Text

Evaluation of TMI Action Plan Application to Fort St Vrain, Monthly Rept for Period Ending 810615
	ML20004F270
Person / Time
Site:	Fort Saint Vrain
Issue date:	06/08/1981
From:	Ball S; OAK RIDGE NATIONAL LABORATORY
To:	;
Shared Package
ML20004F267	List: ML20004F266; ML20004F270;
References
	CON-FIN-B-0762, RTR-NUREG-0660, RTR-NUREG-0696, RTR-NUREG-0737, TASK-1.D.2, TASK-2.B.3, TASK-2.B.4, TASK-2.K.3.17, TASK-TM NUDOCS 8106170151
	Download: ML20004F270 (17)
	v • d • e

-

DRAFT S. J. Ball For Comments 6/8/81 i

E7ALUATION OF TMI ACTION PLAN APPLICATION TO FORT ST. VRAIN !

Requirement I.D.2: Plant Safety Parameter Display Console j l

I. Background The new NRC requirements for a plant safety parameter display system (SPDS) are outlined as part of the Control Room Design task, I.D. in NUREG-0660 (May 1980). i The objective of this task is to improve the ability of the control room operators ,

to prevent and cope with an accident by improving the information provided to them. In conjunction with a design review and upgrade of the control room, all I licensees are required te design and install a SPDS which will:

"diaplay to operating personnel a mini =um set of parameters (safety l state vector) which defines the safety status of the plant. The system should have the capability ci displaydng a full range of i important plant parameters and data trends on demand. In addition,-

the system should provide indication of when process limits are being approached or exceeded."

The licensees were originally required to submit a system design for NRR review by January 1981; however, since the action plan clarification (NUREG-0737) was not issued until Neve=ber 1930, :he deadline for the design (for 757) was reset to July 1, 1981.

The NRC requirements for the SPDS were issued as part of NUREG-0696 in November 1980. The plant functions requiring indication on the SPDS are, at a min 4=n=, reactivity control, reactor core cooling, reactor coolant system integrity, radioactivity containment, and containment activity. The SPDS is to be located in the main centrol room and have additional displays in the technical support center (TSC) and the emergency operations facility (EOF). The required design unavailability goal of the SPDS is 0.001, or 9 hrs / year.

II. Present Status of Implementation at FSV Per a June 4,1981 conversation with the PSC licensing engineer in charge of the emergency response facility implementation, PSC plans to submit a conceptual i

design for the S?DS to NRC by July 1, 1981. The conceptual design will specify what parameters will be read out on the SPDS, and propose means for presenting the operator with trend data and digested information.

e I810617015l' .

- - .. . -- - . - - . - - - -- -..----.--0

I.D.2 2

?

III. Evaluation of Applicability of the SPDS to FSV The design of a " satisfactory" SPDS would be dependent on reactor type (PWR, BWR, or HTGR), and except for minor details, a design that was satisfactory for one plant would probably be suitable for all reactors of that same type.

Thus FSV is at a disadvantage in that the enitre HIGR SPDS development burden would fall on the one plant, while PWR and BWR owners could pool their resources.

On the other hand, plant-type standardization efforts might turn out to be more costly and time consuming than individual efforts.

The objectives and the functional ' requirements for the SPDS for FSV (and others) are clearly vague and therefore allow the licensees considerable latitude in their designs. The objective would be to provide the operators with safety-related information not readily accessible on the main control panels. Such information could include summaries of critical plant parameters, notification of certain combinations of conditions diagnosed to be significaat or potentially dangerous, trend information not apparent from control panel observations, and more sophisticated computations that could indicate potencial problems (such as mass, heat, and reactivity balances). Because of the relatively slow response times of many HTGR parameters, diagnostic algorithms would also have to account for the dynamics of the plant.

The NRC guidelines also appear to be contradictory in that the SPDS pars =ecer set is to be minimized and at the same time be capable of determining the overall status of the plant. Considering the large ou=ber of vital components and subsystams in FSV, it is possible that an SPDS that was poorly conceived with respect to a given accident could be distracting or misleading to an operator' who could otherwise be getting a more complete picture of the situation from the larger and more detailed main control panels. A significant distinction between HTGRs and LWRs with respect to the SPDS is that due to the inherently slower response of the HTGR, its operators would have much more time during an accident sequence to properly assess the plant conditions from the more complex and complete main control panels. Consequently, safety parameter monitoring in FSV equivalent to that in a FWR or 3WR could be effected with a less detailed SPDS.

