ML20087M812
| ML20087M812 | |
| Person / Time | |
|---|---|
| Site: | Fort Saint Vrain  |
| Issue date: | 03/14/1984 |
| From: | Warembourg D PUBLIC SERVICE CO. OF COLORADO |
| To: | Jay Collins NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| References | |
| P-84084, NUDOCS 8404020022 | |
| Download: ML20087M812 (12) | |
Text
-
1 public service company of Cobrado S
ff 16805 WCR 19 1/2, Platteville, Colorado 80651 March 14, 1984 Fort St. Vrain Unit #1 P-84084 3)3@3DWMS o
Mr. John T. Collins g lglgg4 Regional Administrator U. S. Nuclear Regulatory Commission p
611 Ryan Plaza Dr., Suite 1000 4
Arlington, Texas 76011
SUBJECT:
Fort St. Vrain Unit No. 1 NRC Source Term Studies
REFERENCE:
P-84036
Dear Mr. Collins:
As a result of the above referenced letter we received inquiries from your staff that indicated they could not determine who in the NRC was working on -the source tera studies.
The attached articles from Inside NRC are enclosed for your information.
Very truly yours, ht~ W W Don W. Warembourg t/
Manager, Nuclear Production Fort St. Vrain Nuclear Generating. Station DWW/djc Attachments 8404020022 840314
~
PDR ADOCK 05000267 P
/g
1 W.
I INSIDE N.R.C.
elusive report on the U.S. Nuclear Regulatory Commission hee- - -.
BeIIeview Vol. 6, No. 2 - January 23.1984 y-M-% FSV IVERSION TO FOSSIL FUEL SEEN IMMINENT Diamond I" * ** d'" "5 I"* " D ** '"d I *h * "*"* *
- Hill nati Gas & Electric, has told the Cincinnati City Council. Other sources plans to complete it as a nuc! car unit and are working out the economics l"
Diamond aband ning it altogether.
Hill
.lled over rising cost estimates for finishing the plant. CG&E and its part.
l m a
,j smbus & Southern Ohio Electric,have already invested nearly St.7. billion l
u&va ilete. But Bechtel, brought in to complete construction after an NRC-mated St.2 billion to 51.8-billion more will be needed (Nucleonics Week, i
ting that estimate is too low and the Cincinnati City-Council is consider.
f ndon the plant as " clearly unaffordable" for ratepayers. The Ohio Con.
l nsultant Chasles Komanoff and MHB Associates to analyze the Bechtel
~ ffice representing utility customers before the Ohio Public Utilities Com.
l opposing charging ratepayers any costs attributable to mismanagement.
nmer costs. In addition, DP&L and C&SOE have forced CG&E into pri.
f*
osts and possible damages for alleged CG&E mismanagement.
&L and C&SOE met foe seven hours Jan.18 and all refused to speak to From Public AffaI'rs 620
>lanned. Any decisions will have to be approved by all three utilities' boards
- her sources said the decision not to continue with Zimmer as a nuclear Call Changes to X 7389 te now is centered on whether to convert the plant or abandon it. DP&L verting the plant to gas for peak demand.now and converting it to coallater.
The conversion icea a um. o~.; attractive, sources said, by an Ohio law that forbids utilities from recouping costs of a canceled plant from ratepayers.
Rumors of Zimmer's imminent demise were spurred by Public Service Indiana's decision last week to l
cancel the half. completed Marble Hill plant after PSI and a group of municipal cooperatives had spent nearly 52.9-billion (Nucleonics Week,19 Jan.1). PSI decided it could not afford to contmue building. Safety-related work at Zimmer has been stopped since November 1982,when the NRC commissioners ordered a halt until Zimmer management could be audited and revamped, and until an outside company could assess the quality of construction at the 810-Mw Mark II BWR. A management reorganization was approved by NRC last month, with NRC officials saying CG&E and Bechtel had brought in experienced personnel who could get Zimmer built correctly (Nucleonics Week. 22 Dec. '83. 4). - Margaret Ryan STAFF PROPOSES SOURCE-TERM RULE THAT WOULD COVER OPERATING PLANTS
_A source. term rule specifying_the fissionfroduct emission expected for_ severe accidents at all reactors t
now orentingor under_ construction._willbe.ptoposed kv NRf' <taff. The rule would state the maximum risk exeected from any plant. using data derived from six " reference clants." and the need for any backfit pro-l posed to reduce severe accident risks would be judged against the risks in the source-term rule.
That strategy is part of the latest draft of NRC's proposed severe accident policy,which would postpone consideration of measures to mitigate, manage or prevent severe accidents until the source term rule is en-acted. The draft sets out a decision making process by which the staff would weigh relative costs and benefits of possible backfits, using mostly traditional engineering analysis but also probabilistic nsk assessment (PRAl.
It specifies that decisions would be made using reference plants rather than requinng plant. specific analyses INSIDE THIS ISSUE...
'Early second guess'seen easing pipe switches
- p3 Cuomo aide sees no legal basis for I.ilco emergency plans - p7 Commissioners scrunnize research-embargo deals
- p4 Duke's top management gets a break on drdis
- p8 NRC orders Consumers' management appraised
- p4 Watske says nuclear future hinges on better management - p8 Pallacino and Gilinsky split on hydrogen rule
- p5 Aslab examines ' corporate character
- in TMI case
- pIO NRC handling of SES bonuses is criacized
- p6 NRC staff renews PRA's progress ard uncertamnes
- p12
i :.w,
' f,90
[@
st alleperating reactors h also pr: poses a " safety assurance program" for new standard design plants that j could be extended to plants already in operation. Any backfits the staff thmks are worthwhile would be en-W<.
acted through a rule.and, until severe accident decisions are made, the draft would ban Atomic Safety &
Licensms Boards from litigating contentions about a plant's response to severe accidents.
