Text

Email Providing Draft Responses to Request for Additional Information
	ML032030227
Person / Time
Site:	Dresden
Issue date:	07/22/2003
From:	Vanover D; ERIN Engineering & Research
To:	Perch R, Larry Wheeler; NRC/NRR/DRIP/RLEP, Division of Systems Safety and Analysis
	Wheeler L, NRR/DRIP/RLEP, 415-1444
References
	Download: ML032030227 (126)
	v • d • e

- -

- I I Duke Wheeler - Draft Dresden SAMA RI Respone Page 1 I Duke Wheeler - Draft Dresden SAMA RAI Responses Page 1 I From:

Donald E. Vanover <DEVanover@enneng.com>

To:

<dxw~nrc.gov>, <rlp3@nrc.gov>

Date:

7/16103 11:43AM

Subject:

Draft Dresden SAMA RAI Responses Gentlemen:

Bill Maher requested that I send the Draft Dresden SAMA RAI responses to you directly.

See attached.

Don Vanover ERIN Engineering and Research, Inc.

Risk Management Group West Chester, PA Office devanoverE erineng.com 610-431-8260 CC:

"'Bill Maher' <william.maher@ exeloncorp.com>, "'Jeff Gabor"' <jrgabor@erineng.com>

RAI 1

The SAMA analysis is based on the most recent version of the Dresden Nuclear Power Station (DNPS) Probabilistic Safety Assessment (PSA) for internal events, i.e., 2002 Update, which is a modification to the modified IPE submittal transmitted to the NRC in June 1996. Please provide the following information regarding this PSA model:

a. a summary description of any peer reviews of the level 1 and level 2 portions of this PSA beyond the normally-performed internal second checker reviews (e.g., DNPS BWROG Peer Review),

b. a characterization of the findings of these internal and external peer reviews (if any), and the impact of any identified weaknesses on the SAMA identification and evaluation process,

c. a breakdown of the internal events core damage frequency (CDF) by major contributors, initiators and accident classes, such as loss of offsite power (LOOP), station blackout (SBO), transients, anticipated transient without scram (AWS), loss-of-coolant accident (LOCA),

ISLOCA, internal floods, and other, and

d. a description of the major differences from the updated IPE submittal, including the plant and/or modeling changes that have resulted in the new core damage frequency (CDF), along with the corresponding CDF.

Response 1(a):

'[ProvideJ a summary description of any peer reviews of the level 1 and level 2 portions of this PSA beyond the nonnally-performed intemal second checker reviews (e.g.,

DNPS BWROG Peer Review)[.r Three external peer reviews of the Dresden Probabilistic Risk Assessment (PRA) models were conducted.

BWROG Peer Review/Certification Boiling Water Reactor Owners Group (BWROG) PRA Certification Peer Review was conducted in January 1998. (Note, Level 2 analysis of Large, Early release frequency (LERF) was not included in this review). A six-member industry team following the latest BWROG guidance available at the time performed this review.

1

Independent External Review Robert Schmidt performed the independent review with support from Jeff Julius (HRA area). This review was performed in late 1998 and early 1999.

NEI/BWROG Peer Review A six-member industry team performed this review in January 2000 with a report published in March 2000. The review used the Nuclear Energy Institute (NEI) draft,

'Probabilistic Risk Assessment Peer Review Process Guidance." This peer review process was adapted from the review process originally developed and used by the BWROG.

Response 1(b):

[Provide] a characterization of the findings of these internal and external peer reviews (if any), and the impact of any identified weaknesses on the SAMA identification and evaluation process.r BWROG Peer Review/Certification This review evaluated all PSA elements except Level 2 analysis. The evaluation found that all elements were consistently graded as sufficient to support meaningful rankings for the assessment of systems, structures, and components, when combined with deterministic insights. Enhancements were recommended in the following areas:

Completion of Level 2 analysis Treatment of Special Initiators (some special initiators were missing or treatment through the Accident Sequence Evaluation (Event Trees) was judged to require improvements)

Accident Sequence Evaluation (Event Trees) were overly simplified and needed further development to support higher applications Dependency Analysis (Common Cause Factors)

Human Reliability Analysis (HRA) (Operator dependency analysis was judged to require improvement and operator input was necessary).

There is judged to be no impact to the SAMA identification and evaluation process as weaknesses were corrected since the review. Insights were developed and evaluated using the upgraded PRA models. Enhancements included addition of special initiators, upgrading Event Tree Analysis, revision of human reliability analysis (including dependency analysis and operator interviews), update of Common Cause Factors, and completion of Level 2 analysis.

2

Independent External Review This review, primarily performed by Robert Schmidt, was limited to the Level 1, at power, internal events portion of the PSA, excluding the internal flooding analysis. The review focused on Initiating Events, Event Trees, Success Criteria, System Analysis (3 systems), Human Reliability Analysis, and Quantification.

The review found the Dresden Updated PRA to be a high quality Level 1 PSA.

All technical elements reviewed meet or exceed general industry practice. The reviewer found the update process to be well documented in analysis notebooks. No deficiencies were found in the analyses that needed to be corrected immediately. Exelon responded to all of Mr.

Schmidt's comments and made model changes where appropriate.

Areas enhanced in the PRA model to address Mr. Schmidt's comments included the following:

Initiating Events: Loss of a single DC bus was added to the model, Interfacing System Loss of Coolant Accident (ISLOCA) frequency calculation updated, IE frequency updated to reflect improved availability Data: Check Valve failure rate increased to reflect plant data Event Tree: Depressurization added to credit Shutdown Cooling (SDC)

System (Note, Dresden has a SDC system in addition to a Residual Heat Removal (RHR) system (The RHR system is referred to as the Low Pressure Coolant Injection (LPCI/CCSW) system at Dresden). Failure to inhibit ADS was added to several ATWS sequences. A manual scram following IORV was added to the ATWS Event Tree logic. Recovery of containment vents following loss of Instrument air (0.9 failure probability) was added.

HRA: Modified HEP (based on Operator interviews) associated with controlling injection following boron injection. Increased HEP to 0.1 for an operator action to assure dependency was picked up in the cutset recovery process (HEP is returned to normal value following recovery process).

There is judged to be no impact to the SAMA identification and evaluation process as weaknesses (all considered relatively minor) were corrected since the Independent reviewer evaluation. Insights were developed and evaluated using the upgraded PRA models.

NEI/BWROG Peer Review The NEI Certification team gave high marks to the Dresden PRA. The team specifically noted, "The Dresden PSA is consistent with state of the art technology PRAs in scope, methods, data usage, and results. The PSA does not have unique PSA features." Of the eleven elements" evaluated by the team, a Summary Score of "4" was received for the Maintenance and Update element, and Summary Scores of "3" were assigned to the ten 3

other elements. In the words of the review team, These grades are consistent with a very solid PSA program with no major weaknesses." There were no A" level Facts &

Observations (F&Os). There were eight 1B" level F&Os. The 2002 Dresden Update resolved all B" F&Os and a number of "C" F&Os as well.

The most significant recommendations identified weaknesses in the area of Level 2 (LERF) analysis, internal flooding and thermal hydraulic analysis. Special efforts to enhance the PRA model in these three areas have been completed. Further discussion on Level 2 enhancements can be found in response to RAI 3.

There is judged to be no impact to the SAMA identification and evaluation process as weaknesses were corrected since the PEER review evaluation.

Insights were developed and evaluated using the upgraded PRA model.

Response 1(c):

'[Provide) a breakdown of the internal events core damage frequency (CDF) by major contributors, initiators and accident classes, such as loss of offsite power (LOOP),

station blackout (SBO), transients, anticipated transient without scram (ATWS), loss-of-coolant accident (LOCA), ISLOCA, internal floods, and other tcontributors.r Table 1-1 provides a breakdown of the internal events CDF by initiator type, and Table 1-2 provides a breakdown of the internal events CDF by accident class. Note that ATWS and SBO scenarios are not represented by individual initiators but are determined as consequences from an initiating event. The ATWS contribution from the 2002 update model is estimated at 10% of the total CDF, and the SBO contribution is estimated at 22% of the total CDF.

4

Table 1-1 Contribution to Dresden 2002 PRA CDF by Initiator (Truncation Umit = 1 E-1O/yr)

Description

% of Base Single Unit Loss of Offsite Power Initiating Event (LOOP) 26%

Transient With Feedwater (FW) Unavailable and Main Condenser (MC) 22%k Av ilable__

Dual Unit Loss of Offsite Power (DLOOP) 15%k Transient With FW And MC Available 11%

Loss of Multiple DC Buses 8%

Medium LOCA Initiator (MLOCA) 3%

Large LOCA Initiator (LLOCA) 3%

Manual Shutdown Initiating Event 3%

Internal Flooding 3%

Loss of Service Water Initiating Event 2%

Loss of Turbine Building Closed Cooling Water (TBCCW) 0.9%

ISLOCA 0.1%

All other initiators 3%6 Total.

100%

5

Table 1-2 Contribution to CDF by Accident Class (Truncation Limit = 1 E-1 Otyr)

Accident Short Description CDF Updatedt CCDDF(lD Loss of Makeup at High Reactor Pressure Vessel 56.1%

IA/IE (RPV)Pressure 1.06E-06 / yr 56_1_

IBE Early Station Blackout (less than 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months ) 3.01 E-07 / yr 15.9%

IBL Late Station Blackout (greater than 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months ) 1.08E-07 / yr 5.7%

IC Loss of Makeup (ATWS) 1.79E-08 / yr 0.9%

ID Loss of Makeup at Low RPV Pressure (transient Initiators) 2.56E-08 / yr 1.4%

II Loss of Decay Heat Removal 8.15E-08 / yr 4.3%

IIIB Loss of Makeup at High RPV Pressure (LOCA Initiators) 1.53E-08 / yr 0.8%

1110 Loss of Makeup at Low RPV Pressure (LOCA Initiators) 8.20E-08 / yr 4.3%

IIID Loss of Vapor Suppression 1.18E-08 /yr 0.6%

IVA Loss of Reactivity Control (ATWS) 1.86E-07 / yr 9.8%

V Containment Bypass 1.74E-09 / yr 0.1%

Total:

1.89E-06 / yr 10O.0%

Response 1(d):

'[Provide a description of the major differences from the updated IPE submittal, including the plant and/or modeling changes that have resulted in the new core damage frequency (CDF), along with the corresponding CDF."

A summary of the total calculated CDF for each of the relevant models is provided in Table 1-3. As can be seen, the Dresden CDF has been reduced from the Modified IPE CDF to the present. The total reduction in CDF is approximately 44%. Table 1-4 provides the change in CDF contribution from the Modified IPE to the 2002 Update.

Here, it can be seen that the Dresden risk profile has also changed significantly.

Also provided is information that relates modeling methodology, plant data and plant configuration changes to changes in Core Damage Frequency. Examples of each type of change are listed below. These changes are from the 1999 Upgrade.

PRA Methodology Change: Calculating Medium LOCA frequency using the latest EPRI methodology increased the MLOCA frequency.

6

Plant Operating Experience:

The General Transient Frequency was reduced based on operating experience.

This caused a decrease in ATWS contribution by -60%.

Plant Configuration Changes: Installation of Station Blackout Diesel Generators and the Division 1 4kV cross-tie reduced Loss of Off-She Power contribution by 75%.

It is apparent from this information that the present PRA results are significantly different from the Modified IPE. One could conclude that insights from the present model are more valuable than IPE insights at this time.

Table 1-3 Dresden CDF History Model J

Date I

CDF (Per Yr) (1)

Modified IPE 1996 3.4E-06yr Upgraded PRA 1999 2.6E-06/yr 2002 Updated PRA 2002 1.9E-06/yr (2)

Notes to Table 1-3 (1) Results shown are for Unit 2. The Unit 3 results are the same except for the Modified IPE. The Modified IPE Unit 3 CDF was 5E-06/yr. The difference in the CDF estimates between the two units was due to a hardware modification that eliminated a FW trip on loss of DC power as an initiating event at Unit 2. The modification was later installed on Unit 3.

2) The most recent version of the Dresden 2002 PRA model is Revision 3.

However, the SAMA calculations were performed using Revision 0 of the 2002 PRA. For consistency with the SAMA evaluations, the results of the Revision 0 model are reported for the RAI responses. The Revision 3 results are not significantly different than for Revision 0.

7

Table 1-4 Dresden Risk Profile History Initiator Modified IPE (1) 1999 Uggrade 2002 U p d a te MLOCA 39%

21%

3%

DLOOP 24%

23% (3) 15% (4)

LOOP 8%

3%

26% (4)

General Transient 27%

27%

33%

Loss of Service Water 1%

11%

2%

ISLOCA

<0.1%

0.4%

0.1%

Loss of Instrument Air

<0.1%

7%

0.9%

Large LOCA

<0.1%

2%

3%

Excessive LOCA N/A 0.2%

0.3%

Loss of Multiple NIA 3%

8%

125 VDC Buses Loss of TBCCW N/A 0.9/0 0.9%

Manual Shutdown N/A 0.6%

2%

Service Water Flood N/A N/A 3%

All Others

<1%

<3%

Notes to Table 1-4 t

The Modified IPE report gave a separate ATWS result that included contributions from many initiators, but mainly due to General Transients. Therefore, for risk profile comparison purposes, the ATWS contribution is included with the General Transient results above in the Modified IPE column.

Unit 2 results are reported (see note 1 of previous table).

° The third and fourth columns contain the 1999 and 2002 PRA Model Update results with ATWS contributions included with the results for each initiator. For example, the table shows that General Transients contribute 27% of the total CDF in the 1999 Upgrade. General Transients evolving into ATWS events contributed 14.4% of the total CDF.

) Of the 23% contribution of Dual Unit LOOPs in the 1999 Upgrade, approximately two-thirds is due to Station Blackout (SBO) sequences. No other initiators (including Single Unit LOOPs) include any significant CDF contribution to SBO sequences.

(4)

Of the 41% CDF contribution of Single and Dual Unit LOOPs in the 2002 Update, approximately one-half is due to Station Blackout (SBO) sequences.

8

RAI2 The CDF cited and used in the SAMA analysis is based on the isk profile for internal events at DNPS Unit 2. Provide the internal events CDF for Unit 3, and a discussion of the reasons for any differences from Unit 2. Discuss the impact on the SAMA analysis, including the impacts of external events, and results ff the analysis were based on Unit 3 rather than Unit 2.

Response 2 Internal Events Unit 2 CDF The Unit 2 internal events CDF is identical to that of Unit 3: 1.9E-06/yr.

Unit 2 Differences from Unit 3 There are asymmetries between Unit 2 and Unit 3 related to the 125V DC System bus configuration, AC Bus Initiating Event Frequency difference due to water spray, AC power supplies to the SDC pumps, HPCI room cooler and LPCI cooling (Containment Cooling Service Water System (CCSW).

These asymmetries involve highly reliable components, such as Electric Power Busses.

Thus, there is a minimal impact on baseline CDF.

Loss of AC Bus Initiator Frequency:

The Loss of AC Bus Initiator Frequencies includes loss of bus due to water spray. The water spray contributions vary by unit. This asymmetry is not significant.

125V DC System Bus Configuration: The 125V DC System is essentially a plant distribution system: normally supporting equipment on both units in order to provide divisional separation for safety related systems.

Divisional separation is accomplished by having Division I equipment on a unit supplied by the Unit's own 125V DC System while the opposite unit's 125V DC system supplies the Division II loads. For example, the 125V DC battery located in Unit 2 is considered "Division I" power supply for Unit 2 and Division 11' power supply for Unit 3. The two battery divisions have asymmetries in the bus configurations. The electric busses and cable connections are highly reliable and the 125 V DC system Bus asymmetries have an insignificant impact to baseline CDF.

Shutdown Cooling (SDC) SYSTEM:

The power supplies to the SDC pumps are not symmetric. There are three SDC pumps for each unit. The Unit 2 2A and 2C SDC pumps obtain power from Bus 23-1 and the 2B 9

pump obtains power from Bus 24-1. The Unit 3 3A SDC pump obtains power from Bus 33-1 and the 3B and 3C pumps obtain power from Bus 34-1. The impact of this asymmetry is insignificant.

HPCI Room Cooler: The power supply for the HPCI room cooler is not symmetric. The Unit 2 HPCI room cooler is powered from MCC 29-4 and the Unit 3 HPCI room cooler is powered from MCC 39-1. These MCCs are powered from symmetric sources and therefore, this asymmetry is not significant.

LPCI Containment Cooling (CCSW System): The CCSW System for each unit is comprised of two loops with 2 pumps per loop. The Unit 2 CCSW Loop A is partially dependent on MCC 28-2 and the Unit 3 CCSW Loop A is partially dependent on MCC 38-3. The Unit 2 CCSW Loop B is partially dependent on MCC 29-4 and the Unit 3 CCSW Loop B is partially dependent on MCC 39-1. These asymmetries are not significant.

The Unit 3 model uses the same event trees as the Unit 2 model. The Unit 3 model uses the same system logic and database, except as impacted by the items above.

While these differences do appear in low-frequency cutsets, the effects of the fault tree differences are small enough that they do not affect the total internal events CDF.

Therefore, the differences do not affect the SAMA analyses for internal events.

Extemal Events Unit 2 Fire CDF The Unit 2 fire CDF is 1.7E-05/yr compared to a Unit 3 fire CDF of 3.1 E-05/yr.

Fire-Related Unit 3 Differences from Unit 2 A review of the dominant risk contributors shows a few notable asymmetries in the risk profiles.

Self-Initiated Cable Fires Due to Cable Routing Differences The risk contribution from self-initiated cable fires is much higher in Unit 3 (25%) than in Unit 2 (2%). Two thirds (5.OE-06Iyr) of the Unit 3 self-initiated cable fire contribution results from Cable Tunnel fire scenarios. Since Unit 2 cables are generally not routed through this area, a similar exposure does not exist for Unit 2. Of the remaining Unit 3 contribution, one half (1.OE-06/yr) results from fires on the second floor of the Reactor Building. These fires also affect cables whose fire-induced failure disable Suppression Pool Cooling, Shutdown Cooling, and one or more trains of Core Spray and LPCI, 10

depending on the scenario being considered, which increases the significance of these fires. It should also be noted that safe shutdown procedures were not credited in the Dresden fire analysis. Recovery of the Isolation Condenser is addressed in the Safe Shutdown Procedures. Crediting this recovery would reduce the difference in CDF from unit asymmetries.

The common control room is located adjacent to Unit 2.

Unit 3 is located on the opposite side of Unit 2, thus requiring a Unit 3 cable tunnel to the control room area.

The SAMA analysis would not be impacted as rerouting sufficient number of cables to significantly reduce fire risk in Unit 3 is judged not to be cost effective. Crediting safe shutdown procedures in the Dresden fire analysis would reduce the contribution from these scenarios.

Loss of 125 VDC in Unit 2 A large oil fire involving Unit 2 Reactor Feedwater Pump C or a fire involving MCC 26-1 is a dominant contributor to the Unit 2 CDF. This is because of the location of the cables needed for the Unit 2 DC power system. The Unit 3 DC power feed to one train of the Unit 2 DC system and the Unit 2 AC power cable to the battery charger for the redundant DC train are exposed to a common hazard. Although the circuits are located in separate trays, they are stacked vertically. The occurrence of a postulated large fire event requires an operator action to either align the spare battery charger or to connect the spare Unit 2 battery bank.

An option also exists to use the safe shutdown procedures, which specify manual actions to operate the Isolation Condenser. The safe shutdown procedures were not credited in the Dresden Fire PRA model. Fires involving the Unit 2 Reactor Feedwater Pump C and MCC 26-1 contribute approximately 4.OE-06/yr (with no credit given to the safe shutdown procedures).

RWCU pump fires in Unit 2 Fires originating from the Unit 2 RWCU pumps contribute 8.5% (1.OE-06/yr) to the Unit 2 CDF, while Unit 3 RWCU pump fires contribute less than 1% to the Unit 3 CDF. The difference between the two scenarios is primarily due to a fire-induced loss of the Isolation Condenser in the Unit 2 analysis. In particular, the trays affected by the Unit 2 pump fire contain cables whose fire-induced failure disables Reactor Building 25OVDC MCC #2. Such an exposure does not exist in the Unit 3 analysis. Crediting the safe shutdown procedures would allow for recovery of the Isolation Condenser.

Seismic-Related Unit 3 Differences from Unit 2 With modifications to each unit in response to the Seismic Margins Analysis, there is no significant difference in seismic vulnerabilities between the two units.

11

RAI3 In the Extended Power Uprate (EPU) Amendment application, Exelon indicates that the Level 2 analysis is based on NUREG/CR-6595. However, there is no such indication in the SAMA portion of the Environmental Report (ER). Based on the above, provide a description of the following:

a. the changes in the Level 2 methodology since the modified IPE submittal, including major modeling assumptions, containment event tree (CET) structure, and binning of endstates.

b. the methodology and citeria for binning CET endstates into release categories used in the Level 3 analysis. Include the definitions of the release characteristics listed in Column 2 of Table 4-5.

c. each release (consequence) category used in the Level 3 analysis (as listed in Column 1 of Table 4-5), the specific source terns used to represent each release category, and a containment matrix describing the mapping of Level 1 results (plant damage state frequencies) into the various release categories.

Response 3(a):

"[Provide] the changes in the Level 2 methodology since the modified IPE submittal, including major modeling assumptions, containment event tree (CET) structure, and binning of endstates.'

The IPE and modified IPE employed what some would call a simplistic Level 2 methodology. Many accident progression phenomena or failure modes were eliminated from consideration, based on experiments, MAAP calculations, or judgments concerning the likelihood of various phenomena. Core damage end states were coded for sequence characteristics that would affect the remaining phenomena affecting containment performance. Based on those characteristics, it was determined in what time range the vessel would fail, whether the pedestal area was dry or wet, whether containment sprays were operating, whether liner melt-through was likely, and whether containment vent was operated. Based on this information, it was determined which core damage end states resulted in containment failure, and which resulted in LERF.

Because of the limitations of the IPE Level 2 model, the model was revised for the 1999 Dresden PRA Upgrade. It was decided to use a simplified LERF model in the style of NUREG/CR-6595. The 1999 Dresden PRA was used for the Extended Power Uprate (EPU) submittal.

The submittal for License Renewal required Level 3 calculations. Therefore, Exelon decided to develop a full Level 2 PRA model for Dresden that meets standard industry 12

practices. The full Level 2 model was used for the License Renewal analyses, and that model also has now been incorporated in the 2002 Dresden PRA model. It is also the basis for LERF calculations for risk assessment.

A brief summary of the current Level 2 model compared to the 1999 Level 2 model that was used for the EPU submittal follows:

No changes in modeling assumptions CET structure has been enhanced to include more top event nodes Old CET had LERF and non-LERF end states whereas the updated model has several release category bins (see Responses 3(b) and 3(c))

Response 3(b):

"[Provide) the methodology and cteria for binning CET endstates into release categories used in the Level 3 analysis.

Include the definitions of the release characteristics listed in Column 2 of Table 4-5."

Each CET end state can be associated with a radionuclide source term bin, which covers a spectrum of similar potential scenarios and timing. Theoretically, it would be desirable in determining the point estimates of risk to evaluate the source terms for each sequence of each accident plant damage state. However, for purposes of risk presentation, the CET end states can also be characterized in such a manner as to combine similar Mconsequence impact" sequences within a CET end state.

The discrete nature of the radionuclide release categories means that the severe accident spectrum is divided up into bins, which then represent a group of severe accidents that have similar characteristics. These characteristics would imply similar public health consequences.

It has been found in the past that the public health consequences are affected by a large number of governing features. The following portrays the radionuclide release category characterization used for Dresden.

Radionuclide Release CateQories (CET End States)

The spectrum of possible radionuclide release scenarios is represented by a discrete set of categories or bins. The end states of the containment and phenomenological event sequences may be characterized according to certain key quantitative attributes that affect offsite consequences. These attributes include two important factors:

Timing (e.g., early or late releases); and, Total quantity of fission products released.

13

Therefore, the containment event tree end states represent the source term magnitude and relative timing of the radionuclide release. The number of categories used for Dresden (i.e., 13) in the source term characterization offers a level of discrimination similar to that included in numerous published PRAs.

Timing Bins Three timing categories are used, as follows:

Early (E)

Less than time when evacuation is effective Intermediate (I)

Greater than or equal to Early, but less than 24 hrs Late (L)

Greater than or equal to 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

The definition of the categories is based upon past experience concerning offsite accident response:

Early is conservatively assumed to include cases in which minimal offsite protective measures have been observed to be performed in non-nuclear accidents.

Intermediate is a time frame in which much of the offsite nuclear plant protective measures can be assured to be accomplished.

Late (>24 hours) are times at which the offsite measures can be assumed to be fully effective.

Radionuclide Release Magnitude Bins The assessment of plant response under postulated severe accident scenarios is a complex integrated evaluation.

The primary and secondary containment building responses are sensitive to pressures, temperatures, flows, and event timings. These parameters also affect the operator action timings, the radionuclide release timings, and the mitigating system performance assessments.

Therefore, the proper plant specific characterization of the severe accident progression is important to the realistic representation of the plant and highly desirable for the Level 2 assessment.

These deterministic calculations provide the following information:

The pressures and temperatures for various accident scenarios in the RPV, the drywell, the wetwell, and the reactor building; The times to reach these pressures and temperatures which is key to the assessment of recovery; (The time windows available for recovery actions must be estimated.)

The source term magnitude and timing.

14

Five severity classifications associated with volatile or particulate releases are defined as follows:

High (H) - A radionuclide release of sufficient magnitude to have the potential to cause prompt fatalities.

Medium or Moderate (M) - A radionuclide release of sufficient magnitude to cause near-term health effects.

Low (L) - A radionuclide release with the potential for latent health effects.

Low-Low (LL) - A radionuclide release with undetectable or minor health effects.

Negligible (OK) - A radionuclide release that is less than or equal to the containment design base leakage.

A relationship was then developed with the five release severity categories. The results of this partitioning are shown in Table 3-1.

Table 3-1 Release Severity Categorization Release Severity Fraction of Released Csl Fission Products High greater than 10%

Medium/Moderate 1 to 10%

Low 0.1 to 1.0%

Low-Low 1 )

less than 0.1%

Negligible much less than 0.1%

The resulting definitions of the radionuclide release end states are summarized in Table 3-2. The combinations of severity and timing classifications results in one OK release category and 12 other release categories of varying times and magnitudes. These 12 other release categories are shown in Table 3-3. These are the dominant release categories shown in column 2 of Table 4-5 of the Environmental Report.