The design unavailability requirement (0.001) appears to be low enough to dictate the need for backup computers instead of just a single computer, and would escalace the costs and complexity of the system considerably, and perhaps unnecessarily.

Consequently, a cost-benefit analysis should be done to arrive at a justifiable unavailability goal for an HIGR SPDS. ,

l I.D.2 3 !

IV. Summarv of Conclusions and Recommendations to Date

1. Final recomendations will be made af ter further review of NRC PSC correspondance.

2. An assessment of the requirements for the safety parameter display system (SPDS) as outlined in NURECs 0660, 0737, and 0696 has determined that the intended objectives of the action item may not be met because of the vague and somewhat contradictory wording of the NRC reports!

Particularly in the case of FSV, which has long response times relative to LWRs, a SPDS which has only a summary of information otherwise available on the main control panels may be distracting or misleading to an operator who would probably have sufficient time to absorb and analyte a more complete set of data. Because of the tight schedules imposed on the licensee for the design and implementation of the SPDS, it is likely that relatively little analytical capability could be incorporated into it. Consequently, it is recommended that: (1) the summary display of critical parameters requirement be waived for FSV; (2) the licensee and NRC jointly develop design criteria and requir=ents for an analytical capability for the SPDS (such as reactivity anomaly detection, heat and = ass balance ecmputations, component perfor=ance degradation detection, accident progression prediction, etc.) and a schedule that would allow a reasonable amount of time for development.

3. A cost-benefit analysis should be done to justify the design unavailability goal requirement of 0.001 for an HTGR SPDS.

V. Proposed Further Action 3v ORNL

1. Review recent NRC and PSC correspondance on the emergency response facilities (ERF), and on the SPDS in particular.

2. Visit FSV to observe the status of the ERF, and discuss the present PSC SPDS design criteria with operations personnel.

3. Make an independent evaluation of HTGR SPDS detailed functional requirements.

4. Investigate cost-benefit features of single vs backup computers.

5. Develop final recommendations.

l .

I

DRAFT J. C. Conklin Ecr Comments 6/11/81 EVALUATION OF TMI ACTION PLAN APPLICATICN TO FORT ST. VRAIN Item II.B.3: Postaccident Sampling Capability I. Position The basic statements regarding radionuclide analysis can be directly applicable to Fort St. Vrain (FSV). However, the requirement that the reactor coolant spectrum correspond to a Regulatory Guide 1.3 or 1.4 might need reevaluation in light of the entirely different fuel configuration of FSV as compared to a L*4R.

Specifically, the FSV fuel consists of spherical particles of (Th + U)C3 ,

coated with sic, and dispersed in a matrix of graphite so that there is no direct contact of coolant and fuel particles. Therefore, the path for radionuclide transfer to the coolant is not as direct as the L*4R. Placeout of iodine and cesium upon the graphite moderator occurs during normal operation, and might become a source term for these radionuclides during a heatup or moisture ingress accident.

A regulatory guide for FSV comparable to Regulatory Guide 1.3 or 1.4, and 1.7 should be developed, based upon appropriate phenomenological considerations.

The chemical analysts requirement for boren and chloride concentration is inappropriate for FSV. Chemical analysis for foreign gases in the reactor coolant system 'should be substituted. These foreign gases would primarily consist of H3 0, H 3, CO, and CO

2 II. Clarification Items (See Clarification listed in NUREG-0737, I1.3.3)

1. This item is appropriate for FSV. Mcwever, as an HIGR has a substantially increased thermal capacity over an LWR, the three hour time limit for sampling and analysis might be increased if this would result in improved accuracy and measurement reliability.

2. Appropriate for FSV, with modifications

a. unchanged

b. hydrogen, air, 00, and CO2 e ncentrati ns in the coolant and reactor building *

c. inappropriate

d. unchanged.

