ML20203E616
ML20203E616 | |
Person / Time | |
---|---|
Site: | Seabrook ![]() |
Issue date: | 04/30/1986 |
From: | Kreslyon Fleming, Torri A, Woodard K PLG, INC. (FORMERLY PICKARD, LOWE & GARRICK, INC.) |
To: | |
Shared Package | |
ML20203E591 | List: |
References | |
PLG-0465, PLG-465, NUDOCS 8607240188 | |
Download: ML20203E616 (67) | |
Text
, - -
I i
l PLG-0465 i
l SEABROOK STATION I EMERGENCY PLANNING SENSITIVITY STUDY t
I i
i i
I '
Prepared for NEW HAMPSHIRE YANKEE DIVISION [
PUBLIC SERVICE COMPANY OF NEW HAMPSHIRE Seabrook, New Hampshire April 1986 (
I l 1
060724018e 860721 P Dit ADOCK 05000* 3 a
Pickard.,Lowe andGarrick,Inc.
Engineers e Applied Scientists e Management Consultants !
Newport Beach, CA Washington, DC
e i
PLG-0465 I
I l SEABROOK STATION EMERGENCY PLANNING SENSITIVITY STUDY I
I by I Karl N. Fleming Alfred Torri Keith Woodard I Jackie Lewis Thomas J. Miksch!
I I
I Prepared for I NEW HAMPSHIRE YANKEE DIVISION PUBLIC SERVICE COMPANY OF NEW HAMPSHIRE Seabrook, New Hampshire April 1980 I
I I
I Pickard,Lowe andGarrick,Inc.
Engineers e Applied Scientists e Management Considtants Newport Beach, CA Washington, DC
CONTENTS Section Page 1 INTRODUCTION 1-1 1.1 Purpose 1-1 1.2 Background 1-1 1.3 Objectives and Basis 1-4 1.4 Report Guide 1-4 1.5 References 1-4 2 RESULTS AND CONCLUSIONS 2-1 2.1 Sensitivity Results 2-1 2.2 Conclusions 2-4 3 OVERVIEW 3-1 3.1 Overview 3-1 3.2 Uncertainty Distributions for Release Categories S2 and S6 3-1 3.3 Results for Release Categories S1, S3, SS, I and S7 3.4 References 3-3 3-4 4 SOURCE TERMS 4-1 5 SIIE ANALYSIS 5-1 I 5.1 Review of SSPSA Site Model and Modifications for this Study 5.2 CRACIT Postprocessor Function S-1 5-3 5.2.1 Assessment of Dose as a Function of Distance b-3 I. 5.2.2 Conditional Cumulative Distribution Functions b-4 5.3 References b-4 APPENDIX A: CONDITIONAL DOSE VERSUS DISTANCE PLOTS A-1 APPENDIX B: CONDITIONAL HEALTH EFFECTS PLOTS B-1 I 1380P040186 iii
I
- 1. INTRODUCTION 1
1.1 PURPOSE This report presents the results of a sensitivity study to help evaluate emergency planning options for Seabrook Station. In a companion study I (Reference 1-1), a comprehensive risk-based evaluation was performed that supported an emergency planning zone (EPZ) for immediate protective actions no further than 1 mile from the plant. That evaluation was based, in part, on new insights regarding the nature and magnitude of I radioactive release source terms relative to the source term technology that was used to develop the current generic requirement for a 10-mile EPZ (Reference 1-2). The purpose of this study is to determine the radius of the EPZ that can be justified for Seabrook Station without consideration of any advances regarding the source term methodology since the completion of the Reactor Safety Study (Reference 1-3) in 1975. By I using WASH-1400 source term methodology for the transport, deposition, and release of radionuclides in reactor accidents, the question of where to set the distance for planning of an evacuation can be isolated from current issues surrounding source terms. While source terms are a major f actor in planning for immediate protective actions, the use of extremely conservative source term assumptions in tnis study enables a more focused examination of other factors that play an important role in determining the effectiveness of alternative emergency planning strategies.
1.2 BACKGROUND
This sensitivity study is an extension of the Seabrook Station Risk Management and Emergency Planning Study (RMEPS, Reference 1-1). RMEPS used an updated version of a PRA model of Seabrook Station, initially developed in the Seabrook Station Probabilistic Safety Assessment (SSPSA, Reference 1-4), to evaluate emergency planning options. RMEPS demonstrated that, upon comparison with analyses that had been performed in NUREG-0396 (Reference 1-2) to establish the current generic requirement for a 10-mile emergency planning zone (EPZ), an EPZ distance of 1 mile or less can be justified for Seabrook Station. It was shown in RMEPS that an EPZ of 1 mile or less can be justified without violating any available risk acceptance criteria for emergency planning. RMEPS utilized the most up-to-date information on risk factors at Seabrook Station. A distinguished peer review group concurred with these and I other conclusions reached in RMEPS and found that the conclusions were robust in spite of the large uncertainties inherent in risk-based analyses like the ones in RMEPS and Nb3EG-0396.
l An important factor in the evaluation of emergency planning options is i
the consideration of risk acceptance and evaluation criteria. In RMEPS, the criteria that were used in NUREG-0396 to justify a 10-mile EPZ were j adopted and extended to provide a more complete perspective for evaluating alternative protective action strategies. The NUREG-0396 criteria included a consideration of the risks of radiological doses as a function of distance for three categories of potential accidents: design basis accidents, less severe core melt accidents, and more severe core melt accidents. The quantitative results of NUREG-0396 provided one set 1-1 I 1366P040286
I of numerical criteria that were used in RMEPS. The following additional acceptance criteria were used:
e WASH-1400 risk curves for early f atalities and latent cancer fatalities. WASH-1400 assumed a 25-mile evacuation zone.*
e The SSPSA risk curves, which assumed a 10-mile evacuation zone and 1982-1983 PRA technology.
e NRC safety goals for individual risk and societal risk.
e Effectiveness of alternative protective action strategies as a means of reducing risk levels.
The last criterion required a quantification of risk factors as a function of the extent of protective action; i.e., evacuation and/or sheltering distance. In this way, the incremental benefits of each portion of the EPZ were measured in terms of the amount of risk reduction afforded by that element of the overall protective action strategy. A key result of RMEPS was that the total benefits of evacuation are very small because the risk without evacuation is already very small.
Furthermore, most of the benefits realized by evacuation occur within the first mile. In fact, there is no appreciable risk reduction whatsoever for evacuation beyond about 2 miles.
I Importantly, it was found that no evacuation is needed to meet the safety goal for individual risk with more than two orders of magnitude of margin on the basis of mean values.
In comparison with the NUREG-0396, it was determined that a 1 mile evacuation at Seabrook Station would achieve a lower risk of radiological exposure (for 1, S, 50, and 200-rem whole body gamma doses) than what was perceived from the NUREG-0396 results with a 10-mile evacuation.
The primary reasons why RMEPS was able to justify a substantially reduced EPZ distance for Seabrook Station in relation to that established in NUREG-0396 for all reactors include the following:
e The strength and effectiveness of the large, dry, reinforced concrete primary containment building at Seabrook Station reduces the potential for early releases.
e New data and engineering insights about the initiation and progression of sequences involving interfacing systems LOCA.
e A full-scope risk model that fully accounts for specific and unique features of the Seabrook plant and site and takes advantage of advances in data and modeling techniques that have been made since WASH-1400 and NUREG-0396. Key examples of particular importance to emergency planning are:
- The treatment of time-dependent releases and modeling of site-specific characteristics, such as evacuation paths and speeds.
- A more complete comparison of evacuation models is presented in Section 5.1.
1-2 1366P042486
The assessment of containment failure modes and pressure capacity.
The quantification of uncertainty in the risk estimates.
o Results of experimental and analytical research sponsored by the U.S.
Nuclear Regulatory Commission (NRC) (Reference 1-5) and IDCOR (Reference 1-6) that provide an enhanced basis for assessment of radioac.tive material release source terms for a wide spectrum of accident sequences.
RMEPS performed a number of sensitivity calculations that demonstrated the insensitivity of the main conclusion of the study--that an EPZ distance of 1 mile or less was as well justified as the 10-mile distance in NUREG-0396--to assumptions regarding source terms and other areas of uncertainty. Uncertainties in source terms were addressed in RMEPS by defining two sets of source terms for each accident release category.
Six such categories were defined to account for the full spectrum of accident sequences that were analyzed in the risk model . One set of source terms was developed using either the MAAP computer code of the IDCOR program or from the SSPSA, and this set was used to characterize the best estimate results. A second set of source terms was defined from more conservative source terms developed in the SSPSA to establish a conservative upper bound. This conservative set was comparable to source terms published in NUREG-0956 (Reference 1-7). The peer review group I agreed that this set of source terms, and even the " realistic set," used greater releases than those that would be currently regarded as realistic. RMEPS showed that an EPZ distance of 1 mile or less could be justified using either the conservative source terms, the realistic source terms, or their probabilistically weighted average.
The use in RMEPS of conservative source terms and the meeting of all available decision criteria by wide margins by and large divorced the question of where to set the EPZ distance for Seabrook Station from source term issues. However, even the conservative source terms in RMEPS I account for some of the new research in source term technology. That brings us to the purpose of this study--to determine the extent to which the conclusions of RMEPS are dependent on any new source term I technology. Conversely, the purpose is to fully isolate the separate contributions of new source term technology and the other factors cited above to the RMEPS conclusions.
From a purely technical viewpoint, it is important to account for all important risk factors in the development of an emergency plan. A conscientious effort to account for such factors was made in RMEPS. In this study, the key calculations performed in RMEPS are redone using the WASH-1400 source term methodology (Reference 1-8). Using the WASH-1400 source term methodology is meant to imply that the MARCH and CORRAL codes are used to quantify the release of radionuclides, while taking advantage of the models in these codes to represent Seabrook-specific design features. While it is not realistic to ignore the results of all the source term research that was performed in the past decade, the purpose of this study is to determine the extent to which a shortened emergency 1-3 1366PO42286
I plan radius can be justified without any reliance on new source term technology since publication of WASH-1400. If such a justification can be made, the licensing and regulating authorities can more easily separate the issues surrounding source terms from their review of the Seabrook Station emergency plan and operating license application.
1.3 OBJECTIVES AND BASIS The objectives of this study are to determine which emergency plan protectivo action strategies evaluated in RMEPS meet the risk acceptance I criteria used in RMEPS on the basis of WASH-1400 source term methodology. These strategies include no immediate protective actions (i.e., no evacuation), 1-mile evacuation, 2-mile evacuation, 10-mile I evacuation, and 2-mile evacuation with sheltering out to 10 miles. The basis for this sensitivity analysis is to parallel WASH-1400 and NUREG-0396 computations as closely as possible without compromising the enhanced features and plant-specific capabilities of the Seabrook risk model. For consistency with WASH-1400, median risk curves based on median accident frequencies and best estimate consequence modeling assumptions are used for comparison purposes.
1.4 REPORT GUIDE The outline of this report parallels that of RMEPS. The results and conclusions are presented in Section 2. RMEPS developed mean point estimate accident frequencies for the risk model update Uncertainty distributions were developed in this study to determine the median values I for accurate comparison with NUREG-0396 and WASH-1400. These results are in Section 3. In Section 4, the basis for the source terms used in the study and how they differ from those developed in RMEPS are described.
