ML20206B124
| ML20206B124 | |
| Person / Time | |
|---|---|
| Site: | Haddam Neck File:Connecticut Yankee Atomic Power Co icon.png |
| Issue date: | 04/02/1987 |
| From: | Frank Akstulewicz Office of Nuclear Reactor Regulation |
| To: | Mroczka E CONNECTICUT YANKEE ATOMIC POWER CO. |
| References | |
| NUDOCS 8704080345 | |
| Download: ML20206B124 (17) | |
Text
,
O April 2,1987 Docket No. 50-213 Mr. Edward J. Mroczka, Senior Vice President Nuclear Engineering and Operations Connecticut Yankee Atomic Power Company Post Office Box 270 Hartford, Connecticut 06141-0270
Dear Mr. Mroczka:
SUBJECT:
REQUEST FOR ADDITIONAL INFORMATION CONCERNING NORTHEAST UTILITIES' REPORT - NUSCO 140-1 "NUSCO THERMAL HYDRAULIC MODELQUALIFICATIONVOLUME1(RETRAN)"
Re:
Haddam Neck Plant By letter dated July 30, 1984, Northeast Utilities Company (NUSCO) sub-mitted the Topical Report NUSCO 140-1 "NUSCO Thermal Hydraulic Model Qualification Volume 1 (RETRAN)" for NRC review and approval. The staff has completed its preliminary review of this report applicable to the Haddam Plant and concluded that additional information is required in order to complete the review.
A request for additional informatten is enclosed.
In order to complete its review of the RETRAN code prior to the upcoming refueling outage, the staff requests that responses to these questions be submitted within 45 days of receipt of this letter. The staff and its consultants are available to meet with NUSCO in an effort to resolve the issues in an expeditious manner.
If you are unable to meet this schedule yourlicensingdepartmentshouldcontactmeat(301)492-4790 with a pro-posed schedule for submittal of the necessary information.
The reporting and/or recordkeeping requirements contained in this letter affect fewer than ten respondents; therefore, OMB clearance is not required under P.L.96-511.
Sincerely,
@ og y 73@00$13 Francis M. Akstulewicz, Jr., Project Manager PDR Integrated Safety Assessment p
Project Directorate Division of PWR Licensing-B
Enclosure:
DISTRIBUTIO_N As stated
_ U6d et File 7 FMiraglia OGC-Bethesda CYLiang NThon EJordan ISAPD File ACRS(pson 10)
BGrimes cc:
See next page NRC POR LOCAL PDR RJones JPartlow FAkst61e///f
//)i D:!SAPD ISAPD ISAPD wicz:cm PAnde on of CThomas 4/2./87 4/ t/87 4/ t/87
e.
jf...,k UNITED STATES g
NUCLEAR REGULATORY COMMISSION 5
l WASHINGTON, D. C. 20555
%,.....,/
April 2,1987 Docket No. 50-213 Mr. Edward J. Mroczka, Senior Vice President Nuclear Engineering and Operations Connecticut Yankee Atomic Power Company Post Office Box 270 Hartford, Connecticut 06141-0270
Dear Mr. Mroczka:
SUBJECT:
REQUEST FOR ADDITIONAL INFORMATION CONCERNING NORTHEAST UTILITIES' REPORT - NUSCO 140-1 "NUSCO THERMAL HYDRAULIC MODELQUALIFICATIONVOLUME1(RETRAN)"
Re:
Haddam Neck Plant By letter dated July 30, 1984 Northeast Utilities Company (NUSCO) sub-mitted the Topical Report NUSCO 140-1 "NUSCO Thermal Hydraulic Model Qualification Volume 1 (RETRAN)" for NRC review and approval. The staff has completed its preliminary review of this report applicable to the Haddam Plant and concluded that additional information is required in order to complete the review. A request for additional information is enclosed.
In order to complete its review of the RETRAN code prior to the upcoming refueling outage, the staff requests that responses to these questions be submitted within 45 days of receipt of this letter. The staff and its consultants are available to meet with NUSCO in an effort to resolve the issues in an expeditious manner.
If you are unable to meet this schedule your licensing department should contact me at (301) 492-4790 with a pro-posed schedule for submittal of the necessary information.