Robert Bernero head of the sourcederm revision effort. told an Advisory Committee on Reactor hfe.
- guards ( ACRS) subcommittag_at a Jan.1I briefina that his staff was considering a methodology..a set of tables, or both for the sourca-term rule. The tables could specify " acceptable results for tission product trans-port for use in plant-specific calculations." The table idea is complicated. he said.by the fact that different
)
tables are needed for each reactor and containment combination.and that fission product transport calcula-
- tions vary with time and accident sequence. As a result, his staffis considering some combiriation of tables
_i and methodologies which could be used for any olant-specific -tysis. In its final form. however, the rule
~
"will represent the source term for existing plants,' derived from the careful evaluation of severe accident risk characteristics for a representative set of those plants," the latest severe accident draft says.
1he stadt would withdraw the severe accident rab==aring proposed in 1980. because "it has become in-
- creasingly clear that the regulatory concepts in (the 1980 proposal) were undesirably amorphous and unfocused."
4 and modifications made since the Three Mile Island 2 accident have already reduced the risk of severe acci-
. dents. Utilities put tcgether the Industry Degraded Core Ru'emaking(Idcor) program in response to that 1980 -
proposal. and industry officials have been pressing NRC officials te return to the single rule concept (INRC, 26 Dec. '83,6). However, the new proposal also rejects the other extreme - plant-specific analysis for all issues -
though it leaves the option for further analysis open where even the staff finds it appropriate.
The draft proposal parallels earlier versions by stating that no " major redesign needs" for the current generation of plants are expected for ongoing severe accident research. 'It is possible -though not necessar-
,ily likely for any or all classes of nuclear power plants - that new information will demonstrate the desira-
~ bility of certain lesser changes such as improved reliability of some engineered safety features and addition of filtered vents to son.e types of containment and desiltn features that would reduce the risk from sabotage
' and earthquakes." the draft says, adding that research could pinpoint " worthwhile refinements" in current designs as well.
Earlier versions of the severe accident policy were criticized for not laying out the decision making process for any backfits in sufficient detail (INRC.18 April'83,4), and the latest draft sets out three steps for deciding on the safety of existing plants: engineering analysis of performance of existing plants,' analysis of existing PRAs for generic insights, and development of policy papers on important aspects of plant safety.
After those steps, the staff will put together a policy paper stating NRC's expectations about the ability of current plants to cope with severe accidents. Then, according to the draft, the staff will analyze possible im-
- -^
' provements, possibly using a " decision analysis method" yet to be developed, and will make recommenda-tions to the comrmssion. The proposals specifies that any backfits would be enacted through a public rule-
{
making but there is no specific provision for outside resiew of measures the staff does not think are warrant-ed.'an omissum Comrmsmoner James Asselstine objected to in the previous draft.
I Decisions to being backfits to the ca-== will be made in part based on cost-benefit analyses, done l-in line with the analyses required by the Committee to Review Generic Requirements, the draft says. How-
.'ever,it says benefit estimates will not be pegged to any single formula.such as the one used in proposing
- $1.000 per man rem averted as a cost-benefit measure in the commission's proposed safety goals. It also says costs need not be known more precisely than within a factor of three. "Rather, the range and variety of such
. conversion factors will be treated on the same footing as other sources of uncertainty." the draft says.
Recognizing that both engineenng analysis and PRA involve uncertainties, the draft sets out is philo-j.
sophy. Its precepts are:
L
- The most cost 4ffective regulations are preferred.
- Realistic estimates of costs and benefits are preferred, to avoid either over or underregulation.
l
- - Uncertainty reduction has a value, whether it involves a backfit reducing the uncertainty in an engi-
- neerms analysis or a change that will ensure regulations themselves are more stable and less uncertain. Mea-sures that " strengthen defense-in<iepth, or otherwise strengthen the diversity with which safety is assured.
are to be preferred over those which concentrate protection in fewer safety functions." and features that handle broad classes of accidents are preferable to those that handle specific vulnerabilities.
-_ Early implementation is preferred. since the value of measures taken an operating plants is higher -
t if they are done earlier in a plant's operating life.
(.
, Measures that minimize rule changes are preferred. since new requirements that chance the mix of personnel skills needed in both industry and NRC are difficult to accomplish.
L
- Measures that can be really verified are preferred. since ambiguities can lead to controsersy. delays m
L and costs.
1
- Regulations ought not to be unnecessarily prescriptive.When the NRC staff cannot make a clear-2-
INSIDE N.R.C. - January 23.1984 L
.a
.u
1 i
- s.,,y.
- w NUCLEAR SERVICES: SURVIVING THE SHAKEOUT WASHINGTON, D.C. APRIL 29 MA Y 2 AT THE SHOREHAM HOTEL - CALL NOW FOR DETAILS Sidney M. Stoller of the S.M. Stoller Corp., W.G. Counsil of Nonh-east Utilities and the editors of McGraw-Hill's nuclear publications have planned a conference to look even harder at the future of the nuclear services market than we did in last year's enormously suc-e cessful meeting. This time, top executives and policy-makers will
}
be addressing such issues as how the shakeout will proceed, who'll survive, how utilities would like the service community to look when it is over, and much more. We're pretty excited about the papers we've lined up.so far.
For details, phone Peggy Collins toll free at 800-2234180 (in New York State call 212-5124410); telex 232365 MCGRAW PNET.
cut decision among options based on the six principles above, this principle will be brought into play. The draft says,"It is a poor regulatory philosophy to narrow the options of the regulated industry unnecessarily.
