ML20085G271
| ML20085G271 | |
| Person / Time | |
|---|---|
| Site: | 05200002, 05200004 |
| Issue date: | 06/14/1995 |
| From: | Quinn J GENERAL ELECTRIC CO. |
| To: | Quay T NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM), Office of Nuclear Reactor Regulation |
| References | |
| MFN-085-95, MFN-85-95, NUDOCS 9506200146 | |
| Download: ML20085G271 (11) | |
Text
'
+
O GENuclearEnergy Ym 75 C$
$165 San Jose, CA 95125-1014 LEfR B
Aven 408 925-1005 (phone). 408 925-3991 (facsimile)
June 14,1995 MFN 085-95 1
Docket STN 52-004 Document Control Desk U. S. Nuclear Regulatory Commission Washington DC 20555 Attention: Theodore E. Quay, Director Standardization Project Directorate
Subject:
SBWR-RAIs 901.75 - 901.160 on GE LTR NEDE-32177P
Reference:
- 1. Letter, J. H. Wilson (NRC) to J. E. Quinn GE), Request for Additional.
Information (RAI) Regarding The Simplified Boiling Water Reactor (SBWR)
Design (Q901.75-Q901.160).
2.GE Licensing Topical Report (LTR) NED-32177, TRACG Computer Code Qualification, dated January 1993.
3.GE letter MFN No. 071-95, J. E. Quinn (GE) to D. M. Crutchfield (NRC),
Revised Schedule For Review of the SBWR Technology Program, dated May 11, 1995.
I 1
We have received the Reference 1 Requests for Additional Information (RAIs) on GE LTR NED-32177 (Reference 2) submitted to the NRC for review in February,1993. As indicated in the attachment to this letter, we propose to respond to the RAIs in three groups. The first group of 43 RAIs (half of the total) are clear and will be responded to by the end of July, 1995, as requested (see attachment).
I The second group of 29 RAIs are relatively clear but quite detailed and more resource intensive.
As indicated in Reference 3, we understood from the long review and our interactions to date that the review of NED-32177 was nearly complete except for information from the testing program, and accordingly have released some of the relevant TRACG people from the Project.
These RAIs will require arrangements be made to secure additional resources with which to respond; we expect to be able to respond to these RAIs before the end of 1995.
We believe the remaining 14 RAIs need more discussion to assure that we clearly understand the needs of the NRC and their subcontractors beforejvp undertake this significant effort and commit to a schedule. We wish to do all that 1gjecchsg;rf-to support the review of this key document and to make the next issue of this document scheduled for April 1996, a complete document which will expedite your review process. To clarify what is being requested we propose a telephone or video-conference meeting during the week of June 19th,1995. We will contact you separately to arrange this telephone / video-conference meeting.
9506200146 950614
~
PDR ADOCK 05200002 A
, i 1
1
- s If you have questions with regard to these RAls please contact Bharat Shiralkar of our staff on 408-925-6889.
l I
Sincerely l'
2 inn, Projects Manager L
nd SBWR Programs
Attachment:
RAIs on Qualification LTR (901.75 - 901.160).
I cc:
Document Control '(NRC)
Original paper copy P. A. Boehnert (NRC/ACRS) (2 paper copies w/att. plus E-Mail w/att.)
I. Catton (ACRS)
(1 paper copy w/att. plus E-Mail w/att.)
S. Q. Ninh (NRC)
(2 paper copies w/att. plus E-Mail w/att.)
T.E. Quay (NRC)
(E-Mail w/att.)
J. N. Wilson (NRC)
(E-Mail w/att.)
J.- H. Wilson (NRC)
(1 paper copy w/att. plus E-Mail w/att.)
l I
i
RAls on Qualification LTR (901.75 - 901.160)
Commit RAINo.
Request for information Date The word " Qualification" is used throughout the report without ever being defined.
