally distributed.

(2) The previous measurement falls above the 99% upper bound of the new data.(3) This implies that the corrosion may have occurred in the time period covered by this data. Therefore, the corrosion is classified as "Possible".

(4) The current mean thickness

+ standard error is 1056.0+9.1 mils.4.2.4 Bay 17A" 11/25/86 to 12/17/88 The 6"x6' grid data was taken for the first time in December 1988 during the 12R outage. This bay was considered for cathodic protection, but is not within the scope of the cathodic protection being installed.

The primary purpose of this data is to establish a base line to monitor corrosion in the future. However, a previous TDR 948 Rev. 0 Page 20 of 26 I-point measurement was taken in November 1986. The location of this point may have been somewhat removed from the location of the new 6"x6" grid data set. The previous measurement was compared with the new data set using the methods described in paragraph 2.6. The mean thickness was determined as described in paragraph 2.8.3.(1) The new data are not normally distributed.

However, the top three rows and the bottom four rows are each normally distributed.

(2) The previous measurement falls below the 99%confidence interval for the top three rows, and above the 99% confidence interval for the bottom four rows.(3) The corrosion is classified as "Indeterminable".

(4) The current mean thickness

+ standard error is 1133.1+6.9 milsfor the top three rows and 957.4 +9.2 mils for the bottom four rows.4.3 Results for 6"x6" Grids at Upper Elevations 4.3.1 Bay 5 51' Elevation:

11/01/87 to 10/8/88 Three 49-point data sets were available for this bay covering the time period from November 1, 1987 to ,October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.1.(1) Except for the first data set, the data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 99% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 750.0+0.02 mils.(6) The corrosion rate + standard error is -4.3+0.03 mpy.(7) One data point was determined to be statistically different from the mean thickness.

The probability of this occurring due to expected random error is less than 1% at each specific time.

101" ,9 ý)f.,,: 2,C: --TDR 948 Rev. 0 Page 21 of 26 4.3.2 Bay 9 87' Elevation:

11/6/87 to 10/8/88 Three 49-point data sets were available for this bay covering the time period from November 6, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is appropriate than the regression model.(3) There was no significant corrosion from November 6, 1987 to October 8, 1988.(4) The current mean thickness

+ standard error is 620.3+1.0 mile.4.3.3 Bay 13 87' Elevation:

11/10/87 to 10/8/88 Three 49-point data sets were available for this bay covering the time period from November 10, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) There was no significant corrosion from November 10, 1987 to October 8, 1988.(4) The current mean thickness

+ standard error is 635.6+0.7 mile.4.3.4 Bay 15 87' Elevation:

11/10/87 to 10/8/88 Three 49-point data sets were available for this bay covering the time period from November 10, 1987 to October 8, 1988. The data were analyzed as described in paragraphs 2.4 and 2.8.2.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) There was no significant corrosion from November 10, 1987 to October 8, 1988.(4) The current mean thickness

+ standard error is 634.8+0.7 mile.

19 J f. 6 : 2 iDR 948 Rev. 0 Page 22 of 26 4.4 Results for Multiple 6"x6" Grids in Trench 4.4.1 Bay 17D Trench: 12/9/86 to 12/23/88 The two sets of measurements in the Bay 17D Trench were taken on December 9, 1986 and December 23, 1988. The 1986 data is a 7 column by 36 row array. The 1988 data is a 7 column by 42 row array. The 1986 data is at the same elevation as the lower 36 rows of the 1988 data, but is centered about 3-/12 inches to the left of the 1988 data.To compare these two data sets, the 1986 data set and the lower 36 rows of the 1988 data set were each subdivided into six 7 column by 6 row subsets. Each pair of subsets was compared as described in paragraphs 2.5 and 2.8.3.Fourth Subset From The Top: The chi-squared statistic for the fourth subset from the top from the 1986 data set slightly exceeded the critical value for level of significance of 0.05, but was within the critical value for level of significance of 0.01. Also, the F statistic exceeded the critical value for levels of significance of 0.05 and 0.01. Therefore, it is inappropriate to apply the two-tailed t-test based on equal variances.

However, the approximate t-test based on unequal variances can be applied. From the results of this test, it is concluded that the difference between the mean thicknesses is not significant.

This implies that corrosion at this location was not significant.

All Other Subsets: (1) The data are normally distributed.

(2) The variances are equal.(3) Comparison of the means using the two-tailed t-test is appropriate.

(4) The difference between the means of the subsets was not significant.

This implies that there was no significant corrosion in the period from December 9, 1986 to December 23, 1988.(5) The current mean thickness

+ standard error of the top subset is 981.2 +6.7 mils. This is the thinnest area in the trench.

TDR 948 Rev. 0 Page 23 of 26 4.4.2 Bays 17/19 Frame Cutout: December 1988 Two sets of 6"x6" grid measurements were taken in December 1988. The upper one is located 25" below the top of the high curb and the other below the floor. There is no previous data. The upper location has been added to the long term monitoring program. With no prior data, the only possible analysis was to check the data sets for normality using the chi-squared test.The data at the upper location are not normally distributed.

The lack of normality was tentatively attributed to minimal corrosion in the lower half of the 6"x6" grid with more extensive corrosion in the upper half of the grid. To test this hypothesis, each data set was divided into two subsets, with one containing the top three rows and the other containing the bottom four rows. These subsets proved to be normally distributed, thus confirming the hypothesis.

The current mean thickness

+ standard error is 981.7 +4.4 mile for the top three rows and 1003.8+6.6 mile for the bottom four rows.The data at the location below the floor is normally distributed.

Also, the mean thickness is higher than at the upper location.

The mean thickness

+ standard error is 1034.1 +6.8 mils.4.5 Results for 6" Strips in Sand Bed Reqion 4.5.1 Bay 1D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls below the 99% lower bound of the new 7-point data set. Thus, the corrosion rate is class-ified as indeterminable.

The current mean thickness

+standard error is 1114.7 +30.6 mils.4.5.2 Bay 3D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls within the 99% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 1177.7 +5.6 mils.4.5.3 Bay 5D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The 1:2e:23 TDR 948 Rev. 0 Page 24 of 26 previous measurement falls within the 95% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness rate + standard error is 1174.0 +2.2 mils.4.5.4 Bay 7D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data was coupared as described in paragraph 2.7. The previous measurement falls just above the 99% upper bound of the new 7-point data set. This implies that corrosion has possibly occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 1135.1 +4.9 mils.4.5.5 Bay 9A: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1966. The data were compeaed as described in paragraph 2.7. The previous measurement falls below the 99% lower bound of the new 7-point data set. Thus, the corrosion rate is class-ified as indeterminable.

The current mean thickness

+standard error is 1154.6 +4.6 mile.4.5.6 Bay 13C: 11/25/86 to 12/17/8B The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls within the 95% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the tire period covered by the data. The current mean thickness

+ standard error is 1147.4 +3.7 mile.4.5.7 Bay 13D: 11/25/86 to 12/17/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. The data were compared as described in paragraph 2.7. The previous measurement falls within the 95% upper and lower bounds of the new 7-point data set. This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 962.1 +22.3 mile.4.5.8 Day ISA: 11/25/86 to 12/19/88 The 7-point data set was taken in December 1988 and a single point measurement was taken in November 1986. Also, a 6"x6" grid data set was taken on December 2, 1986 at this I0/191C6 TDR 948 Rev. 1 Page 25 of 26 location.

As a best approximation, the first 5 points in the 7-point data set are at the same location as points 38 to 42 of the 6"x6" grid. These five points all fall within the 99% confidence interval of the new 7-point data set.The single measurement falls below the 99% lower bound.This implies that significant corrosion has not occurred at this location in the time period covered by the data. The current mean thickness

+ standard error is 1120.0 +12.6 mils.4.6 Summary of Conclusions Bay & Area Location Corrosion Rate**Mean Thickness"'

4.6.1 6"x6" Grids in Sand Bed Region at Original Locations 1IA 11C 17D 19A 19B 19C Sand Sand Sand Sand Sand Sand Bed Bed Bed Bed Bed Bed Not significant Indeterminable

-27.6 +6.1 spy-23.7 74.3 mpy-29.2 +0.5 mpy-25.9 +4.1 mpy 4.6.2 6"x6" Grids in Sand Bed Region at New Locations 9D 13A 15D 17A Sand Sand Sand Sand Bed Bed Bed Bed Indeterminable*

Not significant*

Possible*Indeterminable*

908.6 916.6 864.8 837.9 856.5 860.9 1021.4 905.3 1056.0 957.4 750.0 620.3 635.6 634.8+5.0 mile+10.4 mile+6.8 mile+4.8 mile+0.5 mile+4.0 mile+9.7+10.1+9.1+9.2 mile mile mile mile 4.6.3 6"x6" Grids at Upper Elevations 5 9 13 15 51' Elev.87' Elev.87' Elev.87' Elev.-4.3 +0.03 mpy Not significant Not significant Not significant