Doing so discourages innovation and diminishes the sense of responsibility in the regulated industry." As &
result, NRC willlean toward asking applicants to come t'p with approaches for new plants and giving utilities with operating reactors alternatives among which to choose. "Where residual uncertainties leave the choice among alternatives ambiguous, performance objectives will be specified by the commission with the choice ofimplementation approaches to be left to the licensees," the draft says.
A major new proposal in the draft is a " safety assurance peopam," proposed by Roger Mattson, head of the severe accident effort, and Frank Rowsome. ~the draft, as did previous versions, proposes that new standardized designs be approved through rulemaking so they can be referred to in future construction per-mit applications. The proposal would use the techniques adapted from reliability engineering in the aerospace industry to make sure that the margins of safety calculated for standard plants when their designs were ap-proved are not degraded by choices of components, construction, or operations. "Since it is contrary to the -
concept of a certified design to call for an extensive new licensing safety analysis of the design as it is readied for commercial service, it is necessary to require in the rule certifying the design those subsequent institutional checks and balances that assure the original safety analysis assumptions remain 5alid with the further (de-finition) of design detail," the draft says.
Those same techniques could be used to tailor severe accident decisions to individual reactors,the draft says. The techniques include plant-specific safety analysis and feedback of experience into plant operations to make sure safety ar.alysis assumptions are valid, the draft says. Pilot programs are under way at Consumers Power's Big Rock Point, Consolidated Edison's indian Point 2, and New York Power Authority's Indian Point 3, the draft says, and NRC is looking at the concept in the Integrated Safety Assessment Program now being ex-piored (INRC,25 July '83,5). - Margarer Ryan DENTON DELIVERS 'EARLY SECOND GUESS' TO PROMOTE BWR PIPE SWITCHES The NRC commissioners have approved a staff plan aimed at allaying concern of BWR owners that " switch-ing out" cracked primary system piping will subject them to hearings on license amendments (INRC.14 Nov.
'83,5). Harold Denton, director of NRC's Office of Nuclear Reactor Regulation, said the "early second guess" of how the agency would treat pipe replacement under its regulations would provide " maximum flexibility" to encourage BWR owners to switch to nuclear. grade pipmg.
So. in a forthcoming letter, NRC staff will tell BWR owners: "We encourage programs to replace piping so as to mmimize the potential for cracking and we will expeditiously review any submittals provided to us so as to not delay this important improvement program... Prior NRC approval is not necessary unless the pro-posed change to the facility involves an unreviewed saiety question or a change in technical specifications."
Licensees, however, will be required to submit a radiation-protection plan to NRC before work is begun. and those who cannot keep estimated cumulative exposures to less than 2.000 man rems will be required to meet with the staff to resolve the issue.
After discussing the proposed letter. the wnmissioners asked the staff to make it clearthat the agency is issuing " selective and not complete guidance" on how uttlities should treat a pipe switch under 10 CFR 5039.
INSIDE N.R.C. - January 23,1984 3
c
f-I InsideNR.C A
An ensinsive report onthe U.S. Nuclear RegulatoryConsmission VoL 5, No. 22 - October 31.1983 TWO REPORTS FIND PRdPOSED PRICE OF AVERTING DOSES TOO HIGH Two draft research reports say that $1,000 per man rem averted - NRC proposed standard for calculat-ing benefits from backfits -is too high by at least a factor of 10. The reports will urge NRC commissioners to reduce the 51,000 to $100 or 550 per man rem averted, about what industry has been urging.
The studies are being done by Science Applications Inc. and the Pacific Northwest Laboratory. The SAI study involves methods for factoring worker exposures into safety analyses,while the PNL study, draft Nureg-0933, covers methods for settling priorities for dealing with generic safety issues. In looking at the dollar value placed on human life in other federal standards, Jerry Cohen of SAI told NRC's safety research meeting last week, SAI researchers found figures as low as $300 used in highway cost-benefit formulas. In contrast, the NRC standard works out to SS-million per death averted. That figure would be reduced to $500,000 by the SAI pro-posal to put the value of a man-rem averted at $100, and to $250,000 by the PNL draft proposal to place it at $50, though the latter researchers say any figure between $10 and $125 would be reasonable.
NRC is using the $1,000 figure now in test cost-benefit - or "value-impact" - studies of proposed new regulations and backfits. The figure is used in the commission's proposed safety goals.- Margaret Ryan THE JUSTICE DEPARTMENT HAS GONE TO COURT TO ENFORCE 26 SUBPOENAS by the NRC of present and for'mer Three Mile Island.2 employees in its investigation into possible falsification of leak, rate reports there before the 1979 accident.
De action follows a Sept. 21 refusal by the NRC commissioners (Commissioner Victor Gilinsky was not present) to quash 47 subpoenas NRC had issued after it decided earlier this year to reopen its investigation
(
into the falsification allegations. In making that decision, the commissioners rejected arguments that reopening l
their civilinvestigation and requiring testimony would conflict with the two-year.old enmmalinvestigation being pursued by a federal grand jury in Pennsylvania.
When NRC asked Justice two years ago to look into possible ermunal wrongdoing at TMI. it was under the impression that Justice did not want the NRC civilinvestigation to continue, the commissioners said. NRC i
has since !eamed that Justice has no objections to continuing the civilinvestigation and NRC wants to investi-l gate the charges before those involved forget what happened. The allegations relate to the ongoing enforcement l
proceeding imotving Three Mile Island and must be resolved before a final decision on TMI.1 restart can be made, they added. The commissioners did agree to allow those no longer living in the TM1 region to respond to their subpoenas in the court nearest them.