901.75
" Validation" and "Verifi:ation" are well defined words. It would appear that " Qualification" 7/24/95 is used in place of" Validation." If this is so, it should be explicitly sta i
Each section of this report should demonstrate by itself, that TRACG has been validated 901.76 for the specific model(s) considered in that section. This requires a sufficiently detailed 7/24/95 description of the experimental data base and the data range, to demonstra j
The SBWR supplement to this report should consider the GE PIRT results in determining, which items must be validated. This systematic approach should result in a comprehensive list of models to be validated and should include a complete matrix of the val Were all model computations in the Qualification Report done with the same version of 901.78 the code? State the code versions used. Describe how GE proposes to demonstrate 7/24/95 that the validations performed in preceding years with earlier versions of the code are in most sections of the Qualification Report, the nodalization of the relevant TRACG 901.79 model is described, but generally without justification. For a meaningful code validation the nodalization should be justified, generally based on an evaluation of speci in many cases " good" or " satisfactory" agreement between simulation and experiment is 901.80 claimed. Such qualitative statements are not adequate. A meaningful quantitative comparison is required. This should include the following three steps: 1. an estima (Section 3.1) The introduction to Section 3 states that the test data used for qualification 901.81 of TRACG for the prediction of void fractions cover a wide range of flow conditions, 7/24/95 pressures, flow rates and inlet subcooling " assuring that all major flow reg (Section 3.1) Describe, in detail, the method used for void fraction measurements in 901.82 12/31/95 Sections 3.1.2,3, and 4, as it can affect the value or weight given to a set of data.
(Section 3.1) No low pressure data are included in Section 3.1. During and after GDCS 901.83 refill, the reactor vessel will be at pressures significantly below the pressure of all the test data provided. Since the vapor-to-liquid density ratio at 2 to 3 bar (Sections 3.1.2 to 4) Provide additional details concerning the test apparatus and, in 901.84 particular, about the void fraction measurement method for the Christensen, Wilson and 12/31/95 Bartolomei, as well as the EBWR data. The information given for each of these 1
RAls on Qualification LTR (901.75 - 901.160) i Commit
{
RAINo.
Request for Information Date (Section 3.1.5) The PSTF Level Swell Tests of Section 3.1.5, refer to the TRACG two-901.85 phase level model(Section 3.2.7 of Model Report). The definition of a two-phase level 7/24/95 requires some arbitrary / reasonable definitions. Here, a two-phase level is assume (Section 3.1.5) Provide a description of how a two-phase levelis measured and what it 901.86 7/24/95 is compared to.
1 (Section 3.2) This section only considers film boiling and core spray heat transfer. In l
901.87 the Model Report, Sections 3.2.9 and 10 consider interfacial and wall heat transfer with
)
many more heat transfer regimes. For interfacial heat transfer, the Model R 1
5 Section 3.2.1) The film boiling evaluations of Section 3.2.1 state that THTF pressures j
901.88 are significantly higher than those of BWR transients. What are the THTF pressures and 7/24/95 l
how much are the correlations affected by the differences between THTF and SBWR i
2 (Section 3.2.1) These high pressure data cannot be applied for code validation of film l
901.89 boiling after depressurization (i.e. in the range of about 2 bar), as would be required if the core were ever uncovered. Provide a justification for applying high pr i
(Section 3.2.1) What is the sensitivity of the film boiling heat transfer to the given rod l
i 901.90 temperature data (i.e., what error in film heat transfer rates would induce what error in 7/24/95 rod temperatures), which clearly lag in response due to thermal capacity 1
(Section 3.3) The comparison of TRACG code predictions against test data from the 901.91 CSHT facility is specific to ECCS liquid coolant entering the fuel bundles, which is not of j
interest in SBWRs. However, CCFL can occur under accident conditions in SBWRs, j
4 j
(Section 3.3) The descriptions of the tests and the model are inadequate. In partichiar, i
901.92 a detailed description of Figures 3.3-3 and 3.3-4 is required to permit a reader to fcllow 7/24/95
)
the results. Provide additional explanation and description of all of th (Section 3.4.1.3) Evaluate the effect of using heat slabs over the time period of 50 to 60 I
901.93
- s. Was this effect quantitatively evaluated, or is the last sentence of this section i
conjecture?
i (Section 3.5) This section assesses TRACG's capability to predict tube bundle pressure 901.94 drops in the range of operating conditions (mass flow range of test data - 140 to 2,040 l
kg/m2s; average full power mass flux ~1,020 kg/m2s). However, it does not cons 2
4 a
4 2
i
RAls on Qualification LTR (901.75 - 901.160) i Commit RAINo.
