ML20084A559
| ML20084A559 | |
| Person / Time | |
|---|---|
| Site: | Catawba, McGuire, Mcguire  |
| Issue date: | 05/12/1995 |
| From: | Tuckman M DUKE POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML19344C918 | List: |
| References | |
| NUDOCS 9505300384 | |
| Download: ML20084A559 (19) | |
Text
. . _ _ . _ . . _ . . . ._
. AAe AmerCavany . M S nmnv '
' l'O Box 1006 ' Senior bkeltesident
.,.. l ' ; s Omrlone, hC282011006 NuclearGeneranon (IN)3822200 DMce
- (imp 824360 Fat MMR .
l l
1 May 12,1995 .
U. S. Nuclear Regulatory Commission ' ,
Washington, D. C. 20555 Attention: Document ControlDesk 4
Subject:
Duke Power Company '
McGuire Nuclear Station Docket Numbers 50-369 and -370 Catawba Nulear Station Docket Numbers 50-413 and -414 Topical Report DPC-3004-P, " Mass and Energy Release and Containment Response Methodology"; Reponse to NRC Questions On September 30,1994, Duke Power Company submitted the subject topical report for review and approval. By letter dated May 3,1995, the NRC staff requested additional information about the .
report. Attachment 11 provides responses to the Staffs questions.
Please note that the responses to several of the questions contain information that Duke considers !
proprietary. In accordance with 10CFR 2.790, Duke requests that this information be withheld from public disclosure. An affidavit which attests to the proprietary nature of this information is included as i Attachment I. Attachment III contains a non-proprietary version of the responses.
The May 3,1995 request for additional information refers to McGuire Units 1 and 2, and Catawba Unit 1 only. It should be emphasized that, as indicated in the September 30,1994 letter which submitted the topical report, the analyses contained therein are also applicable to Catawba Unit 2.
While the primary purpose of the topical report is to support the replacement of the steam generators at McGuire Units 1 and 2, and Catawba Unit 1, the topical will also be used to support activities at all four units which are unrelated to steam generator replacement; such as a change in ice condenser loading. g j i W 9505300384 950512 PDR i P ADOCK 05000369 An w m ,nw m. PDR
{/U{ l
U. S. Nuclear Regulatory Commission i hiay 12,1995 ,
1 Page 2 l
l Ifyou have any questions, or need more information, please call Scott Gewehr at (704) 382-7581. i
- U hi. S. Tuckman cc: Nir. R. E N1artin, Project hianager Oflice of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission hfall Stop 141125, OWFN Washington, D. C. 20555 hir. V. Nerses, Project hianager Ollice of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission hiail Stop 141125, OWFN Washington, D. C. 20555 hir. S. D. Ebneter, Regional Administrator U.S. Nuclear Regulatory Commission - Region 11 101 h1arietta Street, NW - Suite 2900 Atlanta, Georgia 30323 1 1
W. O. Long Oflice of Nuclear Reactor Regulation )
U S. Nuclear Regulatory Commission j N1 ail Stop O- 8117. OWFN l Washington, D. C 20555 i l
l l
1 8 Attachment I Aflidavit to Support Proprietary Designation i
e m
l r . . . - - - , - - - - ,-e-,__------- - -, _ . _ _ _ _ , _ . - - - , , _ - . - - _ _ _ __
)
,~
. AFFIDAVIT OF Ni S. TUCKMAN
- 1. I am Senior Vice President, Nuclear Generation Department, Duke Power Company (" Duke"),
and as such have the responsibility of reviewing the proprietary information sought to be withheld from public disclosure in connection with nuclear plant licensing, and am authorized to apply for its witholding on behalf of Duke.
2 I am making this aflidavit in conformance with the provisions of 10 CFR 2.790 of the regulations of the Nuclear Regulatory Commission ("NRC") and in conjunction with Duke's application for withholding which accompanies this aflidavit.
- 3. I have knowledge of the criteria used by Duke in designating information as proprietary or confidential.
- 4. Pursuant to the provisions of paragraph (b)(4) of 10 CFR 2.790, the following is furnished for consideration by the NRC in detennining whether the infonnation sought to be withheld from public disclosure should be withheld.
(i) The infonnation sought to be withheld from public disclosure is owned by Duke and and has been held in confidence by Duke and its consultants.