3. unchanged 4 inappropriate j

5. inappropriate

6. unchanged

7. inappropriate a

e 8

~

II.B.3 2 !

8. Should be clarified, so as to preclude use of the plant helium purification system or other system designed for power production use at normal operating conditions to meet the requirements of this action item. The normal and accident coolant sampling systems should j be independent. !

9. Appropriate, with modifications

a. FSV versions of Regulatory Guides 1.3, 1.4, and 1.7 need to be developed. The sensitivity requirement for liquid sample analysis is inappropriate, and a value for gaseous sample sensitivity should be substituted.

b. unchanged f

10. Modified as follows:

Accuracy, range, and sensitivity shall be adequate to provide pertinent data to the operator in '. der to describe radiological and chemical ,

status of the reactor coolant system, regardless of system operating conditions. This requirement specifically includes a situation where all coolant circulation has ceased, at any system pressure.

11. unchanged.

Other ite=s unchanged.

III. Procosed Further Work or Action bv ORNL Additional work is needed to identify appropriate radionuclide source terms.

A great deal of work concerning this subject has previously been done by many investigators, and the infor ation could produce a preliminary regulatory guide for radionuclide source terms analogous to Regulatory Guide 1.3 or 1.4. A great deal of work also has been done for combustible gases (H2 , CD) that might arise from moisture ingress accidents and a preliminary regulatory guide analogous to Regulatory Guide 1.7 could be produced. .

.- , . , - , . - . ., . . . _ , , .na -. - -,- - - ,

DRAFT R. M. Harrington Fr.r Comnento 6/8/81 EVALUATION OF TMI ACTION PLAN APPLICATION TO FORT ST. VRAIN Item II.B.4 Training for Mitigating Core Damage

I. Background Ihe occurrence of severe core damage implies that important reactor plant systems have not been available to, or properly used by the operator. However, at some point during a severe accident sequence by optimum use of available systers it is possible that the operator can halt further core damage and/or prevent containment failure, thereby minimising release of radioactivity. A necessary prerequisite to optimum operator decision making is the ability to diagnose the true condition of the reactor from plant instrumentation.

It is not obvious for every accident and for each reactor plant system what the optimum way to control or mitigate the course of the accident is. The NRC is currently funding at a level of S2(10)6/ year a Severe Accident Sequence Assessment (SASA) program for LWRs. One =ain purpose of the SASA program is to develop a base of knowledge that will allow operators of LWRs to determine the best way to control or mitigate a given severa accident. Industry has funded (at $8(10)0 total) an independent program with approximately the same goal. Never-theless, until these research programs are completed it is important for reactor operators to summarize available knowledge of severe accidents gained from the accident at IMI and knowledge of their own reactor plants in order to provide guidance for severe accidents where none has existed in the past.

II. Present In-plant Status at FSV FSV has committed to develop a training program outline that meets the intent of the requirement and to have that program initiated by April 1,1981.

III. Applicability to FSV While FSV is very different from a LWR, severe fuel damage is possible; therefore, the contingency plans for this unlikely possibflity should be =ade a part of the FSV operational training program.

There are a nunber of possible ways in which the FSV core could be damaged.

The steam generators, helium circulators, and reactor core are all inside the Prestressed Concreta Reactor Vessel (?CRV). An inleakage of steam from the ,

higher pressure steam generators into the hot PCRV would cause oxidation of graphite

. _ _ , _ __ , _ , , _ - , . _ , _ - _ _ , _ rr ,,,--.,_mo

II.B.4 2 in coolant holes in the acderator blocks which hold the fuel. However, this >

reaction requires heat and therefore tends to be self-lfmiting. This combined with system design features makes steam ingress an unlikely candidate for severe fuel damage. Air leakage, if possible, would be of much more concern due to the heat liberation of the air + carbon reaction; however, the PCRV is pressurized with helium coolant and the available sources of air are at atmospheric pressure.

