ML20247F189

From kanterella
Jump to navigation Jump to search
Risk Evaluation of Post-LOCA Containment Overpressure Request
ML20247F189
Person / Time
Site: Oyster Creek
Issue date: 05/05/1998
From:
GENERAL PUBLIC UTILITIES CORP.
To:
Shared Package
ML20247F180 List:
References
NUDOCS 9805190192
Download: ML20247F189 (15)
v • d • e


Text

- - - - - - _ - - - - - - _ _ _

l l

Attachment III Risk Evab. nation Of The Post LOCA Containment Overpressure Request Table of Contents 1.0 Introduction 2 2.0 Background 2 3.0 Method 2 3.1 Probability of a Large Drywell LOCA 3 without Containment Overpressure 3.2 Core Damage Frequency Impact 4 3.2.1 Sensitivity Case One - Containment 5 Overpressure not Required 3.2.2 Sensitivity Case Two - Probability of Failure 6 of Overpressure without External Injection 3.2.3 Sensitivity Case Three - Probability of 7 Failure of Overpressure with External Injection 4.0 Results and Conclusions 8 5.0 References 10 6.0 Containment spray Trip System Analyses 11 9805190192 99050529 PDR ADOCK O G

l , .

I 1.0 Introduction The purpose of this evaluation is two-fold. First, to evaluate the probability of a loss of coolant accident (LOCA) coincident with a loss of containment overpressure. Second, to evaluate the change in the core damage frequency as a result of the requirement for containment overpressure following a LOCA in the drywell airspace.

The requirement for containment overpressure is the result of the potential for emergency core cooling system (ECCS) strainer clogging. A proposed modification of the ECCS strainers to decrease the potential for strainer clogging is to be incorporated into Oyster Creek design in the upcoming refueling outage. However, under the most restrictive ECCS strainer clogging conditions and ECCS flows, containment overpressure is required.

2.0 Background Existing plant analyses have studied plant transients that are the result of a LOCA inside the drywell airspace. These LOCAs can result in the relocation of fibrous insulation material from the drywell to the wetwell area. Once in the wetwell, the insulation can cause blockage of the ECCS strainers.

I Blockage of the suction strainers results in the potential for ECCS pump cavitation on low suction pressurc (i.e., lack of adequate net positive suction head (NPSH)). The I primary concern occurs in the early phases of a LOCA, where ECCS flows are the highest.

The purpose of this plant modification is to reduce the potential for a loss of NPSH during 'aLOCA by installing a new suction strainer design with the capacity to filter more debris with less pressure drop.

4 3.0 Method To estimate the probability of a LOCA and loss of containment overpressure, the frequency of LOCAs is combined with the probability of the loss of containment overpressure.

To estimate the change in core dunage frequency as a result of the proposed modification, the Oyster Creek Probabilistic Risk Assessment (OCPRA) is reviewed.

l Sensitivity cases are evaluated associated with assumptions regarding the requirement for containment overpressure.

f  ;

i

?

i w _. . _ _ _ _ _ _ _ _ - _ - _ _ - - _ - _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . __ - __ _ _ - -

l 1940-98-20124 I i

Attachment III Page 3 l 3.1 Probability of a Large Drywell LOCA v/ithout Containment Overpressure The estimation of the probability of a large drywell LOCA without containment overpressure is a two step process.

In the first step, the probability of a large LOCA is estimated. The large LOCAs of I concern are limited to those which cause insulation to become dislodged and plug ECCS suction strainers. Therefore, the initiating event of the sequence of events is a )

LOCA inside the drywell airspace.

Reference 2 provides the OCPRA initiating events groups. Only the large below the core and inside containment loss of coolant accident (identifier: LBI) is evaluated.

{

(

This LOCA is equivalent to the design basis accident. Other large LOCAs are above  !

the core. As such, either injection cources external to the containment are available for

{

reactor injection, or a minimum core spray pump configuration will be used to refill the j system (i.e. one pump with a booster). There is no NPSH concern under this set of l conditions. Therefore, these LOCAs are not considered. Small LOCAs are not included since they will not result in maximum debris loading (smaller zone of I influence).

The second step is the estimation of the probability of the loss of containment I overpressure. Containment overpressure is not available due to failure. Containment overpressure is the result of the LOCA.

Containment overpressure can be lost via an existing hole in the containment or the

{

failure of the automatic and manual containment spray trip on low drywell pressure. l This equation illustrates the estimation of the frequency of a LOCA in the drywell airspace with a loss of the containment overpressure as a result of a pre-existing hole or a failure of the containment spray system trip.