(ii) The information is of a type that would customarily be held in confidence by Duke.
The information consists of analysis methodology details, analysis results, supponing data, and aspects of development programs, relative to a method of analysis that provides a competitive advantage to Duke.
(iii) The infonnation was transmitted to the NRC in confidence and under the provisions of 10 CFR 2.790, it is to be received in confidence by the NRC.
(iv) The infonnation sought to be protected is not available in public to the best of our knowledge and belief. I (v) The proprietary information sought to be withheld in this submittal is that which is !
marked in the proprietary version of the report DPC-NE-3004, " Mass and Energy l Release and Containment Response Methodology" and supponing documentation, and i omitted from the non-proprietary versions.
l
[b M. S. Tuckman 1%
(continued)
L
. AFFIDAVIT OF M. S. TUCKM AN (Page 2)
This infonnation enables Duke to:
(a) Simulate the mass and energy release rates from loss-of-coolant accidents and steam line break accidents in pressurizer water reactors of the Westinghouse design (b) Simulate the response of an ice condenser containment design to a high-energy line break inside containment.
(vi) The proprietary information sought to be withheld from public disclosure has substantial commercial value to Duke.
(a) It allows Duke to reduce vendor and consultant expenses associated with supporting the operation and licensing of nuclear power plants.
(b) Duke intends to sell the infonnation to nuclear utilities, vendors, and consultants for the purpose of supporting the operation and licensing of nuclear power plants.
(c) The subject information could only be duplicated by competitors at similar expense to that incurred by Duke.
- 5. Public disclosure of this information is likely to cause hann to Duke because it would allow ,
competitors in the nuclear industry to benefit from the results of a significant development I program without requiring a commensurate expense or allowing Duke to recoup a portion ofits expenditures or benefit from the sale of the infonnation. l 1
l b- w-M. S. Tuckman l (continued)
L AFFIDAVIT OF hf. S. TUCKMAN (Page 3)
M. S. Tuckman, being duly sworn, on his oath deposes and says that he is the person who subscribed his name to the foregoing statement, and that the matters and facts set Ibrth in the statement are tmc.
- - ~
M. S. Tuckman Swom to and subscribed before me this /5 day of //61 -, ,1995. Witness my hand and ollicial seal. g dlM W Nota'bj Ppic ~
My commission expires Al 22; 5 l
1 l
l l
l
i e.
E
- l 5 .
Attachment III !
Responses to Questions l (Non-Proprietary Version) !
l h . ,- , _w , +- . - .. . sy- --, , y. ..-y.- -- ;--._p - ,y
Response to NRC Questions on DPC-NE-30M Question 1:
Figure 2.1.3-12 of the topical repon illustrates a difference between the FSAR analysis and l RELAP analysis for the mass and energy release rate during the period from 300 to 1500 seconds.
In Section 2.1.3 of the topical report, this difference is attributed to differences in intact S/O heat transfer due to RELAP using mechanistic modeling in which heat transfer is dependent on mass flow rate. Please provkle a more detailed discussion of these differences.
A description of the FSAR methodology " Westinghouse Mass and Energy Release Data for l I
Contaltunent Design", WCAP-8264-P-A (Proprietary) and WCAP-8312 (Non-Proprietary) is provided in the NRC letter of March 12,1975 (reference 1). Also the NRC staffletter of February ,
17,1987 (reference 2) pmvides an evaluation of WCAP-10325 " Westinghouse LOCA Mass and Energy Release Model for Containment Design" relating to the vendor methodology. To what extent am the differences to which you refer related to issues discussed in these two references.
Response
%c March 12,1975 safety evaluation provides details regarding the treatment of heat transfer fmm the steam generators during reflood. The E REFLOOD code is used to calculate the mass and energy release during this portion of the analysis. De H REFLOOD code utilizes a loop hydraulic resistance model and an energy balance model. We steam generators are assumed to be cooled to the temperature of saturated steam at the containment pressure during the reflood phase of the transient. His pmcess is detennined to have been completed after 960 seconds, as stated on page 6. His assumption is based upon "the maximum steam flow based upon the hydraulic resistance and steam generator heat transfer." During the post-reflood phase of the transient, heat transfer from the steam generators is detennined by the containment depressurization rate. The February 17,1987 staff evaluation describes changes to the original methodology. De change made with respect to steam generator heat transfer is to assume that only saturated steam exits the steam generators. Eliminating the release of superheated steam from the steam generator side of the break increases the overall steam mass flow released to containment.
