ML19322C261
| ML19322C261 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 03/13/1975 |
| From: | Hanauer S NRC COMMISSION (OCM) |
| To: | Gilinsky V NRC COMMISSION (OCM) |
| References | |
| TASK-TF, TASK-TMR NUDOCS 8001160707 | |
| Download: ML19322C261 (8) | |
Text
-
n
,9 '.
~
UNITED STATES NUCLEAR REGULATORY COMMisslON I
g WASHINGTON. D. C. 20555 t
March 13, 1975 Comissioner Gilinsky Thru: Acting Executive Director for Operations
'O' e
TECHNICAL ISSUES Attached you will find, in accordance with your oral request, discussion of some technical issues I believe to be important subjects for Commission consideration, although not necessarily in the imediate future.
The list is confined to reactor safety topics.-
I have also appended a list of some reactor safety policy issues that have come to my attention in technical reviews.
These enclosures represent my personal views and have not been staffed out with the organizations normally concerned with such matters.
d
~
d, h
r Stephen H. Hanauer Technical Advisor -
Encis
- 1. Technical Issues
- 2. Policy Issues cc: w/enci Chairman Anders Commissioner Kennedy Comissioner Mason Commissioner Rowden L.V.;Gossick E. Case H. Kouts F. Schroeder A. Giambusso R. Misiogue po'C'O+\\.
i
't h
goonsa p7 p
t'
_(,
f
?
og s
IMPORTANT TECHNICAL REACTOR SAFETY ISSUES FACING THE COMMISSION NOW
~
OR IN THE NEAR FUTURE 1.
Design Objectives and Safety Design Basis for Water Reactors Although your mother-in-law and your Congressman will tell you that the safety goal is zero risk, we know that this is unattainable-and that some non-zero risk must be accepted in all activities. The social question e
involying cost / risk / benefit comparisons of the various alternatives that are realistically available needs t6 be established.
The Rasmussen Study made an important first step in quantitative risk evaluation but the technology is not >yet available to resolve this question in a completely quantitative way. The study has pointed out a disparity between (a) our present " design basis" safety approach in which all potential accidents are either put into the design basis for complete mitigation or remain outside the design basis and have no safeguards compared to (b) the more realistic viewpoint of a spectrum of accidents each with probability and consequences of its own.
Serious consideration sho.uld be given to modifying the present all-or-nothing approach in the light of reality.
~
2.
Design Objectives and Safety Design Basis for Non-Water Reactors
' For non-water reactors, we have neither the operating experience nor the Safety Study to guide us.in developing criteria. The situation is reasonably wall in hand for HTGRs, but the potential for autocatalytic positive feedback leading to core nuclear explosions in LMFBRs is creating great uncertainty regarding their design requirements.
Calculations of such violent events are increasing in scope and sophistication.
- However, the results presently depend to a considerable extent on the phenomena i
postulated to occur.
For the near term, the staff has already decided that a core disassembly accident must be part of the licensing design l
basis. This decision is subject to futyre revision based on further research that ERDA is convinced will show that such events are so improbable they nu.d not be considered.
j Adequate safety must be'provided. Too much safety - added safety equipment not actually needed to provide adequate safety - wastes scarce and valuable resources. Attention to improbable severe postulated events tends to short-change more probable but less severe accidents that should be considered.
An important corollary issue is whether the planned LMFBR safety research programs' meet the totality of NRC needs.
O P
e w-
~
U
~
. ~. '."
- N 2-3.
Reliability and the Single Failure Criteria NRC has not established quantitative reliability criteria for safety-related systems. The operating plants are one of our chief sources of information but we do not know whether the rate of abnormal occurrences now being experienced is a satisfactory one or not.
We do know that nuclear unit availabilities and capacities are not satis' factory.
We need i
to find out whether safety system availability is satisfactory and to
~
improve whatever aspects of reliability need improving.
4.
Human Perfomance Present desiDns'do not make adequate provision for the limitations of people. Means must be found to improve the perfomance of the people on whom we depend and to improve the design of equipment so that it is less independent on human perfomance.
