ML20049A526
| ML20049A526 | |
| Person / Time | |
|---|---|
| Site: | Oconee  |
| Issue date: | 02/27/1980 |
| From: | Basdekas D NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Murley T NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| Shared Package | |
| ML20049A527 | List: |
| References | |
| REF-GTECI-A-49, REF-GTECI-RV NUDOCS 8104290139 | |
| Download: ML20049A526 (1) | |
Text
_
ENCLOSURE 4 UrdITED ETtTES g* ** M cy.fg
!!UCLEAR REGULATORY COMMisstON 4R[l
..m E
WiSHirtCTDN, D. C. 20555 b
FEB 2 71939 iio...
l HEMORANDUM FOR: Thomas E. Hurley, Director Division of Reactor Safety Research, RES FROM:
Demetrios L. Basdekas Experimental Fast Reactor Snfety Research, RES
~
SUBJECT:
FAILURE OF HAIN FEEDWATER' CONTROL SYSTEM RESULTING IN UNACCEPT-ABLE OVERC00 LING OF REACTOR VESSEL Failure of the Main Feedwater Con' trol System ~to reduce feedwater flow l
following a reactor. trip will produce a severely rapid overcooling of the primary coolant system in a PWR resulting in, among other things, overstressing of the fully pressurized primary system, including the reactor vessel and steam generators. '
l Assuming that this transient continues for several minutes, or longer, integrity of the primary system boundary and of the reactor vessel in particular will be subjected to an unacceptable and possibly catastrophic pressurized therTnal shock.
I don't believe it is necessary. to take this event sequence any further.
I recommend that the subject matter be analyzed carefully and with a priority commensurate to its associated high risk.
HUREG-0611 " Generic Evaluation of F/W Transients and Small Break LOCAs in Westinghouse-Designed Operating Plants", for instance, does not address this event sequence.
l Oconee-1 may'be an.approprigte plant for initial analysis.' A neutron fluence corresponding to fuci power operation through 1982 would be proper.
N d 5&2 fW/&,
Demetrios L. Basdekas Experimental Fast Reactor Safety
' Research, RES
, #, dh cc:
R. Budnitz, RES g
\\g.
- , O G. Sauter, OCH bg
,c-
/
w
, _ ~
b
...)$
E 5 D
UNITED STATES L,
['
NUCLEAR REGULATORY COMMISSION g
g p
WASHINGTON, D. C. 20555 R* *.... /
+
October 8,1981 MEMORANDUM FOR:
Chairman Palladino FROM:
Wil1iam J. Dircks Executive Director"for Operations
SUBJECT:
DIFFERING PROFESSIONAL OPINION OF D. BASDEKAS ON PRESSURIZED THERMAL SHOCK At your meeting on September 17, 1981, with Mr. D. Basdekas, he discussed with yod a series of his concerns related to pressurized thermal shock of PRR pressure vessels. After that meeting, you asked Dr. Denwood Ross to provide staff views on the points raised by Mr. Basc'ekas.
In response to your request, responses as provided by the NRR and RES staffs are enclosed.
l
' William J. Dircks Executive Director for Operations
Enclosure:
Answers to Mr. Basdekas' Concerns cc:
Comissioner Gilinsky Comissioner Bradford Comissioner Ahearne Comissioner Roberts D. Basdekas, RES R. Minogue, RES H. Denton, NRR s
j'b{)k SECY OGC q
0
.0PE y
/'
3 f b, 4 /
U,p/
t K
u p
yy gt h
COMMENT 1.
"Some of the steps taken or proposed by the Staff may be necessary, but they are not sufficient to provide an acceptable level of protection to public health and safety.
Contrary to the staff's position, this matter is urgent.
It is also more-extensive than the Staff states it to be."
RESPONSE
The staff regards pressurized thermal shock of pressure vessels as a very important safety issue and believes that the actions taken and planned are appropriate on the basis of our current knowledge.
In the judgment of the staff, there is sufficient time to evaluate the information to be submitted in response to the letter to eight licensees and in the PWR Owners Groups' generic studies before deciding what further[ regulatory action ~is needed.
' The staff report to the Commission, SECY-81-286A (dated September 8,1981) con-veys a sense of urgency.
A complete reading of Enclosure 1 to SECY-81-286A, which is. the minutes of meetings held July 28-30, 1981, indicates that the staff position is for prompt, positive actions to prevent potentially damaging transients or mitigate their effects.
