ML18096B079

Summary Rept of Current Results & Findings of Salem Generating Station Level II Ipe

ML18096B079
Person / Time
Site:	Salem
Issue date:	10/31/1992
From:	Public Service Enterprise Group
To:
Shared Package
ML18096B080	List: ML18096B079 ML20106E123
References
NUDOCS 9211050369
Download: ML18096B079 (23)

Text

ATTACHMENT 2

SUMMARY

REPORT OF THE CURRENT RESULTS AND FINDINGS OF THE SALEM GENERATING STATION LEVEL II INDIVIDUAL PLANT EXAMINATION OCTOBER, 1992

SUMMARY

REPORT OF THE CURRENT RESULTS AND FINDINGS OF SALEM GENERATING STATION LEVEL 2 INDIVIDUAL PLANT EXAMINATION The intent of this report is to summarize the status and current findings of the Salem Generating Station (SGS) Level 2 (or back-end) probabilistic risk assessment (PRA) in support of the Individual Plant Examination (IPE). This summary is organized in the same format as Section 4 of the IPE submittal. The SGS Level 2 PRA table of contents, shown in Table 1, is based on the reporting guidelines provided in NUREG-1335 (Reference 1 ).

Drafts of Sections 4.1 through 4.6, 4. 7.1, 4. 7.2, and 4.9.1 have been completed.

Analyses and documentation in support of the remaining sections are in progress.

1. PLANT DESCRIPTION In Section 4.1, Plant Data and Description, the SGS design features that are expected to impact containment performance during a severe accident are compared with those for the NUREG-1150 (Reference 2) large, dry containment reference plant, Zion Unit 1. A comparison of the reactor coolant system (RCS) design and the associated engineered safeguards as well as the containment design and its associated safeguards is shown in Table 2. In general, the comparison shows the plants to be quite similar. Both plants employ 4-loop large Westinghouse nuclear steam supply systems (NSSS) with similar power ratings. (The thermal power ratings of SGS and Zion are 3,411 and 3,250 MWt, respectively.) Containment dimensions are nearly identical, but the containment structural designs differ in that SGS is of a reinforced concrete design, whereas Zion is a post-tensioned design. From a containment layout point of view, the general layout of both plants is similar, except that the communication paths between the reactor cavity and other parts of the containment are different, as described below.

The SGS reactor cavity has a larger floor area than Zion (585 ft2 versus 471 ft2). Both plants have in-core instrument tubes coming out of the vessel lower head that pass up the inclined instrument tunnel. In SGS, the outlet of the instrument tunnel is at the floor of the in-core instrument room (this room is located directly below the seal table); a 3-foot-wide by 8-foot-high opening in the crane wall provides access to this room. This opening has a secured wire mesh door that will have no effect on water or cerium flow. If containment spray injects the refueling water storage tank (RWST) contents, the water will fill the reactor cavity, immersing the lower third of the reactor vessel, and will completely cover the annular and lower compartment floors with several feet of water. If vessel breach were to occur with high vessel pressure and a dry reactor cavity, cerium would likely be entrained by the vessel blowdown gases, be swept across the cavity floor up the instrument tunnel into the in-core instrument room, and then out through the opening in the crane wall into the annular compartment. The design for Zion appears to direct any entrained cerium radially inward from such a room into the lower compartment. The severe accident implication of this design difference is that, in SGS, cerium carried out of the reactor cavity could come in contact with the containment liner, possibly resulting in thermal attack and liner failure. Because the annular compartment floor is at the lowest elevation of the containment (exclusive of the reactor cavity and the sump), it will be flooded with several feet of water if the RWST is injected. If no RWST injection occurs, there will be several inches of water on the annular compartment floor due to the initial RCS and accumulator water for several hours after vessel breach. This should provide WP0250.DOC.10/08/92 1

PSE&G

some degree of core debris cooling and decrease the likelihood of early containment liner melt-through. As will be described later, the SGS-specific Modular Accident Analysis Program (MAAP) model has been configured to address this design feature so that an accurate determination of the water depth on the annular compartment floor is included in the evaluations.

Containment arrangement drawings have been carefully reviewed, and several containment walkdowns have been made by SGS Level 2 team members to confirm the general layout.

Focused walkdowns have been made to address the reactor cavity communication paths described above as well as to address the concern about the potential for hydrogen "pocketing" noted in Supplement 2 to Generic Letter No. 88-20 (Reference 3). The results of these walkdowns have been incorporated into a narrated video tape, which is available as part of the Tier 2 backup material. The walkdowns have supported the conclusion that hydrogen pocketing is unlikely from either in-vessel or ex-vessel hydrogen generation during a severe accident.

A comparison of SGS Unit 1 and Unit 2 arrangement drawings has shown no differences significant enough to require unit-specific thermal-hydraulic analyses. Therefore, the same containment event tree (CET) structure is being used for both units.

2. PLANT DAMAGE STATES The interface between the Level 1 and Level 2 portions of the SGS IPE is facilitated through the use of plant damage states (PDS), which address RCS conditions and containment safeguards status at the onset of core damage. The Level 1 model addresses the containment isolation status as well as the status of containment safeguards (e.g.,

containment spray and containment fan coolers), and each accident scenario leading to core damage is assigned to a specific PDS in the Level 1 analysis. The PDS matrix employed, shown in Table 3, differentiates between high and low RCS pressure, the status of emergency core cooling system (ECCS) systems, and containment isolation and safeguards status. It also identifies whether an unisolated steam generator tube rupture (SGTR) initiator occurs, and whether an interfacing systems loss of coolant accident (ISLOCA) occurs. The frequencies corresponding to the more significant PDSs and the selected key PDSs using U.S. Nuclear Regulatory Commission (NRC) screening criteria will be addressed in Section 4.6 of the submittal and will be summarized subsequently.

3. CONTAINMENT FAILURE CHARACTERIZATION Because of structural design differences between the SGS and Zion containments (e.g.,

reinforced versus post-tensioned designs), an SGS-specific probabilistic containment pressure capacity evaluation was conducted by ABB lmpell Corporation. The evaluation focused on SGS Unit 1 design information, but review of Unit 2 information confirmed that the results are applicable to Unit 2 as well. A large number of potential failure modes were evaluated. The uncertain failure pressure of each mode was judged to be represented as a lognormal distribution characterized by a median failure pressure (Pmed) and logarithmic standard deviations ~m and ~s to express uncertainties in analytical modeling and material strengths, respectively. Ten failure modes were evaluated in detail (many others were included in the initial scoping evaluation, but were screened out because they had failure WP0250.DOC.10/12/92 2

PSE&G

pressures far in excess of the others), and the results are summarized in Table 4. As noted in the table, the value ~c is the composite of the modeling and strength uncertainties, and HCLPF is the "high confidence of a low probability of failure" pressure, roughly equal to the 95% confidence failure pressure. All containment failure modes*

except the liner tear mode were judged to result in a large break when failure occurs. The liner tear mode is due to strain concentrations developed at liner thickness discontinuities that occur around several of the larger penetrations; this failure mode was judged to result in a self-limiting, controlled leakage corresponding to about 7 in2 in effective leak area.