The court action by Justice is simply the next step in the progression of the investigation, NRC sources said. The U.S. District Court for the Middle District of Pennsylvania was asked to issue a show cause order to the 26 employees stillliving in the area, sources said, and NRC will probably wait to see how that action goes before it pursues the other individuals now scattered "throughout the United States." NRC has no en-forcement authority of its own on the subpoenas but if the court issues a show-cause order and upholds the NRC's requests,they said, anyone failing to respond to the subpoenas can be charged with contempt of court.
1
- Fmnces Seghers NRC'S SOURCE TERM STUDIES RAISE HOST OF NEW QUESTIONS NRC-sponsored source-term research is raising as many questions as it is answering, with researchers re-porting new phenomena needing stady before conclusions can be drawn. At NRC's safety research meeting WE'RE CHANGING OUR PHONES McGraw-Hill's New York offices are changing telephone systems. As of today.Oct. 31. the number for the nuclear publications' editorial offi:e in New York is (212) 512-3194. We may be experiencing technical f
(
shortfalls, so if you have trouble reaching us on -3194, try.3195, -3196 or -3197. For subscription infor-mation,if our toll-free numbers are not in senice, phone (212) 51245410 or 3916;we regret any inconvenience.
~ r :4 ?
'j%.;-
g in Gaithersburg, Md. last week, NRC researchers said they still expect to have recommendati:ns on source.
f,
term changes ta the NRC commissioners in the middle of next year, even th: ugh they have found new ques-p-
tions in the middle of the project. But no one was guessing how much of a change might be recommended,
[?
with new discoveries pushing estimated releases both higher and lower.
Researchers at the Battelle Columbus laboratories are trying to integrate the mechanistic codes describ-i ing different aspects of reactor and fuel behavior in accidents. Battelle's James Gieseke said they are already recalculating their first set of results,of accident sequences at Surry 1, a PWR with a large, dry containment, because of code improvemenu involving better information on primary flow rates and temperatures. The changes produced some lowering of expected releases in some sequences, where releases already were be'ow a thousandth of core inventory of the highest activity fission products. But Gieseke said BCL researchers had learned mostly that large uncertainties remain in the codes and decisions made by researchers along the way have big effects on results.
T.S. Kress of the Oak Ridge National Laboratory agreed with other speakers who said the basic thermal-hydraulic code that feeds into all other codes, NRC's March code, tops the uncertainty list. A major inade-quacy is its failure to model changes in core geometry as a melt progresses, changes that could be a " strong driver" of source terms,he said. No code models strong recirculation flows that researchers expect inside a vessel, Kress said, and research data on cere-concrete interactions and fission product chemical interactions is lacking.
Robert Wichner of Oak Ridge said the group studying fission-product transport had " belatedly" recog-nized that high-energy photons in a vessel would change fission product chemistry " dramatically," and that some experiments had to be redone. Wichner and R.J. Lipinski of Sandia Nationallaboratories said that a possibilityjust recognized is that aerosols containing fission products, which had settled out of the atmos-phere, would be resuspended by the sudden pressure changes when a molten core melted through a reactor vessel Until now, source terms for aerosols uniformly decreased with time, but Lipinski said researchers are looking at the possibility that up to 5,000 kilograms of aerosols could be resuspended late in an accident sequence.
Lipinski added that researchers also are looking at natural circulation from the core to the plenum, which could increase fission-product retention by a factor of 100 over earlier expectations. But they are also looking at whether decay heat might revaporize fission products that plate out on reactor internals.
S.J. Niemczyk of Oak Ridge said researchers had discovered that automatic fire-suppression sprinklers could come on during an accident and provide substantial scrubbing of fission products from the air, even
}
though that is not what they were designed fdr. The problem is predicting what sprinklers will do, she said, j
since they are automatic and come on with heat buildup. In addition, they are only installed in some com-l partments of plants. Niemczyk said researchers, who are using Browns Ferry-1 for their study, are looking at possible " secondary" leakage pathways for fission products, such as escape through main steam line isolation l
valves to the turbine building.
l Dana Powers of Sandia said attempts to pin down fission product chemistry were complicated because researchers codi not anticipate what impurities in coolant water would affect chemical reactions. While ces-lum hydroxide can react with stainless steel and baron carbide (in control rods) to form stable compounds, reducing the source term, cesium iodine reactions with those surfaces appear to produce free iodine, he said.
. What happens to that iodine is not yet known. Another high-activity fission product, tellurium,is released late in a core melt but can bond either with stainless stee' and Incor.el,in which case it would plate out harm-lessly, or with silver or tin from control rods,in which case it would become an aerosol and available for release.
P.C. Owczarski of Pacific Northwest Laboratory said his work on a code modeling supression-pool scrubbing had raised a key question: whether scrubbing continues throughout an accident. If accumula-tions of fission products in a pool act as surfactants," stiffening" the surface,it could significantly reduce the pool's ability to scrub fission products as an accident progresses, he said. Lipinski, who is working on sensi-tivity studies of code uncertainties, said there is uncertainty in some instances whether, when water hits melt material,it produces scrubbing or a steam explosion.
Joseph Jung of Sandia reported hitting analyticallimits in his attempts to determine containment pres-sure limits, a key component of source term estimates. Jung calculated stresses on three different types of containments: a ring stiffened steel containment at Watts Bar;a remforced concrete structure at Maine Yan-kee;and a pre: tressed concrete containment at Bellefonte.While he was able to analyze the two Tennessee Valley Authonty plants fairly well, estimating failure of Watts Bar between 120 and 140 psig and of Bellefonte at 130 to 139 psig, he found inadequacies in the state of the art of concrete analysis in examining Maine Yan-h kee and had to stop the analysis "due to numericalinstability."