4 9
e
)
COMMENT 2.
The precursory operational experience on pressurized thennal shock has been noted, but not heeded by the Staff.
The probability of severe overcooling accident sequences (particularly outside the design basis envelope) are substantially greater than those given by the Staff.
The industry's " bounding" cases are based on design basis accidents.
RESP 0NSE.
The staff has made a study of PWR ope' rational experience to look for severe.
overcooling transients and their precursors.
In 1980, the RES staff reviewed the operating experience of B&W plants and found that there had been a number of overcooling transients in B&W plants.
The most serious transient was that at Rancho Seco on March 20, 1978, in which the coolant temperature dropped from 550*F to 280*F in about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> while the system pressure first dropped, then returned to near its original value.
Based on this experience, a probability
~
of 3 x 10-2/ reactor-year was estimated for a B&W plant to experience an over-cooling -transient as severe or more severe than the Rancho Seco event was estimated.
In addition, B&W plants have experienced several similar, but less severe transients such as occurred at Oconee-3 in November 1979 and Crystal River-3 in February 1980.
These and other small er transients that occurred in B&W plants lead to an estimate of 5 x 10-I small transients per year in B&W plants as described in M. A. Taylor's memorandum of October 29,1980.
Since these occurrences, operators have received special training in transient response.
Babcock and Wilcox plants have added a back-up power supply to the nonnuclear instrumentation bus, whose failure initiated the three transients above.
The NRR staff examined the impact of the improved power supply and operator training and suggested that these improvements might have reduced the
-3 probability to 10 / reactor year for an overcooling transient as severe as the Rancho Seco event for B&W plants.
C0!P.ENT 2 (continued)_
s +
The operating experience of CE and Westinghouse plants has also been examined.
There have been no events like the Rancho Seco transient, but there have been some precursors.
These are events which typically led to secondary steam dump valves or steam bypass valves sticking open, but which did not result in steam flows large enough to produce very severe overcooling transients.
The most severe of these transients occurred at Arkansas Nuclear One-2 (a CE plant) on December 27, 1978, where a main steam relief valve Iifted and failed to reset, thereby causing the reactor co'olant temperature to drop by 107*F in 52 minutes.
Based on this e~ perience, the staff estimates the probability of a severe over-x cooling transient in a CE or W plant due to a large steam line break or its
-4 equivalent to be no greater than about 10 / reactor year.
In summary, the staff estimates the probability of a severe overcooling transient during which the primary pressure remains high is about 10,3/ reactor year for B&W
-4 plants and 10 / reactor year for CE and Westinghouse plants.
There may be a factor of 10 uncertainty associated with these estimates.
Although some of the PWR Owners Groups' calculations have been based on design l
I basis ac~cidents, we have asked for analyses of a wider range of overcooling transients.
These analyses should cover transients 'resulting from multiple failures or operator errors.
e
~
C0!NENT 3.
The uncertainties in critical parameter values are substantial uncertainties far greater than those given 4
culminating in RTNDT by the staff.
Dumestic and foreign experience indicates a trend for higher than estimated RTilDT
- S RESPONSE.
The staff is aware that there are uncertainties in estimating RTNDT, principally from the following sources:
(1) uncertainties in the initial RTNDTI (2) uncertainties ir copper content of the weld metal; (3) uncertainties in' vessel fluence estimates; (4) uncertainties in irradiation temperature; shift to (5) uncertainties in the Regulatory Guide 1.99 curves relating RTNDT fluence and copper c ntent of welds In calculating RTriDT, the uncertainties have been accounted for by using con-This is servative estimates of the factors that are used in calculating RTHDT.
shift based on and the prediction of RTNDT especially true of the initial RTNDT Regulatory Guide 1.99.
Vessel fluence calculations will be carried out with a well benchmarked and calibrated neutron transport code to yield fluence accuracy not worse than 120%,
The copper and the remaining uncertainty will have only a small effect on RTNDT.
content of the weld metal is believed to be known to within 10.03% copper (based on measured data) and this uncertainty will have only a small effect on RTNDT*
shift from sur-There is a substantial amount of information from measured RTNDT veillance tests of welds which shows that the curves in Regulatory Guide 1.99 for shifts are conservative for shifts above 150*F.
Based on this predicting RTHDT l
information, the staff believes that there are substantial conservatisms in the f r shifts above 150*F.