+0.02 mile+1.0 mile+0.7 mile+0.7 mili 4.6.4 Multiple 6"x6" Grids in Trench 17D 17/19 Trench Frame Cutout Not significant*

Indeterminable*

981.2 +6.7 mile 981.7 +4.4 mile II/9(6i :63 TDR 948 Rev. I Page 26 of 26 4.6.5 6" Strips in Sand Bed Region ID 3D 5D 7D 9A 13C 13D 15A Sand Sand Sand Sand Sand Sand Sand Sand Bed Bed Bed Bed Bed Bed Bed Bed Indeterminable*

Not significant*

Not significant*

Possible*Indeterminable*

Not significant*

Not significant*

Not significant*

1114.7 +30.6 1177.7 +5.6 1174.0 +2.2 1135.1 +4.9 1154.6 ;4.8 1147.4 73.7 962.1 +22.3 1120.0 +12.6 mils mils milo milo mile mile mils milo 4.6.6 Evaluation of Individual Measurements Below 800 Mile One data point in Bay 19A and one data point in Bay 5 Elev. 51' fell outside the 99% confidence interval and thus are statistically different from the mean thickness.

• Based on limited data.*"Mean corrosion rate in"**Current mean thickness See text for interpretation.

mile per year + standard error of the mean in mile + standard error of the mean

[ J Nuclear Calculation Sheet SubjeclSTATISTICAL, ANALYSIS OF DRYWELL CRIcNO.THICKNESS DATA THRU 4-24-90 C-1302-7187-A1 Originator Date Reviewed 4 _/.,-?0 1.0 PROBLEM SLATEMENT The basic purpose of this calculation is to update the thickness measurement analyses documented in References 3.7, 3.8, and 3.11 by incorporating the measurements taken in Xarch and April 1990.Specific objectives of this calculation are: (1) Statistically analyze the thickness measurements in the sand bed region to determine the mean thickness and corrosion rate.(2) Analyze the data taken since the 12R outage for Bays 11A, 11C, 17D, 19A, 190, 19C, and the Frame Cutout between Bays 17 and 19 to determine if cathodic protection has reduced the corrosion rate.(3) Statistically analyze the thickness measurements for Bay 5 at elevation 51' and Bays 9, 13 and 15 at elevation 87' to determine the mean thickness and corrosion rate.(4) To the extent possible, analyze the data for the new locations at elevation 51' and elevation 52'.0 001/0004.1 M, flfIA dIfa, Calc. No. C-1302-187-5300-012 Rev. No. 0 Page 2 of 454 7\0 Or RESULTS 2.*1 "nd R~ed 11A 11C Top 11C Bottom 17D 19A 19B 19C Corrosion Rate **Mean Thickness

• • • F-Ratio B.I..... .aJk ~dLA* ~ -~1 1 I~4.mr.gaw.j.

ku Ergkxc-t&mn MA& ME--CA-1S.6 j2.9-35.2 +/-6.8-22.4 +/-4.3-19.0 +/-1.7-24.3+mipy Mipy Mipy Mipy mipy mpy MiPY 870.4 977.0 865.0 829.S 807.6 836.9 825.1+ 5.7+/-12.5+ 7.8+4.0* 3.0+3.2+ 2.3 mile mile mile mils mils mile mile 5.4 4.6 4.9 29.4 39.5 21.3 66.2 2.2 Sand Bed Reoion With\Cathodie aw-untim sinsm neu Neer RR 1lA 11C Top 11C Bottom 17D 9gA 198 19C Not Significan'tt Not Significant-, Not Significant*

-23.7 +/-4.6 mpy-20.6 13.9 mpy-11.8 +/-3.9 mpy-21.5 +/-3.5 mpy 878.0 996.6 878.1 830.1 808.2 841.2 826.3 4.4.+/-4.4.+4.S.9 8.3 5.6 3.8 3.2 3.3 2.9 mile mile mile mile mile mils mils 2.7 2.8 0.9 3.7 2.3 Sand Bed Reoion Frame Cutout 17/19 Top 17/19 Bottom Not Significant*

Not Significant*

986.0 4.4.7 mili 1008.4 ;"3.9 mili 2.4 Sand Bed Region Without Cathodig Protection 9D 13A 13D 15D 17A Top 17A Bottom Not Significant*

-39.1 +/- 3.4 mpy Indeterminate Not Significant*

Not Significant*

Not Significant*

1021.7 853.1 931.9 1056.5 1128.3 745.2+/-+ 8.9+ 2.4:t22.6+2.3+/- 2.2 12.1 mile mils mils Mills mile mile 16.9 1.3* Not statistically significant compared to random variations in measurements" Mean corrosion rate in mil per year +/- standard error of estimate***Best estimate of current mean thickness in mils +/- standard error of the mean 001/0004.2 Calc. No. C-1302-167-5300-011 Rev. No. 1 Page 24of 454 2.0

SUMMARY

Or RESWfTf Say & Crea Corrosion Rate Imvvl Mean Thicknees

• 'Best Estimate*

951 Conf,**2.1 Sand Bed Region With Cathodic Protection

-All Data* F-Ratio F 11A l1C Top 11C Bottom 17D 19A 19B 19C-15.6-35.2-22.4-25.0-21.4-19.0-24.3+/-2.9+6.8+/-4.3+/-2.0+1.5+/-1.7+1.3 mpy mpy mpy mpy mpy mpy Mpy-21.0-48.2-30.5-28.7-24.1-22.3-26.7 870.4 977.0 865.0 829.5 807.6 836.9 825.1+/- 5.7+/-12.5-+/- 7.8+/- 4.0+3.0+/- 3.2+2.3 mil.mile mile mile mils mile Milo 5.4 4.6 4.9 29.4 39.5 21.3 66.2 9 9 9 10 10 9 9 3.I 3.0 3.0 3.0 3.2 3.2 3.0 3.0 1.5 1.5 1.5 1.5 1.5 1.5 1.5 2.2 Sand Bed Recion with Cathodic Protection

-Since October 1988 IIA 11C Top 11C Bottom 17D 19A 19B 19C Not Significant****

Not Significant****

Not Significant****

-23.7 +4.6 mpy-20.6 +3.9 mpy-11.5 +3.9 mpy-21.5 -+/-3.5 mpy-34.2-29.7-21.1-29.5 878.0 996.6 878.1 830.1 808.2 841.2 826.3++4++++5.9 8.3 5.6 3.8 3.2 3.3 2.9 mile mile mile mile mile mile mile 2.7 2.8 0.9 3.7 5 5 5 5 5 5 5 2.3 Sand Bed Region Frame Cutout 17/19 Top Not Significant-**

17/19 Bottom Not Significant****

986.0 +/- 4.7 mile 1005.7 +/- 5.6 mile 5 5 1.3 1.3 1.2.4 Sand Bed Region Without CathoQic Protection 9D 13A 13D 15D 17A Top 17A Bottom Not Significant*"*

-39.1 + 3.4 mpy Indeterminate Not Significant****

Not Significant"***

Not Significant****

-46.4 1021.7 853.1 931.9 1056.5 1128.3 950.8-+ 8.9+ 2.4+/-22.6+ 2.3+/-2.2+/-5.3 milo mile mile mile mile Milo 5 16.9 6 1 5 5 5 1.3 1.4 0 1.5 1.4 1.4 11* Mean corrosion rate in mil. per year +/- standard error of estimate** Upper bound of the one-sided 95% confidence interval*' Best estimate of current mean thickness in mile + standard error of the mean****Not statistically significant compared to random variations in measurements N w Number of data sets Yrs -Years from first to last data set 001/0004.3 DOCUMENTf NO.[ ] Nuclear (M I~ucear I C-1302-187-5300-011 TITLE STATISTICAL ANALYSIS OF DRYWELL THICKNESS THRU 4-24-90 REV

SUMMARY

OF CHANGE APPROVAL DATE I Computed 95% upper bound of the corrosion rate in each bay where regression model is appropriate.

veJ %-ai i V 130 2- 117-0 Computed maximum potential corrosion rate at 9 4-95% confidence for each bay where mean model is appropriate.

Deleted Summary of Apparent Corrosion Rates and added Summary of Xaximum Potential Corrosion Rates at 95% Confidence.