15

Table 3-2 Release Severity And Timing Classification Scheme Release Severity Release Timing l Classification Cs Iodide %

Classification l

rime of Initial Release(1 )

Classification CsIdd lsiiation Relative to Time for General Category Release Category Emergency Declaration High (H)

Greater than 10 Late (L)

Greater than 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months Medium or Moderate 1 to 10 Intermediate (I) 5 to 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months (M)

Low (L) 0.1 to 1 Early (E)

Less than 5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months Low-low (LL)

Less than 0.1 No iodine (OK) 0

) The conditions dictating a General Emergency are used as the surrogate for the time when EALs are exceeded, which in turn is used as the relative time to measure when the release occurs.

Table 3-3 Dresden Release Categories Time of Magnitude of Release Release H

M L

It E

H/E WE LIE LL/E H/A Ml L

LL/I L

HA.

Mt.

L/L LLAL 16

Response 3(c):

7ProvideJ each release (consequence) category used in the Level 3 analysis (as listed in Column 1 of Table 4-5), the specific source terms used to represent each release category, and a containment matrix describing the mapping of Level 1 results (plant damage state frequencies) into the various release categories."

Source Terms used to Represent each Release Categorv As requested, Table 3-4 provides a list of the source terms associated with each of the release categories as listed in Column 1 of Table 4-5 of the ER.

17

Table 3-4 Source Terms Associated with Each Release Category Release Category( 2 1.2-1 I

L2-2 I

1.2-3 I

1.2-4 1.2-5 1.2-1.2-7 12-8 1.2-9 L2-10

=

A

=

MAAP Run DR0024 DR0040 NA DR0034 DR0031 NA DR0028 DR0042 DR0039 DR0043 lme after Scram when General j

Emeraencyr i8 declared 60 min 15 hr NA 1.1 hr 15 hr NA 45 min 15 hr 20 min 60 min Fission Product Group:

1) Noble

_.i Total ReleU% a 36 Hours 95 100 NA 1_00 l

NA l 86 94 100 0.33 Start of Release (hr) 4.1 47.5 NA 1.1 37.8 NA 5.7 5.7 0.28 3

End of Release (hr) 4.1 55 NA 4

45 NA 6

6 2

36 2)-Csl Total Release % at 36 Hours 23 35 NA 1.7 1.8 NA 0.35 0.22 96 6.30E-04 Start of Release (hr) 4.1 47.5 NA 1.9 37.8 NA 5.7 5.7 0.28 3

End of Release (hr) 4.1 60 NA 2

45 NA 5.7 11 2

6

3) TeO2 Total Release % at 36 Hours 18 27 NA 1.6 0.9 NA 0.48 0.39 78 3.20E-05 Start of Release (hr) 4.1 55 NA 1.9 37.8 NA 5.7 5.7 0.28 3

End of Release (hr) 8 65 NA 4

45 NA 5.7 8

2 6

4)SrO 7 _

Total Release % at 36 Hours 3.1 3.1 NA 5.8 0.7 NA 4.4 3.40E-03 4.6 3.20E-05 Start of Release (hr) 4.1 55 NA 7

55 NA 5.7 7

0.28 6

End of Release (hr) 7 60 NA 10 60 NA 8

10 8

8

5) MoO2 Total Release % at36 Hours 9.OOE-04 5.20E-02 NA 0-027 3.00E-03 NA 4.50E-04 1.80E-03 1.9 2.30E-07 Start of Release (hr) 4.1 47.5 NA 1.9 37.8 NA 5.7 12 0.28 3

EndofRelease(hr) 4.1 47.5 NA 2

40 NA 16 16 2

6

6) C s0H I

__I_

Total Release%at36Hours 27 31 NA 2.8 0.8 NA 1.3 1.5 78 2.20E-04 Start of Release (hr) 4.1 55 NA 1.9 37.8 NA 5.7 l

5.7 0.28 l

3 EndofRelease(hr 11 65 NA 2

45 NA 14 18 2

6 18

Table 3-4 Source Terms Associated with Each Release Category Release Category 2" lI ~

~

~~

.I

.I H

1.2-i L2-2 1-2-3 1-2.4 1-2-5 12.6 1-2.7 1.2-B 1-2-9 "2-In MAAP Run DR0024 DR0040 NA DR0034 DR0031 NA DR0028 DR0042 DR0039 DR0043 Time after Scram when General EmergencyI=declared 60 min 15 hr NA 1.1 hr 15 hr NA 45 min 15 hr 20 min 60 min Fission Product Grop:

I___

7) BaO Total Release % at 36 Hours 1.4 1.4 NA 2.5 0.3 NA 1.9 6.80E-03 4.7 1.30E-05 Start of Release (hr) 4.1 55 NA 7

55 NA 5.7 8

0.28 6

End of Release (hr) 7 60 NA 10 60 NA 8

14 8

8

8) a203 Total Release % at 36 Hours 0.4 0.32 NA 0.62 4.OOE-02 NA 0.65 3.10E-03 0.6 6.70E-06 Start o Release (hr) 4.1 55 NA 7

55 NA 5.7 8

0.28 6

End of Release (hr) 7 60 NA 10 60 NA 8

12 8

8

9) CeO2 Total Release % at 36 Hours 2.1 1.9 NA 2.8 0.3 NA 2.3 0.023 2.2 1.60E-05 Start of Release (hr) 4.1 55 NA 7

55 NA 5.7 8

5 6

End of Release (hr) 7 60 NA 10 60 NA 8

12 8

8

10) Sb Total Release % at 36 Hours 74 43 NA 26 21 NA 20 19 88 7.50E-04 Start of Release (hr) 4.1 55 NA 1.9 37.8 NA 5.7 8

0.28 3

End of Release (hr) 14 70 NA 12 70 NA 14 18 4

6 1 1) T e2__

Total Release % at 36 Hours 1.4 1.1 NA 0.52 1.30E-01 NA 0.31 0.38 0.3 1.OOE-05 Start of Release (hr) 4.1 55 NA 7

55 NA 5.7 8

5 6

End of Release (hr) 14 60 NA 8

60 NA 8

12 20 8

12) U02 Total Release % at 36 Hours 1.001E-02 9.00E-03 NA 1.70E-02 1.OOE-03 NA 2.OOE-02 3.40E-03 1.50-02 1.60E-07 StartofRelease(hr) 4.1 55 NA 7

4 55 NA 5.7 8

5 6

End ofRelease r) 6 55 NA 10 60 8

14 8

8 (1) Puff releases are denoted In the table by those entries with equivalent start and end times.

(2) All cases run for 36 hours4.166667e-4 days 0.01 hours 5.952381e-5 weeks 1.3698e-5 months except DR0040 and DR0031 run for 72 hours8.333333e-4 days 0.02 hours 1.190476e-4 weeks 2.7396e-5 months 19

Mapping of Level 1 Results into the Various Release Cateaories One link between the Level 1 PSA accident sequences and the Containment Event Tree occurs in the definition of the Level 1 end states. The definition of the end states are developed to transfer the maximum amount of information regarding the accident sequence characteristics to the CET assessment. What follows summarizes the link between Level 1 end states and the entry condition to the CET such that a mapping of the Level 1 results into the various release categories can be provided.

A broad spectrum of accident sequences have been postulated that could lead to core damage and potentially challenge containment. The Dresden Level 1 PSA has calculated the frequency of those accident sequences that contribute to the core damage frequency for Dresden using system oriented (systemic) event trees. Each of these sequences may result in different challenges to containment.

However, many of these challenges to containment have similarities in their functional failure characteristics. This has been confirmed in individual BWR PRAs including NUREG-1 150.

The result is that these studies have categorized these containment challenges into a finite, discrete group of accident sequence bins, which have similar functional failures.

As pointed out in past BWR PRAs, different portions of the spectrum of postulated core damage accidents represent substantially different challenges to the containment depending upon the system failures and phenomena that have contributed to the sequence. Therefore, the containment event tree response must be capable of reflecting the entire spectrum of challenges to ensure that the following are explicitly incorporated:

System failures in the Level 1 evaluation (including support systems)

Phenomenological interaction due to the type of core melt progression RPV conditions Pressures Decay heat level Containment conditions Timing of the sequence of events (i.e., core damage and containment failure (if applicable)).

Core Damage Functional Classes An event sequence classification into five accident sequence functional classes can be performed using the functional events as a basis for selection of end states.

The description of functional classes is presented here to introduce the terminology to be used in characterizing the basic types of challenges to containment. The reactor pressure vessel condition and containment condition for each of these classes at the time of initial core damage is noted in Table 3-5.

20

Table 3-5 Core Damage Functional Classes (from the Level 1 Analysis)

Core Damage RPV Condition Containment Functional Class Condition Loss of effective coolant inventory (includes high and Intact low pressure inventory losses)

II Loss of effective containment pressure control, e.g.,

Breached or Intact heat removal III LOCA with loss of effective coolant inventory makeup Intact IV.

Failure of effective reactivity control Breached or Intact V

LOCA outside containment Breached (bypassed)

In assessing the ability of the containment and other plant systems to prevent or mitigate radionuclide release, it is desirable to further subdivide these general functional categories. In the second level binning process, the similar accident sequences grouped within each accident functional class are further discriminated into subclasses such that the potential for system recovery can be modeled. The interdependencies that exist between plant system operation and the core melt and radionuclide release phenomena are represented in the release frequencies through the binning process involving these subclasses, as shown in past PRAs and PRA reviews.

The binning process, which consolidates information from the systems' evaluation of accident sequences leading to core damage in preparation for transfer to the containment-source term evaluation, involves the identification of 18 classes and subclasses of accident sequence types. Table 3-6 provides a description of the possible subclasses used in the Dresden analysis.

The Accident Class designators and subclasses listed in Table 3-6 represent the core damage endstate categories from the Level 1 analysis that are grouped together as entry conditions for the Level 2 analysis. Each of the subclasses is then represented by a series of Containment Event Trees (CETs) to determine the Release Categorization for each of the accident scenarios. As such, the end states from the Level 2 analysis are assigned to one of the Release Categories noted in Table 3-3 as part of Response 3(b).

The characterization of the Level 2 results (i.e., as H/E, WI, etc., or Class V or OK) was then used to determine the frequency of the associated Consequence Category shown in Table 4-5 of the ER. Note that in this fashion, the Level 1 results are not directly linked to a release category, but rather the Level 2 endstate results based on the sum of all of the Release Category frequencies comprise the Consequence Category for each Phase II SAMA considered.

21

Table 3-6 Summary of the Core Damage Accident Sequence Subclasses WASH-1400 Accident Class Subclass Definition Designator Designator Example Class I A

Accident sequences involving loss of inventory makeup TOUX in which the reactor pressure remains high.

B Accident sequences involving a station blackout and TEQUV loss of coolant inventory makeup.

C Accident sequences involving a loss of coolant TT'CMQU inventory induced by an ATWS sequence with containment intact.

D Accident sequences involving a loss of coolant TQUV inventory makeup in which reactor pressure has been successfully reduced to 200 psi.; i.e., accident sequences initiated by common mode failures disabling multiple systems (ECCS) leading to loss of coolant inventory makeup.

E Accident sequence involving loss of inventory makeup in which the reactor pressure remains high and DC power is unavailable.

Class 11 A

Accident sequences involving a loss of containment TW heat removal with the RPV initially intact; core damage induced post containment failure L

Accident sequences involving a loss of containment AW heat removal with the RPV breached but no initial core damage; core damage after containment failure.

T Accident sequences involving a loss of containment N/A heat removal with the RPV initially intact; core damage induced post high containment pressure V

Class IIA or IL except that the vent operates as 1W designed; loss of makeup occurs at some time following vent initiation. Suppression pool saturated but intact.

22

Table 3-6 Summary of the Core Damage Accident Sequence Subclasses T

=

WASH-1400 Accident Class Subclass Definition Designator Designator Example Class Ill A

Accident sequences leading to core damage conditions R

(LOCA) initiated by vessel rupture where the containment integrity is not breached in the initial time phase of the accident.

B Accident sequences initiated or resulting in small or SIQUX medium LOCAs for which the reactor cannot be depressurized prior to core damage occurring.

C Accident sequences initiated or resulting in medium or AV large LOCAs for which the reactor is at low pressure and no effective injection is available.

D Accident sequences which are initiated by a LOCA or AD RPV failure and for which the vapor suppression system is inadequate, challenging the containment integrity with subsequent failure of makeup systems.

Class IV A

Accident sequences involving failure of adequate TTCMC 2 (ATWS) shutdown reactivity with the RPV initially intact; core damage induced post containment failure.

L Accident sequences involving a failure of adequate N/A shutdown reactivity with the RPV initially breached (e.g., LOCA or SORV); core damage induced post containment failure.

T Accident sequences involving a failure of adequate N/A shutdown reactivity with the RPV initially intact; core damage induced post high containment pressure.

V Class IV A or L except that the vent operates as N/A designed; loss of makeup occurs at some time following vent initiation. Suppression pool saturated but intact.

Class V Unisolated LOCA outside containment N/A The CET calculation for each cutset uses Boolean logic and fault tree models to process the incoming Level 1 cutsets to ensure that the resulting Radionuclide release frequencies properly reflect the impact on release magnitude and timing of the containment and containment mitigation systems. A typical CET (for Accident Class 1A) is provided in Figure 3-1.

23

Figure 3-1 Typical Dresden Level 2 Containment Event Tree 25

In summary, the Level 1 end states do not translate directly into release categories.

Each Level 1 accident sequence (all of the cutsets) is transferred into the appropriate CET. The CET is then used to determine the resulting frequency for each radionuclide release end state from each incoming cutset. This is typical of a full Level 2 for a binned fault tree model. This approach does not involve a matrix that relates Level 1 sequences directly to Radionuclide end states.

Although not created as part of the normal calculation process, the results of the analysis can be binned to show the contribution to each release category by Level 1 end state. Table 3-7 shows the requested results for the base case 2002 model.

Table 3-7 Matrix of Level I Results with Various Release Categories Base Case (2002 Model)

Level 2 Release Category I Level 3 Consequence Category Level l LA Accidet HtE HW VL un.

un WL 4 LE or Ll, or Class V Intact Total Class (12-1)

(L2-2)

(12-3)

(L2-4)

(L2-5)

(12-6)

L27 LL (L2-9) (12-10)

____I L27 (12-8)

IAtI E 1.04E-07 NA 9.49E-09 1.61 E-08

.22E-08 N/A 2.72E-09 5.43E-08 NtA 8.41 E-07 1.06E-06 IBE 9.82E-09 N/A O.OOE+O O.OOE+00 9.05E-08 O.OOE+O0 4.59E-10 1.23E-09 N/A 1.98E-07 3.01E-07 1BL N/A 4.64E-09 O.OOE+0O N/A 5.79E-08 O.OOE+00 N/A 0.OOE+00 N/A 4.55E-08 1.08E-07 IC O.OOE+00 N/A O.OOE+00 O.OOE+0O O.OOE+O0 O.OOE+00 O.OOE+O0 O.OOE+0O N/A 1.79E-08 1.79E-08 1 D O.OOE+O N/A O.OOE+00 O.OOE+0O 1.88E-08 NA O.OOE+O0 2.03E-09 N/A 4.77E-09 2.56E-08 2 O.OOE+00 6.25E-10 N/A 1.15E-09 7.97E-08 N/A N/A N/A N/A O.OOE+00 8.15E-08 3B O.OOE+00 O.OOE+O0 O.OOE+O N/A O.OOE+00 O.OOE+0O 1.09E-10 O.OOE+0O NtA 1.52E-08 1.53E-08 3C 8.17E-08 N/A N/A NtA N/A N/A N/A NIA N/A 3.49E-10 8.20E-08 3D 1.18E-08 N/A N/A N/A N/A N/A N/A N/A N/A O.OOE+0O 1.18E-08 4A 9.41E-08 N/A N/A 9.15E-08 N/A N/A N/A N/A N/A O.OOE+oo 1.86E-07 5

N/A N/A N/A I

N/A N/A N/A N/A N/A 1.74E-09 O.OOE+OO 1.74E-09 Total:

3.01E-07 5.26E-09 9.49E-09 1.09E-07 2.79E-07 Q.OOE+OO 3.29E-09 5.76E-08 1.74E-09 1.12E46 1.89E06 (1) Included with the H/i Consequence Category (12-2) for evaluation purposes.

(2) Included with the MA Consequence Category (12-5) for evaluation purposes.

26

RAI4 Provide the following information concerning the MELCOR Accident Consequences Code System (MACCS) analyses:

a. the MACCS analysis assumes all releases that occur at ground level and have a thermal content the same as ambient. These assumptions could be non-conservative when estimating offsite consequences.

Provide an assessment of the sensitivity of offsite consequences (doses to the population within 50 miles) to these assumptions.

b. the discussion of meteorology indicates that there are data voids in the 2000 data set used Interpolation was used between hours if only a brief period of data was missing, and hourly observations from the airport were used to fill larger data voids. Provide a characterization of the magnitude and extent of the data voids and the rationale for using the airport data rather than interpolation. Confirn that the 2000 data set is representative of the DNPS site andjustify its use.

c. clarify the time periods used for am and pm for the atmospheric mixing heights (e.g., midnight to noon and noon to midnight, versus sunrise to sunset).

Response 4(a):

MThe MACCS analysis assumes a releases that occur at ground level and have a thermal content the same as ambient. These assumptions could be non-conservative when estimating offsite consequences.

Provide an assessment of the sensitivity of offsite consequences (doses to the population within 50 miles) to these assumptions."

MACCS2 was re-run for all 8 sequences assuming that all plumes originated from the top of the reactor building, at an elevation of 141 feet above grade, rather than ground level (top of reactor building at 658 feet, grade at 517 feet above sea level). Table 4-1 shows the increases that were obtained for each sequence.

As can be seen, the calculated dose increase from the elevated release case compared to the ground level release case leads to an increase in the dose of between 4% and 8%. The cost associated with each consequence category went up by about 5-18% except for the containment intact case where a reduction in cost occurred. The overall impact using the same assumptions that were utilized in the ER is a $27,952 increase (+6.1%) in the calculated maximum averted cost risk.

It is judged that this would not change the results of the SAMA analysis.

27

Table 4-1 Ratio of Dose Results (Elevated to Ground-Level Releases)

Consequence MAAP Run Dose Cost Category L2-1 DR0024 1.04 1.08 12-2 DR0040 1.06 1.05 12-4 DR0034 1.06 1.10 12-5 DR0031 1.07 1.18 L2-7 DR0028 1.06 1.10 L2-8 DR0042 1.05 1.18 L2-9 DR0039 1.05 1.06 12-10 DR0043 1.08 0.73 Response 4(b):

c7The discussion of meteorology indicates that there are data voids in the 2000 data set used Interpolation was used between hours if only a brief period of data was missing, and hourly observations from the airport were used to ill larger data voids. Provide a characterization of the magnitude and extent of the data voids and the rationale for using the airport data rather than interpolation.

Confirm that the 2000 data set is representative of the DNPS site andjustify its use."

The year 2000 meteorological data sets for QCNPS and DNPS were selected due to the fact that they had the least number of data voids (compared to 1998, 1999 and 2001).

The year 2000 DNPS meteorological data set had a total of 14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months of missing data.

Of these 14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months , no more than two consecutive hours were missing. All gaps in the year 2000 meteorological data set for DNPS were filled by using interpolation methods.

Due to the rather small extent of the data voids, it is believed that the data set is representative of the DNPS site.

28

Response 4(c):

"Clarify the time perods used for am and pm for the atmospheric mixing heights (e.g.,

midnight to noon and noon to midnight, versus sunrise to sunset).

The original source (George C. Holworth, 'Mixing Heights, Wind Speeds, and Potential for Urban Air Pollution throughout the Contiguous United States," USEPA Office of Air Programs, January 1972) did not use the words 'am' or pm', but actually referred to

'moming' and "afternoon' mixing heights. This source defined morning as being the four-hour period from 0200 to 0600 Local Standard Time and afternoon as being the four-hour period from 1200 to 1600 Local Standard Time.

The Code Manual for MACCS2: Volume 1 (from Appendix B, page B-2) states the following:

The first of these two values corresponds to the morning mixing height and the second to the afternoon height. In the current implementation, the larger of these two values and the value of the boundary weather mixing height is used by the code.'

min its present form, that atmospheric model implemented in MACCS2 does not allow a change in the mixing layer to occur during transport of the plume. Mixing layer height is assumed to be constant and therefore only a single value is used by the code."

Since the Dresden MACCS2 analyses considered plumes that have durations in excess of 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months (some as long as 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months ), these conditions mean that, for all intents and purposes, only the afternoon mixing height is used since it is always larger than the morning mixing height. Note that the boundary weather mixing height, wind speed and stability category are only used when there is no met data file. These fixed values are ignored by the code when an hourly met data file is supplied by the user, as was the case in the MACCS2 runs for Dresden.

29

RAI5 According to Table F-1 of the Environmental Report (ER), Exelon evaluated 265 SAMA candidates.

Of these 265 candidates, 21 were obtained from DNPS-specific documents. It is not clear that the set of SAMAs evaluated in the ER addresses the major risk contributors for DNPS. In this regard, provide the following:

a. a description of how the dominant risk contributors at DNPS, including dominant sequences and cut sets from the current Probabilistic Risk Assessment (PRA) and equipment failures and operator actions identified through importance analyses (e.g., Fussell-Vesely, Risk Reduction Worth, etc.) were used to identify potential plant-specffic SAMAs for DNPS.

b. the number of sequences and cut sets reviewed/evaluated and what percentage of the total CDF they represent.

c. a listing of equipment failures and human actions that have the greatest potential for reducing isk at DNPS based on importance analysis and cut set screening.

d. for each dominant contributor identified in the current PRA (2002 Update), a cross-reference to the SAMAs evaluated in the ER which addresses that contributor.

If a SAMA was not evaluated for a dominant isk contributor, then justify why SAMAs to further reduce these contributors would not be cost beneficial

e. the reasons for the difference in the number of SAMAs evaluated for Quad Cities Nuclear Power Station (QCNPS) and DNPS (280 v. 265).

f. a general description of the group of 130 insights mentioned in the original IPE and a discussion of how and whether the insights that were not implemented were factored into the SAMA evaluation.

Response 5(a):

'[Provide a description of how the dominant risk contributors at DNPS, including dominant sequences and cut sets from the current Probabilistic Risk Assessment (PRA) and equipment failures and operator actions identified through importance analyses (e.g., Fussell-Vesely, Risk Reduction Worth, etc.) were used to identify potential plant-specific SAMAs for DNPS."

A review of the CDF-based Risk Reduction Worth (RRW) rankings for the current model was performed.

The rankings of these equipment failures, operator actions, and initiating events were checked to determine if any items could be beneficial that were not addressed by the existing SAMA list. The examination of the dominant RRW basic 30

events encompassed the dominant sequences and cut sets from the current PRA model.

RAI response 5(d) provides a more detailed discussion of this importance ranking review.

Response 5(b):

"[Provide] the number of sequences and cut sets reviewed/evaluated and what percentage of the total CDF they represent.'

The CDF-based RRW listing was reviewed down to and including the 1.01 level, which indicates the events below this point would influence the CDF by less than 1.0%. This corresponds to about a $4000 averted cost-risk based on CDF reduction assuming 100% reliability of the associated event.

An evaluation of the top LERF-based contributors to RRW was also performed. It was determined that a similar averted cost of about $4000 would be obtained by examining the LERF-based RRW factors down to a value of 1.03.

RAI response 5(d) provides a more detailed discussion of the importance ranking review and the results.

Response 5(c):

'[Provide) a listing of equipment failures and human actions that have the greatest potential for reducing risk at DNPS based on importance analysis and cut set screening."

RAI response 5(d) provides a listing of equipment failures, human actions, and initiating events that have the greatest potential for reducing risk at DNPS based on importance analysis and cut set screening.

Response 5(d):

[Provide] for each dominant contributor identified in the current PRA (2002 Update), a cross-reference to the SAMAs evaluated in the ER which addresses that contributor. If a SAMA was not evaluated for a dominant risk contributor, then justify why SAMAs to further reduce these contributors would not be cost beneficial" Table 5-1 (for CDF) and Table 5-2 (for LERF) provide a correlation between the events identified in the DNPS PSA model (2002 Update) that are considered to have the greatest potential for reducing risk and their relationship to the SAMAs evaluated in the Environmental Report.

The events included in Table 5-1 are based on the core damage frequency RRW factors down to and including RRW values of 1.01. The events included in Table 5-2 are based on the large early release frequency RRW factors down to an RRW value of 1.03. Both of these RRW factors correspond to potential averted cost risk of about

$4000. The events below this point are judged to be highly unlikely contributors to the identification of cost-beneficial enhancements.

31

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition F-ICALONE 1.OOE+00 1.87 FLAG: IC FAILURE NOT This event represents a sequence marker flag identifying those CAUSED BY FAILURE OF scenarios with Isolation condenser failures. The isolation REACTOR VESSEL MAKE-condenser can provide level control and decay heat removal.

UP Many SAMAs were considered that explored alternate injection and decay heat removal capabilities. No additional SAMAs were suggested.

O-AD-MU1 1.10E-04 1.47 2ADOP-ACT-ADSH--

This event represents the unlikely scenario of combined 21COP-IC-MU1-H--

operator action failures for separate actions that otherwise are evaluated independently. This event is included for completeness as part of the human reliability dependency analysis. Phase I SAMAs 250 and 255 examine potential improvements in operator performance. No additional SAMAs were suggested for this topic.

%LOOP 3.09E-02 1.35 INIT: LOSS OF OFFSITE This event is a single unit loss of offsite power event.

POWER Improvements related to enhanced AC or DC reliability or availability were considered in Phase I SAMAs 90 through 129.

Many other SAMAs were also considered that would provide mitigation benefits in loss of offsite power scenarios including Phase II SAMAs 1, 2, 6, and 10. No additional SAMAs were suggested for this broad topic.

%TF 4.47E-02 1.28 INIT: TRANSIENT WITH FW This event represents the loss of feedwater initiating event UNAVAILABLE AND MC frequency. Industry efforts over the last fifteen years have led AVAILABLE to a significant reduction in the number of plant scrams from all causes. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

2HI-SYSTEMUA-M--

2.13E-02 1.19 HPCI SYSTEM MUA This event represents the probability of the HPCI system in maintenance. Potential improvements to enhance high pressure injection capabilities were considered in Phase I SAMAs 19,178, 179,185,189, 193, 196,198, 201, 203, and 204. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

32

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%DLOOP 9.41 E-03 1.18 INIT: DUAL LOSS OF This event represents the dual unit loss of offsite power initiating OFFSITE POWER event frequency. See disposition above for %LOOP (Single Unit Loss of Offsite Power).

BDCBY125----FCC 4.93E-06 1.14 COMMON CAUSE FAILURE This event represents the common cause failure of the 125V OF UNIT 2 AND UNIT 3 DC batteries. Many SAMAs were Included that address 125VDC BATTERIES potential enhancements for DC reliability and/or alternate (B=9.86E-03) means of providing DC power. Phase I SAMAs 92, 93, 96, 97, 98, 99, 113, 124,125, 126, 127, and 128 are all related to improved DC performance. No additional SAMAs were suggested for this broad topic.