In the McGuire/ Catawba RELAPS model, heat transfer fmm the steam generators is determined, in pan, by the mass flow tiuough each steam generator. The flow split between the broken and intact loops is detennined by the hydraulic resistances in the model, and is not modified in the analysis.
We bmken steam generator rapidly cools down to the temperature of saturated steam in the RCS as expected, due to high flow rates in the affected loop, lleat transfer from the intact steam generators is limited by the mass flow through the primary side of the tube bundle and the degree of superheat that can be achieved. He steam generator secondary side initial conditions are specified to maximize the available stored energy, thereby maximizing the heat transfer to the primary and the mass and energy release out the break.
Question 2:
Refening to the discussion of the ice condenser heat transfer correlations available in GOTillC-4.0/ DUKE (pg. 2-16 of DPC-NE-30N). describe which correlations and/or heat transfer coeflicient values are actually selected. Indicate if an upper limit value for heat transfer coefficient is input to the program. Refer to the staffs SER " Staff Evaluation of the Tests conducted to Demonstrate the Functional Adequacy of the Ice Condenser Design." of April 25.1974 (PDR ttS603040079) in which an upper limit coefficient of 10,000 BTU /hr-sq.ft was established.
Papei
h e
N o
4 9
- n Response:d Ti jis used to detemiine a heat transfer coefficient for the ice condenser heat transfer equations in all GOTillC-4.0/ DUKE containment analyses discussed in DPC-NE-30M. This correlation [
]provided the best matches with the data generated in the Ice Condenser Test Facility heat transfer tests. Ileat transfer fmm the steam to the ice is calculated in GOTlilC for each node containing ice. Separate heat transfer coefficients are calculated for each node. Tids is different then the LOTIC-1 ice condenser heat transferlogic, where the ice melt rate (and hence, the heat transfer coefficient to the ice) is dependent on the steam flow rate into the condenser, assuming constant meltwater temperatures and complete condensation of steam prior to ice meltout. Even if one or more nodes exceeded the 10,000 BTU /hr-sq.ft-F heat transfer coefficient value (the limit imposed on ice condenser heat transfer in LOTIC-1, per PDR #8603040079) in a G0rlilC analysis, the heat transfer coefficient for the entire ice condenser would still be well below the value of 10,000.
I /
There is' las there is in the Uchida correlation. Therefore,it is possible for
~
the heat transfer coefficien( ]to exceed the 10,000 BTU /hr-sq.ft-F value in a single node. In detemiining what the maximum heat transfer coefficient is in the DPC-NE-30M GOTillC analyses, it is observed that thc[ . .
] The highest velocities through the ice condenser occur during the very early stages of blowdown. At this time (0-2 seconds) the highest ice melt rates are also observed. The highest heat transfer coefficient for any ice condenser node was on the order of
( ]DTU/hr-sq.ft-F, directly above the break, at the lowest elevation in the ice condenser. The flow vek) city through this node was the highest of any ice condenserlocation at any time during the transient. 'Ihe 10,000 value is not approached in any transient.
Question 3:
Section 2.3.2 discusses the spray droplet size used, which is significantly less than the size used for Oconec. Was this decrease intentional? If possible relate your selection to the data in WCAP-8258,"SPRAYCO Model 17143A Nozzle Spray Drop-Size Distribution".
Response
The difference in the assumed droplet size for the spray droplets between the McGuire/ Catawba and Oconee containment models was intentional. The spray headers are required at McGuire and Catawba to produce a dmplet size spectrum with a mean diameter ofless than 700 rn. Figure 6-195 of the McGuire FS AR shows the distribution produced by the McGuire spray headers. This figure shows that the assumption of 700 m for an average droplet size is therefore conservative.
At Oconce, there is no corresponding requirement for a maximum spray droplet size. An average droplet size of 700 m is assumed in the Oconce FATI1 OMS base model. This is increased to 7000 pm in the long term containment response analyses, which introduces additional conservatism. 'Ihc FATilOMS containment trsponse is very insensitive to chmges above the 700 pm size.