Revised paragraphs 2.0, 4.5.2, and 4.10 to reflect these changes.C. a-Y ec~4C4 +% pas es Suy"Mal f-eA la N0036 (03-90) 2O/22/o.~

Z~~2:(7 ~) ~92yjpci LJ'*Cale. No. C-1302-187-5300-011 Rev. No. 0 Page 3 of 454 Mean Thickness

• • • E-Ratio e Corrosion RAt 2.5 Eevat o S51 5/D-22 -4.6 +/-: 1.6 5/5 Indeterminate 13/31 Indetexminate 15/23 determinate 2.6 Elevation 52' ", N a **745.2 745.1 750 .8 751.2+/- 2.1+/-3.2+/-11.5+/- 3.6 mile mile mile mile 1.3 7/2S 13/6 13/32 19/13 indeterminate Indetermin~te indeterminate Indeterminat6 715.5 724.9 698.3 712.5+/- 2.9+/- 2.9+ 3.1 2.7 Elevation 87" 9 13 15 Not Significant*

Not Significant*

Not Significant*

619.9 636.5 636.2*_0.6+ 0.8+ 1.1 2.5 Apparent Corrosion Rates Theme estimates of the corrosion rate are based on a least squares fit of the data. In those cases where the F-Ratio is less than 1.0 they should not be used to make future projections.

For bays with cathodic protection, these apparent rates are for the period from October 1988 to April 1990. For the other bays, it is for all data.Apparent Corrosion 2L~e (Mai FrE-aig I" 11A lC Top 11C Bottom 17D 19A 19B 19C 17/19 Top 17/19 Bottom-16.2-2S.0-16.7-23.7-20.6-11.8-21.5-8.2-13.1+/- 8.6+/-10.6-+7.1* 4.6+/- 3.9+/- 3.9+/- 3.5+/-:10.7+/-11.6 0.2 0.6 0.6 2.7 2.8 0.9 3.7 0.1 0.1 9D 13A 15D 17A Top 17A Bottom 5 EL 51'9 EL 87'13 EL 87'15 EL 87'Appaxent Corrosion Rate (m1DyV 1-21.0 t18.1-39.1 +/- 3.4-4.6 +/-4.8-6.8 3.7-17.7 +/- 7.6-4.6 +/- 1.6-0.2 +/- 0.9 zero zero 0.1 16.9 0.1 0.3 0.01 1.3 zero 001/0004.3 Calc. No. C-1302-187-5300-011 Rev. No. 1 Page 34of 454 Bay & Area Qorrosion Pate jMyVj Mean Thickness

"'95% Conf,**F-Ratio Ff Beat Estimate*2.5 Elevation 51'5/D-12 5/5 13/31 15/23-4.6 :t 1.6 mpy Indeterminate Indeterminate Indeterminate

-2.2 745.2 745.1 750.8 751.2 2.6 Elevation 52'7/25 Indeterminate 13/6 Indeterminate 13/32 Indeterminate 19/13 Indeterminate

--2.1+/- 3.2+/-11.5+/- 3.8+/- 2.9+/- 2.9+/-_5.0+/-3.1 mile 1.3 mile mile milo 6 2 2 2 2.5 1.1 1.1 1.1 715.5 724.9 698.3 712.5 619.9 636.5 636.2 mile mils mile mile 1 1 1 1 5 5 5 0 0 0 0 2.7 Elevation 67" 9 13 15 Not Significant****

Not Significant****

Not Significant****

+/-t 0.6 mile+/- 0.8 mile+/-t 1.1 mile 2.4 2.4 2.4 2.8 Potential Corrosion Rates at 95% Confidence For those locations where the corrosion rate is not statistically significant, the possibility does exist that the variability in the data may be masking an actual corrosion rate. The potentially masked corrosion rate at 95% confidence is bounded by the upper bound of the 95% one-sided confidence interval about the slope computed in the regression analysis (see Paragraph 4.10.1).95% Upper Bound Corrosion Rate nElevation MPY) H XUi 11A (Since 10/88)11C Top (Since 10/88)11C Bottom (Since 10/8)17/19 Top 17/19 Bottom 9D 15D 17A Top 17A Bottom 9 13 15 NK s The high value for value on 6/26/89.is -29.2 mpy.Sand Bed Sand Bed Sand Bed Frame Cutout Frame Cutout Sand Bed Sand Bed Sand Bed Sand Bed 87'87'87'-36.4-49.9-33.3-33.4-40.5-63.4-16.0-15.5-35.6-2.2-2.1-0.6 5 5 5 5 5 5 S 5 5 5 5 5 1.5 1.5 1.5 1.3 1.3 1.3 1.4 1.4 1.4 2.4 2.4 2.4 Bay 9D results from one extremely high mean Without this data point, the 95% upper bound 001/0004.4 1II/i1l6C 1-2.32:2f Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 4 of 454'In 2 .)V Evaluation of individual Measurements Exceedina 99%199% Tolerance Interval One data point in Bay 5 Elev. 51, fell outside the 99%/99% tolerance interval and thus is statistically different from the mean thickness.

Based on a linear regression analysis for this point, it is concluded that the corrosion rate in this pit is essentially the same as the overall grid.001/0004.4 Calc. No. C-1302-187-5300-Ol Rev. No. 0 Page 5 of 454 3.0 IEFEMCES 3.1 GPUN Safety Evaluation SE-000243-002, Rev. 0, "Drywell Steel Shell Plate Thickness Reduction at the Base Sand Cushion Entrenchment Region" 3.2 GPUN TDR 854, Rev. 0, "Drywell Corrosion Assessment" 3.3 GPUN TDR 851, Rev. 0, "Assessment of Oyster Creek Drywell Shell" 3.4 GPUN Installation Specification IS-32e227-004, Rev. 3, "Functional Requirements for Drywell Containment Vessel Thickness Examination" 3.5 Applied Regression Analysis, 2nd Edition, N.R. Draper & H. Smith, John Wiley & Sons, 1981 3.6 Statistical Concepts and Methods, G.K. Shattacharyya

& R.A.Johnson, John Wiley & sons, 1977 3.7 GPUN Calculation C-1302-187-5300-005, Rev. 0, "Statistical Analysis of Drywell Thickness Data Thru 12-31-88" 3.8 GPUN TDR 948, Rev. 1, "Statistical Analysis of Drywell Thickness Data" 3.9 Experimental Statistics, Mary Gibbons Natrella, John Wiley & Sons, 1966 Reprint. (National Bureau of Standards Handbook 91)3.10 Fundamental Concepts in the Design of Experiments, Charles C.Hicks, Saunders College Publishing, Fort Worth, 1982 3.11 GPUN Calculation C-1302-187-5300-008, Rev. 0, "Statistical Analysis of Drywell Thickness Data thru 2-8-90, 001/0004.5 10/12/06 12:32:2E Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 6 of 454 4.0 ASSUMPTIONS

& BASIC DATA 4.1 Background The design of the carbon steel drywall includes a sand bed which is located around the outside circumference between elevations 8'-11-114" and 121-3". Leakage was observed from the sand bed drains during the 1980, 1983 and 1986 refueling outages indicating that water had intruded into the annular region between the drywell shell and the concrete shield wall.The drywell shell was inspected in 1986 during the 1OR outage to determine if corrosion was occurring.

The inspection methods, results and conclusions are documented in Ref. 3.1, 3.2, and 3.3.As a result of these inspections it was concluded that a long term monitoring program would be established.

This program includes repetitive Ultrasonic Thickness (UT) measurements in the sand bed region at a nominal elevation of 111-3" in bays 11A, 1IC. 17D, 19A, 19B, and 19C.The continued presence of water in the sand bed raised concerns of potential corrosion at higher elevations.

Therefore, UT measurements were taken at the 51' and 87" elevations in November 1987 during the 11R outage. As a result of these inspections, repetitive measurements in Bay 5 at elevation 51' and in Bays 9, 13 and 15 at the 87' elevation were added to the long term monitoring program to confirm that corrosion In not occurring at these higher elevations.

A cathodic protection system was installed in selected regions of the sand bed during the 12R outage to minimize corrosion of the drywell. The cathodic protection system was placed in service on January 31, 1989. The long term monitoring program was also expanded during the 12R outage to include measurements in the sand bed region of Bays ID, 3D, 5D, 7D, 9A, 13A, 13C, 13D, 15A, 15D and 17A which are not covered by the cathodic protection system. It also includes measurements in the sand bed region between Says 17 and 19 which is covered by the cathodic protection system, but does not have a reference electrode to monitor its effectiveness in this region.The high corrosion rate computed for Bay 13A in the sand bed region through February 1990 (Ref. 3.11) raised concerns about the corrosion rate in the sand bed region of Say 13D. Therefore, the monitoring of this location using a 6"x6" grid was added to the long term monitoring program. In addition, a 2-inch core sample was removed in March 1990 from a location adjacent to the 6"x6" monitored grid in Bay 13A.001/0004.6 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 7 of 454 Measurements taken in Bay 5 Area D-12 at elevation 511 through March 1990 indicated that corrosion is occurring at his location.Therefore, survey measurements were taken to determine the thinnest locations at elevation S1V. An a result, three new locations were added to the long term monitoring program (Bay 5 Area 5, Bay 13 Area 31, and Say 15 Area 213).The indication of ongoing corrosion at elevation 51' raised concerns about potential corrosion of the platen immediately above which have a smaller nominal thickness.