2RPCDRPS-MECHFCC 2.10E-06 1.12 RPS MECHANICAL This event represents the Mechanical Scram failure probability FAILURE based on the NUREG/CR-5500 INEEL evaluation of a representative BWR RPS system. Potential improvements to minimize the risks associated with ATWS scenarios were explored in Phase I SAMAs 213-227, 259, and 260. Phase I SAMAs 259 and 260 were retained as Phase II SAMAs 7 and 8, respectively. No additional SAMAs were suggested for this broad topic.

%TT 1.81 E+00 1.12 INIT: TRANSIENT WITH FW This event represents the turbine trip initiating event frequency.

AND MC AVAILABLE Industry efforts over the last fifteen years have led to a significant reduction in the number of reactor scrams and turbine trips. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

2PLSV-F-RECL-K--

1.50E-01 1.11 FAILURE OF SRVs TO This event represents the likelihood that the SRVs will not RECLOSE ON REDUCED reclose after initially sticking open in response to a pressure PRESSURE transient. The failure value of 0.15 is based on limited industry evidence. See disposition for 2PLSVSORV-NTTK-- (Probability of SORV for non-turbine trip initiators) below.

33

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2PLSVSORV-NTTK--

5.40E-02 1.10 PROBABILITY OF SORV This event represents the likelihood that an SRV will stick open FOR NON TT INITIATORS in response to a pressure transient. For Dresden, this renders the isolation condenser Ineffective. Consequently, early injection from HPCI or depressurization for low pressure injection Is required for success. Many SAMAs considered potential benefits from improved injection capabilities or improved depressurization capabilities. No additional SAMAs were suggested.

%TDC 1.50E-06 1.09 INiT: LOSS OF MULTIPLE This event represents the unlikely initiating event of a complete DC BUSES loss of both 125V DC buses. Many SAMAs were included that address potential enhancements for DC reliability and/or alternate means of providing DC power. Phase I SAMAs 92, 93, 96, 97, 98, 99, 113, 124, 125, 126, 127, and 128 are all related to improved DC performance. No additional SAMAs were suggested for this broad topic.

2DCRX-BUS2RECF-7.10E-01 1.09 DC BUS 2 FAIL TO This event involves failure to recover one of the 125V DC buses RECOVER (GIVEN LOSS given loss of both. See disposition above for %TDC (Loss of OF MULTIPLE DC BUSES Multiple 125V DC Buses Initiating Event).

INITIATOR %TDC) 3DCRX-BUS3RECF--

7.1 OE-01 1.09 DC BUS 3 FAIL TO This event involves failure to recover one of the 125V DC buses RECOVER (GIVEN LOSS given loss of both. See disposition above for %TDC (Loss of OF MULTIPLE DC BUSES Multiple 125V DC Buses Initiating Event).

INITIATOR %TDC)

F-BUS241 1.OOE+00 1.08 FLAG: LOSS OF POWER AT This event Is a sequence marker flag for Bus 24-1 failures.

BUS 24-1 Improvements related to enhanced AC or DC reliability or availability were considered in Phase I SAMAs 90 through 129.

Many other SAMAs were also considered that would provide mitigation benefits in loss of offsite power scenarios including Phase II SAMAs 1, 2, 6, and 10. No additional SAMAs were suggested for this broad topic.

34

Table 5-1 Correlation of COF Importance Listing to Evaluated SAMAs Event Name Probability RRW Baslc Event Description Disposition F-BUS231 1.OOE+0O 1.08 FLAG: LOSS OF POWER AT This event is a sequence marker flag for Bus 23-1 failures. See BUS 23-1 disposition above for F-BUS241 (Sequence marker flag for loss of Bus 24-1).

2ADOP-DEPMADSH--

3.50E-03 1.08 OP ACT: DEPRESS RPV This event represents the human error probability for failing to (MLOCA/SORV) depressurize for low pressure injection given a medium LOCA or SORV event and initial failure of HPCI to inject (thereby requiring depressurization for low pressure injection). Potential improvements to depressurization capabilities were considered in Phase I SAMAs 190, 191, 229, 230, 240, 241, and 247. No additional SAMAs were suggested for this broad topic.

2HITB2301TURBX--

9.60E-03 1.08 HPCI TURBINE FAILS TO This event represents the HPCI turbine failing during its mission RUN time. See disposition above for 2HI-SYSTEMUA-M-- (HPCI system in maintenance).

O-AD-HI-MU1 1.1OE-04 1.02 2ADOP-ACT-ADSH--

This event represents the unlikely scenario of combined 2HIOP-OVRFILLH--

operator action failures for separate actions that otherwise are 2ICOP-IC-MU1 -H--

evaluated independently. This event is included for completeness as part of the human reliability dependency analysis. Phase I SAMAs 250 and 255 examine potential improvements in operator performance. No additional SAMAs were suggested for this topic.

BDGDG-3E-2S--XCC 1.88e-04 1.06 2SBO AND 3EDG FAILURE This event represents the unlikely scenario of a common cause TO RUN CCF failure of the 2SBO and 3EDG. Improvements related to enhanced AC or DC reliability or availability were considered in Phase I SAMAs 90 through 129. Many other SAMAs were also considered that would provide mitigation benefits in loss of offsite power scenarios including Phase II SAMAs 1, 2, 6, and

10. No additional SAMAs were suggested for this broad topic.

RDLOOP4 2.20E-01 1.06 FAILURE TO RECOVER This event represents the probability of not recovering off-site DLOOP WITHIN 4 HOURS power within 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months following a dual unit loss of off-site power.

See disposition above for %DLOOP (Dual Unit Loss of Offsite Power).

35

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2ACBS-UAT-RATF--

1 1.OOE-01 1.04 PROB AC BUS WILL NOT This event represents a pseudo-recovery action in loss of DC TRANSFER TO THE bus or loss of multiple DC bus initiated events. The importance RESERVE AUX of this event would be minimized by reducing the frequency of TRANSFORMER (RAT) loss of DC events. See disposition above for %TDC (Loss of GIVEN LOSS OF DC BUS 2 Multiple 125V DC Buses Initiating Event).

2RXSE-LEAK---L--

1.OOE-01 1.03 PROB REACTOR COOLANT This event represents the likelihood that the recirculation pump LEAKAGE IS SUFFICIENT seals leak sufficiently to require RPV makeup given a loss of TO REQUIRE RPV MAKEUP cooling (e.g., in SBO scenarios). Improvements to the reliability SOURCES of the recirculation pump seals were examined in Phase I SAMA 3 that was retained as Phase II SAMA 1. No additional SAMAs were suggested.

%S1 2.40E-03 1.03 INIT: MEDIUM LOCA This event represents the Medium LOCA initiating event frequency. Mitigation from such-an event would be improved by the existence of more reliable or diverse low pressure injection systems and water sources. Such potential improvements were examined In Phase I SAMAs 177, 184, 187, 194,197, 202, 205, and 208. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

%A 1.90E-04 1.03 INIT: LARGE LOCA This event represents the Large LOCA initiating event frequency. Mitigation from such an event would be Improved by the existence of more reliable or diverse low pressure injection systems and water sources. Such potential improvements were examined in Phase I SAMAs 177, 184, 187, 194, 197, 202, 205, and 208. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

2CAHU25-ABCDLHCC 8.00E-05 1.03 PREINIT: MISCALIBRATE This event represents the unlikely scenario of a common cause LEVEL SWITCHES 263-25 miscalibration of level switches leading to unavailability of the A-B-C & D DUE TO CC -

isolation condenser.

No additional SAMAs are suggested for LOW this topic.

36

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name ProbabIlity RRW Basic Event Description Disposition RDLOOP30 6.40E-01 1.03 FAILURE TO RECOVER This event represents the probability of not recovering off-site DLOOP WITHIN 30 power within 30 minutes following a dual unit loss of off-site MINUTES power. See disposition above for %DLOOP (Dual Unit Loss of Offsite Power).

2HITB-MULT---A--

4.20E-03 1.03 HPCI TURBINE FAILS TO This event represents the HPCI turbine failing when required to START MULTIPLE TIMES start more than once. See disposition above for (1.5 TIMES SINGLE START 2HI-SYSTEMUA-M-- (HPCI system in maintenance).

FAILURE)

%MS 2.68E+00 1.03 INIT: MANUAL SHUTDOWN This event represents the manual shutdown initiating event frequency. Industry efforts over the last fifteen years have led to a significant reduction in the number of manual shutdowns and scrams from all causes. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

%TSW 1.98E-03 1.02 INIT: LOSS OF SERVICE This event is the loss of service water nitiating event. Potential WATER Improvements and enhancements to the service water system were examined in Phase I SAMAs 10, 20, 21, and 23. No additional SAMAs were suggested, and no related SAMAs were retained for Phase II. It is noted that in Phase I SAMA 23, the cost of installing an additional service water pump had been estimated at approximately $5.9 million which is greater than the maximum averted cost risk (even if large uncertainties and external events are considered).

O-AD-CC2-MU1 1.00E-06 1.02 2ADOP-ACT-ADSH--

This event represents the unlikely scenario of combined 2CCOP-CNTC2-H--

operator action failures for separate actions that otherwise are 21COP-IC-MUI-H--

evaluated independently. This event is included for completeness as part of the human reliability dependency analysis. Phase I SAMAs 250 and 255 examine potential improvements in operator performance. No additional SAMAs were suggested for this topic.

37

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2MSOPMSIVINLKH-9.30E-01 1.02 OP ACT: BYPASS LOW This event represents the humanerrorprobabilityof bypassing LEVEL MSIV INTERLOCK the MSIV isolation as directed in the EOPs. This issue was ;

(ATWS) specifically examined in Phase I SAMA 223 that was dispositioned wth reference to existing capabilities. However,,

this action requires the use of jumpers with a limited time availabfe andas such carries a relatively high HEP value.

The potential benefit of Implementing a dedicated low levei interlock switch is explored as part of this RAt response (see

Response

7(c)).

21C-SYS-----M--

7.74E-03 1.02 ISO CNDNSR SYSTEM MUA This event represents the probability that the isolation condenser is in maintenance. The isolation condenser can provide level control and decay heat removal. Many SAMAs were considered that explored alternate injection and decay heat removal capabilities. No additional SAMAs were suggested.

2FW-LDCHIGH-F--

5.00E-02 1.02 CONDITIONAL PROB. OF This event represents the conditional probability of a feedwater FW PUMP TRIP ON HIGH pump trip on high level given a loss of multiple DC bus initiator.

LEVEL As such, the importance of this event would be reduced by minimizing the loss of DC failures. See disposition above for

%TDC (Loss of Multiple 125V DC Buses Initiating Event).

2FW-LDC-LOW-F-5.OOE-02 1.02 CONDITIONAL PROB. OF This event represents the conditional probability of a feedwater FW PUMP TRIP ON LOW pump trip on low level given a loss of multiple DC bus initiator.

LEVEL As such, the importance of this event would be reduced by minimizing the loss of DC failures. See disposition above for

%TDC (Loss of Multiple 125V DC Buses Initiating Event).

2HiTB2301TURBA--

2.80E-03 1.02 HPCI TURBINE FAILS TO This event represents the HPCI turbine failing to start. See START disposition above for 2HI-SYSTEMUA-M-- (HPCI system in maintenance).

38

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2LI--TOP-PEAKF--

2.50E-01 1.02 RX PWR IS TOP PEAKED This event was added to the model in response to a GE SUCH THAT 2/3 RX LVL IS concern about LPCI steam cooling capabilities leading to INSUFF TO COOL CORE elevated temperatures in upper portions of the core.

(CS REQD)

Subsequent clarification on this issue will result in a re-examination of this assumption as part of the next update (i.e.,

LPCI injection with level maintained at 2/3 core height is a success state). No SAMAs were suggested for this issue.

2PVPPWATERBRKR-8.00E-02 1.02 WATER LINE BREAK This event represents the conditional probability of medium MEDIUM LOCA LOCA Initiating events that are water line breaks as opposed to steam line breaks. See disposition above for %S1 (Medium LOCA Initiator).

2SLEV2-1106ABDCC 7.15E-03 1.02 EXPLOSIVE VALVES 2-This event represents the common cause failure of the SLC 1106A&B FAILURE TO system explosive valves. Diversification of the SLC explosive OPEN DUE TO CCF valves was considered in Phase I SAMA 259 which was retained as Phase II SAMA 8. No additional SAMAs were suggested.

BDGDG-3E-2S-ACC 6.32E-05 1.02 2 SBO AND 3EDG FAILURE This event represents the unlikely scenario of a common cause TO START CCF failure of the 2SBO and 3EDG. See disposition above for BDGDG-3E-2S-XCC (2SBO and 3EDG failure to run CCF).

21COP-IC-MU1-H-8.80E-03 1.02 OP ACT: INITIATE IC SHELL This event represents the probability that IC shell side makeup SIDE MAKEUP will not be initiated. The isolation condenser can provide level control and decay heat removal. Many SAMAs were considered that explored alternate injection and decay heat removal capabilities. No additional SAMAs were suggested.

21COP-LODC---H-1.40E-01 1.02 OP ACT: PREVENT LOSS This event represents the probability that the isolation OF IC FOLLOWING condenser will be maintained following battery depletion. The BATTERY DEPLETION isolation condenser can provide level control and decay heat removal. Many SAMAs were considered that explored aftemate injection and decay heat removal capabilities. No additional SAMAs were suggested.

39

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2FWAV3201ABC-DCC 9.65E-04 1.02 RFP RECIRC. (MIN-FLOW) This event represents the unlikely scenario of a common cause VALVES FAIL TO OPEN failure of the RFP min-flow valves rendering Feedwater injection DUE TO COMMON CAUSE unavailable. Potential improvements to enhance high pressure injection capabilities were considered in Phase I SAMAs 19, 178,179, 185, 189, 193, 196,198, 201, 203, and 204. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

2CNPVDWRUPT--R--

6.OOE-02 1.02 LARGE DW CONTAINMENT This event represents the scenario where un-mitigated FAILURE CAUSES LOSS containment pressurization results in a large drywell region OF INJECTION containment failure leading to a loss of all injection systems.

This scenario can be avoided by providing improved decay heat removal methods. Potential improvements for decay heat removal were examined in numerous Phase I SAMAs as well as Phase II SAMAs 2, 3, and 4. No additional SAMAs were suggested for this broad topic.

2H1HU2391-003H--

2.00E-03 1.02 PREINIT: HPCI STM FLOW This event represents a pre-initiator human error that renders MTU 2391-03 the HPCI system unavailable. See disposition above for MISCALIBRATED 2HI-SYSTEMUA-M-- (HPCI system in maintenance).

2H1HU2391-005H--

2.OOE-03 1.02 PREINIT: HPCI STM FLOW This event represents a pre-initiator human error that renders MTU 2391-05 the HPCI system unavailable. See disposition above for MISCALIBRATED 2HI-SYSTEMUA-M- (HPCI system in maintenance).

2HIKV2301-074D--

2.OOE-03 1.02 STOP CHECK VALVE 2-This event represents a valve failure that prevents HPCI system 2301-74 FAILS TO OPEN injection. See disposition above for 2HI-SYSTEMUA-M- (HPCI system in maintenance).

2HIPM2301-AOPA--

2.OOE-03 1.02 AUXILIARY OIL PUMP This event represents an auxiliary failure of the HPCI system.

FAILS TO START See disposition above for 2HI-SYSTEMUA-M- (HPCI system in maintenance).

2HIPM2301 GSCPA--

2.OOE-03 1.02 GSLO CONDENSATE PUMP This event represents an auxiliary failure of the HPCI system.

FAILS TO START See disposition above for 2HI-SYSTEMUA-M-- (HPCI system in maintenance).

40

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition XCSLISUCSTR-FCC 1.OOE-04 1.01 COMMON CAUSE This event represents the unlikely occurrence of a common PLUGGING OF ECCS cause failure of the ECCS suctionstrainers. The Dresden SUCTION STR DURING strainers have recently been upgraded and re-sized such that LOCAS the potential for common cause plugging has been reduced.

No additional SAMAs were suggested.

I,'

SW-CCSW-FACTOR 2.70E-03 1.01 LOSW IE PERCENT This event represents the fraction of loss of SW initiating events FAILING CCSW DUE TO that will also lead to a loss of CCSW due to common causes.

COMMON EFFECTS See disposition above for %TSW (Loss of Service Water Initiating Event).

2CAHU-52-A-B2HCC 8.00E-05 1.01 PREINIT: MISCALIBRATE This event represents the unlikely scenario of a common cause CAS PRESSURE miscalibration of pressure switches leading to unavailability of SWITCHES 52A AND 52B ECCS injection (i.e. failure of RPV low pressure permissive DUE TO CC interlock). This is included for completeness in the model since it has the potential of leading to core damage following a medium or large LOCA initiating event. No additional SAMAs are suggested for this topic.

2ECOP-OCST---H-1.OOE-01 1.01 OP ACT: ALIGN LO PRESS This event epresents the human error probability as ECCS PUMP SUCTION(S) with failure to align ECCS pump suction to the OST. Outset TO CST review ndicates that this action Is important in loss of seave water initiated events. This Idea was considered In Phase 1,

SAMA 188. The potential benefit from improving theHEP value associated with this existing action is explored as patof thisv RAI response (see Response 7(c)).

2HIHU026325AHH--

2.00E-03 1.01 PREINIT: RX Hi LEVEL LIS This event represents a pre-initiator human error that renders 263-25A3 MISCALIBRATED the HPCI system unavailable. See disposition above for

- HIGH 2HI-SYSTEMUA-M- (HPCI system in maintenance).

2HIHU026325BHH--

2.OOE-03 1.01 PREINIT: RX HI LEVEL LIS This event represents a pre-initiator human error that renders 263-25B3 MISCALIBRATED the HPCI system unavailable. See disposition above for

- HIGH 2HI-SYSTEMUA-M-- (HPCI system in maintenance).

41

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%FLSWRB545 6.10E-05 1.01 INIT: SW FLOOD IN RB This event represents the initiating event frequency for a SW ABOVE 545' flood in the reactor building above the 545' elevation. Potential improvements to reduce internal flooding frequency were considered in Phase I SAMAs 153-158. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

%FLSWTB 1.43E-03 1.01 INIT: SW RUPTURE IN TB This event represents the initiating event frequency for a SW rupture in the turbine building. Potential improvements to reduce internal flooding frequency were considered in Phase I SAMAs 153-158. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

2SWPP-RB-UN--R--

9.00E-01 1.01 BREAK IN USILOABLE SW This event represents the conditional probability that the SW PIPE IN RB rupture is not isolatable. See disposition above for %FLSWTB (SW Rupture in turbine building initiating event).

%S2 2.90E-03 1.01 INIT: SMALL BREAK LOCA This event represents the small break LOCA initiating event frequency. Many SAMAs investigated improvements to Improved injection or containment heat removal capabilities that would reduce the importance of this event. No additional SAMAs were suggested.

BDCBS2M&3M---FCC 1.13E-07 1.01 MAIN DC BATTERY BUSES This event represents the unlikely scenario of a common cause 2 AND 3 CCF failure of both main DC battery buses. See disposition above for %TDC (Loss of Multiple 125V DC Buses Initiating Event).

42

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2RXSY-RXFAIL-F--

1.OOE+O0 17.68 FAILURE OF RX (CLASSES This event Is a Level 2 sequence marker flag identifying those ID, IE (OP=F), II, ilA, IIIC, sequences where the RX node has failed (i.e., where core 1I1D, IV) damage was not terminated prior to the time of vessel failure).

The capability to enhance or provide additional Injection systems was examined in Phase I SAMAs 19,177, 178,179, 184,185,.187,189,193,194,196,197,198, 201-205, and 208.

No additional SAMAs were suggested.

2GVPH-INERT--X--

9.90E-01 10.45 CONTAINMENT INERTED; This event is effectively a Level 2 sequence marker flag that VENTING NOT REQUIRED represents the normal operating condition with the containment inerted. No additional SAMAs were suggested.

2SIPHCONTFAILF-1.OOE+00 2.28 DW SHELL MELT-THROUGH FAILURE DUE TO CONT. FAILURE This event represents the evaluated likelihood from the Level 2 analysis that a dry containment floor will lead to containment failure after vessel failure for accident classes 11, IIID, and IV.

The importance of this phenomena would be reduced by the presence of more reliable or diverse injection systems, more reliable or diverse drywell spray systems, and other altemate means to avoid this situation. SAMAs related to improved injection system performance are discussed in the disposition for 2RXSY-RXFAIL-F-- above. Items related to improved drywell spray performance were considered in Phase I SAMAs 35, 36, 52, 54, and 82. Phase I SAMA 35 was retained as Phase II SAMA 3. Aftemate strategies for reducing the potential for drywell shell melt-through were also examined in Phase I SAMAs 43, 44, 47, 48, 50, 56, 57, and 86. None of these, however, were retained for Phase II, and no additional SAMAs were suggested.

2OPPH-NOLOCA-F--

LOCA NOT INDUCED VIA HIGH TEMP, HIGH PRESSURE, OR SORV This event represents a Level 2 phenomena event that would lead to a depressurized state. See disposition below for 20POP-DEPRESSH-- (Operator fails to depressurize in Level 2 given failed in Level 1).

43

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2OPOP-DEPRESSH--

5.20E-01 1.49 OP FAILS TO DEPRESS This event represents the conditional failure probability used in GIVEN OP FAILED IN LVL1 the Level 2 analysis for operators to depressurize prior to vessel OR LOSS OF DC failure given that depressurization was unsuccessful to avert core damage. Potential improvements to the current depressurization capabilities and methods were examined in Phase I SAMAs 190, 191, 229, 230, 240, 241, and 247. None of these, however, were retained for Phase II, and no additional SAMAs were suggested.

2DIDW-ATWSSEQFSU 9.90E-01 1.43 DW INTACT FOR ATWS This event is effectively a Level 2 sequence marker flag that EVENTS (CLASS IV) represents the drywell status at the time of core damage given an ATWS scenario. Note that the evaluated likely failure location for ATWS scenarios is In the wetwell. No additional SAMAs were suggested.

2RPCDRPS-MECHFCC 2.10E-06 1.43 RPS MECHANICAL This event also appears in the CDF importance listing in Table FAILURE 5-1. It represents the Mechanical Scram failure probability based on the NUREG/CR-5500 INEEL evaluation of a representative BWR RPS system. Potential improvements to minimize the risks associated with ATWS scenarios were explored in Phase I SAMAs 213-227, 259, and 260. Phase I SAMAs 259 and 260 were retained as Phase II SAMAs 7 and 8, respectively. No additional SAMAs were suggested for this broad topic.

2CNWW-ATWSSEQF--

5.00E-01 1.42 WW WATER SPACE FAIL.

This event represents the evaluated likelihood that an ATWS FOR ATWS EVENTS scenario with containment failure in the wetwell is located below (CLASS IV) the normal torus water level. Its' importance would be minimized by reducing the potential for ATWS scenarios. See disposition above for 2RPCDRPS-MECHFCC (RPS mechanical failure).

2OPPH-OP8-NOTFSU 9.69E-01 1.41 SUCCESSFUL RPV This event represents the evaluated likelihood that successful DEPRESSURIZATION RPV depressurization occurs in an ATWS. Its' Importance (CLASS IV) would be minimized by reducing the potential for ATWS scenarios. See disposition above for 2RPCDRPS-MECHFCC (RPS mechanical failure).

44

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%TT 1.81 E+00 1.34 INIT: TRANSIENT WITH FW This event also appears in the CDF importance listing in Table AND MC AVAILABLE 5-1. It represents the turbine trip initiating event frequency.

Industry efforts over the last fifteen years have led to a significant reduction in the number of reactor scrams and turbine trips. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

F-ICALONE 1.00E+00 1.31 FLAG: IC FAILURE NOT This event also appears in the CDF importance listing In Table CAUSED BY FAILURE OF 5-1. This event represents a sequence marker flag Identifying REACTOR VESSEL MAKE-those scenarios with isolation condenser failures. The isolation UP condenser can provide level control and decay heat removal.

Many SAMAs were considered that explored alternate injection and decay heat removal capabilities. No additional SAMAs were suggested.

O-AD-MU1 1.10E-04 1.25 2ADOP-ACT-ADSH--

This event also appears in the CDF importance listing in Table 21COP-IC-MU1-H--

5-1. This event represents the unlikely scenario of combined operator action failures for separate actions that otherwise are evaluated independently. This event is included for completeness as part of the human reliability dependency analysis. Phase I SAMAs 250 and 255 examine potential improvements in operator performance. No additional SAMAs were suggested for this topic.

2SIPH-DWHEAD-F-5.OOE-01 1.24 DRYWELL HEAD CLOSURE This event Is a Level 2 phenomena event that represents the FAILS DUE TO probability that a high pressure vessel failure scenario will lead OVERPRESSURE to an early containment failure given that the reactor cavity is wet at the time of vessel failure. The importance of this event would be minimized by reducing the number of high pressure vessel failure scenarios. See disposition above for 20POP-DEPRESSH-- (Operator fails to depressurize given failed in Level 1 or loss of DC). No additional SAMAs were suggested.

45

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%A 1.90E-04 1.19 INIT: LARGE LOCA This event also appears in the CDF importance listing In Table 5-1. It represents the Large LOCA initiating event frequency.

Mitigation from such an event would be improved by the existence of more reliable or diverse low pressure injection systems and water sources. Such potential improvements were examined In Phase I SAMAs 177, 184, 187, 194,197, 202, 205, and 208. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

%S1 2.40E-03 1.17 INIT: MEDIUM LOCA This event also appears In the CDF importance listing in Table 5-1. It represents the Medium LOCA initiating event frequency.

Mitigation from such an event would be improved by the existence of more reliable or diverse low pressure injection systems and water sources. Such potential improvements were examined in Phase I SAMAs 177, 184, 187, 194, 197, 202, 205, and 208. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

%TF 4.47E-02 1.16 INIT: TRANSIENT WITH FW This event also appears in the CDF importance listing In Table UNAVAILABLE AND MC 5-1. It represents the loss of feedwater initiating event AVAILABLE frequency. Industry efforts over the last fifteen years have led to a significant reduction in the number of plant scrams from all causes. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

%LOOP 3.09E-02 1.13 INIT: LOSS OF OFFSITE This event also appears in the CDF importance listing in Table POWER 5-1. It represents a single unit loss of offsite power event.

Improvements related to enhanced AC or DC reliability or availability were considered in Phase I SAMAs 90 through 129.

Many other SAMAs were also considered that would provide mitigation benefits in loss of offsite power scenarios including Phase II SAMAs 1, 2, 6, and 10. No additional SAMAs were suggested for this broad topic.