J NRC researchers have added a study of containment failure to their source term work,with two work-ing groups trying to estimate probability distnbutions for contamment capacities and loading for different l
l 2
INSIDE N.R.C. - October 31.1983 l
n;c 2
- qq%.
- l-Complete proceedings of d
THE NUCLEAR SERVICES BUSINESS CONFERENCE l
held in Washington, D.C., May 17-20,1983 This conference, sponsored by Nucleonics Week, NuclearFuel and Inside N.R.C., drew up-wards of 300 utility and nuclear industry executives to an unprecedented discussion of the changing shape of the nuclear services business,its present and future components, the dol-l lars involved and the competition for them, and the obligation and responsibility of every-one involved to help the utilities hold down costs and improve plant operations.
/
as All the presentations made by the top-notch array of speakers are available now, induding opening remarks by the conference and session chairmen and the
- h_
SI-page complete transcription of the tape-recorded closing panel discussion I
C"_ _'"
and summary.
It's your chance to read and digest for yourself the data, analyses, forecasts, jj I
opmions and perspectiws offered by such experts as John Gray ofInternational i
/
Energy Associates Ltd., Charles Jones of NUS Corp., Sidney Stoller of S.M.
l StollerCorp., Jack Vessely of Florida Power & I.ight, E.P.Wilkmson of the N
Institute of Nuclear Power Operations, and many more.
The proceedings, totaling 338 pages and contained in a handsome three-ring binder, willbe boxed for shipment. Price per set is $220 in the U.S. and Canada, 3235 elsewhere.
To place your order write to Nucleonics Week,1221 Avenue of the Americas, New York, N.Y.10020. Full payment should accompany order. For further information phone Peggy g~
Collins. toll-free at 800 223-6180 (in New York State call 212-997-6410); telex 12-7960 MCGRAW PNET.
types of containments. Gieseke reported that codes had to be modified when PNL researchers got to an ice condenser containment because they found fission product plate-out on the ice bed. Some other results cited by Gieseke included:
- For a Mark I BWR, iodine and cesium compounds were scrubbed out by the suppression pool. How-ever, tellurium, which is released later, was available when a release path through the drywell opened. How.
ever,50% of the tellurium plated out in the drywell.
- For a BWR with a Mark III containment, no bypass paths around the suppression pool have been identified. As a result, fission products are expected to plate out in the reactor coolant system or to be scrub.
j l
bed in the suppression pool, leading to releases below 1% of fission product inventory of cesium iodide, ces-j ium hydroxide and te!!urium.
- Analysis of a sequence leading to basemat melt-through in a PWR with a large, dry containment show.
i ed substantial plate.out in the upper plenum and the primary system leading to very small releases if contain-ment breach is delayed.
THE PROBABILITY IS ' PRETTY HIGH' THAT A TEMPORARY OPERATING LICENSE could be needed for the Catawba plant, Duke Power officials sav, but spokesman at three other utilities listed as potential appli-cants for TOLs aren't so sure.
During recent congressional hearings on the NRC authorization bill for FY-84 and -85, when extension of the provuion allowing the issuance of TOLs was discussed, Catawba 1, Shoreham, Limerick 1 and Comanche Peak.1 were cited as four plants neanng completion for which utilities could possibly need TOLs to load fuel and go to low-power testing. Catawba officials expect to be able to load fuel by next May but don't expect the licensing process for the plant to be complete by then. Hearings on most issues are already going on. but no hearing on emergency planning has ever been scheduled. The utility would like the emergency planning hearings set for early next year but intervenors are pushing for later in the year, a utility spokesman said. increasing the possibility that Duke Power will apply for a TOL to avoid a delay of as much as two months.
The Texas Utilities Generating Co., owner of the Comanche Peak station, hasn't yet made a decision on whether to apply for a TOL but hopes to see the provision for it extended in the authorization act. Esti.
3
. Y, 7 SOURCE TERM STUDIES FIND VALIDITY TO SOME WASH 1400 ' WORST' ACCIDENTS NRC source term research, still in draft f:rm,is sh: wing that the worst case radiation releases predicted by the 1975 Reactor Saf ty Study will n:t materialize in most severe accid:nts at PWRs but could happen in some accidents at older BWRs. NRC contractors are still reviewing their methodology, however, particularly in light of a phenomenon discovered by the industry-sponsored Idcor effort that has led Ideor to tentatively predict hi3 er ieleases in some categories than NRC.
h The draft analyses, produced by Battelle Memorial Institute for NRC, use the results of experiments in fission products and transport in computer models more sophisticated than those available 10 years ago when the Reactor Safety Study, also known as Wash-1400, was being put together. Comparisons with the older study are rough, sources say, because the understanding of severe accidents has changed so much. How-ever, the analyses of selected severe accident sequences at three PWRs show drops in expected releases of an order of magnitude or more in many cases. Researchers now expect fission products in a severe accident to remain in the reactor coolant system or the containment, settling on or chemically bonding with surfaces.
But there are glaring exceptions in accident sequences where the containment is bypassed in PWRs and the suppression pool bypassed in BWRs. Accident sequences called ABS, in which a hot leg pipe breaks and all engineered safety features are lost at the same time containment isolation
SUMMARY
OF RELEASE FRACTIONS FOR A fails, and V,in which a check valve be.