Therefore, survey measurements were taken in April 1990 at the 52' elevation in all bays to determine the thinnest locations.

As a result of this survey, four new locations were added to the long term monitoring plan at elevation S21 (Bay 7 ar~ea 25, Say 13 Area 6, Bay 13 Area 32, and Bay 19 Area 13).Some measurements in the long term monitoring program are to be taken at each outage of opportunity, while others are taken during each refueling outage. The functional requirements for these inspections are documented in Ref. 3.4. The purpose of the UT measurements is to determine the corrosion rate and monitor it over time, and to monitor the effectiveness of the cathodic protection system.4.2 Selection of -Areas to be Monitored A program was Initiated during the 11R outage to characterize the corrosion and to determine its extent. The details of this inspection program are documented in Ref. 3.3. The greatest corrosion was found via UT measurements in the sand bed region at the lowest accessible locations.

Where thinning was detected, additional measurements were made in a cross pattern at the thinnest section to determine the extent in the vertical and horizontal directions.

Having found the thinnest locations, measurements were made over a 6"x6" grid.To determine the vertical profile of the thinning, a trench was excavated into the floor in Bay 17 and Bay 5. Bay 17 was selected since the extent of thinning at the floor level was greatest in that area. It was determined that the thinning below the top of the curb was no more severe than above the curb, and became less severe at the lower portions of the sand cushion. Bay 5 was excavated to determine if the thinning line was lower than the floor level in areas where no thinning was detected above the floor. There were no significant indications of thinning in Bay 5.001/0004.7 10/21/rf 12:32Z26 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 8 of 454 It was on the basis of these findings that the 6"x6" grids in Bays 11A, 1iC, 17D, 19A, 19B and 19C were selected as representative locations for longer term monitoring.

The initial measurements at these locations were taken in December 1986 without a template or markings to identify the location of each measurement.

Subsequently, the location of the 6"x6" grids were permanently marked on the drywell shell and a template in used in conjunction with these markings to locate the UT probe for successive measurements.

Analyses have shown that including the non-template data in the data base creates a significant variability in the thickness data. Therefore, to minimize the effects of probe location, only those data sets taken with the template are included in the analyses.The presence of water in the sand bed also raised concern of potential corrosion at higher elevations.

Therefore, UT measurements were taken at the 51, and 87' elevations in 1987 during the 11M outage. The measurements were taken in a band on 6-inch centers at all accessible regions at these elevations.

Where these measurements indicated potential corrosion, the measurements spacing was reduced to 1-inch on centers. If these additional readings indicated potential corrosion, measurements were taken on a 6"x6" grid using the template.

It was on the basis of these inspections that the 6"x6" grids in Say 5 at elevation 51'and in bays 9, 13 and 15 at the 87' elevation were selected as representative locations for long term monitoring.

A cathodic protection system was installed in the sand bed region of Bays 11A, 12C, 27D, 19A, 19B, 19C, and at the frame between Bays 17 and 19 during the 12R outage. The system was placed in service on January 31, 1989.The long term monitoring program was expanded as follows during the 12R outages (1) Measurements on 6"x6" grids in the sand bed region of Bays 9D, 13A, lSD and 17A. The basis for selecting these locations is that they were originally considered for cathodic protection but are not included in the system being installed.

(2) Measurements on 1-inch centers along a 6-inch horizontal strip in the sand bed region of Bays ID, 3D, 5D, 7D, 9A, 13C, and 15A. These locations were selected on the basis that they are representative of regions which have experienced nominal corrosion and are not within the scope of the cathodic protection system.001/0004.8 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 9 of 454 (3) A 6"x6" grid in the curb cutout between Bays 17 and 19. The purpose of these measurements is to monitor corrosion in this region which is covered by the cathodic protection system but does not have a reference electrode to monitor its performance.

The long term monitoring program was expanded in March 1990 as follows: (1) Measurements in the sand bed region of Bay 13D: This location was added due to the high indicated corrosion rate in the sand bed region of Bay 13A. The measurements taken in March 1990 were taken on a 2'x6" grid. All subsequent measurements are to be taken on a 6"x6" grid.(2) Measurements on 6"x6" grids at the following locations at elevation 5l't Bay 5 Area 5, Bay 13 Area 31, and Bay 15 Area 2/3. These locations were added due to the indication of ongoing corrosion at elevation 51', Bay 5 Area D-1.The long term monitoring program was expanded in April 1990 by adding the following locations at elevation 52': Bay 7 Area 25, Bay 13 Area 6, Bay 13 Area 32, and Bay 19 Area 13. All measurements are taken on 6"x6" grids. These locations were added due to the indication of ongoing corrosion at elevation 51' and the fact that the nominal plate thickness at elevation 52' iu less than at elevation 51'.4.3 UT Measurements The UT measurements within the scope of the long term monitoring program are performed in accordance with Ref. 3.4. This involves taking UT measurements using a template with 49 holes laid out on a 6"x6" grid with l* between centers on both axes. The center row is used in those bays where only 7 measurements are made along a 6-inch horizontal strip.The first set of measurements were made in December 1986 without the use of a template.

Ref. 3.4 specifies that for all subsequent readings, QA shall verify that locations of UT measurements performed are within 11/4" of the location of the 1986 UT measurements.

It also specifies that all subsequent measurements are to be within +/- 1/8' of the designated locations.

001/0004.9 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 10 of 454 4.4 Data at Plug Locations Seven core samples, each approximately two inches in diameter were removed from the drywell vessel shell. These samples were evaluated in Ref. 3.2. Five of these samples were removed within the 6"x6" grids for Bays 11A, 17D, 29A, 19C and Bay 5 at elevation 51'. These locations were repaired by welding a plug in each hole. Since these plugs are not representative of the drywell shell, UT measurements at these locations on the 6"x6" grid must be dropped from each data set.The following specific grid points have been deleted%May Area Points 11A 23, 24, 30, 31 17D 15, 16, 22, 23 19A 24, 25, 31, 32 19C 20, 26, 27, 33, 5 EL 51' 13, 20, 25, 26, 27, 28, 33, 34, 35 The core sample removed in the sand bed region of Bay 13A was not within the monitored 6"x6" grid.4.5 Bases for Statistical Analysis of §"x6' Grid Data 4.5.1 Aesumptions The statistical evaluation of the UT measurement data to determine the corrosion rate at each location is based on the following assumptionsz (1) Characterization of the scattering of data over each 6"x6" grid is such that the thickness measurements are normally distributed.

(2) Once the distribution of data for each 6"x6" grid is found to be normal, then the mean value of the thickness is the appropriate representation of the average condition.

(3) A decrease in the mean value of the thickness with time is representative of the corrosion occurring within the 6"x6" grid.001/0004.10 1012" irllý "".': K : 26 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 11 of 454 (4) If corrosion has ceased, the mean value of the thickness will not vary with time except for random errors in the UT measurements.

(5) If corrosion is continuing at a constant rate, the mean thickness will decrease linearly with time. In this case, linear regression analysis can be used to fit the mean thickness values for a given zone to a straight line as a function of time. The corrosion rate is equal to the slope of the line.The validity of these assumptions is assured by;(a) Using more than 30 data points per 6"x6" grid (b) Testing the data for normality at each G6x6" grid location.(c) Testing the regression equation as an appropriate model to describe the corrosion rate.These tests are discussed in the following section. In cases where one or more of these assumptions proves to be invalid, non-parametric analytical techniques can be used to evaluate the data.4.5.2 Statistical Approach The following steps are performed to test and evaluate the UT measurement data for those locations where 6*x6" grid data has been taken at least three times: (1) Edit each 49-point data set by setting all invalid points to zero. Invalid points are those which are declared invalid by the UT operator or are at a plug location. (The computer programs used in the following steps ignore all zero thickness data points.)(2) Perform a Chi-squared goodness of fit test of each 49 point data set to ensure that the assumption of normality is valid at the 5% and 1% level of significance.

(3) Calculate the mean thickness and variance of each 49 point data set.(4) Perform an Analysis of Variance (ANOVA) F-test to determine if there is a significant difference between the means of the data sets.001/0004.11

ý011.'!315 12:32:2f: EF U-4 E-4 4) A 41V 0 >424 -4 90 r 41.4 E d1 14 M4 PC6 X 4 41 WE4 43V2-4 W -40 420 Ob 4V X 0 C0 1240 00 0 40 W~ U 0 ýCalc. No. C-1302-187-5300-011 Rev. No. ,Jr Page 12 of 454 (5)Using the mean thickness values for each 6"x6* grid, perform linear regression analysis over time at each location.0 x!4 X -.4 C40.0-4 41 415 42.V -A 4 0 .4 41 424.C -4 26 0 U 041 015541 IV0 0 -4 42 W 4 _.AV 4 4 0 A -a 0'4 4 A 0-4 0 0 A 4.. C 4j 4j 41 2:4 C28-4. 0o .5.5 -d $42 -5414 Q 4 00 5420 U 43 .(a) Perform F-test for significance of regression at the 5% level of significance.