The possibility of an Anticipated Transient Without Scram (ATWS) event causing severe core damage seems unlikely due to the unique nature of HTGRs. The backup scram system consists of 1/2-inch diameter poison balls that are dropped into round channals internal to the core. Once actuated, they are in place within seconds. This contrasts to the very slow shutdown achieved by borating the coolant in LWRs. In addition, the FSV core has a very large heat capacity so that there is plenty of time for actuation of a backup scram even for a relatively large discrepancy between heat generation (power level) and heat removeal. The normal full load fuel temperature of about 2000*F is more that 1000*F below the temperature at which the refractory fuel particle coatings begin to fail, allowing volatile fission products to escape. For these reasons ATWS events have a low priority for degraded core considerations at FSV.

Sequences that should receive priority degraded core consideration for FSV involve the loss of forced circulation. The FSV design has steam generators well below the reactor core, therefore even at the full coolant pressure of 700 psia, natural circulation alone will not provide sufficient heat removal. If circulation is available, then adequate heat removal via the steam generators is possible at any helium pressure down to and including atmospheric. Since the steam-driven circulators and the steam generators are normally used for power production, backup safety systems were provided to assure their availability for emergency use.

The fire-water system provides both a redundant drive for the circulators (via the Pelton wheel) and a redundant supply of feedvater to the steam gecarators.

Sequences involving a postulated loss of all 4 circulators are shown in a very schematic fashion on Figure 1. The dominant sequence (No. 1) has already received considerable attention: it is Design Basis Accident No. 1 (DBA-1) and is fully discussed in the FSAR (as amended by PSC submittal P-77250 dated December 1977). When the circulators fail, the reactor trips, and the massive 27 ft. diameter by 23 ft. high reactor core begins to heat up in an approximately adiabatic manner. Figure 2 reproduces the fuel camperature vs time reported for DBA-1 La the FSAR. This figure is included here to illustrate the very slow race of core temperature rise during unrestricted hestup. With respect to core heatup l

l l

i '

II.B.4 3 Figura 1: Exarnoln of Swxra Accidtnt Stquinces for FSV Circulator Failure i Liner 3egin Cooling? Depressurizing in < 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months ?

5 l

E' l Y Y 1. This is reaign Basis Accident ciredators ed a i No. 1 (see FSAR). Effects include: no PCRV failure, fa 0 l moderate fuel damage and a

[

j moderate extended fission product (FP) release. Stable l condition reached vita decay N I heat removal by Liner Cooling.

's i 2. Consecuences Unknown:

1 Possibilities include PCRV failure, total core j destruction, and rapid FP release.

l

' 3. Consequences Unknown:

\ T Possibilities include PCRV

)l failure and total core

}

g destruction.

I I l 1

'I I i 4 Consequences Unknown:

NI Fossib111 ties include PCRV failure, total core

)

destruction and a rapid FP I release.

I I

I l

1 I I I ,

I I l (4 htat should operators do if Liner Cooling is recovered? Concerns:

l Thermal shock to overheated PCRV i l liner, water hax:rner and flashing l due to cold water in hot cooling water 7 s.

I I

i

'--Y btat should operators do if circulators are recovered?

Concerns: Effect on PCRV liner, steam generators.

e

l U

in b

6000 ,

?!

u O

h 5000 -

HAX1 HUM CORE PN TEMPERATURE

^

NH d ." t8

>- a1 nn ow 4000 -

n ze O 4

. wM aa

" w" 3

0r 3000 - TF nn s~

ew rw VOLUMETRIC HEAN ACT!VE #"

eo CORE TEMPERAlURE . rg w" ** $ $

2000 -

u ew 03 $O 4

EU I

g" og

-z 1000 -

0-ww n rr o E

' ' ' I ! I I I i l l i i $

O n l 10 100 1000 $

o

}

TlHE (HR) j Fig. 14.10-1--Haximum and volumetric mean active core teuperature during LOFC accident

,l

53 II.B.4 5 Fig. 3: PCRV Thermal Barrier Arrangemen: (from FSAR) t

.15 ' ,ts + Y A &Y ~ +~

TMta0(8 STUS MSI.C 00 5fv0 t l .

s s' .

ll",i"** t.- i '6 .

ij ::jp

---l (( 'fi '.Sih4;

s%: .

..,,,,,,," s m .. i -

e-

-t; g %: ..rt.$. -

..c. seu..:

toes wtss ;:,,e,' i g. :::u.s *.ss

- ,1 i '

..t in.cs.,/ t i':.':p:.a .32*

yA,;.