Page 2
. n f
.. s
' Question 4:
A low initial containment temperature is generally conservative for the blowdown peak pressure I
detennination. Ilowever, if the limiting peak is relatively late in the event, might a high initial '
containment be conservative due to reduced heat stmeture heat absorption capacity. Provide
- assurance that the initial conditions of 2.3.2 are indeed conservative for the LBLOCA analyses.
Response
A sensitivity run was conducted for the CNS-1 cold leg pump discharge break with the initiallower and upper containment temperatures increased to the upper Tech Spec limits of 120 and 100 *F, respectively. (Re ice condenser initial temperature remained unchanged at 30 *F.) The peak pressure decreased in the sensitivity case from 11.76 psig to 10.79 psig. De effect of an increased 1 air mass (low initial temperature) on the eventual peak pressure following a LOCA is greater than i' that of reduced heat structure heat absorption capacity (high initial temperature), as stated in Section 5.2. It is expected that this trend would hold for all MNS/CNS cases.
Question 5: '
It would be useful to include, for each LBLOCA analysis, a table identifying the peak pressure and time of peak pressure for each identifiable peak (similar to those included in DPC-NE-3003P).
Response
See Table A.
Qacstion 6:
Section 3.3.2.12 states that the long-term analyses are insensitive to a 20 second refill assumption.
Since we have no other information readily available to confimi this, and since ANS 56.4-1983 states that justification should be provided for use of a non-zem refill time in long-tenn analyses, ,
please explain the reason for the non-zem refill assumption.
Response
He 20 second refill time mentioned in Section 3.3.2.12 is obtained from page 3 3 of NSAC/86 (Reference 3-10). He refill phase of a large break LOCA transient is defined as the period of time between the end of blowdown and tir beginning of reficod. The reflood phase of the transient begins when the mixture level in the reactor vessel lower plenum reaches the core inlet. He timing of refill and reficod can be significant in a dry containment design because the peak containment ;
pressure occurs during the early refiood period. For the ice condenser containment design the refill phase and associated phenomena are short-term concems and the peak pressure occurs much later.
Werefore the timing of refill has no effect on the peak pressure response.
Question 7:
Section 3.4 postulates that the LDLOCA case having the greatest integrated steam release will i
pmduce the limiting peak pressure (i.e., insensitivity to timing effects). Ilas this phenomena been shown by analysis?
Respmse:
Analyses have been performed for each of the three break locations for the durations necessary to establish the long tenn mass and energy release trends, and to identify the limiting location. He
! l Page 3
[
' key considerations are die time of ice meltout, and the steaming rate out the break after ice mettou The steaming rate out the break is maximized for the pump discharge break, where ECCS is lost f
0 due to spilling out the break and less steam condensation results.(Refer to the response to QuestionD 7 for details) %e containment pressure is higher for the pump discharge break at the time that the "
long-term steaming trends are established. Derefore,it has been shown by analysis that the pump discharge break is the limiting location.
d Question 8:
l i Please explain the phenomena of the 3000 sec crossover (3.4.1.4) where spilled ECCS fluid causes q the RCP discharge case M&E to be greater than the RCP suction case. g i
Response: . $
Both the pump discharge break case and pump suction case are initiated from the same initial 3 condidons. De initial differences in the integrated break vapor mass and energy release observed n in Figure 3.4.1.4-2 and Figure 3.4.1.4-4 are a result of break location specific blowdown phenomena. Ecse differences are consistent with those presented in the cunent FSAR analyses.
Ec integrated steam release for the pump suction case following blowdown initially exceeds that fmm the pump discharge case. The primary cause for this difference is implicit to the brerk location assumptions. For the pump discharge break location, all ECCS to the broken loop is assumed to be spilled directly to containment and is therefore not available for condensing steam before it reaches the break. ECCS flow is not spilled directly to containment for the pump suction case, allowing steam to be condensed in the broken cold leg prior to reaching the break. He impact of this difference is that the steam release for the pump discharge case eventually exceeds that of the pump suction case. He timing of the crossover is approximately 3000 seconds for the case illustrated in Figures 3.4.1.4-2 and 3.4.1.4-4 He same phenomena is demonstrated in Figure 3,4.2.4-2 and Figure 3.4.2.4-4, although the timing of the crossover differs.
Question 9:
Paragmph 4.4 states that primary and secondary metal structures are initially in equilibrium with the surrounding coolant, with a constant temperature distribution. Explain how this assumption is applied to structures in contact with both primary and secondary coolant.