Site model calculations are summarized in Section 5.
1.5 REFERENCES
1-1. Pickard, Lowe and Garrick, Inc., "Seabrook Station Risk Management and Emergency Planning Study," PLG-0432, prepared for New Hampshire Yankee Division, Public Service Company of New Hampshire, December 1985.
1-2. Collins, H. E., et al., " Planning Basis for the Development of i E State and Local Government Radiological Emergency Response Plans in E Support of Light Water Nuclear Power Plants," prepared for the U.S.
Nuclear Regulatory Commission, NUREG-0396, December 1978.
1-3. U.S. Nuclear Regulatory Commission, " Reactor Safety Study: An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," WASH-1400, NUREG-75/014, October 1975.
1-4. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic Safety Assessment," prepared for Public Service Company of New Hampshire and Yankee Atomic Electric Company, PLG-0300, December 1983.
l 1-4 1366PO40186
I 1-5. Gieseke, J. A., et al ., "Radionuclide Release under Specific LWR Accident Conditions," Volume VI, "PWR Large, Dry Containment Design (Zion Plant)," Battelle Columbus Laboratories, BMI-2104, July 1984, 1-6. Technology for Energy Corporation, " Nuclear Power Plant Response to Severe Accidents, IDCOR Technical Summary Report, November 1984.
1-7. U.S. Nuclear Regulatory Commission, " Reassessment of the Technical Bases for Estimating Source Terms," NUREG-0956, draft report, July 1985.
1-8. Burian, R. I., and P. Cybulskis, " CORRAL-II Users Manual," Battelle Columbus Laboratories, January 1977.
I I
I I
I I
l l
1 l
1-5 1366P040186
I I 2. RESULTS AND CONCLUSIONS A total of five emergency protective action strategies were selected for evaluation at Seabrook Station. These include:
e No Immediate Protective Actions (N0 EV)
I e e
e 1-mile Evacuation (1 EV) 2-mile Evacuation (2 EV) 10-mile Evacuation (10 EV) e 2-mile Evacuation (2 EV) with Sheltering out to 10 Miles (10 SH)
All five of these strategies were evaluated in RMEPS. In this sensitivity study, it was decided to initially develop results for the I first three cases (N0 EV,1 EV, and 2 EV); then, based on these results, the merit of running the remaining two cases would be evaluated.
then determined that the objectives of this sensitivity study could be It was fully met without evaluating the 10 EV and 2 EV/10 SH cases; therefore, only the results of the first three cases are presented in this report.
Following the presentation of the numerical results in Section 2.1, the conclusions of the study are set forth in Section 2.2.
2.1 SENSITIVITY RESULTS In the RMEPS analysis, a full quantification of uncertainties was performed in the risk assessment calculations, including uncertainties in the definition of the source terms, the modeling of consequences, and the estimation of accident frequencies. What is being performed here is a sensitivity calculation that addresses the isolated effect of advances in source term technology apart from other factors relevant to the emergency plan at Seabrook Station; i.e., the inherent capabilities of the systems and containment features specific to Seabrook, the unique site characteristics, evacuation routes, etc. In this sensitivity study, the following results were obtained, e A point estimate for the early fatality risk curves (frequency of exceedance),
e A point estimate for the average individual risk within 1 mile of the site boundary for comparison to the safety goal.
e Point estimates of the frequency of exceedance of dose and distance curves (of the type performed in NUREG-0396) for 1, 5, 50 and 200-rem whole-body doses.
In order to focus on the specific question of how much the justification for a reduced EPZ depends on new source term technology and to be as consistent as possible with NUREG-0396, the following assumptions were I made in the point estimates.
e Median values of accident frequencies were used.
1373P040186
I I e A single set of source terms, based on those calculated in the SSPSA, using the MARCH AND CORRAL codes was used; i.e., one source term for each accident release category. A comparison of the source terms used in this study to those used in RMEPS is presented in Table 2-1.
e A single set of best estimate consequence modeling assumptions was made to develop one conditional risk curve for each accident release category.
e Radiation doses to individuals were accumulated from time of evacuation warning out to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
The above set of assumptions can be regarded as a best point estimate calculation in respect to all factors except source terms. It was judged I inappropciate to superimpose an uncertainty analysis on the source term sensitivity evaluation. Moreover, the value of performing an uncertainty analysis in this study would be questionable because the authors do not lend any credence to the validity of the WASH-1400 source term methodology for predicting realistic source terms in view of the extensive new evidence made available during the last decade. We simply seek to determine the singular effect of the use of that methodology on the results of these calculations. Hence, best estimate point values were used to address all nonsource term-related factors in the sensitivity study. Readers are referred to the RMEPS report for a full I probabilistic cvaluation of uncertainties in accident frequency, source terms and consequence modeling related factors.
I The results for the early fatality risk curves are presented in Figure 2-1. For no immediate protective actions, the sensitivity results for Seabrook Station fall outside the WASH-1400 curve for which a 25-mile evacuation zone was assumed.* However, it would be reasonable to assume I that the comparison would be more favorable for the Seabrook NO EV case if WASH-1400 had assumed, say, a 10-mile evacuation. When a 1-mile evacuation is assumed for Seabrook Station, the risk curves cross, indicating a higher risk of a smaller number of health effects and a lower risk of a larger number of health effects for the Seabrook Sensitivity Study in relation to WASH-1400 results. The mean value of I the risk distribution (i.e., the area under the risk curve) for a 1-mile evacuation at Seabrook is comparable to that for WASH-1400. This result is quantified below in terms of the safety goal comparison.
It is important to note when comparing the results for Seabrook Station against WASH-1400 that a more complete treatment of dependent events such as external events, internal plant hazards and interactions, and system common cause events were included in the Seabrook results. In fact, seismic events, which are not included in the WASH-1400 curves, make a significant contributicn to the Seabrook results. If the WASH-1400 results were enhanced to account for the same level of dependent events as the SSPSA, these comparisons would be more favorable for each of the Seabrook cases evaluated.
I *A more complete comparison of evacuation models is presented in Section 5.1.
2-2 1373P042486
g In the RMEPS analysis, it was shown that the individual risk safety goal 3 was met, even for the case of no immediate protective actions. As shown in Figure 2-2, the safety goal can still be met with WASH-1400-type source terms and best estimate assumptions on other risk factors without taking credit for any immediate protective action, such as evacuation.
The chief difference is that the RMEPS safety goal margin of more than two orders of magnitude is reduced to less than a factor of three for no evacuation. This margin is increased considerably with a 1-mile evacuation to more than an order of magnitude.
The final set of results developed in this sensitivity study are the comparative plots for the NUREG-0396 dose and distance risk curves. The 200-rem and 50-rem curves are compared with the corresponding results from NUREG-0396 that formed part of the basis for the current 10-mile EPZ for all plants. In the sensitivity results for Seabrook Station, the frequency of exceedance of 200 rem is less at 1 mile than the corresponding result from NUREG-0396 is at 10 miles, as illustrated in Figure 2-3. There is no significant frequency of exceeding 200 rem beyond 1.5 miles in the Seabrook sensitivity results. The corresponding distance from NUREG-0396 is 15 miles.
It is important to note that a conservative approach was used in this study and in RMEPS to model evacuation cases of less than 10 miles. For these cases, the evacuation speed was not corrected to account for the reduction in road traffic in comparison with the same population segment in a 10-mile evacuation. Hence, we did not quantify the added positive benefit of speeding up the close-in part of the evacuation when the EPZ is shortened.
The NUREG-0396 comparison for the 1-rem and 5-rem doses is presented in Figure 2-4. Again, the risks of exceeding these doses are lower at 1 mile for Seabrook than are the corresponding results from NUREG-0396 at 10 miles. Hence, as with the safety goal comparison, these sensitivity results, although exhibiting smaller margins than in RMEPS, provide further confidence that the conclusions of RMEPS are insensitive to source term assumptions.
It was fully recognized in RMEPS that there are several key factors in I addition to updated source term technology that provide a different perspective on the risk reduction benefit of emergency planning options for Seabrook Station than can be inferred from NUREG-0396. The most important of these nonsource term-related factors can be explained from Table 2-2, which is reproduced from RMEPS. While the core melt frequency in the updated Seabrook Station results is somewhat higher (because of a more complete assessment of dependent events and component failure rates), the percent of core melts of principal concern in emergency planning (i.e., those with an assumed gross early containment failure or bypass) is more than 300 times lower for Seabrook Station. This difference, in turn, stems from three major factors:
e The effectiveness of the Seabrook Station primary containment to either remain intact or to maintain its fission product retention capability for periods much longer than required for even delayed, ad hoc protective actions.
2-3 1373P040186
I e A more realistic assessment of the strength and failure modes of the Seabrook containment than was possible within the state-of-the-art of PRA when WASH-1400 was completed. ;
I e A more realistic treatment of the initiation and progression of I interfacing systems LOCA sequences in RMEPS.
What has been demonstrated in these sensitivity analyses is that the above factors are sufficient to justify an EPZ of 1 mile or less for Seabrook Station without taking account of a single bit of what has been learned about source terms in the past 10 years. In this regard, the authors wish to emphasize that these results should not be regarded as a replacement for, but rather as an extension of, a more realistic and complete evaluation of emergency planning options in RMEPS.
2.2 CONCLUSION
S The principal conclusion of this sensitivity study and RMEPS is that an emergency planning zone at Seabrook Station of 1-mile radius or less is more fully justified in terms of its risk management effectiveness than the current 10-mile EPZ was justified by the results of NUREG-0396. In RMEPS, this conclusion was based on the most current information on all the risk factors of importance in emergency planning at Seabrook Station including the most up-to-date PRA technology regarding source terms. In this sensitivity study, a 1-mile EPZ can still be fully justified without accounting for any new insights about source terms since WASH-1400.
Because of significant differences between this study and RMEPS regarding source terms, the absolute numerical results are correspondingly different. The principal difference is that the margins between the results and the acceptance criteria are smaller in this study. It was I determined here, on the basis of WASH-1400 source term methodology and best estimate assumptions regarding other risk factors, that:
e The individual risk of early fatalities in the population within 1 mile of the site boundary with no immediate protective actions is less than the NRC safety goal. This individual risk is substantially less when a 1-mile evacuation is assumed.
e The risk of early fatalities with a 1-mile evacuation is comparable to the WASH-1400 results, which assumed a 25-mile evacuation. The I Seabrook Station results for a 2-mile evacuation are substantially less than those for WASH-1400.
I e The risk of radiological exposures for 1, 5, 50 and 200-rem whole body doses with no immediate protective actions is less at 1 mile than the corresponding NUREG-0396 results at 10 miles.
At the inception of RMEPS, many in the industry felt that advances in l source term technology since WASH-1400 dictated a reexamination of the
! risk bases for the current generic requirement for a 10-mile EPZ. While the results of RMEPS demonstrated the validity of this assessment for the Seabrook Station, the results of this study provide a new insight that was not fully appreciated. The new insight is that advances in other 2-4 1373P040186
I I areas of PRA technology and the consideration of plant-specific safety features at the Seabrook Station have proved to be every bit as important. The most important nonsource term-related differences between the Seabrook Station risk model and the one used in NUREG-0396 are the I same ones that provide a more than two orders of magnitude decrease in the frequency of accidents of primary interest in emergency planning; i .e., core melt sequences with an early gross containment failure or bypass. The factors responsible for this decrease include the strength of Seabrook's primary containment building to withstand overpressurization, a more realistic assessment of this strength and containment failure modes, and a more realistic treatment of the interfacing systems LOCA class of scenarios.
