ML031960554

Email - Draft Quad Cities SAMA RAI Responses

ML031960554
Person / Time
Site:	Quad Cities
Issue date:	06/30/2003
From:	Maher W Exelon Corp
To:	Mcdowell B, Palla R, Larry Wheeler Office of Nuclear Reactor Regulation, State of IL
Wheeler L, NRR/DRIP/RLEP, 415-1444
References
Download: ML031960554 (105)

Text

I Duke Wheeler - Draft Quad Cities SAMA RAI Resnonses Paae 1 a1 SAMA I Duke Wheeler - Draft Quad Cities RAI Res~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~onses Paae 1I~~~

From: "Maher, William D." <william.maher@exeloncorp.com>

To: 'Duke Wheeler (E-mail)" <dxw nrc.gov>, "McDowell Bruce (E-mail)'

<mcdowell5@llnl.gov>, 'Bob Palla (E-mail)' <rrp3@nrc.gov>

Date: 6/30/03 3:46PM

Subject:

Draft Quad Cities SAMA RAI Responses Attached you will find the draft RAI responses to the SAMA portion of Quad Cities RAls.

You will notice that the responses to Question #4 are not included. We will be able to speak to them during our phonecall Wed. and we believe that the impact to the overall evaluation is minimal.

If you should have any questions in the meantime, please feel free to contact me at any time.

Bill

<<<Quad~itiesRAIResponse-Draft 6-30.doc>>

This e-mail and any of its attachments may contain Exelon Corporation proprietary information, which is privileged, confidential, or subject to copyright belonging to the Exelon Corporation family of Companies.

This e-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this e-mail, you are hereby notified that any dissemination, distribution, copying, or action taken in relation to the contents of and attachments to this e-mail is strictly prohibited and may be unlawful. If you have received this e-mail in error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. Thank You.

CC: 'Jeff R. Gabor (E-mail)' <JRGaborGerineng.com>, 'Polaski, Fred W.'

<fred.polaski~exeloncorp.com>, 'Fulvio, Albert A.' <albert.fulvio~exeloncorp.com>, 'Nosko, John M."

<john.nosko@ exeloncorp.com>, 'Tzomes, Chancellor' <ca.tzomes@ exeloncorp.com>, 'Hersey, Kevin K.' <kevin.hersey@ exeloncorp.com>

RAI 1

The Severe Accident Mitigation Alternatives (SAMA) analysis is based on the most recent version of the Quad Cities Nuclear Power Station (QCNPS) Probabilistic Safety Assessment (PSA) for internal events, i.e., Revision 02B, which is a modification to the updated individual plant examination (1PE) submittal transmitted to the U.S. Nuclear Regulatory Commission (NRC) in December 1996. Please provide the following information regarding this PSA model:

a. a summary description of any peer reviews of the Level 1 and Level 2 portions of this PSA beyond the normally-perforned internal second checker reviews (e.g., QCNPS BWROG Peer Review, Independent Peer Review),

b. a characterization of the findings of these internal and external peer reviews (if any), and the impact of any identified weaknesses on the SAMA identification and evaluation process,

c. a breakdown of the internal events core damage frequency (CDF) by major contributors, initiators and accident classes, such as loss of offsite power (LOOP) [both single- and dual-unit], station blackout (SBO) [both single- and dual-unit], transients, anticipated transients without scram (ATWS), loss-of-coolant accident (LOCA), interfacing-systems loss-of-coolant accident (ISLOCA), internal floods, and other,

d. a description of the major differences from the updated IPE submittal, including the plant and/or modeling changes that have resulted in the new core damage frequency (CDF), along with the corresponding CDF.

Response 1(a):

[Provide] a summary description of any peer reviews of the Level 1 and Level 2 portions of this PSA beyond the nornally-performed internal second checker reviews (e.g., QCNPS BWROG Peer Review, Independent Peer Review)[.T Two external peer reviews of the 1999 Quad Cities Upgrade PRA were conducted.

NEI/BWROG Peer Review/Certification Conducted in the fall of 1999, with the report published in February of 2000, this review was performed by a six-member industry team following the latest NEI guidance available at the time.

1

Independent External Review Robert Schmidt of Scientech conducted a thorough external independent review of every aspect of the QC 1999 model, following a checklist of his own.

Response 1(b):

"(Provide) a characterization of the findings of these internal and external peer reviews (if any), and the impact of any identified weaknesses on the SAMA identification and evaluation processf.r NEI/BWROG Peer Review/Certification The NEI Certification team rated the QC PRA very well. The team specifically noted,

'The QUAD CITIES PSA is consistent with other industry PSAs in scope, methods, data usage, and results. The PSA does not have unique PSA features." Of the eleven "elements" evaluated by the team, a Summary Score of "4" was received for Systems Analysis. Summary Scores of "3"were assigned to all other elements. In the words of the review team, "These grades are consistent with a very solid PSA program with no major weaknesses.' There were no WA" level Facts & Observations (F&Os). There were a number of "B" level F&O's. The 2002 QC Update resolved all "B" F&Os and a number of 'C" F&O's, as well.

IndeDendent External Review The independent review by Robert Schmidt was conducted during 1999, with the report published in March of 2000. Mr. Schmidt's overall conclusion was 'The Quad Cities Updated PSA is a high quality Level I plus LERF PSA. All the technical elements meet or exceed general industry practice. The update process is well documented in analysis notebooks. No deficiencies were found in the analyses that need to be corrected immediately." In the 2002 update, EGC responded to all 29 of the comments Mr.

Schmidt recommended treating at the next update, plus the 13 that he recommended be treated some time in the future.

Response 1(c):

'[Provide)a breakdown of the internal events core damage frequency (CDF) by major contributors, initiators and accident classes, such as loss of offsite power (LOOP) lboth single- and dual-unit], station blackout (SBO) (both single- and dual-unit], transients, anticipated transients without scram (ATWS), loss-of-coolant accident (LOCA),

interfacing-systems loss-of-coolant accident (ISLOCA), internal floods, and other (contributors]."

2

The contribution to CDF by each initiator in the 2002 PRA Update is shown in Table 1-1.

Table 1-1 Contribution to CDF by Initiator Event Name Basic Event Description 2002 CDF (Iyr) % of 2002 CDF

%TDC LOSS OF 125VDC BUSES 1AND 2 7.6E-7 35.0%/O

%DLOOP DUAL UNIT LOSS OF OFFSITE POWER 3.7E-7 17.0%h

%TSW LOSS OF SERVICE WATER 3.OE-7 13.9%

%oTT TURBINE TRIP WITH BYPASS 1.2E-7 5.5%

%hTBCCW LOSS OF TBCCW 1.0E-7 4.8%

%S1 MEDIUM LOCA (WATER) 1.OE-7 4.8%

%hTIA LOSS OF INSTRUMENT AIR 6.8E-8 3.2%

%MS MANUAL SHUTDOWN 6.6E-8 3.0%

%TC LOSS OF CONDENSER VACUUM 5.4E-8 2.5%

%LOOP LOSS OF OFFSITE POWER 5.2E-8 2.4%

%A LARGE LOCA INITIATOR 4.5E-8 2.1%

%TF LOSS OF FEEDWATER 4.4E-8 2.0%

Other Other Initiating Events 8.3E-8 3.8%

Total 2.2E-6 100.0%

The ISLOCA CDF is 2.31 E-08/yr., or 1% of the Level 1 CDF.

ATWS is treated as a consequential event, not an initiator. The ATWS contribution is determined by the sum of the F-V importance of the mechanical failure to SCRAM and the electrical failure to SCRAM, which is 8% of the CDF.

SBO is a subset of all LOOP events. The contribution to the SBO event tree endstate (i.e., Class IB) is approximately 3.4E-7/yr, or 15% of the CDF.

Internal floods are not included in the 2002 QC internal events model. However, a separate flooding analysis recently completed yields a flooding CDF of 4.67E-7/yr. If this were added to the above internal events CDF, then the flooding contribution would be 18%.

3

Response 1(d):

'"ProvideJ a description of the major differences from the updated IPE submittal, including the plant and/or modeling changes that have resulted in the new core damage frequency (CDF), along with the corresponding CDF."

Plant Changes since Undated IPE Submittal

• Extended Power Uprate

• EOP and Miscellaneous Other Procedure Improvements

• Significant reduction in number of SCRAM's and significant improvement in equipment reliability and availability.

PRA Changes since Undated IPE Submittal

• Increased detail in loss of DC bus initiator

• Revised ATWS modeling

• Extended Power Uprate (EPU) plant configuration and MAAP 4.0.4 analysis

• Revised human reliability analysis (HRA) based on the most recent operator interviews and comments of Site Risk Management Engineer

• Completed URE, OPEX, and NON review efforts

• Maintenance unavailability data based on the most recent plant operating experience

• Bayesian updated initiating event frequencies utilizing Quad Cities most recent operating experience

• Individual component random failure probabilities Bayesian updated (as applicable) based upon the most recent plant specific data and the most current generic sources

• Common cause failure (CCF) calculations revised to incorporate the updated individual random basic event probabilities and the most up to date Multiple Greek Letter (MGL) parameters from NUREG/CR-5497 and NUREG/CR-5485

• Revised LOOP/DLOOP analysis for initiating event frequencies and non-recovery probabilities based upon a Midwest regional data filtering approach

• Revised DC distribution system CCF modeling (CCF events set to zero) to prevent double counting 4

• Revised mechanical and electrical ATWS probabilities, based on information in NUREG/CR-5500

• Response to Quad Cities BWROG Peer Review comments using the NEI PRA Peer Review Process (NEI 00-02)

• Response to additional independent Peer Review Comments

• Other open item comments from the review of the 1999 draft model

• Credit for repair/recovery of RHR for long term loss of DHR events It is not possible to determine the CDF change associated with each one of these model changes. However, a summary of the total calculated CDF for each of the relevant models is provided in Table 1-2.

Table 1-2 Quad Cities CDF History Model l Date I CDF (Per Yr)

• IPE 12/93 1.2E-06/yr

• Modified IPE 8/96 2.2E-06/yr

• Updated IPE 12/96 2.2E-06Iyr

• Conversion/Update 4/99 4.6E-06Iyr (1998 - 99 Update)

• Update Revision 02A 4/02 3.9E-06/yr

• Revision 02B 5/02 2.2E-06/yr 5

RAI2 The CDF cited and used in the SAMA analysis is based on the risk profile for intemal events at QCNPS Unit 1. Please provide the internal events CDF for Unit 2, and a discussion of the reasons for any differences from Unit 1. Discuss the impact on the SAMA analysis, including the impacts of external events, and results if the analysis were based on Unit 2 rather than Unit 1.

Response (2):

Internal Events Unit 2 CDF The Unit 2 internal events CDF is identical to that of Unit 1: 2.2 x E-06/yr.

Unit 2 Differences from Unit 1 There are several minor differences in plant configuration related to the internal events model.

• SSMP SYSTEM. There is an asymmetry in that the normalpreferred supply to Bus 31 for SSMP power is from Unit 1 (AC and DC) and the Unit 2 supply is the alternate (AC and DC). The power realignment for both AC and DC is manual and requires operator intervention. The Unit 2 PRA has Unit-2-specific logic modules for the power supplies to Bus 31 (AC and DC) to account for the preferred (non-symmetric) alignment to Unit 1.

• ADS SYSTEM. Unit 2 has one additional pressure control valve (PCV) in the air supply to each of the PCVs 1(2)-4722A and 1(2)-4722B (supply to Target Rock ADS valve 203-3A). These PCVs rely on the air system for motive power and require no other support systems. The Unit 2 model has a Unit-2-specific logic module for the air supply to the Target Rock ADS valve (2-0203-3A). In addition, the Unit 1 ADS system is comprised of four Electromatic Relief Valves (ERVs) and one Target Rock SRV. On Unit 2, the four ERVs have been replaced by Target Rock PORV's.

However, this has no effect on the PRA model.

• RHRSW SYSTEM. The power supply for MOV 1001-187A is not symmetric. The Unit 1 valve is powered from MCC 18-1A and the equivalent Unit 2 valve is powered from MCC 28-1 B. Since only spurious operation of this valve is modeled, there is no power dependency modeled, and no model changes were required.

• INSTRUMENT AIR SYSTEM. There are three asymmetries associated with the instrument air system. First, there is no equivalent Unit 2 6

component for the 1B instrument air receiver. Therefore, these failures are eliminated from the Unit 2 model. Second, there are three service air compressors at Quad (1A, 11B, and 2), and their output is always cross-tied. Only two of three Service Air Compressors (1A and 1B) are credited in the Unit 1 model, and they are powered from Unit 1. In order to take credit for two SACs for the Unit 2 PRA model, the 1B compressor is credited in the Unit 2 model. To ensure the correct power supply was identified in the Unit 2 quantification, a dependency was inserted into the logic. Finally, the swing IAC is powered only from Unit 1 (MCC 18).

• ATWS LOGIC POWER. Power to the Unit 1, Div 1 ARI/RPT logic is from 125 VDC Reactor Building Distribution Panel #1 (ckt. #15). Power to Unit 1, Div 2 ARI/RPT logic is from 125 VDC Turbine Building Bus 1B-1 (ckt.

1. 32). Power for the Unit 2, Div 1 ARI/RPT logic is from 125 VDC Turbine Building Main Bus 2A-1 (ckt. #4). Power to Unit 2 Div 2 ARI/RPT logic is from 125VDC Turbine Building Bus 2B-1 (ckt. #32). This identifies a minor asymmetry in the Div 1 ARI/RPT power supplies. Turbine Building Main 125VDC Bus 1A(2A) supplies the Div 1 power supplies for both units.

However, each Unit's Div 1 ATWS logic is powered from different sub panels. Unit 1 Div 1 ATWS logic is powered from RB Distribution Panel #1 which is fed by the Turbine Building Main 125VDC Bus 2A-1 which is fed by Turbine Building Main 125VDC Bus 2A. This is resolved by adding the failures of Bus 1(2) A-1 and its feed breaker to supply ARI DIV 1 control power (CKT BKR 8).

The Unit 2 model uses the same event trees and reliability database as the Unit 1 model. While these differences do appear in low-frequency cutsets, the effects of the fault tree differences are small enough that they do not affect the total internal events CDF. Therefore, the differences do not affect the SAMA analyses for internal events.

Extemal Events Unit 2 Fire CDF The fire CDF for Unit 2 as reported in the IPEEE is 7.1 x 1i0/yr., compared to a Unit 1 fire CDF of 6.6 x 10-5/yr.

Fire-Related Unit 2 Differences from Unit 1 Cable routing is not identical for Unit 1 and Unit 2. Two notable asymmetries in the risk profile result. The risk contribution from reactor feed pump fires in Unit 2 is approximately 10% higher than the corresponding contribution from Unit 1. This is because of the specific cable routing of the power supply circuit to MCC 29-2 in Unit 2, which is challenged by postulated Unit 2 RFP fires. The equivalent MCC in Unit 1 (MCC 19-2) is not exposed to such a challenge. The Unit 2 results also show a 4% risk 7

contribution form a postulated air compressor fire because of the proximity of cable trays containing critical circuits for Unit 2 HPCI, for SSMP, and for one train each of Unit 2 CS and RHR. Such exposure does not exist in the Unit 1 analysis.

These differences in fire risk profile are not large enough to affect the SAMA analysis.

Seismic-Related Unit 2 Differences from Unit 1 With modifications to each unit in response to the Seismic Margins Analysis, there is no significant difference in seismic vulnerabilities between the two units.

8

RAI3 In the Extended Power Uprate (EPU) Amendment application, EXELON indicates that the Level 2 analysis is based on NUREG/CR-6595. However, there is no such indication in the SAMA portion of the Environmental Report (ER). Based on the above, please provide a description of the following:

a. the changes in the Level 2 methodology since the updated IPE submittal, including major modeling assumptions, containment event tree (CEV) structure, binning of end states.

b. the methodology and criteria for binning CET endstates into release categories used in the Level 3 analysis. Include the definitions of the release characteristics listed in Column 2 of Table 4-5.

c. each release (consequence) category used in the Level 3 analysis (as listed in Column 1 of Table 4-5), the specific source tenms used to represent each release category, and a containment matrix describing the mapping of Level 1 results (plant damage state frequencies) into the various release categories.

Response 3(a):

"[Provide] the changes in the Level 2 methodology since the updated IPE submittal, including major modeling assumptions, containment event tree (CEV) structure, binning of end statesf.r The IPE, modified IPE, and updated IPE employed what some would call a simplistic Level 2 methodology. Many accident progression phenomena or failure modes were eliminated from consideration, based on experiments, MAAP calculations, or judgments concerning the likelihood of various phenomena. Core damage end states were coded for sequence characteristics that would affect the remaining phenomena affecting containment performance. Based on those characteristics, it was determined in what time range the vessel would fail, whether the pedestal area was dry or wet, whether containment sprays were operating, whether liner melt-through was likely, and whether containment vent was operated. Based on this information, it was determined which core damage end states resulted in containment failure, and which resulted in LERF.

Because of the limitations of the IPE Level 2 model, the model was revised for the 1999 QC PRA Upgrade. It was decided to use a simplified LERF model in the style of NUREG/CR-6595. The 1999 QC PRA was used for the Extended Power Uprate (EPU) submittal.

The submittal for License Renewal required Level 3 calculations. Therefore, Exelon decided to develop a full Level 2 PRA model for Quad Cities that meets standard 9

industry practices. The full Level 2 model was used for the License Renewal analyses, and that model also has now been incorporated in the 2002 QC PRA model. It is also the basis for LERF calculations for risk assessment.

A brief summary of the current Level 2 model compared to the 1999 Level 2 model that was used for the EPU submittal follows:

• No changes in modeling assumptions

• CET structure has been enhanced to include more top event nodes

• Old CET had LERF and non-LERF end states whereas the updated model has several release category bins (see Responses 3(b) and 3(c))

Response 3(b):

'[Provide] the methodology and criteria for binning CET endstates into release categories used in the Level 3 analysis. Include the definitions of the release characteristics listed in Column 2 of Table 4-5. M Each CET end state can be associated with a radionuclide source term bin, which covers a spectrum of similar potential scenarios and timing. Theoretically, it would be desirable in determining the point estimates of risk to evaluate the source terms for each sequence of each accident plant damage state. However, for purposes of risk presentation, the CET end states can also be characterized in such a manner as to combine similar "consequence impact" sequences within a CET end state.