The result of this test indicates whether or not the regression model is more appropriate than the mean model. In other words, it tests to see if the variation of the regression model is statistically significant over that of a mean model.(b) Calculate the ratio of the observed F value to the critical F value at 5% level of significance.

For data sets where the Residual Degrees of Freedom in ANOVA is 4 to 9, this 4f O F-Ratio should be at least 8 for the regression rd.hAo to be considered opposed to w-Sa1-T-"significant." _ --- p -0 i , " 1 (c) Calculate the coefficient of determination (R ) to assess how well the regression model a explains the percentage of total error and thus 4 how useful the regression line will be as a 4 -a, predictor.

541 42 4 (d) Determine if the residual values for the regression equations are normally distributed.

1 5 c v ,.lt.4 > 0 -4 (e) If the regression model is found to be jA, & 0 appropriate, calculate the y-intercept, the !W C U 0.0 slope and their respective standard errors. lo o $The y-intercept represents the fitted mean V 41 0 4 C4 IV4 0a q-4 thickness at time zero, the elope represents M 0 the corrosion rate, and the standard errors V 0 4 represent the uncertainty or random error of 0 0 .a these two parameters.

0 0^ O Op 4j Q.-.4 $4 (6) Use a K factor from Table A-7 of Reference 3.9 and h 0 the standard deviation to establish a one-sided 42 99%/99% tolerance limit about the mean thickness " 0 K values for each 6"x6" grid location to determine 0 V r, whether low thickness measurements or "outliers" are .V.4 statistically significant.

If the data points are A .greater than the 99%/99% lower tolerance limit, then 04 a the difference between the value and the mean is deemed to be due to expected random error. However, if the data point is less than the lower 99%/99%tolerance limit, this implies that the difference is statistically significant and is probably not due to chance.C 001/0004.12

2

32:2E Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 13 of 454 4.6 Analysis of Two 6"X6" Grid Data Sets Regression analysis is inappropriate when data is available at only two points in time. However, the t-test can be used to determine if the means of the two data sets are statistically different.

4.6.1 Aso.mptione This analysis is based upon the following assumptions:

(1) The data in each data set is normally distributed.

(2) The variances of the two data sets are equal.4.6.2 Statistical Aovroach The evaluation takes place in three stepst (1) Perform a chi-squared test of each data set at 5% and 2% levels of significance to ensure that the assumption of normality is valid.(2) Perform an F-test at 5% and 1% level of significance of the two data sets being compared to ensure that the assumption of equal variances is valid.(3) Perform a two-tailed t-test for two independent samples at the 5% and 1% levels of significance to determine if the means of the two data sets are statistically different.

A conclusion that the means are not statistically different is interpreted to mean that significant corrosion did not occur over the time period represented by the data.However, if equality of the means is rejected, this implies that the difference is statistically significant and could be due to corrosion.

4.7 Analysis of Single 6"x6" Grid Data Set In those cases where a 6"x6" data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken at an earlier time. For the most part, these are single point measurements which were taken in the vicinity of the 49-point data set, but not at the exact location.

Therefore, rigorous statistical analysis of these mingle data sets is impossible.

However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 49-point data set, the t-test can be used to compare the means of the two data sets as described in paragraph 4.5.001/0004.13 i2~'~f~:2 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 14 of 454 When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed.4.7.1 Assumptions The comparison of a single 49-point data sets with previous data from the same vicinity is based on the following assumptions:

(1) Characterization of the scattering of data over the 6"x6" grid Is such that the thickness measurements are normally distributed.

(2) Once the distribution of data for the 6"x6" grid is found to be normal, then the mean value of the thickness is the appropriate representation of the average condition.

(3) The prior data is representative of the condition at this location at the earlier date.4.7.2 Statistical Approach The evaluation takes place in four stepst (1) Perform a chi-squared test of each data set to ensure that the assumption of normality is valid at the 95%and 99% confidence levels.(2) Calculate the mean and the standard error of the mean of the 49-point data set.(3) Determine the two-tailed t value from a t distribution table at levels of significance of 0.05 and 0.01 for n-l degrees of freedom.(4) Use the t value and the standard error of the mean to calculate the 95% and 99% confidence intervals about the mean of the 49-point data set.(5) Compare the prior data point(s) with these confidence intervals about the mean of the 49-point data sets.If the prior data falls within the 95% confidence intervals, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If it falls within the 99%confidence limits but not within the 95% confidence limits, this implication is not as strong. In either case, the corrosion rate will be interpreted to be "Not Significant".

001/0004.14

'Cý'C110'

'2:32t2t Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 15 of 454 If the prior data falls above the upper 99% confidence limit, it could mean either of two things: (1) significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 49-point data set in 1986. There is no way to differentiate between the two.In this case, the corrosion rate will be interpreted to be"Possible".

If the prior data falls below the lower 99% confidence limit, it means that it is not representative of the condition at this location at the earlier date. In this case, the corrosion rate will be interpreted to be"Indeterminable".

4.8 Analysis of Single 7-Point Data Set In those cases where a 7-point data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken at an earlier time to identify the thinnest regions of the drywell shell in the sand bed region. For the most part, these are single point measurements which were taken in the vicinity of the 7-point data sets, but not at the exact locations.

However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 7-point data set, the t-test can be used to compare the means of the two data sets as described in paragraph 4.5.When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed.4.8.1 Aseumptions The comparison of a single 7-point data sets with previous data from the same vicinity is based on the following assumptionst (1) The corrosion in the region of each 7-point data set is normally distributed.

(2) The prior data is representative of the condition at this location at the earlier date.The validity of these assumptions cannot be verified.001/0004.15 I " 2, f ".iý 2: :2E Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 16 of 454 4.8.2. Statistical Approach The evaluation takes place in four steps: (1) Calculate the mean and the standard error of the mean of the 7-point data set.(2) Determine the two-tailed t value using the t distribution tables at levels of significance of 0.05 and 0.01 for n-1 degrees of freedom.(3) Use the t value and the standard error of the mean to calculate the 95% and 99% confidence intervals about the mean of the 7-point data set.(4) Compare the prior data point(s) with these confidence intervals about the mean of the 7-point data sets.If the prior data falls within the 95% confidence intervals, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If it falls within the 99%confidence limits but not within the 95% confidence limits, this implication is not as strong. In either case, the corrosion rate will be interpreted to be "Not Significant*.

If the prior data falls above the upper 99% confidence interval, it could mean either of two things: (1)significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 7-point data set in 1986. There is no way to differentiate between the two. In this case, the corrosion rate will be interpreted to be "Possible".

If the prior data falls below the lower 99% confidence limit, it means that it is not representative of the condition at this location at the earlier date. In this case, the corrosion rate will be interpreted to be"Indeterminable".

4.9 Evaluation of Drywell Mean Thickness This section defines the methods used to evaluate the drywell thickness at each location within the scope of the long term monitoring program.001/0004.16 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 17 of 454 4.9.1 EvAluation of Mean Thickness Using Regression Analysis The following procedure is used to evaluate the drywell mean thickness at those locations where regression analysis has been deemed to be more appropriate than the mean model.(1) The best estimate of the mean thickness at these locations is the point on the regression line corresponding to the time when the most recent set of measurements was taken. In the SAS Regression Analysis output (App. 6.2), this is the last value in the column labeled "PREDICT VALUE".(2) The best estimate of the standard error of the mean thickness is the standard error of the predicted value used above. In the SAS Regression Analysis output, this is the last value in the column labeled"STD ERR PREDICT".(3) The two-sided 95% confidence interval about the mean thickness is equal to the mean thickness plus or minus t times the estimated standard error of the mean. This is the interval for which we have 95%confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for tails at n-2 degrees of freedom and 0.05 level of significance, where n is the number of sets of measurements used in the regression analysis.

The degrees of freedom is equal to n-2 because two parameters (the y-intercept and the slope) are calculated in the regression analysis with n mean thicknesses as input.(4) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95%confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for one at n-2 degrees of freedom and 0.05 level of significance.

4.9.2 Evaluation of Mean Thickness Uuing Mean Model The following procedure is used to evaluate the drywell mean thickness at those locations where the mean model is deemed to be more appropriate than the linear regression model. This method is consistent with that used to evaluate the mean thickness using the regression model.001/0004.17 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 18 of 454 (1) Calculate the mean of each set of UT thickness measurements.