Response
his assumption applies to structures that art only in contact with either the primary or secondary coolant, but not both. %e only structures that are physically in contact with bodi the primary and secondary coolant are the steam generator tubes and the tubesheet. The tubesheet metal is at die temperature of the pnmary coolant flowing through it. He small amount of primary-to-secondary heat transfer that occurs at the top of the tubesheet is neglected. Therefore, die tubesheet is modeled by a one-sided conductor connected to the primary side. De steam generator tubes are modeled with a linear tempemture gradient across the heat conductors, which is determined by the adjacent water temperatures.
l l
I' age 4 l
Question 10:
Referring to Fig 4-24, explain why such a large percentage of the total AFW flow is delivered to the intact S/ds (instead of the faulted S/G which is at a much lower pressure). At what time is AFW flow to the faulted S/G terminated? Explain the differences between the AFW flow rate curves for W and BW1 S/Gs (Figs 4.5.3 and 4.5.17).
Response
The AFW flow rates plotted in Figures 4.5.3 and 4.5.17 represent the total AFW flow to the dure intact steam generators and the AFW flow to the single faulted steam generator. 'Ihus, AFW flow )
to the faulted steam generator (~ 1200 gpm)is nearly double that seen by each intact steam l generaror (~ 2000 gpm total or 670 gpm each). l AFW flow to the faulted stea generator is not terminated prior to the end of the mass and energy release analysis. Isolating AFW i'ow to the faulted steam generator would non-conservatively terminate the flow of steam thmuda the break. 'Thus, the mass and energy release is analyzed with continued AFW flow past the time of peak containment temperature.
AFW flow is specified in the RETRAN input deck as a function of steam generator pressure with an enthalpy corirsponding to a given temperature. For the Westinghouse preheater steam generators, a purge volume of hot water is assumed to be delivered before the cold AFW reaches the steam generators. For break sizes above 0.6 ft 2, pressure in the faulted steam generator dmps below the saturation pressure of the hot AFW before the purge of the hot AFW is completed.
When pressure in the faulted generator falls below the saturation pressuit of the hot AFW,it is assumed that the water in the AFW piping flashes and is added to the faulted generator as steam.
Thus, a sharp increase in the volumetric AFW flow rate is seen between 65 and 85 seconds in Figure 4.5.17 due to the much lower density of the steam. Figure i shows the AFW mass flow rate to the faulted steam generator that corresponds to the volumetric flow rate shown in Figure 4.5.17.
Question 11:
For MSLB temperature analysis, it is the reviewer's understanding that the limiting break size is normally that at which entrainment stans to occur. Provide a discussion of how entrainment is accounted for in the RETRAN model for large MSLBs. Also, can you provide the staff with infonnation regarding the sensitivity of MSLB temperature to A(T)[ref page 2-16 of DPC-3004).
Response
From a peak containment temperature pelspective, it is conservative to assume that all of the break flow is released in the form of steam. Therefore, the steam dome volume is modeled as a bubble rise volume with a very large separation velocity. This provides nearly instantaneous and complete separation between the liquid and vapor phases in that volume and precludes any liquid fmm being entrained in the flow leaving through the steam generator outlet nozzle.
'the( )issumed in the GOTIIIC conminment model has a negligible impact on the lower containment temperature response following a MSLB. The peak temperature in lower containment is reached in the first few minutes following a steam line break. This is well before l it is also well before the time when the time of ice mcitout( flower containment would be me ice aluve any one sector o .
)
Page 5 )
l 1
1
=
} 1 V
v ;
N l T [s J
Question 12:
Section 5.2 indicates that the initial ice mass is the Technical Specification (TS) value (which includes a sublimation allowance). Is the sublimation ice mass allowance assumed to be available for containment heat removal? Is the ice mass assumption for McGuire greater than FSAR/LOTIC-1 values?
Response
ne allowance for ice sublimation is not assumed to be available in the GOTlilC containment model. Although this sublimed ice may still be present in the condenser, in the fonn of water vapor / frost, no credit is taken for it in the GOTillC model. The ice mass in the GOTlilC analyses is 13.9% below the Tech Spec value of 2.475 E6 lbm. De GOTHIC ice mass assumptions in both McGuire and Catawba analyses are the same as those in the most recent respective FSAR analyses, performed with LOTIC-1.