The discrete nature of the radionuclide release categories means that the severe accident spectrum is divided up into bins, which then represent a group of severe accidents that have similar characteristics. These characteristics would imply similar public health consequences. It has been found in the past that the public health consequences are affected by a large number of governing features. The following portrays the radionuclide release category characterization used for Quad Cities.

Radionuclide Release Categories (CET End States)

The spectrum of possible radionuclide release scenarios is represented by a discrete set of categories or bins. The end states of the containment and phenomenological event sequences may be characterized according to certain key quantitative attributes that affect offsite consequences. These attributes include two important factors:

• Timing (e.g., early or late releases); and,

• Total quantity of fission products released.

10

Therefore, the containment event tree end states represent the source term magnitude and relative timing of the radionuclide release. The number of categories used for Quad Cities (i.e., 13) in the source term characterization offers a level of discrimination similar to that included in numerous published PRAs.

Timing Bins Three timing categories are used, as follows:

• Early (E) Less than time when evacuation is effective

• Intermediate (I) Greater than or equal to Early, but less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />

• Late (L) Greater than or equal to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

The definition of the categories is based upon past experience concerning offsite accident response:

• Early is conservatively assumed to include cases in which minimal offsite protective measures have been observed to be performed in non-nuclear accidents.

• Intermediate is a time frame in which much of the offsite nuclear plant protective measures can be assured to be accomplished.

• Late (>24 hours) are times at which the offsite measures can be assumed to be fully effective.

Radionuclide Release Magnitude Bins The assessment of plant response under postulated severe accident scenarios is a complex integrated evaluation. The primary and secondary containment building responses are sensitive to pressures, temperatures, flows, and event timings. These parameters also affect the operator action timings, the radionuclide release timings, and the mitigating system performance assessments. Therefore, the proper plant specific characterization of the severe accident progression is important to the realistic representation of the plant and highly desirable for the Level 2 assessment. These deterministic calculations provide the following information:

• The pressures and temperatures for various accident scenarios in the RPV, the drywell, the wetwell, and the reactor building; 11

• The times to reach these pressures and temperatures which is key to the assessment of recovery; (The time windows available for recovery actions must be estimated.)

• The source term magnitude and timing.

Five severity classifications associated with volatile or particulate releases are defined as follows:

• Hiqh (H) - A radionuclide release of sufficient magnitude to have the potential to cause prompt fatalities.

• Medium or Moderate (M) - A radionuclide release of sufficient magnitude to cause near-term health effects.

• Low (L) - A radionuclide release with the potential for latent health effects.

• Low-Low (LL) - A radionuclide release with undetectable or minor health effects.

• Negligible (OK) - A radionuclide release that is less than or equal to the containment design base leakage.

A relationship was then developed with the five release severity categories. The results of this partitioning are shown in Table 3-1.

Table 3-1 Release Severity Categorization Release Severity I Fraction of Released Csl Fission Products High greater than 10%

Medium/Moderate 1 to 10%0l Low 0.1 to .0%

Low-Low 1 ) less than 0.1%

Negligible much less than 0.1%

The resulting definitions of the radionuclide release end states are summarized in Table 3-2. The combinations of severity and timing classifications results in one OK release category and 12 other release categories of varying times and magnitudes. These 12 other release categories are shown in Table 3-3. These are the dominant release categories shown incolumn 2 of Table 4-5 of the Environmental Report.

12

Table 3-2 Release Severity And Timing Classification Scheme Release Severity Release Timing Classification Cs Iodide % Classification Rime of Initial Releaseral Category j Release Category REmergency Declaration High (H) Greater than 10 Late (L) Greater than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> Medium or Moderate 1 to 10 Intermediate (I) 5 to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> (M)

Low (L) 0.1 to 1 Early (E) Less than 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> Low-low (LL) Less than 0.1 No iodine (OK) 0 l (1) The conditions dictating a General Emergency are used as the surrogate for the time when EALs are exceeded, which in turn is used as the relative time to measure when the release occurs.

Table 3-3 Quad Cities Release Categories Time of Magnitude of Release Release H M L LL E WHE E IE LU/E I H/l MWII/l L/I L H/. MA. L11 LUI 13

Response 3(c):

fProvide] each release (consequence) category used in the Level 3 analysis (as listed in Column 1 of Table 4-5), the specific source terms used to represent each release category, and a containment matrix describing the mapping of Level 1 results (plant damage state frequencies) into the various release categories."

Source Terms used to Represent each Release Cateaorv As requested, Table 3-4 provides a list of the source terms associated with each of the release categories as listed inColumn 1 of Table 4-5 of the ER.

14

Table 3-4 Source Terms Assoclated with Each Release Category Release Category"1 11 1.2-11 I 1.2-2 I 1.2-3 L2-4 12-5 I 1.26 12-7 1.2-8 12-9 I 12-10 MAAP Run 000053 QC0082 NA OCOO85 000061 NA 0C0057 0C0058 0C0070 0C0074 Time after Scram when General Emergency Isdeclared 60 min 15 hr NA 55 min 15 hr NA 45 min 15 hr 20 min 60 min Fission Product Group: -=

1) Noble . .__

Total Release %at 36 Hours 94 100 NA 100 100 NA 86 100 100 0.31 Start of Release (hr) 4.4 hr 51.4 hr NA 55 min 39.3 hr NA 5.7 hr 25.9 hr 17 min 3.0 hr End of Release (hr) 4.4 hr 60 hr NA 4 hr 39.3 hr NA 5.7 hr 25.9 hr 1 hr 36.0 hr

2) Csl Total Release %at 36 Hours 28 33 NA 8.4 3.6 NA 1 0.14 96 2.00E-04 Start of Release (*r) 4.4 hr 60.7 hr NA 55 min 39.3 hr NA 5.7 hr 30.0 hr 17 min 3.0 hr End of Release (hr 4.4 hr 60.7 hr NA 2 hr 48.0 hr NA 10.0 hr 36.0 hr 1.0 hr 6.0 hr

3) TeO2 Total Release %at 36 Hours 16 12 NA 6.7 0.76 NA 0.28 0.26 77 8.50E-06 Start of Release (*r) 4.4 hr 60.7 hr NA 55 min 39.3 hr NA 5.7 hr 32.0 hr 17 min 3.0 hr End of Release (hr) 9.0 hr 65.0 hr NA 2 hr 48.0 hr NA 5.7 hr 36.0 hr 1.0 hr 6.0 hr

4) SrO - - - = -

Total Release % at 36 Hours 1.9 2 NA 2.7 0.41 NA 3.2 0.99 4.2 4.90E-05 Start of Release (hr) 4.4 hr 60.7 hr NA 7.2 hr 60.1 hr NA 5.7 hr 32.0 hr 17 min 6.0 hr End of Release (hr) 7.0 hr 65.0 hr NA 9.0 hr 65.0 hr NA 5.7 hr 36.0 hr 1.0 hr 6.0 hr

5) MoO2 . _ _

Total Release %at 36 Hours 3.00E-04 8.40E-04 NA 0.15 6.50E-03 NA 1.70E-04 3.10E-07 2.2 1.80E-07 Start of Release (hr) 4.4 hr 60.7 hr NA 55 min 39.3 hr NA 5.7 hr 25.9 hr 17 min 3.0 hr End of Release (hr) 4.4 hr 60.7 hr NA 6.0 hr 39.3 hr NA 5.7 hr 25.9 hr 1.0 hr 6.0 hr

6) CsOH _ III Total Release%at36 Hours 20 20 NA 7.8 l 1 NA 0.89 0.11 74 1.10E-04 Start of Release (hr) 4.4 hr 60.7 hr NA 55 min 39.3 hr NA 5.7 hr 32.0 hr 17 min 3.0 hr End of Release (hr) 10.0 hr 70.0 hr NA 2 hr l 48.0 hr NA 5.7 hr 36.0 hr 1.0 hr 6.0 hr 15

Table 3-4 Source Terms Associated with Each Release Category Release Category" 2 )

12-1 I 12-2 I 12-3 I 12-4 I L2-5 I L2-6 I 12-7 I 12-8 12-9 12-10 MAAP Run 0C0053 OC0082 NA Q 00085 000061 NA 00057 000058 0C0070 C00074 Time after Scram when General lEmerg=rc is declared 60 min 15 hr NA 55 min 15 hr NA 45 min 15 hr 20 min 60 min Fission Product Group: -

7) BaO . _

Total Release %at 36 Hours 0.83 0.87 NA 1.4 0.19 NA 1.4 0.43 4.7 2.OOE-05 lStar of Release (hr) 4.4 hr 60.7 hr NA 7.2 hr 60.1 hr NA 5.7 hr 32.0 hr 17 min 6.0 hr End of Release (hr) 7.0 hr 65.0 hr NA 9.0 hr 65.0 hr NA 5.7 hr 36.0 hr 6.0 hr 6.0 hr

8) La203 Total Release %at 36 Hours 0.23 0.25 NA 0.43 3.70E-02 NA 0.5 0.02 0.58 9.OOE 06 Start of Release (hr) 4.4 hr 60.7 hr NA 7.2 hr 60.1 hr NA 5.7 hr 32.0 hr 17 min 6.0 hr End of Release (hr) 7.0 hr 60.7 hr NA l 7.2 hr 65.0 hr NA 5.7 hr 36.0 hr 6.0 hr 6.0 hr

9) CeO2 _ _ 1.

Total Release %at 36 Hours 1.4 1.5 NA 1.9 0.27 NA 1.6 0.19 1.8 2.30E-05 Start of Release (hr) 4.4 hr 60.7 hr NA 7.2 hr 60.1 hr NA 5.7 hr 32.0 hr 5.5 hr 6.0 hr End of Release (hr) 7.0 hr 60.7 hr NA 7.2 hr 65.0 hr NA 5.7 hr 36.0 hr 8.0 hr 6.0 hr

10) Sb Total Release %at 36 Hours 44 24 NA 24 6.8 I NA 20 1.5 75 6.90E-04 Start of Release (hr) 4.4 hr 60.7 hr NA 55 min 60.1 hr NA 5.7 hr 32.0 hr 17 min 3.0 hr End of Release (hr) 14.0 hr 70.0 hr NA 10 hr 72.0 hr NA 5.7 hr 36.0 hr 1.0 hr 6.0 hr

-Te2l -

Total Release %at 36 Hours 0.77 0.83 NA 0.21 5.60E-02 NA 0.41 0.17 0.28 2.40E-05 Start of Release (hr) 4.4 hr 60.7 hr NA 7.2 hr 60.1 hr NA 5.7 hr 32.0 hr 5.5 hr 6.0 hr End of Release (hr) 14.0 hr 60.7 hr NA 7.2 hr 72.0 hr NA 5.7 hr 36.0 hr 5.5 hr 6.0 hr

12) U02 I I [.._

Total Release % at 36 Hours 7.OOE-03 7.OOE-03 NA 1.00E-02 1.20E-03 NA 1.20E-02 6.OOE-04 1.30E-02 2.20E-07 Start of Release (hr) 4.4 hr 60.7 hr NA 7.2 hr 60.1 hr l NA 5.7 hr 32.0 hr 5.5 hr 6.0 hr End of Rea r 6.0 hr 60.7 hr NA 7.2 hr 72.0 hr l NA 57 hr 36.0 hr 5.5 hr 6.0 hr (1) Puff releases are denoted In the table by those entries with equivalent start and end times.

(2) All cases run for 36 hrs. except 0C0082 and 0C0061 run for 72 hrs.

16

MapDing of Level 1 Results into the Various Release Categories One link between the Level 1 PSA accident sequences and the Containment Event Tree occurs in the definition of the Level 1 end states. The definition of the end states are developed to transfer the maximum amount of information regarding the accident sequence characteristics to the CET assessment. What follows summarizes the link between Level 1 end states and the entry condition to the CET such that a mapping of the Level 1 results into the various release categories can be provided.

A broad spectrum of accident sequences have been postulated that could lead to core damage and potentially challenge containment. The Quad Cities Level 1 PSA has calculated the frequency of those accident sequences that contribute to the core damage frequency for Quad Cities using system oriented (systemic) event trees. Each of these sequences may result in different challenges to containment. However, many of these challenges to containment have similarities in their functional failure characteristics. This has been confirmed in individual BWR PRAs including NUREG-1 150. The result is that these studies have categorized these containment challenges into a finite, discrete group of accident sequence bins, which have similar functional failures.

As pointed out in past BWR PRAs, different portions of the spectrum of postulated core damage accidents represent substantially different challenges to the containment depending upon the system failures and phenomena that have contributed to the sequence. Therefore, the containment event tree response must be capable of reflecting the entire spectrum of challenges to ensure that the following are explicitly incorporated:

• System failures inthe Level 1 evaluation (including support systems)

• Phenomenological interaction due to the type of core melt progression

• RPV conditions

- Pressures

- Decay heat level

• Containment conditions

• Timing of the sequence of events (i.e., core damage and containment failure (if applicable)).

Core Damage Functional Classes An event sequence classification into five accident sequence functional classes can be performed using the functional events as a basis for selection of end states. The description of functional classes is presented here to introduce the terminology to be used in characterizing the basic types of challenges to containment. The reactor pressure vessel condition and containment condition for each of these classes at the time of initial core damage is noted inTable 3-5.

17

Table 3-5 Core Damage Functional Classes (from the Level 1 Analysis)

Core Damage l RPV Condition Containment Functional Class [ Condition I Loss of effective coolant inventory (includes high and Intact low pressure inventory losses)

If Loss of effective containment pressure control, e.g., Breached or Intact heat removal III LOCA with loss of effective coolant inventory makeup Intact IV Failure of effective reactivity control Breached or Intact V LOCA outside containment Breached (bypassed)

In assessing the ability of the containment and other plant systems to prevent or mitigate radionuclide release, it is desirable to further subdivide these general functional categories. In the second level binning process, the similar accident sequences grouped within each accident functional class are further discriminated into subclasses such that the potential for system recovery can be modeled. The interdependencies that exist between plant system operation and the core melt and radionuclide release phenomena are represented in the release frequencies through the binning process involving these subclasses, as shown in past PRAs and PRA reviews. The binning process, which consolidates information from the systems' evaluation of accident sequences leading to core damage in preparation for transfer to the containment-source term evaluation, involves the identification of 18 classes and subclasses of accident sequence types. Table 3-6 provides a description of the possible subclasses used in the Quad Cities analysis.

The Accident Class designators and subclasses listed in Table 3-6 represent the core damage endstate categories from the Level 1 analysis that are grouped together as entry conditions for the Level 2 analysis. Each of the subclasses is then represented by a series of Containment Event Trees (CETs) to determine the Release Categorization for each of the accident scenarios. As such, the end states from the Level 2 analysis are assigned to one of the Release Categories noted in Table 3-3 as part of Response 3(b). The characterization of the Level 2 results (i.e., as H/E, MWI, etc., or Class V or OK) was then used to determine the frequency of the associated Consequence Category shown in Table 4-5 of the ER. Note that in this fashion, the Level 1 results are not directly linked to a release category, but rather the Level 2 endstate results based on the sum of all of the Release Category frequencies comprise the Consequence Category for each Phase II SAMA considered.

18

Table 3-6 Summary of the Core Damage Accident Sequence Subclasses WASH-1400 Accident Class Subclass Definition Designator Designator Example Class I A Accident sequences involving loss of inventory makeup TOUX in which the reactor pressure remains high.

B Accident sequences involving a station blackout and TgQUV loss of coolant inventory makeup.

C Accident sequences involving a loss of coolant TT'CMQU inventory induced by an ATWS sequence with containment intact.

D Accident sequences involving a loss of coolant TQUV inventory makeup in which reactor pressure has been successfully reduced to 200 psi.; i.e., accident sequences initiated by common mode failures disabling multiple systems (ECCS) leading to loss of coolant inventory makeup.

E Accident sequence involving loss of inventory makeup in which the reactor pressure remains high and DC power is unavailable.

Class II A Accident sequences involving a loss of containment TW heat removal with the RPV initially intact; core damage induced post containment failure L Accident sequences involving a loss of containment AW heat removal with the RPV breached but no initial core damage; core damage after containment failure.

T Accident sequences involving a loss of containment N/A heat removal with the RPV initially intact; core damage induced post high containment pressure V Class 11A or ILexcept that the vent operates as TW designed; loss of makeup occurs at some time following vent initiation. Suppression pool saturated but intact.

19

Table 3-6 Summary of the Core Damage Accident Sequence Subclasses

_ WASH-1400 Accident Class Subclass Definition Designator Designator Example Class Ill A Accident sequences leading to core damage conditions R (LOCA) initiated by vessel rupture where the containment integrity is not breached in the initial time phase of the accident.

B Accident sequences initiated or resulting in small or SIOUX medium LOCAs for which the reactor cannot be depressurized prior to core damage occurring.

C Accident sequences initiated or resulting in medium or AV large LOCAs for which the reactor is at low pressure and no effective injection is available.

D Accident sequences which are initiated by a LOCA or AD RPV failure and for which the vapor suppression system is inadequate, challenging the containment integrity with subsequent failure of makeup systems.

Class IV A Accident sequences involving failure of adequate TTCMC2 (ATWS) shutdown reactivity with the RPV initially intact; core damage induced post containment failure.

L Accident sequences involving a failure of adequate N/A shutdown reactivity with the RPV initially breached (e.g., LOCA or SORV); core damage induced post containment failure.

T Accident sequences involving a failure of adequate NWA shutdown reactivity with the RPV initially intact; core damage induced post high containment pressure.

V Class IV A or L except that the vent operates as N/A designed; loss of makeup occurs at some time following vent initiation. Suppression pool saturated but intact.

Class V Unisolated LOCA outside containment N/A 20

The CET calculation for each cutset uses Boolean logic and fault tree models to process the incoming Level 1 cutsets to ensure that the resulting Radionuclide release frequencies properly reflect the impact on release magnitude and timing of the containment and containment mitigation systems. A typical CET (for Accident Class 1A) is provided in Figure 3-1.