The reporting and/or recordkeeping requirements contained in this letter affect fewer than ten respondents; therefore, OMB clearance is not required under P.L.96-511.
Sincerely,
. b w ;**
Francis M. Akstulewicz, Jk., Project Manager Integrated Safety Assessment Project Directorate Division of PWR Licensing-B
Enclosure:
As stated cc:
See next page l
i
Mr. Edward J. Mroczka Connecticut Yankee Atomic Power Company Haddam Neck Plant cc:
Gerald Garfield, Esquire Kevin McCarthy, Director Day, Berry & Howard Radiation Control Unit Counselors at Law Department of Environmental City Place Protection Hartford, Connecticut 06103-3499 State Office Building Hartford, Connecticut 06106 Superintendent Haddam Neck Plant Richard M. Kacich, Manager RFD #1 Generation Facilities Licensing Post Office Box 127E Northeast Utilities Service Company East Hampton, Connecticut 06424 Post Office Box 270 Hartford, Connecticut 06141-0270 Wayne D. Romberg Vice President, Nuclear Operations Northeast Utilities Service Company Post Office Box 270 Hartford, Connecticut 06141-0270 Board of Selectmen Town Hall Haddam, Connecticut 06103 State of Connecticut Office of Policy and Management ATTN: Under Secretary Energy Division 80 Washington Street Hartford, Connecticut 06106 Resident Inspector Haddam Neck Nuclear Power Station c/o U.S. NRC P. O. Box 116 East Haddam Post Office East Haddam, Connecticut 06423 Regional Administrator, Region I U.S. Nuclear Regulatory Connission 631 Park Avenue King of Prussia, Pennsylvania 19406
= m. u. _ i,
- m ENCLOSURE 1 REVIEW 0FNORTHEASTUTILITIESSERVICECOMPANY'S(NUSCO)
RETRAN CODE FOR FADDAM NECK PLANT REQUEST FOR ADDITIONAL INFORMATION A.
General 1.
It is stated in the topical report that "ancther inportant effect for asymmetrical cooldown events... is the mixing of primary coolart in the reactor vessel." However, this effect was ret addressed in the report.
Provide an evaluation of the necessity to model reactor vessel mixing for asymmetric cooldown transients including the stean.
line break accident.
B.
hocalization 2.
The report describes the use of fill jur.ctions to model pressurizer spray. The RETRAN manual advises use of a time dependart volume instead of a fill function. Demonstrate that this moceling gives proper flew vs pressure drop across the spray line and explain why it deviates from the recommended procedure.
3.
hCSCO used the " local conditions heat transfer" option in the steam generator.
Explain what is beir.g attempted to be modeled with this option and how this code cartbility accomplishes NUSCO's goal?
i 4
~ ~ -. - _, _.. - _. -, -. _. - -, - _, - -, _, - - _ _ - -, + -, - - - - -,.. - - - - - +. - _ _
.. -... - -. - - - _ _. - _ _, - _. _... _ - - - - - - - _ - -, _. _ _ _ _. ~. - -. - -. -
. ~ - -
2 C.
Steam Line Break 4.
Explain how reducing the height of U-tubes affects the heat-transfer computation. Discuss how the two-phase mixture level on the secondary side impacts the computation and what the effect of using the greater height of U-tubes would be.
5.
Compare the reactivity insertion from 1% power to that for full power. Also, discuss the impact of assuming RCP trip at the break opening instead of waiting until trip signal and delay.
6.
Explain the reasons for selecting and not selecting the parameters for sensitivity studies, e.g., why was the secondary U-tube noding not varied? Discuss whether the parameters analyzed are independent of each other enough to vary separately.
l 7.
Provide details of each sensitivity study by providing plots of primary pressure, hot and cold leg and core outlet temperatures, l
secondary pressure, temperature and mixture level. Compare the l
plots of each of these variables against those of the base case.
8.
Figure II.B.1 shows that during a steam line break transient j
the core coolant temperature continues to decrease while the primary pressure remains relatively constant after about 40 seconds.
Explain why.
. 9.