This is the only failure mode that exhibits the "leak before break" characterization. The dome meridional failure mode has the lowest median failure pressure and is due to discontinuities in the meridional rebar pattern near the apex of the top head. If this failure mode were to occur, a large fracture in the containment would be expected, with a direct release into the environment. The basemat flexure failure mode is due to high "diaphragm" bending loads in the basemat in the area where the stiff reactor cavity region joins with the relatively thinner containment basemat. This failure mode is judged to result in a large structural failure area; however, since the failure location is well below grade, any released radionuclides must traverse approximately 80 feet of soil before being dispersed into the atmosphere. It is anticipated that the release fractions of the nongaseous radionuclides will be significantly reduced due to "soil scrubbing," such that this failure mode is differentiated from the above-grade modes in the containment event tree. The remaining shell membrane failure modes are judged to be correlated with the dome mode so that they are not controlling and need not be included in the overall containment pressure capacity evaluation. The basemat shear failure mode is correlated with the basemat flexure failure mode and is therefore not controlling. The hatch and airlock failure modes are correlated, and only the controlling personnel airlock mode need be considered in the overall*

evaluation.

The probability of containment failure due to two pressure loading conditions will be evaluated in the SGS CET. The first is a short-term (i.e., of a few tens of seconds duration) pressure transient load caused by direct containment heating at the time of vessel breach or by hydrogen burns. These evaluations will use both SGS-specific MAAP results (to determine the pre-transient baseline containment pressure level) and the containment loads expert elicitation results reported in Reference 4 for the pressure rise due to direct containment heating (DCH) effects. These sequence-specific imposed load distributions will be statistically compared with the pressure capacity distributions using Monte Carlo sampling techniques to determine the probability and mode of containment failure. The second loading condition evaluated in the CET is a slow rise in containment pressure (and usually temperature) if containment heat removal is unavailable. Such slow pressure rise evaluations have been done for containment atmosphere temperatures ranging from 300°F to 800°F. The evaluation determined the temperature profile through the cylinder and dome walls, and used the appropriate temperature-dependent strengths for the embedded reinforcing steel. An example output for the 300°F case, shown in Figure 1 (a), provides the probability of failure for each of the four independent controlling modes as well as the total probability of failure. Note that the lower tail of the basemat failure mode dominates the lower pressure range, while the dome failure mode is more dominant in the higher pressure range. The personnel airlock failure mode is unlikely, and the asymptotic fraction of controlled leakage failures (due to liner tear) is about 16%.

Thus, the probability of leak-before-break for slow pressure rise cases is relatively small for SGS. Figure 1 (b) shows the total probability of failure as a function of pressure for several containment atmosphere temperature levels. It is interesting to note that the median (i.e.,

WP0250.DOC.10/08/92 3

PSE&G

the 50 percentile} failure pressure decreases by less than 10% when the containment atmosphere temperature increases from 300° F to 800°F.

4. CONTAINMENT EVENT TREE STRUCTURE The SGS CET is used to evaluate the progression of severe accidents, beginning with the onset of core damage (defined as sustained uncovery of the top of the active fuel), and addresses the events and physical processes that are important in determining the probability of containment failure and, if failure occurs, the time, size, and location of the failure. The CET also addresses phenomena that can influence the amount of radioactive fission products released into the environment. The entry conditions for CET analyses are the representative accident scenarios associated with the key plant damage states (KPDS}.

The CET is structured to address accident scenarios that result in core damage for (1) nonbypass events with successful containment isolation, (2) interfacing systems LOCA events (which result in containment bypass), and (3) containment isolation failure events (which, if isolation failures are large, could preclude further challenges to containment integrity). As will be described below, the majority of the SGS. core damage frequency is in the first category. The evaluation of the later two categories focuses more on source term issues rather than on containment response.

The SGS CET addresses events occurring prior to vessel breach (including the potential for in-vessel recovery of the damaged core), the phenomena associated with both the in-vessel and ex-vessel progression of the accident, containment integrity challenges, and the potential for containment failure. If containment failure does occur, the CET characterizes both the timing and the mode of failure. By mode of failure, we mean the containment pressure boundary failure size (e.g., a small, controlled leak or a large, uncontrolled break) as well as the failure location; e.g., direct atmospheric releases, sub-soil releases, or releases into the auxiliary building. Large containment structural failures can occur either in the dome (a direct, elevated release into the environment) or in the basemat (a subsoil release wherein the material must pass through many tens of feet of soil before reaching grade level). Because it would be impractical to develop radiological source terms for each CET sequence, individual CET sequences are binned into a limited number of release categories that characterize the major attributes important to establish accident source terms.* This process will be described below.

The SGS CET provides a logic structure and a systematic framework for analyzing containment performance for each KPDS scenario. The CET logic encompasses all major phenomena between the CET KPDS entry conditions and the release category definition.

Based on insights gained from.reviewing the Zion Accident Progression Event Tree, as well as other large, dry pressurized water reactor (PWR} containment performance studies, a set of 30 top events (excluding the entry state) was selected for the SGS CET. The number of top events was balanced to provide sufficient detail where deemed to be appropriate, while, at the same time, maintaining completeness and scrutability. Greater detail was avoided to preserve the CET as a meaningful tool to communicate key containment response considerations at a practical level of complexity for use in future accident management applications.

The SGS CET is shown in condensed form in Figure 2. By condensed form, it is meant that portions of the tree that have identical structure (but not necessarily the same split WP0250.DOC.10/08/92 4

PSE&G

fraction values} are indicated as subtree transfers. Dashed lines on the SGS CET indicate that a transfer is made to a repeated subtree structure. For example, the ninth sequence in the condensed tree shown in Figure 4.5-1 transfers to subtree 1 (as denoted by the transfer "X 1" in the next to last column}. Subtree 1, as noted in the tree structure, is a small subtree that includes sequences 4 through 8. On the other hand, the largest subtree, subtree 13, contains 8,333 actual sequences. By repetitively defining subtree transfers, the CET shown in Figure 2 with only 45 condensed sequences represents over 33,000 actual sequences.