Figure 3-1 Typical Quad Cities Level 2 Containment Event Tree 21

In summary, the Level 1 end states do not translate directly into release categories.

Each Level 1 accident sequence (all of the cutsets) is transferred into the appropriate CET. The CET is then used to determine the resulting frequency for each radionuclide release end state from each incoming cutset. This is typical of a full Level 2 for a binned fault tree model. This approach does not involve a matrix that relates Level 1 sequences directly to Radionuclide end states..

Although not created as part of the normal calculation process, the results of the analysis can be binned to show the contribution to each release category by Level 1 end state. Table 3-7 shows the requested results for the base case 02B model.

Table 3-7 Matrix of Level I Results with Various Release Categories Base Case (02B Model)

Level 2 Release Category / Level 3 Consequence Category Level 1 IJI, LL/I Accident HIE HJI H/L° M/E Ml LZ LLE or IL, or Class V Intact Total Class (L2-1) (L2-2) (L2-3) (L2-4) (L2-5) (L2-6) (127) L2L- (L2-9) (L2-10) tA/1 E 1.2E-07 NWA 54E-09 5.9E-08 3.4E-08 NVA 9.6E-09 3.1 E-07 N/A 3.3E-07 8.7E-07 1BE 6.3E-10 NWA 0O.OE+00 O.OE+O0 6.2E-09 O.OE+0O 4.0E-11 1.6E-10 N/A I.5E-08 2.2E-08 1BL N/A 1.7E-08 O.OE+00 N/A 1.6E-07 O.OE+00 N/A 2.3E-09 N/A 1.3E-07 3.1E-07 iC O.OE+00 N/A O.OE+0O O.OE+00 O.OE+00 .OE+00 6.8E-12 0.OE+00 N/A 4.OE409 4.0E-09 1D O.OE+00 N/A O.OE+00 1.8E-11 1.9E-10 N/A O.OE+00 2.9E-11 N/A 9.7E-10 1.2E-09 2 2.4E-10 1.8E-08 N/A 3.3E-08 5.9E-07 N/A N/A N/A N/A O.OE+00 6.5E-07 3B 5.1E-10 O.OE+00 2.7E-11 N/A 1.2E-09 7.OE-1 1 3.9E-11 1.9E-09 N/A 9.7E-09 1.3E-08 3C 1.1E-07 N/A O.OE+00 NWA O.OE+00 O.OE+00 N/A O.OE+00 N/A O.OE+00 1.1E-07 3D 1.2E-08 N/A N/A N/A N/A N/A N/A N/A N/A O.OE+00 1.2E408 4A 6.8E-09 N/A NWA 1.6E-07 N/A N/A N/A N/A N/A O.OE+00 1.7E-07 5 N/A N/A N/A N/A N/A N/A WNA N/A 1.8E-08 O.OE+00 1.8E-08 Total: 2.5E-07 3.6E-08 5.5E-09 (') 2.5E-07 8.0E-07 7.0E-11 9.7E-09 3.2E-07 1.8E-08 5.0E-07 2.2E-06 (1)Included with the HA Consequence Category (L2-2) for evaluation purposes.

2 Included with the MW Consequence Category (L2-5) for evaluation purposes.

22

RAI4 Please provide the following information concerning the MELCOR Accident Consequences Code System (MACCS) analyses:

a. The MACCS analysis assumes all releases that occur at ground level and has a thermal content the same as ambient. These assumptions could be non-conservative when estimating offsite consequences.

Please provide an assessment of the sensitivity of offsite consequences (doses to the population within 50 miles) to these assumptions.

b. The discussion of meteorology indicates that there are data voids in the 2000 data set used. Interpolation was used between hours if only a brief period of data was missing, and hourly observations from the airport were used to fill larger data voids. Provide a characterization of the magnitude and extent of the data voids and the rationale for using the airport data rather than interpolation. Confirm that the 2000 data set is representative of the QCNPS site andjustify its use.

c. Clarffy the time periods used foram andpm forthe atmospheric mixing heights, (e.g., midnight to noon and noon to midnight, versus sunrise to sunset.)

Response 4(a):

[To be provided - TTNUS]

Response 4(b):

[To be provided - TTNUS]

Response 4(c):

[To be provided - TTNUS]

23

RAIS According to Table F-1 of the Environmental Report (ER), Exelon evaluated 280 SAMA candidates. Of these 280 candidates, 30 were obtained from QCNPS-specffic documents. It is not clear that the set of SAMAs evaluated in the ER addresses the major risk contributors for QCNPS. In this regard, please provide the following:

a. a description of how the dominant risk contributors at QCNPS, including dominant sequences and cut sets from the current Probabilistic Risk Assessment (PRA) and equipment failures and operator actions identified through importance analyses (e.g., Fussell-Vesely, Risk Reduction Worth, etc.) were used to identify potential plant-specific SAMAs for QCNPS.

b. the number of sequences and cut sets reviewed/evaluated and what percentage of the total CDF they represent

c. a listing of equipment failures and human actions that have the greatest potential for reducing risk at QCNPS based on importance analysis and cut set screening.

d. for each dominant contributor identified in the current PRA (Revision 02B), a cross-reference to the SAMAs evaluated in the ER which addresses that contributor. If a SAMA was not evaluated for a dominant risk contributor, justify why SAMAs to further reduce these contributors would not be cost beneficial.

e. a general description of the group of 81 insights mentioned in the original IPE and a discussion of how and whether insights not implemented were factored into the SAMA evaluation.

Response 5(a):

'[ProvideJ a description of how the dominant risk contributors at QCNPS, including dominant sequences and cut sets from the current Probabilistic Risk Assessment (PR4) and equipment failures and operator actions identified through importance analyses (e.g., Fussell-Vesely, Risk Reduction Worth, etc.) were used to identify potential plant-specific SAMAs for QCNPS' A review of the CDF-based Risk Reduction Worth (RRW) rankings for the current model was performed. The rankings of these equipment failures, operator actions, and initiating events were checked to determine if any items could be beneficial that were not addressed by the existing SAMA list. The examination of the dominant RRW basic events encompassed the dominant sequences and cut sets from the current PRA model. RAI response 5(d) provides a more detailed discussion of this importance ranking review.

24

Response 5(b):

"[Provide]the number of sequences and cut sets reviewed/evaluated and what percentage of the total CDF they represent" The CDF-based RRW listing was reviewed down to and including the 1.02 level, which indicates the events below this point would influence the CDF by less than 2.0%. This corresponds to about a $2,000 averted cost-risk based on CDF reduction assuming 100% reliability of the associated event. An evaluation of the top LERF-based contributors to RRW was also performed. It was determined that a similar averted cost of about $2,000 would be obtained by examining the LERF-based RRW factors down to 1.10. RAI response 5(d) provides a more detailed discussion of the importance ranking review and the results.

Response 5(c):

"[Provide]a listing of equipment failures and human actions that have the greatest potential for reducing risk at QCNPS based on importance analysis and cut set screening.'

RAI response 5(d) provides a listing of equipment failures, human actions, and initiating events that have the greatest potential for reducing risk at QCNPS based on importance analysis and cut set screening.

Response 5(d):

'[Provide)for each dominant contributor identified in the current PRA (Revision 02B), a cross-reference to the SAMAs evaluated in the ER which addresses that contributor. If a SAMA was not evaluated for a dominant risk contributor, justify why SAMAs to further reduce these contributors would not be cost beneficial."

Table 5-1 (for CDF) and Table 5-2 (for LERF) provide a correlation between the events identified in the QCNPS PSA model (Revision 02B) that are considered to have the greatest potential for reducing risk and their relationship to the SAMAs evaluated in the Environmental Report.

The events included in Table 5-1 are based on the core damage frequency RRW factors down to and including RRW values of 1.02. The events included in Table 5-2 are based on the large early release frequency RRW factors down to an RRW value of 1.10. Both of these RRW factors correspond to potential averted cost risk of about

$2,000. The events below this point are judged to be highly unlikely contributors to the identification of cost-beneficial enhancements.

25

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%TDC 1.50E-06 1.54 LOSS OF 125VDC BUSES 1 This event represents the unlikely initiating event of a complete AND 2 INITIATING EVENT loss of both 125V DC buses. Many SAMAs were included that address potential enhancements for DC reliability and/or alternate means of providing DC power. Phase I SAMAs 93, 94, 97, 98, 99, 100, 114, 125, 126, 127, 128, 129, and 131 are all related to improved DC performance. Phase I SAMAs 94 and 131 were retained for further examination as Phase II SAMAs 3 and 6, respectively. No additional SAMAs were suggested for this broad topic.

1DCRX-BUS1 RECF- 7.1 OE-01 1.54 FAILURE TO RECOVER This event involves failure to recover one of the 125V DC buses UNIT 1 BATTERY BUS #1 given loss of both. See disposition above for %TDC (Loss of 125V DC Buses 1 and 2 Initiating Event).

2DCRX-BUS2RECF- 7.1OE-01 1.54 FAILURE TO RECOVER This event involves failure to recover one of the 125V DC buses UNIT 2 BATTERY BUS #2 given loss of both. See disposition above for %TDC (Loss of 125V DC Buses 1 and 2 Initiating Event).

1RHOPREPAIRTRH-- 2.60E-01 1.27 FAILURE TO RECOVER/ This event represents the failure to recover or repair REPAIR SPC BEFORE suppression pool cooling prior to venting. Potential VENT (TRANSIENT/IORV) improvements to the reliability of the RHR heat exchangers were examined in Phase I SAMAs 20 and 22. Alternate means of providing containment heat removal were also examined in Phase I SAMAs 35, 36, 37, 38, 39, 40, 53, 55, 66, 74, 75, 76, 83, 213, 214, and 265. Improvements in the response to containment heat removal events were examined in Phase I SAMAs 277, 278, 279, and 280. Phase I SAMAs 36, 265, and 279 were retained as Phase II SAMAs 2, 13, and 14, respectively. No additional SAMAs were suggested for this broad topic.

26

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%DLOOP 1.20E-02 1.20 DUAL UNIT LOSS OF This event is a dual unit loss of offsite power event.

OFFSITE POWER Improvements related to enhanced AC or DC reliability or availability were considered in Phase I SAMAs 91 through 131.

Many other SAMAs were also considered that would provide mitigation benefits in loss of offsite power scenarios including Phase II SAMAs 1, 2, 3, 4, 5, 6, 8,10, and 13. No additional SAMAs were suggested for this broad topic.

%TSW 5.27E-03 1.16 LOSS OF SERVICE WATER This event is the loss of service water initiating event. Potential INITIATING EVENT improvements and enhancements to the service water system were examined in Phase I SAMAs 10, 20, 21, and 23. No additional SAMAs were suggested, and no related SAMAs were retained for Phase II. It is noted that in Phase I SAMA 23, the cost of installing an additional service water pump had been estimated at approximately $5.9 million which Is greater than the maximum averted cost risk (even if large uncertainties and external events are considered).

BACRXDLOOP4HRH-- 2.20E-01 1.16 FAILURE TO RECOVER This event signifies the time available to recover power prior to DLOOP WITHIN 4 HRS battery depletion. Potential improvement to battery life by using fuel cells instead of lead-acid batteries was examined in Phase I SAMA 94 which was retained as Phase II SAMA 3. The cost benefit analysis indicated a potential averted cost-risk of $4,406.

The benefit would not be much greater from including fire external events since the Quad Fire PRA results are dominated by loss of decay heat removal scenarios, for which extended battery life would not come into play. The relatively low benefit also excluded other potential low cost alternatives to extending battery life such as portable chargers.

27

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition BSSOPSSRMCLNGH-- 1.1OE-01 1.10 OP ACT: ALIGN FP TO This event represents the human error probability of providing SSMP ROOM COOLERS the alternate SSMP room cooling via manual alignment to the (OCOP 2900-02) Fire Protection System. Phase I SAMA 32 included an examination of providing alternate SSMP room cooling. This SAMA was retained as Phase II SAMA 1that resulted in a potential averted cost risk of $11,303. It was estimated that the cost of implementing a backup or automating the existing backup system would be substantially higher than the potential averted cost. Also see revised Phase II SAMA disposition in Table 7-3.

1RPCDRPS-MECHFCC 2.10E-06 1.09 MECHANICAL SCRAM This event represents the Mechanical Scram failure probability FAILURE based on the NUREG/CR-5500 INEEL evaluation of a representative BWR RPS system. Potential improvements to minimize the risks associated with ATWS scenarios were explored in Phase I SAMAs 227-243. Phase I SAMAs 242 and 243 were retained as Phase II SAMAs 11 and 12, respectively.

No additional SAMAs were suggested for this broad topic.

BDGCBEDG/SBOSKCC 4.83E-05 1.06 CCF OF ALL EDG/SBO This event represents the unlikely event of all of the diesel OUTPUT CIRCUIT generator output breakers failing to close leading to an SBO BREAKERS TO CLOSE scenario. See disposition above for %DLOOP (Dual unit Loss of Offsite Power).

%TT 8.81 E-01 1.06 TURBINE TRIP WITH This event represents the turbine trip initiating event frequency.

BYPASS Industry efforts over the last fifteen years have led to a significant reduction in the number of reactor scrams and turbine trips. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

28

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description ' Disposition 1CNPVDWRUPT--R-- 6.00E-02 1.06 LARGE DW CONTAINMENT This event represents the scenario where an unmitigated FAILURE CAUSES LOSS containment pressurization results in a large drywell region OF INJECTION containment failure leading to a loss of all Injection systems.

This scenario can be avoided by providing improved decay heat removal methods. See disposition above for 1RHOPREPAIRTRH-- (Failure to recover/repair SPC before vent).

%TBCCW 4.92E-03 1.05 LOSS OF TBCCW This event represents the loss of TBCCW initiating event INITIATING EVENT frequency. Phase I SAMA 20 explored enhanced procedural guidance for use of cross-tied component cooling or service water pumps. The current procedural guidance was deemed adequate for service water, DGCW, and RHRSW, but inter-unit cross-tie capability does not exist for RBCCW or TBCCW. A separate analysis examines the potential cost-benefit of implementing an inter-unit TBCCW cross-tie capabilities (see Response 7(c)).

%S1 3.80E-04 1.05 MEDIUM LOCA (WATER) This event represents the medium LOCA water line break INITIATOR initiating event frequency. The dominant cutsets associated with this initiator include common cause failures of ECCS strainers or pre-initiator HEPs for miscalibration of pressure switches. Both of these types of events are extremely unlikely, but are included in the model for completeness. No additional SAMAs were suggested.

1MSOPMSIVINLKH-- 9.10E-01 1.05 OP ACT: BYPASS LOW This event represents the human error probability of bypassing LEVEL MSIV INTERLOCK the MSIV isolation as directed in the EOPs. This action requires GIVEN FAILURE TO SCRAM the use of jumpers with a limited time available, and as such carries a relatively high HEP value. A dedicated switch for bypassing the low level interlock would be desirable. This issue was specifically examined in Phase I SAMA 237 that was listed as retained, but did not specifically involve a Phase II SAMA analysis. The potential benefit of implementing a dedicated low level interlock switch is also examined (see Response 7(c)).

29

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description l Disposition 1RSMV1 001 -5ABDCC 2.OOE-04 1.04 RHR HX RWRSW OUTLET This event represents the unlikely failure of the RHR heat VALVES MOV 1-1001-5A exchanger RHRSW outlet valves leading to a loss of AND 5B AND FAIL TO OPEN suppression pool cooling capabilities. See disposition above for 1RHOPREPAIRTRH-- (Failure to recover/repair SPC before vent).

BDGDGRUN-----XCC 2.94E-05 1.04 CCFTR OF ALL EDGs & This event represents the unlikely failure of all of the diesel BOTH SBOs generators failing to run leading to an SBO scenario. See disposition above for %DLOOP (Dual unit Loss of Offsite Power).

BSS--MAINT---M-- 2.26E-02 1.04 SSMP SYSTEM This events represents the SSMP Maintenance unavailability UNAVAILABLE DUE TO probability. SSMP is a risk significant system with performance MAINTENANCE monitored as part of the Maintenance Rule activities. Potential improvements to SSMP reliability/operation were examined in Phase I SAMAs 32, 217, and 218. Altemate means of providing injection to the RPV were examined in Phase I SAMAs 184, 185,186,192, 208, 210, 211, 212, and 215. Phase I SAMA 32 was retained as Phase II SAMA 1. No other SAMAs were suggested.

%TIA 1.22E-02 1.03 LOSS OF INSTRUMENT AIR This event represents the loss of instrument air initiating event INITIATOR frequency. Potential improvements to air/gas systems were examined in Phase I SAMAs 222-226. No SAMAs were initially retained for Phase II, and no additional SAMAs were suggested.

However, a more thorough examination of the Quad Cities Fire PRA leads to a potential benefit being identified by providing an aftemate air source to the containment vent valves. The potential benefit of implementing such a change Is also explored (see Phase II SAMA 17).

30

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%MS 3.07E+00 1.03 MANUAL SHUTDOWN This event represents the manual shutdown initiating event INITIATING EVENT frequency. Industry efforts over the last fifteen years have led to a significant reduction in the number of manual shutdowns and scrams from all causes. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

1CNFLMLLOCA--PCC 1.OOE-04 1.03 COMMON CAUSE This event represents the unlikely occurrence of a common PLUGGING OF ECCS cause failure of the ECCS suction strainers. The Quad Cities SUCTION STRAINERS strainers have recently been upgraded and re-sized such that the potential for common cause plugging has been reduced. No additional SAMAs were suggested.

%TC 7.90E-02 1.03 LOSS OF CONDENSER This event represents the loss of condenser vacuum initiating VACUUM event frequency. Industry efforts over the last fifteen years have led to a significant reduction in the number of plant scrams from all causes. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

1IARXRCOVERIAH-- 1.48E-01 1.03 OP ACT: RESTORE lAS This event represents the restoration of instrument air given AFTER IE OR RANDOM instrument air system loss in time for containment venting. See FAILURE FOR VENTING disposition above for %TIA (Loss of Instrument Air Initiator).

%LOOP 1.35E-02 1.02 LOSS OF OFFSITE POWER This event represents the single unit loss of offsite power INITIATING EVENT initiating event frequency. See disposition above for %DLOOP (Dual unit Loss of Offsite Power).