The potential for internal and external sabotage constituting a public safetyhazardJ;ndthedegreetowhichdesignandoperationneedstotake sabotage into account, need to be delineated. Studies now underway should help, but some of the issues are non-technic-1 In spite of this difficulty, technical criteria are needed.
The relative roles of human operation and automation (both with and withot.t cin-line computers) should be clarified.
Criteria are needed regarding allowatle computerized safety-related functions and computer hardware and software requirements for safety-related applicat, ions.
Plutonium Dose Criteria Present accident dose guidelines values are given only for whole-body and thyroid doses.
Otherdosecomponen,ts(lung,GItract, bone)shouldbe covered by similar guidelines. A number (or numbers) for plutonium is particularly badly,needed and will be particularly hard to establish.
6.
Siting Present criteria for siti-ng are in need of improvement in the following areas:
The design basis external event's now in use for. licensing are foundeo a.
on various schemes for estimating a " probable maximum" event. We do not have any good ;way of estimating the return interval or the frequency of the earthquake or flood calculated in this way.
Furthemore wh are not likely to develcip good methods for doing so in the near future because of the short l
y e
wm -~
- W
)
B
.~
~3-history (a few hundred years at best) and the long recurrence interval desired
~
(sometimes we talk about a million years). Various developmental methods for estimating frequencies of design basis events, chosen as we choose them, give recurrence intervals substantially shorter than a million years.
The
. lack of knowledge and the desire.to be conservative is going to make resolution of this problem very. difficult.
b.
Our population siting criteria are indefinite at best. The applicant is required to study population distributions around a site and to project them for the life of the plant which, of course, he can do only very crudely, but our criterion for population ' distribution surrounding the plant are very vague.
Recent attempts to be more quantitative in this area met with great
~esistance from the industry and from the old AEC. They tend to be over-
- implified, but I believe we could do better than has been done. A related problem.is our present total lack of control over what goes in near the plant after' the site is approved. We have some vague words about the licensee's responsibility to stay informed about subdivisions, ammunition plants, U;G terminals and other post construction materialization of things that would have made the site unacceptable if known before licensing.
Someday some operating reactor is going to have a new neighbor of a really abominable kind and we are going to have trouble coping with it.
I believe we are not being serious enough about siting alternatives c.
that may offer substantial safety improvements.
An obvious example is underground siting about which we are just starting a study in RES.
7.
_ Degree of Detail and Realism in Safety Evaluations The great improvecent in computer codes available for use in analyzing the course and consequences of postulated accidents has rather naturally led to a corresponding increase in the depth and detail of Regulatory review of these accidents.
On the face of it this is a good thing.
It leads to better technical understanding and increased realism in evaluations.
But is overall safety review enhanced.by such detailed examination of certain design basis accidents?
It is at least arguable that a broad brush treatment, with plenty of arbitrary conservatisms, gives at least as much safety with a lot less work on everybody's part. A recent and obvious example is the new ECCS
. regulation, which specifies in gory detail exactly how these calculations-are to be mada. There are many arguments for and against use of such details and tha subject is abou.t right for reopening, in my opinion.
A related subject is the very large increase in the capability of the fiP4 s:aff to make independent calculations in many accident areas. This has proved to be invaluable in increasing the staff's technical
- understanding and should be continued even if some of the details are recognized as too datailed for licensing.
/
Pe
~.
m J
.e.
8, Funi Parformanco The performance of light water reactor fuel in normal service has been disappointing to say the least. One would have thought that by this time fuel technology.would be well developed. The appearance of such difficulties as densification, hydriding, hot pellets, and the recent incident at Dresden where a transient, well within all limits, resulted in unexpected fuel failures - all tell us that fuel technology is not in as good a' state as we thought. Yhe related technology of establishing fuel damage limits under accident conditions is even less well established, principally because PBF is so many years late.
3.
Pu Recycle This is not primarily a reactor problem. The reactor aspects seem to me to be adequately in hand.
de e
O O
e 9
g 4
e e
O 9
4 e
e 1
4 G
I) i W
e l
~
Q g
I
~
REACTOR SAFETY POLICY ISSUES 1.