Request for Information Date j
(Section 3.6.1) This section refers to "five rod groups" and to the rod that first showed 901.95 BT being treated as a separate group. This description is unclear. Is a sub-channel 7/24/95 analysis being applied here, with a separation into five groups? Is the nodal (Section 3.6.1) Describe the relative locatien of thermocouples 2 through 6 in Figure 3.6-901.96
- 1. They are apparently allin the upper section of the bundle, but the response of Tc 4 is 12/31/95 flat, while Tc 2 shows the largest oscillation amplitude. Provide an ex (Section 3.6.1) Fig 3.6-2 is presented as TRACG results for comparison with the 901.97 experimental results of Figure 3.6-1. However, this figure does not really present a validation of the test data, except to show that the CPR frequency is identical to the f j
(Section 3.6.2) How was the pressure increased in these tests? Figure 3.6-3 implies 901.98 that this was achieved through an increase in inlet flow, if so, why was a prescribed 7/24/95 l
pressure vs. time imposed on the simulation, rather than an inlet flow vs. time pr l
(Section 3.6.2) The report stated that the TRACG model used here is the "same" as for 901.99 the oscillation tests. However, a different tube bundle was used here (GE9 vs. GE11).
7/24/95 l
Provide clarification of this discrepancy.
l j
(Section 3.6.2) A plot of CPR with a scale truncated at 1.0 does not provide a validation 901 100 for the comparison of the experiment and the TRACG simulation. Provide justification why parameters equivalent to those of Figure 3.6-3 were not used.
l l
901.101 (Section 3.7) Identify the FRIGG test that was selected.
7/24/95 1
(Section 3.7) FRIGG experiments that used a pseudo random binary signal are l
901.102 7/24/95 l
available. Provide justification for not using one of those experiments for validation.
l (Section 4.2) Provide an explanation for why carryover and carry-under are only l
901.103 reported as functions of inlet quality. Mass flow rates or a related parameter, such as 7/24/95 l
stagnation pressure, should be considered as a separate, independent parameter, l
(Section 4.2) The validation presented is in the range of normal operating data only.
901.104 Consideration of LOCA conditions is missing. In particular during an MSLB scenario a.1d 7/24/95 during ADS operation, carryover could significantly er6ance the amount of liqui l
l l
l i
l l RAls on Qualificrtion LTR (901.75 - 901.160) l l
Commit
- RAINo, Request for Information Date i
(Sectior 4.4.3) One purpose of the steady state heat transfer test was to evaluate "the 901.105 effect of ea:ying nitrogen concentration and steam flow rates on the PCC heat transfer characteristics". The comparison given does not adequately describe the variati (Section 4.4.5) The methodology describing how this degradation parameter was 901.106 established in the experiments is unclear. The text states that it was " calculated by 7/24/95 measuring the condensate flow rate", which is in itself contradictory. To experimentally l
l (Section 5.1.1) The " mixture level" mentioned here is a two-phase level and r.ot a 901.107 collapsed level. Its progression downward during the test run is estimated basad on the 12/31/95 differential pressure measurements, using the taps indicated in Figure 5.1-2. To o (Section 5.1.1) For the TRACG model, Figure 5.1-13 of Section 5.1.2 shows a bundle 901.108 nodalization using 26 nodes. However, the test data of Figures 5.1-3 to 6 are for " Data 12/31/95 I
Nodes" 28 to 31. The results section (Section 5.1.1.3) implies that these " data n l
l (Section 5.1.1) Since this is a natural convection loop, the mass flows through the tube 901.109 bundle, the bypass region and through the downcomer should be determined. Are any 12/31/95 data for these available, and if so, how do the TRACG predictions compare to these (Section 5.1.1) The reported void fraction distributions and the average rod temperature a generaHy show sausfac% agmement behen spedment and sWadon.