BDGDGSTART---ACC 1.88E-05 1.02 CCFTS OF ALL EDGs & This event represents the unlikely failure of all of the diesel BOTH SBOs generators failing to start leading to an SBO scenario. See disposition above for %DLOOP (Dual unit Loss of Offsite Power).

31

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 1--RX-SPC-SSCH-- 1.OOE-06 1.02 OP FAILS TO INITIATE SPC, This event represents the unlikely scenario of combined CONTROL CCST, AND operator action failures for three separate actions that otherwise ALIGN FP TO SSMP are evaluated independently. This event is included for completeness as part of the human reliability dependency analysis. Phase I SAMAs 266 and 271 examine potential improvements in operator performance. No additional SAMAs were suggested for this topic.

1--RX-HPI-ADSH-- 1.1OE-04 1.02 OPERATOR FAILS TO This event represents the unlikely scenario of combined INITIATE HPCI/RCIC/SSMP operator action failures for separate actions that otherwise are AND ADS evaluated independently. This event is included for completeness as part of the human reliability dependency analysis. Phase I SAMAs 266 and 271 examine potential improvements in operator performance. No additional SAMAs were suggested for this topic.

1RHMV1 6AB----KCC 1.1OE-04 1.02 RHR-HX BYPASS VALVES This event represents the unlikely failure of the RHR heat 16A AND 16B FAIL TO exchanger bypass valves leading to a loss of suppression pool CLOSE DUE TO COMMON cooling capabilities. See disposition above for CAUSE 1RHOPREPAIRTRH-- (Failure to recover/repair SPC before vent).

1CAHU263-52ABHCC 8.OOE-05 1.02 PREINIT: CAS PRESSURE This event represents the unlikely scenario of miscalibration of SWITCHES 52A AND 52B pressure switches leading to unavailability of ECCS injection.

MISCALIBRATED This Is included for completeness inthe model since it has the potential of leading to core damage following a medium or large LOCA initiating event. Improvement to maintenance procedures/ manuals was examined in Phase I SAMA 259. No additional SAMAs are suggested for this topic.

32

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition

%A 1.80E-04 1.02 LARGE LOCA INITIATOR This event represents the Large LOCA initiating event frequency. Mitigation from such an event would be improved by the existence of more reliable or diverse low pressure Injection systems and water sources. Such potential improvements were examined in Phase I SAMAs 60, 170,182, 184,187,188, 195, 201, 204, 212, 215, and 250. None of these SAMAs were maintained for Phase II,and no additional SAMAs were suggested.

BDCBY125VDC--FCC 1.24E-06 1.02 COMMON CAUSE FAILURE This event represents the unlikely scenario with common cause OF UNIT 1 AND UNIT 2 failure of both 125V DC batteries. See disposition above for 125VDC BATTERIES %TDC (Loss of 125V DC Buses 1 and 2 Initiating Event).

%TF 1.90E-02 1.02 LOSS OF FEEDWATER This event represents the loss of feedwater initiating event frequency. Industry efforts over the last fifteen years have led to a significant reduction in the number of plant scrams from all causes. Many of the SAMAs explored potential benefits for mitigation from these events. No additional SAMAs were suggested for this broad topic.

1RHMVI 8AB----DCC 1.01 E-04 1.02 MIN-FLOW MOVS 18A AND This event represents the unlikely failure of the RHR min-flow 18B FAIL TO OPEN DUE TO valves leading to a loss of suppression pool cooling capabilities.

COMMON CAUSE See disposition above for 1RHOPREPAIRTRH-- (Failure to recover/repair SPC before vent).

1LIOP-LPFILL-H-- 1.80E-02 1.02 OP ACT: PRVNT OVRFL OF This event represents the human error probability to prevent RPV DUE TO UNCNTRLD uncontrolled injection and overfill in ATWS scenarios. Many INJCTION W/ DPRS & USE potential improvements to minimize the risks associated with 0 ATWS scenarios were explored in Phase I SAMAs 227-243.

Phase I SAMAs 242 and 243 were retained as Phase II SAMAs 11 and 12, respectively. No additional SAMAs were suggested for this broad topic.

33

Table 5-1 Correlation of CDF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 1RSHU-MISCAL1 HCC 8.OOE-05 1.02 PREINIT: RHRSW PUMPs A, This event represents the unlikely pre-initiator failure of the B, C, and D RUNNING RHRSW pumps leading to a loss of suppression pool cooling LOGIC COMMON MISCAL. capabilities. See disposition above for 1RHOPREPAIRTRH--

(Failure to recover/repair SPC before vent).

1SLEV-1 106A/BDCC 1.40E-02 1.02 SBLC EXPLOSIVE VALVES This event represents the common cause failure of the SBLC FAILURE TO OPEN DUE TO explosive valves. Phase 1 SAMA 242 specifically examined the CCF potential benefit from diversifying the SBLC explosive valve operation. This SAMA was retained as Phase II SAMA 11. The averted cost-risk was determined to be $2,390, and it was judged that any hardware changes to the SBLC explosive valves would exceed this potential averted cost.

34

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 1RXSY-RXFAIL-FSU 1.OOE+00 7.89 FAILURE OF RX (CLASSES This event is a Level 2 sequence marker flag identifying those ID,IE (OP=F), II, IIIA, IIIC, sequences where the RX node has failed (i.e., where core hID, IV) damage was not terminated prior to the time of vessel failure).

The capability to enhance or provide additional injection systems was examined in Phase I SAMAs 19, 32, 172,182, 184-188, 191, 192,194-196, 200, 201, 203-205, 207-212, 215, 217, and 219-221. Phase I SAMAs 32, 219, 220, and 221 were retained as Phase II SAMAs 1, 8, 9, and 10, respectively. No additional SAMAs were suggested.

1GVPH-INERT--X- 9.90E-01 6.09 CONTAINMENT INERTED; This event is effectively a Level 2 sequence marker flag that VENTING REQUIRED represents the normal operating condition with the containment inerted. No additional SAMAs were suggested.

1SIPHCONTFAILF-- 1.00E+00 1.69 DW SHELL MELT- This event represents the evaluated likelihood from the Level 2 THROUGH FAILURE DUE analysis that a dry containment floor will lead to shell liner TO CONT. FAILURE failure (i.e., containment failure) after vessel failure for accident classes 11,1I1D, and IV. The importance of this phenomena would be reduced by the presence of more reliable or diverse injection systems, more reliable or diverse drywell spray systems, and other aHemate means to avoid this situation.

SAMAs related to improved injection system performance are discussed Inthe disposition for 1RXSY-RXFAIL-FSU above.

Items related to improved drywell spray performance were considered in Phase I SAMAs 36, 37, 53, 55, and 83. Phase I SAMA 36 was retained as Phase II SAMA 2. Altemate strategies for reducing the potential for drywell shell melt-through were also examined in Phase I SAMAs 44, 45, 48, 49, 51, 57, 58, and 87. None of these, however, were retained for Phase II, and no additional SAMAs were suggested.

35

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 1OPPH-PRESBK-F-- 8.00E-01 1.68 PRESSURE TRANSIENT This event represents a Level 2 phenomena event that would DOES NOT FAIL lead to a depressurized state. Potential improvements to the MECHANICAL SYSTEMS current depressurization capabilities and methods were examined in Phase I SAMAs 197, 198, 224, 245, 246, 253, 256, 257, and 263. None of these, however, were retained for Phase II, and no additional SAMAs were suggested.

1OPPH-SORV---F-- 5.50E-01 1.68 SRVs DO NOT FAIL OPEN This event also represents a Level 2 phenomena event that DURING CORE MELT would lead to a depressurized state. See disposition above for PROGRESSION IOPPH-PRESBK-F-- (Pressure transient does not fail mechanical systems).

1OPPH-TEMPBK-F-- 7.OOE-01 1.68 HIGH PRIM SYS TEMP This event also represents a Level 2 phenomena event that DOES NOT CAUSE FAIL OF would lead to a depressurized state. See disposition above for RCS PRESS. BOUND 1OPPH-PRESBK-F-- (Pressure transient does not fail mechanical systems).

%TDC 1.50E-06 1.67 LOSS OF 125VDC BUSES 1 This event also appears in the CDF importance listing in Table AND 2 INITIATING EVENT 5-1. It represents the unlikely initiating event of a complete loss of both 125V DC buses. Many SAMAs were included that address potential enhancements for DC reliability and/or altemate means of providing DC power. Phase I SAMAs 93, 94, 97, 98, 99,100,114,125,126, 127, 128,129, and 131 are all related to improved DC performance. Phase I SAMAs 94 and 131 were retained for further examination as Phase II SAMAs 3 and 6, respectively. No additional SAMAs were suggested for this broad topic.

1DCRX-BUS1 RECF-- 7.10E-01 1.67 FAILURE TO RECOVER This event also appears in the CDF importance listing in Table UNIT 1 BATTERY BUS #1 5-1. It involves failure to recover one of the 125V DC buses given loss of both. See disposition above for %TDC (Loss of 125V DC Buses 1 and 2 Initiating Event).

36

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 2DCRX-BUS2RECF-- 7.10E-01 1.67 FAILURE TO RECOVER This event also appears in the CDF importance listing in Table UNIT 2 BATTERY BUS #2 5-1. It involves failure to recover one of the 125V DC buses given loss of both. See disposition above for %TDC (Loss of 125V DC Buses 1 and 2 Initiating Event).

1OPOP-DEPRESSH-- 5.20E-01 1.63 OP FAILS TO DEPRESS This event represents the conditional failure probability used in GIVEN OP FAILED INLVL1 the Level 2 analysis for operators to depressurize prior to vessel OR LOSS OF DC failure given that depressurization was unsuccessful to avert core damage. Potential improvements to the current depressurization capabilities and methods were examined in Phase I SAMAs 197, 198, 224, 245, 246, 253, 256, 257, and 263. None of these, however, were retained for Phase II, and no additional SAMAs were suggested.

%S1 3.80E-04 1.39 MEDIUM LOCA (WATER) This event also appears in the CDF importance listing in Table INITIATOR 5-1. It represents the medium LOCA water line break initiating event frequency. The dominant cutsets associated with this initiator include common cause failures of ECCS strainers or pre-initiator HEPs for miscalibration of pressure switches. Both of these types of events are extremely unlikely, but are included in the model for completeness. No additional SAMAs were suggested.

1SIPH-DWHEAD-F-- 5.OOE-01 1.30 DRYWELL HEAD CLOSURE This event is a Level 2 phenomena event that represents the FAILS DUE TO probability that a high pressure vessel failure scenario will lead OVERPRESSURE to an early containment failure given that water exists on the drywell floor at the time of vessel failure. The importance of this event would be minimized by reducing the number of high pressure vessel failure scenarios. See disposition above for 1OPOP-DEPRESSH- (Operator fails to depressurize given failed in Level 1 or loss of DC). No additional SAMAs were suggested.

37

Table 5-2 Correlation of LERF Importance Listing to Evaluated SAMAs Event Name Probability RRW Basic Event Description Disposition 1CNFLMLLOCA--PCC 1.00E-04 1.27 COMMON CAUSE This event also appears in the CDF importance listing in Table PLUGGING OF ECCS 5-1. It represents the unlikely occurrence of a common cause SUCTION STRAINERS failure of the ECCS suction strainers. The Quad Cities strainers have recently been upgraded and re-sized such that the potential for common cause plugging has been reduced. No additional SAMAs were suggested.

1CAHU263-52ABHCC 8.OOE-05 1.21 PREINIT: CAS PRESSURE This event also appears in the CDF importance listing in Table SWITCHES 52A AND 52B 5-1. It represents the unlikely scenario of miscalibration of MISCALIBRATED pressure switches leading to unavailability of ECCS injection.

This is included for completeness in the model since it has the potential of leading to core damage following a medium or large LOCA initiating event. Improvement to maintenance procedures/ manuals was examined in Phase I SAMA 259. No additional SAMAs are suggested for this topic.

%A 1.80E-04 1.20 LARGE LOCA INITIATOR This event also appears in the CDF importance listing InTable 5-1. It represents the Large LOCA Initiating event frequency.

Mitigation from such an event would be Improved by the existence of more reliable or diverse low pressure injection systems and water sources. Such potential improvements were examined in Phase I SAMAs 60,170,182, 184, 187,188, 195, 201, 204, 212, 215, and 250. None of these SAMAs were maintained for Phase II, and no additional SAMAs were suggested.

1SIPH-SI2-NOTFSU 5.OOE-01 1.11 DRYWELL SHELL INTACT This event represents the complement to the Level 2 (OP=F) phenomena event 1SIPH-DWHEAD-F-- discussed above. As such, no additional SAMAs were suggested.

38

Response 5(e):

"[Provide)a general description of the group of 81 insights mentioned in the original IPE and a discussion of how and whether insights not implemented were factored into the SAMA evaluation.'

One of the important means of identifying plant specific improvements for the Quad Cities SAMA analysis was a review of the plant's IPE. As part of the IPE, an analysis of the cutsets and importance rankings was performed in order to identify plant weaknesses and to suggest changes that would address the weaknesses identified.

There were a total of 172 items that were developed from the IPE that were later categorized as IPE or Accident Management insights. These items generally consisted of the following types of improvements:

• Accident Management insights (70)

• Potential procedural enhancements (57)

• Potential hardware modifications (24)

• Mention of good practices (13)

• Recommendations for better data tracking of reliability performance (4)

• Suggestions for training or analysis (2)

• Simple information only (2)

A review of these insights indicates that the disposition is as follows:

• Accident management insights from several sites including Quad Cities were carefully considered by the BWROG in developing the EOPs and SAMGs that have been subsequently implemented at Quad Cities.

Authors of the plant-specific QC SAMG's also reviewed and incorporated, as appropriate, the Quad Accident Management Insights from the Quad IPE. No additional action required.

• Of the 57 potential procedural enhancements, 13 were found to have been addressed with subsequent revisions of the procedures. Of the remaining 44 procedural insights, 21 were found to have been addressed in other procedures, 14 were found to provide superfluous information to existing procedures, and 9 were found to be too specific to provide useful information in the symptom-based procedures. No additional action required.

• Of the 24 hardware modifications, 7 were determined to be unnecessary and 2 have been made irrelevant through implementation of the Maintenance Rule. The remaining 15 hardware modifications are safety improvements. However, given the current risk profile and current equipment performance, they have minimal safety benefit and, therefore, are not cost effective. No additional action required.

39

• The mention of good practices did not require a response. No action required.

• The other 4 recommendations are now considered part of the ordinary Maintenance Rule activities. No additional action required.

• The 2 suggestions for training or analysis are in error.

• The 2 related to providing information are not related to SAMA.

Therefore, no further action for SAMA is appropriate for the 81 IPE insights.

More recent insights from the updated PRA models were factored directly into the SAMA list. Thirty of the Phase 1 SAMAs include the 'Risk Perspectives on Quad Cities" as the reference source (i.e., indicated in Table F-1 of the ER as Reference 83).

These thirty items were specifically developed following the completion of the 1999 PRA model update. The completion of the 2002 model update did not lead to any additional insights, as the results did not dramatically change. In any event, a correlation between importance parameters for both CDF and LERF from the 2002 (02B) model and their relationship to the SAMA analysis is provided in Response 5(d). In summary, it was judged that these more recent insights were sufficient and appropriate for supplementing the generic SAMA lists with plant-specific insights. Exelon review of the 81 IPE insights in response to this RAI confirms that judgment.

40

RAI6 The SAMA analysis did not include an assessment of SAMAs for external events. The QCNPS IPE for External Events (IPEEE) has shown that the CDF due to internal fire initiated events is about 7x1 0 per reactor year, which is substantially greater than the internal events CDF on which the SAMA evaluation is based. The risk analyses at other commercial nuclear power plants also indicate that external events could be large contributors to CDF and the overall risk to the public. In this regard, the following additional information is needed:

a. NUREG-1742 ("Perspectives Gained From the IPEEE Program," Final Report, 4/02), lists the significant fire area CDFs for QCNPS (pages 3-24 and 3-24 of Volume 2). While these fire-related CDF estimates may be conservative, they are still large relative to the QCNPS internal events CDF. For each fire area or dominant fire sequence, please explain what measures were taken to further reduce risk, and explain why these CDFs can not be further reduced in a cost effective manner.

b. NUREG-1742 lists seismic outliers and improvements for QCNPS (Tables 2.7 and 2.12 of Volume 2). Please confirm that all of the 'Plant improvements' that address the outliers have been implemented. If not, please explain why within the context of this SAMA study.

c. In the IPEEE submittal, Exelon estimated that after the resolution of the seismic outliers, the plant high confidence in low probability of failure (HCLPF) would be at least 0.24g which is less than the 0.3g review level earthquake used in the IPEEE. During the EPU evaluation, the staff noted that if the HCLPF capacity was increased to 0.3g, the resulting CDF would be about an order of magnitude reduction in risk from the IPEEE plant condition. Please identify the systems, structures, and components (SSCs) that limit the plant HCLPF. For those SSCs below 0.3g, justify why modifications to increase seismic capacity would not be cost beneficial when evaluated consistent with the regulatory analysis guidelines.

41

Response 6(a):

INUREG-1742 (Perspectives Gained From the IPEEE Program," Final Report, 4/02),

lists the significant fire area CDFs for QCNPS (pages 3-24 and f3-25] of Volume 2).

While these fire-related CDF estimates may be conservative, they are still large relative to the QCNPS internal events CDF. For each fire area or dominant fire sequence, please explain what measures were taken to further reduce risk, and explain why these CDFs can not be further reduced in a cost effective manner."

As an IPEEE, the QC fire study was performed primarily to develop risk insights. It was done in the traditional style of fire PRAs, and as such, employs conservatism and involves some level of uncertainty (also see Attachment A that provides more details on the types of conservatisms and uncertainties associated with the use of quantitative results from Fire PRAs). Therefore, it cannot be used directly to provide a realistic cost-benefit analysis as part of the SAMA evaluations.