Question 13:
De RETRAN-02 code has been approved by the staff for generic use in 1984 and 1991 (references 3 and 4).Section II.C of the Safety Evaluation (SE) identifies " General Limitations" mganiing use of the code. Provide a general discussion of the extent, if any, to which your use of j RETRAN-02 deviates from the limitations. 5
Response
We conclusions regarding the limitations specific to the steam line bmak mass and energy release analysis art given below. De numben and letters correspond to those listed in the SERs.
Liinilations from the MODS.O SER:
2.1 ne general transport model is used to simulate the injection of boron into the primary system. 'Ihe conservative application of this model with respect to purge volumes and assumed borun concentrations is discussed in Section Sc3.2.5 of DPC-NE-3001.
2.2 he 1979 ANS Standant decay heat model is used in the analysis with an added two-sigrna uncertainty to bound all uncenainties associated with the input parameters to the decay heat model.
Litni.lations from the MOD 2.0 SER:
1, n3. [
and energy release results are relatively unaffected dby the use of this m is not allowed to exit the steam generator as discussed in the response to Question 10 above. %c A13V flow rate and primary-to-secondary heat transfer are of primary imponance in this analysis, and their conservative application produces a limiting mass and energy release.
Page 6
.. .~
Question 14:
Provide a general discussion of the extent to which DPC's use of the RELAP 5 code deviates, if any, frum the recommended uses and practices described in the Code Manual (NUREG/CR-5535, Vol. 5 User's Guidelines).
Resp (mse:
In general, the practices presented in the Usefs Guidelines (Vol. 5 of the RELAPS/ MOD 3 Code Manual) represent the basic infomiation required to perform transient analyses. These practices are consistent with those used at Duke Power. A review of Vol. 5 identified three instances where our use of RELAPS differed from the authofs recommendations. 'lhese instances are discussed below.
In Section 4.6.8.1, it is recommended that the frictional torque specined for a pump component be divided equally betwecn TFD and TF2. The frictional torque input to the McGuire/ Catawba RELAP5 model is specified entirely as TF2, The pump model used in the RELAP5 model is the same as that described in the Duke Power RETRAN model for McGuire/ Catawba, which is documented in the NRC-approved topical report DPC-NE-3000-PA.
In Section 4.6.9, the author of Vol.5 recommends against using the multiple junction component, based upon the potential for confusion in identifying the location of a specific junction in the output. The multiple junction component is used in the secondary side of the tube bundle in the BWI FSG model.
In Section 5.1.8, the fm' al recommendations made in this section are to use a two-component representation for the core if simulation of the high-powered fuel rod behavior is important in meeting the analysis objectives, to not model crossflow in the downcomer, and use a simplified system nodalization if possible. 'the first recommendation does not apply for a mass and criergy release analysis. In the McGuire/ Catawba RELAPS large break LOCA model, the reactor vessel
[
were botit necessary to accurately model the LOCA mass and energy release.
Question 15:
Pmvide a discussion of the significant reasons for the change in the b tak location of the limiting LBLOCA fmm the pump suction to pump discharge.
Response
'Ihc mass and energy release analyses in the current FSAR cncompass the blowdown phase for the hot leg break, the blowdown and rellood phase for the pump discharge break and long-term for the pump suction break case only. If the limiting tmnsient were to be selected based solely on the current FS AR analyses, the same conclusion would be reached. The RELAP5 analyses presented for cach break h) cation are analyzed further out in time, thereby gaining additional insights into the long-temt response for dif ferent break locations. A crossover in the integrated mass and energy tricasc occurs for the two cold leg break locations during the cold leg recirculation phase of the transient. The phenomena behind the crossover in the integrated mass and energy release are discussed in the response to Question #7 above.
{
i Page 7 l
l l
J
1 I
' ' Quest' ion 16: )
The new LBLOCA analyses produce significantly lower peak pressure. This can be attributed to I loth the mass and energy release modeling and the containment modeling. lias the GOTillC model been run with the FSAR mass and energy data to determine the relative contribution of each?