I I
I I
I 2-5 1373PO40186
W W M TABLE 2-1. COMPARIS0N OF RMEPS AND WASH-1400 METHODOLOGY SOURCE TERMS FOR SEABROOK STATION RELEASE CATEGORIES Release Description
- Release Time 6" 10 Calories Category Start Duration Warning Per Second Xe I Cs Te Sr Ru La S
j This Study - 24 Hours 2.5 0.5 1.0 12 0.9 .7 .5 .3 .06 .02 .004
- Total 2.5 0.5 1.0 12 0.9 .7 .5 .3 .06 .02 .004 Containment RMEPS - Conservative 1 14 .5 <10 0.9 .135 .135 .032 .01 6 0056 6(-4)
Failure - Best Estimate 2 12 1 <10 0.9 .052 .052 .013 .006 005 2(-4)
S This Study - 24 Hours 5 24 0.5 0 .12 .010 .20 19 .022 01 5 0025 2
- Total 5 76 0.5 0 1 022 .31 .32 .034 .025 .0042 Early RMEPS - Conservative 5 51 .6 <10 1 .025 .025 .008 .003 .0018 3(-4)
Leakage - Best Estimate 13 76 5 <10 1 .013 .013 .004 .002 9(-4) 1(-4)
S This Study - 24 Hours 6 24 2 0 5(-4) 4(-5) 1.7(-4) 1.5(-4) 1.9(-5) 1.2(-5) 8(-6) 3
- Total 89 0 74 0 1 .017 .024 03 .0026 0023 4(-4)
Late RMEPS - Conservative 54 0 42 <10 1 .002 .002 .01 2(-4) 2(-4) 3(-5)
- Best Estimate 89 0 74 <10 1 001 .001 .002 1(-5) 1(-5) 1 (-5 )
S This Study - 24 Hours 4.3 24 .6 <10 .014 5(-7) 5(-7) 1(-7) 6(-8) 2(-8) 2(-9)
S
- Total 4.3 24 6 <10 .014 5(-7) 5(-7) 1(-7) 6(-8) 2(-8) 2(-9)
Containment RMEPS - Conservative 2 24 4 <10 014 5(-7) 5(-7) 1(-7) 6(-8) 2(-8) 2(-9)
Intact - Best Estimate 4.3 24 .6 <10 .009 4(-8) 4(-8) 6(-9) 4(-9) 1(-9) 1 (-10)
S This Study - 24 Hours 2 14 1.5 0 0.9 .19 43 40 .047 033 005 6
- Total 2 14 1.5 0 1 .19 .43 40 047 033 005 Open RMEPS - Conservative 2 12 1 <10 1 .052 .052 .033 .062 .005 2(-4)
Purge - Best Estimate 4 16 3 <10 1 .01 .01 3(-4) 6(-4) 6(-5) 6(-5)
S This ! '1y - 24 Hours 8.5 7 2 0 0.9 7(-4) 5(-4) 3(-4) 6(-5) 2(-5) 4(-6) 7
- Total 8.5 7 2 0 0.9 7(-4) 5(-4) 3(-4) 6(-5) 2(-5) 4(-6) p p RMEPS - Conservative 8.5 7 2 <10 1 094 .094 .083 2(-4) 4(-4) 4(-4)
Bypass
- Best Estimate 8.5 7 5.5 <10 1 3(-4) 3(-4) 2(-4) 1(-6) 3(-6) 3(-6)
- Source terms for this study are given for the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after warning and for the total release. Source terms from RMEPS are given for total release only.
NOTE: Exponential notation is indicated in abbreviated form; i.e., 6(-4) = 6 x 10-4 1367P040286
l I
I TABLE 2-2. COMPARISON OF CORE MELT FREQUENCIES AND DISTRIBUTIONS OF RELEASE TYPES I Risk Parameter WASH-1400 PWR SSPSA Updated Results*
e Mean Core Melt Frequency (events 9.9-5** 2.3-4 2.7-4 per reactor-year) e Percent Contribution of Release Types Gross, Early Containment 34 1 0.1 Failure Gradual Containment 66 73 60 Overpressurization or Melt-Through Containment Intact 0 26 40
- Based on RMEPS (PLG-0432).
- Based on WASH-1400 uncertainty ranges.
I NOTE: Exponential notation is indicated in abbreviated form; i.e., 9.9-5 = 9.9 x 10-5, i
~7 1367P033186
10-3 I
LEGEND
--- SEABROOK STATION I $
PEF. RMEPS AND WASH-1400 SOURCE TERM METHODOLOGY g (MEDIAN RESULTS)
I d o 10-5 -
WASH-1400 (MEDIAN RESULTS)
N I e m
E '% NO IMMEDIATE PROTECTIVE g N ACTIONS I $ 10-6 o
N 1 - MILE N
\
EVACUATION \
I W E
\ \
\
\
10-7
{
so \ \
\
I
\
\
10-8 -
2 - MILE k EVACUATION {
\
l RMEPS RESULTS N \
E OFF SCALE N \
\' -
l I
I I I I 10~9 3
100 10 1
10 2 10 10 4 10 5 EAR LY FATALITIES I FIGURE 2-1. COMPARIS0N OF MEDIAN RISK 0F EARLY FATALITIES AT SEABROOK STATION FOR DIFFERENT EMERGENCY PLANNING OPTIONS I
2-8
I I
10-2 Y
I o E 10-3 -
BACKGROUND ACCIDENTAL FATALITY RISK (5 FATALITIES PER 10,000 POPULATION PER YEAR)
O I i b 104 -
I A N 10-5 I
g 3 SAFETY GOAL (.001 TIMES 9 BACKGROUND RISK) 8 THIS STUDY FOR I < 10-6 g
SEABROOK STATION WITH NO IMMEDIATE PROTECTIVE ACTIONS Q 10'I -
$ WITH I MILE I g N
g 10-8 -
EVACUATION RMEPS RESULTS WITH NO IMMEDIATE I z '
< PROTECTIVE ACTIONS
- N 10'9 '"#
I I
l I FIGURE 2-2. COMPARIS0N OF SEABROOK STATION SENSITIVITY RESULTS USING WASH-1400 SOURCE TERM METHODOLOGY WITH BACKGROUND, SAFETY G0AL INDIVIDUAL AND RMEPS RISK LEVELS I
I 2-9 I
I i i i iiiiiil i ii e iig i i i i ie i i.
_~ NUREG-0396 OZ - ----- THIS STUDY FOR _
8o SEABROOK STATION o3
>o
""" *""*"" RMEPS RESULTS FOR
@j SEABROOK STATION l
co _a (200 REM CURVE OFF
$y SCALE) ow ho 0.1 -- -
cU _
E< - -
o2 -
ww _
I% _
w& - \ _
Ok k l \ 50 REM 2
$Ei \pu g _
wo g i 3o 1 Ow \ i Eg 0.01 --
g )
BS : i 1 en -
1 l 80 l
l 1
!e5 I -
c 5.,, i
-\
' I ' ' ' l ' ' ' ' ' ' ' '
1 10 100 1,000 DISTANCE (MILES)
FIGURE 2-3. COMPARIS0N OF SEABROOK STATION RESULTS IN THIS STUDY AND RMEPS WITH NUREG-0396 - 200-REM AND 50-REM WHOLE BODY DOSE PLOTS FOR NO IMMEDIATE PROTECTIVE ACTIONS 1
2-10
I i i i iiisog i i i iiiil i i ii i s i_
_ NUREG-0396 [
oo
- ---- THIS STUDY FOR -
SEABROOK STATION o3 -
oN ~
RMEPS RESULTS Og FOR SEABROOK Ww STATION i
2 1 REM O$ g 1 l_ 5 REM I go eu z<
52
$$ ~3 -
~
ug ,_\_.._._-__----s s -
x.
ww
., L N wo -
. g oz .
% \ -
\
Om -
- \
\
\
55 \
s*
\ \
3o s ow '*.. , , 1
\
EE 10 0.01
\ -
q, \
- .,* ., \ -
z<
o _
~
E$ _ I. \* SREM( \1 REM os i* .
- \
8; s I
uo
- l \
.,, *1 REM l _
5 REM *
- l
., g g,go, i i i i r. i i ll it i i i !i il i i i isiii i 1 10 100 1,000 DISTANCE (MILES)
I FIGURE 2-4. COMPARIS0N OF SEABROOK STATION RESULTS WITH NUREG-0396 -
1-REM AND 5-REM WHOLE B0DY DOSE PLOTS FOR NO IMMEDIATE PROTECTIVE ACTION AND WASH-1400 SOURCE TERM METHODOLOGY I
2-11
- 3. PLANT ANALYSIS RESULTS 3.1 OVERVIEW The purpose of this section is to document the calculation of median release category frequencies. These frequencies were originally calculated in the SSPSA. An update of the SSPSA plant analysis was performed in RMEPS (Reference 3-1). RMEPS included a reassessment of interfacing systems LOCA sequences, some seismically induced sequences with degraded containment isolation, and consideration of additional recovery actions in relation to the SSPSA results. Only point estimates of the mean frequencies of accident sequences and release categories were calculated in RMEPS. To support this sensitivity study, uncertainty distributions for the frequencies of release categories were calculated so that reasonable estimates of the median values of these distributions could be obtained. As explained more fully in Section 1, median frequencies are used for consistency with the calculations that were performed in NUREG-0396. These median frequencies are used in the risk results in Section 2.
The technical approach to the development of uncertainty distributions for accident release category frequencies is fully described in Section 13.1.2 of the SSPSA. The process consists of the development of a simplified model of the dominant accident sequences in terms of an equation that relates accident sequences first to initiating events and system top events, then to underlying parameters in the data base (e.g.,
component failure rates) whose uncertainties have been separately quar.tified in one or more PRA tasks. For example, the data analysis task documented in Section 6 of the SSPSA quantifies uncertainty distributions I for failure rates, common cause parameters, maintenance unavailabilities, and initiating event frequencies. Using a Monte Carlo process, these parameter distributions are propagated through the simplified risk model to obtain uncertainty distributions for accident sequence frequencies.
A summary of the release category uncertainty distribution results is presented in Table 3-1. Two different computer programs are used to develop these distributions, one to analyze sequences initiated by seismic events and the other to analyze nonseismically induced sequences and to combine the overall results. Special software is needed to analyze seismic sequences because the initiating event and component failure frequencies in this case are keyed to a parameter, the peak horizontal ground acceleration. The uncertainty distributions must be integrated over the range of the ground acceleration at the same time the convolution is performed. It turned out that seismic-induced sequences dominated release categories S2 and S6 and made negligible contributions to the remaining categories. The results for S2 and S6 are described in Section 3.2 below. The results for the remaining categories are presented in Section 3.3.