Exelon has, however, used the fire PRA to develop ideas for plant improvement. A large oil fire involving the reactor feedwater pumps was the dominant risk contributor from the IPEEE fire study because the location of combustibles in proximity to cables and circuits associated with RHR Service Water. In response to this insight, the Station performed a modification to improve the response time of the sprinkler heads in the reactor feedwater pump areas. A sensitivity study with the fire PRA shows that a 25%

reduction in fire CDF could be obtained for this modification, alone. Loss of decay heat removal was also identified as important in many fire scenarios. Because of this, another plant enhancement is providing an alternate or redundant air supply for the containment vent valves. Perfect reliability of this redundant air supply had been estimated to reduce the fire CDF by 17%; however, with the sprinkler head modification done, it could be of reduced effectiveness. Nevertheless, the Station is planning to implement a method to provide alternative air supplies in the case of failure of instrument air. Since such a change has not yet been implemented at the site, the idea has been revised to 'retained" status in the Phase I SAMA analysis (see Response 7(b),

Table 7-2, #225), and is now included as Phase II SAMA 17 (see Response 7(c), Table 7-3).

Fourteen other plant modification ideas from the fire IPEEE were analyzed for potential fire CDF reduction [Reference 6-1]. The results from the sensitivity cases fell into three categories. The majority of the cases (9) were shown to have less than 1% benefit in fire CDF risk reduction, and therefore the potential improvement was not pursued. In the other five cases, a fire CDF reduction estimate was not directly available, but in three of the cases, the potential enhancement was qualitatively determined to have minimal risk benefit, and therefore were not pursued further. The final two potential enhancements were for providing control room or alternate local control station access for select RHR and RCIC valves. These were also not pursued because they would require extensive design engineering and analysis work, and the actual benefit could not be readily measured for the fire CDF. Hence, these were also qualitatively evaluated such that the cost exceeds the potential benefit, and were also not pursued 42

further. Therefore, Exelon believes that all of the potentially worthwhile improvement ideas from the fire IPEEE have been identified. An additional fire-area-by-fire-area search for improvement ideas will not be productive until Fire PRA technology advances to the point that a direct comparison of the Fire CDF results and the internal events CDF results is possible.

REFERENCE

[6-1] ERIN Engineerng and Research, Inc., 'Quad Cities Fire IPEEE Insights and Sensitivities,' ERIN Report No. R1i34-98-04.R08, June 1999.

Response 6(b):

WNUREG-1742 lists seismic outliers and improvements for QCNPS (Tables 2.7 and 2.12 of Volume 2). Please confirm that all of the 'Plant improvements' that address the outliers have been implemented. If not, please explain why within the context of this SAMA study."

As indicated in NUREG-1742, an extensive number of plant improvements or other actions were planned to resolve the USI A-46 outliers. These improvements pertained primarily to enhancing anchorage/support capacity and reducing or eliminating the potential for adverse interactions. Quad Cities recently informed the NRC that all of the outliers have been resolved. Reference letter from Timothy J. Tulon, Quad Cities Nuclear Power Station, Completion of Actions Associated With Supplement No. 1 to Generic Letter 87-02: Verification of Seismic Adequacy of Mechanical and Electrical Equipment in Operating Reactors, Unresolved Safety Issue (USI) A-46 (TAC Nos.

M69476 and M69477), SVP-03-0033, dated February 28, 2003.

Response 6(c):

"in the IPEEE submittal, Exelon estimated that after the resolution of the seismic outliers, the plant high confidence in low probability of failure (HCLPF) would be at least 0.24g which is less than the 0.3g review level earthquake used in the IPEEE. During the EPU evaluation, the staff noted that if the HCLPF capacity was increased to 0.3g, the resulting CDF would be about an order of magnitude reduction in risk from the IPEEE plant condition. Please identify the systems, structures, and components (SSCs) that limit the plant HCLPF. For those SSCs below 0.3g, justify why modifications to increase seismic capacity would not be cost beneficial when evaluated consistent with the regulatory analysis guidelines."

43

In the IPEEE submittal, EGC estimated that, after the resolution of the seismic outliers, the plant high confidence in low probability of failure (HCLPF) would be at least 0.24g.

This is less than the 0.3g review level earthquake used in the IPEEE. During the EPU evaluation, the staff noted that if the HCLPF capacity were increased to 0.3g, the resulting CDF would be about an order of magnitude reduction in risk from the IPEEE condition. This order of magnitude change would be before the seismic outliers identified in the IPEEE were modified.

10 CFR 54 defines the requirements for renewal of operating licenses for nuclear power plants. The Rule is founded on two principles. The first principle of license renewal is that with the possible exception of the detrimental effects of aging on the functionality of certain plant systems, structures, and components in the period of extended operation and possibly a few other issues related to safety only during the period of extended operation, the regulatory process is adequate to ensure that the licensing bases of all currently operating plants provides and maintains an acceptable level of safety so that operation will not be inimical to public health and safety or common defense and security. The second and equally important principle of license renewal holds that the plant-specific licensing basis must be maintained during the renewal term in the same manner and to the same extent as during the original licensing term.

The license renewal rule at 10 CFR 54.30 specifies matters that are not subject to NRC review and that may not be contested in a hearing for license renewal. The intent of the provision in 10 CFR 54.30 is to clarify that safety matters of noncompliance for the current operating term should not be the subject of the renewal application or the subject of a hearing in a renewal proceeding, absent specific Commission direction.

Issues concerning operation during the currently authorized term of operation should be addressed as part of the current license in accordance with the Commission's current regulatory process rather than deferred until a renewal review (which will not occur if the licensee chooses not to renew its operating license). Furthermore, 10 CFR 54.30 is intended to make clear that aging issues discovered during the renewal review for the structures and components that are reviewed in 10 CFR 54.21 (a)(3) or 54.21 (c)(1) and that raise questions about the capability of these structures and components to perform their intended function during the current term of operation must be addressed under the current license. However, an applicant for renewal is not relieved from addressing the issue relevant to the period of extended operation as part of its renewal application.

Section 54.30 does not require a general demonstration of compliance with the current licensing basis (CLB) as a prerequisite for issuing a renewed license. Section 54.30 discusses the applicant's responsibilities for addressing safety matters under its current license, which are not within the scope of the renewal review.

In addition, in NUREG-1 437, Section 5.3.3.1, it states that the 'NRC's earthquake design standards have been conservatively developed to ensure protection of the public health and safety from earthquakes whose magnitudes are well above the most likely earthquake magnitude when considering the collective earthquakes history for specific plant sites in the United States. Therefore, earthquakes exceeding NRC seismic design 44

standards are extremely unlikely. However, in the unlikely event of such an earthquake, there would be substantial damage to older residential structures, commercial structures, and high-hazard facilities such as dams whose seismic design standards are below nuclear seismic design standards. The societal impact due to the non-nuclear losses alone from an earthquake larger than the design basis of a nuclear plant, including property damage, injuries, and fatalities, would be major. The technology for assessing losses from such large earthquakes is a developing one, and there are several ongoing studies of this technology, including work at the United States Geological Survey. Presently there is no agreed-upon method for performing such assessments, although a recent report of the National Academy of Sciences suggests some broad guidelines. The NRC has not developed a method for assessing the societal losses from large earthquakes such that the reactor contribution to accident consequences can be quantitatively compared with the non-nuclear losses. However, as supported by at least one study, the commission expects that the reactor accident contribution to the losses from large beyond design basis earthquakes would be small relative to the non-nuclear losses. While this in itself does not mean the reactor consequences from such an earthquake would be small, the commission concludes that even with potentially high consequences from a beyond design basis earthquake, the extremely low probability of such earthquake yields a small risk from beyond design basis earthquakes at existing nuclear power plants'.

NUREG-1437 goes on to say that '...the commission concludes that the risk from sabotage and beyond design basis earthquakes at existing nuclear power plants is small and additionally, that the risks form other external events, are adequately addressed by a generic consideration of internally initiated severe accidents.'

As stated in the Quad Cities UFSAR, the Design Basis Earthquake is 0.24 g. The above listed question asks for a justification as to why modifications to increase the seismic capacity would not be cost beneficial when evaluated consistent with the regulatory analysis guidelines (0.3g). This seismic capacity is beyond the current licensing basis for Quad Cities.

EGC believes that the analysis contained within the environmental report and the answers to these questions adequately support a staff conclusion that the adverse environmental impacts associated with license renewal are not so great that preserving the option of license renewal for energy planning purposes is unreasonable and that the conclusion of small impacts associated with severe accidents, as outlined in NUREG-1437, is defensible.

45

RAI7 The SAMA analysis did not include an assessment of the impact that PRA uncertainties and external event risk considerations would have on the conclusions of the study.

Some license renewal applicants have opted to double the estimated benefits (for internal events) to accommodate any contributions for other initiators when sound reasons exist to support such a numerical adjustment, and to incorporate additional margin in the SAMA screening criteria to address uncertainties in other parts of the analysis (e.g., an additional factor of two in comparing costs and benefits of each SAMA). At QCNPS, external events (both fire and seismic) are dominant contributors to the total CDF, and are over a factor of 10 greater than internal event contributions. On that basis, please provide the following information to address these concerns:

a. an estimate of the uncertainties associated with the calculated core damage frequency (e.g., the mean and median internal events CDF estimates and the 5e and 95th percentile values of the uncertainty distribution),

b. an assessment of the impact on the Phase 1 screening if risk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated with external events (as applicable), and

c. an assessment of the impact on the Phase 2 evaluation if risk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated with external events (as applicable). Please consider the uncertainties due to both the averted cost-risk and the cost of implementation to determine changes in the net value for these SAMAs.

Response 7(a):

"IProvide] an estimate of the uncertainties associated with the calculated core damage frequency (e.g., the mean and median internal events CDF estimates and the 5U and 95m percentile values of the uncertainty distribution)[.r Revision 02B of the Quad Cities PRA model was utilized as the basis for the SAMA analysis performed in support of the environmental report. This version of the model was not populated with uncertainty distributions for the data input parameters.

Conse9 uently, development of the median internal events CDF estimates and the 5 t and 95 percentile values of the uncertainty distribution are not readily available. (Note that population of the uncertainty distribution parameters is anticipated for a future model revision update) In any event, Table 7-1 provides estimates of internal events Level 1 CDF uncertainty distributions that were obtained for other plants from various sources.

46

Table 7-1 Representative Core Damage Frequency Uncertainty Distributions Plant / Mean 5 Median 95 95" / Error Reference Model Value Percentile Value Percentile Mean Factor Value Ratio Peach 4.5E-6 3.5E-7 1.9E-6 1.3E-5 2.9 6.1 NUREG/CR-4551, Bottom Volume 4, Rev. 1, Part 1 (Table S-la)

Grand 4.1 E-6 1.8E-7 1.1 E-6 1.4E-5 3.4 8.8 NUREG/CR-4551, Gulf Volume 6, Rev. 1, Part 1 (Table S-2)

LaSalle 4.4E-5 2.1 E-6 1.6E-5 1.4E-4 3.2 8.2 NUREG/CR-4832, Volume 2 (RMIEP),

(Table 3.1)

LaSalle 6.64E-6 (1) 2.82E-6 5.20E-6 1.39E-5 2.1 2.2 LS-PSA-014, LaSalle 6.88E-6 (2) Quantification Notebook, Revision 2, June 2003 (Appendix G)

Sequoyah 5.6E-5 1.5E-5 3.9E-5 1.6E-4 2.9 3.3 NUREG/CR-4551, Volume 5, Rev.1, Part 1 (Table S.2)

H.B. 4.5E-5 1.5E-5 3.3E-5 1.1 E-4 2.4 2.7 Docket No. 50/261 Robinson (Response to Request for Additional Information Regarding SAMA Analysis)

V.C. 5.6E-5 1.9E-5 4.4E-5 1.3E-4 2.3 2.6 Docket No. 50/395 Summer (Response to SAMA Request for Additional Information)

(t) Point estimate mean value

° Parametric uncertainty mean value The collective information shown in Table 7-1 indicates that a factor of 3 increase from the calculated point estimate mean internal events CDF with an error factor of 6 is a reasonable estimate to approximate the uncertainty distribution. This correlates to an estimated 95t percentile value of 6.6E-6/yr for the Quad Cities internal events core damage frequency. The 95t percentile value is assumed to represent an upper bound estimate in the uncertainty analysis described in Responses 7(b) and 7(c). Additionally, the assumed error factor of 6 can be used to approximate the median and 5th percentile values as well as is shown below.

47

Quad Cities Approximated Uncertainty Distribution:

95th Percentile: 3 * (Point Estimate Mean) = 6.6E-6/yr Median: 95t1 / EF = 6.6E-6Iyr / 6 = 1.1 E-6/yr 5t Percentile: Median / EF = 1.1 E-6/yr / 6= 1.8E-7/yr Response 7(b):

"(Provide]an assessment of the impact on the Phase 1 screening if risk reduction estimates are increased to account for uncertainties in the risk assessment and the additionalbenefits associatedwith extemal events (as appficable)f.r As indicated in Response 7(a), it is estimated that the 95t percentile value would be approximately a factor of 3 higher than the reported mean CDF value of 2.2E-6. This can be assumed to correspond to an internal events upper bound value of about 6.6E-6.

The Quad Cities Internal Fire risk model was updated in 1999 as part of the revised IPEEE submittal report. The CDF contribution to internal fires was estimated at 6.6E-5/yr for Unit 1 and 7.3E-5/yr for Unit 2. However, plant improvements have occurred since that time as identified in Response 6(a), and the methodology invoked to determine the fire CDF is judged to be somewhat conservative. The seismic portion of the IPEEE program was completed in conjunction with the SQUG program. Quad Cities performed a seismic margins assessment (SMA) following the guidance of NUREG-1407 and EPRI NP-6041. The SMA is a deterministic evaluation that does not calculate risk on a probabilistic basis. No core damage frequency sequences were quantified as part of the seismic risk evaluation. However, an extensive number of plant improvements were identified and these have all been resolved as is noted in Response 6(b).

Consequently, to account for both uncertainties in the risk assessment and the potential additional benefits associated with external events, the Phase I screening was re-performed assuming a factor of almost five increase to the base cost risk for QCNPS to

$500K (compared to the base internal events cost-risk of $103,000 used in the ER).

The screening criteria utilized in Table F-1 of the Quad Cities ER includes the following categories:

1. 1 - Not applicable to the QC design

1. 2 - Similar item is addressed under other proposed SAMAs

1. 3 - Already implemented at QC

1. 4 - No significant safety benefit associated with this SAMA for QC

1. 5 - Cost of implementation clearly greater than the maximum averted cost risk

1. 6 - Retained for Phase II analysis 48

1. 7- Not used

1. 8 - ABWR design issue, not practical For the revised Phase I screening, SAMA items that previously screened by Criteria #1 or #8 were not re-examined. SAMA items that previously screened by Criteria #2 or #3 were only looked at from the potential impact of additional benefits that might be afforded by including external events in the analysis. SAMA items that previously screened by Criteria #4 or #5 were all re-examined, and the previously retained items (i.e., Criteria #6) were still retained and were subject to re-analysis as described in Response 7(c). The results of the revised Phase I screening are included in Table 7-2.

As can be seen, three additional SAMAs are now retained for Phase 11(See Phase I SAMA 20, Phase I SAMA 225, and Phase I SAMA 237).

49

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original DIsposmon Revised Disposition Phase 11i SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 19 Use fire protection system SAMA would reduce the #5 - Cost would be Fire protection is a low head system at The cost is considered to N/A pumps as a backup seal frequency of the RCP more than risk Quad Cites and cannot currentiy be used be greater than the upper Injection and high- seal LOCA and the SBO benefit as a HP Injection source. Given that recirc bound maxdmum averted pressure makeup. CDF. pump seal failure Is a negligible contributor cost risk of $500K. No to Quad Cities risk, no consideration is change to the screening given to modifying the FP system to criteria category.

provide seal cooling. The ability to provide high pressure Injection during an SBO would be beneficial, but the cost of the required modifications would be high.

Installation of new high pressure piping, a high head, high flow pump (as it would also have to support the fire system) and a supporting diesel generator or pump motor Is similar in scope to SAMA 185. The cost is also considered to be simflar ($5 milion to $10 million) and Is greater than the maximum averted cost-risk for Quad Cities.

20 Enhance procedural SAMA would reduce the #3 - Already At Quad Cites, Service Water is Investigate potential 15.

guidance for use of cross- frequency of the loss of implemented at completely cross-tled (between units and benefit from Improving bed component cooling or component cooling water Quad Cities. divisions). Inter-unit RHRSW and DGCW TBCCW performance service water pumps. and service water. cross-ties are available via manual valves based on CDF RRW Revised to: which are normally closed. The TBCCW factor review from

1. 6 - Retain pumps discharge to a common header for Response 5(d).

a given unit, but no Inter-unit cross-tie capability currently exists. The same is true or RBCCW.

Procedural guidance Is adequate.

23 8.a. Additional Service SAMA would conceivably #5 - Cost would be The cost of Implementing this SAMA has The cost Isconsidered to N/A Water Pump reduce common cause more than risk been estimated at approximately $5.9 be greater than the upper dependendes from SW benefit million and is greater than the maxrmum bound maxdmum averted system and thus reduce averted cost-risk for OC. cost risk of $500K. No plant risk through system change to the screening reliability Improvement criteria category, 50

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $50OK)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase il SAMIA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 25 Provide reliable power to SAMA would Increase #4 - No significant Control Room HVAC has reliable power Considering uncertainty N/A control building fans. availability of control room safety benefit souroes. The B HVAC train Is powered by and potential impacts ventilation on a loss of the swing EDG in the event of a loss of from external events does power. offsite power. The A Division is from the not introduce any unit diesel. In addition, Control Room significant changes. No HVAC is not required for successful change to the screening accident mitigation. criteria category, 26 Provide a redundant train SAMA would increase the #5 - Cost would be It has been determined that room cooling The cost is considered to N/A of ventilation. availability of components more than risk is not required for successful operation of be greater than the upper dependent on room benefit RHR and Core Spray at Quad Crites. bound maximum averted cooling. RCIC does not require room cooling given cost risk of $500K. No that It is not run concurrently with Core change to the screening Spray, which is assumed to be true in the criteria category.