46

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2HI-SYSTEMUA-M--

2.13E-02 1.13 HPCI SYSTEM MUA This event also appears in the CDF importance listing in Table 5-1. It represents the probability of the HPCI system in maintenance. Potential improvements to enhance high pressure injection capabilities were considered in Phase I SAMAs 19,178,179,185,189,193,196,198, 201, 203, and 204. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

2SIPH-SI2-NOTFSU 5.00E-01 1.10 DRYWELL SHELL INTACT This event represents the complement to the Level 2 (OP=F) phenomena event 2SIPH-DWHEAD-F-- discussed above. As such, no additional SAMAs were suggested.

2LI--TOP-PEAKF--

2.50E-01 1.10 RX PWR IS TOP PEAKED This event also appears In the CDF Importance listing in Table SUCH THAT 2/3 RX LVL IS 5-1. It was added to the model in response to a GE concern INSUFF TO COOL CORE about LPCI steam cooling capabilities leading to elevated (CS R temperatures In upper portions of the core. Subsequent clarification on this issue will result in a re-examination of this assumption as part of the next update (i.e., LPCI injection with level maintained at 2/3 core height is a success state). No SAMAs were suggested for this issue.

2PVPPWATERBRKR--

8.00E-02 1.10 WATER LINE BREAK This event also appears in the CDF importance listing in Table MEDIUM LOCA 5-1. It represents the conditional probability of medium LOCA initiating events that are water line breaks as opposed to steam line breaks. See disposition above for %S1 (Medium LOCA Initiator).

XCSLISUCSTR-FCC 1

1.OOE-04 1.09 COMMON CAUSE This event al so appears in the CD importan liOting in Table.

PLUGGING OF ECCS 5-1.f Tis event representsthe unlikel ocurrence of a SUCTION STR DURING common cause failre of the ECCS suctionstriners. ',TheI LOCASDresden strainers have recently been upgraded and re-sized such that the potential for common cause pu lugng h as bee n 0 reduced. No additional SAMAswere suggested.

47

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2FCPH-FC1-NOTFSU 3.80E-01 1.09 CONT. FLOODING This event represents the evaluated likelihood that containment INITIATED (CLASS IAIC) flooding is initiated in accident class A or 1C scenarios.

Potential improvements to existing containment flooding capabilities were considered In Phase I SAMAs 45, 47, 48, 57, 61, 62, 81, and 86. No additional SAMAs were suggested.

BDCBY1 25----FCC 4.93E-06 1.08 COMMON CAUSE FAILURE This event also appears in the CDF importance listing in Table OF UNIT 2 AND UNIT 3 5-1. This event represents the common cause failure of the 125VDC BATTERIES (BETA 125V DC batteries. Many SAMAs were included that address

= 9.86E-03) potential enhancements for DC reliability and/or alternate means of providing DC power. Phase I SAMAs 92, 93, 96, 97, 98, 99, 113, 124, 125, 126, 127, and 128 are all related to improved DC performance. No additional SAMAs were suggested for this broad topic.

%TDC 1.50E-06 1.08 INIT: LOSS OF MULTIPLE This event also appears in the CDF importance listing in Table DC BUSES 5-1. It represents the unlikely initiating event of a complete loss of both 125V DC buses. Many SAMAs were Included that address potential enhancements for DC reliability and/or alternate means of providing DC power. Phase I SAMAs 92, 93, 96, 97, 98, 99, 113, 124, 125, 126, 127, and 128 are all related to improved DC performance. No additional SAMAs were suggested for this broad topic.

2DCRX-BUS2RECF-7.10E-01 1.08 DC BUS 2 FAIL TO This event also appears in the CDF Importance listing in Table RECOVER (GIVEN LOSS 5-1. It involves failure to recover one of the 125V DC buses OF MULTIPLE DC BUSES given loss of both. See disposition above for %TDC (Loss of INITIATOR %

Multiple 125V DC Buses Initiating Event).

3DCRX-BUS3RECF--

7.1OE-01 1.08 DC BUS 3 FAIL TO This event also appears in the CDF importance listing in Table RECOVER (GIVEN LOSS 5-1. It involves failure to recover one of the 125V DC buses OF MULTIPLE DC BUSES given loss of both. See disposition above for %TDC (Loss of INITIATOR %

Multiple 125V DC Buses Initiating Event).

48

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2CAHU-52-A-B2HCC 8.00E-05 1.07 PREINIT: MISCALIBRATE This event also appears in the CDF importance listing in Table CAS PRESSURE 5-1. It represents the unlikely scenario of miscalibration of SWTICHES 52A AND 52B pressure switches leading to unavailability of ECCS injection DUE TO CC (i.e., failure of RPV low pressure permissive interlock). This is included for completeness in the model since it has the potential of leading to core damage following a medium or large LOCA initiating event. No additional SAMAs are suggested for this topic.

2MSOPMSIVINLKH--

9.30E-01 1.07 OP ACT: BYPASS LOW is event Also appears In t isfing, iTable LEVEL MSIV INTERLOCK 5-1. it represents the human error probability of bypassing the,-,

(ATWS)

MSIV Isolat directed in the EOPs.This issue was '

specifically examined In Phase I SAMA 223 that was dispositioned with reference to existing capabilities. However, thisaction requires the use of jumpers with a limited time available, and as such carries a relativelyhigh HEPJvalue.

Thepotential benefit of Implementing a dedicated low level interlock switch Is explored as part of this RAI response (seeai Response 7(c)).

2ACBS-UAT-RATF--

1.OOE-01 1.07 PROB AC BUS WILL NOT This event also appears in the CDF importance listing In Table XFER TO RAT GIVEN LOSS 5-1. It represents a pseudo-recovery action in loss of DC bus or OF DC BUS 2 loss of multiple DC bus initiated events. Its' importance would be minimized by reducing the frequency of loss of DC events.

See disposition above for %TDC (Loss of Multiple 125V DC Buses Initiating Event).

2SLEV2-1106ABDCC 7.15E-03 1.06 EXPLOSIVE VALVES 2-This event also appears in the CDF importance listing in Table 1106A&B FAILURE TO 5-1. It represents the common cause failure of the SLC system OPEN DUE TO CCF explosive valves. Diversification of the SLC explosive valves was considered in Phase I SAMA 259 which was retained as Phase II SAMA 8. No additional SAMAs were suggested.

49

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2CZPH-HPBDVS3F--

1.00E+00 1.05 HIGH PRESSURE This event represents the evaluated likelihood that given the BLOWDOWN unlikely scenario of a vessel rupture or a vapor suppression OVERWHELMS VAPOR failure, that the RPV blowdown will indeed fail vapor SUPPRESSION suppression. Its' importance would be reduced by reducing the probability of vapor suppression failures. Improvements to the vacuum breakers at Dresden would reduce the probability of vapor suppression failures. Potential vacuum breaker improvements were explored in Phase 1 SAMA 68. No additional SAMAs were suggested.

2PVPPSTEAMBRKR--

9.20E-01 1.05 STEAM LINE BREAK This event represents the conditional probability of medium MEDIUM LOCA LOCA initiating events that are steam line breaks as opposed to water line breaks. See disposition above for %S1 (Medium LOCA Initiator).

2HITB2301TURBX--

9.60E-03 1.05 HPCI TURBINE FAILS TO This event also appears in the CDF importance listing in Table RUN 5-1. It represents the HPCI turbine failing during its mission time. See disposition above for 2HI-SYSTEMUA-M- (HPCI system in maintenance).

2SIHU-RCVR---H--

9.OOE-01 1.04 FAILURE TO RECOVER A This event represents the evaluated likelihood that an injection WATER SYSTEM system will not be recovered prior to drywell shell melt through.

See disposition above for 2SIPHCONTFAILF-- (Drywell Shell Melt-Through Fails Containment).

2SIPH-BARRIS-F-1.OOE+00 1.04 DW BARRIERS FAIL TO This event represents the evaluated likelihood that drywell PREVENT DEBRIS FROM barriers would prevent debris from contacting the shell, thereby CONTACTING SHELL preventing drywell shell melt-through. See disposition above for 2SIPHCONTFAILF- (Drywell shell melt-through fails containment).

2SIPH-SUMPOV-F--

1.001E+00 1.04 MELT OVERFLOWS SUMP This event represents the evaluated likelihood that core debris will overflow the sump and contact the drywell wall liner. See disposition above for 2SIPHCONTFAILF- (Drywell Shell Melt-Through Fails Containment).

50

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2DIDW-LOSSVSSF--

1.00E+00 1.04 DW NOT INTACT FOR This event represents the evaluated likelihood that vapor LOSS OF VAPOR SUPP.

suppression failures in LOCA scenarios would lead to (CLASS ID) containment failure. Its' importance would be reduced by reducing the probability of vapor suppression failures.

Improvements to the vacuum breakers at Dresden would reduce the probability of vapor suppression failures. Potential vacuum breaker improvements were explored in Phase I SAMA

68. No additional SAMAs were suggested.

2GVPHSTMINERTX--

1.001E+00 1.04 COMBUSTIBILE GAS This is a Level 2 phenomena event representing the evaluated VENTING NOT REQUIRED likelihood that a vapor suppression failure scenario would result (STEAM INERTED - CLASS in a steam inerted environment in containment thereby IIID) precluding the need for combustible gas venting. See disposition above for 2DIDW-LOSSVSSF-- (Vapor suppression failures lead to containment failure).

2PLSV-F-RECL-K--

1.50E-01 1.04 FAILURE OF SRVs TO This event also appears in the CDF importance listing In Table RECLOSE ON REDUCED 5-1. It represents the likelihood that the SRVs will not reclose PRESSURE after initially sticking open in response to a pressure transient.

The failure value of 0.15 is based on limited industry evidence.

See disposition for 2PLSVSORV-NTTK-- (Probability of SORV for non-turbine trip initiators) below.

2PLSVSORV-NTTK--

5.40E-02 1.04 PROBABILITY OF SORV This event also appears in the CDF importance listing in Table FOR NON TT INITIATORS 5-1. It represents the likelihood that an SRV will stick open in response to a pressure transient. For Dresden, this renders the isolation condenser ineffective. Consequently, early Injection from HPCI or depressurization for low pressure injection is required for success. Many SAMAs considered potential benefits from improved injection capabilities or improved depressurization capabilities. No additional SAMAs were suggested.

51

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2ADOP-DEPMADSH--

3.50E-03 1.04 OP ACT: DEPRESS RPV This event also appears in the CDF importance listing In Table (MLOCASORV) 5-1. It represents the human error probability for failing to depressurize for low pressure injection given a medium LOCA or SORV event and initial failure of HPCI to inject (thereby requiring depressurization for low pressure injection). Potential improvements to depressurization capabilities were considered in Phase I SAMAs 190, 191, 229, 230, 240, 241, and 247. No additional SAMAs were suggested for this broad topic.

O-AD-HI-MU1 1.10E-04 1.03 2ADOP-ACT-ADSH--

This event represents the unlikely scenario of combined 2HIOP-OVRFILLH-- 21COP-operator action failures for separate actions that otherwise are IC-MU1-H--

evaluated independently. This event is included for completeness as part of the human reliability dependency analysis. Phase I SAMAs 250 and 255 examine potential improvements in operator performance. No additional SAMAs were suggested for this topic.

%DLOOP 9.41 E-03 1.03 INIT: DUAL LOSS OF This event also appears in the CDF importance listing in Table OFFSITE POWER 5-1. It represents the dual unit loss of offsite power initiating event frequency. See disposition above for %LOOP (Single Unit Loss of Offsite Power).

52

Response 5(e):

"(Provide] the reasons for the difference in the number of SAMAs evaluated for Quad Cities Nuclear Power Station (QCNPS) and DNPS (280 v. 265)."

Quad Cities included 30 plant-specific insights in addition to 250 generic insights as part of the SAMA list development. 19 of these plant-specific insights were not applicable to Dresden (e.g., they related to the SSMP or were specific to the IPEEE for Quad Cities),

and as such were not included for Dresden. Two additional SAMAs that were PWR specific were included on the list for Quad, but not for Dresden. This means that 259 of the 265 Dresden SAMAs were also on the Quad list. The remaining 6 SAMAs were unique to Dresden. Phase 1 SAMAs 257, 258 related to the isolation condenser, and therefore were not applicable to Quad Cities, and SAMAs 261, 262, 263, and 265 were obtained from Dresden PRA Insights, and were not included in the Quad Cities SAMA list.

Response 5(f):

'[Providel a general description of the group of 130 insights mentioned in the original IPE and a discussion of how and whether the insights that were not implemented were factored into the SAMA evaluation."

One of the important means of identifying plant specific improvements for the Dresden SAMA analysis was a review of the plant's IPE. As part of the IPE, an analysis of the cutsets and importance rankings was performed in order to identify plant weaknesses and to suggest changes that would address the weaknesses identified. The original Dresden IPE submittal report stated that over 130 IPE insights and over 60 Accident Management (AM) insights had been identified. Subsequent to that report, several additional insights were identified.

In summary, the original IPE included a commitment to implement two IPE insights.

Procedure revisions were completed in 1993 and 1994 that implemented those two insights. In 1994, the Dresden PRA group identified 11 other IPE insights as having a major benefit. Further evaluation indicated that action was not warranted on most, but action was taken on two of those insights. It was also found that 12 of the IPE insights had been fully or partially implemented via procedure revisions, operator aids, or changes in control room staffing. The revised IPE submittal report prepared in 1996 (Reference [5-11, Section 4.7.3) indicated that 'a vigorous pursuit of further enhancements is not warranted at this time."

The Accident Management insights from several sites including Dresden were carefully considered by the BWROG in developing the EOPs and SAMGs that have been subsequently implemented at Dresden. No additional action was required.

53

Although the IPE insights were not directly used as input into the SAMA analysis, more recent insights from the updated PRA models were factored directly into the SAMA list.

Seventeen of the Phase 1 SAMAs include the Dresden Risk Management Insights' as the reference source (i.e., indicated in Table F-1 of the ER as Reference 64) and four others were based on IPEEE insights.

These twenty-one items were specifically developed following the completion of the 1999 PRA model update and 2000 Fire Risk Model. The completion of the 2002 model update did not lead to any additional insights as the results did not dramatically change.

In any event, a correlation between importance parameters for both CDF and LERF from the 2002 model and their relationship to the SAMA analysis is provided in Response 5(d). In summary, it was judged that these more recent insights were sufficient and appropriate for supplementing the generic SAMA lists with plant-specific insights.

References

[5-1] ComEd, Dresden Station Individual Plant Examination Submittal Report, Revision 1, June 1996.

54

RAI6 The SAMA analysis did not include an assessment of SAMAs for external events. The DNPS IPE for External Events (IPEEE) has shown that the CDF due to internal fire initiated events is 1. 7x10-5 per reactor year for Unit 2 and 3. 1x10-5 per reactor year for Unit 3. The risk analyses at other commercial nuclear power plants also indicate that external events could be large contributors to CDF and the overall risk to the public. In this regard, provide the following:

a. NUREG-1742 (Perspectives Gained From the IPEEE Program,' Final Report, 4/02), lists the significant fire area CDFs for DNPS (pages 3-15 and 3-16 of Volume 2). While these fire-related CDF estimates may be conservative, they are stifl large relative to the DNPS internal events CDF.

For each fire area or dominant fire sequence, explain what measures were taken to further reduce risk, and explain why these CDFs can not be further reduced in a cost effective manner.

b. the IPEEE Safety Evaluation Report (SER), Extended Power Uprate (EPU) SER, and NUREG-1 742 (Tables 2.7 and 2.12) identify seismic outliers and improvements for DNPS.

Confirm that all of the plant improvements that address the outliers have been implemented. If not, then discuss the rationale within the context of this SAMA study.

For those improvements still pending (e.g., seismically-verified makeup path to the isolation condenser, and modifications to improve the reliability of the containment cooling service water cooling function),

provide a brief description of each improvement and its status.

c. Exelon states that Phase 2 SAMA 5 remains under investigation for resolution as part of the DNPS closeout of the IPEEE commitments.

Describe the improvements under investigation, their status, and expected implementation schedule. As part of this response, identify the systems, structures, and components (SSCs) that limit the plant high confidence in low probability of failure (HCLPF).

Justify why modifications to increase seismic capacity would not be cost-beneficial when evaluated consistent with the regulatory analysis guidelines for those structures, systems and components fSSCSJ below 0.3g yet not expected to be modified.

Response 6(a):

INUREG-1742 Perspectives Gained From the IPEEE Program, Final Report, 4/02),

lists the significant fire area CDFs for DNPS (pages 3-15 and 3-16 of Volume 2). While these fire-related CDF estimates may be conservative, they are still large relative to the DNPS internal events CDF. For each fire area or dominant fire sequence, explain what 55

measures were taken to further reduce risk, and explain why these CDFs can not be further reduced in a cost effective manner."

As an IPEEE, the Dresden fire study was performed primarily to develop risk insights. It was done in the traditional style of Fire PRAs, and as such, employs conservatism and involves some level of uncertainty (also see Attachment A that provides more details on the types of conservatisms and uncertainties associated with the use of quantitative results from Fire PRAs). Therefore, it cannot be used directly to provide a realistic cost-benefit analysis as part of the SAMA evaluations.

In any event, a review of the Dresden Fire PRA model cutsets was performed to determine the dominant sequence types. Excluding the control room severe fire, it was determined that although there are many different scenarios and initiating events, there are just three dominant sequence types: loss of decay heat removal (1W), loss of injection at high pressure (TQUX), and loss of injection at low pressure (TQUV). These three scenarios are also significant contributors to the internal events calculated core damage frequency.

Potential improvements to respond to the three dominant Fire PRA sequence types were examined in many portions of the SAMA analysis.

This included potential improvements to high pressure injection capabilities, RPV depressurization capabilities, low pressure injection capabilities, and decay heat removal capabilities. As such, it is judged that any improvements that could be justified using the internal events CDF as a measure (with extra margin considered to account for potential benefits from external events as described in Response 7(c)), is the best use of available capabilities to determine the estimated averted costs and benefits. An additional fire-area-by-fire-area search for improvement ideas will not be productive until Fire PRA technology advances to the point that a direct comparison of the Fire CDF results and the internal events CDF results is possible.

Response 6(b):

77he IPEEE Safety Evaluation Report (SER), Extended Power Uprate (EPU) SER, and NUREG-1742 (Tables 2.7 and 2.12) identify seismic outliers and improvements for DNPS. Confirm that all of the plant improvements that address the outliers have been implemented.

f not, then discuss the rationale within the context of this SAMA study.

For those improvements still pending (e.g., seismically-verified makeup path to the isolation condenser, and modifications to improve the reliability of the containment cooling service water cooling function), provide a brief description of each improvement and its status.'

56

IPEEE Safety Evaluation Report and NUREG-1742 Seismic Outliers and Improvement Status As indicated in NUREG-1742, an extensive number of plant improvements or other actions were planned to resolve the USI A-46 outliers. These improvements pertained primarily to enhancing anchorage/support capacity and reducing or eliminating the potential for adverse interactions. Dresden recently informed the NRC that all of the outliers have either been resolved or will be completed no later than the end of the Unit 2 refueling outage scheduled for October 2003 except for those listed in Table 6-1 which will be completed by the end of the Unit 3 refueling outage scheduled for fall 2004. Reference letter from R. J. Hovey, Dresden Nuclear Power Station, Delays in Completion of Unresolved Safety Issue (USI) A-46 Commitment, RHLTR 03-0046, dated July XX, (Need date from licensing, current submittal schedule is July 17th) 2003.

Remaining unresolved issues and scheduled completion dates are shown below.

Table 6-1 Unresolved Safety Issue Status Description Completion Schedule Unt 3 Modifications to five 250 volt direct current (VDC)

D3R18 Scheduled forfall 2004 Motor Control Centers (MCCs)

EPU SER Seismic Outlier and Improvement Status

'The NRC SER on the DNPS IPEEE indicates that the licensee had implemented a number of improvements during the resolution of unresolved safety issue (USI) A-46, "Verification of Seismic Adequacy of Equipment in Operating Plants," and that a number of additional improvements were still under consideration. In particular, the SER states that the licensee was developing a concept for providing a seismically-qualified/verified make-up path to each unit's isolation condenser and that this design change would be implemented in conjunction with the approved schedule for resolution of the USI A-46 outliers. The DNPS IPEEE SMA took credit for this modification for the scenario in which the dam fails during a seismic event, but the modification has not been implemented at this time."

Additional background:

Dresden responded to an NRC EPU RAI regarding seismic capability in a letter from K.

A. Ainger, RS-01-208, dated September 26, 2001, "Additional Information Supporting the License Amendment Request to Permit Uprated Power Operation at Dresden Nuclear Power Station."

"The sources of makeup water to the IC shell side are not seismically qualified, but given the redundancy and diversity of these sources, there is a high confidence that at 57

least one source will be available following a seismic event. The current sources include initial makeup from on-site tanks and the Unit 1 fire pump, and makeup from the ultimate heat sink (UHS). The DNPS response to the Individual Plant Examination of Extemal Events (PEEE)

(Reference 1) included a commitment to provide a seismic makeup path to the IC by November 2003."

This commitment to provide a seismic makeup path to the IC is on schedule to be completed by November 2003.

In the same document, the following is stated:

'Question 2: Provide additional discussion regarding the results of the study to confirm the adequacy of the isolation condenser to provide suppression pool cooling following a small break LOCA with a dam failure, and the acceptability of proceeding with the power uprate based on the results of this study.

Response

The study for the small break loss of coolant accident (SBLOCA) coincident with a dam failure has been completed for EPU conditions. The study assumed a one inch small break, consistent with the guidance in EPRI NP-6041-SL, A Methodology for Assessment of Nuclear Power Plant Seismic Margin (Revision 1)." EPU decay heat was used in the analysis. The analysis demonstrates that the IC and available emergency core cooling systems (ECCS) (i.e., high pressure coolant injection (HPCI) and low pressure coolant injection (LPCI)) are sufficient to mitigate a seismically induced SBLOCA for a 24-hour period. The study shows that additional equipment, specifically a cooling water supply to the CCSW heat exchangers, will be required 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months after the onset of the event to supply suppression pool cooling.

DNPS has developed a conceptual design using large portable pumps that would be used to restore the required CCSW cooling flow via suction from the intake canal.

These pumps would be stored in an area that could withstand the postulated seismic event, and would be staged with hose connections to the CCSW piping. The necessary fittings will be installed on the existing CCSW piping. Power for the portable pumps will be supplied either by portable diesel engines or by temporary power connections to the available existing electrical buses. Procedures will be developed to ensure that the necessary actions will be taken within the 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months period to establish suppression pool cooling flow. These actions will provide the capability to mitigate the seismically induced SBLOCA for the 72 hour8.333333e-4 days 0.02 hours 1.190476e-4 weeks 2.7396e-5 months time frame given in EPRI NP-6041-SL. These actions will be completed on the same schedule as the modification to provide a seismically qualified makeup path to the IC as described in Reference."

The CCSW fitting modification and development of Procedures to ensure that the necessary actions will be taken within the 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months period to establish suppression pool 58

cooling flow are scheduled to be completed on the same schedule as the IC make-up seismic upgrade modification and will be completed by November 2003.

Response 6(c):

'Exelon states that Phase 2 SAMA 5 remains under investigation for resolution as part of the DNPS closeout of the IPEEE commitments. Describe the improvements under investigation, their status, and expected implementation schedule.

As part of this response, identify the systems, structures, and components (SSCs) that limit the plant high confidence in low probability of failure (HCLPF).

Justify why modifications to increase seismic capacity would not be cost-beneficial when evaluated consistent with the regulatory analysis guidelines for those structures, systems and components (S[S]CS) below 0.3g yet not expected to be modified."

See Response 6(b) for the improvements that have been or will be made, their status, and expected implementation schedule.

Table 6-2 [Reference 6-1] shows the new HCLPF capacity of items listed on page 1-3 and 1-4 of the original IPEEE submittal. The new HCLPF capacities are based on additional evaluations and improvements that have been made or are scheduled to be made as identified in Response 6(b).

Table 6-2 HCLPF Capacities of Previously Identified Outliers Original New Description Basis for New Capacity Capacity Capacity (pga)

(pga) 0.15g

>0.3g Cable Trays-Turbine, Reactor & Service More rigorous I________

Bldgs., El. 517' (GIP LAR 007) evaluation.

0.17g 0.202g Buses - D03-8303B---M05, D02-8302B----

Additional evaluation.

M05, D03-8303A---M05, Dist. Panel D03-83125---PO6 0.17g 0.27g Distribution Panel - D02-83125---P06 and Additional evaluation.

Bus D02-8302A---M05 0.17g

>0.3g Cabinet - D02-2252-0010 Anchorage Modification.

0.20g No Condensate Storage Tanks - DOO-3303-A---

Original evaluation change T05, DOO-3303-B----T05 0.22g

>0.3g Control Panels D02-0902-0004, 0015, 0017, Modification.

0019 & 0036, D03-0903-0004, 0015, 0017, 0019 & 0036 0.23g

>0.3g Control Panels D02-0902-0028 & -0003, D03-Additional evaluation.

0903-0028 59

Table 6-2 HCLPF Capacities of Previously Identified Outliers Original New Description Basis for New Capacity Capacity Capacity (pga)

(pga) 0.26g No Diesel Fuel Oil Storage Day Tank DOO-5202-Original evaluation.

change T05 0.27g No Battery Charger - D02-8300-2A--B05 Original evaluation.

change 0.27g No Distribution Panels - D02-9802-A & B---P06 Original evaluation.

change 0.27g No Switchgear - D02-7328--S35 & D02-7329---

Original evaluation.

change S35 0.27g No Bus #2A D02-8302A1---P06 Original evaluation.

change 0.27g No 125V DC/TB Battery Bus #2 D02-83125 Original evaluation.

change P06 0.27g No 125V DC/Battery Charger #2 D02-8300-2--

Original evaluation.

change B05 0.28g No 125V DC Battery Charger - D03-8300-3A--

Original evaluation.

change B05 0.28g No Unit 2&3 Torus Suppression Chambers Original evaluation.

change 0.28g

>0.3g Cabinet - D02-2252-0021 Anchorage modification.

0.29g No Motor Control Centers D02-83250---M05 &

Original evaluation.

change D02-7826-4--M05 0.29g No Bus #2B D02-83021-1 -- P06 Original evaluation.

change 0.29g No 125V DC/TB Res Bus #2 D02-83125-1--P06 Original evaluation.

change As can be seen in Table 6-1, there are a limited number of components with HCLPF capacities that fall into the range of 0.2g to 0.3g.

In fact, the majority of SSCs at Dresden already have HCLPF values of at least 0.3g.

EPRI has estimated that the SQUG modifications resulted in expenses of $1.4M per plant, but it is estimated that Dresden had more SQUG outliers than the average plant.

To address all of the remaining items listed above, it is estimated that this would require a similar effort to the SQUG modifications, or more than $2.0M.

Limited benefit would be obtained by improving the plant HCLPF to 0.3g for all SSCs.