3.2 UNCERTAINTY DISTRIBUTIONS FOR RELEASE CATEGORIES S2 AND S6 The uncertainty analysis for release categories S2 and S6 was done using the SEIS4 (Reference 3-2) computer code. As noted above, release 3-1 1375PO40186
categories S2 and S6 are dominated by seismic events. SEIS4 enables the .
analyst to combine fragility curves for individual elements of the system into a fragility curve for the failure of the system as a whole. System fragilities are combined, as dictated by the scenario, and assembled with the seismicity curve to obtain the distribution for the frequency of i failure for a given state. l The frequency of release category S2 is approximately equal to the sum of the frequencies of plant damage states 1FP, 3FP, and 7FP. Release category S6 is equal to the sum of the frequencies of plant damage I states 3F and 7F. The dominant accident sequences to these plant damage states in terms of the initiating events and system top events are described in Table 3-3 of RMEPS.
The procedure for developing an uncertainty distribution is the same for S2 and S6. The first step is to algebraically reduce the dominant scenarios contributing to a release category to a single equation. To perform this step, it is important to understand the difference between the point estimate quantification, as summarized in Taole 3-3 of RMEPS, and the uncertainty quantification of these same scenarios using SEIS4.
For the point estimate quantification, the seismicity curve was discretized and the plant model analyzed for six different levels of ground acceleration for the seismically induced general transient and ATWS events, and two different acceleration levels for the large LOCA seismic event. The different seismicity levels are accounted for by the convolution of the family of seismicity curves with the component fragility families within SEIS4. Similarly, most of the split fractions for a given top event are essentially a single split fraction. The only difference is due to the discretization of the seismicity curve. After the list of scenarios has been reduced as much as possible algebraically, I the reduced equation must be input to SEIS4 along with the seismicity curve, the fragility curves for components involved in the equation, and distributions for any nonseismic failures involved in the equation.
Table 3-2 gives a list of components or structures whose seismic failure I fractions contributed to release category S2 or S6. Nonseismic failure fractions included in the SEIS4 run for S2 and S6 are listed in Table 3-3.
Table 3-4 shows how the top events were quantified in the uncertainty analysis using SEIS4. For the plant model, it was assumed that all seismic events would cause either a general tran nt event or a large LOCA. A large LOCA will result when components or fail due to the seismic event (see Table 3-2). All other sei iceve@ntsareassumed to cause a general transient event. It is important to note that the rare event approximation is not applicable to the quantification of seismic events due to the magnitude of the failure fractions. This is especially true at the higher ground acceleration levels.
3-2 1375PO42286
The algebraic reduction of scenarios contributing to the frequency release category S2 produced the following equation for input to SEIS4:
S2 = 0G[GT[EFW[(GA)(GB)((1 - SA)(1 - SB)(2 - (RW)(1 + RT))
+ (1 - GA)(1 - GB)SA(1 - RT)((1 - PA)(1 -PB)(RW + (H2)(1 - RW))
+ PA) + (GA1)(GBA)] + (1 - EFW)[(GA)(GB)(2 - RW - RT)
+ (GA2)(WB4) + (GB2)(WA4) + (GA1)(GBA)
+ (1 - GA)(1 - GB)((WA3)(WBC)
+ (SA)(PA)(2 - RT(1 - RW) - RW))))
+ LLOCA[(GA)(GB)((1 - SA)(1 - SB) + SA)
+ (1 -GA)(1 - GB)SA((1 - RW) + PA
The frequency of release category S6 was input as the following equation S6 = GT(1 - OG)SA(1 - PA)(1 - PB)C2[1 + EFW(1 - RW) + RW]
The results of the quantification of release categories S2 and S6 are listed in Table 3-1. Reference 3-3 contains information to aid in locating and identifying the computer files related to the SEIS4 quantification of release categories S2 and S6.
3.3 RESULTS FOR RELEASE CATEGORIES S1, S3, SS, AND S7 The frequency of any release category Sj is first calculated in terms of separate contributors from the plant analysis, which includes summary results of the plant event tree quantification and the containment analysis which includes summary results of the containment event tree I quantification.
Es Id) " f (k)F (k,j) p where Fs(j) = frequency of release category Sj Fp (k) = frequency of plant damage state k Fc (k,j) = conditional frequency that an accident in plant damage state k will result in release category J I
3-3 1375P040286
3 From Table 4-17 in RMEPS, which includes point values of each term in the 3 above equation, it is easily seen that the frequencies of release categories S1, S3, SS, and S7 can be expressed by the following Fs(1) = F(1FV) + 3.77 x 10-10 Fs(3) = F(30) + F(70) + F(8D)
Fs(5) = F(2A) + F(4A) + F(8A)
Fs(7) = F(1FPV) + F(7FPV)
The constant term in the above equation for Fs(1) makes a correction for plant damage states 2A, 3D, 7D, 4A, 8A, and 8D, which make such small contributions individually, that it is not necessary to model them explicitly for the purpose of uncertainty quantification. This term is needed to make the means of the uncertainty distributions match the point estimate means.
The next step is to express the plant damage state frequencies in terms of the frequencies of individual accident sequences, which in turn are expressed in terms of initiating events and system states. The major contributing accident sequences to each plant damage state are presented in Table 3-3 of RMEPS. The first step is to break down the system state I events into equations that relate those states to component parameters, such as failure rates, common cause factors, and maintenance frequencies. Uncertainty distributions for the parameters are developed in Section 6 of the SSPSA. The results of propagating these parameter distributions through the model described above via Monte Carlo sampling using the STADIC4 program (Reference 3-4) are presented in Table 3-1.
3.4 REFERENCES
3-1. Pickard, Lowe and Garrick, Inc., "Seabrook Station Risk Management and Emergency Planning Study," PLG-0432, prepared for New Hampshire Yankee Division, Public Service Company of New Hampshire, December 1985.
3-2. Lin, J. C., and S. Kaplan, "SEIS4 Computer Code Users Manual,"
PLG-0287, Proprietary, December 1985.
3-3. Pickard, Lowe and Garrick, Inc., Computer Data Files, "S2.0VT and S6.0UT; PLG PRIME directory; < PURPLE >SEABR00K.2," March 5,1986.
3-4. Wakefield, D. J., and K. N. Fleming, "STADIC4 Computer Code Users Manual," PLG-0438, Proprietary, October 1985.
I 3-4 1375P040186
M M M M M M TABLE 3-1.
SUMMARY
OF RELEASE CATEGORY FREQUENCY UNCERTAINTY DISTRIBUTIONS Annual Frequency Release Category Lower Bound Median Mean 5th Percentile 50th Percentile Estimate 95th Percentile S1 - Early Containment Failure 1.5(-9) 1.5(-9) 4.0(-9)* 5.2(-9)* 1.5(-9)
S2 - Early Containment Leakage 3.5(-7) 7.5(-6) 2.1(-5) 2.0(-5) 1.0(-4) w S3 - Late Containment
& Overpressurization 5.1(-5) 8.3(-5) 1.4(-4) 1.4(-4) 2. 3( -4 )
S5 - Containment Intact 5.5(-5) 7.7(-5) 1.1(-4) 1.1(-4) 1.8(-4)
S6 - Containment Purge Isolation Failure <10(-10) 1.5(-8) 6.5(-7) 3.2(-7) 4.4(-6)
S7 - RHR Pump Seal Bypass 9.8(-10) 4.5(-9) 6.3(-8) 3.9(-8) 1.4(-7)
- Mean influenced by right tail of distribution beyond the 95th percentile.
NOTE: Exponential notation is indicated in abbreviated form; i.e.,1.5(-9) = 1.5 x 10-9 1376P040186
TABLE 3-2. SEABROOK KEY COMPONENTS FOR SEISMIC ANALYSIS Median Acceleration Capacity Impacted Symbo! Structure / Equipment Top Events
@ Unit Auxiliary Transformers 0.30 0.25 0.62 OG h Switchyard 0.40 0.25 0.54 OG h Steam-Driven Emergency Feed Pump 0.66 0.40 0.56 EFW h 120V AC Instrument Buses 0.75 0.42 0.36 SA, SB h RWST 0.86 0.40 0.33 RW
@ PCC Heat Exchangers 0.99 0.37 0.49 PA, PB h Diesel Fuel Oil Day Tanks 1.03 0.39 0.48 GA, GB h RHR Pumps 1.07 0.34 0.65 LA, LB h Safety Injection Pumps 1.07 0.34 0.65 H2 h Reactor Internals 1.50 0.38 0.44 RT h Diesel Generators 1.51 0.36 0.35 GA, GB h Steam Generators 1.71 0.36 0.39 Large LOCA*
h Reactor Coolant Pumps 1.74 0.35 0.32 Large LOCA*
- Failure causes a large LOCA.
I 1376PO40186 3-6
I TABLE 3-3. NONSEISMIC DISTRIBUTION USED IN THE UNCERTAINTY ANALYSIS OF RELEASE CAfEGORIES S2 AND S6 Distribution Name* Impacted Top Event GA1 GA, GB EP3 GA, GB GB1 GB EP2 GB GBA GB GBB GB EF2 EFW WA3 WA WA4 WA WBC WB WB4 WB PA2 PA P01 (Point Value Only) C2
- Variables represent failure frequency distributions.
3-7 1376P040186
MOM MO TABLE 3-4.
SUMMARY
OF EVALUATION OF TOP EVENTS FOR S2 AND S6 ANALYSIS i Sheet 1 of 2 nr u ors Assumptions for SEIS4 Top Split **" " "*" #" "
Event Fractions Analysis of 52 and 56 Nonseismic Seismic OG (all) OG1 @,@ OG1 ( V ) @V@
OGl = .66-EFW (all ) EF2 @ --
(EF2) V @ l H2 H27. H28 H21 h H21 K H21 = 1 C2 C25 P01 -- -- (P01 )
l RT (all) RT3 h RT3 RT3 = 5. -6 Q
RW (all) RW3 h RW3 RW3 = 1. -7 l W
[n WA WA3 (WA3) -- -- (WA3)
WA4 (WA4) -- -- (WA4)
WB4 (WA4) -- -- (WA4)
PA (all} PA2 h --
(PA2) V @
PB PB, given PA Failure h Failure of PB, given PA has failed is assumed equal 1
to 1.0.
PAPB Success of PA and PB PA2 h PAPB = h h SA f all) SAGT or @ SAGT< 8 - SAGT = 2.9-6
- SALLOCA = 4.0-4 SALLOCA SALLOCA<
SB 58, given --
@ Failure of SB, given SA has 1 SA Failure failed is assumed equal to 1.0.
- The operator (V) is a logical OR. Example: A V B = F(A) + F(B) - F(A)F(B); i.e., the rare event approximation is not used.
A bar above an event indicates the complement or logical NOT of the event.