Frequency Probability Probability of l

Frequency of I of. of Pre- + Containment = LOCA with Large Existing Spray Trip Loss of LOCA in Containment Failure Containment the Drywell Leak Overpressure i 5.67x10d * ( 4.1x104 + 3.93x10 d ) = 2.46x104 per year l

l

, l t . _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _

z .

1940-98-20124 t Attachment III Page 4 The probability of a pre-existing hole in the drywell is taken from NRC sponsored research (reference 5) and is equal to 4.1x104.

l The probability of the failure of the containment spray trip function is evaluated in the Level'1 OCPRA'in split fraction CF1. The split fraction CF1 is defined as the probability of the failure of the trip of 2 of 2 running containment spray pumps and is l

equal to 3.93x104. This split fraction includes the contribution of operator action to trip the pumps following 'he failure of the automatic action.

The frequency of 2.46x10 per year represents the occurrence of a large LOCA and L loss of containment overpressure (either through a pre-existing leak or as a result of the L containment spray system failure to trip).

l l 3.2 Core Damage Frequency Impact To estimate the core damage frequency increase due to the proposed change, the l OCPRA risk model is reviewed. The success critetia for a large below core and inside l containment LOCA is defined in Section 8 of the OCPRA (reference 2). The success ,

criteria (adapted from reference 2) are given as: j INIT=LBI

  • CS=S
  • OS=S * (CC=S + OV=S) l This success criteria for initiating event group large below the core and inside containment LOCA (INIT=LBI) can be expressed as:

4 success of the core spray system (CS=S) and 4 operator action to align core spray to the condensate storage tank (OS=S) and 4 containment heet removal via either the

- containment spray / emergency service water system (CC=S) or

External injection to the reactor vessel from the condensate storage tank is required for the long term mitigation of the large below core and inside containment initiating event.

External injection is also directed by the Emergency Operating Procedures (EOPs).

The proposed change to require containment overpressure does not change the current LOCA mitigation strategy and external injection would continue to be required.

Once core spray is shifted to external injection sources, containment overpressure would no longer be required. Containment overpressure is not required until the insulation debris sufficiently clog the ECCS strainers. Maximum debris loading is not expected for 30 minutes, i

i l

i

o ( ,

1940-98-20124 Attachment III Page 5 Given the requirement for external injection, and the time available to perform the alignment to external sources (maximum debris loading is approximately 30 minutes), it can be assumed that the requirement for containment overpressure has no impact on the success criteria for the mitigation of the LBI initiating event. Therefore, the proposed containment overpressure does not affect the core damage frequency.

In fact, the proposed modification to the ECCS suction strainers may reduce the core damage frequency as the availability of containment overpressure may preclude the need for alignment to external injection sources.

, Three sensitivity cases were performed and are described in the following sub-sections.

In Sensitivity Case 1, " Containment Overpressure Not Required", the decrease in core damage frequency is estimated given that containment overpressure and alignment to external injection sources are not required.

Sensitivity Case 2, " Probability of Overpressure Failure Without External Injection",

assumes that containment overpressure is required for the mitigation of the design basis LOCA regardless of external injection. Case 2 is the conservative bounding case, where containment overpressure is required for success of core spray and containment spray systems following a large below core LOCA.

Sensitivity Case 3, " Probability of Overpressure Failure with External Injection"' most closely reflects the current design and operating practices as well as the proposed change. That is, containment overpressure is required for the long term success of the core spray system. Without containment overpressure, the core spray system will be available in the short term and injection via the fire protection system or alignment of the core spray to th: condensate storage tank remain viable alternatives to satisfy long term core cooling requirements. Sensitivity case 3 represents a decrease in the core damage frequency.

3.2.1 Sensitivity Case 1 - Containment Overpressure Not Required The first sensitivity case evaluates the reduction in core damage frequency as a result of the guaranteed availability of containment overpressure. In this sensitivity case, the base OCPRA risk model's success criteria for the LBI initiating event group is adjusted to remove the requirement for the alignment of the core spray system to the condensate storage tank. This sensitivity case evaluates the core damage frequency where containment overpressure is not required for the mitigation of large below core LOCAs.

This sensitivity case does not include the current requirement of the alignment of the core spray system to the condensate storage tank but does assume that containment overpressure is not required.

1 L_ i__ . .. . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ _  !

  • 1940-98-20124 4 Attachment III Page 6 Table 1 - Sensitivity Case 1 Results Initiating Event Group Base Case CASE 1 Percent Description OCPRA Guaranteed Change Overpressure (from base) 4 All Transients and LOCAs 3.689x10 3.689x10' 0%

Except LBI Large Below Core and 1.089x10-7 1.081x10 - 0.8 %

Inside Containment LOCA (LBI)

TOTALS 3.798x10' 3.797x10' ~0%

J 3.2.2 Sensitivity Case 2 - Probability of Failure of Overpressure without External Injection  ;

i The second sensitivity case evaluates the core damage frequency given the probability  ;

of failure of containment overpressure (based on the failure of the containment spray j pump trip and pre-existing leak in containment).