During the CET quantification, the frequency associated with each of the accident sequences is assigned to an appropriate release category using sequence end state binning logic rules. The release category assignment attributes will be described in Section 4.9 in the IPE submittal and will be briefly described below. Each release category represents a unique combination of containment failure mode and timing as well as other factors that influence the radiological release into the environment. The Salem release category descriptions are intended to be consistent with the NUREG-1150 release category (containment failure bins} designations and nomenclature for Zion.

The first three sequences on the CET involve degraded core recovery; i.e., sequences in which cooling to the core is recovered before irreversible core degradation and vessel breach occurs. Delayed recovery of electric power is a principal contributor to in-vessel recovery. For the remaining sequences, core cooling is not restored, core melting is assumed to progress to vessel failure, and the remaining sequences address challenges to the containment pressure boundary and, should containment failure occur, factors that affect the timing and extent of offsite radionuclide release.

The first two CET top events are invoked for interfacing systems LOCA events. The next top event questions the extent of reactor coolant pump (RCP} seal leakage, which is important for scenarios when seal injection and thermal barrier cooling are unavailable. As previously mentioned, the SGS Westinghouse RCPs employ the "new 0-ring elastomeric material" discussed in the RCP seal leakage expert elicitation in Reference 5. This is particularly important in slow station blackout scenarios (caused either by loss of offsite power initiating events or by events involving switchgear failure that can occur due to ventilation failures or internal floods) wherein the turbine-driven auxiliary feedwater (AFW) pump is available. The next top event (CS} is the only operator action explicitly included in the CET and addresses longer term operator actions to maintain turbine-driven AFW and steam generator depressurization when the station batteries are discharged. Top Event IR addresses recovery of the damaged core prior to vessel breach. If Top Event IR fails, vessel breach will eventually occur. The next three top events (SV, IS, and 18) are questioned in scenarios wherein the RCS pressure is at or near the system setpoint such that pressurizer power-operated relief valves (PORV) cycle or induced hot leg creep rupture or SGTR is a concern.

The next two top events (HP and MP) differentiate as to whether high (i.e., greater than 2,000 psi}, medium (200 psi to 2,000 psi}, and low (i.e., less than 200 psi} RCS pressure is expected at the time of vessel breach. Top Events C1 and L 1 address containment isolation failures. The next two top events (AL and ME} question vessel failure and corium dispersal phenomena either just prior to or at the time of vessel breach. The next three top events (C2, L2, and Z2} question the probability of containment failure due to DCH phenomena at or shortly after vessel breach. The next four top events (S2, DC, LM, WP0250.DOC.10/08/92 5

PSE&G

and F2) address whether containment spray and/or fan cooler continue to operate after vessel breach, whether the corium debris in the reactor cavity and/or on the containment floor is adequately cooled with overlying water so as to prevent concrete thermal decomposition, and whether the containment liner opposite the in-core instrument room opening is failed due to corium thermal attack. The remaining top events question longer term challenges to containment integrity, due to hydrogen burns, slow pressure rise if containment heat removal is not available, or whether long-term containment basemat melt-through occurs.

5. SELECTION OF KEY PLANT DAMAGE STATES AND REPRESENTATIVE ACCIDENT SEQUENCES This activity involves evaluating the Level 1 results to select a manageable number of key PDSs and their representative accident scenarios so as to provide a good representation of the SGS risk profile. The selection process considers both frequency and containment performance status, and uses a conservative application of the screening criteria presented in the NRC Generic Letter No. 88-20 and in NUREG-1335. The NRC Level 1 screening criteria apply to both systemic and functional sequences; the PDS criteria are denoted in Section 2.2.25 of NUREG-1335. The SGS Level 1 analysis employed the linked fault tree (as opposed to the large event tree) methodology, so that its resulting sequences would be categorized as functional. The Level 1 results include all internal initiators as well as internal floods. The core damage frequency (CDF) due to internal events is 4.45 x 1 o-5 per year for SGS Unit 1. The CDF due to internal floods is 1.55 x 1 o-5 per year for SGS Unit 1. (SGS Unit 2 has similar CDFs that will be addressed separately in the IPE submittal.) PSE&G has conservatively used the systemic screening criteria to select the significant PDSs shown in Table 5. Here, all PDSs whose constituent core damage sequences have frequencies greater than 1 o-7 per year are included as well as all containment bypass, containment isolation failure, and SGTR PDSs with whose constituent core damage sequences have frequencies greater than 1 o-8 per year are included.

Eighteen significant PDSs result whose cumulative frequency represent 99.85% of the total CDF. (Note that the successful containment isolation PDSs account for 97% of the total CDF.) These 18 significant PDSs were then combined into nine key PDSs using a conservative binning process, as depicted in Table 6. The frequency percentage contribution for the seven dominant PDSs are shown in pie chart format in Figure 3.

Representative accident scenarios were selected for each KPDS; sometimes, more than one scenario was selected for each KPDS.

6. ANALYSIS OF SEVERE ACCIDENT PROGRESSION Public Service Electric and Gas Company (PSE&G) has selected the Modular Accident Analysis Program, Version 3.08, Revision 17.02, as the computer code used to model SGS severe accident behavior (Reference 6). PSE&G is a member of the MAAP Users Group (MUG) and has a staff member dedicated to developing and exercising the SGS MAAP model. As a MUG member, PSE&G has used the services of the MAAP maintenance contractor, Fauske and Associates, Inc. (FAI), and also has maintained a contract with Gabor, Kenton and Associates, Inc. (GKA), to assist with specialized modeling issues.

WP0250.DOC.10/08/92 6

PSE&G

The SGS MAAP model represents the reactor coolant system (which includes both the primary and the secondary coolant systems), the emergency core cooling system, the primary containment, and containment safeguards systems. The RCS and ECCS models are similar to those for the MAAP reference plant (Zion) and are fairly standard. The containment is subdivided into four global regions. The upper compartment is the large, open volume above the refueling deck (which is at Elevation 130'), comprising about 77%

of the total containment free volume of 2.65 x 106 ft3. The lower compartment is the region below the refueling deck and inside the crane wall; this volume comprises about 12 % of the total volume. The lower compartment floor is at Elevation 81 '. The annular compartment is the fairly open region outside the crane wall and below the refueling deck elevation, and comprises about 11 % of the total volume. The annular compartment floor is at Elevation 78'. The floor has a slight drain angle away from the containment cylinder wall, and a small, sloping trench (with perforated metal cover plates) is located just outside the crane wall, which slopes down to the 8-foot-deep main containment sump. The main sump is about 90 ° away from the in-core instrument room opening through the crane wall mentioned earlier. The reactor cavity includes the open volume around and below the reactor vessel and the sloped instrument tunnel. The cavity volume is 9,800 ft3, or about 0.4% of the total volume. The cavity floor is at Elevation 54.25', and the upper end of the instrument tunnel, which opens into the in-core instrument room, has a 9-inch-high curb, which is at Elevation 81. 7 5'.