PSA model. HPCI, Feedwater, the SSMP, RHRSW, and the EDO rooms require room cooling for success over the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> mission time. The cost of installing a redundant, diverse train of HVAC for a Switchgear Room has been estimated at

$10 million (Reference 19) and for exceeds the maximum averted cost-risk for Quad Cities ($0.1 million). Providing a redundant train of HVAC for HPCI, Feedwater, the SSMP, and RHRSW is similar In scope and Isjudged to cost approximately the same; thus, these changes are also screened.

51

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 29 Create ability to switch SAMA would allow #4 - No significant During a postulated SBO, HPCI and RCIC Considering uncertainty N/A fan power supply to DC In continued operation In an safety benefit can operate for the duration of the event and potential Impacts an SBO event. SBO event. This SAMA which Is limited by DC battery life. Use of from external events does was created for reactor a DC powered fan would Increase the not Introduce any core Isolation cooling drain on the batteries with no impact on significant changes. No system room at the reliability of the HPCI or RCIC systems change to the screening Fitzpatrick Nuclear Power as long as there is no gland seal failure. criteria category.

Plant. For the low probability event of an SBO and gland seal failure the crew Is directed to bypass high temperature room trips.

This would avoid the trip of HPCI and RCIC. Component failures of these systems could also occur, but this Is judged to represent a negligible risk Impact. As such there Is no measurable safety benefit associated with this SAMA. l 32 Provide means for The SSMP requires room #6- Retain SSMP has alternate room cooling via a Still retained.

alternate SSMP room cooling at extended times. manual alignment to FPS. The SAMA cooling This SAMA would allow would be yet a further enhancement.

SSMP operation late in accidents when normal Evaluate the benefit of providing alternate room cooling has failed. SSMP room cooling. These options may Include:

• Controls in the Main Control Room for remote alignment of SW or FPS to SSMP room cooling

- Procedures for opening SSMP room doors and using portable fans for SSMP room cooling 35 Install an independent SAMA would decrease #5 - Cost would be Installation of a new, Independent, The cost is considered to N/A method of suppression the probability of loss of more than risk suppression pool cooling system Is similar be greater than the upper pool cooling. containment heat benefit in scope to installing a new containment bound maximum averted removal. For PWRs, a spray system, which has been estimated cost risk of $500K. No potential similar to cost approximately $5.8 million. This change to the screening enhancement would be to exceeds the maximum averted cost-risk criteria category.

Install an Independent for Quad Cities.

cooling system for sump water.

52

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 36 Develop an enhanced SAMA would provide a #6 - Retain The Fire Protection system can already Still retained. 2 drywelI spray system. redundant source of water provide water to the RHR system at Quad to the containment to Cites; however, no procedures have been control containment developed to use It as a containment pressure, when used in spray source. The containment spray conjunction with function could be further enhanced at containment heat Quad Cities.

removal. __ ii i I 37 Provide dedicated SAMA would provide a #5 - Cost would be Installation of a new, Independent, The cost is considered to N/A existing drywell spray source of water to the more than risk containment spray system has been be greater than the upper system. containment to control benefit. estimated to cost approximately $5.8 bound maximum averted containment pressure, million. This exceeds the maximum cost risk of $500K. No when used In conjunction averted cost-risk for Quad Cities. change to the screening with containment heat criteria category.

removal. This would use an existing spray loop Instead of developing a new spray system.

39 Install a filtered SAMA would provide an #5 - Cost would be Potential to improve both the Level 1 and The cost is considered to N/A containment vent to aiternate decay heat more than risk Level 2 results. be greater than the upper remove decay heat. removal method for non- benefit bound maximum averted ATWS events, with the cost risk of $500K. No released fission products change to the screening being scrubbed. criteria category.

Option 1: Gravel Bed Filter Option 2: Multiple Venturi Scrubber 40 Install a containment vent Assuming that injection Is #5- Cost would be Quad Cities does not have a hard pipe The cost is considered to N/A large enough to remove available, this SAMA more than risk vent of sufficient capacity to mitigate be greater than the upper ATWS decay heat. would provide altemate benefit. ATWS pressurization unless other bound maximum averted decay heat removal In an mitigation steps are successful. The cost cost risk of $BOOK. No ATWS event, of a larger vent Is estimated to be in change to the screening excess of $3 million. This exceeds the criteria category.

maximum averted cost-risk for Quad

_ __ __ __ _ _ _ _ _ _ _ __ __ _ __ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Citie s._ _ _ _ _ _ _ _ _ _ _ _ _

53

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 44 Create a large concrete SAMA would ensure that #5 - Cost would be Core retention devices have been The cost is considered to N/A crucible with heat removal molten core debris more than risk Investigated in previous studies. IDCOR be greater than the upper potential under the escaping from the vessel benefit concluded that 'core retention devices are bound maximum averted basemat to contain would be contained within not effective risk reduction devices for cost risk of $500K. No molten core debris. the crucible. The water degraded core events'. Other evaluations change to the screening cooling mechanism would have shown the worth value for a core criteria category.

cool the molten core, retention device to be on the order of preventing a melt-through $7000 (averted cost-risk) compared to an of the basemat. estimated Implementation cost of over $1 million (Der unit).

45 Create a water-cooled SAMA would contain #5 - Cost would be Core retention devices have been The cost is considered to N/A rubble bed on the molten core debris more than risk Investigated In previous studies. IDCOR be greater than the upper pedestal. dropping on to the benefit concluded that Ocore retention devices are bound maximum averted pedestal and would allow not effective risk reduction devices for cost risk of $500K. No the debris to be cooled. degraded core events'. Other evaluations change to the screening have shown the worth value for a core criteria category.

retention device to be on the order of

$7000 (averted cost-risk) compared to an estimated Implementation cost of over $1 million (per unit).

46 Provide modification for SAMA would help #4 - No significant BWR Mark I risk Is typically dominated by Considering uncertainty N/A flooding the drywell head. mitigate accidents that safety benefit events that result in early failure of the and potential Impacts result in the leakage dryweil shell due to direct contact with from external events does through the drywell head core debris and events that bypass the not Introduce any seal. containment. This is also true at Quad significant changes. No Cities. The head flooding system would, change to the screening therefore, not be expected to have any criteria category.

significant impact on the overall risk.

The potential for competing risks due to Reactor Building flooding is considered to eliminate any positive safety benefit.

54

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA tMe Result of potential Original I Revised Original Disposition Revised DlsposMon Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 47 Enhance fire protection SAMA would Improve #4 - No significant Current Standby Gas Treatment Systems Considering uncertainty N/A system and/or standby fission product scrubbing safety benefit do not have sufficient capacity to handle and potential Impacts gas treatment system in severe accidents. the loads from severe accidents that result from external events does hardware and In a bypass or breach of the containment. not Introduce any procedures. Loads produced as a result of RPV or significant changes. No containment blowdown would require large change to the screening filtering capacities. These filtered vented criteria category.

systems have been previously ivnstigated and found not to provide sufficient cost benefit Quad Cities has limited fire protection sprinkler systems In the Reactor Building.

Use of these for fission product scrubbing In the R.B. could create competing risks associated with spray failures and flooding of equipment with very limited potential

_________ benefit. I 51 Create a core melt source SAMA would provide #5 - Cost would be Core retention devices have been The cost is considered to N/A reduction system. cooling and containment more than risk Investigated In previous studies. IDCOR be greater than the upper of molten core debris. benefit concluded that core retention devices are bound maxdmum averted Refractory material would not effective risk reduction devices for cost risk of $500K. No be placed underneath the degraded core events'. Other evaluations change to the screening reactor vessel such that a have shown the worth value for a core criteria category.

molten core falling on the retention device to be on the order of material would melt and $7000 compared to an estimated combine with the materiai. implementation cost of over $1 miflion.

Subsequent spreading and heat removal from the vitrified compound would be facilitated and concrete attack would not l ________ _______________________ occu r 55

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $SOCK)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 54 Install a secondary SAMA would filter fission #5 - Cost would be Secondary containment at Quad Cities The cost Is considered to N/A containment filtered vent. products released from more than risk makes extensive use of blow out panels to be greater then the upper primary containment, benefit protect the structural Integrity of the bound maximum averted building In the event of Internal pressure cost risk of S500K. No challenges such as steam line breaks In change to the screening the reactor building or external pressure criteria category.

challenges such as tornadoes. Major structural redesign of the reactor building would be required to make the reactor building capable of retaining and processing a primary containment failure.

55 Install a passive SAMA would provide #5 - Cost would be See SAMAs 36 and 53. A passive system The cost Is considered to N/A containment spray redundant containment more than risk is another alternative enhancement for the be greater than the upper system. spray method without benefit. Containment Spray function. See #36. bound maximum averted high cost cost risk of $500K. No change to the screening

_______________________________ criteria category.

56 Strengthen SAMA would reduce the #5 -Cost would be Reference 17 discusses the cost of The cost Is considered to N/A primary/secondary probability of containment more than risk increasing the containment pressure and be greater than the upper containment. overpressurization to benefit temperature capacity, which is effectively bound maximum averted failure. strengthening the containment. This cost cost risk of $500K. No Isestimated assuming the change Is made change to the screening during the design phase whereas for Quad criteria category.

Cites, the changes would have to be made as a retrofit The cost estimated for the ABWR was $12 millIon and it is judged that retrofitting an existing containment would cost more. The cost of implementation for this SAMA exceeds the maximum averted cost-risk for Quad l_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ __ _ _ __ _ __ _ Cities.

57 Increase the depth of the SAMA would prevent #5 - Cost would be Core retention devices have been The cost is considered to N/A concrete basemat or use basemat melt-through. more than risk Investigated in previous studies. IDCOR be greater than the upper an alternative concrete benefit concluded that core retention devices are bound maximum averted material to ensure melt- not effective risk reduction devices for cost risk of $500K. No through does not occur. degraded core events'. Other evaluations change to the screening have shown the worth value for a core criteria category.

retention device to be on the order of

$7000 compared to an estimated I _______ ____________________ __ _________________ _____ __________ implement i ncost ofover m~im /on dem entao cost of over $1 m illion/site.

56

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase 11 SAMA IDl enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 58 Provide a reactor vessel SAMA would provide the #5 - Cost would be This has been estimated to cost $2.5 The cost Is considered to N/A exterior cooling system. potential to cool a molten more than risk million and exceeds the maximum averted be greater than the upper core before it causes benefit cost-risk for Quad Cities defined In Section bound maximum averted vessel failure, if the lower F.4.7. ORNL [871 has performed thermal cost risk of $500K. No head could be submerged hydraulic calculations on BWR external change to the screening in water. cooling methods and determined that the criteria category.

current BWR RPV support skirt design makes It Impractial to cool the RPV by external cooling to prevent RPV breach.

Therefore, the modification would require RPV support skirt modification and reanalysis to allow the external cooling to be effective.

59 Construct a building to be SAMA would provide a #5 - Cost would be Based on engineering judgement, the cost The cost Isconsidered to N/A connected to method to depressurize more than risk of this enhancement is expected to greatly be greater than the upper primary/secondary containment and reduce benefit exceed the maximum averted cost risk for bound maximum averted containment that Is fission product release. Quad Cities. cost risk of $500K No maintained at a vacuum. change to the screening criteria catenory.

65 1.h. Simulator Training for SAMA would lead to #4 - No significant Simulators could be upgraded and used to Considering uncertainty N/A Severe Accident improved arrest of core safety benefit provide operator training for severe and potential Impacts melt progress and accidents; however, these scenarios are from external events does prevention of containment Previously rare and the Instruction time would not introduce any failure assessed by the compete with time required to train significant changes. No NRC as not operators on more likely scenarios that are change to the screening required to support severe accident precursors. The benefit of criteria category.

Accident simulator training Is difficult to quantify as management the results would be based on the because of Improved reliability of human actions In the marginal cost mitigation of severe accidents. Training benefit. can positively Influence the values of HEPs, but the Impact Issmall. In addition, the TSC wouid be manned In a severe accident evolution and could provide additional support by personnel familiar I _ __ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ ____________ with the SAMFs.

w_ e S A M G s._ _ _ __ _ _ _ _ht l 57

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 67 3.a. Larger Volume SAMA increases time #5 - Cost would be Enlargement of the containment would be The cost is considered to N/A Containment before containment failure more than risk similar in scope to the ABWR design be greater than the upper and Increases time for benefit change SAMA to Implement a larger bound maximum averted recovery volume containment but would likely cost risk of $500K. No exceed the $8 million estimate for that change to the screening change as a retrofit would be required. criteria category.

This Is greater than the maximum averted cost-risk defined In F.4.7.

69 3.c. Improved Vacuum SAMA reduces the #5- Cost would be The Quad Cities plant has twelve (12) Considering uncertainty N/A Breakers (redundant probability of a stuck open more than risk Indhvdual vacuum breaker lines with a and potential Impacts valves In each line) vacuum breaker. benefit single vacuum breaker In each line. from external events does Providing redundant vacuum breakers In not Introduce any each line would decrease the potential for changes to the original vapor suppression failure and suppression disposition (Vapor pool bypass. This plant modification suppression failures are requires new valves, the structural not significant contributors changes to implement the modification, to external events). No and the outage time to Install. Based on change to the screening the PRA results that vapor suppression criteria category.

failure and pool bypass are negligible risk contributors and the apparent extremely high cost, this proposed SAMA Is not I _______ ____________________ considered cost effective. _iderecoste_____e 94 Use fuel cells Instead of SAMA would extend DC #6- Retain Improving battery capacity may be cost Still retained. HS lead-acid batteries. power availability in an beneficial for Quad Cities. Further SBO. extension of battery life with fuel cells Is estimated to have a small impact on the I_______ ___________________ _______________ Quad Cities residual risk profile. _sresidual___skp _f__e_

58

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 96 Improve 4.16-kV bus Enhance procedures to #6 - Retain Manual cross-tie between AC buses is Still retained.

cross-tle ability. direct 4kV bus cross-tie. proceduralized for certain buses If this procedural step depending on the available AC source already exists, Investigate (e.g., offsite power, SBO DIG). These installation of hardware cross-ties are effective and further risk that would perform an reduction from auto cross-toe Is of marginal automatic cross-iVe to the benefit, and could produce competing opposite 4kV bus given risks.

failure of the dedicated diesel. Automatic cross-tie could be implemented at Quad Cities. In addition, procedures could be developed that would allow the following cross-ties to be performed:

-Bus 14-1 to Bus 24-1 from EDG I

-Bus 24-1 to Bus 14-1 from EDO 2

-EDG 1/2 to Buses 13-1 and 23-1 - VX  :

107 Install gas turbine SAMA would Improve #5 - Cost would be The cost of installing a diverse, redundant, The cost is considered to N/A generator. onsite AC power reliability more than risk gas turbine generator Is similar In scope to be greater than the upper by providing a redundant benefit Installing a new diesel generator. The cost bound maximum averted and diverse emergency of Installing an additional diesel generator cost risk of $500K. No power system. has been estimated at over $20 million In change to the screening Reference 19. This cost of criteria category.

implementaMon for this SAMA greatly exceeds the maximum averted cost-risk for Quad Cities defined In Section F.4.7.

In addition, Quad Cities already has five diverse on-site AC power sources.

Installing a gas turbine would provide minimal safety benefit 108 Create a backup source This SAMA would provide #6 - Retain An additional EDG cooling source may be Still retained =

for diesel cooling. (Not a redundant and diverse cost beneficial for Ouad Cities. This load from existing system) source of cooling for the path also Includes ECCS room cooling.

diesel generators, which would contribute to enhanced diesel l _____ ___ __ ____ ___ ___ ___ ___ ___ _ reilablil t . ____ ___ __ __A V__

59

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 110 Provide a connection to SAMA would reduce the #5 - Cost would be Offsite power lines would be exposed to The cost Is considered to N/A an alternate source of probability of a loss of more than risk severe weather at some point along the be greater than the upper offsite power. offsIte power event. benefit offsite power line route. While the actual bound maximum averted cost of this SAMA will vary depending on cost risk of $500K. No site characteristics, the cost of connecting change to the screening to an alternate source of power has been criteria category.

estimated at >$25 million for another commercial US nuclear plant.

Implementing this SAMA at Quad Cities Is considered to be within the same order of magnitude and exceeds the maximum averted cost-risk for Quad Cities as defined In Section F.4.7. In addition, Quad Cities has multiple offsle sources and multiple, diverse on-site AC power sources. Providing additional AC power sources would provide minimal safety I_ _ _ _ _ _ _ benefit 111 Bury offsite power lines. SAMA could improve #5 - Cost would be While the actual cost of this SAMA will The cost is considered to N/A offsite power reliability, more than risk vary depending on site characteristics the be greater than the upper particularly during severe benefit cost of burying offste power lines has bound maximum averted weather. been estimated at a cost significantly cost risk of $500K. No greater than $25 million for another change to the screening commercial US nuclear plant. criteria category.

Implementing this SAMA at Quad Cities is considered to be within the same order of magnitude and exceeds the maximum averted cost-risk for Quad Cities as I _______ _______________ defined in defined In Section F.4.7. l 114 Provide DC power to the SAMA would increase the #4 - No significant 1) Loss of 120V AC Is not an Initiating Considering uncertainty N/A 120/240-V vital AC reliability of the 120-VAC safety benefit Event and potential Impacts system from the Class 1E Bus. 2) 120 VAC Is not a risk significant from external events does station service battery support system not introduce any system Instead of ts own significant changes. No battery. change to the screening I _ _ _ __ _ _ ___________________ ___________________ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ criteria cate gory.

60

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMAtMe Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 121 9.f. Improved SAMA would provide #4 - No significant 1) Loss of 120V AC Is not an Initiating Considering uncertainty /A Uninterruptable Power increased reliability of safety benefit Event and potential Impacts Supplies power supplies supporting 2) 120 VAC Is not a risk significant from external events does front-line equipment, thus support system not Introduce any reducing core damage significant changes. No and release frequencies. change to the screening criteria category.

125 10.a. Dedicated DC This SAMA addresses the #5 -Cost would be The cost of Implementation for this mod Is The cost is considered to N/A Power Supply use of a diverse DC more than risk estimated at $3 million, which Is greater be greater than the upper power system such as an benefit than the maximum averted cost-risk for bound maximum averted additional battery or fuel Quad Cities as defined in Section F.4.7. cost risk of $500K. No cell for the purpose of See also SAMAs 93, 94, 97, 98, 99, and change to the screening providing motive power to 100. criteria category.

certain components (e.g.,

RCIC).