(2) Sum the means of the sets and divide by the number of sets to calculate the grand mean. This is the best estimate of the mean thickness.

In the SAS Regression Analysis output, this is the value labelled "DEP MEAN".(3) Using the means of the sets from 11) as input, calculate the standard about t__e mean. This is the best estimate of the standard error of the mean thickness.

(4) The two-sided 95% confidence interval about the mean thickness is equal to the mean thickness plus or minus t times the estimated standard error of the mean. This is the interval for which we have 95%confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for eaual tails at n-i degrees of freedom and 0.05 level of significance.

(5) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95%confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for one tail at n-l degrees of freedom and 0.05 level of significance.

4.9.3 Evaluation of Moan Thickness Using Single Data Set The following procedure is used to evaluate the drywell thickness at those locations where only one set of measurements is available.

(1) Calculate the mean of the set of UT thickness measurements.

This is the best estimate of the mean thickness.

(2) Calculate the standard error of the mean for the set of UT measurements.

This is the best estimate of the standard error of the mean thickness.

Confidence intervals about the mean thickness cannot be calculated with only one data set available.

001/0004.18 0-4 4 0 V V U E 0 0 0, 0 in 41 -4~ ~ 1 0% £40 M4£ 4 Ui4J 0J 0.MO~ C 0 OVA 0 X- 3 4) W -4 0.1-4 4j 0 44 a 0 M4q Ko *00 0£C 01 1 000 0 0 a-4j~ A '0 .1.4 eft 04 &14*-41 0 0 z .4 4)A 14 H 0.0Id4'.-4 *.£W0oC 34 c 41 c 0. 00 4) 0 0 4 0 00.40£ 4j 00 0 A A.44 C-.1 X v 4 -4 00 i 4( 104104A %0U a4 C6 C 6 JC IV 00 0 4'J 0 C% E 0I-4C S at W t ID 0V '0'.0 .-4 4.)O -4 M V 3 140Ev 4.19 iCalc. No. C-1302-187-5300-011 Rev. No.f.245 Page 19 of 454 valuation of Drvwell Corrosion Rate 4.20.1 Mean Model If the ratio of the observed F value to the critical F value is les than 1 for the F-test for the significance of regression, it indicates that the mean model is more appropriate than the regression model at the 5% level of significance.

In other words, the variation in mean thickness with time can be explained solely by the random variations in the measurements.

This means that the corrosion rate is not significant compared to the random variations.

I this case, an F-test is performed to compare the vari ility of the data set means between data sets with the var*ility of individual measurements within the data sets. if C observed F value is less than the critical F value, it con that the ean model is appro~riate.

If the F-test indica tha e means is significant, the Lea Significant Difference (LSD) is computed.

This is the max m difference between data set mean thicknesses that can be aributed to random variation in the measurements.

If the dif nce between the means of data sets exceeds LSD, it indicat that difference is significant.

The difference between me is subtracted from LSD and the result is divided by the t between measurements to estimate the Significant Corr !on Rate" in mils per year (mpy). If the difference betwee the means does not exceed LSD, then it is concluded that significant corrosion occurred during that period of ti 4.10.2 Regression Model If the ratio of the observed F value to the critical F value is 1 or greater, it indicates that the regression model is more appropriate than the mean model at the 5%level of significance.

In other words, the variation in mean thickness with time cannot be explained solely by the random variations in the measurements.

This means that the corrosion rate is significant compared to the random variations.

Although a ratio of 1 or greater indicates that regression is significant, it does not mean that the slope of the regression line is an accurate prediction of the corrosion rate. The ratio should be at least 4 or 5 to consider the slope to be a useful predictor of the corrosion rate (Ref.

Calc. No. C-1302-187-5300-011 Rev. No. Page 20 of 454 3.5, pp. 93, 129-133).

A ratio of 4 or 5 means that the variation from the mean due to regression is approximately twice the standard deviation of the residuals of the regression.

To have a high degree of confidence in the predicted corrosion rate, the ratio should be at least 8 or 9 (Ref.3.5, pp. 129-133).In t instances, four sets of measurements over a per~kid'of abou ne year do not provide a significant re sion model which be used to predict future thi ease,.However, a least area fit of the four a points does easonable imate of the ent corrosion rate. arts i tin rl aluable for assigteefciees ahdcprotection and the draining of the sand bed ion. ce a linear regression analysis performs a ear least quari it of the data,the best estimat f the recent corrosion e is the slope from the re sion analysis for the period of' erest.The values are tabulated as the 'Apparent Corrosion e-paragraph 2.5.The upper bound of the 95% One-sided confidence interval I about the computed slope is an estimate of the maximum probable corrosion rate at 95% confidence.

The 95% upper bound is equal to the computed slope plus the one-sided t t-table value times the standard error of the slope. The value of t is determined for n-2 degrees of freedom.001/0004.20 Calc. No. C-1302-187-5300-011 Rev. No.,&' I Page 21 of 454 5.0 OLCMATIONS 5.1 6"x6* Grids in Sand Bed Region With Cathodic Protection 5.1.1 Bay 11~5.1.1.1 Day 11A: 5/2/87 to E-fte-Nine 49-point data sets were available for this bay covering 4/24/90 period. Since a plug lies within this region, four of the points were voided in each data set. The data were analyzed as described in paragraphs 4.4, 4.5.1 and 4.6.1.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 78.3% of the variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness t standard error is 870.4 1 5.7 mils.(6) The corrosion rate .standard error is-15.6 +/- 2.9 mile per year.(7) F/F critical -5.4.(8) The measurement below 800 mile was tested and determined not to be statistically different from the mean thickness.

5.1.1.2 Bay 11A: 1018/88 to 4/24/90 Five 49-point data sets were available for this bay covering this period.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) The F-test for the significant of the difference between the means shows that the difference between the mean thickness are not significant.

fi3/4~~~~001/0004.21 I , I10 2 :2 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 22 of 454 (4) The t-test of the last two data sets shows that the difference between the mean thickness is not significant.

(5) The current thickness based on the mean model is 878.9 + 5.9 mile.(6) These analyses indicate that the corrosion rate with cathodic protection is not significant compared to random variations in the measurements.

(7) The beat estimate of the corrosion rate during the period based on a least squares fit is -16.2 + 8.6 mils per year.5.1.2 Ba 1 5.1.2.1 Bay 12C: 5/1/87 to 4/24/90 Nine 49-point data sets were available for this bay covering this period. The initial analysis of this data indicated that the data are not normally distributed.

The lack of normality was tentatively attributed to minimal corrosion in the upper half of the 6"x6" grid with more extensive corrosion in the lower half of the grid. To test this hypothesis, each data set was divided into two subsets, with one containing the top three rows and the other containing the bottom four rows.Top 3 Rowe (1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 79% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 977.0 +/- 12.5 mils.(6) The corrosion rate is -35.2 +/- 6.8 mils per year.(7) F/F critical -4.6.001/0004.22 Caic. No. C-1302-187-5300-011 Rev. No. 0 Page 23 of 454 Bottom 4 Rows (1) Seven of the nine data sets are normally distributed.

The other two are skewed toward the thinner side of the mean. The Chi-square test shows that they are close to being normally distributed at the 1%level of significance.

(2) The regression model is appropriate.

(3) The regression model explains 80% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 865.0 1 7.8 mils.(6) The corrosion rate + standard error is-22.4 +/- 4.3 mils per year.(7) F/F critical -4.9 5.1.2.2 Bay IlCi 10/8/88 to 4/24/90 Five 49-point data sets were available for this period. These data were divided into two subsets as described above.Toy 3 Rowe (1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) The F-test for the significance of the difference between the means shows that the differences between the mean thicknesses are not significant.

(4) The t-test of the last two data sets shows that there is no statistical difference between their means.(5) These analyses indicate that the current corrosion rate with cathodic protection is not significant compared to random variations in the measurements.

001/0004.23 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 24 of 454 (6) Based on the mean model, the current thickness I standard error is 996.6 +8.3 mils.(7) The best estimate of corrosion rate during this period based on a least squares fit is -25.0 +/- 10.6 mile per year.Bottom 4 Rowe (1) Four of the five data sets are normally distributed. (See 5.1.2.1 above).(2) The mean model is more appropriate than the regression model.(3) The F-test for the significance of the difference between the means shows that the differences between the mean thicknesses are significant.

(4) The t-test of the last two data sets shows that there is no significant statistical difference between their means.(5) Based on the mean model, the current thickness

standard error is 878.1 +5.6 mile.(6) Based upon examination of the distribution of the five data set mean values, it is concluded that the current corrosion rate is not significant compared to random variations in the measurements.