If the RWST is injected into containment, the reactor cavity would be flooded and the water level would be between about Elevations 82.5' and 84', depending on whether the RWST level is at the minimum or maximum Technical Specification value. Adding the RCS and accumulator water (about 750,000 lbm) would increase level by slightly less than 1 additional foot. If the RCS and accumulator water were distributed over the annular compartment floor, the depth would be about 1.8 feet.

The MAAP containment model for SGS is straightforward, except for accurately modeling the communication path out of the reactor cavity. As mentioned above, if vessel breach occurs at high vessel pressure and the reactor cavity is dry (i.e., cases wherein the RWST has not been injected), a substantial fraction of the corium released at vessel breach could be entrained by the blowdown gases and swept into the annular compartment. Since the annular compartment is at a low elevation, RCS and accumulator water that condenses on containment structural heat sinks will tend to drain down to the annular compartment floor, not only providing cooling to the corium on the floor but also resulting in more steam generation and a more rapid containment pressurization rate. The MAAP code only allows the user to specify reactor cavity corium ejection to either the lower or upper compartments.. To properly model the SGS communication paths and relative elevations, the input data for the annular and lower compartments were interchanged. A MAAP modeling expert from GKA assisted in this activity.

As many of the severe accident phenomenological issues evaluated in MAAP have considerable uncertainty, MAAP allows the user to vary selected input parameters to assess the sensitivity of containment response to variations in these parameters.

Recommended base case parameter values, as well as ranges of parameter values for sensitivity analyses, are provided by the MUG (Reference 7). Selected sensitivity evaluations noted in this reference are planned for the risk-dominant SGS-specific accident scenarios. Along these lines, a few other MAAP modeling variations are worth noting.

WP0250.DOC.10/08/92 7

PSE&G

For slow station blackout (SBO) scenarios using recommended MAAP parameters, the MAAP results had 100% of the core U02 inventory being swept out of the reactor cavity at the time of vessel breach. Whereas it is plausible that essentially all of the core material that is molten at the time of vessel breach could be ejected from the vessel and swept out of the cavity, it is difficult to conceive that the unmolten (and relatively unmobile) core material would be swept out. The expert elicitation presented in Reference 6 predicts a median ejection fraction of 0.28 for such scenarios. Accordingly, MAAP parameters were adjusted to allow 30% of the core material (U02 1 Zircaloy, and a fraction of the stainless steel lower reactor internals materials) to be swept out of the cavity, with the remainder being spread over the reactor cavity floor to simulate delayed meltdown after the vessel depressurizes. The corium spread area on the annular compartment floor (modeled as the lower compartment as noted above) was selected from a judgmentally based spread area histogram. For low vessel pressure cases, none of the corium is swept out of the reactor cavity.

Another MAAP variable was modeling the equivalent reactor coolant pump seal leakage area for scenarios wherein seal injection and thermal barrier cooling were unavailable. SGS RCP seals have the "new" Westinghouse 0-ring material (References 8 and 9). The probabilities for the equivalent seal leakage rates were selected from the expert elicitation material presented in Reference 4. We have assumed in evaluating the equivalent leak areas that the noted volumetric leakage rates (in gpm per pump) are based on the pressure and temperature levels at the RCP during normal, full-power operation.

MAAP analyses have been completed for the majority of the representative accident scenarios for each KPDS. As MAAP is a deterministic code, several variations were typically run for each scenario. For example, in an SBO where the turbine-driven AFW pumps are available and the operators reduce the steam generator pressure as directed by procedures, cases were run in which: (1) the AFW pumps stop and the steam generator dump valves close on loss of DC power (taken as 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> fron:t SGS-specific battery depletion evaluations), and (2) the operators locally control the AFW pumps and/or locally maintain the dump valves open. Detailed time-dependent plots are prepared for selected key parameters and the timing of key events; e.g., time of core uncovery, time of RCS pressure at vessel breach, total hydrogen produced, time of containment failure if it occurs, and the degree of concrete ablation on the cavity and annular compartment floors.

These deterministic MAAP results will be used to quantify the CET for each KPDS.

7. CONTAINMENT EVENT TREE QUANTIFICATION The SGS CET, shown in Figure 2, will be used to provide a point estimate (i.e., mean value) quantification of each representative accident scenario for all of the key PDSs.

Each scenario will have a unique "name" and frequency. The quantification process involves calculating the mean frequencies of each CET sequence. The quantification will use the RISKMAN software developed by PLG. The process starts with the specified frequency of the initiating event (e.g., the specific key'PDS scenario) and progresses by evaluating the frequencies of both the success and failure branches of each node in the tree. The failure fraction at each node is referred to as the top event split fraction; the corresponding success fraction at a particular node is simply the complement of the failure fraction. Different top event split fraction values are typically encountered for different,

initiating events and different paths through the CET. The process involves evaluating the WP0250.DOC. 10/08/92 8

PSE&G

various split fraction values for each CET top event and providing Boolean logic statements to assign the proper split fraction value to the correct node. This Boolean logic is based on the initiating event and the success or failure status of top events that have previously been questioned in the sequence being quantified; i.e., the logic can be based on the status of CET top events to the left of the top event split fraction in question. As the CET will be repeatedly quantified for a large number of key PDS scenarios, the process involves developing a master frequency file of all top event split fractions and a split fraction assignment file that describes the Boolean assignment logic.

The SGS CET quantification is currently in progress. However, it should be pointed out that PSE&G is attempting to use some of the relevant Zion data from Reference 10.

Unfortunately, this report is only a preliminary draft (issued in October 1990). The principal author of this work (Brookhaven National Laboratory) has indicated that "the results presented in this draft will not change significantly in the final publication." It is a presumption that this is the case. As a result, PSE&G does not intend to revise the SGS Level 2 evaluation when the final Reference 10 report is available.