901.110 12/31/95 hewever, two anomalies should be addressed: Why do the TRACG predictions consdently (Section 5.1.2) This section gives a partialjustification of the nodalization. Under 901.111 "further subdivisions" "for more accurate representation," the nodalization of the lower plenum region is discussed. Two axiallevels are used, since one level wou (Section 5.1.2) The early disagreement between TRACG break flow predictions and test 901.112 data in Figure 5.1-19 is well discussed. However, a rationale for the significant 12/31/95 underprediction of the break flow between 50 and 150 s should be provided.
(Section 5.1.2) For CCFL breakdown times between 90 and 130 s, " noticeable changes" n upper plenum pressure drop are mentioned on Page 5-17. In particular for the 901.113 12/31/95 measured data, no such changes can be readily identified in Figure 5.1-23. A more direc 4
RAls on Qualification LTR (901.75 - 901.160)
Commit RAINo.
Request for Information Date l
(Section 5.1.2) Clarify the description and discussion of rod heat-up on Page 5-18.
901.114 "Little or no heat-up" at the top elevation applies for the test data (when referred to the 12/31/95 initial operating temperature) and for the " average rod" temperature predicti (Section 5.1.2.3, page 5-17) A TRACG " hot rod" model is introduced and described.
901.115 However, the staff could not find any reference to this " hot rod" model in the Model 7/24/95 Report. A description of this model should be included in the revised Model Report.
(Section 5.1.3, Figure 5.1-33) Provide an explanation for the test data bypass pressure 901.116 12/31/95 drop spike at about 17 s.
(e n
)
re erences to omer sedons oms repod and to references sM be 901.117 7/24/95 reviewed and corrected.
(Section 5.1.4) Various ECC flow rates are here given as fractions of those in previous 901.118 tests, but the absolute values are not given in either place. A table with the relevant flow 12/31/95 rates and temperatures should be provided.
(Section 5.1.4) The text of Section 5.1.4.3, describing Figure 5.1-40 (core inlet flow),
901.119 refers to unreliable core inlet flow data and states that dynamic effects a'fected the 12/31/95 accuracy o' the density determination. How is the test core inlet flow estabi (Secticrb.1.4) Figure 5.1-45 shows two separate test data traces, differiag from most 901.120 other figures "<ith rod temperatures. What is the second (lower) trace? (See also 12/31/95 Figures 51 0 a 4 21.)
(Section 5.1.4) Although no void fraction data were presented with the results of this 901.121 section, the conclusions of Section 5.1.4.4 claim good perfoTnance of the void 12/31/95 distribution models. Further, Page 5-42 refers to Figure 5.1-40 and claims that the lo (Section 5.2) This report should be free-standing and should demonstrate qualification 901.122 of the TRACG cocc. This section contains insufficient detailin the description of the test 7/24/95 facility and of the TRACG model. No rationale for the selected TRACG noda (Section 5.2.2) The discussion of Figure 5.2-3 states that an attenuation of the 901.123 depressurization rate at 11.5 s was observed. A slightly lower depressurization rate was 12/31/95 observed at about 18 s, but not at 11.5 s. Provide a basis substwaatng the clai (Section 5.2.2) Figures 5.2-6 and 7 are claimed to show a " faster reduction" of the 901.124 bypass and bundle inventories for the time period of 40 to 64 si If DP is accepted as a 12/31/95 measure of inventory, then the test data show a very slight decrease for this tim 5
RAls on Qualification LTR (901.75 - 909/i60)
Commit RAINo.