Using the methodology from the Dresden EPU RAI responses, the maximum benefit 60

from increasing the plant HCLPF from 0.2g to 0.3g is conservatively estimated at about 5E-06/yr, but practically the actual maximum benefit is quite less. The cost estimate of more than $2.OM precludes this as being cost-beneficial. Cost benefits from individual improvements can also not be easily made at this time without extensive analysis efforts. As such, it is judged that further modifications to increase seismic capacity are not warranted.

REFERENCES

[6-1] Letter from P. Swafford to USNRC, Request for Additional information Regarding Individual plant Examination of External Events," dated March 30, 2000.

61

RAI7 The SAMA analysis did not include an assessment of the impact that PRA uncertainties and external event risk considerations would have on the conclusions of the study.

Some license renewal applicants have opted to double the estimated benefits (for internal events) to accommodate any contributions for other initiators when sound reasons exist to support such a numerical adjustment, and to incorporate additional margin in the SAMA screening criteria to address uncertainties in other parts of the analysis (e.g., an additional factor of two in comparing costs and benefits of each SAMA). At DNPS, external events (both fire and seismic) are dominant contributors to the total CDF, and are over a factor of 10 greater than internal event contributions. On that basis, provide the following information to address these concems:

a. an estimate of the uncertainties associated with the calculated core damage frequency (e.g., the mean and median internal events CDF estimates and the 5th and 95th percentile values of the uncertainty distribution).

b. an assessment of the impact on the Phase 1 screening if risk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated with external events (as applicable).

c. an assessment of the impact on the Phase 2 evaluation if risk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated with external events (as applicable).

Consider the uncertainties due to both the averted cost-risk and the cost of implementation to determine changes in the net value for these SAMAs.

Response 7(a):

'[Provide) an estimate of the uncertainties associated with the calculated core damage frequency (e.g., the mean and median internal events CDF estimates and the 5th and 95th percentile values of the uncertainty distribution)."

The 2002 update of the Dresden PRA model was utilized as the basis for the SAMA analysis performed in support of the environmental report. This version of the model was not populated with uncertainty distributions for the data input parameters.

Conse91uently, development of the median internal events CDF estimates and the 5h and 95 percentile values of the uncertainty distribution are not readily available. (Note that population of the uncertainty distribution parameters is anticipated for a future model revision update.) Table 7-1 provides estimates of internal events Level 1 CDF uncertainty distributions that were obtained for other plants from various sources.

62

Table 7-1 Representative Core Damage Frequency Uncertainty Distributions Plant I Point Para-5 Median 9 5P 95 Error Reference Model Estimate metric Percentile Value Percentile P.E.

Factor Mean Mean Value Value Mean Value Value Ratio Peach 3.6E-6 ()

4.5E-6 3.5E-7 1.9E-6 1.3E-5 3.6 6.1 NUREG/CR-Bottom

4551, Volume 4, Rev. 1, Part 1 (Table S-1a)

Grand Gul 2.OE-6 ()

4.E-6 1.8E-7 1.1E-6 1.4E-5 7.0 8.8 NUREG/CR-

4551, Volume 6, Rev. 1, Part 1 (Table S-2)

LaSalle /

3.1 E-5 4.4E-5 2.1 E-6 1.6E-5 1.4E-4 4.5 8.2 NUREG/CR-RMIEP

4832, Volume 2 (RMIEP),

(Table 3.1)

LaSalle I 6.64E-6 6.88E-6 2.82E-6 5.20E-6 1.39E-5 2.1 2.2 LS-PSA-014, Current LaSalle Quantification

Notebook, Revision 2, June 2003 (Appendix G)

H.B.

4.3E-5 4.5E-5 1.5E-5 3.3E-5 1.1E-4 2.6 2.7 Docket No.

Robinson 50/261 (Response to Request for Additional Information Regarding SAMA Analysis)

V.C.

5.6E-5 5.6E-5 1.9E-5 4.4E-5 1.3E-4 2.3 2.6 Docket No.

Summer 501395 (Response to SAMA Request for Additional Information)

(1) From NUREG/CR-4550, Vol. 4, Rev. 1, Part 1, Page 5-1.

(2) From NUREG/CR-4550, Vol. 6, Rev. 1, Part 1, Page 5-1.

63

The collective information shown in Table 7-1 indicates that the point estimate to mean ratio could be as little as 2 or as large as 7. The LaSalle/RMIEP distribution parameters are chosen as representative since they represent the second-most broadest distribution. Therefore, a factor of 4.5 increase from the calculated point estimate mean internal events CDF with an error factor of 8 is used as a reasonably conservative estimate to approximate the uncertainty distribution.

This correlates to an estimated 95t percentile value of about 8.6E-6/yr for the Dresden internal events core damage frequency. Additionally, the assumed error factor of 8 can be used to approximate the median and 5t percentile values as well as is shown below.

Dresden Approximated Uncertainty Distribution:

95th Percentile:

4.5 * (Point Estimate Mean)

= 8.5E-6Iyr Median:

95hI EF = 8.5E-6Iyr/ 8

= 1.1 E-6/yr 5t Percentile:

Median / EF = 1.1 E-6Iyr / 8

= 1.3E-7/yr Response 7(b):

'Provide] an assessment of the impact on the Phase 1 screening f sk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated with external events (as applicable)."

As indicated in Response 7(a), it is estimated that the 95t percentile value would be approximately a factor of 4.5 higher than the reported mean CDF value of 1.9E-6. This can be assumed to correspond to an internal events upper bound value of about 8.5E-6.

The Dresden Internal Fire risk model was updated in 1999 as part of the revised IPEEE submittal report. The CDF contribution to internal fires was estimated at 1.7E-5/yr for Unit 2 and 3.OE-5/yr for Unit 3. However, the methodology invoked to determine the fire CDF is judged to be highly conservative, and therefore it is judged that it is not appropriate at this time to directly compare internal events CDF values with the reported Fire CDF. 2 The seismic portion of the IPEEE program was completed in conjunction with the SQUG program.

Dresden performed a seismic margins assessment (SMA) following the guidance of NUREG-1407 and EPRI NP-6041. The SMA is a deterministic evaluation that does not calculate risk on a probabilistic basis.

No core damage frequency sequences were quantified as part of the seismic risk evaluation.

However, an extensive number of plant improvements were identified and these have are being resolved as is noted in Response 6(b).

2 Attachment A provides an assessment of the use of quantitative risk estimates from Fire PRAs, and why it is judged that the calculated CDF values should not be directly compared at this time.

64

Consequently, to account for both uncertainties in the risk assessment and the potential additional benefits associated with external events, the Phase I screening was re-performed assuming a factor of almost five increase to the base cost risk for DNPS to

$2.OM (compared to the base internal events cost-risk of $457,000 used in the ER).

The screening criteria utilized in Table F-i of the Dresden ER includes the following categories:

1 - Not applicable to the Dresden design

2 - Similar item is addressed under other proposed SAMAs

3 - Already implemented at Dresden

4 - No significant safety benefit associated with this SAMA for Dresden

5 - Cost of implementation clearly greater than the maximum averted cost risk

6-Retained for Phase II analysis

7-Requested additional information from Dresden (Not Used)

8 - ABWR design issue, not practical For the revised Phase I screening, SAMA items that previously screened by Criteria #1 or #8 were not re-examined. SAMA items that previously screened by Criteria #2 or #3 were also re-examined to see if an alternative approach to addressing the SAMA could be potentially beneficial, and to look at the potential impact of additional benefits that might be afforded by including external events in the analysis.

SAMA items that previously screened by Criteria #4 or #5 were also all re-examined, and the previously retained items (i.e., Criteria #6) were still retained and were subject to re-analysis as described in Response 7(c).

The results of the revised Phase I screening for all previous criteria #4, #5, and #6 entries are included in Table 7-2. Criteria #2 or #3 entries are only included in Table 7-2 if the disposition is changed. As can be seen, two additional SAMAs are now retained for Phase II (See Phase I SAMA 188 and Phase I SAMA 223) where the revised disposition column is noted as being the key for noting changes compared to the ER.

65

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening CrIterla Including Uncertainty and SAMA ID number External Events number 1

Cap downstream piping of SAMA would reduce the

4 - No significant The ROCCW system and the SW system Considering uncertainty and N/A normally closed component frequency of a loss of safety benefit vent and drain valves are not observed to be potential Impacts from cooling water drain and vent component cooling event, a failure modes at Dresden. Their failures are external events does not valves.

large portion of which was not included in the Dresden PSA. The risk Introduce any significant derived from catastrophic impact of vent and drain valve failures Is changes. No change to the failure of one of the many estimated to be negligible at Dresden.

screening criteria category.

single isolation valves.

3 Enhance loss of component SAMA would reduce the

6 - Retain SiNl retained.

cooling procedure to potential for RCP seal present desirabilIty of failure.

cooling down reactor coolant system (RCS) prior to seal LOCA.

66

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential original / Revised Original DIsposition Revised Disposition IPhase SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 7

Proceduralize shedding SAMA would Increase time

4 - No significant PWR RCP seal leakage Issue. The Considering uncertainty and N/A component cooling water before the loss of safety benefit competing risks associated with shedding peIal mpacts from loads to extend component cooling (and other RBCCW loads Is not considered external events does not component cooling heatup reactor coolant pump seal justified. Therefore, this SAMA Is not Introduce any significant on loss of essential raw failure) In the loss of pursued.

changes. No change to the cooling water.

essential raw cooling water Dresden has the following features that screening criteria category.

sequences.

reduce the Impact of loss of Recirculation Pump seal cooling:

- Minimal Seal leakage might occur If both 11 Create an Independent SAMA would add

5 - Cost would be the cooling from RBCCW and the purge Considering uncertainty and N/A RCP seal Injection redundancy to RCP seal more than risk flow from CRD become unavailable.

potential Impacts from system, with a dedicated cooling alternatives, benefit This Is postulated for SBO events or external events does not diesel.

reducing CDF from loss of loss of SW events.

introduce any significant component cooling or

- a new improved Recirculation pump seal changes. No change to the service water or from a with significantly reduced potential for screening criteria category.

station blackout event leakage (12.5 gpm/pump versus some PWR estimates of 480gpm/pump) 12 Use existing hydro-test SAMA would provide an

5 -Cost would be

- multiple high pressure injection systems Considering uncertainty and WA pump for RCP seal Independent seal injection more than risk that provide RPV makeup capability to oential impacts from assure adequate RPV Inventory. These PotenilIpcsfo injection.

source, without the cost of a benefit Include:

external events does not new system.

introduce any significant

- HPCI (turbine driven system) changes. No change to the

- CRD (Unit 2 and Unit 3) screening criteria category.

- SLC from test tank or SBLC tank

- Feedwater

- HPCI and SOLC are Independent of SW and RBCCW failure

FW and CRD are independent of RBCCW failure Because of the availabilIty of multiple high pressure Injection systems, the small Recirculation Pump seal leakage Is not a significant contributor to the risk profile.

67

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 19 Use fire protection system SAMA would reduce the

5 - Cost would be Fire protection Is a low head system at The cost Is considered to be N/A pumps as a backup seal frequency of the RCP seal more than risk Dresden and cannot currently be used as greater than the upper injection and high-LOCA and the SBO CDF.

benefit a HP injection source. The ability to bound maxdmum averted pressure makeup.

provide high pressure njection during an cost risk of $2.OM. No SBO may be beneficial, but the cost of the change to the screening required modifications would be high.

criteria category.

Installation of new high pressure piping, a high head, high flow pump (as It would also have to support the fire system) and a supporting diesel generator or pump motor Is similar in scope to SAMA 185. The cost Is also considered to be similar ($5 million to $10 million) and Is greater than the maxdmum averted cost-risk for Dresden

($457,000).

See also SAMA 178.

22 Improved ability to cool SAMA would reduce the

6 - Retain Dresden has redundant methods of decay StNli retained.

2 the residual heat removal probability of a loss of heat removal including:

heat exchangers.

decay heat removal by LPCI In torus cooling Implementing procedure and hardware modifications SDC (separate system) to allow manual alignment Venting of the fire protection system Main Condenser or by installing a component cooling water cross-tle.

LPCI In torus cooling Is cooled by the CCSW from the ntake.0a A portable diesel-driven C

fo t i

pump Is under consideration to provide Dresden's Shutdown Cooling system has cooling water to a LPCI heat exchangers that are cooled by heat exchanger. This was RBCCW and SW from the intake. Plant discussed in the EPU capability and procedures are available to correspondence as the allow cross-tle to the opposite unks tentative plan for dealing RBCCW system.

with the seismic outiler of The portable diesel-driven pump s Dresden Island Lock &

considered to deal with large reduction In Dam, i.e., loss of UHS, by intake level.

Fall 2003.

68

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA tMe Result of potential Original Revised Original Disposition Revised Dispos Won Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 23 8.a. Additional Service SAMA would conceivably

5-Cost would be The cost of Implementing this SAMA has The cost is considered to be N/A Water Pump reduce common cause more than risk been estimated at approximately $5.9 greater than the upper dependencies from SW benefit million and Is greater than the maximum bound maximum averted system and thus reduce averted cost-risk for Dresden ($457,000).

cost risk of $2.0M. No plant risk through system change to the screening reliability Improvement.

criteria category.

24 Create an independent This SAMA would add

4 - No significant The recirculation pump seal leakage at Considering uncertainty and N/A RCP seal injection redundancy to RCP seal safety benefit Dresden could compromise the long term potential Impacts from system, without dedicated cooling altematives, success of the Isolation Condenser. An external events does not diesel reducing the CDF from loss Independent safety related seal cooling Introduce any significant of CC or SW, but not SBO.

system could reduce this Impact however, changes. No change to the the risk Impact of the recirculation seal screening criteria category.

leak Is already very low.

25 Provide reliable power to SAMA would Increase

4 - No significant Control Room HVAC Is powered by Non-Considering uncertainty and N/A control building fans.

availability of control room safety benefit ESS buses that can be powered by EDGs potential Impacts from ventilation on a loss of given a LOOP. Control Room HVAC Is external events does not power.

not required for successful accident introduce any significant mitigation.

changes. No change to the screening criteria category.

26 Provide a redundant train SAMA would Increase the

5 - Cost would be The cost of Installing a redundant, diverse The cost Is considered to be N/A of ventilation.

availability of components more than risk train of HVAC for a Switchgear Room has greater than the upper dependent on room cooling. benefit been estimated at $10 million (Reference bound maximum averted 19). This estimate far exceeds the cost risk of $2.OM. No maximum averted cost-risk for Dresden change to the screening

($457,000). Assuming the cost to Install a criteria category.

redundant train of HVAC in other areas Is approximately equivalent to this estimate providing a redundant train of HVAC would not be cost beneficial for any system and Is screened from further analysis.

l 69

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 number Externa Events number

- ~~

29 Create ability to switch fan power supply to DC In an SBO event.

SAMA would allow continued operation in an SBO event This SAMA was created for reactor core Isolation cooling system room at Fitzpatrck Nuclear Power Plant.

4 - No significant safety benefit The systems that require room cooling and have the capability of operating during an SBO Include only HPCI (no IC room cooling dependency). During a postulated SBO, HPCI can operate for the duration of the event which Is limited by DC battery life. Use of a DC powered fan would increase the drain on the batteries with no Impact on the reliability of the HPCI systems as long as there is no gland seal failure. For the low probability event of gland seal failure the crew is directed to bypass high temperature room trips. This would avoid the trip of HPCI. Component failures of these systems could also occur, but this is judged to represent a negligible risk impact. As such there is no measurable safety benefit associated with this SAMA.

Considering uncertainty and potential impacts from external events does not introduce any significant changes. No change to the screening criteria category.

N/A Install an Independent SAMA would decrease the

5 - Cost would be Installation of a new, independent, The cost Is considered to be N/A method of suppression probability of loss of more than risk suppression pool cooling system Is similar greater than the upper pool cooling.

containment heat removal.

benefit In scope to Installing a new containment bound maximum averted For PWRs, a potential spray system, which has been estimated cost risk of $2.OM. No similar enhancement would to cost approximately $5.8 million. This change to the screening be to Install an independent exceeds the maximum averted cost-risk criteria category.

cooling system for sump for Dresden ($457,000).

water.

Develop an enhanced SAMA would provide a

6 - Retain A potential enhancement would be to Still retained. Consider drywell spray system.

redundant source of water proceduralize the crosstie between the benefit that could be onEo=

to the containment to containment spray path of one unit to the obtained by the addition of control containment LPCI system of the opposite unit Another a connection between the pressure, when used in alternative Is the addition of a connection containment spray and the conjunction with between containment spray and the plants fire protection containment heat removal.

plant's fire protection system.

system. Also consider lower cost aitemative of procedurallzlng existing (See DEOP 0500-03).

capabilities from other unit

-.ff LPCI cross-tie.

70

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SANMA ID number External Events number 36 Provide dedicated existing SAMA would provide a

5 - Cost would be Installation of a new, independent, The cost is considered to be N/A drywelI spray system.

source of water to the more than isk containment spray system, has been greater than the upper containment to control benefit estimated to cost approximately $5.8 bound maximur averted containment pressure, when million. This exceeds the maximum cost risk of $2.OM. No used in conjunction with averted cost-risk for Dresden ($457,000).

change to the screening containment heat removal.

criteria category.

This would use an existing spray loop instead of developing a new spray system.

l 38 Install a filtered SAMA would provide an

5 - Cost would be Potential to Improve both the Level I and The cost Is considered to be N/A containment vent to alternate decay heat more than sk Level 2 results. Cost expected to exceed greater than the upper remove decay heat.

removal method for non-benefit the maximum averted cost-risk for bound maximum averted ATWS events, with the Dresden ($457,000) cost risk of $2.OM. No released fission products change to the screening being scrubbed.

criteria category.

Option 1: Gravel Bed Filter Option 2: Multiple Venturi Scrubber 39 Install a containment vent Assuming that injection is

5 - Cost would be Dresden does not have a hard pipe vent of The cost is considered to be N/A large enough to remove available, this SAMA would more than risk sufficient capacity to mitigate ATWS greater than the upper ATWS decay heat.

provide alternate decay benefit pressurization unless other mitigation bound maximum averted heat removal in an ATWS steps are successful. Cost expected to cost risk of $2.OM. No event exceed the maximum averted cost-risk for change to the screening Dresden ($457,000) criteria category.

71

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number w_

.~~~~~~~

40 Create/enhance hydrogen recombiners with Independent power supply.

SAMA would reduce hydrogen detonation at lower cost, Use either

4 - No significant safety benefit

1) a new Independent power supply

2) a non-safety-grade portable generator

3) existing station batteries

4) existing AC/DC Independent power supplies.

1*

t t

41 Install hydrogen recombiners.

SAMA would provide a means to reduce the chance of hydrogen detonation.

4 - No significant safety benefit.

The Dresden primary containment Is nerL The Nitrogen Make-up system maintains an nerted atmosphere within containment during normal operation. In accident conditions, It provides a feed and bleed function which purges the containment atmosphere of accumulated combustible gases including oxygen and hydrogen, etc.) and replaces them with nitrogen.

Nitrogen Containment Atmospheric Dilution (NCAD) this modification has been installed on both units. This system provides a reliable source of Nitrogen for combustible gas control following an accident. It would be used should the normal make-up flow path not be available during post-accident conditions. The design flow rate Is 29 scfm through each line at 31 pslg.

The NCAD system is designed to control the 02 and H2 concentrations by venting and purging with nitrogen. In addition, hydrogen recomblners are precluded from operating in conditions with high hydrogen, I.e., severe accidents. In addition, because of their small processing capacity are ineffective In treating the dominant contributors to severe accident risk.

Hydrogen recombiners are precluded from operating In conditions with high hydrogen, I.e., severe accidents.

Negligible Impact on risk results from adding hydrogen recombiners.

Considering uncertainty and potential Impacts from external events does not introduce any significant changes. No change to the screening criteria category.

NA Considering uncertainty and potential Impacts from external events does not introduce any significant changes. No change to the screening criteria category.

N/A I

A I

I 72

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised DisposItion Phase II SAMA ID enhancement I Screening Criteria Including Uncertainty and SAMA ID number

[

External Events number 43 Create a large concrete SAMA would ensure that

5 - Cost would be Core retention devices have been Considering uncertainty and N/A crucible with heat removal molten core debris escaping more than risk investigated In previous studies. IDCOR potential impacts from potential under the from the vessel would be benefit concluded that 'core retention devices are external events does not basemat to contain molten contained within the not effective risk reduction devices for Introduce any significant core debris.

crucible. The water cooling degraded core events'. Other evaluations changes. No change to the mechanism would cool the have shown the worth value for a core screening criteria category.

molten core, preventing a retention device to be on the order of melt-through of the

$7000 (averted cost-risk) compared to an basemat estimated Implementation cost of over $1 million (per unit).

44 Create a water-cooled SAMA would contain molten #5 - Cost would be Core retention devices have been Considering uncertainty and N/A nubble bed on the core debris dropping on to more than risk investigated In previous studies. IDCOR potential Impacts from pedestal.

the pedestal and would benefit concluded that core retention devices are external events does not allow the debris to be not effective risk reduction devices for introduce any significant cooled.

degraded core events'. Other evaluations changes. No change to the have shown the worth value for a core screening criteria category.

retention device to be on the order of

$7000 (averted cost-risk) compared to an estimated implementation cost of over $1 million (per unit).

45 Provide modification for SAMA would help mitigate

4 - No significant BWR Mark I risk Is typically dominated by Considering uncertainty and N/A flooding the drywell head.

accidents that result In the safety benefit events that result In early failure of the potential Impacts from leakage through the dryweil drywellshell due to direct contact with external events does not head seal.

core debtis and events that bypass the Introduce any significant containment. This is also tMe at Dresden.

changes. No change to the The head flooding system would, screening criteria category.

therefore, not be expected to have any significant Impact on the overall risk.

The potential for competing risks due to Reactor Building flooding Is considered to eliminate any positive safety benefit.

73

Table 7-2 Revised Phase I SAMA Dispositlon (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original / Revised Original DisposItIon Revised Disposion Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 46 Enhance fire protection SAMA would improve

4 - No significant Current Standby Gas Treatment Systems Considering uncertainty and N/A system and/or standby fission product scrubbing In safety benefit do not have sufficient capacity to handle potential Impacts from gas treatment system severe accidents.

the loads from severe accidents that result external events does not hardware and procedures.

In a bypass or breach of the containment Introduce any significant Loads produced as a result of RPV or changes. No change to the containment blowdown would require large screening criteria category.

filtering capacities. These filtered vented systems have been previously investigated and found not to provide sufficient cost benefit Dresden has limited fire protection sprinkler systems In the Reactor Building.

Use of these for fission product scrubbing In the R.B. could create competing risks associated with spray failures and flooding of equipment with very limited potential benefit 50 Create a core melt source SAMA would provide

5 - Cost would be Core retetion devices have been Considering uncertainty and N/A reduction system.

cooling and containment of more than risk Investigated in previous studies. IDCOR potential Impacts from molten core debris.

benefit concluded that core retention devices are external events does not Refractory material would not effective risk reduction devices for introduce any significant be placed underneath the degraded core events'. Other evaluations changes. No change to the reactor vessel such that a have shown the worth value for a core screening criteria category.

molten core falling on the retention device to be on the order of material would melt and

$7000 compared to an estimated combine with the material.

implementation cost of over $1 miNion.

Subsequent spreading and heat removal form the vitrified compound would be facilitated, and concrete attack would not occur 74

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / RevIsed Original Disposition Revised Disposition Phase i SAMA ID enhancement Screening Criteria including Uncertainty and SAMA ID number External Events number 53 Install a secondary SAMA would filter fission

5 - Cost would be Secondary containment at Dresden makes Considering uncertainty and N/A containment filter vent.

products released from more than isk extensive use of blow out panels to protect potential impacts from primary containment, benefit the structural integrity of the building In the external events does not event of nternal pressure challenges such Introduce any significant as stearnilne breaks In the reactor building changes. No change to the or external pressure challenges such as screening critera category.

tornadoes. Major structural redesign of the reactor building would be required to make the reactor building capable of retaining and processing a primary containment failure.

54 Install a passive SAMA would provide

5 - Cost would be A passive system is another alternative Considering uncertainty and N/A containment spray redundant containment more than risk enhancement for the Contalnment Spray potential impacts from system.

spray method without high benefit function. See SAMA 35. Cost expected to external events does not cost exceed the maximum averted cost-risk for Introduce any significant Dresden ($457,000) changes. No change to the l_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

_______________________________ screening criteria category.

55 Strengthen SAMA would reduce the

5 - Cost would be BWR Mark I risk Is typically dominated by The cost Is considered to be N/A primary/secondary probability of containment more than risk events that result In early failure of the greater than the upper containment.

overpressurization to failure. benefit drywell shell due to direct contact with bound maximum averted core debris and events that bypass the cost risk of $2.OM. No containment. Strengthening the primary change to the screening

/secondary containment would have a criteria category.

small Impact on the overall risk of these accidents. Reference 17 discusses the cost of increasing the containment pressure and temperature capacity, which Is effectively strengthening the containment. This cost Is estimated assuming the change Is made during the design phase whereas for Dresden, the changes would have to be made as a retrofit The cost estimated for the ABWR was $12 million and t Is judged that retrofitting an existing containment would cost more. The cost of Implementation for this SAMA exceeds the maximum averted cost-risk for Dresden ($457,000).

75

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 56 Increase the depth of the SAMA would prevent

5 - Cost would be Core retention devices have been Considering uncertainty and N/A concrete basemat or use basemat melt-through.

more than risk Investigated In previous studies. IDCOR potential Impacts from an alternative concrete benefit concluded that core retention devices are external events does not material to ensure melt-not effective risk reduction devices for introduce any significant through does not occur.

degraded core events'. Other evaluations changes. No change to the have shown the worth value for a core screening criteria category.

retention device to be on the order of

$7000 compared to an estimated Implementation cost of over $1 mililon/site.

57 Provide a reactor vessel SAMA would provide the

5 - Cost would be This has been estimated to cost $2.5 The cost Is considered to be N/A exterior cooling system.

potential to cool a molten more than risk million and exceeds the maximum averted greater than the upper core before it causes vessel benefit cost-risk for Dresden ($457,000). ORNL bound maximum averted failure, If the lower head (35] has performed thermal hydraulic cost risk of $2.OM. No could be submerged In calculations on BWR external cooling change to the screening water.

methods and determined that the current criteria category.

BWR RPV support skirt design makes it impractical to cool the RPV by external cooling to prevent RPV breach.

Therefore, the modification would require RPV support skirt modification and reanalysis to allow the external cooling to be effective.

58 Construct a building to be SAMA would provide a

5 - Cost would be Based on engineering udgement, the cost The cost Is considered to be N/A connected to method to depressurize more than risk of this enhancement Is expected to greatly greater than the upper primary/secondary containment and reduce benefit exceed the maximum averted cost risk bound maximum averted containment that Is fission product release.