8. RADIONUCLIDE RELEASE CATEGORIES AND SOURCE TERMS Release categories are used to characterize qualitatively each CET accident scenario so that it can be assigned to an appropriate category based on sequence-specific attributes that are judged to affect the offsite radionuclide release and consequences. The approach being used for the SGS Level 2 evaluation is to employ the attributes used in the ZISOR program as described in Reference 10. These attributes are associated with several possible states for the following items:

Containment failure mode and timing.

Timing and status of containment spray.

Extent of core-concrete interaction (CCI).

RCS pressure at the time of vessel breach.

Mode.of vessel failure.

If a steam generator tube rupture occurs and whether the faulted steam generator relief valve sticks open.

The fraction of molten core participating in CCI.

Amount of Zircaloy oxidized in vessel.

The extent of high pressure melt ejection effects.

The size and location of containment failure, including subsoil releases.

The timing of core damage relative to accident initiation.

WP0250.DOC.10/08/92 9

PSE&G

PSE&G intends to simplify the number of outcomes for each of the above attributes relative to those addressed in Reference 10.

The RISKMAN program has CET sequence binning logic that will be employed during the quantification. This binning logic is similar to that described earlier for the.top event split fraction assignment process; e.g., Boolean logic statements based on initiating event and the success-or-failure status of the CET top event in each sequence. Once the CET quantification process is completed, key release categories (KRC) will be selected based on both frequency and consequence considerations. Although this process has yet to be completed, it is being attempted to keep it as simple as possible, and approximately 12 KRCs are envisioned.

For each KRC, quantitative source terms will be evaluated based on MAAP information, possibly supplemented with ZISOR computations. (The SGS project has programmed the ZISOR program listed in Reference 10 and verified the results with those reported in Figures 3.5-1 through 3.5-4 of Reference 10.) The source term information provides both timing and duration of the atmospheric releases into the environment as well as the release fractions for selected radionuclide groups.

9. BACK-END ANALYSIS RES UL TS This report section will provide the detailed quantitative output for the SGS Level 2 evaluation and is still in progress. The detailed results will be presented in the higher level release category groups, as noted in Table 1. More emphasis will be given to the large, early release group as it generally represents the greatest acute fatality offsite risk if an offsite consequence evaluation were to be made. CET split fraction importance information will likely be useful in deriving severe accident insights. Only limited sensitivity

_ calculations (over and above those done with the MAAP program noted in Section 4.7.3 in Table 1) and KRC frequency uncertainties are currently envisioned.

10. SEVERE ACCIDENT INSIGHTS AND POSSIBLE ACCIDENT MANAGEMENT ISSUES Through the quantification of the CET, severe accident insights are being developed and noted. However, this data will not be formalized until this quantification is completed.

After formalization, the insights will be used for severe accident management planning.

11. REFERENCES

1.

U.S. Nuclear Regulatory Commission, "Individual Plant Examination: Submittal Guidance," NUREG-1335, August 1989.

2.

• U.S. Nuclear Regulatory Commission, "Severe Accident Risk: An Assessment of Five U.S. Nuclear Power Plants," NUREG-1150, December 1990.

WP0250.DOC.10/08/92 10 PSE&G

3.

U.S Nuclear Regulatory Commission, "Individual Plant Examination for Severe Accident Vulnerabilities," 1 OCFR50.54, Generic Letter No. 88-20, November 23, 1988.

4.

"Evaluation of Severe Accident Risks: Quantification of Major Input Parameters.

Experts Determination of Containment Loads and Molten Core Containment Interaction Issues," NUREG/CR-4551, Vol. 2, Rev. 1, Part 2, April 1991.

5.

"Analysis of Core Damage Frequency From Internal Events: Expert Judgment Elicitation," NUREG/CR-4550, Vol. 2, April 1989.

6.

Henry, R. E., and M. G. Plys, "MAAP-3.0B - Modular Accident Analysis Program for LWR Power Plants," Electric Power Research Institute, EPRl-NP-7071-CCML, November 1990.

7.

Kenton, M.A., and J. R. Gabor, "Recommended Sensitivity Analyses for an Individual Plant Examination Using MAAP 3.0 B, Gabor, Kenton and Associates, 1991.

8.

"Reactor Coolant Pump Seal Performance Following the Loss of All AC Power,"

WCAP-10541, Rev. 2, November 1986.

9.

"Analysis and Transient Behavior of the Westinghouse 8-inch Design Number 2 RCP Seal During a Loss of All Seal Cooling Event Representative of a Loss of All AC Power," WCAP-10541, Rev. 2, Supplement 2, June 1988.

10.

"Evaluation of Severe Accident Risks: Zion Unit 1," NUREG/CR-4551, Vol. 7, Rev. 1, October 1990.

WP0250.DOC.10/08/92 11 PSE&G

Table 1 (Page 1 of 2). Table of Contents for Salem Nuclear Generating Station IPE Back-End Submittal 4.0 BACK-END ANALYSIS 4.1 PLANT DATA AND DESCRIPTION 4.1.1 Comparison of Salem and Zion (NUREG-1150 Reference Plant) 4.1.2 Containment and Reactor Building Walk-Through 4.1.3 Containment and Reactor Building System Analyses

4. 1.4 Equipment Survivability during Severe Accident 4.1.5 Evaluation of Hydrogen Pocketing Potential

4. 1.6 Differences between Unit 1 and Unit 2 4.2 PLANT MODELS AND METHODS FOR PHENOMENOLOGICAL EVALUATIONS 4.2.1 Analogy to NUREG-1150 Reference Plant (Zion)

. 4.2.2 Salem Unique Phenomenological Issues and Analyses 4.3 PLANT DAMAGE STATES 4.3. 1 Selection of Plant Damage State Parameters 4.3.2 Plant Damage State Matrix for Salem 4.4 CONTAINMENT STRUCTURAL EVALUATION AND FAILURE CHARACTERIZATION 4.4. 1 Probabilistic Characterization of Pressure Capacity 4.4.2 Results of Salem-Specific Pressure Capacity Analysis 4.5 CONTAINMENT EVENT TREE 4.5.1 Containment Event Tree Logic 4.5.2 Containment Event Tree Structure and Top Event Description 4.6 KEY PLANT DAMAGE STATES AND REPRESENTATIVE SEQUENCES 4.6.1 Selection of Key Plant Damage States from Level 1 Results 4.6.2 Selection of Representative Key Plant Damage State Accident Sequences for Severe Accident Analysis 4.6.3 SGS Unit 2 Key Plant Damage States and Representative Sequences