Internal Quality Assurance
. We are not taking our own medicine with regard to a quality assurance Jrogram in Reg. We do not have a quality assurance organization, independent of the line, reporting to higher management and we have very little auditing and QA in the line.
If 10 CFR 50, Appendix B, is good stuff, then it 'should be applie' to the NRC organization. This must be applied to the quality i
d of our product - safety decisions as well as the quantity and timeliness
.cf our output.
~
- Y
~
~
~
?.
Making Better, Faster and More Generic Decisions Our recent record is mixed. A good example is ATWS and a bad example is turbine missiles, about which we seem not to be able to make up our minds.
1%ture technical safety review should not be endless and mindless repetition of what we have been doing for the past couple of years but rather consolidation into general decisions and general principles, better identification of what is truly im automation of routine evaluations.portant (risk evaluation?),. and i.ncreasing 3.
Stabilization of Regulation Requirements and Standardization of Designs Our recent reviews of the standardized designs that have been submitted and recent discussions on standardization (and piggy-back) show the following:
The standardization designs submitted are not consolidations ~ of a.
previous experience. The proposed standard designs include a large number of " improvements" not yet actually designed.
So, these first standard cps will be based on a bunch of promises, eve,n more than recent custom cps.
b.
New information from design and operating experience and safety research programs, and new insights as a result of this experience and research have pointed the way to im while and in some cases necessary. provements in safety that seem worth-The pace and guidelines of the standard reviews has not permitted implementation of these, so they are hanging over cur heads as a serious threat to standardization.
Asia result of a. and b. and of the long time lag between today's c.
lunch of promises and construction and operation of standard plants, more z.ttention needs to be paid to the execution of standardization over the next several years and stabilization of Reg requirements.
I Y
I f
--+
.o i
~
m n_
i i
~
2 4.
Too Many Surprises This is closely related to Item 3.
In the past couple of years surprises
.have come both from operating experience and from improved understanding by both Reg and the industry of safety problems we thought were put t6 bed.
.S.n obvions example is all the trouble we had with ECCS evaluation models.
Innovation by applicants will continue to generate surprises.
We must develop methods for daaling with these surprises, in cases and generically, without having a fire drill each time.
O e
e e
e 9
e
- +
.. ~.
4
Tksw pkTae 2Rdn Safety o'f BreederReactors Questioned Dr. Hannum added that the!
By DAVID Bt;RNHAM ii g ast breeder reactor's design l.was. Hanauer's memorandum f
Dr a aniern.p t-arw.
one of 6l reports -pla ed WASHINGTON. Feb.15-A was such that scientists in! in the commission's public docu-leading Government expert has his agency,were convincedl ment room last week after it
=
said that possible " core nuclear that " core dispersal could bee had been cited by Robert A.
- l explosions" in experimental re. : contained without violating the Pollard. a commission project actors that the Government ex ' seals of the reactor vessel" manager to support his arga.
8 pects to te built all over the. In a recent paper calling for racnt that the commissica " sun-8 United States "is creating great more research on this questinn-passes th existence M unt's.
l howeve uncertainty regarding thesr de, agency'r. four scientists at the:s Ims Alamos scientific l solved safety. Questions ar.d sign requirements" fa Is to resolve these protiems The discussion of, the possi. laboratory wrote:
bihty of such an ace: dent in the ' methods for analyz, "Present, prior to allowing reacto.s to mg hypo '
ep,,, g,,=
i ast breeder reactor. a source thetical core disruptive acci.:
Mr. Pottard has submitted his f
of power that the Fo-d Admin dents cannot show concia.
resignation to the commin.on-istration and most nuclear of. sively that such accidents rio to go to work for the Union of ficials feel is essential to the not lead to the Concerned Scientists. an organ-icontinued growth of nucleart preswevessel., rupture of the t
jer.ergy. was contained to a re -
Plu* onium is so toxic that' Ization opposed to present nu.
. port wntten last March 13 by the accidental release to the at.
- i'g*3 h;'*f'P*rt. Dr. Huaun
)
Dr. Stephen H. Hanaust, one m.m.here of several hun.tred discussd a number of ques.
of the most serdor technical grams of this material woul.d.
tions that have been rar ed re-experts on the staff of the he a raatter of serious concern.