($0.4 million).

cost risk of $2.OM. No maintained at a vacuum.

change to the screening criteria category.

76

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 64 1.h. Simulator Training for SAMA would lead to

4 - No significant Simulators could be upgraded and used to Considering uncertainty and N/A Severe Accident improved arrest of core melt safety benefit provide operator training for severe potential impacts from progress and prevention of accidents; however, these scenarios are external events does not containment failure rare and the instruction time would introduce any significant compete with time required to train changes. No change to the operators on more likely scenarios that are screening criteria category.

severe accident precursors. The benefit of simulator training is difficult to quantify as the results would be based on the improved reliability of human actions In the mitigation of severe accidents. Training can positively influence the values of HEPs, but the impact Is small. In addition, the TSC would be manned in a severe accident evolution and could provide additional support by personnel familiar with the SAMOs.

Previously assessed by the NRC as not required to support Accident management because of marginal cost benefit 66 3.a. Larger Volume SAMA increases time

5 - Cost would be Enlargement of the containment would be The cost Is considered to be N/A Containment before containment failure more than risk similar In scope to the ABWR design greater than the upper and increases time for benefit change SAMA to Implement a iarger bound maximum averted recovery volume containment, but would likely cost risk of $2.OM. No exceed the $8 million estimate for that change to the screening change as a retrofit would be required.

criteria category.

This is greater than the maximum averted l ________

cost-risk ($457,000).

___________________45_0 0

77

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMAtle Result of potential Original / Revised Original Disposition Revised Disposition Phase i SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 68 3.c. Improved Vacuum SAMA reduces the

5 - Cost would be The Dresden plant has six (6) Individual Considering uncertainty and N/A Breakers (redundant probability of a stuck open more than risk vacuum breaker lines with two vacuum potential Impacts from valves In each line) vacuum breaker.

benefit breakers In parallel In each line. Providing external events does not redundant vacuum breakers in each line introduce any significant would decrease the potential for vapor changes. No change to the See Table 6 and Section suppression failure and suppression pool screening criteria category.

A.4.3.3 of ABWR SAMDAs.

bypass. Ths plant modification requires new valves, the structural changes to Implement the modification, and the outage time to Install. Based on the PRA results that vapor suppression failure and pool bypass are negligible risk contributors and the apparent extremely high cost, this proposed SAMA Is not considered cost effective.

92 Provide additional DC SAMA would ensure longer

3 - Already Dresden already has Included spare Considering uncertainty and N/A battery capacity.

battery capability during an Installed.

batteries. These can be used to extend IC potential Impacts from SBO, reducing the operability and allow more credit for AC external events does not frequency of long-term SBO power recovery. This would decrease the introduce any significant sequences.

frequency of core damage and ofisIte changes. No change to the releases.

screening criteria category.

4 - No significant safety benefit.

The addition of 250V DC batteries could be evaluated to provide all the HPCI DC power requirements. However, room cooling and torus cooling would be more limiting.

93 Use fuel cells instead of SAMA would extend DC

5 - Cost would be Further extension of battery lfe with fuel The anticipated N/A lead-add batteries.

power availability in an more than risk cells Is estimated to have a small Impact Implementation cost is SBO.

benefit on the Dresden residual risk profile. In judged to exceed the addition, the cost of hardware (fuel cells),

benefit even If the benefit Is engineering, and hazard analysis Is increased by almost a factor expected to exceed the maximum cost of five to account for averted of $457,000.

uncertainty and potential impacts from external events. No change to the screening criteria category.

78

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised DIsposItIon Phase 1 SAMA ID enhancement Screening Criteria including Uncertainty and SAMA ID number External Events number 107 Create a backup source This SAMA would provide a

5-Cost would be A new system for diesel cooling would The anticipated N/A for diesel cooling. (Not redundant and diverse more than risk require extensive engineering, safety Implementation cost Is from existing system) source of cooling for the benefit analysis, hardware and labor for Judged to exceed the diesel generators, which Installation. This would exceed the benefit even If the benefit Is would contribute to

$457,000 maximum averted cost.

increased by almost a factor enhanced diesel reliablity.

of five to account for uncertainty and potential Impacts from external events. No change to the screening criteria category.

110 Bury offsite power lines.

SAMA could improve olfsite #5 - Cost would be While the actual cost of this SAMA will The cost Is considered to be NWA power reliability, particularly more than risk vary depending on site characteristics, the greater than the upper during severe weather.

benefit cost of burying offslte power lines has bound maximum averted been estimated at a cost significantly cost risk of $2.OM. No greater than $25 million for another change to the screening commercial US nuclear plant.

criteria category.

Implementing this SAMA at Dresden Is considered to be within the same order of magnitude and exceeds the maximum averted cost-risk for the plant ($457,000).

l 113 Provide DC power to the SAMA would increase the

4 - No significant

1) Loss of 120V AC Is not an Initiating Considering uncertainty and N/A 120/240-V vital AC system reliability of the 120-VAC safety benefit Event potential Impacts from from the Class IE station Bus.

2) 120 VAC s not a risk significant external events does not service battery system support system (from a risk reduction introduce any significant instead of Its own battery.

worth perspective that is key for Me changes. No change to the SAMA analysis]

screening criteria category.

120 9.1. Improved SAMA would provide

4-No significant

1) Loss of 120V AC Is not an Initiating Considering uncertainty and N/A Uninterruptable Power Increased reliability of safety benefit Event potential Impacts from Supplies power supplies supporting

2) 120 VAC Is not a risk significant external events does not front-line equipment, thus support system [from a risk reduction Introduce any significant reducing core damage and worth perspective that Is key for the changes. No change to the release frequencies.

SAMA analysis]

screening criteria category.

79

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title I

Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 124 10.a. Dedicated DC This SAMA addresses the

5 - Cost would be Dresden has the capability to operate the The cost Is considered to be N/A Power Supply use of a diverse DC power more than risk Isolation Condenser (once Initiated) greater than the upper system such as an benefit without DC power. This is included In the bound maximum averted additional battery or fuel cell Dresden PRA as a success path. The cost risk of $2.OM. No for the purpose of providing cost of mplementation for this mod Is change to the screening motive power to certain estimated at $3 million, which Is greater criteria category.

components (e.g., HPCI).

than the maximum averted cost-risk for Dresden ($457,000).

129 Add an automatic bus Plants are typically sensitive #4 - No significant

1) Loss of 120V AC Is not an Initiating Considering uncertainty and N/A transfer feature to allow to the loss of one or more safety benefit Event potential Impacts from the automatic transfer of 120V vital AC buses.

2) 120 VAC Is not a risk significant external events does not the 120V vital AC bus Manual transfers to support system [from a risk reduction introduce any significant from the on-line unit to the aitemate power supplies worth perspective that Is key for e

changes. No change to the standby unit could be enhanced to SAMA analysis]creenng critea category.

transfer automatically.

138 Locate residual heat SAMA would prevent

5 - Cost would be Competing risks associated with such a The cost is considered to be N/A removal (RHR) Inside of intersystem LOCA more than risk design are manifold and would require greater than the upper containment.

(ISLOCA) out the RHR benefit extensive analysis to demonstrate bound maximum averted pathway.

capability. For an existing plant, the cost cost risk of $2.OM. No of moving an entire system is judged to change to the screening greatly exceed the maximum averted cost-criteria category.

risk for Dresden ($457,000). Related to mitigation of an ISLOCA. Per IN-92-36, and ts additional supplement, ISLOCA contributes little risk for BWRs, because of the lower primary system pressures.

139 Install additional SAMA would decrease

4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty and N/A Instrumentation for ISLOCA frequency by safety benefit IN-92-36, and Rs additional supplement, potential impacts from ISLOCAs.

installing pressure of leak ISLOCA contributes little risk for BWRs, external events does not monitoring instruments In because of the lower primary system introduce any significant between the first two pressures.

changes. No change to the pressure Isolation valves on screening criteria category.

low-pressure Inject lines, RHR suction lines, and HPSI lines.

80

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original Revised Original Disposition Revised Disposition Phase i SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 140 Increase frequency for SAMA could reduce

4-No significant Related to mitigation of an ISLOCA. Per Considering uncertainty and NMA valve leak testing.

ISLOCA frequency.

safety benefit IN-92-36, and Its additional supplement, potential Impacts from ISLOCA contributes little risk for BWRs, external events does not because of the lower primary system introduce any significant pressures.

changes. No change to the screening criteria category.

141 Improve operator training SAMA would decrease

4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty and NA on ISLOCA coping.

ISLOCA effects.

safety benefit IN-92-36, and Its additional supplement, potential Impacts from ISLOCA contributes Me risk for BWRs, external events does not because of the lower primary system introduce any significant pressures.

changes. No change to the screening criteria category.

In addition, the Dresden EOPs provide secondary containment monitoring parameters which Include room specific temperature, room specific radiation, vent radiation, and room specific water level.

The instrumentation and procedural guidance help locate and Isolate breaks which have bypassed primary containment.

143 Provide leak testing of SAMA would help reduce

4-No significant Related to mitigation of an ISLOCA. Per Considering uncertainty and N/A valves In ISLOCA paths.

ISLOCA frequency. At safety benefit IN-92-36, and Its additional supplement, potential Impacts from Kewaunee Nuclear Power ISLOCA contributes little risk for BWRs, external events does not Plant, four MOVs isolating because of the lower primary system introduce any significant RHR from the RCS were pressures.

changes. No change to the not leak tested.

screening criteria category.

81

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 145 Ensure all ISLOCA SAMA would scrub all

4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty and N/A releases are scrubbed.

ISLOCA releases. One safety benefit IN-92-36, and its additional supplement, potential Impacts from example Is to plug drains in ISLOCA contributes little risk for OWRs, external events does not the break area so that the because of the lower primary system introduce any significant break point would cover pressures.

changes. No change to the with water.

screening criteria category.

The cost of performing the analysis to Identify aN ISLOCA pathways and to ensure that any physical modifications implemented to mitigate ISLOCAs are not detrimental to the plant (e.g.. cause flooding hazards) combined with the cost of Installing the required equipment s judged to greatly exceed any benefit Additonally, the suggested enhancement of plugging drain lines would not guarantee a release would be scrubbed as the release may occur above the break location. Room flooding equipment and waterproofing of mitigative components would be required to make this SAMA potentially effective. Such changes would be extremely costly and potential competing risk appears to signifIcantly outweigh any possible safety benefit.

146 Add redundant and SAMA could reduce the

4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty and NA diverse limit switches to frequency of containment safety benefit IN-92-36, and Its additional supplement, potential mpacts from each containment Isolation failure and ISLOCA contributes litffle risk for BWRs, external events does not isolation valve.

ISLOCAs through enhanced because of the lower primary system introduce any significant isolation valve position pressures.

changes. No change to the Indication.

screening criteria category.

147 Early detection and SAMA would limit the

4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty and N/A mitigation of ISLOCA effects of ISLOCA accidents safety benefit IN-92-36, and Its additional supplement, potential impacts from by early detection and ISLOCA contributes little risk for BWRs, external events does not isolation because of the lower primary system Introduce any significant pressures.

changes. No change to the screening criteria category.

82

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA tMe Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 148 8.e. Improved MSIV This SAMA would decrease #4 - No significant Redundant MSIVs are designed to Isolate Considering uncertainty and N/A Design the likelihood of safety benefit on severe accidents that could lead to potential Impacts from containment bypass radionuclide release and bypass external events does not scenarios.

containment. These Include breaks introduce any significant outside containment The MSIVs are leak changes. No change to the tested to ensure their adequacy. The screening criteria category.

maintenance Rule program monitors the performances of the MSIVs providing early feedback on any degradation.

The PRA has determined that the risk contribution from MSIV failures to Isolate is l _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

very sm a ll.

153 Modify swing direction of SAMA would prevent flood

4 - No significant Dresden plant configuration Is not Considering uncertainty and N/A doors separating turbine propagation, for a plant safety benefit susceptible to flood propagation from the potential Impacts from building basement from where internal flooding foi Turbine Building to adjacent buildings with external events does not areas containing turbine building to safety equipment Flooding from Turbine introduce any significant safeguards equipment safeguards areas Is a Hall Into adjacent buildings considered to changes. No change to the concern.

have negligible Impact screening criteria category.

155 Implement nternal flood This SAMA would reduce

4 - No significant The total contribution to CDF from ntemal Table 1-2 In Response 1(b)

N/A prevention and mitigation the consequences of safety benefit flooding Is 1.8E-7/yr or less than 10% of indicates that the current enhancements.

internal flooding.

the total nternal events CDF. nternal contribution from nternal flood Is not considered to be a dominant flooding Is about 3%.

contributor to the CDF at Dresden and Considering uncertainty and adequate precautions and training are potential Impacts from believed to be In place to prevent and external events does not respond to postulated flood.

introduce any significant changes. No change to the screening criteria category.

83

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 157 Shield electrical SAMA would decrease risk

5 - Cost would be Protecting equipment from spray may be a The anticipated N/A equipment from potential associated with seismically more than sk cost beneficial means of reducing risk at implementation cost s water spray induced internal flooding benefit Dresden. However, there are very few, If judged to exceed the any, locations that can be effectively benefit even If the benefit Is protected from water spray adverse effects Increased by almost a factor that are not already protected. This fact of five to account for coupled with the knowledge that the total uncertainty and potential CDF from all internal floods is so low, impacts from external means that any plant modification is nearly events. No change to the impossible to justify. The 4-kV emergency screening criteria category.

buses in Reactor Building have water hoods. Some MCCs have small hoods.

Additional spray protection could be provided to swltchgear In Turbine Building.

Main risk reduction would be from providing water spray protection to Unit 3 125 VDC battery bus and switchgear In cage outside of Unit 3 Battery Charger room.

84

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase i SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 164 Install a new condensate Elter replace the existing

4 - No significant For SBO conditions, the CST contains Considering uncertainty and N/A storage tank (CST) tank with a larger one, or safety benefit enough water to allow make-up injection potential Impacts from install a back-up tank.

from HPCI for a period longer than Rs external events does not estimated operability (based on battery Introduce any significant life). The 1A, 2/3A and 2t3B CSTs have a changes. No change to the combined nominal water volume (typical) screening criteria category.

of 410,000 gallons. For LOCA initators, the CST does not contain enough water to provide Injection for the 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months mission time. The CST makeup systems do not currently have the capacty to match the inventory loss for a LOCA. Feedwater has connections to unlimited water supplies (SBCS) not dependent on the CST.

CST connections to Core Spray and LPCI already exist. The ability to refill the CST from external water sources Is considered both desirable and not difficult The Technical Support Guidelines (TSGs)

Appendix J provides the makeup sources available to Dresden to allow CST refill.

The Isolation Condenser (IC) which Is a separate mitigation system also has significant makeup capabilities Independent of the CST. The TSG Appendix K cites the systems that can make-up to the shell side of the IC. This represents a significant benefit over other plants without an IC.

85

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 165 Provide cooling of the This SAMA would improve

5 - Cost would be AFW is a PWR system for steam The anticipated N/A steam-driven AFW pump success probability in an more than risk generator make-up njection. The HPCI Implementation cost is In an SBO event SBO by (1) using the FP benefit pump at Dresden Is equivalent in many judged to exceed the system to cool the pump, or respects to the PWR AFW pump. The benefit even if the benefit Is (2) making the pump self HPCI turbine requires room cooling over a increased by almost a factor cooled, or (3) providing a 24 hour2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months mission time or the SBO mission of five to account for fan cooling capability.

time of 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months . Installation of an uncertainty and potential additional room cooling system for HPCI Impacts from external that would be Independent of AC and DC events. No change to the power would be the only type of Isystem screening criteria category.

that would change the risk profile. This additional system Is expected to cost more than the maximum cost averted of

$457,000 and therefore to not be cost beneficial.

86

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase F SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 166 Proceduralize iocal This SAMA would lengthen

4 - No significant AFW is a PWR system for steam Considering uncertainty and N/A manual operation of AFW AFW availability In an SBO. safety benefit generator make-up injection. HPCI is the potential mpacts from when control power Is lost. Also provides a success turbine driven Injection system for external events does not path should AFW control Dresden. The available njection time for introduce any significant power be lost in non-SBO these systems is limited by factors such as changes. No change to the sequences.

battery life, depressurization on HCTL.

screening criteria category.

and injection source volume. HCTL is 167 Provide portable This SAMA would extend

4 - No significant reached In the suppression pool at Considering uncertainty and NM generators to be hooked AFW availability in an SBO safety benefit approximately 7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months after the Initiating potential Impacts from Into the turbine driven (assuming the turbine event of an SBO without IC operation.

external events does not AFW, after battery driven AFW requires DC Providing local, manual control capability introduce any significant depletion.

power) for the HPCi system (removing the DC changes. No change to the dependence) could extend Injection an screening criteria category.

additional three hours beyond the 4 hour4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months battery life. However, hardware changes would be necessary In addition to procedure updates for Dresden.

For SBOs with the IC operating, HPCI could extend the time of adequate core cooling (by providing RPV makeup for seal LOCA events). This operation of HPCI will allow adequate core cooling to be extended as long as the battery supply of DC can be preserved or the battery (DC) requirement bypassed by manual action.

HPCI room cooling Is the limiting condition under this scenario.

DC power Is not the limiting support system for HPCI operation. The room cooling requirement for AC power for the HPCI fan Is most limiting. This SAMA for local generaton of HPCI without DC does not result In any noticeable change In CDF because of the small failure profitability of DC and the presence of more limiting failure modes (i.e., room cooling).

Therefore, the potential benefit for this modification Is very small.

87

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 172 Install an Independent This SAMA would allow

4 - No significant HPCI Is the turbine driven injection system Considering uncertainty and N/A diesel generator for the continued inventory make-safety benefit for Dresden. The 1A, 2/3A and 2M3B CSTs potential impacts from CST make-up pumps up to the CST during an have a combined nominal water volume external events does not SBO.

(typical) of 410,000 gallons. Given a introduce any significant battery ife of 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months (required for HPCI changes. No change to the operation), no additional water source screening criteria category.

would be required for injection during the 4 hour4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months SBO mission time. Mnimal benefit would be gained from this SAMA.

Even if CST water Is exhausted, the switchover of suction from the CST to the torus would continue to allow HPCI Injection. The limitlng time and action for HPCI effectiveness in an SBO (other than batteries) or other accident sequences without DHR is the torus water temperature greater than HCTL This leads to RPV depressurization and the unavailability of HPCI as an effective RPV make up method regardless of CST volume. Therefore, there is negligible risk benefit associated with increasing CST make up capability under SBO conditions.

The Technical Support Guidelines (TSGs)

Appendix J provides the makeup sources available to Dresden to allow CST refill.

The Isolation Condenser (IC) which is a separate mitigation system also has significant makeup capabilities independent of the CST. The TSG Appendix K cites the systems that can make-up to the shell side of the IC. This represents a significant benefit at Dresden over other plants without an IC.

88

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Dispositlon Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 178 Provide an addfflonal This SAMA would reduce

5 - Cost would be This is primarily a PWR insight where RPV The anticipated N/A HPSI pump with an the frequency of core melt more than risk depressurization Is not as easily available.

implementation cost is Independent diesel from smaN LOCA and SBO benefit The availability of an additional high judged to exceed the sequences pressure water injection source is not a benefit even if the benefit is significant risk reduction measure for Increased by almost a factor Dresden because of the existing design.

of five to account for uncertainty and potential Dresden has substantial high pressure impacts from external RPV Inventory control methods. These events. No change to the include:

screening criteria category.

HPCI Feedwater (motor driven)

Isolation Condenser CRD pumps These methods represent substantial high pressure inventory control methods including active HPSI from the turbine driven HPCI system which Is independent of AC power initially.

Dresden has a turbine driven high pressure Injection with the capability to provide a supplement or an alternative to the Isolation Condenser (IC) system for safe shutdown.

FW depends on offsite AC power to provide high-pressure Injection.

Onsite AC power Is available from either unit EDG the swing EDG, or either S0 DG (5 sources) to support CRD operation.

Because of the cost associated with this SAMA and the existing Dresden capability, a negligible change in risk is calculated.

Even the maximum cost averted

($457,000) could not justify the engineering and hardware of an additional pump.

89

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised DIsposition Phase i SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 184 Upgrade Chemical and For a plant like the AP600

5 - Cost would be A potential functional equivalent for The cost Is considered to be N/A Volume Control System to where the Chemical and more than risk Dresden would be the enhancement of the greater than the upper mitigate small LOCAs.

Volume Control System benefit RWCU system such that Injection flow bound maximum averted cannot mitigate a Small rates on the order of 1000 gprn were cost risk of $2.OM. No LOCA, an upgrade would possible. This change is considered to be change to the screening decrease the Small LOCA similar In function, scope, and cost to criteria category.

CDF contribution.

SAMA 185 ($5-$10 million) with the exception of the Independent power source. However, new power circuits and wiring would likely be needed for the larger pumps. The low end of the cost of implementation estimate ($5 million) Is judged to be applicable for this SAMA, which Is greater than the maximum averted cost risk for Dresden ($457,000).

187 Replace 2 of the 4 safety This SAMA would reduce

4-No significant Dresden has a diverse set of injection Considering uncertainty and N/A injection (SI) pumps with the Si system common safety benefit systems and more than one method of potential impacts from diesel-powered pumps.

cause failure probability.

containment heat removal. Common external events does not This SAMA was intended cause failure of the 4 train LPCI system Is Introduce any significant for the System 80+, which a low contributor to risk and removing the changes. No change to the has four trains of SI.

4/4 system failures would have minimal screening criteria category.

Impact on the results. The CCF of all four LPCI pumps to fall to start or run (2LIPM-2ABCD14ACC, 2LIPM-2ABCD14XCC) does not appear in any CDF cutsets above the truncation limit for the plant model and would not impact the results I it were Improved.

188 Align low pressure core This SAMA would help to

3 - Already This Is already directed at Dresden.

However, a cutset review 11 injection or core spray to ensure low pressure ECCS Implemented at Indicates that this action s the CST on loss of can be maintained In loss of Dresden Important In loss of service suppression pool cooling.

suppression pool cooling water initiated events. The scenarios.

potential benefit from Revise to:

improving the HEP value

6 Retain associated with this existing action is explored as part of this RAI response.

90

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 189 Raise high pressure core This SAMA would ensure

4 - No significant The HPCI high backpressure trip Is Considering uncertainty and N/A Injectionlreactor core high pressure core safety benefit already set at a pressure above the potential Impacts from isolation cooling injection/reactor core containment ultimate pressure; thus, external events does not backpressure trip isolation cooling avaliability raising the trip limits would have no introduce any significant setpoints when high suppression pool Impact.

changes. No change to the temperatures exist.

screening criteria category.

190 Improve the reflablity of This SAMA would reduce

5 - Cost would be High pressure melt scenarios are The anticipated N/A the automatic the frequency of high more than risk significant contributors to the Dresden Implementation cost Is depressurizatlon system.

pressure core damage benefit CDF. The SAMA Is Interpreted to mean Judged to exceed the sequences.

Improved reliability of the ERVs and benefit even if the benefit Is Target Rock SRVs and their support increased by almost a factor systems. A plant modification to eliminate of five to account for dependence on DC power to Increase the uncertainty and potential success probability of these valves would impacts from external reduce the high pressure Inlection events. No change to the accident classes of IA and IE.

screening criteria category.

No such design is currently available. This would require a research and development project and would exceed the maxdmum cost averted of $457,000.

91

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria including Uncertainty and SAMA ID number External Events number 193 Proceduralize Intermittent SAMA would allow for

4 - No significant Umitatons on HPCI operation In an SOO Considering uncertainty and N/A operation of HPCI.

extended duration of HPCI safety benefit are based on battery depletion. Multiple potential Impacts from availability.

starts and stops of the system are a larger external events does not drain on the battery than continuous introduce any significant operation with excess flow directed to the changes. No change to the torus. In addition, multiple starts of the screening criteria category.

system Introduce additional start demands which may Increase the system failure probability for a given period of operation.

The principal sequence dependent limitation for operation of HPCI Is battery life In SO and HCTL In other sequences where LPCI suppression pool cooling Is not available. Negligible benefit has been Identified for this SAMA at Dresden.

HPCI pump operation must be controlled for SO to preclude the minimum flow valve operation from dumping excessive amounts of CST water to the torus. HPCI In the CST pressure control mode Is recommended and currently preferred operating mode of HPCI.

l 194 Increase available net SAMA increases the

5 - Cost would be Requires major plant changes such as The anticipated N/A positive suction head probability that these pumps more than risk new LPCI/CS pumps, moving the LPCI Implementation cost Is (NPSH) for Injection will be available to Inject benefit pumps, a new suppression pool design, a judged to exceed the pumps.

coolant Into the vessel by larger CST (only applicable for Injection benefit even If the benefit is increasing the available phase), or an additional containment Increased by almost a factor NPSH for the injection cooling system. The cost of these changes of five to account for pumps.

would exceed the maximum averted cost-uncertainty and potential risk for Dresden.

Impacts from external events. No change to the screening criteria category.

l 92

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Soreening Criteria Including Uncertainty and SAMA ID number External Events number 195 Modify Reactor Water SAMA would provide an

5 - Cost would be RWCU heat removal capacity Is too low The cost Is considered to be N/A Cleanup (RWCU) for use add itonal source of decay more than risk for decay heat removal.

greater than the upper as a decay heat removal heat removal.

benefit bound maximum averted system and proceduralize cost risk of $2.OM. No use.

In order to make RWCU a viable heat change to the screening removal system, the piping, pumps, heat criteria category.

exchangers, and power sources would have to be upgraded. This SAMA is considered to be similar in scope to SAMA 191. The cost of Implementation for such a change (approximately $5 million) s greater than the maximum averted cost-risk for Dresden ($457,000).

199 Re-open MSIVs SAMA to regain the main

6 - Retain There are two important aspects of the Still retained.

4 condenser as a heat sink by MSIV closure response:

re-opening the MSIVs.

For non-ATWS conditions, the ability to rapidly respond to MSIV closure and restore the main condenser as a heat sink Is not explicitly directed.

For ATWS conditions, Dresden EOPs direct MSIV low level closure bypass n order to retain the main condenser as a heat sink; however, this assumes the MSIVs have not yet closed.

For both cases, explicit procedural direction to re open the MSIVs could be included.

93

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original Revised Original Disposition Revised DisposItion Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 201 2.a. Passive High SAMA will Improve

3 - Already Dresden has an IC which provides the The anticipated N/A Pressure System prevention of core melt Instalted.

capability for passive Inventory control for Implementation cost Is sequences by providing a short time following scram. Active judged to exceed the additional high pressure systems are used for IC shell makeup and benefit even If the benefit Is capability to remove decay RPV makeup due to Recirculation pump Increased by almost a factor heat through an Isolation

5 - Cost would be seal leakage.

of five to account for condenser type system more than risk uncertainty and potential benefit Impacts from external The addition of tanks for IC makeup and events. No change to the another Active system for RPV makeup screening criteria category.

make the 'passivew feature not cost beneficial.

