ML20209H839
| ML20209H839 | |
| Person / Time | |
|---|---|
| Issue date: | 08/06/1986 |
| From: | Stello V NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO) |
| To: | |
| Shared Package | |
| ML20209H803 | List: |
| References | |
| FOIA-86-864, TASK-PII, TASK-SE SECY-86-228, NUDOCS 8702060150 | |
| Download: ML20209H839 (7) | |
Text
_ ,__ ____ - -___ ____ _ ___ - _-___-- - - _
,f*"%
.....l August 6, 1986 POLICY ISSUE Secy.ee.228 (Information)
FOR: The Commissioners FROM: Victor Stello, Jr.,
Executive Director for Operations
SUBJECT:
INTRODUCTION OF REALISTIC SOURCE TERM ESTIMATES INTO LICENSING PURPOSE: To inform the Comissioners of the Staft's approach to the use of new source-term information for future applications (Program Element 3 of SECY-86-76.)
SUMMARY
- The postulated limiting accident currently used by the staff to assess site suitability as well as to evaluate the adequacy of plant mitigation and other safety systems is derived from a 25 year old report (TID-14844) that is now regarded as outmoded.
This paper presents the staff's plan to treat the releases
(" source terms") from core damage and core-melt accidents in a more realistic fashion in licensing future plants. Implementation will be primarily through revisions in the Standard Review Plan and Regulatory Guides. Existing plants would not be affected unless their owners proposed license amendments calling for a review under a revised section of the Standard Review Plan.
To accomplish this, the staff will select a small number of severe accident sequences and will use the source tem code package (SICP) methodology described in NUREG-0956, to compute the rates of release of fission products in the containment during these sequences. The sequences selected will be chosen to represent those severe accidents which, by virtue of their probability, are considered to dominate degraded core and core-melt events. Thus, for future plants, the limiting design basis accident will be derived from a set of core-melt events.
The releases into containment from these sequences will be used to set the performance levels of certain engineered safety features, and to determine containment leakage limits and site acceptability, replacing the assumptions in TID-14844 presently used in the Standard Review Plan.
Emergency planning requirements would not be determined from this and will be dealt with separately following review of the Chernobyl accident.
CONTACT: Zoltan R. Rosztoczy, NRR x28016 l l
I 8702060150 870202 P PDR FOIA POR WILLIAM 86-864 _
g g y - - -;-- 5,; w
i
, ' The Commissioners Extremelylowprobabilitysequences(thosewhosesumfora given plant meet the guidance proposed by Comission for further l
staff consideration for implementation of the safety goal) would.
be excluded from consideration. Sequences involving early con-tainment failure or bypass would be required to be sufficiently reduced in likelihood that they would have an extremely low j probability of occurrence. The remainder would be required to be acceptably mitigated sufficiently to assure that the con-sequences would be below 10 CFR Part 100 guidelines. ,
The source tems, or fission product releases from containment into the environment, will be calculated more realistically than currently. In order to account for uncertainties, explicit margins of safety will be developed and applied to the results of realistic calculations. Present regulations require the use of offsite dose calculations and the staff plans to continue this use. The dose guidelines of 10 CFR Part 100 would be retained as acceptance criteria for judging the adequacy of engineered safety features and siting, but with the possible addition of other organ doses to correct the current domination of thyroid doses due to iodine. For future standardized plants, we expect that dose calculations would be done at the initial application j only to establish acceptability of plant design and site envelope.
BACKGROUND: In response to the severe accident policy statement, the staff developed a plan for introducing new source-tenn information into regulatory use. This plan was presented to the i
Comission as SECY-86-76, and contains three major elements:
(1) the examination of existing plants for risk outliers, (2) defining the role of PRA in new applications, and (3) the modification of rules, guides and practices. Under the third of these elements, SECY-86-76 scheduled two further comission infomation papers: the present paper, specifying the means of selecting severe accidents for use in regulatory requirements, and a second to propose containment perfonnance criteria, due in February, 1987. The approach outlined in the present paper would be implemented in the form of draft Standard Review Plan and Regulatory Guide revisions. By December, 1986, revisions to the two pivotal Regulatory Guides, 1.3 and 1.4, will be proposed.
These along with a technical and policy analysis of the course of action proposed will be submitted to the Comissioners for their approval.
NRC's regulatory philosophy is based upon consideration of both accident prevention and mitigation. Accident prevention, which occupies a major share of the NRC's regulatory activities, relies upon numerous specific plant design and operational requirements to prevent the occurrence of an accident. Despite these measures, NRC regulations also require that appropriate mitigation features must be provided to cope with an accident,
The Commissioners should it occur. The principal accident mitigation feature in current plants is the containment and its supporting systems.