SURRY UNIT tween the reactor coolant system and Fraction of Wash-1400 the emergency core cooling system fails inventory release and opens a path for fission products Sequence Species released fraction to the auxiliary building, produced the ABe 1
4.8 x 10-5 8 x 10-4 highest potential releases in an analysis Cs 4.8 x 10-5 8 x 10-4 using Virgiria Electric Power's Surry Te 4.0 x 10-5 i x 10-3 plant as an example. In the higher V AB7 I
8.7 x 10-2 o,7 sequence,52% of the iodme (as cesium Cs 8.5 x 10-2 0.5 i dide),43% of the cesium (as cesium Te 7.0 x 10-2 o,3 hydroxide) and 5.8% of the tellurium are estimated to reach the auxiliary ABS I
8.7 x 10-2 0.7 building. In Wash 1400, those releases Cs 8.5 x 10-2 o,3 were put at 70%,50% and 30% respectively.
Te 7.0 x 10-2 0.3 However,in the latest draft of the Batte!Ie ABS
- I 5 x 10-2 0.7 study, designated BMI-2104, researchers y
Cs 5 x 10-2 0.5 have attempted to estimate how much of Te 4 x 10-2 0.3 those key fission products would be re.
abs **
I o.37 o,7 tained in the auxiliary building.They Cs 0.37 0.5 show only about 7% of the cesium and Te 0.19 0.3 iodine compounds and 4.4% of the tel-
- O*
'"C' TMLB '6 I
4.7 x 10-2 o,7 tention of up to 30% of the cesium and Cs 4.4 x 10-2 0.5 iodine and 12% of the tellurium are Te 0.11 0.3 estimated for the ABS sequence if the TMI.B 'e I
2.6 x 10-3 8 x 10-4 containment and auxiliary buildings are Cs 5.3 x 10-4 8 x 10-4 divided into two model areas, and about Te 7.9 x 10-2 1 x 10-3 4% more each if they are divided into S Dy(Hot Leg)
I 4.2 x 10 5 four areas.
2 Cs 6.4 x 10 5 BWR sequences are being recalculated Te 3.3 x 10-2 and NRC sources expect the numbers S De(Cold Leg)
I 1.5 x 10-8 2 x 10-5 there to change, but they say the recalcu-2 Cs 1.4 x 10-8 1 x 10-5 lations cannot avoid the suppression Te 7.7 x 10-8 2 x 10-5 pool bypass problem. In Mark IIIs, that bypass is considered a very low proba-V I
0.52 0.7 btlity,and the first draft of analyses using Cs 0.43 0.5 Mississippi Power & Light's Grand Gulf l Te 5.8 x 10,-
0.3 as an example did not include any se-
- Uses containment model with four control areas:ABg quences in which the pool was bypassed.
user two.
NRC researchers are now analyzmg a se-
- With no credit taken for attenuation in auxiliary building.
quence that includes a partial bypass.
But analyses of a Mark I plant, using tNSIDE N.R.C. - February 20,1984 3g
-- q l
Philadelphia Electric's Peach Bottom-2 as an example, find bypassing a potential problem in each of three h
sequences.
M The drafts, and NRC and industry sources, all caution there are many modeling uncertainties for BWRs,
' ~@
including how to handle lower plenum and control rod drive structures, molten core behavior, and suppression y
pool scrubbing efficiency. But in the early Peach Bottom analyses in event TW-gamma prime, which involves
}
failure of the residual heat removal system and a consequent pressure buildup that
SUMMARY
OF RELEASE FRACTIONS FOR A causes containment failure directly to the SEQUOYAH UNIT atmosphere, researchers estimated as Fraction of Wash-1400 much as 28% of the cesium and iodme inventory release compounds, and 18% of the tellurium, Sequence Species released fraction could escape. In sequence AE, involving TMLB'y 1
1.7 x 10-2 0.7 to 0.2 a large loss-of. coolant accident (loca) 2.2 x 10j Cs 0.5 to 0.2 combined with enyggency core cooling Te 1.4 x 10s 0.3 to 0.3 injection failure and failure of the dry.
TMLB3 1
3.9 x 104 0.7 to 0.2 well,up to 67% of the tellurium and 21%
Cs 4.4 x 104 0.5 to 0.2 of the cesium and iodine could be re-Te 2.2 x 10-3 0.3 to 0.3 leased. The worst-case BWR accidents in Wash-1400 predicted 40% of the cesium S HF7 1
6.0 x 10-3 0.7 2
and iodine would be released and 70%
Cs 6.0 x 10-3 0.5 of the tellurium in one set of sequences, Te 2.9 x 10-2 0.3 and 90% of the iodine,50% of the cesium TML7 1
1.3 x 10-3 0.2 and 30% of the tellurium in another set.
Cs 6.5 x 10-3 0.2 Those early BWR drafts said further Te 5.5 x 10-3 0.3 analysis of structures at individual BWRs TMIA 1
6.9 x 10 9 3 x 10-2 could affect their conclusions. For in.
Cs 7.4 x 10-9 9 x 10 3 stance, the Peach Bottom report said,"If Te 1.6 x 10-8 5 x 10-3 the reactor building were expected to withstand blowdown forces from the failure of primary containment with a high degree of confidence, the risk could be reduced substantially." The Grand Gulf draft suggested some l
Mark I reactor buildings could retain more fission products than this Mark 111, because of the " limited strength" I
of Grand Gulf's shield building and the lower capacity ofits standby gas treatment system.
Researchers on all sides say there are major uncertamties left in all the analyses, despite the advances made in the last several years. Participants in the Idcor program have expressed some frustration in two tech-nical meetings because NRC research has occasionally produced data outdating their analyses by the time they report it. But Idcor researchers have come up with one finding that has NRC contractors going back to their models for a new lock. After Idcor and NRC researchers met in Baltimore Feb. 7-8, Idcor researchers reported that in their computer code, fission products that settle on the surfaces inside the reactor continue to produce heat through fission decay. Under some circumstances, they said, that heat could be sufficient to resuspend the particles as aerosols, increasing the fission product aerosolinventory at times not previously accounted for. Timing is a key question, since all releases to the environment depend on what is in the reactor atmosphere at the time that a leak path develops to the outside.