. gpy

.u.

1 '*"*"% l %',, .

' .if l ;

l a.

i,.,. . . ,..u....

,'.',g a '

I . .

- - a .....

s,.s inir%+: -

+/;

M. uU.:

t' Mr

$, m ?$;/s.

2222225 _.

3' A. ..

r' j W::W t! ::,,: %t.*N N; i . /"s r' 9 % .c $r$y.

l l ll ' ::! ; * :

MA"" 5*'""

l} !!l}5: 5i lif I

l kl.' }/.E, [3[,

,, ., A

!.'d NM,

,k 3k ) emcas secree. . au sesusario. i. .asa a,

.. l ..* !

.r . .';. .. ! .: *

,' ., e. .r" . ' ..

_ / ts .

l n

JM ;,' w. s.. ....

. e '[

4 ., . . ,

l c.

j l *> ., . ...

,1 w . . . ......

,/,

, i ., .w .

,, ., ww . .. . . .

s ,* ., s. .. .

{ . w .. ..

... w . . . . .

. . . , = . .

, J ... . ..

g I . .I

.. ... .. .. . .I e U %". h I"""' \

,./ V > < r 9.,. ,.6.......

i.,....

s ,' ,.,,...... . . . . . .

..,- l ..

,; 1 , v

g. , t. .. . . . . . .. 1

,, .u. . ,, .

we

.~.4.s.*.F..

. e. 6

, Fig. 3.9-3--Ther=al barrie r arrangement

i II.B.4 6 Table 1 PCRV Material Properties

7 Temperature Component Property ( 'F )

Cover Plates failure 1500 to 2000 (o.25" thickness steel)

Kaowool Insulation Melting 3200 i Liner (0.75" thickness failure 2000 steel)

Corcrete design temp. 150 free water loss 190 to 275 decarboxylation 1630 to 1710 (CO2 production) zero strength 1800 melting point 2000 to 2700 ,

Reference:

HTGR AI?A Status Report: Phase II Risk Assessment, GA-A15000, iC-77, April 1978. ,

s i a

l I

i l

e

, , , , . _ . , - . , . , , , , - , -r - . . . - - - -, . - ., _ . . , . , - , . . , ...n.. .,

w-..v. . ~ . , ,

. II.B.4 7 rate, the initial portion of Figure 2 would be applicable to all the sequences shown on Figure 1. Also shown by Figure 2 is the fact that the reactor can possibly reach a stable operating condition after DBA-1 without ever regaining the use of the circulators. Although the system was deprassurized during the early part of DBA-1 to limit heat transfer to the liner, af ter about 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months the outer surface temperatures of the core are high enough such that radiant heat transfer directly to the liner exceeds the core heat generation race, which by 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months is din 1nished by natural decay and by loss of volatile fission products (i.e. When temperature exceeded the 3100*F fuel particle coating rupture temperature).

Liner cooling flow is assumed available throughout DBA-1. Radiological consequences of DBA-1 could be very light if the PCRV depressurization is completed before a significant number of fuel particles begin to rupture.

In addition to high temperature effects on the reactor core, a severe accident annlysis for FSV must consider also the high temperature effects on the PCRV.

Concrete is clearly not a high temperature material. As shown on Figure 3, the concrete is protected from the high temperature helium envireament by a 3/4 inch thick carbon steel liner and varying thicknesses of insulation (depending on design heat load for each location). The liner is cooled by water flowing through tubes ,

welded to the outer surface of the liner. Table 1 gives high temperature properties of selected PCRV components.

The PCRV is depressurized* to 5 psig during DBA-1 in order to protect the liner and PCRV. The concern is that, without depressuri:stion, natural circulation convectivc heat transfar from the hot coolant would cause failure of the 1/4 inch thick insulation cover plates in the top head region. The temperature from hotter refueling regions can exceed 2000*F after LOFC. If the cover plates failed, the insulation could then drop away, exposing the liner to excessive temperature and possibly causing liner failure and degradation of the PCRV concrete.

This situation would not be self-limiting. Any degradation products could fall direc;1y onto the top of the core, further complicating the ci.ancas for recovery.