Nucteer Regulatory Commis. the scayntysts said. Noting that
!c,"m issi n.
such s th e E'
s;on each liquid rnetal fast breeder "This would be nothing like re:ictor may contain more than l nuclear ofg,g;als og the v, enera; an atomic bomb but would in *2.000 pounds of plutomum, volve a vast release of energy." they argued that the release of Electric Company who quit tb Dr. Hanauer said, when asked as little as *0 0f percent of weeks ago to woric aga:nst
.J o comment on his report. en-the availab!e plutonium may be nuclear power irl Cal:fornia'eir t
- titled. "Im rtant ' Technical cause (or alarm."
The three men stressed th -
Reactor Sa y hsues Facmg Because the fast breeder is *oncern that nuclear reactors the Commission Now or in the; designed to create more f el Near Future" than it ba n. it is reen as es*
fa 1ure. "r. Han"aun ma h
Other experts. both favoring'sential 12 the I ng-term use of same point when he sa,de th v'r.
and opposed to nue: car energy. nudear reactors.
id th!
4
$a(ent reactor "desiens do noc Pre a
agreed that the k.ed of acci l *Without the breeders." Dr.
tions of, perp!e p epm e
dent referred to by Dr. Hanauer Hann fm said. "we can tap only li !
woutd not le2d ta the physicalI percent of the potential o'f
'.Means trust be found." he -
destruction of the area sur-uranium. and uraniurn thus said. "to improve the p+r.
!roundmg a fast breeder reactor..vould be eouivalent to our oil; formance 6f the people mi Such an accident however. etervet.which are runningout.
most agreed, cou*d lead to the.W::h the brefoer. the power whom we depend and,to i-n-sive contaimne-t ca;ac:ty and, prove the design of equiprnest breaching of the reactor's mas 9otential within existing ura.( iso it is less dependent on.ha.
phere of vapori:ed p:utoniurn.: niurn reserves ca,n be increased; man performance **
the injection into the atmoM He added that "the pote:-
A furthef concern was that 50 to 60 titres.
a tial for internal and external liquefied phteniurn ml::htsome e Dr. Hannum said the Gov-1satiotage constitutir'e a puh.
how reac:a underground sourcest:ernment beheved that within;:lic safety hazard, ar.d the de.
8'50 years the United States'; gree to which desi:n and m of drinking wa*er.
hwould be dotted with hun eration reed to take sabota :e Cancee Fears involved breeder reactors."*
account, need to be Extremery sma:1 arounts of,.creds ofi Ground levelmg for 'the first into delineated."
plutonium, the bas.c fuel ofi. demonstration model, of the: On the subject of 'where rh the fast breeder, have beenl-shown to cause 1.tng and other* fast. breeder reactor is sched. actors could be safely situate' uled to begm at a site on the;Dr. Hanauer said many of 15 types of cancef-Chnch River in Tennence latericommission's guidelin'es "a v Dr. Wilham Mannuro. a senior official in the Energy Research, this year.Dr. Hanauer,in his March 13. indefinite at best."
s
- We have sorne va2ue words to Victor Gitinsky. a bout the nuclear reactor li-and Development Admmatra. report a
tion, which as of June will have member of the, Nuclear Rega censee's respnnsibility in stav spent $2.69 bit!!nn developinit latory Commission. wrote thatinformed about subdaisioni the fast breeder, said in ank "the potential for autocatalyticammunition plants. 1:cuefred interuew that Dr. Hansuer's fcerthack leading toriatural gas terminals and other e
emplosion,e term "c, ore nuclear;. positivehyuid metal fpt breeder reac post constructin nuclear explocons in ue of th technically l : ore materiahra-was tion of things that cou:d mae irrorrect--we call it getic core disruption.,an ener j tors is creatmg great uncer a site unacceptable if krinwa
. tainty regard;ng their designhefore licensing" Dr. Hanauc.t requirements.
wrote.
"Some day some operati:4 reactor is gning to hase a nos
' neighbor of a realle abnen.
a Inable kind ard we are peq 1
have trouble cbpmg wat e
t
. Enclosure 2
,