130 Add an automatic bus Plants are typically #4 - No significant 1) Loss of 120V AC Is not an Initiating Considering uncertainty /A transfer feature to allow sensitive to the loss of safety benefit Event and potential impacts the automatic transfer of one or more 120V vital 2) 120 VAC is not a risk significant from external events does the 120V vital AC bus AC buses. Manual support system not introduce any from the on-line unit to the transfers to aitemate significant changes. No standby unit power supplies could be change to the screening enhanced to transfer criteria category.

automatically.

131 Provide procedures for (a) This SAMA would allow #6 - Retain While DC buses are reliable, procedure Still retained.

bypassing major DC for powering specific changes may be cost beneficial given the buses; (b) locally starting loads given a DC bus importance of DC power.

equipment failure and/or the ability to start equipment locally that normally requires DC power for a control room start.

132 Provide procedures to This would provide #5 - Cost would be A procedure change may be a cost Additionally, the dominant N/A allow cross-tie of the 1/2 additional diversity Inthe more than risk beneficial enhancement for Quad Cities. failure mechanisms for EDG to a bus which can SSMPs power supply. benefit However, the ability to cross-tie among the SSMP do not Involve supply the SSMP (14-1, divisions has so many competing risks and electrical or electrical 24-1, or 31) requires hardware changes that make this support failures. As such, SAMA unacceptable given the low implementation of such a maximum averted for Quad C~ies. procedure would have minimal Impact on the CDF results. No change to the screening criteria category.

61

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 141 Locate residual heat SAMA would prevent #5 - Cost would be Competing risks associated with such a The cost Isconsidered to N/A removal (RHR) inside of Intersystem LOCA more than risk design are manifold and would require be greater than the upper containment. (ISLOCA) out the RHR benefit extensive analysis to demonstrate bound maximum averted pathway. capability. For an existing plart, the cost cost risk of $500K No of moving an entire system Is Judged to change to the screening greatly exceed the maximum averted cost- criteria category.

risk for Quad Cites as defined in Section F.4.7.

142 Install addltional SAMA would decrease #4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty N/A Instrumentation for ISLOCA frequency by safety benefit IN-92-36 and Its additional supplement, and potential impacts ISLOCAs. Instafling leak monitoring ISLOCA contributes Hile risk for BWRs. from external events does Instruments in between For Quad Cities, ISLOCA and Large Break not Introduce any the first two pressure Outside Containment have CDF based changes to the original isolation valves on Iow- Risk Reduction Worth values of 1.005 and disposition (ISLOCAs are pressure Inject lines and 1.000, respectively. ISLOCA sequences not significant contributors RHR suction Ones. comprise less than 1%of the LERF at to external events). No Quad Cities. change to the screening criteria category.

143 Increase frequency for SAMA could reduce #4 - No significant The PIV Interface valves at Quad Cities Considering uncertainty N/A valve leak testing. ISLOCA frequency. safety benefit are leak tested. Related to mitigation of and potential Impacts an ISLOCA. Per IN-92-36 and its from external events does additional supplement, ISLOCA not Introduce any contributes little risk for BWRs. For Quad changes to the original Cities, ISLOCA and Large Break Outside disposition (ISLOCAs are Containment have CDF based Risk not significant contributors Reduction Worth values of 1.005 and to external events). No 1.000, respectively. ISLOCA sequences change to the screening comprise less than 1%of the LERF at criteria category.

Quad Cities. Competing Risk: Valve leak testing may actually increase risk because

_ ___ ______ ______ __ ___ ______ _____ on-line valve m anipulation is required._ _ _ _ _ _ _ _ _ __ _ _ _ _

62

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 144 Improve operator training SAMA would decrease #4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty N/A on ISLOCA coping. ISLOCA effects. safety benefit IN-92.36 and its additional supplement, and potential impacts ISLOCA contributes Aittle risk for BWRs. from external events does For Quad Cities, ISLOCA and Large Break not introduce any Outside Containment have CDF based changes to the original Risk Reduction Worth values of 1.005 and disposition (ISLOCAs are 1.000, respectively. ISLOCA sequences not significant contributors comprise less than 1%of the LERF at to external events). No Quad Cities. change to the screening criteria category.

In addition, the Quad Cities EOPs provide secondary containment monitoring parameters which Include room specific temperature, room specific radiation, vent radiation, and room specific water level.

The instrumentation and procedural guidance help locate and isolate breaks which have bypassed primary l _________ ________________________ __________________ containment.

146 Provide leak testing of SAMA would help reduce #4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty N/A valves In ISLOCA paths. ISLOCA frequency. At safety benefit IN-92-36 and Its additional supplement and potential Impacts Kewaunee Nuclear Power ISLOCA contributes litrie risk for BWRs. from external events does Plant, four MOVs isolating For Quad Cities, ISLOCA and Large Break not Introduce any RHR from the RCS were Outside Containment have CDF based changes to the original not leak tested. Risk Reduction Worth values of 1.005 and disposition (ISLOCAs are 1.000, respectively. ISLOCA sequences not significant contributors comprise less than 1%of the LERF at to external events). No Quad Cities. Competing Risk: Valve leak change to the screening testing may actually Increase risk because criteria category.

____________________ _______________ on-line valve manipulation is required.alvemanipulationIsrequired.

63

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition IIPhase SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 148 Ensure all ISLOCA SAMA would scrub all #4- No significant ISLOCA and Large Break Outside Considering uncertainty NIA releases are scrubbed. ISLOCA releases. One safety benefit Containment have CDF based Risk and potential impacts example Is to plug drains Reduction Worth values of 1.005 and from external events does In the break area so that 1.000, respectively. ISLOCA sequences not introduce any the break point would be comprise less than 1%of the LERF at changes to the original covered with water. Quad CiOfes. The cost of performing the disposition (ISLOCAs are analysis to identify all ISLOCA pathways not significant contributors and to ensure that any physical to external events). No modifications implemented to mitigate change to the screening ISLOCAs are not detrimental to the plant criteria category.

(e.g., cause flooding hazards) combined with the cost of Installing the required equipment Is judged to greatly exceed any benefit. Additionally, the suggested enhancement of plugging drain lines would not guarantee a release would be scrubbed as the release may occur prior to the submergence of the break. Room flooding equipment and waterproofing of mitigative components would be required to make this SAMA potentially effective.

Such changes would be extremely costly and potential competing risk appears to significantly outweigh any possible safety benefit_

149 Add redundant and SAMA could reduce the #4 - No significant Related to mitigation of an ISLOCA. Per Considering uncertainty N/A diverse limit switches to frequency of containment safety benefit IN-92-36 and Its additional supplement, and potential Impacts each containment Isolation failure and ISLOCA contributes little risk for BWRs. from external events does Isolation valve. ISLOCAs through For Quad Cities, ISLOCA and Large Break not Introduce any enhanced Isolation valve Outside Containment have CDF based changes to the original position Indication. Risk Reduction Worth values of 1.005 and disposition. No change to 1.000, respectively. ISLOCA sequences the screening criteria comprise less than 1%of the LERF at category.

Quad Cities.

64

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original DisposItIon Revised DispositIon Phase 11 l

SAMA ID en h ancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 151 8.e. Improved MSIV This SAMA would #4 - No significant There Is no evidence of poor MSIV Considering uncertainty NA Design decrease the likelihood of safety benefit performance. Redundant MSIVs are and potential Impacts containment bypass designed to isolate on severe accidents from external events does scenarios. that could lead to radlonuclide release and not Introduce any bypass containment. These include changes to the original breaks outside containment. The MSIVs disposition. No change to are leak tested to ensure their adequacy. the screening criteria The Maintenance Rule program monitors category.

the perfomiances of the MSIVs providing early feedback on any degradation.

The PRA has determined that the risk contribution from MSIV failures to isolate Is very small.

156 Modify swing direction of SAMA would prevent #4 - No significant Quad Cites plant Is not susceptible to Considering uncertainty N/A doors separating turbine flood propagation, for a safety benefit flood propagation from the turbine building and potential Impacts building basement from plant where Internal to adjacent buildings with safety from external events does areas containing flooding from turbine equipment Flooding from Turbine Hall not Introduce any safeguards equipment. building to safeguards Into adjacent buildings considered to have changes to the original areas Is a concern. negligible Impact Electrical Equipment disposition. No change to (MCCs, diesel generators, batteries, the screening criteria SSMP) are located at the 595' El. or category.

above. There are Turbine Building access

'rall-up doors at the 595' El. Flooding Is not expected to reach the 595' El.; if it does, then discharge to the outside should preclude any further rise.

65

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase ii SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 158 Implement internal flood This SAMA would reduce #5 - Cost would be The Quad Cities Internal Flooding Analysis Considering uncertainty N/A prevention and mitigation the consequences of more than risk states that there do not appear to be any and potential Impacts enhancements. Internal flooding. benefit. flood specific response procedures for from external events does catastrophic flood events. The existing not introduce any procedures appear to be completely changes to the original adequate for small leaks; however, they disposition. No change to are judged not to provide specific the screening criteria directions to respond to large flow rate category.

breaks. As a result, relatively high failure probabilities are estimated for the mitigative actions required to prevent extensive damage. Internal flood enhancements would Include:

- Curbs around the comer room stairwells to the RHR compartments

- Coping procedures for SW floods In the Reactor Building For example, a specific pipe break scenario has been postulated that would disable 4kV buses 13 and 14. Given the consequential failure of Unit 1 TBCCW ,

several compensatory options exist The internal flood evaluation In the IPE calculated a CDF that would be less than 10% of the current Quad Cities CDF. This translates Into approximately $10,000 as the maximum cost that can be shown to be cost beneficial. No procedures or plant modification Isjudged to be possible for this cost and therefore this SAMA Is found I __ ____ ___________ _____ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ ____________ not to be b cost c s beneficial.

b ei a _ _ _ _ _ _ _ _ _ _ _ _ __nt 66

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 162 Review Circulating Water This is a Quad Cities #4 - No significant Risk contribution is so low due to this Considering uncertainty N/A Pump Auto Trip specific SAMA that is safety benefit postulated scenario that cost cannot be and potential impacts procedure to determine Its related to the procedural justified. from external events does applicability to a direction to start the not introduce any condenser pit flooding standby Circulating Water changes to the original scenario pump on tip of the Initially disposition. No change to running pump given high the screening criteria Condenser Pit level. Use category.

of the current procedure may exacerbate the flooding and result In an overflow Into the Turbine Basement (which contains the condensate pumps and RHRSW vaults).

163 Consider dual unit flood The current Quad Cities #4 - No significant Quad Cities flood Induced risk is quite low Considering uncertainty N/A effects In the EOPs EOPs (QGAs) do not safety benefit and that due to any dual unit issues and potential Impacts consider the impact of a negligible. Changes cannot be from external events does flooding event In the implemented on a cost beneficial basis. not Introduce any opposite unit on the changes to the original equipment of the given disposition. No change to unit. A flood In certain the screeing criteria compartments of one unit category.

will result In a challenge to equipment In the opposite unit due to plant configuration. Updating the QGAs to account for the potential loss of equipment given a flood In the opposite unit will allow the operators to prepare for a scram and plan for the use of appropriate alternative systems. l 67

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 164 Examine the potential for The RHRSW vaults at #5 - Cost would be The internal flood probabilistic analysis Considering uncertainty N/A RHRSW vault failure and Quad Cities contain more than risk includes the quantification of the RHRSW and potential impacts consequential Turbine piping from the discharge benefit. pipe breaks and the resulting from external events does Basement flooding from one or more other quantification shows that the subject not introduce any RHRSW pumps. A break insight has a negligible Impact on plant changes to the original in the piping not co- risk. The estimated cost of structural disposition. No change to located with the pump will analysis, structural changes, Instrument the screening criteria flood the RHRSW vault changes, or procedure changes would not category.

and result In an Internal be cost justified, i.e., would be far in pressure build up. The excess of the total internal flood risk potential exists for the contribution >>$1 0,000.

vault to collapse and result In Turbine Basement flooding.

Resolution of this SAMA would decrease the contribution of internal flooding In this area.

170 Install a new condensate Either replace the existing #5- Cost would be Installation of an additional CST may be a The cost Isconsidered to N/A storage tank (CST) tank with a larger one, or more than risk cost beneficial means of reducing risk at be greater than the upper Install a back-up tank. benefit Quad Cities. The availability of bound maximum averted significantly larger CST volume could be cost risk of $500K. No used by LPCI or CS to provide continuous change to the screening RPV Injection regardless of torus criteria category.

conditions.

178 Install an Independent This SAMA would allow #4- No significant HPCI and RCIC are the turbine driven Considering uncertainty N/A diesel generator for the continued Inventory safety benefit Injection systems for Quad Cities. The and potential Impacts CST make-up pumps make-up to the CST CCSTs each have a nominal water supply from external events does during an SBO. of 260,000 gallons and the reserved not Introduce any volume (only accessible by SSMP, HPCI, significant changes. No and RCIC) Is 90,000 gallons. Given a change to the screening battery life of 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> (required for criteria category.

HPCI/RCIC operation) and an Initial volume of 90,000 gallons, no additional water source would be required for injection during the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> SBO mission time. Minimal benefit would be gained from this SAMA.

Similar Item Is addressed under proposed SAMA #60. . -

68

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening CrIteria Including Uncertainty and SAMA ID number External Events number 191 Upgrade Chemical and For a plant like the AP600 #5 - Cost would be A potential functional equivalent for Quad The cost Is considered to N/A Volume Control System to where the Chemical and more than risk Cities would be the enhancement of the be greater than the upper mitigate small LOCAs. Volume Control System benefit RWCU system such that injection flow bound maxdmum averted cannot mitigate a Small rates on the order of 1000 gpm were cost risk of $500K. No LOCA, an upgrade would possible. This change Is considered to be change to the screening decrease the Small LOCA similar In function, scope, and cost to criteria category.

CDF contribution. SAMA 185 ($5-10 million) with the exception of the independent power source. However, new power circuits and wiring would likely be needed for the larger pumps. The low end of the cost of implementation estimate ($5 million) Is judged to be applicable for this SAMA, which Is greater than the maximum averted cost risk for Quad Cities as defined In Section F.4.7. l 194 Replace 2 of the 4 safety This SAMA would reduce # - No significant Quad Cities has a diverse set of Injection Installation of N/A injection (SI) pumps with the SI system common safety benefit systems and more than one method of Independent RHR diesel-powered pumps. cause failure probability. containment heat removal. Common RHRSW pumps that This SAMA was intended cause failure of the 4 train RHR system Is could provide an alternate for the System 80+, which Revised to: a low contributor to risk and removing the means of containment has four trains of SI. #5 - Cost would be 4/4 system failures would have minimal heat removal would be more than risk Impact on the results. The CCF of all four beneficial to reduce the benefit RHR pumps to run (1RHPMlABCD- Fire CDF that Is largely XCC) has a Risk Reduction Worth of dominated by loss of 1.000 (with respect to CDF). The CCF of decay heat removal all four RHR pumps to fali to start scenarios. However, the (1RHPM1ABCD-ACC) does not appear cost to implement such a In any CDF cutsets above the truncation system Is considered to limit for the plant model and would not be greater than the upper impact the results if it were improved. bound maximum averted cost risk of $500K. l 69

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential OrIginal I Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAIVIA ID number External Events number 196 Raise high pressure core This SAMA would ensure #4 - No significant The HPCI high backpressure trip Is Considering uncertainty N/A injectionlreactor core high pressure core safety benefit already set at a pressure above the and potential Impacts isolation cooling Injection/reactor core containment ultimate pressure; thus, from external events does backpressure trip isolation cooling raising the trip limit would have very not Introduce any setpoints avaliability when high limited impact. The RCIC tip limit could significant changes. No suppression pool be increased or bypassed, but the benefit change to the screening temperatures exist. would also be small because RPV criteria category.

depressurization is required before containment conditions are above these back pressure set points. Therefore, no benefit Isgained from Increasing these numerical values.

197 Improve the reliablity of This SAMA would reduce #5 - Cost would be High pressure melt scenarios are Considering uncertainty /A the automatic the frequency of high more than risk significant contributors to the Quad CAties and potential Impacts depressunization system. pressure core damage benefit CDF. The SAMA is interpreted to mean from external events does sequences. improved reliability of the ERVs and not introduce any Target Rock SRVs and their support significant changes. No systems. A plant modification to eliminate change to the screening dependence on DC power to Increase the criteria category.

success probability of these valves would reduce the high pressure Injection accident classes of IA and IE.

No such design is currently available. This would require a research and development prolect.

201 Increase available net SAMA increases the #5 - Cost would be Requires major plant changes such as The cost is considered to N/A positive suction head probability that these more than risk new RHR pumps, moving the RHR be greater than the upper (NPSH) for injection pumps will be available to benefit pumps, a new suppression pool design, a bound maximum averted pumps. Inject coolant into the larger CCST (only applicable for injection cost risk of $50OK. No vessel by increasing the phase), or an additional containment change to the screening available NPSH for the cooling system. The cost of these changes criteria category.

Injection pumps. would exceed the maximum averted cost-risk for Quad Cities as defined In Section F.4.7.

70

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original I Revised Original Disposition Revised Disposition Phase lI SAMA ID enhancement Screening Criteria Including Uncertalnty and SAMA ID number External Events number 202 Modify Reactor Water SAMA would provide an #5 - Cost would be In order to make RWCU a viable heat The cost is considered to N/A Cleanup (RWCU) for use additional source of decay more than risk removal system, the piping, pumps, heat be greater than the upper as a decay heat removal heat removal. benefit exchangers, and power sources would bound maximum averted system and proceduralize have to be upgraded. This SAMA is cost risk of $500K. No use. considered to be similar In scope to SAMA change to the screening 191. The cost of Implementation for such criteria category.

a change (approximately $5 million) Is greater than the maximum averted cost-risk for Quad Cities.