In this sensitivity case, the base OCPRA risk model's success criteria are adjusted to remove the requirement for the alignment of the core spray system to the condensate storage tank, as in case 1. In addition, the top event CF is added to the LBI success criteria. Top event CF is redefined as the probability of the failure of the containment spray pump trip and the probability of a pre-existing leak in containment. Therefore,  !

top event CF represents the failure of containment overpressure.

The failure of containment overpressure is assumed to fail the core spray and containment spray systems. This sensitivity case does not include the current requirement of the alignment of the core spray system to the condensate storage tank and therefore does not credit flow management strategies.

Therefore, this sensitivity case models a requirement for containment overpressure to ensure the availability of the core spray and containment spray systems. No credit is given for the potential for operators to align the core spray system to external sources or injection of the fire protection system.

l L_ _-__ _ _ _ _ _ - --- -_- - - ___ _ ____-_- _-__-_-_ - _ -- ---- - -- --_ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

,... l 1940-98-20124 Attachment III Page 7 Table 2 - Sensitivity Case 2 Results Initiating Event Group Base Case 1 Case 2 Percent Description OCPRA Probability Change of (from Overpressure base) 4 All Transients and LOCAs 3.689x10' 3.689x10 0%

Except LBI Large Below Core and 1.089x10-7 3.541x10-' 225 %

Inside Containment LOCA (LBI)

TOTALS 3.798x104 4.043x104 6.5 %

i These results agree closely with the results of the equation estimate of the frequency of overpressure and a large below core LOCA produced in Section 3.1. That is:

CASE 2 OCPRA Equation LBI Initiator -

LBI Initiator = Estimation Core Damage Core Damage Of Overpressure i Frequency Frequency Probability 3.54x10-7 per year -

1.09x10-' per year = 2.45x10-' per year The results agree closely since the assumptions made in sensitivity case 2 result in assumed core damage when a large below core LOCA occurs and containment overpressure is not available.

l 3.2.3 Sensitivity Case 3 - Probability of Failure of Overpressure with Ex.ternal Injection Sensitivity case 3 evaluates the core damage frequency change as a result of the proposed modification assuming that overpressure is required for long term success of core spray. Short term reactor vessel inventory must be provided by core spray system regardless of overpressure. Long term (greater than 30 minutes) reactor vessel inventory can be provided by the fire protection system given short term success of the core spray system. The success criteria can be stated as:

INIT= LBI * ((CS =S

  • CF=S) + (CS =S
  • CF= F
  • FS =S)) * (CC=S + OV =S)  :

i

.4 , .

1940-98-20124 Attachment III Page 8 Where success mitigation is of the large below core inside containmer;t initiating event (INIT=LBI) is defined as either:

(CS=S

  • CF=S) success of core spray injection and containment overpressure.

Or (CS = S

  • FS = S) success of the core spray system (early), failure of the containment overpressure and fire protection injection through core spray (late).

With successful containment heat removal via the containment spray / emergency service water system or the hardened vent (CC=S + OV=S).

Table 3 - Sensitivity Case 3 Resuks Initiating Event Group Base Case Case 3 Percent Description OCPRA Probability of Change i Overpressure (from with External base)

Injection All Transients and LOCAs 3.689x10 4 3.689x10 4 0%

Except LBI Large Below Core and 1.089x10-' 1.081x10 - 0.7 %

Inside Containment LOCA (LBI)

TOTALS 3.798x10 4 3.797x10' ~0%

4.0 Results And Conclusions Given the current mitigative measures (e.g., training emphasis on the use of the fire protection system in the case of ECCS suction strainer clogging), sensitivity case 3 is i considered the best representation of the proposed modification as well as current '

design and operating practices.

1 Sensitivity Case 1 presents the resulting core damage frequency if containment overpressure was not required for mitigation of the large below core and inside containment initiating event (design basis LOCA).

1940-98-20124 Attachment III Page 9 l

1 Since the aligonent of the core spray system to the condensate storage tank is currently required in the OCPRA, this case represents a decrease in the core damage frequency from the OCPRA base results. The OCPRA requires the alignment of the core spray system to exte.nal sources since the emergency

)

operating procedures currently require containment flooding if reactor vessel level cannot be determined.