. predicted values for RT!iDT
C0l1 MENT 4.
The scenario related thermohydrodynamic assumptions are not realistically conservative.
The designation of the Rancho Seco event as the bounding benchmark for I&C systems initiated overcooling transients is not realistic.
RESPONSE.
The staff is evaluating a wide range of possible overcooling transients, as presented by the licensees as well as those done by us*, and will assess their '
expected frequency as well as severity.
Until this evaluation is completed, the staff has selected the March 20, 1978, Rancho Seco transient as a bench-mark for use in fracture mechanics calculations of pressure vessels.
The
~
Rancho Seco transient is not intended to be a bounding overcooling transient.
, The final choice of overcooling transients to use as benchmarks for fracture
' mechanics calculations will be made on the basis of the staff's judgment of the probability of the transients occurring and their severity of overcooling with appropriate consideration of the uncertainties associated with such estimates.
- The staff uses several system analysis models to predict the temperature and pressure histories of a primary cooling. system in a transient state.
_ COMMENT 5.
"A request for design information on control systems to selected utilities was blocked by HRR management."
RESPONSE.
Mr. Basdekas recommended that the following information be requested in the August 21, 1981, letter to eight licensees:
" design descriptions, including system functional block diagrans and single-line schematics of the plant's control systems, and the associated P&ID's and system flow diagrams."
Mr. Basdekas believed that this information was required for the staff to perform a complete. analysis of the potential for pressurized thermal shock incidents occurring at these plants and also to provide some of the data for a research contract to investigate the safety significance of control system failures for which he is the IUhC contract monitor.
During development of the final version of the letters, NRR Managers concluded that this request for control system design details, which are in excess of those normally reviewed in the licensing process, was not appropriate at this time, in the context of the letters on pressurized thermal shock.
(The letters do request the licensees to " provide any failure modes and effects analyses of control systems currently available or reference any such analyses already submitted...")
After the staff has reviewed the information to be submitted in response to the August 21 letters, and the reports of the Owners Groups due by the end of the year, a decision will be made regarding the depth of staff review of control systems needed to resolve the issue of pressurized thermal shock.
As to obtaining this type of information for the research contract to investigate the safety significance of control system failures, a different, more appropriate method to obtain this information will be utilized.
~
l-It should also be noted that a Task Action Plan is being developed for Unresolved Safety Issue A-47, " Safety Implications of Control Systems."
Consideration is being given in the development of that plan to the identification of control l
l system failures that can contribute to reactor vessel overcooling transients, and to the' development of criteria for plant-specific reviews of control systems.
Mr. Basdekas' comments have been, and will continue to be, requested and con-sidered in the development of the plan.
8 4
k 4
e M
e I
l '.
O O
e I. ~._
COMMENT 6.
The significance of the synergistic effect of Nickel 'on the rate of increase of RT is not accounted for in Regulatory Guide 1.99.
HDT RESPONSE.
1 The chemical content (including nickel)is known for all specimens from which data were derived and used in generating the curves in Regulatory Guide 1.99.
Thus, the effect of nickel on the RT shift is accounted for in Regulatory NDT Guide 1.99.
Past and recent information from measured RT shifts for sur-NDT veillance tests of weld material shows that the Regulatory Guide 1.99 curves for predicting RT shifts are conservative for materials with high and low NDT nickel content for shifts above 150*F.
9 e
e
COMMENT 7.
One assumption on which the staff based the in-vessel materials surveillance program was that.the welds were not going to be the critical elements for embrittlement considerations.
Hence, the circumferential placement of sample capsules were designed with that in mind.
It turns out that in most instances, welds are the data are critical elements.
Furthermore, fluence and RTNDT obtained in cycles of 5' 6 years.
RESPONSE.
Surveillance capsules which consist of weld, heat affected zone and base mate-rials, are placed in reactor vessels to provide lead time information on RT NDT It is
' shifts and to. provide data to benchmark heutron fluence calculations.
not necessary that the surveillance capsules be placed at the weld locations, since the data obtained can be extrapolated to the correct vessel locations by means of calculations.
In fact, locating the surveillance capsules at the
~ longit'bdinal welds would interfere with inservice inspection of the welds, and would actually increase the rate of'e' brittlement of the weld because there m
is a peak in the fast neutron flux just behind the capsule on the inside wall of the vessel.