4. 7 ANALYSIS OF SEVERE ACCIDENT PROGRESSION

4. 7. 1 Analytical Models for Salem Severe Accident Analysis

4. 7.2 Results of Accident Progression Analysis

4. 7.3 Results of MAAP Sensitivity Analyses 4.8 CONTAINMENT EVENT TREE QUANTIFICATION 4.8.1 Description of CET Quantification Process 4.8.2 Evaluation of Containment Event Tree Split Fractions 4.8.3 Split Fraction Assignment Looic WP0250.DOC.10/08/92 12 PSE&G

Table 1 (Page 2 of 2). Table of Contents for Salem Nuclear Generating Station IPE Back-End Submittal 4.9 RADIONUCLIDE RELEASE CATEGORIES AND SOURCE TERMS 4_.9. 1 Release Category Definition and Assignment 4.9.2 Selection of Key Release Categories for Source Term Analysis 4.9.3 Determination of Source Terms for Key Release Categories 4.10 BACK-END ANALYSIS RES UL TS

4. 10.1 Summary of Results by Release Category Groups

4. 10.2 Release Category Group I - Large, Early Containment Failure or Large Containment Bypass

4. 10.3 Release Category Group II - Small, Early Containment Failure or Small Containment Bypass 4.10.4 Release Category Groups Ill and IV - Late Containment Failure and Containment Intact 4.10.5 Containment Event Tree Split Fraction Importance 4.10.6 Sensitivity Calculations 4.10. 7 Release Category Group Uncertainties 4.10.8 SGS Unit 2 Back-End Analysis Results 4.11 SEVERE ACCIDENT INSIGHTS AND POSSIBLE ACCIDENT MANAGEMENT ISSUES

4. 12 REFERENCES APPENDIX A - ASSORTED ENGINEERING ANALYSES IN SUPPORT OF CONTAINMENT PERFORMANCE EVALUATION WP0250.DOC.10/08/92 13 PSE&G

Table 2 (Page 1 of 2). Comparison of Plant Design of Salem Unit 1 and NUREG-1150 Reference Plant. Zion Unit 1 Salem-1 Zion-111 I Type of Reactor PWR PWR Manufacturer Westinghouse Westinghouse Date of Commercial Operation 1973 Reactor Core Nominal Power 3,411 MWt 3,250 MWt Number of Fuel Assemblies 193 193 Fuel Rods per Assembly 264(17 x 17) 204 Number of Fuel Rods 50,952 39,372 Core Weight, Total 276,000 lb.

Uranium Dioxide 224,400 lb.

216,660 lb.

Zircaloy 45,600 lb.

44,550 lb.

Miscellaneous 15,200 lb.

14,850 lb.

Reactor Vessel Inside Diameter 173 in.

173 in.

Overall Height 43.8 ft.

44.0 ft.

Thickness at Beltline (Exel. Clad) 8.50 in.

8.441 in.

Bottom Head Thickness (Exel. Clad) 5.375 in.

5.375 in.

Water Capacity with Core and Internals in 4,945 ft3 4,945 ft3 Place Reactor Coolant System Volume (Nominal, Including PZR) 12,612 tt3 12,710 tt3 Water in System (Nominal) 534,000 lb.

537,900 lb.

Operating Temperature (Outlet) 610°F 594°F Operating Pressure (Nominal) 2,250 psia 2,250 psia PORV Setpoint (Nominal) 2,330 psia 2,500 psia Number of Reactor Coolant Pumps 4

4 Number of Steam Generators 4

4 Recirculation Spray Pumps Number 2

3 Design Flow (Each) 3,000 gpm 3,000 gpm Design Head 350 ft.

477 ft (100 psia)

Recirculation Spray Heat Exchangers Number 2

2 Design Capacity (Each)

34. 1 5 E6 Btu/hr 28.0 E6 Btu/hr Accumulator (Number) 4 4

Pressure 650 psia 650 psig Water Volume (total) 850 x 4 = 3,400 ft3 (min.)

5,400 tt3 (3)

Refueling Water Storage Tank 350,000 gal. (min.)

350,000 gal.

WP0250.DOC.10/08/92 14 PSE&G

Table 2 (Page 2 of 2). Comparison of Plant Design of Salem Unit 1 and NUREG-1150 Reference Plant, Zion Unit 1 Salem-1 Zion-1 (1 >

Containment Inside Diameter 140 ft.

140 ft.

Maximum Inside Height 210 ft.

212 ft.

Free Volume 2,650,000 tt3 2,736,000 ft3 Design Leak Rate 0.10%/day 0.10%/day Design Pressure 47 psig 47 psig Operating Pressure 14.7 psia 14.7 psia Operating Temperature 120°F 100oF Construction Reinforced Concrete Reinforced Concretel2l Wall Thickness 4.5 ft 3.5 ft.

Dome Thickness 3.5 ft 2.75 ft.

Basemat Thickness 18.0 ft.

9.0 ft.

Floor Thickness above Liner (Outside 2.0 ft.

2.0 ft.

Cavity *only)

Pressure Boundary Steel Liner Steel Liner Liner Thickness, Walls 0.375 in.

0.25 in.

Liner Thickness, Dome 0.5 in.

0.25 in.

Liner Thickness, Floor Outside Cavity 0.25 in.

0.250.in.

Liner Thickness, Cavity Floor 0.25 in.

0.750 in.

Atmosphere (Normal Operation)

Nitrogen 4,940 lb. moles 2,355 lb-molesl3l Oxygen 1,314 lb. moles 626 lb-molesl3)

Reactor Cavity Cavity Floor Area 585 ft2 471 ft2 Water Capacity 9,250 ft3 9,207 ft3 Fan Coolers (Number) 5 Not Specified Flow Rate 47,000 CFM, Each(4)

Not Specified Heat Removal Rate 81 E6 Btu/hr, Each(4)

Not Specified Notes:

1.

From Table A.3, NUREG/CR-4551, Vol. 7, Rev. 1, Part 2A.

2.

Zion containment is post-tensioned.

3.

Zion values appear to be in error.

4.

For accident conditions.