Since the three secondary noding models analyzed all had different bubblerise(BR)models,andthereforedifferentfluidvolumevs height which would result in differences in heat transfer, it is difficult to compare the separate effect of the nodalization changes, i
Therefore, provide the following:
1)
Explain what would be expected from simply increasing the number of nodes.
ii)
Explain why an infinite BR model was not used for three volume and eight volume secondary nodalizations.
l iii)
The conclusion that a single volume secondary is conser-vative does not appear to be warranted because it is based upon different assumptions. Reassess this conclusion after completion of the analyses indicated above.
10.
The comparison of the RETRAN results to the results of the contain-ment system experiments (CSE) tests (Fig. II.B.8), indicate a sub-stantial difference in both level and in phenomenological behavior.
In addition, results showing impact of many nodes vs. one node for ER analysis was not presented. NUSCO should provide the basis to support its conclusion stated in Page 25 of the topical report that one node j
is adequate for SG modeling.- Furthermore, the RETRAN dynamic slip results indicate alternating vertical void fraction profile l
- =
.4 (Fig.II.B.11)symptomaticofflowregimemapproblemsaswere observed in early versions of RELAP5. Explain how this comparison 1
shows that RETRAN is acceptable.
- 11. Explain the trend in Fig. III.A.8 as compared with Fig. III.A.9.
It is not clear why the pressure computed with an infinite bubble rise velocity falls between the pressure computed with bubble rise velocities of 0.5 ft/s and 3 ft/s.
12.
Figure III.A.10 shows that there is substantial differences in the steam generator pressure transient obtained by RETRAN and NULAPS.
Provide an explanation for this deviation.
13.
Ordinarily for a steam line break analysis, it is assumed that adequate phase separation takes place in the steam generator to ensure pure steam exiting into the steam lines, and thereby provide a conservative analysis in terms of cooldown. Present the data which demonstrates that the separators will be flooded (p. 64) and discuss the impact of two phase flow exiting into the steam lines on the transient results.
l 14
'Since RETRAN gave phenomenologically very different results with the dynamic slip model than it did with the BR model, and both of these appear to be qualitatively different from NULAP, which model is appropriate and why?
-.~.-
i.
1_._ __ a.a.. _ _
w--._.c_
.n.
+
e 15.
Are the CSE tests being used as a basis for the statement that both the dynamic slip and bubble rise velocity with I ft/s 4v 43 ft/s are " adequate" for SG blowdown during SLB? Provide the basis to supportthisconclusion(page66).
16.
(Fig. III.B.1 & 2) Explain why primary pressure and temperature decrease fastei nher a finite BR velocity as opposed to the case with an infinite BR velocity. How does this fact correlate with the secondary pressure which comes down slower for finite BR velocities?
.One would expect two-phase blowdown could bring mass down faster because it depletes inventory faster, but single phase steain blowdown ordinarily tends to cause more rapid energy removal and therefore a faster temperaturo decrease. This trade-off is unclear but results of Fig. III.B.1 & 2 seem inconsistent with III.B.3.
17.
Explain why small BR velocity is conservative for small SLB and why an infinite BR velocity is conservative for large SLB's. See Page 87 of this topical report for details.
l L
18.
Explain Fig. III.B.11, provide sequence of events, and explain why trends are so different between these cases analyzed. Discuss why the 1 ft/sec BR velocity inserts positive reactivity while the i
infinite RB velocity inserts negative reactivity.
19.
For the hot zero power (HZP) case it is stated that a faster RCS depressurization is obtained with finite BR velocities than with
.~:., -.
y,.
a
. infinite BR velocity. However, SG depressurization is faster with infinite BR velocity than with the finite BR velocity as seen for the hot full pcwer (HFP) case (see p. 86, and also Fig. III.B.19).
If SG depressurization is slower, one would expect the primary to depressurize slower. Explain why the results do not indicate this behavior.
20.
Onp.88itisstatedthat"blowdownisconsiderablyextended...[and]
a lower minimum RCS pressure is obtained for the infinite case."
Does.this not give greater positive moderator reactivity feedback and therefore, is this not a worse case? Is the bubble ris'e model used for the steam line break calculation realistic or a reasonable approximation of expected secondary side behavior? Justify the statement that " unrealistically large bubble rise velocities will significantly alter plant behavior and should not be used" (p. 91) for licensing calculations.
21.