NRC researchers agreed that their codes, considered generally more complex than Ideor's, do not account for resuspension of aerosols and said they want to look at the effect of that phenomenon on both their thermal-hydraulic and fission product transport codes. It is uncertain what differences the
SUMMARY
OF RELEASE FRACTIONS FOR A phenomenon could make since Idcor has ZION UNIT not given its plant-specific analyses to Fraction of Wash 1400 NRC with the work on core heat.up and inventory release on iission product transport phenomena.
Sequence Species released fraction Core heat-up, including steam and hydro-TMLB '
I 1.9 x 10-6 8 x 10-4 gen production that can endanger the Cs 1.9 x 10-6 8 x 104 contamment, were debated by NRC and Te 7.8 x 10-5 1 x 10-3 Ideor technical;xperts at Harpers Ferry, S3De (Hot Leg)
I 2.5 x 10-8 8 x 104 W. Va. late last year and there were few i
Cs 2.3 x 10-8 8 x 104 substantial areas of disagreement (INRC.
~
Te 3.6 x 10-8 1 x 10-3 12 Dec. '83,6). But at the Baltimore meeting. Idcor speakers said their calcula.
12 INSIDE N.R.C. - February 20.1984
R
'hi tions t f potential fission product releases were comms in hishir,in some cases, than th Btttelle draft num-y 4
bers, and sources said the resuspension phen:menon was rpparently a majgr contributor.The meeting on the plant. specific analyses, once c:t for March,has now been moved back to May because both sides want to do[
i more worki I
s De usajor dismysements between Idcor and NRC contractoes identified at the Baltimore meeting
. laciuded:
. De timing of teNuruun reisenes. ldcor experts think tellurium will be released earlier than NRC's f
experts, who say some teBurium will be temporarily retained by uncxidized zirconium cladding. Both sides
^
agreed Idcor's model called for release of other fission products et lower temperatures - earlier - than research now indicates the release will occur.
. Whosher there is siher in the aerosols. NRC analysis of contammation in the Three Mile Island.2 reactor indicates a significent portion of the aerosol mass was silver from the control rods. Idcor's models show the silver melting and running off before there is significant vaporization. Paul Nakayama of Jaycor Corp., an Idcor contractor, said he has been unable to figure a way for the silver to be available for aerosol formation,given the temperatures and pressures at core melt. Mike Kuhlman of Battelle said silver could repre-sent'atmost half the aerosol mass, and Phil Mcdonald of EG&G Idaho, an NRC contractor, said the aerosol quantity and the timing of its formation could make a big difference in the amount of fission products car-ried out of a damaged reactor vessel.
Distributions of different slass of particles in vapors. Idcor's model assumes a more uniform distri-e'
- bution than those used by NRC's contractors. Since the particle size - measured in microns - is directly re.
lated to how long the particle stays suspended in vapors and available for release, assumptions about size can affect results throughout calculations for fission product transport, through both the reactor coolant system t'
- and the contamment.
e - Use of a constant decone==i==*ian favor (DF) for suppression pool scrubbing in BWRs. Idcor and Gen 6tal Electric experts said they had used the most conservative factor in their calculations, and that tests showed the suppression pool could be expected to scrub out fission products much more effectively. NRC contractors questioned whether the DF should be altered over time, as temperatures increase and as the sup.
pression pool fBis with debris.
NRC contractors acknowledged that they have found some phenomena more completely modeled by
. Idcor's codes than by NRC's. Several observers said Idcor's work was winning "more respect" from NRC for
- tackling such phenomena. However, sources on both sides predict " blood on the floor" for the May meeting when Idcor's application ofits codes to specific plants 30 head.to head with the Battelle analyses done for NRC. Idcor has not released its studies, and some elements are being reviewed, but Idcor announced last fall that the studies showed all U.S. nuclear plants could withstand se_ vere accidents safely.
Industry has been hoping the source term research would both stave off severe accident backfits by proving current plants are safe enough, and provide some basis for reducing emergency planning requireme by showing the worst accidents have far less severe consequences than Wash 1400 predicted. The dr done by Battelle and Ideor must be combined with other studies by NRC before their exact effects will be known. One group in NRC is looking at the probability of each accident sequence, to see if the risk of the worst sequences occurring is a real one. Two other groups are looking at containment behavior to ascertain how real the risk of each type of containment failure is. No group has completed its work.
Battelle tested codes on two PWRs 'with large, dry containments, Surry and Commonwealth Edison's Zion. B8th plants have had probabilistic risk assessments (PRAs) and NRC's contractors took sequences t ranked high on their4 ras' risk lists to test their codes. According to drafts given to NRC in January, both
~
l were analyzed for the *e sequence - station blackout with failure of the secondary system relief valves that cannot be corrected before the core melts through the containment. Researchers estimated there would be substantially more " scavenging" of fission products by the reactor coolant system in the Zion case than at 7
. Surry, with fission product releases more than 100 times lower than predicted in Wash.100. At Surry, how
, ever,the numbers are slightly higher than the Wash 1400 predictions, with less than 1% of the cesiu iodine available by the time containment is breached but some 7.9% of the tellurium available. At Surry, the
_ same sequence with an early containment failure due to overpressurization, TMLB '8, was calcula l.
i results were far lower than Wash la00. Instead of 70% of the iodine,50% of the cesium and 30% of the tel-lurium escaping to the atmosphere, the researchers now estimate only 4-5% of the iodine and cesium a of the tellurium would escape the reactor coolant system and containment, even with the early containment breach. Early breaches are generally considered the most serious accidents, though there is debate abo and how such breaches could occur in reality.-
Other accident sequences analyzed,with all showing significant reductions in expected releases includ i
I-l-
13 i'
INSIDE N.R.C. - Fetmaary 20,1984 l
NQ
- [h AB, a hot leg loca f;11 owed by f
- ilure of all engineered safety features, combined with e (melt through t.y
. (failure by hydrogen burn), and # (frilure by inadequate isolation) containment failure modes:S D, a small-Mg 2
break loca with an emergency core cooling injection failure, analyzed with the 7 and e failure modes.