If PCRV degradation were sufficient to cause failure of the vessel, there would be a sudden releass of the PCRV contents. This is the cause of the much more i severe sequences shown on Figure 1: in order to prevent severe consequences it ;

is necessary to protect the PCRV. If the PCRV failure occurs at high pressure then there is a much greater chance of a large rapid release.

To miniace radiological consequences and liner damage, the depressuri:stion ;

sust start within 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months of the loss of forced convection, and be completed over a time period of about 7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months . .

i

II.B.4 8 ,

a i

Although the chance of severe ac:idents as shown on Figure 1 may be diminishingly small (especially for sequences 2,3, and 4), these ara exanyles of the sequences that must be considered in the requirement for training for cont.ol and mitigation of degraded core accidents at FSV. The quantitative -

aspects of each major sequence must be investigated sufficiently such that time -

to reach various branch points is known, and temperatures of critical system components are known from best estimate computer analyses. For any given sequence written operational recommendations should be available concerning: ,

1. Desired (or likely) stable end condition

2. Recommended recovery strategy as a function of time when failed systems become available. '

Recovery strategies are important because it is more likely that capabilities will be restored during an accident, especially for a slow-responding HTGR system. 3 In addition, it is very improtant that restored systems be brought on in such i a manner that damage is not caused by thermal shock or other unanticipated effects.

In order for the operators to make proper mitigation and control decisions they mus: ba able to accurstely assess the condition of the reactor core and the PCRV. Operational training and written procedures should specify how to do this. I Topics covered should include:

1. How the core outlet ther=occuple readings relate to interior core temperatures during severe heat-up transients. !

2. Which radiation moniotrs can be used to estimate release of fission :

products from the fuel into the helium coolant within the PCRV. ,

3. Effects of high radiation levels on instrumentation readings.

4. How the condition of the PCRV liner and concrete can be determined.

Clearly, the liner cooling water tube outlet temperatures are valuable, f especially if there is forced cooling water flow. If forced flow were interrupted, these outlet temperatures might provide indication of liner heat-up by measuring the amount of superheat in stean produced as the contenta of the cubes boil off.

5. The effect of liner heat-sp on the nuclear instrumentation ind1 cations u; a function of PCRV temperature.

l IV. Summarv of Conclusions and Recommendations-Short Range ;

The requirement for training for control and mitigation of degraded core l evenes can best be met by development of contingency precedures and background information on:

e

II.3.4 9

1. Responsa of plant instrumentation during severe accident sequences, including the effects of accident conditions such as high radiation and temperature.

2. Use of plant instrumentation to assess the condition of the core and PCRV and to determine the amount of fission products released from the core to the reactor coolant.

3. Optimum use of plant systems as a function of ti=e in severe accident sequences.

4. llecessary special proceduras for startup of failed plant systems if they do tecoma available at some point in a potentially severe accident sequence.

Long Range A sufficient d'aca base of calculational and other investigative results does not exist to permit adequate ctusideration of the full range of severe accident sequences possible at FSV. Therefore, a task of assessment of accident sequences specific to FSV should be initiated. The AIPA study conducted at gal should provide an excellent starting point for such a study, but will not of itself satisfy this I

requirement because it:

1. Is specific to the 3000 W HTGRs and not to the 842 W FSV HIGR.

2. Is concerned more with the calculation of overall statistically expected radiological consequences of severe and non-severe accidents rather than consequences of severe accident sequences.

V. Proposal for Future Work at ORE, 1

Short Term

Future work at ORNL should consist of a review of the procedures for training for control and nitigation of degraded core accidents developed at FSV. After this review, and after consultation, as necessary, with NRC, PSC and GA personnel then ORNI. can: nake final recom=endations about the adequacy of the training procedures.

1"HTGR Acciden Initiation and Progression Analysis Status Report," Volumes I thru VII, GA-A13617 and GA-A15000, October 1975 through April 1973.

e

II.B.4 10 Long Range As. indicated in a previous se.tien, there is a need for continuing study ,

of FSV severe accident sequences. ORNL can contribute to this study in a number of ways:

1. Futher study of severe accident sequences possible at FSV (pact of an existing RSR-Sponsored program).

2. Review of GA models used for calculation of thermal-hydraulic and radiological consequences of severe accidents.

3. Modification of ORNL thermal-hydraulic codes to allow calculation of severe accident effects on the PCRV and liner cooling system.