Sensitivity Case 2 presents the resulting core damage frsquency assuming that overpressure is required and no external injection sources are available for mitigation of the event. This case represents the bounding case, where containment overpressure is required for success of core spray and containment spray systems following a large below core LOCA.

I Sensitivky Case 3 presents most closely reflects the current design and operating i practices as well as the proposed modification. That is, containment overpressure is I required for the long term success of the core spray system. Without containment {

overpressure, the core spray system will be available in the short term and injection j via the fire protection system or alignment of the core spray to the condensate J storage tank remain viable alternatives to satisfy long term core cooling  !

requirements. Sensitivity case 3 represents a decrease in the core damage frequency. The decrease in core damage frequency is the result of the credit of the 1 core spray system, without external injection, as a success path given containment I overpressure in addition to currently modeled success path of the core spray system l with external injection.

Probability of a 12rne Drvwell LOCA without Containment Overpressure The probability oflarge LOCA with the failure of containment overpressure is 2.46x10' per year. 'Where the probability of:

the failure of containment spray pump trip system is 3.93x104 per demand a pre-existing leak of the containment is 4.1x104 per year e a large below core LOCA is 5.67x104 per year The probability of the failure of containment overpressure is given as 4.34x10d .

Core Damane Freauency Imnact Sensitivity case 3 is judged to best represent both the current design and operating practices as well as the design and operating practices following the incorporation of the ,

modification. Therefore, the core damage frequency impact is approximately 0% and  !

the modification is judged to be risk neutral.

.* 1940-98-20124 Attachment III Page 10

' Table 4 - Sensitivity Case Results Sensitivity Case Description Core Damage Percent Frequency increase Sensitivity Case * - Containment Overpressure Not Required for 3.797x104 0%

Successful Mitigation Sensitivity Case 2 - Probability of Failure of Overpressure without External 4.043x104 + 6.5 %

Injection Sensitivity Case 3 - Probability of Failure of Overpressure with External 3.797x10 4 0%

Injection 5.0 References

1. GPU Nuclear Corporation, Memorandum, J. Fornicola to E. O'Donnell, "GORB Action Item", GORB-505, October 16,1997.
2. GPU Nuclear Corporation, " Oyster Creek Probabilistic Risk Assessment (Level 1)", Volumes 1 through 6, November 1991.
3. PLG Corporation, " Database for Probabilistic Risk Assessment of Light Water Nuclear Power Plants (BWR Initiators)", PLG-0500, Volume 7, Revision 0, July 1989.
4. BWR Owners Group, " Pipe Break Probabilities in Boiling Water Reactors", November 1993.
5. P.J. Pelto, K.R. Ames, R.H. Gallucci, " Reliability Analysis of Containment Isolation Systems", NUREG/CR-4220, Pacific Northwest Laboratory, June 1985.
6. GPU Nuclear Corporation, Oyster Creek Procedure and Policy Manual, "On-Line Maintenance Risk Management",2000-ADM-3022.01, Revision 4, September 5,1997.

4 l

4 - ,

.. 1940-98-20124 Attachment III Page 11 6.0 Containment Spray Trip Analyses 6.1 Discussion The top event 'CF' models the trip of the containment spray pumps to prevent continued containment pressure drop and the potential to reduce containment overpressure below 1.25 psig.

The split fraction 'CF1' evaluates the trip of 2 of 2 running containment spray pumps. Both manual and automatic actions are modeled. The probability of containment spray trip failure is 3.93x10-7. Split fraction CFA, which is equal to 'CF1' plus the pre-existing containment leak probability (4.1x105), is used in the revised risk model.

Figure 1 (on pages 12 and 13) presents the fault tree used to quantify split fraction 'CF1'. The results of, and input to, the quantification are displayed on Tables 5 (on page 14) and 6 (on page 15).

l L

a e , ,

=

.. 1940-98-20124 Attachment III Page 12 1

Figure 1 - Containment Spray Pump Trip Fault Tree (Page 1 of 2)

O eI N 4,:'n' LH Feasofts I

esenNN

.n= N!,"; i j

1 I I Ej A ed  !!st f4 fh I I setav , ump a Ye"

eu ese e-et-st Jurui,. "'=td"*' 'i=:."

h h eSeSAe IIACF1 NV311.

I I I I seN!n Naa a e svN!rN!sa  ! N n.. ...e

.t= usu. .-.

I i

b

(_ _ _ _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - ___ ----

,, 1940-98-20124 Attachment III  !