R
__equest for Information Date i
l (Section 5.3.2) Table 5.3-1 lists tt,e tests to be modeled by TRACG, along with the 901.125 rationale for why these ti.sts were r, elected. Tests B01 and 807 are both main steam 7/24/95 line breaks, however, there is no description of what the difference between these two i
j (Section 5.3.2) Aside from the arrangement of cooling water inventories, there are 901.126 additional differences between the SBWR design for which GIST was built, and the 7/24/95 current design. In particular, the reference to ADS components, (very different from the (Sectiom 5.3.4) Table 5.3-1 states that Test 807 (core uncovery and subsequent heat-up) would be the only SBWR-related test where these phenomena were observed.
901.127 7/24/95 However, this was not discussed in the text and no results of core flows and/or temperatures J
l (Section 5.3.4, Page 5-76) The report states that Figure 5.3-12 shows that test and I
901.128 calculation show onset of GDCS flow within 12 s of each other. However, the figure 7/24/95 j
shows at least 70 s (less than 450 s vs. more than 520 s). Explain this discrepancy.
(Section 5.3.4) Figure 5.3-13 shows some disagreement for the annulus pressure drop j
901.1" between test and predictions during depressurization ( 50 to 200 s) and, in Figure 5.3-12/31/95 4, the core pressure drop disagrees after GDCS onset. However, neither is discuss i
4 1
I i
(Section 5.3.4) Provide an explantion for why a GDCS flow over-prediction after 901.130 calibration against test data can be justified by a test data uncertainty of 10%. Since the 12/31/95 test data were used as a basis, better agreement would be expected, regardless of (Section 5.3.4) There are some disagreements in the annulus pressure drop data, in 901.131 particular in the 50 to 150 s range, and some in the post-30f; s range. These should be 12/31/95 discussed and quantified to substantiate the subpctive conclusion that these are "
(Section 5.4.3) The detailed qualitative description of the.'ccidant scenario points to the 901.132 importance of keeping the lower plenum two-phase level below tite side entry orifices
-12/31/95 (SEO). However, the results section stated that this was the case, without s (Section 5.4.3) The discussion of Figure 5.4-7 is not clear. The figure shows vapor 901.133 velocities, which turn negativa at about 13 s for the peripheral bundles. It is t-ot clear 12/31/95 how this indicates transition to liquid downflow at about 10 s, and transition 6
RAls on Qualifbation LTR (901.75 - 901.160)
Commit RAINo.
Request for information Date (Section 5.4.3) Apparently the SEO pressure drop is not an orifice DP, but a measurement across taps 0.567 m apart. The text description and data for Figure 5.4-8 901.134 12/31/95 differ. Oniv at close to 20 s (not 10 s) does the measured pressure difference exceed 0.5 (bectica 5.4.3) The calculated pressure drop data of Figures 5.4-8 to 14 show 901.135 oscillations of differing frequencies, which are not observed in the test data. Provide a 12/31/95 rationale expMning this behavior.
(Sections 5.5.1 & 5.5.2) The description of the facility, the test procedure (and also of 901.136 the TRACG model in Section 5.5.3.1) are adequate, but, the test itself is never identified.
7/24/95 It is initially referred to as "a system response test. The test proc (Section 5.5.3) Nine almost straight lines are presented on nine graphs and, at the 901.137 selected scales, agreement between test and experiment appears excellent. Since the test was run to demonstrate the operation of the PCCS, the actual heat removal rates, (Section 5.5.3, Figure 5.5-5) The report states that the initial peak pressure in the 901.138 drywell is well-predicted, however, the figure does not support this conclusion. The 12/31/95 report should provide a blow-up of that region and should explain why such a maxim (Section 5.5.3) Explain the differences between Figures 5.5-7 and 8 (same title, same straight lines).
(Section 5.5.3) Using the pool level as an indicator of heat removalis a relatively 901.140 inaccurate integral reading, telling little about the actual heat removal conditions. Also, 12/31/95 over the time period of 17 hr the predicted pool level drop is almost 20% la (Section 5.5.3) TRACG apparently over-predicts the PCCS heat removal rate. This 901.141 should reduce the drywell pressure. Explain why the drywell pressure agreement is 12/31/95 good, with an over-prediction in PCCS heat removal.