~

4. Study of and development of computer methods for prediation of radiation transport during severe ETGR accidents.

5. Performance of calculations to check predictions of severe accident consequences calculated by GA for FSV.

4 d

I

DRAFT S. J. Ball Fcr Commento 6/8/31 EVALUATION OF TMI ACTION PLAN APPLICATION TO FORT ST. VRAIN Requirement II.K.3.17: Report on Outages of Emergency Core Cooling Systems I. Background The NRC requirement for licensees to report historical data (for the last 5 years) on cumulative outage times for ECCS components is part of a systems reliability analysis task, II.K, in NUREG-0660 (May 1980). The objective of this task is to improve emergency operating procedures and operator training to improve the capability of plants to mitigate the consequences of loss-or-coolant accidents (LCCA) and loss-of-feedwater events. The present requirement was one of the items recommended by the NRR Sulletins and Orders (B&O) cask force. The purpose is to provide NRC with data to evaluate critical component unavailability, are thus determine if tech specs are needed on cumulative outage times. The requirements of the licensee report are given in NUREG-0737, which specifies a Jan. 1, 1981 deadline. PSC responded (P-80441, Dec. 26,1980) by giving outage histories for the PCRV cooling systems (System 46) and the standby diesel generators (System 92), noting that these are the only two ECCS-related systems for which 1

the tech specs permit substantial outage times. Other FSV ECCS systems, such as the circulators and steam generators, are part of the normal plant operation cooling systems. Hence PSC claims that their reliability and availability is continuously demonstrated by plant operation, and so no reports en their unavailability are necessary.

II. Evaluation of the Applicability of Item II.K.3.17 to FSV The Diesel generator emergency power supply system and the PCRV liner cooling system are clearly critical parts of the FSV ECCS. PSC's report on their unavailability stated that neither of the two (redundant) liner cooling systems had any downtime in the last 5 years, so its historical unavailability is zero.

Analysis of PSC data for the two (redundant) diesel generators shewed an average single-unit unavailability due to all causes (including routine =aintenance downtimes) of 0.0088. The average single-unit unavailability due to forced outages was 8.2 x 10~3 .

Hence an approximate historical unavailability for emergency diesel power would be the product of the above nu=bers, 7.2 x 10~7 , or less than one in a million. Other observations about the PSC data: the forced outage ti=es for the diesels ranged frem N2 to 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months ; and there was no apparent significant deterioration with time. '

- - , - . v y ,, - ~ - - ~ - - - - ,,- -c- , - - - , ,r ,, ,--,n - - , 3

II.K.3.17 2 It is recognized that the PCRV liner cooling system is the ultimate ECCS for FSV, and that, except for the use of the firewater system as a backup cooling system, the diesel generators are needed as the emergency power source for the PCRV coolant circulator pumps. However, the main helium circulators and steam generators are also part of the ECCS, and even though they are used for normal pant operation, they have not demonstrated a cero unavailability history during post shutdown periods. Thus in order to more comple'..ely evaluate total ECCS unavailability, historical data on other subsystem and component performance would be useful, including the circulators, the emergency feedwater and firewater systens, and the steam generator normal and emergency cooling water supply systems.

III. Summary of Conclusions and Recommendations

1. Final recommendations will be =ade after further discussions with NRC and PSC.

2. The last-five-year historical record indicates excellent (low) unavailability for the PCRV liner cooling systems and the emergency diesels. No tech spec changes appear to be necessary if the good records persist.

3. To evaluata the total ECCS unavailability, historical downtime data -

is needed for the circulators and steam generators and those subsystems used (including backups) to operate them during decay heat re= oval periods.

IV. Procosed Further Action by ORNL

1. Discuss the need for more data with NRC and PSC.

2. Draw up final recommendations.

l

3. Analyze unavailability data.

\s e

_ -r -

ML20004F270

Text

Reference:

Navigation menu

Search