NOTE: Exponential notation is indicated in abbreviated form; f.e.. 2.66-4 = 2.66 x 10-4 1376PO40186
M M -
M TABLE 3-4 (continued)
Sheet 2 of 2 Top Split Assumptions for SEIS4 ean r an cadon*
Event Fractions Analysis of S2 and 56 Nonseismic Seismic SASB Success of S4 and SB
@ SASB=h h GA GA1 GA1 -- --
(GA1)
GA2 EP3 --
(E )
(all others) EP3 @-, h --
(EP3)V V@
GB GBl GB1 -- --
(GBl)
GB2 EP3 -- --
(EP3)
GBA GBA -- --
(GBA)
GBB GBB --
(GBB)
(all otters) EP2,GA1 h,h --
(EP2)V@V@)l (GA1)V(EP2)V@V@*
LLOCA -- --
@,@ Failure of will cause a@largeor @2 LOCA @V@
y Factor GT or ATWS -- --
@,@ Seismi events at do not (@V@)
Factor fail 0 or will cause a general tran ent event.
- The operator (V) is a logical OR. Example: A V B = F( A) + F(B) - F(A)F(B); i.e., the rare event approximation is not used.
A bar above an event indicates the complement or logical not of the event.
- Vertical bar indicates a condition.
1376PO40186
- 4. SOURCE TERMS In the RMEPS study, PLG-0432, use was made of advanced source term information from a range of sources, including the SSPSA, the IDCOR program and from NUREG-0956. Uncertainti?s in the source terms were estimated in the form of best estimate and conservative estimate source terms, which were probabilistically weighted in the base (mean value) case. Several sensitivity cases were investigated, which reflected assumptions of varying degree of conservatism. As an additional test of the robustness of the RMEPS conclusions with respect to source term uncertainties, the current study is based on source terms directly derived from the WASH-1400 source term methodology. This section describes how the WASH-1400 methodology source terms were derived.
The purpose of the current analysis is to determine whether and to what extent the conclusions in the RMEPS study depended on advancements in the generic state of knowledge of accident source terms. Tnerefore, a set of WASH-1400 based source terms was determined by employing the WASH-1400 source term analysis methodology (i.e., the MARCH code and the CORRAL
% code). These analyses modeled the actual Seabrook configuration and b containment. This was done in the original SSPSA analysis, except that the containment heat transfer model in the MARCH code was replaced by the LOC 0 CLASS 9 code developed by Westinghouse. The results predicted by the COC0 CLASS 9 code have been thoroughly reviewed by several NRC contractors, particularly in the review of the Zion and the Indian Point PRAs, as well as in the reviews of several other PRAs, including the Seabrook PRA and I the Millstone 3 PRA. The C0C0 CLASS 9 code predicts results that are very similar to the results obtained from the MARCH containment model. For the primary system response, the MARCH code was used, and for the behavior and release of radionuclides, the CORRAL code was used.
Concrete penetration by the debris was analyzed using the CORCON code instead of the INTER code. However, since the Seabrook concrete employs
< basaltic aggregate, concrete penetration by the debris is not a si1nificant physical phenomena. For the purpose of source term dev lopment, basemat melt-through scenarios behave very similarly to conte, ment overpressurization; i.e., there is no significant reduction in sours a terms due to filtration as with soil based contairments.
From the orminal accident progression analyses in the SSPSA that utilized thest. codes, a WASH-1400 based source term was derived for each of the six release categories used in the RMEF5 study. For consistency with NUREG-0396, only releases during the acute accident time of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> were considered. Seabrook-specific desigr. information was used, as follows, in determining the WASH-1400 based source terms:
e The dominant accident sequences that contribute to each release category were determined from the results of the plant model update, documented in the RMEPS study. The timing of the source term release was then based on the accident progression for these dominant accident sequences.
I 1378PO40286 4-1
e The primary and secondary system model corresponded to the actual Seabrook Station de-ign. The large secondary water inventory in the Seabrook steam generators has a significant influence on the time to core uncovery by delaying the time to steam generator dryout.
e The Seabrook containment was modeled as designed, including the containment enclosure building. The pressure capacity of the containment building and of the enclosure building was based on the SSPSA analysis. The high pressure capacity and the identified failure modes for the Seabrook containment had a significant impact on the WASH-1400 based source terms. The basaltic aggregate used in the concrete also helped to reduce the containment pressurization rate for the dominant accident sequences.
The six source term categories used in this study are identical to those used in the RMEPS study. They are identified in Table 4-1 (identical to Table 4-4 in the RMEPS report). In order to distinguish the current I source terms from previously used source terms, the WASH-1400 based source terms are designated with a suffix "W"; i .e., S1W to S7W. The following step by step procedure was used to determine the WASH-1400 based source terms:
- 1. For each source term category, the frequency dominant plant damage state was determined for the revised C-matrix shown as Table 4-2 (identical to Table 4-17 in the RMEPS report).
- 2. The dominant accident sequences contributing to each plant damage state were determined from Tables 3-3a to 3-30 in the RMEPS report.
- 3. The dominant sequences were associated with the most appropriate accident progression analysed in the SSPSA. The accident progression timing from this analysis was used to detemine the timing parameters for the source term release. The release timing was modified if it was affected by the acute accident time duration.
I 4. The time dependent radionuclide release fractions for the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the warning time were obtained from the original I CORRAL code calculations for the corresponding containment f ailure mode and for the most applicable accident progression analyzed in the SSPSA.
- 5. For release categories where the release duration is short compared to the evacuation time, or for very small releases, a single puff radionuclide release was used to model the source term. For all other release categories a three puff release was used to model the source term. The first two puff releases are typically intended to cover the time during which evacuation would be in progress or completed, whereas the third puff would cover the time period to the end of the acute accident time.
Release category SlW includes all accident sequences that lead to an I early gross failure of the containment. The dominant plant damage state is IFV. A large fraction of this plant damage state is made up of 4-2 I 1378PO40286
I V-sequences where a gross RHR pipe rupture occurs. It is assumed that the rupture occurs above the water level in the RHR vault after the RWST is injected, so that the release is not mitigated by water pool scrubbing. A significant fraction of the RHR piping is located in the RHR vault at elevations that would be flooded. The WASH-1400 release category PWR-2 is used to represent this release, which is designated S1W.
I Release category S2W includes all accident sequences that lead to an early increased leakage rate from the containment. In the SSPSA, it was shown that the Seabrook containment has a high pressure capacity. Even for the highest calculated pressure spike at the time of vessel breach, the pressure was well below the median pressure capacity of the lowest pressure failure mode. The lowest capacity failure mode at 181 psia involved a failure of three Schedule 160 piping penetrations. The leak area associated with each penetration is 0.5 square inches for a total leak area of 1.5 square inches, if complete correlation between the uncertainties in the failure pressure is assumed. It was concluded that if a pressure spike at vessel breach caused failure of the containment, this failure most likely involved a leak area not exceeding 1.5 square inches. At the failure pressure of these penetrations and considering the geometry of the leak path, it was concluded that the effect of this f ailure mode was an increase in the containment leak rate from the design basis. leak rate of 0.1% per day to a maximum of 40% per day. Release category S2W thus was modeled at an increased leak rate of 40% per day at I the time of vessel breach. Given this leak rate, the containment would not overpressurize and fail at a higher leak rate failure mode for at least several days, if at all . For a 24-hour acute accident time the leak rate thus never would exceed 40% per day.
The dominant plant damage states for release category S2W are 7FP and 3FP. These two plant damage states constitute 93% of the total I frequency of release category S2W. The dominant sequences in these two plant damage states are earthquake induced station blackout transients.
Therefore, release category S2W was modeled as a three puff release based I on the timing of a transient sequence without any power, and, in addition, the turbine-driven auxiliary feedwater pump was conservatively assumed to be unavailable (as in sequences in 3FP). In the SSPSA, this sequence was designated TE and implied a dry containment condition without RWST injection and without containment heat removal. This condition guaranteed concrete penetration into the reactor cavity basemat following vessel breach.
Release category S3W includes all accident sequences where the containment remains intact throughout the early phase of the accident progression but there is no containment heat removal. Therefore, the containment slowly pressurizes and fails after several days due to overpressure. The dominant plant damage states for S3W are 80 (1.0E-4) and 3D/7D. Plant damage state 8D is dominated by transient sequences with failure of PCC or service water and, thus, no ECCS core cooling and no containment cooling. These sequences are charcterized by a pump seal LOCA over an extended period with secondary side cooling before core uncovery and a wet containment condition that prevents concrete penetration by the debris. Plant damage states 30/7D are dominated by 4-3 I 1378PO40286
transient sequences without injection and no containment cooling, and by ATWS sequences. Overall, ~73% of all accident sequences in release category S3W are transients with RWST injection, 20% are transients without injection, and 7% are ATWS sequences. The acute accident source terms for transients with an extended feed and bleed period would be considerably lower than for transients without injection. Release category S3W was conservatively modeled as a single puff release based on the timing of a transient without injection, without auxiliary feedwater, and with a dry containment without containment cooling.
Release category S5W includes all accident sequences where the containment remains intact throughout the accident progression. The source term for this release category is always low. The dominant sequence in SSW is a station blackout sequence with successful containment recovery and with the turbine-driven feedwater pump operating. The total frequency of release category SSW is less than the frequency of S3W. Furthermore, since in SSW the containment spray system is operating, the 24-hour release fractions for S5W are much lower than those for S3W. Tnerefore, the risk contribution from SSW is negligible compared to that from S3W and the consequence of an SSW release was not actually calculateo.
Release category S6W includes all accident sequences where the containment is not isolated with an opening of 3 inches in diameter or greater. Plant damage state 3F contributes 87% to the total frequency of S6W. Overall, 94% of the accident sequences in S6W are earthquake induced ATWS sequences combined with a failure of one of the containment purge valves to close. These sequences were classified as ATWS sequences because of seismic damage to the solid state protection system.
Realistically, seismic damage would result in the generation of a reactor I trip signal to the breakers. Therefore, release category S6W was modeled as a three puff release based on the timing of a small LOCA sequence without injection, without auxiliary feedwater and with a dry containment without containment heat removal. The containment leak was modeled as an 8-inch diameter opening (size of purge valve) directly to the environment.
Release category S7W is dominated by V-sequences where the leak location is in the RHR vault at an elevation below the flooding level following l RWST ain ection. The dominant leak location is at the RHR pump seal, which has a leak location approximately 30 feet belcw the flood level.
The thermal hydraulic response and the tining of the accident progression was based on the calculations reported in the RMEP5 report that modeled the primary coolant system with a small LOCA corresponding to the size of i the RHR pump seal leak and accounting for additional leakage through the l RHR relief valves. Since no source terms exist that reflect this failure mode and that are based on the WASH-1400 source term methodology, the release fractions for S7W were based on the WASH-1400 release category I PWR-2. The source term mitigation resulting from a subcooled 30 feet deep suppression pool was modeled as a decontamination factor of 1,000 for all release fractions except the noble gases. All V-sequence scenarios for which there is no pool decontamination effect are included I in release category S1W, which is identical to PWR-2 in WASH-1400.
4-4 1378PO42286
I I The release categories usea in this study are summarized in Table 4-3.