Sequoyah, chosen to represent ice condenser containment PWRs. was analyzed for the TMI.B sequence j
with 7 and 6 containment failures. The latter is a failure by overpressurization and was postulated to take k
place almost seven hours later than the y failure, estimated to occur about 3.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> into the accident. Also analyzed at Sequoyah were: the TML sequence, an initiating transient followed by failure of the diesel gener-ators, secondary system relief valves and auxiliary feedwater, for both the y and 5 containment failure modes; and theS HF7 sequence, a small-break loca with failure of the emergency core cooling system and contain-2 ment sprays leading to containment failure from a hydrogen burn. In all cases at Sequoyah, releases were sub-stantially below the Wash-1400 predictions, with the ice and its compartment structures estimated to catch anywhere from 6% to 25% of various fission products, though the reactor coolant system retained the bulk of those containments as it did at the other PWRs. - Margaret Ryan IHSI MAY REMEDY CRACKING PROBLEMS AT SOME BWRS, NRC STAFF SAYS Induction heating stress improvement (IHSI)"may provide an acceptable remedy for long term opera-tion" of BWRs that do not have "significant" primary system cracks, the NRC staff told the commissioners last week in an update on long range planning for resolving the problem of intergranular stress corrosion cra'ck-ing (IGSCC). The staff said it will have specific recommendations for guiding utilities in adopting long range strategies by May, but it noted that several utilities are " actively considering" pipe replacement.
The staff plans to be able to " anticipate alllikely long term solutions" that BWR owners may propose, the staff said, so that it can readily determine the " acceptability and conditions which might accompany ac-ceptability" of those proposals. One potential conflict the staff foresees with industry is over deciding how remedial techniques such as IHS! will impact inspection frequency and radiation exposure to inspectors over the lifetime of a plant. To expedite resolution of such conflicts, the staff,in conjunction with outside consul-tants,is developing specific policy positions on:
- replacement of all or part of susceptible piping;
- IHS1 of piping for recently licensed plants;
-IHS1 of piping for older plants where cracking has not been detected;
- use of hydrogen injection techniques in combination with other mitigation measures;
- long-term operation with shallow cracks and/or crack repairs; and
- combinations of the above approaches on a plant-specific basis.
Meanwhile, at a recent, closed meeting of the Organization for Economic Cooperation & Development (OECD) Nuclear Ener*y Agency in Paris, there appeared to be international agreement that IGSCC is a " safety-relevant" problem with only one certain solution - pipe replacement. However, sources said, none of the nu-clear regulators said they favor such a radical policy. Rather, regulators worldwide are grappling with the prob-lem of deciding which of several short term solutions should be adopted. Sources said that no regulatory body has yet taken a firm position on this issue. At the meeting - the first concrete follow-up to OECD's caucus I
last June of senior regulatory officials - U.S., Japanese, Swedish, West German, Italian, Finnish, Swiss and I
Spanish officials gave formal presentations. Representatives from the United Kingdom, France, Norway and Holland also attended.
The participants in the Paris meeting were told that worldwide about 10.000 BWR welds - or some 10 kilometers of welding - are susceptible to IGSCC, sources said. It is estimated that a maximum of 25% of those welds actually are affected by stress corrosion cracking. The most conservative repair, a sweeping change of all BWR pipes, would cost $50-million per unit, excluding outage time, sources said. If regulators are will-ing to accept a less-than-perfect solution, they are faced with the necessity of deciding what size cracks are acceptable and how to reliably detect and measure them.
So far, the sources said, only the Japanese appear to have taken a position, albeit informal, on the ques-tion of what size cracks should trigger repair or replacement of BWR piping. Japanese regulators reported that they are working to a " rule of thumb" under which cracks reaching one third of the pipe wall thickness would trigger a regulatory response, the sources said. However, the regulators stressed that they ara treating each plant individually. In the 1970s, German utilities were ordered to change BWR piping, but for a different reason (a cold cracking problem stemming from steel manufacture), with the result that German BWRs have not been affected by IGSCC. Sources noted that other regulatory bodies appear to be waiting to see what the U.S. plans to do, h
Participants reported that the " big hit" of the meeting was IHSI, which involves heating welds from the outside via electric coils while cooling water moves through the pipe, thereby reversing the stress in the weld. This technique has been used in the U.S., where General Electric has been competing with Japan's Ishi-14 INSIDE N.R.C. - February 20,1984
INTER-DEPARTMENT MEMO - PUBLIC SERVICE COMPANY OF COLORADO PPC-84-0693
- +
DATE:
March 14, 1984 TO:
Mr. LeRoy W. Singleton, Superintendent of QA Operations, FSV FROM:
Jerry J. McCauley, Results Engineering Supervisor, FSV ATTN:
SUBJ:
CAAR 233 The target' date-for our completion of CAAR 233, concerning the incorporation _of relay settings and tolerances into the Master Setpoint List,1s August 15, 1984.
M ii tM /
Jgery g gauley
~~ ~Q JJM/djc l
4 S
- f..
i
\\
, ~ ~. _. _ -..,.,, _.,