Included among the latter are fission product cleenup systems such as containment sprays and filters. .
To test whether or not the degree of mitigation is adequate under present NRC practice, the applicable regulation (Part 100.11) directs that an applicant " assume a fission product release from the core" and, using the characteristics of the plant mitigation features as well as the site characteristics, determine the radiological doses to hypothetical individuals at selected locations for predetermined time periods. The plant mitigation features together with the site are deemed acceptable if the calculated doses do not exceed the guideline values (25 rem to the whole body or 300 rem to the thyroid) given in Part 100. The fission product release postulated as a test of plant mitigative capability is characterized, in a footnote to Part 100.11, as one which:
"...would result in potential hazards not exceeded by those from any accident considered credible. Such accidents have generally been assumed to result in sub-stantial melt down of the core with subsequent release of appreciable quantities of fission products."
When Part 100 was promulgated in 1962 Technical Information Document (TID)-14844, published separately, was noted as con-taining a procedural method that reflected current practices of the Comission. TID-14844 also contained a number of assump-tions regarding the radiological dose methodology, as well as a quantification of the postulated fission product release from the core into the containment. Over the years, the NRC staff has modified its dose calculation methodology from that given in TID-14844. The staff's current methodology, given in Regalatory Guides 1.3 and 1.4, nevertheless makes use of a fission product release and its escape into the environmint that are essentially unchanged from those given in TID-14844.
Use of the TID-14844 release has not been confined to a deter-mination of plant and site suitability alone. The regulatory applications of this release cover a wide range, including the basis for a) the radiation accident environment for which safety-relatedequipmentshouldbequalifiedb) post-accidenthabit-ability requirements for the control room, c) performance of important fission product cleanup systems such as sprays and filters, and d) post-accident sampling systems and accessibility.
- Regulatory Guides 1.3 and 1.4 retained the same release fractions for the noble gases and iodines as TID-14844, with additional specification of the iodine chemical form. The 1% solid fission products present in TID-14844 was used only to calculate a direct gama dose transmitted through the thin contain-cent walls of 1960-type designs. TID-14844 did not consider the escape of aerosols into the environment.
. The Comissioners DISCUSSION:
It is clear that the " postulated fission product release"
~
required by Part 100 was intended to be representative of a core-melt accident. Although the present assumed release is very substantial and has resulted in a very high level of plant mitigation capability, nonetheless, based upon the large amount of infomation obtained on severe accidents since the publica-tion of TID-14844 almost 25 years ago, the present recipe is no longer compatible with a realistic understanding of severe acci-dents. Its assumption of an instantaneous release into contain-ment is physically impossible and may lead to unduly conservative i requirements in areas such as the time required for containment isolation systems to actuate. At the same time, the present formulation may be non-conservative in that it considers the release into the environment of only the noble gases and iodine isotopes, neglecting other important isotopes likely to be
- released, such as cesium. Because of this, some engineered safety features designed to mitigate dose consequences may not be optimally designed with regard to the actual severe accident environment that our more recent research leads us to expect within containment. In addition, the present formulation may be inconsistent since the containment temperature and pressure design conditions are derived from that of a major pipe rupture, while the radiological environment is more nearly that of a core-melt. For these reasons, the staff considers it t.ppropriate to replace this obsolete fomulation and to incorporate a more realistic treatment.
Several alternative actions have been considered for a more realistic treatment of fission product releases. These actions range from minor revision of TID-14844 assumptions to a depar-l ture from the use of off-site dose criteria in setting regula-tory requirements. While generic regulatory requirements eliminating the need for dose criteria could be developed for many plant features, the use of dose criteria provides a readily understandable measure by which the efficacy of different plant systems can be compared and tested, and which allows for possible plant innovation as well. The approach selected for revising current practice will retain the use of dose consequences as acceptance criteria, but will use realistic calculations perfomed for important severe accidents for each type of plant to establish time-dependent fission product inventories within containment, rather than the non-mechanistic instantaneous release assumed in TID-14844. To retain the advantages of simple assumptions currently enjoyed, simplified tables of limiting fission product releases will be developed for use in setting such requirements as containment leak rate, engineered safety feature performance, and equipment qualification.
l Emergency planning involves a major policy question as well as technical considerations and will be dealt with separately.
l
'. The Commissioners Current practice uses assumed fission product releases into !
enclosed plant volumes as tests of design acceptability, but does not use physical and chemical processes in realistic models of fission product behavior. The availability of the source term code package (STCP) methodology permits the sta'ff to use realistic models to examine predicted accident sequences, and to compute the effects of processes by which fission products would be dispersed in an accident and of processes that would retain them. The STCP alone, however, does not constitute a replacement for TID-14844 without additional specification of the accident sequences to be used in computing off-site consequences. The staff must supply new regulatory guidance to define the sequence selection, and new review plans to describe the computation of in-plant transport and off-site consequences and the consideration of uncertainties.