Table 4-4 shows a comparison of the single puff equivalent source terms for this study and for the RMEPS study. Table 4-4 also shows the effect of the 24-hour acute accident time on the release categories by listing I both the 24-hour release and the total release. It is noted that a release based on acute accident time of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> has a noticeable effect on the release fractions for release categories S2W and S3W, whereas I there is no impact on the other release categories because the total release occurs in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or less. However, since the frequency of release categories S2W and S3W is higher than for release categories S1W, S6W, and S7W, the effect on risk can be significant, I
I I
4-5 1378PO42286
I l
l I TABLE 4-1. SOURCE TERM CATEGORIES I I '
Source Analyzed Analyzed Term Containment Failure Mode in the in This Category SSPSA Study S1 Early Containment Failure Yes Yes S2 Early Increased Containment Leakage Yes Yes 53 Late Overpressure Failure Yes Yes S4 Basemat Melt-through Yes No*
SS Containment Intact Yes Yes S6 Containment Not Isolated Yes Yes S7 Contcinment Bypassed (V-sequence) No Yes
- Based on the SSPSA results, basemat melt-through sequences were assigned to category S3 in this study.
l I
I I 1300PO40186 4-6
TABLE 4-2. REVISED C-MATRIX FOR NEW SOURCE TERM CATEGORIES P ant S urce Term Category D ge S1 S2 S3 SS S6 S7 (f q ency) 1F 1.0 (2.0-8) (2.0-8) 1FV 1.0 (4.6-9) (4.6-9) 1.0 I 1FP (1.4-6) (l.4-6) 1FPV 1.0
( 2.7-8 ) ( 2.7-8 )
2A 3.4-5 1.4-4 1.0-2 0.99 (1.9-6) (6.5-11) ( 2.7 -10 ) (l.9-8) (1.9-6) 30/70 2.0-6 8.0-5 0.95 0.05 (3.8-5) (7.6-11) ( 3.0-9 ) (3.6-b) (1.9-6) 3F/7F 1.0 (3.0-7) ( 3.0-7 )
3FP/7FP 1.0 (1.9-5) (1.9-5) 4A/8A 3.1-6 1.3-4 5.2-3 0.995 (1.1-4) (3.3-10) (1.4-8) (5.5-7) (1.1-4) 7FPV 1.0 (1.2-8) (l.2-8) 80 1.1-6 3.1-b 0.9999 (1.0-4) (l.1-10) (3.2-9) (l.0-4)
Total 5.2-9 2.0-5 1.4-4 1.1-4 3.2-7 3.9-8 Frequency NOTES:
- 1. Exponential notation is indicated in abbreviated form; i .e., 2.0-8 = 2.0 x 10-8 I 2. Numbers inside parentheses are unconditional frequencies (events per reactor year) based on mean values. Numbers not inside parentheses are conditional frequencies of source term categories, given the indicated plant damage state, also based on mean values. Median values of source term categories are presented in Section 3.
4-7 1300P040186
m -
M TABLE 4-3. RELEASE CATEGORIES FOR SEABROOK STATION BASED ON WASH-1400 SOURCE TERM METHODOLOGY l
Release Release Warning Energy Release Fractions Relene Time Duration Time Release
- 9 ##
(hours) (hours) (hours) (MCA/S) XE 0.I. I-2 CS TE BA RU LA S1 W 2.5 0.5 1.0 11.9 0.9 7-3 .7 .5 .3 .06 .02 4-3 S2W-1 4.8 2.0 0.5 0 .03 2.1-4 4.3-3 .023 4.2-3 2.8-3 8.4-4 8.4-5 S2W-2 6.8 4.0 2.5 0 .07 5.0-4 1.3-3 .048 .039 5.5-3 3.4-3 5.2-4 i S2W-3 19.8 18.0 15.5 0 .023 1.6-3 2.3-3 .126 .147 .014 .011 1.9-3 i TOTAL 4.8 24.0 0.5 0 .123 2.3-3 7.9-3 .20 .19 .022 2.5-3 9
m
.01 5 53W 6.0 24.0 2.0 0 4.7-4 3.3-6 3.2-5 1.7-4 1.5-4 1.9-5 1.2-5 2.0-6 56W-1 1.75 1.0 1.5 0 .15 1.1-3 .10 .11 .02 .014 4.1-3 4.1 -4 56W-2 2.75 4.0 2.5 0 .42 2.9-3 .07 .19 .063 .022 .009 .001 56W-3 15.75 18.5 15.5 0 .32 2.2-3 .01 .13 .32 .011 .020 3.8-3 TOTAL 1.75 23.5 1.5 0 .9 6.2-3 .18 .43 40 .047 .033 5.2-3 ;
S7W 8.5 7.0 2.0 0 .9 7-6 7-4 5-4 3-4 6-5 2-5 4-6 I I NOTE: Exponential notation is indicated in abbreviated form; f.e., 7-3 = 7 x 10-3, 1
1377P040286 ,
_-_t
M M TABLE 4-4. COMPARIS0N OF RMEPS AND WASH-1400 METHODOLOGY SOURCE TERMS FOR SEABROOK STATION RELEASE CATEGORIES Release Release Time Energy Release Fractions Description
- 10 Calories Category Start Duration Warning Per Second Xe I Cs Te Sr Ru La S This Study - 24 Hours 2.5 0.5 1.0 12 0.9 .7 .5 .3 .06 .02 .004
- Total 2.5 0.5 1.0 12 0.9 .7 .5 .3 .06 .02 .004 ogy,,,,,,, RMEe5 - Conse,vative i i4 .5 <io 0.9 .i35 .i35 .032 .0ie 005e 6(-4)
- Best Estimate 2 12 1 <10 0.9 .052 .052 .013 006 005 2(-4)
Fd h 5 This Study - 24 Hours 5 24 0.5 0 .12 .010 .20 .19 .022 015 .0025 2
- Total 5 76 0.5 0 1 .022 .31 .32 034 .025 .0042 Early RMEPS - Conservative 5 51 .6 <10 1 .025 .025 .008 .003 .0018 3(-4)
Leakage - Best Estimate 13 76 5 <10 1 .013 .013 .004 002 9(-4) 1(-4)
S This Study - 24 Hours 6 24 2 0 5(-4) 4(-5) 1.7(-4) 1.5(-4) 1.9(-5) 1.2(-5) 8(-6) 3 a - Total 89 0 74 0 1 017 .024 .03 .0026 0023 4(-4)
$ Late RMEPS - Conservative 54 0 42 <10 1 .002 .002 01 2(-4) 2(-4) 3(-5)
- Best Estimate 89 0 74 <10 1 .001 .001 002 1(-5) 1(-5) 1(-5)
S This Study - 24 Hours 4.3 24 .6 <10 014 5(-7) 5(-7) 1(-7) 6(-8) 2(-8) 2(-9) 5
- Total 4.3 24 .6 <10 .014 5(-7) 5(-7) 1(-7) 6(-8) 2(-8) 2(-9)
Containment RMEPS - Conservative 2 24 .4 <10 .014 5(-7) 5(-7) 1(-7) 6(-8) 2(-8) 2(-9)
Intact - Best Estimate 4.3 24 .6 <10 .009 4(-8) 4(-8) 6(-9) 4(-9) 1(-9) 1 (-10)
S This Study - 24 Hours 2 14 1.5 0 0.9 .19 .43 .40 047 033 005 6
- Total 2 14 1.5 0 1 .19 43 40 047 .033 .005 Open RMEPS - Conservative 2 12 1 <10 1 .052 .052 .033 .062 .005 2(-4)
Purge - Best Estimate 4 16 3 <10 1 .01 .01 3(-4) 6(-4) 6(-5) 6(-5)
S This Study - 24 Hours 8.5 7 2 0 0.9 7(-4) 5(-4) 3(-4) 6(-5) 2(-5) 4(-6) 7
- Total 8.5 7 2 0 0.9 7(-4) 5(-4) 3(-4) 6(-5) 2(-5) 4(-6) I p p RMEPS - Conservative 8.5 7 2 <10 1 094 .094 083 2(-4) 4(-4) 4(-4) l Bypass
- Best Estimate 8.5 7 5.5 <10 1 3(-4) 3(-4) 2(-4) 1(-6) 3(-6) 3(-6)
- Source terms for this study are given for the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after warning and for the total release. Source terms from RMEPS are given for total release only.
NOTE: Exponential notation is indicated in abbreviated form; f.e., 6(-4) = 6 x 10-4 1367P040286
- 5. SITE ANALYSIS The site (or consequence) analysis calculations for this study were made by using the CRACIT (calculation of reactor accident consequences including trajectories) computer program. The CRACIT program is described in detail in the SSPSA (Reference 5-1); therefore, only a brief summary of the model is presented in this section. CRACIT runs made in this study used essentially the same input information as was used in the SSPSA analysis with the exception of certain release category and site-specific assumptions that are discussed in this section.
The primary intent of this section is to present the results of individual CRACIT runs. Thus, this section is intended to characterize results that are conditional only on the occurrence of the individual release category being run. The combination of these results with the results obtained for the plant and the containment analysis to form frequency-weighted total risk curves is presented in Secticn 2.
This section also describes modifications made to CRACIT and associated output (postprocessor) routines to enable spatial evaluation of doses.
Plotted CRACIT output obtained for individual release categories is described and results for all runs are summarized in Appendices A and B, 5.1 REVIEW OF SSPSA SITE MODEL AND 110DIFICATIONS FOR THIS STUDY The site model used in the SSPSA incorporated Seabrook site features including population distributions, meteorological data (from the site tower and other regional sources), estimates of evacuation trajectories based on highway locations, and evacuation times. The CRACIT program uses these data to evaluate accident consequences for 96 randomly selected weather scenarios for each release category. In each scenario, doses are computed taking into account the time-dependent plume and receptor (evacuee) locations. Results of these scenarios are combined and plots of frequency versus number of health effects are produced.
CRACIT features the ability to treat wind direction and evacuation track changes; thus, it is referred to as a variable trajectory model . Options I in the CRACIT program are used to account for other effects of interest in this study, such as delay time before evacuation and evacuation distances, which are discussed in this section.
The site (consequence) analysis methodology for this study is the same as that described in detail in the SSPSA (Reference 5-1, Section 4.5, " Site Model Analytical Procedure"). Section 12 of the SSPSA, " Site Consequence I Analysis," provides further details concerning the consequence analysis methodology. Analytical details regarding the underlying models can be found in Appendix I of the SSPSA entitled " Site Model." In this study, only exposures received in the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following the accident are considered since there is ample time to mitigate long-term exposures.
5-1 I 1371PO40186
One objective of the analytical technique used in this study is to quantify the site matrix part of the master assembly equation that was described in Section 4.2 of the SSPSA using revised source terms. The elements of the site matrix computed in this study are the conditional frequencies of exceeding damage levels (see Section 5.2) for health effect components of accident consequences; i.e., acute fatalities. All results presented in this section are conditioned on the occurrence of an I accident characterized by the " release category" being considered. The major differences between the SSPSA consequence analysis and this study are the revised source term characteristics resulting from the conservative use of WASH-1400 methodolgy presented in Section 4.
The 96 meteorological scenarios and the evacuation trajectories that were used are all identical to those used in the SSPSA and in RMEPS (Reference b-2). Changes to the CRACIf input compared with that used in the SSPSA are described below.
e in the SSPSA, uncertainty in the consequence analysis was expressed by calculating three sets of consequence analysis results (H, M, and L for high, medium, and low, respectively) and assigning probabilities to each (see SSPSA, Section 12.4). For tnis study, only medium point estimates were used (Table b-1) to be consistent with WASH-1400 (Reference b-3) and NUREG-0396 (Reference 5-4).