Smaller SLB - explain why subsequent steady states are independent of l
break size. Does the mixture level reach the break? If it did, it is expected that the two-phase break flow would have a different net energy removal rate from the single-phase break flow, and therefore, would reach a different subsequent steady state.
22.
Explain the impact of the transient initialization (Table III.B.2) with different SG inventories as a function of BR velocity upon the cooldown curves shown in Figures III.B.1 through III.B.16. Why does
- - ~ - ~ -
. not the 25% difference in the initial inventory cause significant difference in the results of the steam line break transient particu-larly for the large steam line break case? How do these different initializations impact the transient.
23.
Explainwhatismeantby"realisticheatslabheights"(page91)and how the choicc influences the transient.
24.
What are the minimum and maximum heat transfer rates between primary to secondary calculated for throughout each sensitivity study and are they within the ranges of validity or approval granted 'to RETRAN02-UGRA?
25.
On page 164 of this topical report, the stated purpose for RETRAN comparison with the FDSA transients is to " demonstrate NUSCO's under-standing of licensing analysis of design basis events and the capability to reproduce the design basis analysis with RETRAN02."
Therefore, more thorough explanations and analyses should be given.
The statement that the failure to agree well with the FDSA analysis results after peak of total steam flow (page 193) "may be due to differences in modeling the secondary system does not provide sufficient basis for the staff evaluation.
To support these objectives, (i) explain, on the basis of detailed references to modeling differences and cross reference to various
b:
e
. parameters, the reasons the RETRAN analysis does not agree with the FDSA analysis; and (ii) either (a) alter the RETRAN modeling to match the FDSA result or (b) justify the RETRAN modeling as being more accurate or more acceptable from a licensing perspective.
26.
During a main steam line break transient, injection of borated water was assumed. Has analysis been performed to justify use of the user specified boron transport model? If so, provide the results, if not, provide justification for the use of this model.
27.
For a main steam line break transient, NUSCO used 103 ft/se'c BR velocity to simulate separation. Did the analysis in the FDSA model the separator? If so, what are the results regarding the steam cenerator pressure and water inventory during a MSLB transient? How does this correlate with the statement on p. 64 that separators do not separate if flooded? Explain why with the higher steam flow the FDSA did net show continuously higher power?
l 28.
Figure IV.E.8 shows that there is substanstial difference in the results obtained from RETRAN and the FDSA analysis in the average l
heat flux during a MSLB transient. Explain why.
l P
1
r 9-D.
Reactor Vessel Head Modeling i
29.
Comparing to data at the plant, NUSCO found that the upper head circulation flow rate is between 5 and 10 lb/s and is clearly a non-linear relationship. Explain why 1-2 lb/s was felt to be an appropriate flow (p.136) in comparing to WFLASH but approximately 10 lb/s was required (p. 139) to get close to data?
30.
When the code computes flow in the upper head it gets about 2 lb/s.
This disagrees with data, but is conservative. However, with pumps running flow in the upperhead should be 50 lb/sec, but only 40 lb/s was obtained with the code. Explain why the code initialized at 40 lb/s and what upper head flow result would be obtained after pump trip if it was initialized at 50 lb/s.
31.
If RETRAN is initialized with valves to isolate the RV head, what transient is induced by* simply opening the valves and how is this effectseperatedfromthgactualtransient? Why was the use of such valves necessary?
E, HNP Natural Circulation Test of 1981 1 's i-
't
- 32.
What caused the oscillation in the RETRAN computation in Fig. V.D.2?
?
y
.?
O t-
. F.
Loss of Load 33.
What led NUSCO to believe steam line nodalization and temperature transport were important independent variables for this transient?
Why is not SG U-tube region heat transfer more important? Why is not SG relief valve two-phase flow modeling important?
34.
What is the physical significance and accuracy of the reasoning behind modeling the use of pressurizer sprays assuming the spray does not de-superheat the steam? What does the spray do-if it does not de-superheat? What is the range of validity of th'e spray rainout model and was this computation within the range?
G.
Haddam Neck - FDSA Comparison For Loss of Load Transient 35.
What core flows and reactor high pressure temperature setpoint were used in FDSA analysis for loss of load? Was charging and letdown used in FDSA? Why did the pressure in the FDSA analyses (Fig. IV.D.1) not begin to increase for approximately 5 seconds while the pressure predicted by RETRAN begins increasing immediately?