Respntse:
The GOTillC runs conducted to benchmark the containment simulation model, described in Section 2.3.3, use the FSAR mass and energy release data. As mentioned at the bottom of p. 2-22, there are some uncertainties as to the exact data used in some phases of the LOTIC-1 FSAR analyses. The results of this run indicate appmximately how much the change in containment analysis methods alone impmved the peak pressure result. In this GOTHIC run, a peak pressure of 12.90 psig was calculated. Compared with the FSAR result of 14.07 psig, a decrease of 1.17 psig is achieved in the peak contairunent pressure when utilizing GOTHIC alone, without new mass and energy release data.
The CNS-2 cold leg pump discharge mn documented in Section 5.4.2 utilizes RELAP5 mass and energy release data and the GOTHIC containment analysis model. In addition, the CNS-2 analysis is for the Westinghouse preheater steam generator type; the impact of the BWI feedring S/G is not present in this analysis. The results of this analysis shows the significance of the RELAP5/GO1111C analysis package, without the differing S/G geometry effect. As documented in Section 5.4.2, the CNS-2 cold leg pump discharge analysis resulted in a peak containment pressure of 10.29 psig. When subtracted from the Catawba FSAR peak pressure of 14.05 psig, a decrease of 3.76 psig is achieved when utilizing the RELAPS/ GOTHIC analysis package. This is mughly 3 times the decrease when using GOTHIC alone, as mentioned above. Therefort, the relative contribution of each code may be generalized as 2 parts REl.APS,1 part GOTHIC.
Question 17:
Describe DPC's intentions with respect to Appendix K minimum pressure analyses. Does DPC seek appruval for use of GOTHIC to establish a higher Appendix K minimum pressure? Would structural heat transfer coefficients be consistent with the guidance of ANS-56.4-1983? Is the 700F spray temperature Key Assumption in Section 5.6 conservative for winter conditions?
Response
The minimum containment pressure analysis methodology described in Section 5.6 will be used to calculate the minimum containment backpressure boundary condition for future LOCA PCT analyses. The GOTillC model and inputs used for the peak pressure analyses will be significantly modified as described in order to conservatively predict a minimum pressure response. The predicted pressure response may be higher or lower than the current FS AR analysis, depending on the initial and boundary conditions assumed.
Due to the large heat sink effect of the ice condenser, the modeling of structural heat sinks plays a much smaller role than in conventional dry contakunent design. A +5% allowance for stmetural heat sinks is included. The ANS 56.4-1983 heat transfer coefficient modeling guidance is unnecessary, and is not used.
l Page 8
' The 700F spray ternperature assumption is consistent with the minimum temperature specified in Technical Specification 3.5.4 for the Catawba Nuclear Station and Technical Specification 3.5.5 for the McGuire Nuclear Station.
Question 18:
Section 6.1 states that "once the affected steam generator is isolated, the release of steam to die containment is essentially finished." is this staternent intended to refer to the feedwater flow?
Response
Yes. Main feedwater flow is au omatically isolated, and once the auxiliary feedwater flow is manually isolated fmm the faulted generator, the steam sticase to containment is finished.
Question 19:
Fmm Section 6.5 it is not clear as to what was the highest peak temperature among all the cells of the lower compartment. The fact that the break volume was cooled by Jetting raises the question as to whether other areas were subjected to increased local temperatures. What is the peak temperature experienced in the lower compartment?
Response
The steam crane wall. line inbreak As discussed is assumed Section 6.2, there are[ to occur in thethenode GOTIIIC closest to the c steam line break containment model. Due to the high pressure at which the steam is injected through the break in the early stages of the transient, all[ 3areat essentially the same temperature. The cooler air coming in behind the break due to the jetting effect is forced along with the steam to the next node downstream of the break until hitting the crane wall. Allh ]in this sector of lower containment, are at the same pressure, temperature, and steam / air concentration. Upon hitting the crane wall, the steam / air jet is re-directed in all directions and mixes around containment. The break compartment peak temperature is referred to as the peak temperature in lower containment, although the nodes adjacent to the break compartment in the direction of the break flow are at the same temperature.
This does not hold true throughout the transient, however. As the steam line pressure decreases, the jetting forte decreases, and the jetting flows decrease relative to the break flow.1he node adjacent to the crane wall is slightly hotter than the node at the break location later in the transient, llowever, the peak temperature for the entire transient has already been reached by the time this jetting decrease occurs.
Question 20:
The Safety Evaluation Report, NUREG-0847 Supplement 7 for Watts Bar describes COBR A-NC analyses perfomied by Westinghouse for Catawba and subsequently applied also to Watts Bar.