Values selected for these parameters are discussed in Reference 5-1; however, a summary is presented below due to their importance to this study.
The assumptions of emergency response for the " medium" case allows for an extended delay of evacuation in up to 10% of the weather scenarios. The extended delay can be used to represent the impact on risk of severe weather effects or ineffective evacuation for other reasons, such as earthquakes. Delays of this type would be expected only rarely.
For all cases, the assumption of no emergency response for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> beyond the evacuation zone is pessimistic. Even in the absence of emergency response planning, relocation of the affected population beyond the evacuation zone could be expected before 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, e As a result of the plant and source term analyses described in Sections 3 and 4, some release categories result in long duration releases that required multiphase (or multipuff) treatment in the CRACIT calculation.
e Population distributions near the plant were reviewed to ensure that there were no significant anomalies; e.g., people located in sectors that were over water. This review confirmed that the population distributions used in the SSPSA were appropriate (Reference b-b).
The only population change from the SSPSA was the deletion of the 2,000 Unit 2 workers assumed to be within the first 1/2-mile in the direction of Sector 25. Note that by deleting this population, the spatial distribution of risk from the site is more realistically determined. This deletion has no significant impact on the 5-2 1371PO40186 L
I conclusions presented in Section 2. The effect of this change on evacuation trajectories and timing in CRACIT was not taken into account because it was judged to be minimal.
e In the SSPSA, the evacuation distance was always assumed to be 10 miles with shelter to 60 miles and normal activities beyond 50 miles. In this study, evacuation zone is varied and sheltering attributable to " normal activities" is assumed beyond the evacuation zone (see Table 5-2).
e In the SSPSA, the assumption was made that evacuees traveled to the edge of the evacuation zone, then received an additional 4-hour dose. For this study, it was considered to be realittic to assume evacuees would continue to travel beyond the evacuation zone. This is particularly true for the smaller evacuation distances studied; e.g., 1 or 2 miles.
The CRACIT consequence model simulates the actual evacuation trajectories and speeds that are defined and calculated for the plant emergency response plan. In the SSPSA, the evacuation trajectories and speeds were based on a 10-mile evacuation distance. In this study and in RMEPS, the same speeds and trajectories were used as those used in the SSPSA, except that the evacuation was terminated at various distances.
The above approach can be compared with the evacuation model employed in the Reactor Safety Study using the CRAC program, the predecessor of CRACIT. The RSS model simulated evacuation for all people within b miles of the plant in all directions, as well as those within 45 of the initial wind direction, out to 25 miles from the plant. fo reflect the uncertainty in the speed of evacuation and to account for ineffective evacuation, 30% of the people in the evacuation area were assumed to remain unevacuated, 40% moved radially outward at 1.2 niiles per hour, and the remaining 30% moved radially outward at 7.0 miles per hour.
A summary of important CRACIT input parameters used in this study is provided in Table 5-2.
b.2 CRACIT POSTPROCESSOR FUNCTION The following discussion describes the CRACIT postprocessor calculations and resulting evaluations. Risk point estimates are summarized on I spreadsheets, as discussed in Appendix C.
5.2.1 ASSESSMENT OF DOSE AS A FUNCTION OF DISTANCE I
The frequency of exceeding whole-body dose levels as a function of distance, assuming no immediate protective action, was calculated for each release category. Exposures were allowed to continue for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> af ter the time of release. These calculations were made for comparisons with the dose versus distance calculations presented in NUREG-0396 (Reference S-4), which were used for developing emergency planning i
1 l
5-3 l 1371PO42486 1
strategies. The 1, 5, 50, and 200-rem whole-body doses were included in the computation. The calculation proceeded as follows:
e For each scenario, if the whole body dose exceeded the selected dose level at any one of the population grid distances in CRACIT (see Appendix I of the SSPSA for distances), a counter is then incremented for that location.
e These occurrences are accumulated over all scenarios for each grid distance and divided by the total number of scenarios to determine frequency of exceeding each dose level at each distance.
e Since a dose in excess of the given level in any direction will increment the counter for the given distance, the results are independent of direction.
e Doses are only evaluated at locations on the grid where at least one person is located in the population table. The results are not population-weighted; i .e., results are not related to the number of people who receive a dose at or above the given dose levels.
Plots of dose versus distance are provided for each release category in Appendix A. Doses in these plots assume that residents do not take protective actions for a period of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> af ter the release starts. An I example of a dose versus distance plot used for screening purposes is shown as Figure 6-1 for a typical release category, which was found to have a significant contribution to the risk of early health effects.
Dose versus distance curves cannot be used individually for risk assessments; rather, they should be weighted by the frequencies attributable to each release category and summed as described in Section 2.
Although thyroid dose and its effects were computed in this study, thyroid dose results are not presented. The assessment of protective I action strategies in this report are based primarily on calculations of whole-body dose and early fatality risk.
5.2.2 CONDITIONAL CUMULATIVE DISTRIBUTION FUNCTIONS The most common way to express risk in consequence analyses is through the use of CCDFs, which are tables or curves representing the probability I (based on the number of weather scenarios run) versus the number of effects (e.g., acute fatalities) conditional on the release. A typical CCDF is illustrated for a single release cateory in Figure b-2. The CCUFs are generated in CRACIT as describcd in Section 12 of the SSPSA.
Separate distributions are computed for each release category, uncertainty level, and mitigation strategy. CCOFs are provided f or all CRACIT runs in Appendix B.
Total risk curves that use individual CCOFs to account for the frequencies of occurrence for each release category are provided in Section 2. The methodology uses the CC0Fs in the same way as described in Section 13 of the SSPSA.
5-4 i 1371PO42386
I S.3 REFERENCES S-1. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic Safety Assessment," prepared for Public Service Company of New I Hampshire and Yankee Atomic Electric Company, PLG-0300, December 1983.
5-2. Pickard, Lowe and Garrick, Inc., "Seabrook Station Risk Management and Emergency Planning Study," PLG-0432, prepared for New Hampshire Yankee Division, Public Service Company of New Hampshire, December 1985.
i S-3. U.S. Nuclear Regulatory Commission, " Reactor Safety Study: An Assessment of Accident Risks in U.S. Commercial Nuclear Power Pl ants ," WASH-1400, NUREG-/b/014, October 1975. l S-4. Collins, H. E., et al ., " Planning Basis for the Development of
' I State and Local Government Radiological Emergency Response Plans in Support of Light Water Nuclear Power Plants," prepared for the U.S.
Nuclear Regulatory Commission, NUREG-0396, December 1978.
5-5. Lee, Dr. S., Yankee Atomic Electric Company, letter to K. Woodard, Pickard, Lowe and Garrick, Inc., July 29, 1985.
5-6. Aldrich, D. C. , D. M. Ericson, Jr. , and J. D. Johnson, "Public Protection Strategies for Potential Nuclear Reactor Accidents:
Sheltering Concepts with Existing Public and Private Structures,"
SAN 077-1726, February 1978.
l l
l l
I iI 5-5 1371P042386
I I l TABLE 5-1. PARAMETERS RELATED TO PROTECTIVE ACTION ASSUMPTIONS USED IN CONSEQUENCE CALCULATIONS Emergency Response Parameter Medium Uncertainty Case Meteorological Scer:ario Sampling Random Stratified Tails Weighted (See SSPSA Section 12)
Incremental Delay in 1.0 in 90% of Weather Scenarios Evacuation (hour)* 4.0 in 7% of Weather Scenarios 6.0 in 3% of Weather Scenarios Fraction of Population Sheltered 0 (assumed normal activities Beyond Evacuation Distance beyond evacuation distance)
Ground Dose Period for Population I beyond Evacuation Distance (hour)**
24
- In the Medium Case, the delays are for the entire evacuee population after warning is given to government authorities by plant personnel.
Delays are in addition to delay times given in SSPSA Table I-3, which are based on the evacuation study discussed in Section 12.3.1.2 of the SSPSA.
100% evacuation is assumed.
- Period between the beginning of exposure and relocation to an unaffected area. During this period, dose accumulates due to exposure to radiation from material deposited on the ground and other surfaces.
I I
I I
I 1372P031486 5-6
I I
TABLE S-2. ADDITIONAL PARAMETERS RELATED TO PROTECTIVE ACTIONS USED IN CRACIT RUNS Parameter CRACIT Values Fraction of Evacuation Zone Population Evacuating 0 for Nonevacuation Cases; 1.0 for All Evacuation Cases Fraction of Sheltering Zone Population Sheltered N/A Maximum Distance of Sheltering Zone (miles) 0 '
Last Evacuation Element Stay Time (hours) O Starting Distance Segment Number 1 Ending Distance Segment Number 34
)
Maximum Evacuation Distance Segment Number 0 = No Evacuation, 2 = 1 Mile, 4 = 2 Miles, Cloud Shielding for Evacuees 1.0 Cloud Shielding for Normal Activities 0.75 Ground Shielding during Evacuation 0.5 Ground Shielding for Normal Activities 0.33 I Cloud Shielding for Sheltered Nonevacuees N/A Ground Shielding for Sheltered Nonevacuees N/A Power Level Fraction of 3,300 MWth 0.9997 l l
Ground Dose Exposure Time for Nonevacuees (hours) 24 1 l I
, I 1
! {
r
. I i
i I
l S-7 I 1372P031486
I I
I I N
=
' tEp = 7= T. . .
N W
I
.. , ^^
E' g i, , O C3 H D O =UO I W I W
<N W
W E
=E I sz u.s WH I .-
..=w-m<O D
C>
W"
^ uH I tr)
W d
E
~
l2 W q"
H H m@
8 a-O l u i (f) w aa w O tM Q >E W ME M
- O 4EP
..*=
oOg I < .
A 7
m W
.M M
I EEEE wwww EMEm D
C3 W
LA.
W I tEP EEP e tEF =tEP (El
- tEPtEP tEl(EP e e EU W Wu-X+4e e
tEP I '; '. .
, .. .v ,
m
. ?= 'g E * 'g A008 310HM 013S00 N3dID DNIG333X3 30 ^1IlI8080Nd I
5-8
m num m m m i
1 l
e 1ee .;e3 .;eA ..; es is . . . _. . . ; e t- - . y.a_
-. . . . . . . . . .. . . . . . _ , ; e2 ..,,1 g
.. x w ' .' 'm y ACUTE FATALITIES a=
1 R 9.C u,JE,,,IH;l,U,,IE S, 1 g'I TOTAL LATENT EFF E wU 18' .. '
a w \ '.'
d' ; :
u o I .
NU m=
\\t -
o a \'-
-2 -2 U =>
= 18 ._ __16 z3 --
\. *:. -:
m ao o m
\:
I W \i "w
a o
n '
a l
E-8 3 5
- -3 o E 18 _"_ -- 16 a .