.a..
-.-......w..a..
-- - =
36.
If RETRAN predicts peak loop average temperature to be 6*F less than FDSA peak, why is the pressurizer pressure peak only 22 psi lower and the pressurizer volume only 40 ft8 different? A simplified calculation could be used to demonstrate consistency of these numbers or explain.the apparent lack of consistency.
37.
The loss of load is a secondary dominated transient. Present and discuss the secondary side behavior predicted by RETRAN (pressure, temperature, flows,mixturelevel)andcomparetothe FDSA results. Also, compare primary to secondary heat transfer predictions and discuss whether or not the two-phase transf'er correlations used in RETRAN were used within the range for which approval wes given to UGRA.
H.
HNP 30% Load Rejection i
38.
This comparison shows good agreement. However, this comparison is based upon a spray flow which was input rather than computed.
Justify the use of 500*F spray temperature and 200 gpm flow rate on the basis of actual plant heat lost from the spray line and actual as designed plant spray line friction losses and delta-P.
(
What flows would be computed based on plant friction losses and the plant delta-P? What spray flow would actually be computed by RETRAN if NUSCO used a volume with junctions to model the spray line, instead of a pair of fill junctions with forced flow?
......~.w..-..-.....
c 12 -
1 39.
What impact would the computed spray flow from Question 40 have on the computed pressure vs time?
I.
Loss of Feedwater 40.
The loss of feedwater transient is dominated by the secondary side behavior. Present plots of secondary mixture level, pressure and temperature with respect to time and correlate the secondary behavior to the primary behavior.
4 41.
Since changing primary nodalization in the SG made a substa~ntial difference in natural circulation rate, explain what would happen if a corresponding change were made in secondary nodalization. Justify the adequacy of the secondary nodalization which is used. The statement was made that 6 primary steam generator nodes would be used to ensure a conservative DNB calculation. Discuss the impact of similar change on the secondary noding, since DNB would be non-conservatively predicted by higher natural circulation flow rate.
l l
42.
Is the secondary mixture boiled up to the relief valves? Explain differences between boil up expected in this transient from that expected in a MSLB.
l 43.
What secondary flow area was varied? If it was the relief valve area, explain why it has only a 3 second impact and discuss the l
l other impacts on the transient.
.._......a
- - ~ -
,7,...~._,.
. 44.
For the partial loss of feedwater transient, explain why the plant data stops at different times for different parameters (60 see for pressurizer pressure, 80 sec of pressurizer level but 560 sec for LoopT,y,). If available, plot the balance of data out to 500 sec.
J.
Loss of Flow 45.
What caused the changes of slope at approximately 3 and 4 seconds in core power vs time in Fig. II.E.17 46.
What junction inertias were varied for this transient?
K.
Dropped Rod 47.
There is substantial difference in the results obtained from RETRAN and FDSA in the core power transient during a dropped rod event.
l Explain why.
48.
Present plots of primary pressure and temperature (core inlet),
secondary pressure and temperature, reactivity and power vs time for all the sensitivity studies and discuss the differences in those plots.
49.
What conservative assumptions does NUSCO intend to make for design basis analysis as a result of these sensitivity studies.
i
- 3
- ,
o.
L.
HNP Rod Rejection - FDSA Comparison 50.
Although modeling differences are briefly discussed, we cannot evaluate accuracy or understanding without a more thorough dis-cussion. What are the details of the differences between FDSA and 1
RETRAN modeling of doppler reactivity feedback and moderator density reactivity feedback that cause the factor of two differences in results plotted in Figure IV.B.2? What is NUSCO's analysis of the computations and what conclusion does HUSCO reach about the accuracy of the RETRAN and FDSA modeling 1.e., which is correct and why?
M.
HNP Inadvertent ECCS Actuation 51.
What plant data indicated non-equilibrium pressurizer behavior l
and how much?
52.
What quantification does Ref. 9 give to the interfacial heat l
transfer in the pressurizer?
53.
If the RETRAN non-equilibrium model is conservative on outsurge, is it non-conservative on insurge? Discuss the limitations that are applied to the RETRAN code documented in this topical report and demonstrate that this RETRAN code was used within its range of applicability for Haddam Neck plant.
-