ML19310A217
| ML19310A217 | |
| Person / Time | |
|---|---|
| Issue date: | 12/04/1979 |
| From: | Ross D NRC - TMI-2 BULLETINS & ORDERS TASK FORCE |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8006060296 | |
| Download: ML19310A217 (5) | |
Text
.
g" "'%g j
,1
,o UNITED STATES
+
l y
,#((gq NUCLEAR REGULATORY COMMISSION
- /
E WA$HING TON, D. C. 20555
. d g' f !
o\\14/
usa e
MEMORANDUM JOR:
H. Denton, Director, Office of Nuclear Reactor Regulation FROM:
D. Ross, Jr., Director, Bulletins & Orders Task Force
SUBJECT:
B&O REPORT ON W PLANTS I have completed my review of the subject report. During my review I concluded that several reconinendations would not be included in the final version of the report, for the reasons stated below:
1.
Conservatism in SBLOCA It was recomended that the NRC review of the conservatism in LOCA calculations be accelerated, with a position on SBLOCA by 6/80.
I took this out on grounds of relevance.
It belongs, if anywhere, in the NRC action plan, since it involves prioritization of resources.
I intend to delete this frc:n the other reports also.
2.
Loss of Feedwater (All)
It was recomended that either the capac'ty of the relief valves to provide depressurization in a case of a complete loss of feedwater ce demonstrated (with due account of input uncertainties and calculational uncertainties) or continued operation of the affected plants be conditioned on timely design changes.
I deleted this because:
(i) we have already required improvements in AFW; and, (ii)
I believe it needs more careful attention under USI.
S. Hanauer agrees with this position.
I intend to delete this from CE and B&W reports. The ability of relief valves to pass vapor or 2-0 mixtures is the subject of STLL.
3.
PORY Ooeration It was recomended that the likelihood of stuck-open PORV be reduced by a combination of items which are attached here as Appendix A.
I decided not to embrace these recomendations because:
(1) Through STLL there is emergency power for PORY and block valves, and ;:erformance testing for PORV, and direct positien indication for PCR~/;
scososo 23 b
H. Denton CCC dN (ii) Through B&O the plant procedures have been upgraded to account for stuck-open PORV; (iii) One opening mechanism (see Appendix B) is being eliminated; (iv) Safety classifications and qualifications are being upgraded through point 9 of LTLL; and, (v)
I have decided to include a new item which calls for auto-closure of block valve on low pressure (for all PWRs).
It is my judgment that these five items are sufficient and superior to Appendix A.
I embraced Appendix B.
I also deleted a related recommendation to do some detailed comparison of plant experience vs. analytical predictions.
If this is to be done at all it should be in an orderly manner under the topical report program 'or LOFTRAN.
4.
Small Break Methodoloay There were many deficiencies noted in the small-break methodology. Most of these stayed in; in context, they will reed attention during the 1980 Appendix X phase.
5.
Loads l
There was a concern about loads due to injection of cold water.
I asked for a staff judgment, absent conclusive data.
6.
Pump Seals The recomendation was that the worst failure of RCP seals be assumed with the worst small break.
I asked fer an "or" clause on the ability of the pump seals to survive during SBLOCA.
I believe this represents the important elements.
For historical purposes there will be available, in due course, a complete manuscript of Appendi-VIII (Analysis) as I receive it which may be compared with the final as-pri d version.
All of the changes that I made were discussed with the review team and there was no disagreement, except possibly from Dr. Rosztoczy, who may ultimately have a differing viewpoint.
There is, however, unanimity on the conclusion that:
The small break analysis methods used by W are satisfactory for the purpose of predicting trends in pTant behavior following a small LOCA. The results of the analyses can be used to develop improved emergency procedures and for : raining of reactor operators.
4N H. Denton cf.C As that was the principal charter of B&O, I believe thc: we can confidently close our task force effort on analysis. Further work can proceed under a different organization unit.
/
r
/
D.
. Rois, Jr., Director Bulletins & Orders Task Force cc: All B&O Task Force Members R. Mattson S. Hanauer ACRS (16)
E. Case Attac,tments :
Appendix A & B i
1 l
Apoendir, A (NotAdopted)
Recommendations :
a.
Frequent overpressure transients should not result in opening of the relief valves. Licensees of Westinghouse plants should:
(1) install an anticipatory reactor trip on turbine trip if it is not already present; (2) show that overpressure feedwater transients challenge the relief valves only in exceptional cases (less than 5% of transients).
If the latter point cannot be supported by the 150 years of operating experierce presently available, appropriate design changes should be introduced. Changes in relief valve and in high pressure reactor trip setpoints and the installatien of additional anticipatory reactor trips should be considered. The licensees must comply with this requirement by March 1,1980.
b.
Licensees of Westinghouse plants up to date reported approxima*ely 300 feedwater transients. The peak reactor system pressure reached during these transients and possible indication of relief valve opening during these transients should be reported to NRC prior to March 1,1980.
c.
All future relief valve challenges should be recorded and reported to NRC.
One possible way to completely eliminate tne risk associated with the failure of relief valve: is to operate the plants with the block valves closed. This mode of operation, however, could result in some increase in the lift frequency of one safety valve. The licensees so far have failed to provide infonnation on the observed failure rate of safet.v valves. Consequently, neither the desirability nor the acceptability of this node of operation can be evaluated at this time.
Recommendatiens:
(a) Each licensee should be required to submit a report on safety valve challenge rate and safety valve failure rate based on past history of the plant. These reports should be submitted to NRC prior to February 1, 1980.
(5) All future safety valve challenges should be recorded and reported to NRC.
Appendix B (Adopted)
The derivative control system, installed on at least one relief valve 'in most Westinghouse plants, has caused spurious valve actuations. This resulted in a considerable number of relief valve challenges. Westinghouse recently recommended the elimination of derivative control.
Recommendation:
The derivative control, if it is still in use, should be replaced with a constant pressure setpoint system. This change should be accomplished within 30 days from the issuance of this report.
Relief valves supplied by Contro? Components, Inc. were used, at the.first time, on the McGuire plant, owned by Duke Power. One of these valves failed during hot functional testing. Following failure the manufacturer of the valves recommended modifications to the valves.
Recommendation:
The McGuire relief valves should eitner be replaced with valves which have an operational data base, or should be tested under design conditions prior to startup of the plant to assure their reliability, 4
d l
2 -.m
..