5 o
-4 -A 10 ... .. .. ..i ..i 16 0' g 1 ' '@' 1 g 3 ' "'18 4 ' "'185' ' ig6 HUMBER OF HEALTH EFFECTS FIGURE 5-2. HEALTH EFFECTS CCDF FOR RELEASE CATEGORY S2W FOR NO IMMEDIATE PROTECTIVE ACTION
E E E E E i
l l
l l
18 1
= -
$ =
E I 3
-1 -1 l E 18 __ _18 W
8 l 5 a
5 2 -2
, 18 __ , _18 5
5
? U
~
5 5 -3 -3 3 18 -- , _18 b
s E x288.88 REM a 58.88 REN
@ +5.88 4 REN ne o
1.88 REN 4 _4 18 i i 18 18-1 '188 '181 182 DISTANCE (MILES)
FIGURE A-1. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S1W FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 447)
! l i W M -
M M M M i
1 l
s 1g 1s ,
3 i
8 m
-t -1 R 16 __ .18 i w
! E -
1 1 5 I E o -2 -2 m 16 _18
, m
! o j > d i A M w
- l
'l Ei -3 -3 i > is __ __1s j Z a
l E a
x266. 66 REN 4
+56.68 REN j @ 5.86 REN j = 1.68 REN
- -4 -4 i
16 e '
16 i 16-1 'i ss 'si i isa DISTANCE O1It.ES) l FIGURE A-2. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S6W
] FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 452) 4
! E E E 1 l l
l 16 -
1f 5
' E 3
-1 E 16__ - 16 N
l g j
x:
5 5 _a -8 w 16 _-_
-'_16
> 0 L y 5 'l 5 -3 ~
g 16 _, -_16 5
x266.66 REN
$ +56.66 REN 5.66 REN
}
g 4
. 66 REN
- n. _4
-4 16 ' 16 16-1 '166 '161 162 I DISTANCE (NILES)
I FIGURE A-3. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S2W l FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 453)
I l
1 1
E 16 tk lim I N U
-1 -1 5 16 -- 16 M
i 8
! 5 N 1 a 1 es -2 -2 i
e,16 . _ 16
- e i
?
O 5
Ei -3 -3 3 16 16 I
C a
E x266.66 REN
- <z +56.66 REN I l @ 4 5.06 REN cx 1.66 REN l n. -4 1
-4 16 i i 16 16-1 '166 '161 162 i DISTANCE (NILES)
,' FIGURE A-4. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S7W l FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 454) i
M M M M i
I 16 1 l 0 m
! w 4
d I
5
~
i 5 16' __ - 18 w
E o : -::
1 E i 2
- I l
- e 16" --
- a 18 E
i >
l ln $$
Ei -3 -3 3 16 _ -- 18 D
a
- x268.66REN a
- +58.88 REN
@ a5.88 REN
, = 1.88 REN
, n. 4 -4 16 i ' 18 1g2 16-1 '166 '181 DISTANCE (NILES)
FIGURE A-5. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S3W FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 455)
M M M M -
M 16 10 ,161, ., . ,162 , ,,163 . . ,164 . . ,16 5 r- . w . m -4 i ~i "i g
w .......................,,,,,, ........N...,%
E _
'h : ACUTE FATALITIES Eg E , \., .,
A.C. U.T. E. . . .I.N. .J.U. .R.I. E. .S. .
l E d 16 ,-- \', ,
16 , TOTAL LATENT EFF E=
m \.', . -:
I "mo \ ',
X ** "
W y s
- oh k' U "o 16 ,16 Su m o i m E :
' w$
N m_ .
_., o .
1 E d -3 o :> 16
-3
_ . 16 t 'i :
E 5 :
-4 -4
, 16 ..' .. .' ' ..'
16
'tgG 160' '"161 ' "1 g2 ' "183' "'18 4 ' "'185' NUMBER OF HEALTH EFFECTS FIGURE B-1. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S1W FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 447)
M M i le s les ... .Js1 .,t e2 .383 ..,t e 4 ..,t s5
- - - . , . . 2. , g. ,
m 3 -
E r
Em
'\ : ACUTE FATALITIES
\ -
~I a.C U,JE,,J,y,U 3,[E,S, E $ 18~ __
k*. -_16 TOTAL LATENT EFF 2= -
yw $ '. -
m --. -
55
- 3 E -2 *
. -2 g ga 16 _
g 18 w o .
n \ ., -
! m a
Sr
=g na u_ i. .
a o \ \ .
l E" -3 k -3 o 5 16 __ .
_ 16 t" -:
5 :
-4 -4 16 ' ' .' > ' 18 l 168' "'t 01 ' 182 ' 18 3 ' "'18 4 ' "'1 g5 ' ' igG NUMBER OF HEALTH EFFECTS FIGURE B-2. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6W FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 452) 4
M M M M M M M -
M i
18 0 166 , , , , , , , ,1 g 1
- , ,,1 g2 , 3 . ,,16 4 ,,,165
.,j 6
- - - -- -- --m i . . . 4 . . . . . , , , , w . 4 i .
w
'., . N,
.g q E %
, , ,'3_ l ACUTE FATALITIES w
Eg ( -
A.C.U.T.E...I.N..J.U.3.I.E.S. .
E, ,
~ =d 16 , - l- _16 ,
TOTAL LATENT EFF O -
k *' .
du u o
\'
, 30 O
! b5 -2 'l. I.
-2 3
ug= 16 __ \i __18 1
- 5u
==
\:
1 co i
0m
=g \.i.: :
w w s:'
a o
- E -8 -3 *
- -3 o $ 16 16
[U .
E 5 :
-4
-4 16 ' >
16 160' "'181 ' "'162 ' 18 3 ' "'18 4 ' " '1 g 5 ' '186 NUMBER OF HEALTH EFFECTS FIGURE B-3. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY 32W FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 453)
I n n n n R FR R F1 FR R F1 FR F1 F1 ll R F1 R F~
j$ .. .. .i . ..ib3
- ,s ..s t64 ..n 165
.r- . g ,- g g i y -
w
's. :
N E : ACUTE FATALITIES r 'N -
mE \ a.Cu,JE,,J,MJ,URJES,
~I '
E $ 16 __ '\ _ 18 TOTAL LATENT EFF E= -
d' g
\ :
l ............., \ .
wa '.-
(
g5 . .
E -2 l -2 j __18 h 816 __
ES { 5
? 55 i.
A u_
a o .
E-' -3 -3 o 5 16 __ 16 t;"i :
E 8 :
-4 -4 16 ... .' .' ..i ..i 16 186' "'161' 182 ' 16 3 ' "'16 4 ' " '165' '166 NUMBER OF HEALTH EFFECTS FTGURE B-4. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S7W FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 454) 4 z
O s
- 1. e
M M M M M M M M 168 166 J63 . .,16 4 . . ,165
. . ,,161
.. .,; 62 m
i Ee
- ACUTE FATALITIES m -
EE - .A.CU,TE,,,I,N,J,U,R I,,E,S,
-1 E 5 16' '._\. _16 TOTAL LATENT EFF EE \ :
i Og u '\, :
6g
\
u5 o m
\ -
-a -2 sz = 18 -
u5 - _._16 goo o z -
i
?
m Es u-a e .
Ed -3 -3 5 :> 16 __ _,16 t;$ :
E 5 :
-4 -4 16 > > > > >
16 1g6' "'181 ' 1 g2 ' 183 ' "'t g4 ' "'165 ' '1g6 NUMBEF. OF HEALTH EFFECTS
'l i FIGURE B-5. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS l
RISK FOR RELEASE CATEGORY S3W FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 455) i iI
M M M M M M i
i le s tee . . . . . . . ,1 e 1 . . ,1 e2 .1e3 . .,1 e 4 ..;e5
-.-------.........,,, w, ... ... i ...
w
'., N, w
E . '
ACUTE FATALITIES r m ,...\.
j E= -
'l.
A.C.U.T.E...I.N..J.U. R.I.E.S..
EgJ 18 ,
1 _1 g , TOTAL LATENT EFF
.w .
o=
w
{., :
w .. -
5o P. :
Nd 'f. -
o5 -2 O. .
-2 Umg 16 j 1, _18 Z g
- e .
1 Wo o
- 1. ,. :
m w5 \* -
- a *Bo ' d.
_.s i .
! EJ -3 '. -3 l
5 16 __
- 18
- [" .
! = :
1 o .
u .
-4 -4 10 -
- ' ::' ::' 10 160 '161 :182 :163 '184 . '165 . 1 86 MUMBER OF HEALTH EFFECTS FIGURE B-6. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2W FOR 1-MILE EVACUATION (RUN NUMBER 456) i
W M M -
M 8 . . ,t e t ,t e2 , .,1 s 3 ..,t es 18 _!ss .J, s4
-...,...=,g.
w g*N :
3 r
. ACUTE FATALITIES m N.
a= \ -
a.C U,JE,,,I,y,U R,IE S, g w" ggd,,, __ g g' I TOTAL LATENT EFF i
o m '1. ' :
l 5u g, g= i ., .
wy - --- -
w m oE i
D E 16' _'~_ i 18 S*S o
\
i
'E i m 3 "53 j i -
O -
i ~
a= 1 i -
E d o :) 16
-3 ) i 16
-3 t; " :
E 5 :
I -
! -4 -4 16 ..i .. .i ..i ... 18
! 168' "'161 ' 182 ' '1g3' "'18 4 ' "'185' '1g6 Mut1BER OF HEALTH EFFECTS l
FIGURE B-7. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6W FOR 1-MILE EVACUATION (RUN NUMBER 457) i l
i
- M M M i
168 166 ,,,t g 1
..g
, , , , , , , ,132 ,
,,163 .,164 , , ,1 65
,, ., ..g ..g ,
4
%.N, :
w . . . . . . . . , , , - . . , ' ' " 'N -
E z
~ 'N : ACUTE FATALITIES
$ w=
I.
g , , A.C. U.T.E. ..I.N..J.U.
... . . R.I.E.S. .
E " 16 ,_ _
N- $ -
16 , TOTAL LATENT EFF EY ., k :
=uO - -
u $
X w *
~
j W u "
u r o a '.
>= -2 u => 16 __
16 2
z8
t Eo : ':
m i
SE mg 5
, m u- :
a o : .
i E" -3 5
-3 5 16 __16 i.
I _w
~a .
@ i :
8 5 .... :
-4 -4 16 ' ' ' ' ' 16 1g6' "'t 81 ' "162 ' "163' "'164 ' "'185' 'tg6 NUMBER OF HEALTH EFFECTS FIGURE B-8. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2W FOR 2-MILE EVACUATION (RUN NUMBER 459)
i M M M l
l 1
166 186 . . ,161 , . , ,,162 .J 63 ,,p g4 .j 65
--...v,p,,,s,N-
! w '.
.)
4 E
r " -). ACUTE FATALITIES
.\*..
$ w= -
j g , .,g' _
A.C.U.T.E...I.N..J.U.R..I.E.S..
i "E d 16 , - _1 g , TOTAL LATENT EFF o a -h. .
Og \ ':
M : --
I w w {
$b 5 -
E -2 I. -2 18 b 8 18 -- - 1 -
5u s o m
s Sz a w o uRo ,
a : -
Ed -3 .
-3 j
5 r w
>16 .". .
16 mJ -
E I I. '
1 o *
\ -
' -4 -4 16 > .' -i > ..' 16
- 186 ' "'161 ' "1 g2 ' 'tg3' "'164 ' "'165 ' 'ggG HUMBER OF HEALTH EFFECTS l
4 i
FIGURE B-9 COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6W FOR 2-MILE EVACUATION (RUN NUMBER 458) i
_