The cost of this enhancement has been estimated to be $1.7 million In Reference

17. This Is greater than the maximum averted cost-risk for Dresden ($457,000).

202 2.c. Suppression Pool SAMA will Improve

5-Cost would be From a review of the contributors to the The anticipated N/A Jockey Pump prevention of core melt more than risk Dresden risk profile, It Is found that the Implementation cost Is sequences by providing a benefit availability of low pressure pumps for RPV judged to exceed the small makeup pump to make up Is not a dominant contributor.

benefit even the benefit s provide low pressure decay The low pressure pump availability for Increased by almost a factor heat removal from the RPV RPV Injection Is a negligible contributor to of five to account for using the suppression pool the risk profile. The expense of adding uncertainty and potential as a source of water.

another low pressure Injection system Impacts from extemal without Introducing severe competing risks events. No change to the Is expected to be high. it can be screening criteria category.

concluded that the cost will not be able to be justified.

94

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11i SAMA ID enhancement Screening Criteria including Uncertainty and SAMA ID number External Events number 207 4.c. High Flow SAMA would improve

5 - Cost would be The Suppression Pool Cooling system is The cost Is considered to be N/A Suppression Pool Cooling suppression pool cooling.

more than risk already sized to accommodate flow to greater than the upper for ATWS response benefit remove all decay heat and operate under bound maximum averted ATWS conditions with SBLC Injection cost risk of $2.OM. No success.

change to the screening criteria category.

Increasing the capabilities of suppression pool would require new pumps, heat exchangers, piping, and other equipment.

The mplementation cost of this change is considered to be approximately equivalent to SAMA 35 ($5.8 million) and is screened from further review as it Is significantly greater than the maximum averted cost-risk for Dresden ($457,000).

211 Install nitrogen bottles as This SAMA would extend

4 - No significant Dresden depressurization capability Is Considering uncertainty and N/A a back-up gas supply for operation of safety relief safety benefit primarily supported by DC power. The potential Impacts from safety relief valves.

valves during an SBO and EMRVs are powered by 125V DC and are external events does not loss of air events (WRs).

available during an SBO. The single Introduce any significant Target Rock SRV uses nitrogen pneumatic changes. No change to the supply as the motive power to open the screening criteria category.

valve against spring pressure, but 125V DC Is still required for valve control. An accumulator is available to allow a limited number of SRV openings after loss of Drywoll Air.

Because of the SRV redundancy with the EMRVs, only a negligible change In risk would be achieved.

95

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement ScreenIng Criteria Including Uncertainty and SAMA ID number External Events number 215 Create cross-connect This SAMA would improve

5 - Cost would be Each units SLC system has two trains The anticipated N/A ablity for standby liquid reliability for boron injection more than risk which have common suction and implementation cost is control trains during an ATWS event.

benefit discharge headers. Redundant suction judged to exceed the and discharge paths exist beyond these benefit even If the benefit Is headers, which can be Isolated, If Increased by almost a factor required. No further cross connection is of five to account for beneficial between the trains of a given uncertainty and potential unit An inter unit cross-tie Is a potential Impacts from external enhancement However, because the events. No change to the SLC system response Is dominated by screening criteria category.

common cause failures of the explosive valves and the operator action to Initiate SLC, the ability for use of a cross tie will have limited benefit In the risk profile. This small change in the small ATWS contribution results in little potential safety Improvement, but a substantial cost 223 Bypass MSIV isolation In SAMA will afford operators

3 -Already BWROG EPC Issue 98-07 addresses this However, this action 12 Turbine Trip ATWS more time to perform Instafled.

issue. The bypass of the MSIV Isolation requires the use of Jumpers scenarios actions. The discharge of a was moved upward In the flowchart, with a limited time available, substantial fraction of steam rendering It more Important. Bypass of and as such carries a to the main condenser (i.e.,

Revise to:

MSIV solation Is procedurally directed in relatively high HEP value.

as opposed to into the

- Retain the DEOPs under failure to scram The potential benefit of primary containment) conditions.

implementing a dedicated affords the operator more low level interlock switch is time to perform actions explored as part of this RAI (e.g., SLC Injection, lower response.

water level, depressurize RPV) than If the main condenser was unavailable, resulting In lower human error probabilities 96

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original DlsposMon Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Unoertainty and SAMA ID number External Events number 229 Create/enhance RCS With either a new

5 -Cost would be PWR Issue related to the limited The anticipated N/A depressurization ability depressurization system, or more than risk depressurization capability of the PWR. In Implementation cost Is with existing PORVs, head benefit addition, reference 19 estimates the cost judged to exceed the vents, and secondary side of this SAMA to range between $500,000 benefit even If the benefit Is valve, RCS and $4.6 milion. For Dresden, more increased by almost a factor depressurization would effective depressurization capabilities of five to account for allow earlier low pressure would require significant hardware uncertainty and potential ECCS Injection. Even If changes and/or additions on top of the impacts from external core damage occurs, low analysis that would be required to events. No change to the RCS pressure would Implement the change. The cost estimate screening criteria category.

alleviate some concerns for the modification Is considered to be on about high pressure melt the high end of the range provided in election.

Reference 19. The cost of implementation for this SAMA Is Judged to greatly exceed the maximum averted cost-risk for Dresden ($457.000) l 233 Install secondary side This SAMA would prevent

5 -Cost would be This Is primarily a PWR issue. The steam The anticipated N/A guard pipes up to the secondary side more than risk lines for a BWR inside the inside MSIV are implementation cost Is MSIVs depressurization should a benefit completely within the containment judged to exceed the steam line break occur requiring no guard pipe. Between the two benefit even if the benefit Is upstream of the main steam MSIVs is a very short length of pipe that Increased by almost a factor isolation valves. This contributes a negligible amount to the CDF of five to account for SAMA would also guard and LERF. The addition of a guard pipe to uncertainty and potential against or prevent the steam tunnel for the short pipe length Impacts from extemal consequential multiple Is judged to be very expensive and events. No change to the SGTR following a Main substantially in excess of any potential screening criteria category.

Steam Une Break event benefit associated with risk reduction.

239 Increase seismic SAMA would Increase the

6 - Retain Components were Identified In the IPEEE Still retained.

ruggedness of plant availability of necessary whose seismic ruggedness could be components.

plant equipment during and Improved.

after seismic events.

Increase the seismic capacity of components Extends the safe shutdown on the safe shutdown path seismic capacity to at paths with capacities less least 0.3g.

than 0.3g to 0.3g.£i--_

97

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase SAMA ID enhancement Screening Criteria l Including Uncertainty and SAMA ID number External Events number 244 I.e. Improved Accident SAMA will improve

5 - Cost would be The risk as measured by CDF, LERF, and The anticipated

/A Management prevention of core melt more than risk population dose is low. The implementation cost Is Instrumentation sequences by making benefit instrurnentation available to the operating Judged to exceed the operator actions more crew at Dresden is comparable to that benefit even the benefit is reliable.

available at other BWRs. Based on a Increased by almost a factor review of the accident sequences that of five to account for contribute to the Dresden risk profile, the uncertainty and potential estimated risk reduction associated with Impacts from external additional accident mitigation events. No change to the instrumentation is judged to be negligible.

screening criteria category.

248 2.h. Safety Related SAMA will improve

5 - Cost would be The HPCI system has a safety related The anticipated N/A Condensate Storage Tank availablity of CST following more than risk water source from the torus. The cost of implementation cost is a Seismic event benefit engineering, installation, and safety judged to exceed the analysis of an additional large water benefit even the benefit Is source Is significantly greater than the Increased by almost a factor maximum cost averted $457,000.

of five to account for uncertainty and potential Impacts from external events. No change to the screening criteria category.

249 4.d. Passive Overpressure This SAMA will prevent

- Retain Dresden has Installed a hard piped Still retained.

Relief catastrophic failure of the containment vent system that provides a containment. Controlled controlled means of containment relief through a selected overpressure relief. The passive feature vent path has a greater of adding a rupture disk to this system potential for reducing the Introduces competing risks that limit the release of radioactive usefulness of the vent over the spectrum material than through a of severe accidents.

random break.

255 Train operations crew for This SAMA would Improve

4 - No significant The 120V AC system Is not risk significant Considering uncertainty and N/A response to Inadvertent chances of a successful safety benefit at Dresden [from a risk reduction worth potential Impacts from actuation signals response to the loss of two perspective that Is key for the SAMA external events does not 120V AC buses, which may analysis]. While other plants have introduce any significant cause inadvertent signal Identified specific 120V AC failure changes. No change to the generation.

scenarios that would lead the generation screening criteria category.

of Inadvertent signals, no comparable vulnerablities have been Identified at Dresden.

98

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I 1 SAMA title Result of potential Original Revised Original Disposition Revsed Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncerainty and SAMA ID number External Events number 256 Install tornado protection This SAMA would improve

4 - No significant No gas turbines on-site. Additional Considering uncertainty and N/A on gas turbine generators onsito AC power reliability.

safety benefit measures could be taken to improve the potential impacts from protection of other on-site AC power external events does not sources; however, the PEEE investigated Introduce any significant risk from high wind events and found It to changes. No change to the be n egligible.

screening criteria category.

259 Diversify the explosive An alternate means of

6-Retain SBLC Injection failure Is a dominant Still retained.

valve operation opening a pathway to the contributor to ATWS mitigation failure.

RPV for SBLC injection Evaluate SBLC system mprovements.

would improve the success probability for reactor shutdown.

260 Enrich Boron The Increased boron

6 - Retain Increasing the boron concentration for Still retained.

8 concentration will reduce SBLC may be a cost effective means of the time required to achieve reducing AIWS risk.

the shutdown concentration.

This will provide increased margin in the accident timeline for successful operator activation of SOLC.

_=

261 Bypass Low Pressure LPCI and CS injection

6-Retain A reduction in this CCF will result in a Still retained.

9 Permissive valves require a permissive small decrease In CDF.

signal from the same 2 pressure sensors In order to open. The nstruments are

. currently specified as diverse. However, because this Is a pinch point' for all CS and LPCI njection, it Is Judged prudent to consider a plant modification to allow a bypass switch (1/division) to Insert the permissive If the sensors fal to perform their function. A few other BWRs currently have this l ________

____________________ capability (e.g., Peny).

99

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 262 Modify R.B. Blowout The Reactor Building

4 - No significant No change in CDF is calculated and no Considering uncertainty and N/A Panels blowout panels are safety benefit Impact on LERF.

potential Impacts from designed to blow free from external events does not their normal positions.

Introduce any significant Hinging the Reactor Other risk measures would be affected in changes. No change to the Building blowout panels so a negligible way.

screening criteria category.

they reclose once the reactor building to environment pressure differential subsides has several advantages:

Prevents frigid external air If present from entering the reactor building Urnits reactor building accelerated circulation that could reduce radionuclide residence time In the Reactor Building May contribute to improved SCTS operation in the long term where late revolatilization of ClI could be effectively mitigated.

l 100

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase I SAMA title Result of potential Original I Revised Original DIspositIon Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 263 Supplemental Air Supply The containment vent

6 - Retain Possible Alternatives:

Still retained.

10 for the Containment Vent function Is among the last resort methods currently Ar o specified In BWRs to Air or N2 bott s oeld near the o

remove heat from OstacabermtlvaedIo containment and control the AOVs to allow AOV operation.

containment pressure under extremely adverse or cIrcumstances. The Dresden air compressors are required to support the Air supply line connections into the containment vent function.

Reactor Building from external to the The air compressors n turn reactor building to allow Air Bottles or require cooling, normally pneumatic supply trucks to supply the from TBCCW/SW. An required air pressure for AOV altemative method to su operation.

air to the vent valves for Zf v opening would be desirable If SW were to become inadequate.

101

Response 7(c):

'Provide] an assessment of the impact on the Phase 2 evaluation if risk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated wfith extemal events (as applicable).

Consider the uncertainties due to both the averted cost-risk and the cost of implementation to determine changes in the net value for these SAMAs.'

To perform this assessment, a two-step approach was taken. The first step was to reexamine the Phase II evaluation utilizing an upper bound maximum averted cost estimate of $2.OM consistent with the revised Phase I screening.

This revised screening would then result in a set of potential plant changes that could be cost beneficial when compared to the upper bound estimate of the averted cost. For these potential enhancements, a comparison was then made to a more realistic estimated averted cost to determine if the proposed change would be cost beneficial.

To provide an upper bound estimate on the risk reduction estimates to account for potential uncertainties on the risk assessment and the additional benefits associated with external events, each of the previously retained Phase II SAMAs plus the additional retained SAMAs from the revised Phase I screening in Response 7(b) have been reassessed. The reassessment assumes that the maximum averted cost risk is about

$2.OM compared to the original maximum averted cost of $457K used in the ER. Table 7-3 shows the results of this reassessment with each of the previously calculated averted costs multiplied by a factor of 5.

Additional Phase II SAMA Analyses The revised Phase I screening described in Response 7(b) resulted in two additional SAMAs being carried forward to Phase 2. Additional Phase II SAMA analyses were performed to support the revised screening provided in Table 7-3. Each of these is described below.

PHASE II SAMA NUMBER 11

==

Description:==

Align low pressure core injection or core spray to the CST on loss of suppression pool cooling.

Model Changes: Reduce HEP for aligning ECCS pump suction from base PRA model value of 0.1 to 1E-2.

Results: The results from this case indicate a reduction from the base CDF of 2.1 E-8/yr that applies primarily to loss of DHR scenarios (Class II) because the operator action is credited to support long term injection for loss of DHR events. There was no reduction in LERF (base LERF = 3.03E-7/yr). This would lead to an averted cost-risk of $3,652 utilizing the same methodology and assumptions that were utilized in the ER.

102

PHASE II SAMA NUMBER 12

==

Description:==

Enhance bypass of MSIV isolation interlock (ATWS)

Model Changes:

Reduce HEP for operator failure to bypass MSIV low RPV level interlock (ATWS) from 0.93 to 1 E-2.

In addition, increase complementary HEP for operator successful bypass of MSIV low RPV level interlock (ATWS) from 7E-2 to 0.99.

Results: The results from this case indicate a reduction from the base CDF of 2.OE-8/yr that applies only to ATWS scenarios (Class IVA and IC). Maintaining the availability of the main condenser for decay heat removal enhances the ability for successful mitigation of ATWS events. The LERF decreased from the base LERF of 3.03E-7/yr to 2.99E-7/yr.

This would lead to an averted cost-risk of $6,067 utilizing the same methodology and assumptions that were utilized in the ER.

The results of the reassessment including the two new Phase II SAMA analyses are provided in Table 7-3.

The potential costs are consistent with those provided in Response 1 1.

103

Table 7-3 Revised Phase 11 SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase II Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMA tItle enhancement Estimate Potential Cost Revised Dispositfon i

3 Enhance loss of SAMA would reduce the 5 $8,31 8

$50-1OOK for When the upper boud averted cost estimate Is component cooling potential for RCP seal failure.

- $41 590 procedural appled to two units at the sie, this procedural procedure to present 2

enhancements change may be cost beneficial. Retain for more desirability of cooling 2 Units with engineering detailed cost benefit analysis (see Table 7-4).

down reactor coolant

=$83,180 analysis system (RCS) prior to required.

seal LOCA.

2 22 Improved ability to SAMA would reduce the 5 $7,713

$50-1OOK for Not cost beneficial. Implementation of this SAMA cool the residual heat probability of a loss of decay

= $38,565 procedural would Involve procedural and hardware changes that removal heat heat removal by Implementing 2

enhancements would exceed the upper bound averted cost exchangers.

procedure and hardware 2 Unis with engineering estimate.

modifications to allow manual

= $77,130 analysis alignment of the fire protection required, plus system or by nstalling a

$100K minimum component cooling water cross-for hardware tie.

changes.

A portable diesel-driven pump is under consideration to provide cooling water to a LPCI heat exchanger. This was discussed in the EPU correspondence as the tentative plan for dealing with the seismic outlier of Dresden Island Lock &

Dam, i.e., loss of UHS, by Fall 2003.

3a 35 Develop an enhanced SAMA would provide a 5 $68,950

>$265K as The fire prtion sy (FPS c drywell spray system.

redundant source of water to

= $344750 reported in ER water to the RN s at DNPS through FP the containment to control for procedural

-drain valves, but hardware and procedures have not, containment pressure, when 2 Units enhancements been developed to use N

through the RHR system as used in conjunction with

= $689,500 with engineering an RPV Injection source or a containment spray containment heat removal.

analysis and source. Assuring the viability of such a proposed hardware change would also require extensive engineering changes analysis. However, deveopment of such capabnies required.

maybe beneficial when compared to the upper bound averted cost estmate. Retain formore detailed cost benefit analysis(see Table 7.4).

104

Table 7-3 Revised Phase 11 SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase II Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMAtie enhancement Estimate Potential Cost Revised Disposition 3b 35 Develop an enhanced SAMA would provide a 5 * $68,950

$50-1 00K for Dresden has capabilies to use LPCI cross-fie from drywall spray system.

redundant source of water to

- $344,750 procedural other unit This is currently procedurally directed for the containment to control enhancements altemate Injection to the RPV, but procedures have:

containment pressure, when 2 Units with engineering not been developed to use it as an alternate used In conjunction with

- $689500 analysis containment spray source. Retain formore detailed containment heat removal.

required.

costbenefit analysis (see Table 74). A t Ho--

4 199 Re-open MSIVs SAMA to regain the main 5 Negligible Not required.

Not cost beneficial. Minimal benefit Is obtained and condenser as a heat sink by re-Negligible associated Implementation costs would easily opening the MSIVs.

exceed the upper bound averted cost estimate.

5 239 Increase seismic SAMA would increase the Not calculated

>2.OM to bring all See Response 6(c).

ruggedness of plant availability of necessary plant SSCs to 0.3g.

components.

equipment during and after selsmic events.

Increase the seismic capacity of Extends the safe shutdown path components on the seismic capacity to at least safe shutdown paths 0.3g.

with capacities less than 0.3g to 0.3g.

l 6

249 4.d. Passive This SAMA will prevent 5 $6,369

>$100K unit Not cost beneficial. Implementation of this SAMA Overpressure Relief catastrophic failure of the

- $31,845 would Involve extensive hardware changes that containment. Controlled relief would exceed the upper bound averted cost through a selected vent path estimate.

has a greater potential for reducing the release of radioactive material than through a random break.

7 259 Diversify the explosive An alternate means of opening 5 * $24,515

>$100K / unit Not cost beneficial. Any hardware change would valve operation a pathway to the RPV for SBLC

= $122575 easily exceed the minimum hardware cost of S1OOK injection would improve the for this type of change, and therefore would exceed success probability for reactor the upper bound averted cost estimate.

shutdown.

105

Table 7-3 Revised Phase 11 SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.0M)

Phase II Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMA tIto enhancement Estimate Potential Cost Revised Disposition 8

260 Enrich Boron The increased boron 5 * $1,439 Not Required Not cost beneficial. Minimal benefit Is obtained and concentration will reduce the

- $7195 associated implementation costs would easily time required to achieve the exceed the upper bound averted cost estimate.

shutdown concentration. This will provide ncreased margin In the accident timellne for successful operator activation of SBLC.

9 261 Bypass Low Pressure LPCI and CS Injection valves 5 * $24,609

>$1 00K /unit Not cost beneficial. Any hardware change would Permissive require a permissive signal from

= $123,045 easily exceed the minimum hardware cost of $1 00K the same 2 pressure sensors In for this type of change, and therefore would exceed order to open. The Instruments the upper bound averted cost estimate.

are currently specified as diverse. However, because this Is a pinch point' for all CS and LPCI Injection, It Is judged prudent to consider a plant modification to allow a bypass switch (1/dIvision) to insert the permissive if the sensors fall to perform their function. A few other BWRs currently have this capability (e.g., Perry).

10 263 Supplemental Air The containment vent function 5^ $6,026 Lower cost Whn t uer averm Supply for the Is among the last resort

- $30,130 altemative of applied totwuniat the site, this enhancement Containment Vent methods currently specified In providing backup may be cost benefidalRetan for more detailed BWRs to remove heat from 2 Units botties or cost beneftinalysis (see Table 7-4).

containment and control

-$60,260 portable air containment pressure under compressors extremely adverse estimatedat circumstances. The Dresden

$50-1OOK for air compressors are required to procedural support the containment vent enhancements, function. The air compressors training, and In tum require cooling, normally hardware from TBCCW/SW. An modifications.

altemative method to supply air to the vent valves for opening would be desirable If SW were to become Inadequate.

106

Table 7-3 Revised Phase II SAMA Disposition (Assuming Maximum Averted Cost Risk of $2.OM)

Phase II1 Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMA title enhancement Estimate Potential Cost Revised Disposition 110) 188 Align low pressure This SAMA would help to 5 $3,65 2 4

$25-50K for When te upper bound aver cost estiate is core Injection or core ensure low pressure ECCS can

.$18,26 procedural applied to two units at the site this procedural spray to the CST on be maintained In loss of enhancements.

Xenhancement may be cost benfidcal. RetaIn for loss of suppression suppression pool cooling 2 Units more detailed estimate cost benefit analysis(see pool cooling.

scenarios.

$36,520 Table 7-4).

12")

237 Bypass MSIV solation SAMA will afford operators 5* $6,067

$50-1 OOK for Not cost beneficial. Implementaton of this SAMA In Turbine Trip ATWS more time to perform actions.

- $30,335 procedural would Involve procedural and hardware changes to scenarios The discharge of a substantial enhancements Implement a dedicated low level Interlock switch that fraction of steam to the main 2 Units with engineering would exceed the upper bound averted cost condenser (I.e., as opposed

. $60,670 analysIs estimate.

to Into the primary required, plus containment) affords the

$100K minimum operator more time to perform for hardware actions (e.g., SLC Injection, changesto lower water level, Implement depressurize RPV) than If the automatic MSIV main condenser was isolation bypass unavailable, resulting In lower capabilities.

human error probabilIties Notes to Table 7-3

(')This is a new Phase II SAMA dentifier that was not included in the ER.

(2) Detailed development of the PRA model changes made for this Phase II SAMA investigation are provided prior to the table.

107

Response 7(c) - continued:

'[Provide] an assessment of the impact on the Phase 2 evaluation ff risk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated with external events (as applicable). Consider the uncertainties due to both the averted cost-risk and the cost of implementation to determine changes in the net value for these SAMAs.'

As can be seen in Table 7-3, five of the Phase II SAMAs could be categorized as cost beneficial when compared to the upper bound averted cost estimate.

It should be noted, however, that there are many factors to consider when looking at the benefits of the SAMA candidates.

Plant specific implementation of SAMA candidates may be complicated by space limitations, outage costs, regulatory requirements, and other considerations.

These factors tend to result in underestimation of the costs.

Additionally, the specific PSA analyses that were performed in addressing specific SAMA candidates were done optimistically.

That is, the potential cost-benefit was derived from a case that maximized the CDF (and/or offsite release) reduction that would result from implementation of the SAMA. Both of these factors would, in effect, offset the uncertainties associated with the CDF estimates.

A factor of 5 is judged as a reasonable value to account for uncertainty and to account for potential contributions from external events that were not included in the averted cost estimates in the ER. Attachment A includes information about why a factor of three is more appropriate than a factor of more than 10 that would be obtained if the unmodified Fire PRA results were used directly.3 The remaining portion (from a factor of 3 up to 5) is to account for uncertainty, and the potential contributions from other external events.

Additionally, each SAMA case was re-examined to ensure that the better estimated averted cost from the internal events model was appropriately representing the potential benefit rather than representing the maximum benefit as was typically done for screening purposes. This includes a re-examination of the assumptions utilized in the initial screening analysis as well as recognizing existing model limitations that could lead to over-estimation of the averted costs. In some cases, the implementation costs were also refined to better reflect the potential cost benefit. The results of this additional screening are illustrated in Table 7-4.

Re-analysis of Phase II SAMA 3a and 3b For Phase II SAMA 3, the averted cost estimate was determined by making the drywell spray system perfectly reliable for all cases in the Level 2 analysis where it is currently considered (i.e., all accident classes except for Class II, IIID, IV, and V). In practice, though, the proposed modifications (either by establishing a means for using the fire 3 Attachment A provides an assessment of the use of quantitative risk estimates from Fire PRAs, and why it is judged that the calculated CDF values should not be directly compared at this time.

108

system or by utilizing existing LPCI cross-tie capabilities from the other unit) would not alter the release categorization in two scenarios that accounted for much of the calculated averted cost. These two scenarios are as follows:

Station blackout or loss of multiple DC bus scenarios where power would not be available to operate the drywell spray valves independent of the source of water.

Accident Class 111C scenarios with LPCI pumps available that conservatively did not credit use of the existing LPCI pumps for the drywell spray function (e.g., low pressure permissive failures that would disable the injection function, but would not disable the drywell spray function for these pumps).

A more realistic averted cost estimate can be obtained for this SAMA by excluding these cases as benefiting from the proposed modification. In that case, consistent with the ER, there is still no reduction in the CDF, but the LERF decrease goes from the base case value of 3.03E-7/yr to 2.85E-7/yr (instead of down to 2.43E-7/yr), and other release category changes occur as well. With these changes, the averted cost estimate drops from the originally calculated value of $68,950 to $7,601 using the same methodology and assumptions that were utilized in the ER.

109

Table 7-4 Refined Phase 11 SAMA Disposition of Remaining Dresden SAMA Candidates Phase 11 Phase I SAMA ID SAMA ID Result of potential Better Estimated Better Estimated number number SAMA title enhancement Averted Cost Potential Cost Better Estimate Disposition 3

Enhance loss of SAMA would reduce the 5 * $8,318

>$1OOK for Not cost beneficial. Procedural changes to reduce component cooling potential for RCP seal failure.

- $41 590 procedural RPV pressure to minimize seal leakage would be procedure to present enhancements contrary to current BWROG EOP strategies.

desirability of cooing 2Units with very Validating a recommended approach (such as down reactor coolant

$83,180 extensive depressurizing the RPV to 200 pslg) would Involve system (RCS) prior to engineering extensive analysis to determine acceptable seal LOCA.

analysis and conditions to Implement such an approach.

training required.

Consequently, any changes would require very extensive engineering analysis and justification to provide the viability and acceptability of such an approach.

Performing extensive engineering analysis, establishing a procedure, and providing training for the recommended approach would likely lead to potential costs that could easily exceed the upper bound of the estimated potential cost, or >1 OO l

This would lead to overall mplementation costs that are higher than the estimated averted cost.