WP0250.DOC.10/08/92 15 PSE&G

Table 3. Salem Nuclear Generating Station Plant Damage State Matrix CONT. ISOL STATUS ISOLATED NOT ISOLA TED RCS PRESSURE AT TIME OF CORE CONT. SGs STATUS FC +CS FC ONLY CS ONLY NONE FC +CS FC ONLY CS ONLY NONE UNCOVERY ECCS STATUS A

B c

D E

F G

H INJECTION O'NL Y 1 HIGH INJECTION AND 2

RECIRCULATION NONE 3

INJECTION ONLY 4 LOW INJECTION AND 5

RECIRCULATION NONE 6

7 INTERFACING SYSTEM LOCA (RHRSI Abbreviations :

SGs = Safeguards FC = Containment Fan Coolers CS = Containment Spray (Injection)

Notes:

Key Plant Damage State (KPDS) designators preceded by "K".

Primed PDS's similar to unprimed but designate unisolated steam generator tube rupture initiators.

Table 4. Pressure Capacities for Dominant Failure Modes Temperature Condition #1 Temperature Condition #2 Temperature Condition #3 FAILURE MODE Ti=300°F; To,c=90°F; To,d=95° Ti=550°F; To,c=110°F; To,d=120 Ti= 800°F;To,c = 130°F;To,d = 145°F Pmed Bm r!s Be HCLPF Pmed Bm r!s Be HCLPF Pmed Bm r!s Be HC~:*

(psig)

(psig)

(psig)

(psi a)

(psig)

(osi

1. DOME MERIDIONAL MEMBRANE 123 0.13 0.07 0.15 96 117 0.15 0.07 0.17 89 109 0.16 0.07 0.17 82 2~ BASEMAT FLEXURE 133 0.23 0.09 0.25 89 133 0.23 0.09 0.25 89 133 0.23 0.09 0.25 89

3. LINER TEAR (STRAIN CONCENTRATION 145 0.19 0.05 0.20 105 142 0.22 0.05 0.23 98 135 0.23 0.05 0;24 92

4. CYLINDER HOOP MEMBRANE 170 0.21 0.05 0.22 119 166 0.22 0.05 0.23 115 158 0.24 0.05 0.25 106

5. CYLINDER MERIDIONAL MEMBRANE 175 0.17 0.06 0.18 130 171 0.19 0.06 0.20 123 163 0.2 0.07 0.21 115

6. DOME HOOP MEMBRANE 181 0.17 0.07 0.18 134 175 0.18 0.07 0.19 127 165 0.19 0.08 0.21 118

7. BASEMAT SHEAR 187 0.26 0.08 0.27 120 184 0.26 0.08 0.27 118 181 0.26 0.09 0.28 115

8. WALL BASEMAT JUNCTION SHEAR 252 0.26 0.11 0.28 158 247 0.26 0.12 0.29 154 237 0.26 0.12 0.29 148

9. EQUIPMENT HATCH 230 0.16 180 210 0.18 160 190 0.14 150

10. PERSONNEL AIRLOCK (INNER) 184 0.19 135 164 0.21 116 147 0.18 109 Notes: Ti, To,c and To,d Denote inside surface, outer cylinder and outer dome temperatures, respectively.

Table 5. Salem Unit 1 PDS Frequencies Vessel Reactor Cont. Heat Plant Damage State Cavity Containment Status Pressure Floor Removal C3D(F)*

HIGH DRY NO ISOLATED C6D LOW DRY NO ISOLATED C4A LOW WET YES ISOLATED C3A HIGH WET YES ISOLATED C3B(F)*

HIGH DRY YES ISOLATED C1A HIGH WET YES ISOLATED C6A LOW WET YES ISOLATED C3H HIGH DRY NO UNISOLATED C7 V-SEQ C3C HIGH WET NO ISOLATED C4C LOW WET NO ISOLATED ClB' SGTR C4B' SGTR C3G HIGH WET NO UN ISOLATED C6H LOW DRY NO UNISOLATED C4E LOW WET YES UNISOLATED C3E HIGH WET YES UNISOLATED C1E HIGH WET YES UNISOLATED SCREENED TOTAL CDF

• PDS's listed as C3D(F) or C3B(F) include contribution from internal floods.

' Primed PDS's signify unisolated SGTR accident initiatiors Percentage Cum Frequency Percentage Contribution Contribution 1.5E-05 25.00%

25.00%

1.3E-05 21.17%

46.17%

1.1 E-05 18.50%

64.67%

9.2E-06 15.37%

80.03%

4.9E-06 8.20%

88.23%

2.7E-06 4.47%

92.70%

1.8E-06 3.02%

95.72%

6.5E-07 1.08%

96.80%

5.6E-07 0.94%

97.73%

3.9E-07 0.64%

98.38%

3.7E-07 0.61%

98.99%

2.2E-07 0.37%

99.36%

7.8E-08 0.13%

99.49%

7.5E-08 0.13%

99.61 %

6.3E-08 0.11%

99.72%

4.2E-08 0.07%

99.79%

3.0E-08 0.05%

99.84%

1.1 E-08 0.02%

99.85%

8.7E-08 0.15%

100.00%

6.00E-05 100.00%

Table 6. Salem Generating Station Plant Damage State Matrix with Superimposed Significant and Key Plant Damage States CONT. ISOL STATUS ISOLATED NOT ISOLA TED RCS PRESSURE AT TIME OF CORE CONT. S.G.'s STATUS FC +CS UN CO VERY FC + CS FC ONLY CS ONLY FC. ONLY CS ONLY NONE NONE ECCS STATUS A

B C

D E

F G

H INJECTION ONLY 1 HIGH INJECTION AND 2

RECIRCULATION NONE 3

ID INJECTION ONLY 4 LOW INJECTION AND 5

RECIRCULATION NONE 6

7 Abbreviations :

SG's = Safeguards FC = Containment Fan Coolers

= Isolated Key PDS.

CS = Containment Spray (Injection) j = Unisolated, Interfacing LOCA, or SGTR Key PDS.

Notes:

Key Plant Damage State (KPDS) designators preceded by "K".

Primed PDS's similar to unprimed but designate unisolated steam generator tube rupture initiators.

.. s =

if 0

-~

a 1.00E+OO

~ 1.00E-01

£

• ~..

"'5 E a 1.00E-02 60

(.('.umulative Probabilty of Failure at 300.

• DOME a BASEMAT-subsoil

• LINER TEAR-small o PERSONNEL AIRLOCK

• TOTAL

~

~

/

~-

/

~

~

~

~ ~

u I

/

/

/

/

,/

.,r

/.ff

/

/

/,P

/

/

7 I

/

I I I

~

70 80 90 100 110 120 130 140 Containment Pras*ure (pslgl (b)

Cumulative Probabilty of Failure at Various Temperatures

-~ =

1.00E+OO

l 1.00E-01

£

~

"'5 E a 1.00E-02 60

• 300 a 425

• 550 o 675

• 800 I a

~ -

~~JI::::.