Page 13 Figure 1 - Containment Spray Pump Trip Fault Tree (Page 2 of 2)

.- 3:::"ei l I aNaY ed Es A A i I waat pour e mesms" mov t-si-s

=dit."liiN L'It't.!"!' ".lll%"

JU'nia escoce sacrt avasse i I I I sv$Nas a e svNEso MBf1305 M1880 SL36BD RLtage Masap RL3ese I

( . _ . _ . _ _ . _ _ . _ _ . . _ . _ _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _

1 s* .r l ,,

. 1940-98-20124 I Attachment III Page 14 Table 5: Containment Spray Pump Trip Results l

l J MODEL Name: STRAINER l Cause Table for Top Event CF and Split Fraction CF1 l PE Value of CF1 = 3.9298E-04 l

No. Cutsets Value Percent Percent Importance Cumulative l

1 CBCSCO

  • HACF1 6.56E-05 16.6905 16.6905 j 2 CBCSAO
  • HACF1 6.56E-05 16.6905 33.3809 l 3 HACF1
  • PS15CD 2.72E-05 6.9113 40.2922 i 4 HACFI
  • PS15BD 2.72E-05 6.9113 47.2036 5 HACF1
  • PS15AD 2.72E-05 6.9113 54.1149 6 HACF1
  • PS15DD 2.72E-05 6.9113 61.0262 7 HACF1
  • RL26AD 2.44E-05 6.1988 67.225 8 HACF1
  • RL25BD 2.44E-05 6.1988 73.4238 9 HACF1
  • RL25AD 2.44E-05 6.1988 79.6226 10 HACF1
  • RLK2BD 2.44E-05 6.1988 85.8214 11 HACF1
  • RL26BD 2.44E-05 6.1988 92.0202 12 HACF1
  • RLK2AD 2.44E-05 6.1988 98.219 13 CBCSAO
  • MV211D 1.19E-06 0.3026 98.5216 14 CBCSCO
  • MV215D 1.19E-06 0.3026 98.8242 15 MV215D
  • PS15BD 4.93E-07 0.1253 98.9495 16 MV215D
  • PS15DD 4.93E-07 0.1253 99.0748 17 MV211D
  • PS15CD 4.93E-07 0.1253 99.2001 18 MV211D
  • PS15AD 4.93E-07 0.1253 99.3255 19 MV215D
  • RLK2BD 4.42E-07 0.1124 99.4379 20 MV215D
  • RL26BD 4.42E-07 0.1124 99.5503 21 MV215D
  • RL25BD 4.42E-07 0.1124 99.6627 22 MV211D
  • RLK2AD 4.42E-07 0.1124 99.7752 23 MV211D
  • RL26AD 4.42E-07 0.1124 99.8876 24 MV211D
  • RL25AD 4.42E-07 0.1124 100 1

a* sa

.,

  • 1940-98-20124 Attachment III Page 15 Table 6: Containment Spray Pump Trip Results MODEL Name: STRAINER Basic Event Report for Top Event CF Basic Description Database Variable Event and Value CBCSAO SPRAY PUMP A CIRCUIT BREAKER TRIPS ZTCB10 = 6.4938E-04 CBCSCO SPRAY PUMP C CIRCUIT BREAKER TRIPS ZTCBIO = 6.4938E-04 HACF1 OPERATOR CLOSES DRYWELL INLET VALVE ZHECF1 = 1.0100E-01 V-21-5 MV211D MOV V-21-11 CLOSES ON DEMAND ZTVMOD = 1.8315E-03 MV215D MOV V-21-5 CLOSES ON DEMAND ZTVMOD = 1.8315E-03 PS15AD PRESSURE SWITCH IP15A ZTSWPD = 2.6891E-04 PS15BD PRESSURE SWITCH IP158 ZTSWPD = 2.6891E-04 PS15CD PRESSURE SWITCH IP15C ZTSWPD = 2.6891E-04 PS15DD PRESSURE SWITCH IP15D ZTSWPD = 2.6891E-04 RL25AD LOW DRYWELL PRESSURE RELAY 16K25A ZTRL1D = 2.4120E-04 RL25BD LOW DRYWELL PRESSURE RELAY 16K25B ZTRLID = 2.4120E-04 RL26AD LOW DRYWELL PRESSURE RELAY 16K26A ZTRLID = 2.4120E-04 RL26BD LOW DRYWELL PRESSURE RELAY 16K26B ZTRLID = 2.4120E-04 RLK2AD SPRAY PUMP START RELAY 16K2A ZTRL1D = 2.4120E-04 DEENERGIZES RLK2BD SPRAY PUMP START RELAY 16K2B ZTRLID = 2.4120E-04 DEENERGIZES l

,

Retrieved from "https://kanterella.com/w/index.php?title=ML20247F189&oldid=3158310"