The measurements alternated as follows: 897, 877, 891, 869, 863. Therefore the difference must be due to variations other than corrosion.

(7) The best estimate of the corrosion rate during this period based on a least squares fit is -16.7 ' 7.1 mile per year.001/0004.24 Cale. No. C-1302-187-5300-011 Rev. No. 0 Page 25 of 454 5.1.3 Bay 17D 5.1.3.1 Bay 17D: 2/17187 to 4/24190 Ten 49-point data sets were available for this period. Since a plug lies within this region, four of the points were voided in each data set. Point 24 in the 2/8/90 data was voided since it is characteristic of the plug thickness.

(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 95% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 829.5 + 4.0 mils.(6) The corrosion rate + standard error is-25.0 +/- 2.0 mils per year.(7) F/F critical = 29.4 (8) The measurements below B00 mile were tested and determined not to be statistically different from the mean thickness.

Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The regression model is more appropriate than the mean model.(3) The regression model explains 90% of the variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 830.1 +/- 3.8 mile.001/0004.25 1 ': / -- , i ý- 6 1, 2 : '3 1' : ý 6 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 26 of 454 (6) The corrosion rate + standard error is-23.7 + 4.6 mpy.(7) F/F critical -2.7 5.1.4 Bay 19A 5.1.4.1 Bay 19A: 2/17/87 to 4/24190 Ten 49-point data sets were available for this period. Since a plug lies within this region, four of the points were voided in each data set.(1) The data are normally distributed at the 1% level of significance.

(2) The regression model is appropriate (3) The regression model explains 96% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+/- standard error is 807.6 +/- 3.0 mils.(6) The corrosion rate +/- standard error is-21.4

• 1.5 mpy.(7) F/F critical -39.5 (8) The data points that were below 800 mils were tested and determined not to be statistically different from the mean thickness.

5.1.4.2 Bay 19A: 10/8/88 to 4/24/90 Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The regression model is more appropriate than the mean model.001/0004.26 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 27 of 454 (3) The regression model explains 90% of the variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 808.2 + 3.2 mile.(6) The corrosion rate +/- standard error is-20.6 +/- 3.9 mpy.(7) F/F critical -2.8 5.1.5 Bay 19B 5.1.5.1 Day 198: 5/1/87 to 4124/90 Nine 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 94% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 836.9 + 3.2 mile.(6) The corrosion rate +/- standard error is-19.0 +/- 1.7 mpy.(7) F/F critical -21.3 (8) The measurements below 800 mils were tested and determined not to be statistically different from the mean thickness.

5.1.5.2 Bay 199: 10/8/88 to 4/24/90 Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The regression model is more appropriate than the mean model.001/0004.27 I L12 I F-f 2 12 :2C2, 5.1.6 Bay 19C 5.1.6.1 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 28 of 454 (3) The regression model explains 75% of the variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 841.2 +/-t 3.3 mile.(6) The corrosion rate +/- standard error is-11.8 +/- 3.9 mpy.(7) F/F critical = 0.9 Bay 19C: 5/1/87 to 4/24/90 Nine 49-point data sets were available for this period. Since a plug lies within this region, four of the points were voided in each data set.(1) The data are normally distributed at the 1% level of significance, but appears to be developing two peaks.(2) The regression model is appropriate.

(3) The regression model explains 98% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error in 825.1 + 2.3 mile.(6) The corrosion rate + standard error is-24.3 + 1.3 mpy.(7) F/F critical = 66.2 (8) The measurements below 800 mile were tested and determined not to be statistically different from the mean thickness.

001/0004.28 10/2' Puf i 2 :--2 : -' 6 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 29 of 454 5.1.6.2 Bay 19C: 1018188 to 4/24/90 Five 49-point data sets were available for this period.(1) The data are normally distributed at the 1% level of significance.

(2) The F-test for significance of regression indicates that the regression model is appropriate.

(3) The regression model explains 93% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+ standard error is 826.3 +/- 2.9 mile.(6) The corrosion rate +/- standard error is-21.5 + 3.5 mpy.(6) F/F critical -3.7.5.1.7 Says 17119 Frame Cutout: 12130188 to 4/24/90 Two sets of 6"x6" grid measurements were taken in December 1988. The upper one is located 251 below the top of the high curb and the other below the floor. There is no previous data. The upper location was added to the long term monitoring program.Five 49-point data sets were available for this period.These data were analyzed as described in 4.4, 4.5.2 and 4.6.1. The initial analysis of this data indicated that the first and last data sets are not normally distributed.

The lack of normality was tentatively attributed to more extensive corrosion in the upper half of the grid than the bottom half. To test this hypothesis, each data set was divided into two subsets, with one containing the top three rows and the other containing the bottom four rows.001/0004.29

-C1/2"' 2:32:2 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 30 of 454 Ton 3 Rows (1) Four of the five subsets are normally distributed at the I% level of significance but one is not.(2) The mean model is appropriate.

(3) The F-test for the significance of the difference between the means shows that the differences between the mean thicknesses are not significant at 1% level of significance.

(4) These analyses indicate that the corrosion rate is not significant compared to the random variations in the measurements.

(5) Based on the mean model, the current thickness

+standard error is 996.0 + 4.7 mils.(6) The best estimate of the corrosion rate during this period based on a least squares fit is -8.2 +/- 10.7 mile per year.Bottom 4 Rows (1) Four of the five subsets are normally distributed at the 5% level of significance, and one at the 1% level of significance.

(2) The mean model is appropriate.

(3) The F-test for the significance of the difference between the means shows that the differences between the mean thicknesses are not significant at 1% level of significance.

(4) These analyses indicate that the corrosion rate is not significant compared to the random variations i the measurements.

(5) Based on the mean model, the current thickness standard error is 1005.7 + 5.6 mile.(6) The best estimate of the corrosion rate during this period based on a least squares fit in -13.1 +/- 11.6 mile per year.001/0004A.1 Calc. No. C-1302-187-5300-011 Rev. Ko. 0 Page 31 of 454 5.2 6"x6" Grids in Sand Bed Region Without Cathodic Protection 5.2.1 Bay 9D: 12/19/88 to 4/24/90 Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) The current mean thickness is 1021.7 +/- 8.9 mile.(4) The F-test for the significance of the difference between the mean thicknesses indicates that the differences between the means are significant.

The LSD analysis shows that this is due to the second measurement on 6/26/89 which is 33 to 52.3 mile higher than the other four.(5) The t-test of the last two data sets shows that the difference between the mean thicknesses is not significant.

(6) The overall analysis indicates that there was no significant corrosion from December 19, 1988 to April 24, 1990.(7) The best estimate of the corrosion rate during this period based on a least squares fit is -21.0 +/- 18.1 mile per year.5.2.2 Day 13A: 12/17/88 to 4/24/90 Seven 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 97% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+/- standard error is 853.1 t 2.4 mile.001/0004A.2 D/ 2 1 /(, f I Z : '.ý2 : 2f.Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 32 of 454 (6) The indicated corrosion rate I standard error is-39.1 + 3.4 mile per year.(7) F/F critical -16.9 (8) The measurements below 800 milo were tested and determined not to be statistically different from the mean thickness.

5.2.3 Bay 13D: 3/28/90 to 4/25/90 One 7-point data set and one 49-point data set are available for this bay covering this period.(1) The 7-point data set is normally distributed at 5%level of significance.

The 49-point data set is normally distributed at 1% level of significance.

However, there is a diagonal line of demarcation separating a zone of minimal corrosion at the top from a corroded zone at the bottom. Thus, corrosion has occurred at this location.(2) The mean of the 7-point data set is not significantly different from the mean of the corresponding 7 points in the 49-point data set.(3) The current means thickness is 931.9 + 22.6 mils.It is concluded that corrosion has occurred at this location.

However, with minimal data over a one-month period, it is impossible to determine the current corrosion rate.5.2.4 Bay 15D: 12/17/88 to 4/24/90 Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) The current mean thickness

+/- standard error is 1056.5: 2.3 mils.(4) The F-test for the significance of the difference between the mean thicknesses indicates that the differences between the means are not significant.

001/0004A.3 1, / 06 12: 12 :.ZiE Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 33 of 454 (5) The t-test of the last two data sets shows that the difference between the mean thicknesses Is not significant.

(6) There was no significant corrosion from December 17, 1988 to April 24, 1990.(7) The best estimate of the corrosion rate during this period based on a least squares fit is -4.6 mils per year.5.2.5 Bay 17A: 12117/88 to 4/24/90 Five 49-point data sets were available for this period.The initial analysis of this data indicated that the data are not normally distributed.