~

~~

Y"Y°'~

././......-: vi........-

/ 'l/,?/"':

/~ Vp//

W#, v

~ w

. ~

I

.,, II

/

1" I/

/

/ /

~

I fl'"/

II I/.( I

/}J II 111/J v I/II 70 80 90 100 110 120 130 140 Containment Pressure (pslgJ Figure 1. Cumulative Probability of Containment Failure versus Pressure 20

~

150 150

{

.i I 1

' 1

• 1

' i P-No. 1 llOOEL..., SALEM Event Tree1 CU71J2 13120111 02 OCT 1992 1£ BT LB SL cs IR IS IS HP C1 LI AL cz, L2 Z2 S2 DC F2 HE C3 L3 C4 L4 Z4 Bl

--"T""---....,--.--~13 Damaged Core Recovered* In Vessel

--~---------------------------------------------

'2..

Ill m

5i E

c s s u

0 m II)

L.

of..

GI N

'i:

I Ill..

CD L..

Q.

Cll a:

!.J II) u

'ti g

c..

iii GI

'i II)

u.

0..

if u

a:

'----.---.--....-~-H_l=-gh_P--1112:----11111-....--1~11---...--n--.rll:Z:--------111---------

s Q.

0..

it

.Y...

II) *

'i

~

~..

II)

L.

0 a::

0 0..

it m L.

m GI c

J QI E'

I II)

L.

0 gi

.J a:

I-0 :c

!.J II)

'ti Cll

'ti u

GI

.g u.g E

E CD L.

~

m

~

... 'i 5i L.

.a CD E

~

~

c j

Ill

~

c c.2 Cll J!

0...

it u i 8

w s a:

a:i w

u L.

~

0

&i GI ti

c L.

0..

OI

i.

0:

Q.

E

I II)

OI.s 8

u..

'i:

J:I GI Q

0 z L.

GI c

J

~81

v'" I

-~.; ~ Lffe 4

Sub-Soll j S..

~' _o~"~'"!i£~:!-

u

-.-J:O=X..-------=.~.!i:...... ~.....................

c C5!I c:

~~

... ~.~.......................................................

X5 iM!M:I

~J'---.,....---1

c c:::.............. :::::::::::::::::::::::::::::::::::

L.............. =

'--TiM~,.,,*:&.t~~!:'::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

Med. Press.

---

rr--'-:i:t.~~:~;:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

Small Large

'--T'=='------~* ;;*......................................................................................... ******...................................................................................

I e~s;ss Scrubbed

....... ~e~...,,~s~s_u...._n~sc=...ru

§~§~e~a,,_ _____________________ _

1 2

3 4

5 6

7 a

9 1D 11 12 13 14 15 16 17 ta 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

• 34 35 36 37 38 39 40 41 42 43 44 45 I

2 3

4 5

6 7

a X1 9*13 14 15 16 X2 17*29 X3 30*55 56 57 58 59 X4

~-

115 X6 X7 73-119 xa 9o-t75 X9 176*347 348 349 XIO 350*695 696 X11 697*1389 1390 1391 x12 1392*277'9 X12 2780*4167 Xl2 4168-5555 5556 5557 Xl2 5558*6945 Xt2 6946-8333 X13 8334-16666 Xt3 16667-24999 X13 25000*33332 33333 33334 33335 X14 33336:33338 Top Event DHl-tor ***** Top Ewnt DHcrlptlon.............................

Top Ewnt DHl-tor ***** Top Ewnt DHcrlptlon.............................

Top Event Dnllinator..... Top Event Description............................,

1£ In LI SL cs II IS II HP MP C1 PLANT DAMAGE STATE CONTAINMENT NOT BTPASSBI PRIOll TO COii£ DAMAGE (PDS)

BTPASS PlllOll TO COii£ DAMAGE IS llOT LARGE (POS) llO SIGNIFICANT RCP SEAL LEAKAGE OPER.CNTRLS.SO COOLING AFTER BATTERY DEPL IN*VESSEL RECOVERY OF DAMAGED COii£ PRESSURIZER POllVS & SVS RESEAT AFTER CYCLING llO INDUCED STEAi! GENEllATOll Tl/BE RUPTllRE NO INDUCED BRKO OF HOT LEO Oii SUllGI! LINE HIGH RCS PRESSURE AT VESSEL BREACH (>2000 PSl,PDS+)

11£1>1111 RCS PIESSUllE AT VESSEL BREACH (200*2000 PSI) llO CONTAINMENT FAILURE PRIOll TO VESSEL BREACH (PDS)

Figura 2. Salam Containment Event Tree L1 AL 1!£ CZ LZ zz sz DC LN.

FZ HE C3 CONTAINllENT FAILUllE PlllOll TO VI IS SllALL llO CONTAlllllENT FAILURE DUE TO IN-VESSEL STtAll EXPLOSION OR ROCiET llOOE NO ftlGH PRESSURE llELT EJECTION lfO CONTAlllllENT FAILURE AT VESSEL BREACH lfO LARGI! COllTAlllMENT FAILURE AT VESSEL BREACH LARGE CONTAINMENT FAILURE BELOW GllADE Oii V*SEQ.SC11\\JBBED COllTAlllllENT SPRATS OPERATE AFTER VESSEL BREACH COii£ DEBRIS COOLED (PDS) llO CONTAlllllENT LINER MEL TTHRCUGR DUE TO COllllJM THERMAL ATTACK FAN COOLERS OPERABLE AFTER VESSEL BREACH lfO MTDllOGEN BUllN UITHIN 4 HRS AFTER VESSEL BREACH NO.CONTAINMENT FAILURE DUE TO EARLT BURN (P SPIKE)

L3 Z3 HR C4 L4 Z4 Bl llO WGE CONTAINMENT FAILURE DUE TO EARLY BURN WGE COllTAllll1ENT FAILURE IS BELOW GRADE LONG TERN CONTAlllllUT HEAT IEllOVAL AVAILABLE NO LATE CONTAINl!ENT FAILURE (UP TO 48 HRS)

LONG TERM COllTAINMENT FAILUllE IS SMALL LARGE CONTAINMENT FAILURE IS BELOW GRADE llO BASEllAT llELT*THROUGH

C3A 15.4%

Figure 3. Significant PDS Characterization C4A 18.5%

Others 4.28%

C3D(F)*

25.0%

C6D 21.2%

~