(Section 5.6) Provide a discussion explaining why the reference natural circulation j
901.142 experiment was not simulated, since it is much more relevant to SBWR applications than 7/24/95 l
the forced flow experiment. Apparently, the information for both experiments is in (Section 7.2) The data shown in Figures 7.2-2 and 3 appear to apply during recirculation 901.143 pump coast-down time. Only the transition to a natural circulation mode and the flow 7/24/95 under those conditions are of interest for SBWR applications. Since this plant (Sections 7.2 & 7.3) Provide a detailed description of the test conuitions and the data 901.144 12/31/95 measured during the plant tests and for the different time periods.
7
1 RAls on Qualification LTF; (901.75 - 901.160) i Commit RAINo.
Request for Inferriatien Date (Section 7.3) During the Hatch MSIV Closure Test, the quenching of steam by incoming 901.145 feedwater played an essential role. What actually happened to the feedwater flow during 12/31/95
)
this test? Was it generally set to match steamline flow; had it been tripped?
In Section 3.2.12, the Model Report points out need for a quenching heat transfer model 901.146 during reflood, which apparently is available as an option in TRACG. However, the 7/24/95 Qualification does not mention this model. Has it been validated and/or what are th (Figures 3.1-6 to 3.1-9) In the plots of void fraction versus quality (Figures 3.1-6 to 3.1-901.147 9), the measured data and the predictions show good agreement. However, to verify 7/24/95 this a reader must copy and superimpose the figures. The report should include 901.148 (Figures 3.1-18 & 19) Provide the units of vapor flux on the abscissa scale.
7/24/95 l
(Section 3.1.5) Table 3.1-8(5) lists four tests with top break and two with bottom break.
l 901.149 Why are only the first two of the top break tests simulated with the TRACG model ? Was 7/24/95 l
there a prcblem with the other two top break tests (5801-19 & 5702-16) ?
901.150 (Section 3.3) Provide the name of the "CSHT" facility.
7/24/95 (Equation 3.3-1) No notation is provided for this ec'uation. The staff notes that the 901.151 corresponding equation in the Model Report (Equation 3.2-54) differs slightly and is only 7/24/95 partly defined there. Provide notation for the equation and define it in the (Se tion 3.4) This section should be enhanced by including figures of the vessel 901.152 pressure vs. time.
(Table 3.1-8) Determine whether the third entry in Table 3.1-8 should be Test 5801-19 901.153 7/24/95 or -17, as referenced on Pages 3-54 & 55.
(Page 3-99) Apparently the figure titles on this page are reversed. Figure 3.9-3 shows 901.154 " Energy Release", which is more commonly referred to in the text as " integrated power".
7/24/95 For clarity, also indicate that it is integrated power in the figure title (o (Figures 4.2-2 & 5) Provide revised figure titles. Both figure titles are mislabelod; 7/24/95 carryover is shown.
(Page 5-3) The mixture level of 0.254 m (top of Page 5-3) appears to be in error.
901.156 Confirm that level is 2.54 m (100 in.), which would roughly correspond to the elevation 7/24/95 shown in Figure 5.1-6.
(5.1.3.3 should be to Figure 5.1-35 (not 5.3.1.9).
e n.
.)
n m ame rekence to a Ogwe in me nehasHne in Sedon 901.157 7/24/95 0
RAls on Qualific: tion LTR (901.75 - 901.160)
~
Co;nmit j
RAINo.
Request for information Date (Section 5.1.4, Page 5-43) The reference to Figures 5.1-44 and 45 in the text appears 901.158 wrong. Confirm that figure numbers and discussion refer to mid plane and upper 7/24/95 j
elevation (not 0.89 m elevation).
)
(Section 5.3.1) is the reference to recirculation line breaks in the second paragraph of 901.159 this section an error, since GIST was established for SBWR-related tests. If no' nrovide 7/24/95 l
an explanation.
1 i
901.160 (Page 5-130, first line) in a TRACG model, how does one fill the system with water?
7/24/95 t
.i i
i i
1 l
4 i
i i
i 4
b i
}
4 4
9 4
m ----,