The lack of normality was tentatively attributed to minimal corrosion in the upper half of the 6"x6" grid with more extensive corrosion in the lower half of the grid. To test this hypothesis, each data set was divided into two subsets, with one containing the top three rows and the other containing the bottom four rows.ToR 3 Rows (1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) The current mean thickness

+/- standard error is 1128.3+/- 2.2 mils.(4) The F-test for the significance of the difference between the mean thicknesses indicates the differences between the means are not significant.

(5) The t-test of the last two data sets indicates that the difference between the mean thicknesses is not significant.

(6) There was no significant corrosion during this period.(7) The best estimate of the corrosion rate during this period based on a least squares fit is -6.8 +/- 3.7 mile per year.001/0004A.4 Calc. No. c-1302-187-5300-011 Rev. No. 0 Page 34 of 454 Bottom 4 Rows (1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) The current mean thickness

+/- standard error 950.83+/- 5.3 mile.(4) The F-test for the significance of the difference between the mean thicknesses indicates that the differences between the means are not significant.

(5) The t-test of the last two data sets indicates that the difference between the mean thicknesses is not significant.

(6) There was no significant corrosion during this period.(7) The best estimate of the corrosion rate during this period based on a least squares fit is -17.7 + 7.6 mile per year.5.3 6161 Grids at 51' Elevation 5.3.1 Bay 5 Area D-l 2 51' Elevation:

11/1/87 to 4/24/90 Eight 49-point data sets were available for this period.The initial analysis of this data indicated that the data are not normally distributed.

These data sets names start with E. The following adjustments were made to the data: (1) Point 29 in the 9/13/89 data is much greater than the preceding or succeeding measurements.

Therefore, this reading was dropped from the analysis.(2) Point 9 is a significant pit. Therefore, it was dropped from the overall analysis and is evaluated separately.

(3) Points 13 and 25 are extremely variable and are located adjacent to the plug which was removed from this grid. They were also dropped from the analysis.(4) Point 43 in the 11/01/87 data is much less than any succeeding measurement.

Therefore, this reading was dropped from the analysis.001/0004A.5 Calc. No. C-1302-187-5300-01O Rev. No. 0 Page 35 of 454 With these adjustments, the first and last data sets are normally distributed at the 1% level of significance and the other five at 5%. These data set names start with F.It was noted that the D-Meter calibration at 0.750- yielded readings which ranged from -1 mil for one set of measurements to + 4 mile for another. The data was adjusted to eliminate these biases. These data set names start with G. The final analyses are based on these adjusted data sets.(1) The data are normally distributed.

(2) The regression model is appropriate.

(3) The regression model explains 57% of the total variation about the mean.(4) The residuals are normally distributed.

(5) The current mean thickness

+/- standard error is 745.2+/- 2.1 mile.(6) The indicated corrosion rate +/- standard error is -4.6+/- 1.6 mile per year.(7) F/F critical -1.3. Thus, the regression is just barely significant.

(8) The F-test for significance of the difference between the mean thickness indicates that the differences are significant.

(9) The t-test of the last two data sets shows that the difference between the mean thickness is not significant.

(10) The measurements of the pit at point 9 were 706, 746, 696, 694, 700, 688, 699 and 689 mile. The mean value of these measurements is 702.3 + 6.5 mile. A least squares fit shows that the best estimate of the corrosion rate during this period is -11.5 mils per year with R 2-31%. The second measurement is much higher than the others. Dropping this point, the mean of the remaining measurements is 696.0 +/- 2.4 mile, and the best estimate of the corrosion rate is-4.9 mils per year with R 2 .49%. Recognizing that the variability of single measurements will be about 6 times the variability of the mean of 40 measure-ments, it is concluded that the corrosion rate in the pit is essentially the same as the overall grid.001/0004A.6 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 36 of 454 5.3.2 Bay S Area 51-5 at 511 Elevation:

3/31/90 to 4/25/90 Two 49-point data sets are available for this time period.11) The data are not normally distributed.

This is due to a large corroded patch near the center of the grid, and several small patches on the periphery.

When the data less than the grand mean were segregated, it was found that these subsets are normally distributed.

(2) The t-tests of the two complete data sets and the two subsets indicate that the difference between the mean thicknesses are not significant.

(3) The current mean thickness

+ standard error is 745.1 t 3.2 mils.It is concluded that corrosion has occurred at this location.

However, with minimal data over such a brief period, it is impossible to determine the current corrosion rate.5.3.3 Bay 13 Area 31 Elevation 511: 3/31/90 to 4/25/90 Two 49-point data sets are available for this time period.(I) The data are to normally distributed.

This is due to a large corroded patch at the left edge of the grid.When the data less than the grand mean were segregated, it was found that these subsets are normally distributed.

(2) The t-test of the two complete data sets indicate that the difference between the means is statistically significant.

However, the difference between the means of the two subsets is not statistically significant.

{3) The current mean thickness is +/- standard error is 750.8

• 11.5 mils.It is concluded that corrosion has occurred at this location.

However, with minimal data over such a brief period, it is impossible to determine the current corrosion rate.001/0004A.7

,/2 1 /ý,ir 1 ":,2 : 2 6 Calc. No. C-1302-187-S300-011 Rev. No. 0 Page 37 of 454 5.3.4 Bay 15 Area 23 Elevation 51:I 3132/90 to 4/25/90 Two 49-point data sets are available for this time period.(1) The data are not normally distributed.

This is due to a large corroded patch.When the data less than the grand mean were segregated, it was found that these two subsets are normally distributed.

(2) The t-tests of the two complete data sets and the two subsets indicate that the differences between the mean thicknesses are not significant.

(3) The current mean thickness

+/- standard error is 751.2+/- 3.8 mile.It is concluded that corrosion has occurred at this location.

However, with minimal data over such a brief period, it is impossible to determine the current corrosion rate.5.4 6" x 6" Grids at 52' Elevation 5.4.1 Bay 7 Area 25 Elevation 52': 4/26190 One 49-point data set is available.

(1) The data are not normally distributed.

The subset of the data less than the mean thickness is not normally distributed.

When four points below 700 mile were dropped from the data set, the remaining data was found to be normally distributed.

Therefore, the lack of normality of the complete data set is attributed to these thinner points. Three of these could be considered to be pits (626, 657 and 676 mile) since they deviate from the mean by more than 3 sigma.(2) The current mean thickness

+/- standard is 715.5 + 2.9 mile.It is concluded that corrosion has occurred at this location.001/0004A.8

'C/-l/03; 12:32: Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 38 of 454 5.4.2 Bay 13 Area 6 Elevation 52*1 4/26/90 One 49-point data set is available.

(1) The data are not normally distributed.

The subset of the data less than the mean thickness is normally distributed.

Thus, the lack of normality of the complete data set is attributed to a large corroded patch at the left side of the grid.(2) The current mean thickness

+/- standard error in 724.9; 2.9 mils.(3) It is concluded that corrosion has occurred at this location.5.4.3 Bay 13 Area 32 Elevation 521: 4/26/90 One 49-point data set is available.

(1) The data are not normally distributed.

The subset of the data less than the mean thickness is normally distributed.

Thus, the lack of normality of the complete data set is attributed to these corrosion patches.(2) The current mean thickness

+/- standard error is 698.3+ 5.0 mils.It is concluded that corrosion has occurred at this location.5.4.4 Bay 19 Area 13 Elevation 521: 4/26/90 One 49-point data set is available.

(1) The data are normally distributed.

However, two adjacent points differ from the mean by 3 sigma and 5 sigma. Thus, there is a pit.(2) The current means thickness

+/- standard error is 712.5+ 3.1 mils.It is concluded that some corrosion has occurred at this location.001/0004A.9 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 39 of 454 5.5 k" x (5.5.1 6" Grids at 87' Elevation Say 9 87' Elevation:

11/6/87 to 3/28190 Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) There was no significant corrosion during this period.(4) The current mean thickness

+ standard error is 619.9+/- 0.6 mile.(5) The beat estimate of the corrosion rate during this period based on a least squares fit is -0.2 +/- 0.9 mile per year.5.5.2 Bay 13 87' Elevation:

11/10/87 to 3/28/90 Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.(3) There was no significant corrosion during this period.(4) The current mean thickness

+/- standard error is 636.5+/- 0.8 mile.(5) The best estimate of the corrosion rate during this period based on a least squares fit is zero mile per year.5.5.3 Bay 15 87* Elevation:

11/10/87 to 3/28/90 Five 49-point data sets were available for this period.(1) The data are normally distributed.

(2) The mean model is more appropriate than the regression model.001/0004A.10 1 ý/." 1II/ý 4 : ": 32 :216 Calc. No. C-1302-187-5300-011 Rev. No. 0 Page 40 of 454 (3) There was no significant corrosion during this period.(4) The current mean thickness

+ standard error is 636.2÷ 1.1 mile.(5) The best estimate of the corrosion rate during this period based on a least squares fit is zero mils per year.001/0004A.11