3a 35 Develop an enhanced SAMA would provide a 5* $7,601

>$265K as Not cost beneficial. The fire protection system (FPS) drywell spray system.

redundant source of water to

= $8,00 reported In ER can already provide water to the RPV system at the containment to control for procedural DNPS through the RFP drain valves, but hardware containment pressure, when 2 Units enhancements and procedures have not been developed to use it used in conjunction with

= $76,010 with engineering through the RHR system as an RPV Injection source containment heat removal.

analysis and or a containment spray source. Assuring the viability hardware of such a proposed change would also require changes extensive engineering analysis. Overall required.

Implementation costs Including hardware modifications would exceed the estimated averted cost 111

Table 7-4 Refined Phase 11 SAMA Disposition of Remaining Dresden SAMA Candidates Phase 11 Phase I SAMA ID SAMA ID Result of potential Better Estimated Better Estimated number number SAMA title enhancement Averted Cost Potential Cost Better EstImate Disposition 3b 35 Develop an enhanced SAMA would provide a 5 * $7,601(1)

$50- OOK for Not cost beneficial. Dresden has capabilities to use drywell spray systemn.

redundant source of water to

/ 2 with less procedural LPCI cross-tie from other unit. This Is currently the containment to control conervative enhancements procedurally directed for altemate Injection to the containment pressure, when treatment of with engineering RPV. but procedures have not been developed to used In conjunction with global failure of analysis use It as an alternate containment spray source.

containment heat removal, the suppression required.

A detailed review of the cutsets that contribute to the pool suction averted cost Indicates that the currently calculated strainers benefit Is totally dependent on the assigned value for

- $19,003 common cause failure of the suppression pool

n~ts suction strainer failures which Is currently assigned a 2 Units 1.01E24 value for LOCA scenarios based on

= $38,006 engineering judgment. This is believed to be conservative since the strainers have been enhanced and replaced at Dresden similar to changes made at other BWRs, and since new requirements exist for control of fibrous materials inside containment and water cleanliness.

Given these considerations, It Is estimated that the averted cost estimate Is high by at least a factor of two for these scenarios due to the conservatisms and uncertainty associated with the very unlikely global common cause failure value of all of the suppression pool suction strainers. The revised best estimate averted cost Includes this reduction factor.

Consequently, this would lead to potential costs that are higher than even the lower bound value of the estimated averted cost.

112

Table 7-4 Refined Phase 11 SAMA Disposition of Remaining Dresden SAMA Candidates Phase I Phase I SAMA ID SAMA ID Result of potential I Better Estimated Batter Estimated number number SAMA tiMe enhancement Averted Cost Potential Cost Better Estimate Disposition 10 263 Supplemental Air The containment vent 5* $6,026 Lower cost Not cost beneficial. Very minimal credit is currently Supply for the function is among the last

/3 with Ies altemative of taken for recovery of instrument air In the Dresden Containment Vent resort methods currently conservative providing backup model. The SAMA analysis changed the current specified In BWRs to remove credit for existing bottles or value of 0.9 to 0.0 to estimate the averted cost heat from containment and capablities portabie air benefit. For comparison, the Quad Cities model control containment pressure compressors currently uses a recovery value of 0.148 for recovery under extremely adverse

= $10,043 estimated at of Instrument air In support of verting.

circumstances. The Dresden

2 Units

$50- 00K for Given these considerations, It Is estimated that the air compressors are required

= $20,086 procedural averted cost estimate is high by at least a factor of to support the containment enhancements, three for these scenarios compared to the vent function. The air training, and capabilities that already exist and could be more compressors in turn require hardware realistically credited. The revised best estimate cooling, nomnally from modifications.

averted cost Includes this reduction factor.

TBCCW/SW. An alternative method to supply air to the Consequently, this would lead to potential costs that vent valves for opening would are higher than even the lower bound value of the be desirable if SW were to estimated averted cost become inadequate.

l 11 188 Align low pressure This SAMA would help to 5* $3,652

$50K for Not cost beneflcial. Current procedures exist to core injection or core ensure low pressure ECCS

= $18 260 procedural perform such actions at Dresden. The relatively high spray to the CST on can be maintained In loss of enhancements.

HEP value of 0.1 Is largely based on uncertainty loss of suppression suppression pool cooling

2 Units associated with environmental conditions that may pool cooling.

scenarios.

. $36,520 exist when performing the actions in the reactor building. Improvements to existing procedures would not Justily a significant reduction In the HEP value.

Larger benefit could only by significant restructuring of the procedures and EOPs to make this action always viable before environmental conditions put its performance in doubt. This would require procedural enhancements at the upper end of the estimated potential cost, or $S50K. This would lead to overall Implementation costs that are higher than the estimated averted cost Revised from original analysis to reflect a better estimated averted cost based on a re-analysis of the scenarios that could actually benefit from the proposed modifications.

113

RAIB For certain SAMAs considered in the ER, there may be lower cost alternatives that could achieve much of the risk reduction, such as adding a diesel-driven battery charger. Confirm that low cost alternatives to Phase 2 SAMAs were considered, and provide a brief discussion of these alternatives.

Response 8:

Lower cost alternatives were considered in both the initial Phase I screening all the way through to the final revised Phase II screening. Examples included a portable generator to provide prolonged battery capacity (see Table 7-2, Phase I SAMA 167), and backup bottles or portable compressors for supplementing instrument air capabilities (see Table 7-3, Phase II SAMA 10). Other lower cost alternatives were also explored in the form of potential procedural changes (see Table 7-3, Phase II SAMAs 1, 3b, 4, and 11). While many of these may only involve procedural changes in concept, a more thorough investigation leads to the finding that more costs would actually be incurred when considering that the procedure changes may also require engineering analysis, experimentation, and extensive training (see also Response 11). Additionally, a more refined evaluation of the initial averted cost estimates indicate, that in most of the cases, analysis simplifications or existing model limitations tend towards an overestimation of the averted cost. The identified modeling limitations are not considered significant when considering the typical uses of the PRA models, but come to the forefront when specific risk reduction values are calculated. As such, none of the remaining SAMAs (including lower cost alternatives) were determined to be cost beneficial.

114

RAI9 During the review of the EPU application, the staff noted several areas where the PSA should be modified to reflect modifications to the plant or changes in success paths.

These include: a plant modification to install a recirculating pump run back control circuit; a plant modification to trip the condensate/booster pump D in the event of a LOCA to prevent an overload condition from occuring; a change in success criteria for reactor pressure vessel (R. V.) depressurization in a transient without a stuck open relief valve (two valves under EPU conditions); a change in success criteria for R. V. over pressure protection in ATWS sequences (12 of 13 valves under EPU conditions).

Confirm whether these model changes, as well as others, have been incorporated in the PSA used for the SAMA analysis. For those not incorporated, provide an assessment of the impact that the model change would have on the SAMA analysis.

Response 9:

The model was revised to include all appropriate EPU changes:

The purpose of the recirc. pump runback control circuit is to prevent the reactor trip frequency from increasing due to EPU. The recirc. pump runback is needed because there no longer are sparew condensate pumps or feedwater pumps. Due to this modification, the transient initiating event frequency is not expected to change. However, effects on the plant can only be incorporated in the PRA after some plant experience via the next periodic update of initiating event frequencies.

The potential risk impact of the recirc. runback modification was addressed in a response to a NRC RAI to support the EPU application

[Reference 9-1]. The response to the RAI addressed both 1) the failure of the recirc. runback to operate as designed, and 2) spurious recirc.

runback. The RAI judged that the incorporation of the recirc. runback modification would result in a negligible risk increase.

The circuit to trip condensate/condensate booster pump D" on a LOCA signal is expected to be very reliable. The risk impact of the condensate/condensate booster pump D" trip logic was also addressed in Reference 9-1. The risk impact was calculated to be 1.7E-10/yr. Due to the minor contribution to CDF, this failure mode was not explicitly included in the PRA model.

The success criterion for RPV depressurization is reflected in the revised transient without SORV model.

The success criterion for ATWS overpressure protection is reflected in the revised ATVIS model.

The higher decay heat load due to power uprate reduces the time available for certain operator actions. This has been reflected in revised HEP's for those actions.

115

REFERENCE

[9-1] Letterfrom K.A. Ainger, Exelon Generation Company, to U.S. NRC, uAdditional Risk Information Supporting the License Amendment Request to Permit Uprated Power Operation at Dresden Nuclear Power Station and Quad Cities Nuclear Power Station", RS-01 -168, August 14, 2001.

116

RAI 10

During the review of the EPU application, the staff noted that a new means of inducing a LOOP initiating event potentially exists under EPU conditions. The end result could be an overbusy condition on the unit auxiliary or reserve auxiliary transformer. Given this new condition, provide an evaluation of the costs and benefits associated with the replacement of the affected transformer with a higher capacity transformer.

Response 10:

The risk impact of the induced LOOP initiating event was addressed in a response to a NRC RAI to support the EPU application [Reference 10-1].

Information from the response to the RAI is summarized below.

BACKGROUND During normal operation the station loads are distributed between the Unit Auxiliary Transformer (UAT) and the Reserve Auxiliary Transformer (RAT). Normally, the loads for two non-essential 4kV buses are aligned to the UAT and the loads for the other two non-essential 4kV buses are aligned to the RAT. If either the UAT or RAT become unavailable during normal operation without a reactor scram, the increased loads for the EPU configuration may result in an overload condition for the remaining transformer's bus duct connection to the 4kV buses.

The scenario of concern is a loss of the UAT or RAT due to transformer failure, failure of protective relaying (e.g., false fast transfer signal), or spurious opening of multiple circuit breakers [see note (1)], causing a fast transfer of all running loads to the other transformer.

Under these conditions, certain bus duct segments are overloaded, requiring operator action within one hour to reduce load to within the bus duct rating.

This action will be procedurally directed. The one hour time frame for load reduction was determined based on an Exelon Generation Company (EGC), LLC evaluation of a General Electric Company study on short term overload conditions for the bus ducts.

The simplifying assumption is made that failure to take this action would lead to a loss of offsite power (LOOP). In reality, overload of the bus duct results in heating above the allowable temperature limits if ambient temperature is at the design value.

No deterministic evaluation has been conducted to determine if overheating will result in complete failure of the bus duct, thereby causing a LOOP.

(1) Spurious opening of an individual circuit breaker to an individual 4kV bus would cause a fast transfer of the individual 4kV bus loads to the alternate transformer. However, based on the estimated EPU loads, the transfer of loads for a single 4kV bus (i.e., loads from three 4kV buses on a single transformer) would not place the transformer bus ducts in an overload condition.

117

RESULTS The induced LOOP initiating event is calculated to result in a 6E-9/yr increase in the Dresden Level 1 CDF. The risk evaluation accounts for the estimated frequency of the transformer overduty condition and failure of the plant or operating staff to mitigate the event.

CONCLUSIONS FOR SAMA Based on the minor risk impact, the costs associated with the replacement of the affected transformer or associated electrical equipment (e.g., 4kV bus duct connections) is judged not to be warranted.

Additional details of the risk calculation can be found in Reference [10-1].

REFERENCE

[10-1] Letter from T. W. Simpkin (Exelon Generation Company) to U. S. NRC, Additional Information Supporting the License Amendment Request to Permit Uprated Power Operation, Dresden Nuclear Power Station and Quad Cities Nuclear Power Station," RS-01 -200, dated September 19, 2000.

118

RAI 11

In Section 4.20.5 of the ER, Exelon states that a preliminary cost estimate was prepared for each of the remaining candidates (remaining after the initial screening). However, implementation costs were provided for only one of the Phase 2 SAMAs. Provide the estimated implementation costs (preliminary cost estimates) for the Phase 2 SAMAs, so that the staff can readily determine whether any of these SAMAs are potentially cost-beneficial when considering the impact of extemal events and uncertainties. In addition, indicate the minimal cost assumptions used for procedure and hardware changes.

Response 11:

For all of the Phase 2 SAMAs evaluated in Section 4.20.5 of the ER, only one of them had a benefit that was close to the potential implementation cost. Therefore, only one estimated cost was supplied (i.e., >$265K for overall implementation of allowing FPS to act as an aternate drywell spray system).

As a supplement to the original SAMA evaluation, Exelon has developed the following estimated implementation costs for use in Response 7(c).

These costs have been estimated based on existing SAMA evaluations and have addressed the following cost elements:

Procedural changes Engineering evaluations Hardware modifications Testing to support engineering evaluations and/or training to support procedural modifications The following references have been used to assign an appropriate cost to these elements.

REFERENCES

[11-1] NUREG-1437, Generic Environmental Impact Statement for License Renewal of Nuclear Plants, Oconee Nuclear Station", Supplement 2, U.S. Nuclear Regulatory Commission, Washington, D.C., December 1999.

[11-2] Peach Bottom SAMA Evaluation and RAI Responses

[11-3] HB Robinson SAMA Evaluation and RAI Response

[11-4] VC Summer SAMA Evaluation and RAI Response

[11-5] GE Nuclear Energy, Technical Support Document for the ABWR,m 25A5680, Rev. 1, November 1994.

119

PROCEDURAL CHANGES Procedure development and modification requires preparation by a System Engineer, technical review and validation, oversight review, and a variety of additional plant reviews prior to release.

In addition, plant staff will need to be trained prior to implementation. A few examples of other procedure change estimates are provided below.

ABWR [11-5] indicates that improvements to existing maintenance procedures would cost approximately $300K.

PB [11-2] describes a procedural modification to allow for cross-tie of CCW at an estimated implementation cost of $50K.

For the Dresden SAMA analyses, a range for procedural changes is estimated to cost from $25K to $50K. The lower estimate is judged to be more appropriate for changes to existing procedures, and the upper estimate is judged to be more appropriate for the development of new procedures.

ENGINEERING EVALUATIONS In support of procedural and hardware modifications, an engineering evaluation will be required.

For a procedural modification, the engineering requirements could easily double the cost of the change. This would increase the procedural change cost to an estimated range of $50K to $IOOK.

HARDWARE MODIFICATIONS The following provides examples from previous SAMA evaluations.

PB [11-2] evaluated alternate methods to provide cooling to the RHR pumps at an estimated implementation cost of $250K.

PB [11-2] also estimated a cost of $1600K to replace all 8 station batteries.

Numerous hardware changes were evaluated for the ABWR [11-5] at a cost range from $1000K to $6000K.

Hardware modifications were evaluated for Oconee [11-1] including automatic refill systems for the refueling water storage tank, automatic switchover of HPI to the spent fuel pool, and others ranging from $1000K to $5000K.

120

For the Dresden SAMA analysis, several hardware modifications have been evaluated and range in cost from $100K to over $1000K. A minimum of $100K is used to account for engineering analysis, purchase, and maintenance of any proposed hardware modification.

TESTING/TRAINING Similar to engineering costs to support a procedural change, testing of a plant system to establish operating limits or extensive training requirements to implement the procedure modification is estimated to double the cost of the procedural change. An example of this would be for a proposed SAMA to justify the operation of RCIC at low RPV pressures (such as was explored for Quad Cities), or to implement a containment venting strategy within prescribed limits. Procedural changes in addition to potential testing/training costs could increase the overall implementation cost to a range of $100K to $200K.

SUMMARY

OF IMPLEMENTATION COST Based on a review of previous SAMA evaluations and an evaluation of expected implementation costs at Dresden, Table 11-1 provides the estimated costs for each potential element of the proposed SAMA implementation. Depending on the individual elements involved with each proposed SAMA, these estimates are then used to determine the total implementation cost with the remaining Phase II SAMAs as described in Response 7(c).

Table 1 -1 Estimated Implementation Costs Type of Change Estimated Cost Range Procedural only

$25K-$50K Procedural change with engineering required

$50K-$100K Procedural change with engineering and testing/training $100K-$200K required Hardware modification

$100K to > $1000K 121

RAI 12

For Phase 2 SAMAs 3, 6, 7, and 10, hardware modifications, as well as procedural changes, are necessary. However, the hardware modifications are not fully described.

Briefly describe the proposed hardware modifications.

Response 12:

The following briefly describes the hardware modifications required to implement Phase 2 SAMAs 3, 6, 7, and 10.

Phase 2 SAMA #3: This SAMA addresses the use of the Fire Protection System as a source of water for Drywell Sprays. This modification would require the addition of a spool piece and piping to allow for a connection between FPS and the RHR system. As described in Section 4.20.6.3 of the ER, this capability would also require procedural changes along with engineering analysis to show the capability of FPS to remove heat from the Drywell atmosphere in this new mode of operation.

Phase 2 SAMA #6: Implementation of this SAMA would require installation of a rupture disk in the existing containment vent path or the addition of a completely new vent pathway. If the existing vent piping was to be used, then the valves currently installed in that line would have to be locked open or removed to allow for proper functioning of the rupture disk. If the existing valves were to remain in the vent path, then logic would have to be added to allow for opening of these valves at the proper time to allow for the rupture disk to function.

Phase 2 SAMA #7: Implementation of this SAMA would require either replacement of the existing valves to allow for a more reliable method of opening the path for SBLC, or to install a new bypass pathway using explosive valves to provide SBLC injection to the RPV.

Phase 2 SAMA #10: This SAMA would require the use of portable air bottles with the installation of dedicated tie-in points for quick connection in the event of loss of normal instrument air. The capability could also be achieved using a portable compressor with the same dedicated tie-in points.

Phase 2 SAMA #12: This is a new Phase 2 SAMA identified in the response to RAI 7c and would involve an enhancement to the capability for the operator to bypass the MSIV isolation interlock for an ATWS.

One possible hardware modification to provide this benefit would be the installation of a dedicated low level interlock bypass switch.

122

ATTACHMENT A FIRE PRA AND USE OF QUANTITATIVE RISK ESTIMATES Overview The following summarizes the fire PRA topics where quantification of the associated figure of merit, CDF, may introduce different levels of modeling uncertainty than the internal events PRA.

The uncertainties generally reflect the following:

lack of adequate data for initiating events lack of realistic fire modeling capabilities including mitigation lack of ability to track all cables (e.g., BOP cables) uncertainty in crew response, especially for control room fires, and their modeling limited peer reviews that examine the need for realism instead of conservatism In many cases, analysts choose to address these uncertainties by incorporating margin into the analysis (i.e., conservative assumptions).

Elements of Fire PRA Fire PRAs are useful tools to identify design or procedural items that could be clear areas of focus for improving the safety of the plant. Fire PRAs use a structure and quantification technique similar to that used in the internal events PRA.

Since less attention historically has been paid to fire PRAs, conservative modeling is common in a number of areas of the fire analysis to provide a bounding, methodology for fires. This concept is contrary to the base internal events PRA which has had more analytical development and is judged to be closer to a realistic assessment (i.e., not conservative) of the plant.

There are a number of fire PRA topics involving technical inputs, data, and modeling that prevent the effective comparison of the calculated core damage frequency figure of merit between the internal events PRA and the fire PRA. These areas are identified as follows:

Initiating Events:

The frequency of fires and their severity are generally conservatively overestimated. A revised NRC fire events database indicates the trend toward lower frequency and less severe fires.

This trend reflects the improved housekeeping, reduction in transient fire hazards, and other improved fire protection steps at utilities.

123

System Response:

Fire Modeling:

Fire protection measures such a sprinklers, C02, fire brigades may be given minimal (conservative) credit in their ability to limit the spread of a fire.

Cable routings are typically characterized conservatively because of the lack of data regarding the routing of cables or the lack of the analytic modeling to represent the different routings. This leads to limited credit for balance of plant systems that are extremely important in CDF mitigation.

Fire damage and fire spread are conservatively characterized. Fire modeling presents bounding approaches regarding the fire immediate effects (e.g., all cables in a tray are always failed for a cable tray fire) and fire propagation.

There is little industry experience with crew actions under conditions of the types of fires modeled in fire PRAs. This has led to conservative characterization of crew actions in fire PRAs.

Because the CDF is strongly correlated with crew actions, this conservatism has a profound influence on the calculated fire PRA results.

HRA:

Level of Detail:

Quality of Model:

The fire PRAs may have reduced level of detail in the mitigation of the initiating event and consequential system damage.

The peer review process for fire PRAs is less well developed than for internal events PRAs. For example, no industry standard, such as NEI 00-02, exists for the structured peer review of a fire PRA.

This may lead to less assurance of the realism of the model.

Summary and Conclusions The fire PRA may be subject to more modeling uncertainty than the internal events PRA evaluations. While the fire PRA is generally self-consistent within its calculational framework, the fire PRA does not compare well with internal events PRAs because of the number of conservatisms that have been included in the fire PRA process.

Therefore, the use of the fire PRA figure of merit as a reflection of CDF may be inappropriate. Any use of fire PRA results and insights should consider areas where the

'state of the art" in fire PRAs is less evolved than other PRA topics.

Relative modeling uncertainty is expected to narrow substantially in the future as more experience is gained in the development and implementation of methods and techniques for modeling fire accident progression and the underlying data.

Until that time, however, the following assessment is made to provide a methodology for estimating the conservatisms included in the reported Fire PRA CDF numbers for Dresden when compared to the internal events CDF numbers.

124

Initiating Events A review of a recent NRC report [Reference A-1] was made to obtain an estimate of potential reductions in the fire initiating event frequencies that may occur if more recent and less conservative data were utilized in the Dresden analysis. Note that the NRC report only presents the data in the form of fire frequency by major plant location (it does not provide a breakdown by component such as that which was utilized for the Dresden analysis). As such, a direct comparison is not possible, but if all of the areas listed for each plant location are added up for Dresden and placed into one of the categories provided in the NRC report, then an approximate comparison can be made.

Table A-1 provides the comparison, and as can be seen, in all areas, the NRC reported frequency per area is lower than that which was utilized in the Dresden analysis.

Table A-I Comparison of Recent NRC Report Fire Initiating Event Frequencies with Dresden IPEEE Values Location NRC [A-1]

Dresden Ratio (Dresden NRC)

Reactor Building 2.8E-2 1.OE-1 / (2 Units) = 5.0E-2 1.8 Turbine Building 4.1 E-2 3.6E-1 / (2 Units) = 1.8E-1 4.4 Control Room 7.2E-3 2.4E-2 3.3 Cable Spreading Room 8.4E-4 2.7E-3 3.2 Switchgear Rooms 5.1 E-3 7.2E-2 / (2 Units) = -3.6E-2 7.1 EDG Building 1.4E-2

- 3.OE-2 per room 2.1 SWS Pumphouse 7.2E-3 2.9E-2 4.0 Battery Room 8.4E-4

- 3.5E-3 per room 4.2 Other NIA 0.12 N/A Therefore, based on the comparison provided in Table A-1, it is judged that a factor of two reduction on the Initiating Event / System Response portion of the Fire CDF can be made as a reasonable assumption to make to provide a more accurate comparison to the internal events CDF.

System Response/ Fire Modeling The Dresden Fire modeling typically utilized bounding approaches regarding the fire immediate effects (e.g., all cables in a tray are always failed for a cable tray fire, and all failed cables lead to failure states of the associated equipment).

In the analysis, 125

severity factors were utilized in some cases to distinguish between large versus small fires, and therefore the consequences associated with each. However, the complement of the severity factor was also maintained in the Dresden analysis such that the total frequency was always accounted.

The NRC data would support lower initial fire frequencies and lower severity factors in an updated analysis that would lead to lower frequencies associated with many of the dominant fire scenarios.

While no direct comparison can be made to approximate the effects this has on the Fire CDF, it is estimated that this modeling approach can also be characterized by at least a factor of two reduction in the Fire CDF to provide a more accurate comparison to the internal events CDF.

HRA ILevel of Detail An examination of the dominant fire scenarios for Dresden from the IPEEE indicates that approximately 26% (Unit 2) and 44% (Unit 3) of the reported CDF (excluding Control Room fires) is due to Loss of Containment Heat Removal scenarios. These scenarios are conservative in nature since they involve many hours to evolve (i.e., >24 hours) at which time many ad hoc procedures could be written or previously failed systems could be recovered. In the Dresden fire analysis, system recovery was not credited at all for these scenarios.

Other PRA models have also credited recovery of failed systems (e.g., RHR pumps or Instrument Air) in support of scenarios such as the dominant loss of containment heat removal scenarios. Such recoveries were also excluded from the reported Dresden Fire CDF since the fire damage could preclude such recovery actions.

However, such recovery actions are not precluded per se from other (i.e., non fire-related) failures that exist in the cutsets in leading to core damage. Typical recovery values for these types of scenarios range from 0.1 to 0.4.

Other dominant scenarios in the Dresden fire model included operator action failures that are based solely on the direction provided in the EOPs and Off-normal procedures that are credited in the internal events model.

Additionally, the Safe Shutdown Procedures that exist for potential fires in all fire areas were not credited at all in the Dresden fire analysis. Credit for these procedures also has the potential for reducing the HEP values utilized in the Fire analysis since they may provide more timely cues or actions to consider given a fire in a specific area compared to the cues that would arise from the symptom-based EOPs.

Considering all of these effects together, it is judged that the simplified HRA modeling and lack of sufficient level of detail in the model can easily lead to an additional factor of 1.5 reduction in the in the Fire CDF to provide a more accurate comparison to the internal events CDF. This can be supported by noting that a 0.2 recovery factor on the Loss of Containment Heat Removal cases alone would lead to about a factor of 1.5 reduction in the total Fire CDF for Dresden Unit 3.

126

Combined Impact for Comparison to the Intemal Events CDF The CDF contribution to internal fires was estimated at 1.7E-5/yr for Unit 2 and 3.1 E-5/yr for Unit 3 in the Dresden IPEEE submittal. Using the Unit 3 value as a bounding case, and the reduction factors provided above, the following assessment is made.

Reported Fire CDF:

3.1 E-5/ yr Reduction from Conservatisms in the Initiating Event frequencies and System Response (2):

3.1 E-5/yr/ 2 = 1.55E-5/yr Reduction from Conservatisms in Fire Modeling (2):

1.55E-5/yr / 2 = 7.75E-6/yr Reduction from HRA Simplifications and Lack of Detail in the Scenario Modeling (1.5):

7.75E-6/yr/ 1.5 = 5.17E-6/yr Considering all of the conservatisms in the reported Fire CDF indicates that if the fire results were reported in a more realistic fashion for Dresden, then the actual result would be no more than a factor of 3 (i.e., 5.2E-6/yr / 1.9E-6/yr = 2.7, or approximately 3) higher than the internal events CDF. This conclusion is supported by the discussion above.

REFERENCES

[A-1] U.S.

Nuclear Regulatory Commission (Division of Risk Analysis and Applications), Fire Events - Update of U.S. Operating Experience, 1986-1999; Commercial Power Reactors", RES/OERAB/S02-01, January 2002.

127

ML032030227

Contents

Text

Subject:

RAI 1

Response

Response

RAI 10

RAI 11

SUMMARY

RAI 12

Navigation menu

ML032030227

Text

Subject:

RAI 1

Response

Response

RAI 10

RAI 11

SUMMARY

RAI 12

Navigation menu

Search