With respect to timing of surveillance capsule examination, a capsule is Rules for withdrawal typically pulled at the first or 'second refueling.
schedules are given in 10 CFR 50, Appendix H.
s
i COMMENT 8.
TherE are important espects of the foreign experience, concerns,
~'
and measures, taken on this problem that need to be examined carefully.
RESPONSE.
In the context of this answer, foreign experience refers to overcooling transients at foreign reactors.
The concerns ar.e, of course, the same:
high thermal stress; undercalculation of fast fluence; loss of fracture toughness due to irradiation.
The countermeasures are the same as being considered in the US:
reduce fluence, reduce challenge; restore and maintain ductility.
i The RES ahd NRR staffs have had regular contact with foreign experts on pressure l
I vessel integrity for years.
An extensive set of questions relating to pressure vessel thermal shock has been sent to foreign regulatory authorities, and the staff has reviewed the responses received to date.
As an example of the type of discussions with foreign experts, the staff. met with representatives of the German Ministry of the Interior (BMI) and the Reactor Safety Commission (RSK) in Bethesda on September 29, 1981, on this subject.
l
C0!'JiENT 9.
The in situ annealing capability requirement (Appendix G IV-C) is not met by many PWRs.
RESPONSE.
The 10 CFR 50 (Appendix G, IV-C) requires that the reactor pressure vessel be designed to permit in-place annealing to recover material fracture toughness properties if calculations predict that neutron irradiation may increase the RT to 200*F or more before the end of plant life.
For licenses issued HDT after the effective date of this rule (August 16, 1973), licensees who predict end-of-life RTNDT > 200*F have asserted that they have the capability to anneal.
Since the' requirement was not imposed until August 16, 1973, some older plants have never been asked'to respond to the requirement of Appendix G.
The staff judgment, isowever, is that the similarity in PWR designs is such that there should be no design aspects that would preclude in, situ RPV annealing in any
~
plant. -
The eight licensees
- have been specifically asked to provide the basis for demonstrating that their plants meet the requirements in 10 CFR 50, Appendix G, -
IV-C.
When these responses are received in January 1982, the staff will have a better basis for assessing the capabilities of these PWRs for in_ situ annealing.
"TMI-1, Ft. Calhoun, Robinson 2, San Onofre 1,14aine Yankee, Oconee 1, Turkey Point 4, Calvert Cliffs 1
8 9
F C0!'J1EliT 10.
The RTilDT mentioned possible operational limit of 300*-F is very question 6ble.
~
RESPONSE.,
The staff has considered the possibility of establishing an upper limit on RT f r operating PWR pressure vessels.
Although the value of 300*F has riDT been mentioned as an example, many mo're analyses of transient frequency and j
severity are needed before a limit can be established.
Furthermore, if an upper limit on RT is to be imposed, it may well turn out that the limit NDT will have to be a plant-specific limit.
e M
O I
O m
a M-
a Not all reasonable options appear to have been considered by
~
COMMENT 11.
the staff seriously.
The impact of possible shutdowns must be determined if it_ has not already.
RESPONSE.
The staff has concluded on the basis of current information that corrective action is not necessary at this time.
The actions the staff has initiated and the information we have requested from licensees and owners groups are intended to lead to decisions on any required corrective actions by the summer of 1982, or earlier if indicated.
The option of plant shutdowns is always available should the staff judge such actions are necessary to protect the public health and safety.
War 9
9 O
e o
4 s,
[.
COMMENT 12.
For all the above, and in relation to Item No.1 on this list, it is recommended that an ad hoc group, including experts out-side NRC, be charged to study this matter and report to the Commission with short-and long-term recommendations for dealing with it.
In the interim, the Commission may consider all reason-able options available to it to assure.an acceptable level of protection of public health and safety, something the staff measures;do not provide. -
RESPONSE.
- Experts outside the NRC have been assisting the staff on the subject of pressure vessel integrity.for a number of years.
Major contributions have been made by experts from ORNL, NRL, US Naval Academy, Naval Ship R&D Center, University of Maryland,.BCL, HEDL, and NBS. We have also had critical contributions from researchers in Belgium, Germany, France and En5 and.
1 Our long-standing practice of involving the most knowledgeable experts continues
'n this case.
Of course, it is the staff that must be responsible for ultimate i
~
decisions.
We intend to discuss the conclusions with outside experts and with the ACRS before final disposition by the Commission.
i 9
O O
1 4
=
I
-e