The staff will select a set of diverse accident sequences to represent the spectrum of severe accidents at each plant. For new applications selection of the sequences would be based upon the plant-specific PRA. For existing licensees wishing to use this new approach in an application for a license amendment, sequences could be selected from the applicable reference plant evaluation, using appropriate sequence selection criteria. The staff will permit applicants and licensees to use the STCP or equivalent approved methods to calculate source tenns for their plants.
To adequately represent plant hazards, the chosen sequences must include all likely paths to core-melt. In addition, the staff will examine other sequences and also include any found to have low assigned probabilities only because of assumptions made in the analysis, such as thermal-hydraulic success criteria or an assumed high probability that operator actions would be success-ful. It is expected that for the purpose of source term calcu-lations a few of the selected sequences might serve to envelop the remainder, which will permit some simplification. Simulations of these sequences would be used to fill the roles currently held by the TID-14844 assumptions.
Extremely low probability sequences would be excluded from con-sideration. Such sequences would be those, which when summed for a given plant, do not exceed the guidance proposed by the Commission for further staff consideration for implementation of the safety l goal. For future plants, all sequences either would be prevented sufficiently to be excluded on grounds of low probability, or would be mitigated sufficiently to meet the Part 100 dose guidelines using realistic dose calculations. Existing plants have had their containments designed to meet the temperature and pressure conditions i
l
- )
The Comissioners i of the loss of coolant accident (LOCA) described in 10 CFR Part 50. ,
In general, they would not be expected to meet the same require- 1 ments as future plants, and the backfit rule would be applied to l determine the required degrees of prevention and mitigation.
Rather than the instantaneous release and transport assumed in TID-14844, the STCP methodology will be used to obtain more realistic rates at which fission products are released into the containment atmosphere. The STCP methodology derived rates will be input to an existing staff licensing computer code which models the transport and removal of fission products within containment from both natural deposition as well as cleanup systems such as sprays and filters, and which also calculates off-site doses from assumed containment leakage. The results of this evaluation will be used to judge the performance of engineered safety features and the adequacy of plant design and siting, in compliance with 10 CFR Part 100. The heat energy, gases, ar.d aerosols also released in these sequences will be similarly used in realistic calculations of the temperatures, aerosol loadings and pressures within the plant for use in setting perfomance capabilities and in assessing the rate of fission product leakage into the environ-ment.
The staff intends to revise the applicable regulatory guides and standard review plans in order to make the calculations of off-site radiological consequences more realistic. The dose guide-lines of 10 CFR Part 100 would be retained as acceptance criteria for judging the adequacy of engineered safety features and siting. However, since Part 100 considers only doses to the whole body and thyroid, it may be necessary to add other organ doses, such as bone and lung and corresponding dose pathways such as from ground contamination, as well. The authors of TID-14844 avoided considering uncertainties by deliberate conservatism in all factors used in dose calculations. By the estimates of those conservatisms given in TID-14844 itself, the calculated doses may have been increased by three or more orders of magnitude in some applications. The new approach will avoid the non-mechanistic j
artificial constructions which prompted such conservatisms.
However, there are still large uncertainties associated with the i
' STCP methodology and with the calculation of consequences. Also, some severe accident phenomena have not yet been modeled, and, consequently, are not accounted for in the STCP. Consequently, the results of realistic calculations will have applied to them an explicitly stated margin of safety to account for possible uncertainties and for incompleteness.
The approach outlined above will change the nature of the severe accidents against which mitigative engineered safety features of l
+
o' The Commissioners future plants are to be judged. This change will provide for more direct assurance of protection against potential severe accidents than could be achieved by the use of arbitrary assump-tions, and will rely upon realistic calculations using modern techniques, and appropriately conservative safety margins. At the same time, however, the new approach will retain some of the simplicity of application of the old approach.
C
/5Wi-ctor Stello, J .
Executive Direc or for Operations