TVA found that the MSt.B hot spot locations are not locations containing envimnmentally-qualified equipment. To what degree can DPC state that the Catawba COBRA-NC analyses are consistent with DPC GOTillC analyses? To what extent can DPC state that the COBRA-NC methods described in WCAP-10988-P (proprietary version) and WCAP-10989 (non. proprietary version) are consistent with DPC's GOTillC code and models?
l l
Page 9
,, ,s
Response
%c location of envirbnmentally-qualified equipment within the Catawba /McGuire containment buildings wonid not be significant since the maximum temperatures calculated by GOTIIIC following a MSLB do not exceed the EQ limit of 340aF.
. %c GOTillC analyses are consistent with the COBRA-NC results in the Westinghouse WCAP. l 10988 P insofar as the COBRA-NC code is a precursorof the GOTillC code. He prediction of ;
the jetting effect and its subsequent cooling oflower contairunent are consistent between the l analyses, although the magnitude of the cooling may be different between analyses. Other consistencies include the closing of the ice condenser doors above the break due to thejet-induced pressure decrease. De maximum bulk temperature of 29l'F in the COBRA-NC analysis (Model '
- 2) is very close to the 297+F peak in the GOTHIC analysis. ne peak temperature is about 15+F wanner in the COBRA NC results, which can be attributed to the smaller node sizes used.
Many other modeling factors are involved which could impact the differences in flow pattems and longer-tenn temperature increase in the COBRA-NC analysis which is not present in the GOTHIC analysis. Among these factors is th(
3 i
%e GOTillC code itselfis technically more advanced than COBRA-NC. De separate set of energy equations present in GOTHIC for the droplet phase could have a major impact on the analysis results. The modeling of flow paths (junctions) in GOTHIC is completely different than COBRA NC. In which these flow paths were simply left as gaps in the calculational mesh. De interfacial heat transfer routines have been fine-tuned with an additional ten years of comparisons with test data during the development from COBRA NC to FATHOMS to GOTIIIC 4.0. All .
differences between the COBRA-NC and GOTHIC codes would provide a higher degree of }
accuracy and certainty with the GOTHIC code. De conservatism required to ensure a :
conservatively high peak building temperature is applied through the selection of conservative initial and boundary conditions throughout all GOTHIC analyses, as well as in the mass and energy release calculations.
Question 21:
McGuire Unit 2 LER 85-29 of October 31,1985 describes how spray (NS) pump switchover ,
initiation time and delay interval could cause LOCA peak pressure to exceed the containment design pressure depending on initial FWST level. Does the DPC-NE-3004-P methodology account .
i for the concems identified in this LER7
Response
%c primary concem identified is the potential for ice mettout prior to completing the transfer of the containment spray pump suction from the FWST to the containment sump. His would result in a time frame during which no ice remained without containment spray. De time at which auxiliary containment spray is initiated was changed from 60 minutes to 50 minutes to address this concem. The same timing assumptions regarding auxiliary containment spray initiation are made in the DPC-Nil-30lM methottology.
Page 10 l
. . - _ - . . - , _ - _ . . . - . -. . .. ..- . _ - - - -- _._____-i
Westinghouse Preheater SG 1.4 Ft2 Steam Line Break Mass and Energy Release 160 =
~
M 140 '
7 :
=i :
2 120 7
S :
~ . #
3 100 _
c 2
I? -
~
I $ 80 '
i a:
O
~
40 g :
- -s -
[ 20[
0 . . . . . . . . . . . .
0 50 100 150 200 250 300 350 400 Time (seconds)
- - :n r__
TABLE A DPC-NE-3004 GOTIllC Analyses - Peak Pressures Figure Peak pressure Time of Title peak pressure (sec)
No. (psig) 2.3.31 12.90 4900 MNS FSAR CLPS Break 5.4.4.1-1 11,77 5600 CNS-1 Pump Discharge Break 5.4.2.1-1 10.29 6850 CNS-2 Pump Discharge Break Note: Only cases run past the time ofice mettout have peak pressures reponed. Cases which were not the limiting cases were not run to this point. Initial peaks (within a few seconds of the end of initial NC System blowdown) are not reponed, as the maxunum pressure in all cases is expected to be reached after ice mettout.