208 2.a. Passive High SAMA will Improve #5 -Cost would be The cost of this enhancement has been The cost is considered to N/A Pressure System prevention of core melt more than risk estimated to be $1.7 million In Reference be greater than the upper sequences by providing benefit 17. This Is greater than the maximum bound maximum averted additional high pressure averted cost-rdsk for Quad Cities as cost risk of $500K. No capability to remove defined In Section F.4.7. change to the screening decay heat through an criteria category.

Isolation condenser type system_ _

209 2.c. Suppression Pool SAMA will Improve #5 - Cost would be From a review of the contributors to the Loss of all low pressure N/A Jockey Pump prevention of core melt more than risk Quad Cities risk profile it Is found that the Injection Is also not a sequences by providing a benefit availability of low pressure pumps for RPV dominant contributor to small makeup pump to make up Is not a dominant contributor. the external events provide low pressure The low pressure pump availability for analysis. As such, decay heat removal from RPV injection is a negligible contributor to considering uncertainty the RPV using the the risk profile. The expense of adding and potential impacts suppression pool as a another low pressure Injection system from external events does source of water. without Introducing severe competing risks not Introduce any Is expected to be high. It can be significant changes. No concluded tat the cost will not be able to change to the screening be lustifled. criteria category.

214 4.c. High Flow SAMA would Improve #5 - Cost would be Increasing the capabilities of suppression The cost Isconsidered to N/A Suppression Pool Cooling suppression pool cooling more than risk pool would require new pumps, heat be greater than the upper for ATWS response. benefit exchangers, piping, and other equipment. bound maximum averted The Implementation cost of this change Is cost risk of $500K No considered to be approximately equivalent change to the screening to SAMA 35 ($5.8 million) and Is screened criteria category.

from further review as It Is significantly greater than the maximum averted cost-risk for Quad Cities as defined In Section F.4.7. .

71

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 216 Delete High DW Pressure This SAMA would allow #6 - Retain SDC could be used for DHR In conditions Still retained. 7I Signal from SDC isolation the Initiation of SDC when where it is currently precluded from use.

the drywell is at elevated Removal of this logic is not a cost pressures. beneficial modification but would be a safety enhancement if justified on other bases.

217 Use SSMP to provide The SSMP provides #4 - No significant This SAMA only applies to dual unit Considering uncertainty N/A injection to Unit 1and Injection to one unit at a safety benefit initiators. For single unit Initiators, SSMP and potential impacts Unit 2 simultaneously time. Injection to both can be dedicated to the shutdown unit. from external events does units simultaneously not Introduce any could be beneficial in The SSMP flow rate is sufficient to support significant changes. No cases where only SSMP a single unit for adequate core cooling if it change to the screening Injection Is available. This is the sole injection source and the event criteria category.

would eliminate the need resembles an MSIV closure from full to alternate Injection power. In that case, sharing of SSMP Is between the units. not an effective option.

For other less severe cases (e.g., reduced power operation, other Injection sources available), the SSMP Is suffident to refill the RPV to Level 8. Therefore, the number of SSMP cycles to alternate between units is relatively low, i.e.,

approximately ten over the 24-hour mission time. The SSMP can be easily switched from one unit to the other through the manipulation of two MOVe. In addition to the MOVs, there are four check valves that also need to open per 'cycle.'

This results in a small change In SSMP failure probability of 6.4E124 (12% of the SSMP unavailability not counting the support systems) and a negligible change to the Quad Cmites risk profile.

(1) Consistent with the assessment of subsequent MOV and check valve movements the failure probability Is set at a factor of ten lower than the initial failure probability on a per demand (cycle) basis.

72

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA thiseResult of potential Original I Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 218 Install a high level SSMP This would help prevent #5 - Cost would be The Impact of this SAMA Is very low. Considering unicertainty N/A pump trip to avoid water inadvertent more than risk Water solid over-pressurization Iscurrently and potential impacts solid operation of the overpressurization of the benefit modeled In the PSA to be a negligible from external events does RPV. RPV. contributor to risk. not Introduce any significant changes. No change to the screening I _______ criteria catecory.

I_____________

219 Develop procedures to This SAMA increases the #6 - Retain Evaluate the benefit of improved Still retained. 8 control Feedwater flow functionality of Feedwater Feedwater level control given loss of DC.

without 125 VDC power to In loss of DC scenarios -::--

prevent tripping and Increases the Feedwater on High/Low probability of successful level level control.

220 Remove Loop Select In the event that there is #6 - Retain Evaluate the benefit removal or bypass of Still retained. 9 Logic no break In the recirc LPCI Loop Select Logic.

loops and there Isa Loop B'injection path failure, the Loop "A Injection path is precluded from use. Removal of the LPCI Loop Select Logic or Installation of a bypass switch would allow use of the 'AZ loop for injection in the event of a 'B injection

___________________ path failure.

221 Demonstrate RCIC This SAMA would #6 - Retain Determine if demonstrating the operability Still retained. Xt operability following Increase the operators' of RCIC after depressurization Is a cost-depressurization options for low pressure beneficial effort Altematively, Emergency vessel Injection. depressurization could be directed to be

________ ____________________ ____________________ stopped p at at 100 I peiq.A0 m_________so 73

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revsed Disposition Phase IX SAMA ID enhancement Screening Criteria Including Uncertainty and SAM A ID number External Events number 225 Allow cross connection of SAMA would increase the #3 - Already An inter-unit Instrument Air crosstle valve Mods EC 335806 and EC 17 uninterruptable ability to vent containment implemented at already exists at Quad Cities and can be 335807 have been compressed air supply to using the hardened vent. Quad Cities opened locally. A connection to the cancelled due to large opposite unit. Service Air System also exists for each scope of needed equipment Revised to: unit (the unit Service Air compressors changes.

1. 6 - Retain output to a common header such that the two units are normally fully cross-tied). Now pursuing hookup of temporary compressor to A plant modification Is already approved to existing IA connections. A Increase Instrument air reliability for such technical evaluation (EC things as venting for long-term sequences, 339420) has been by providing for connection of a truck- performed that includes the mounted compressor. Unit 1 &2 necessary requirements for Instrument Air Mods (EC 335806 and the temporary air hose, EC335807, respectively) add ability to tie Including a description of In truck-mounted IA compressor to IA the flow path and the system to allow opening of containment connections to the air vents In cases of extended loss of header.

lA/containment heat removal. The modification to be installed by 12/31/02 This SAMA Is now retained provides the necessary piping and to determine the potential supports to permit temporary hook-up of a cost benefit of such a 1600 CFM, diesel Driven, Air Compressor change.

to a 3 NPT Threaded connection on the Instrument Air System. Several area rental facilities have been contacted and all have stated that they have the ability to provide a temporary compressor wIthin 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of notification regardless of the day or time. With this hookup Installed, it can reasonably be expected that te system can be pressurized well before the containment venting valves are required to

_____ ____ ___ _____ _____ _____ _____ ____ ___ ___ ___ ___ ___ ___ o p er ate.j:

. _S_S_ _= _ _ _ _ _X __:Xrt 74

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original DIsposition Revised Disposition Phase II SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 237 Bypass MSIV Isolation in SAMA will afford #6 - Retain Bypass of MSIV Isolation is procedurally Still retained. Now 16 Turbine Trip ATWS operators more time to directed in the EOPs; however, this action subject to analysis.

scenarios perform actions. The requires the use of jumpers. A dedicated discharge of a substantial switch for bypassing the low level interlock fraction of steam to the would be desirable.

main condenser (.e., as opposed to into the primary containment) affords the operator more time to perform actions (e.g., SLC injection, lower water level, depressurize RPV) than I the main condenser was unavailable, resulting In lower human error probabilities 242 Diversify the explosive An altemate means of #6 - Retain SBLC injection failure Is a dominant Still retained 11 valve operation opening a pathway to the contributor to ATWS mitigation failure.

RPV for SBLC injection Evaluate SBLC system improvements.

would Improve the success probability for reactor shutdown. _ l 243 Enrich Boron The Increased boron #6 - Retain Increasing the boron concentration for Still retained. 12 concentration will reduce SBLC may be a cost effective means of the time required to reducing ATWS risk.

achieve the shutdown concentration. This will provide Increased margin In the accident timeline for successful operator I _______ ____________________ activation of SBLC. _

75

Table 7-2 Revised Phase I SAMA Dispositlon (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potentIal OrigInal I Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 245 Create/enhance RCS With either a new #5 - Cost would be PWR Issue related to the limited The cost is considered to NWA depressurizatlon ability depressurization system, more than risk depressurization capability of the PWR. In be greater than the upper or with existing PORVs, benefit addition, reference 19 estimates the cost bound maximum averted head vents, and of this SAMA to range between $500,000 cost risk of $500K. No secondary side valve, and $4.6 million. For Ouad Cities, more change to the screening RCS depressurization effective depressurization capabilites criteria category.

would allow earlier low would require significant hardware pressure ECCS injection. changes and/or additions on top of the Even If core damage analysis that would be required to occurs, low RCS pressure implement the change. The cost estimate would alleviate some for the modification is considered to be on concerns about high the high end of the range provided In pressure melt ejection. Reference 19. The cost of Implementation for this SAMA is judged to greatly exceed the maximum averted cost-risk for Quad Ciies as defined in Section F.4.7.

249 Install secondary side This SAMA would prevent #5 - Cost would be This Is primarily a PWR issue. The steam Considering uncertainty N/A guard pipes up to the secondary side more than risk lines for a BWR Inside the inboard MSIV and potential impacts MSiVs depressurizatlon should a benefit are completely within the containment from external events does steam line break occur requiring no guard pipe. Between the two not Introduce any upstream of the main MSIVs Isa very short length of pipe that significant changes. No steam isolation valves. contributes a negligibie amount to the CDF change to the screening This SAMA would also and LERF. The addition of a guard pipe to criteria category.

guard against or prevent the steam tunnel for the short pipe length consequential multiple is judged to be very expensive and SGTR following a Main substantially In excess of any potential Steam Une Break event, benefit associated with risk reduction.

250 Instali digital large break Upgrade plant #5 - Cost would be Large break LOCA risk Is low. Upgraded Considering uncertainty N/A LOCA protection instrumentation and logic more than risk Instrumentation is unproven, benefit Isnot and potential Impacts to improve the capability benefit known, cost Is highly uncertain. The from external events does to Identify Implementation could not be realistically not introduce any symptoms/precursors of a justified. significant changes. No large break LOCA (leak change to the screening before break). criteria category. I 76

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500CK)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised Disposition Phase 11 SAMA ID enhancement Screening Criteria Including Uncertainty and SAMA ID number External Events number 255 Increase seismic SAMA would increase the #3 - Already Refer to SAMA 251. Also see Response 6(b). N/A ruggedness of plant availability of necessary Implemented at Seismic issues were examined In the components. plant equipment during Quad Cities Quad Cities IPEEE and the cost-effective and after seismic events. means of reducing plant risk were implemented as part of the program.

These changes include:

Replacing mercury switches In the Fire Protection System Improving MCC mounting and anchor welds Enhancing battery restraints 260 I.e. Improved Accident SAMA will Improve #5 - Cost would be The risk as measured by CDF, LERF, and Considering uncertainty N/A Management prevention of core melt more than risk population dose is low. The and potential impacts Instrumentation sequences by making benefit Instnrmentation available to the operating from external events does operator actions more crew at Quad Cities is comparable to that not introduce any reliable. available at other BWRs. Based on a significant changes. No review of the accident sequences that change to the screening contribute to the Quad Cities risk profile, criteria category.

the estimated risk reduction associated with additional accident mitigation Instrumentation Is judged to be neglilgible.

265 4.d. Passive This SAMA will prevent #6 - Retain This SAMA may be a cost effective means Still retained. 13 Overpressure Relief catastrophic failure of the of reducing risk at Quad Cities.

containment. Controlled X- I relief through a selected Quad Cities has installed a hard piped vent path has a greater containment vent system that provides a potential for reducing the controlled means of containment release of radioactive overpressure relief. The passive feature material than through a of adding a rupture disk to this system random break. Introduces competing risks that limit the usefulness of the vent over the spectrum

____ _ ______ ______ of severe accidents. _ _ _ _ _ _ _ _ _ _ _ _ _ _

271 Train operations crew for This SAMA would #4- No significant The 120V AC system Is not risk significant Considering uncertainty N/A response to inadvertent improve chances of a safety benefit at Quad Cities. While other plants have and potential impacts actuation signals successful response to identified specific 120V AC failure from external events does the loss of two 120V AC scenarios that would lead the generation not Introduce any buses, which may cause of inadvertent signals, no comparable significant changes. No Inadvertent signal vulnerabilities have been identified at change to the screening l__________________ generation. Quad Cities. criteria category. I _ _

77

Table 7-2 Revised Phase I SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase I SAMA title Result of potential Original / Revised Original Disposition Revised DIsposition Phase II SAMA ID enhancement Screening Criteria Including UncertaInty and SAMA ID number External Events number 272 Install tornado protection This SAMA would #4 - No significant Additional measures could be taken to Considering uncertainty N/A on gas turbine generators improve onsite AC power safety benefit Improve the protection of the on-site AC and potential Impacts reliability. power sources; however, the IPEEE from external events does Investigated risk from high wind events not Introduce any and found It to be negligible. Specifically, significant changes. No the emergency diesel generators are In change to the screening safety category I structures. criteria cateqory.

277 Use RHRSW cross tie This SAMA was identified #4 - No significant The physical capability to establish the The RHRSW cross tie N/A from opposite unit as part of the risk insights safety benefit cross tie exdsts. There are system from the opposite unit is from the Quad Cities procedures to perform the alignment. The credited In the Internal PRA. Insight merely is to establish additional events and fire portion of training and to specify when It can be the PRA model. The HEP used. This Insight while considered useful values are based on the for further Investigation is a safety procedural direction enhancement that results In a small provided In QCOA-1000, unmeasurable risk reduction benefit. QCOP-1000-15, OCOP-1000-20, and OCOP-

__ _ _ _ __ __ __ _ _ _ _ _ _ __ _ _ _ _ _ __ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 1000 -3 0. _ _ _ _

278 Provide mechanical stops This SAMA seeks to #4- No significant Calculation for BWR containment Considering uncertainty N/A on AOVs for venting physically prevent rapid safety benefit depressurization rates show that such and potential Impacts containment physical stops are not adequate by from external events does depressurization during themselves for this purpose. not Introduce any venting by imposing significant changes. No physical stops on the vent change to the screening valves. criteria category.

279 Control containment This SAMA was derived #6 - Retain There is a minor potential risk reduction Still retained. 14 venting within a narrow from the Quad Cities Risk associated with the SAMA and a cost band of pressure Insights document to associated with procedure changes, establish a narrow training, and documentation.

pressure control band that would thereby prevent rapid containment depressurization when venting is Implemented thus avoiding adverse Impacts on the low pressure ECCS Injection systems taking suction

- - _____________________ from the torus. ;_._._i_______

78

Response 7(c):

"[Provide) an assessment of the impact on the Phase 2 evaluation if risk reduction estimates are increased to account for uncertainties in the risk assessment and the additional benefits associated with extemal events (as applicable). Please consider the uncertainties due to both the averted cost-risk and the cost of implementation to determine changes in the net value for these SAMAs."

To perform this assessment, a two-step approach was taken. The first step was to reexamine the Phase II evaluation utilizing an upper bound maximum averted cost estimate of $500K consistent with the revised Phase I screening. This revised screening would then result in a set of potential plant changes that could be cost beneficial when compared to the upper bound estimate of the averted cost. For these potential enhancements, a comparison was then made to a more realistic best estimate averted cost to determine if the proposed change would be cost beneficial.

To provide an upper bound estimate on the risk reduction estimates to account for potential uncertainties on the risk assessment and the additional benefits associated with external events, each of the previously retained Phase II SAMAs plus the additional retained SAMAs from the revised Phase I screening in Response 7(b) have been reassessed. The reassessment assumes that the maximum averted cost risk is $500K compared to the original maximum averted cost of $103K used in the ER. If the proposed SAMA would provide benefit to both the internal events CDF and the Fire CDF, then the upper bound estimate for the averted cost-risk is scaled accordingly (i.e.,

by a factor of $500K/$103K = 4.85). If the proposed SAMA is noted as having benefit mostly to the internal events CDF, and would offer minimal improvement to the fire CDF, then the upper bound estimate for the averted cost risk is obtained from a factor of 3 that represents the estimated 95th percentile value of the internal events CDF as indicated in Respbnse 7(a).

Additional Phase II SAMA Analyses The revised Phase I screening described in Response 7(b) resulted in three additional SAMAs being carried forward to Phase 2. One of those SAMAs was judged to be adequately characterized by another SAMA investigation to estimate the potential cost benefit. However, two additional Phase II SAMA analyses were also performed to support the revised screening provided in Table 7-3. Each of these is described below.

PHASE II SAMA NUMBER 15

==

Description:==

Provide means for inter-unit crosstie for TBCCW Model Changes: Set TBCCW initiating event frequency and all TBCCW component failures to 0.0.

79

Results: The results from this case indicate a decrease from the base CDF of 2.16E-6/yr to 2.05E-6yr. The decrease in CDF (reduction of 1.1 E-7/yr) applies primarily to loss of DHR and ATWS scenarios (Class II and IVA) due to the dependence of BOP systems on TBCCW. The main condenser and containment venting are DHR systems that are dependent on TBCCW. In addition, the main condenser and Feedwater systems support ATWS mitigation. There was no reduction in LERF (base LERF =

2.67E-7/yr). This would lead to an averted cost-risk of $5,757 utilizing the same methodology and assumptions that were utilized in the ER.

PHASE II SAMA NUMBER 16

==

Description:==

Enhance bypass of MSIV isolation interlock (ATWS)

Model Changes: Reduce HEP for operator failure to bypass MSIV low RPV level interlock (ATWS) from 0.91 to 1E-2. In addition, increase complementary HEP for operator successful bypass MSIV low RPV level interlock (ATWS) from 9E-2 to 0.99.

Results: The results from this case indicate a decrease from the base CDF of 2.16E-6/yr to 2.09E-6/yr. The decrease in CDF (reduction of 6.5E-8/yr) applies only to ATWS scenarios (Class IVA and IC). Maintaining the availability of the main condenser for decay heat removal enhances the ability for successful mitigation of ATWS events. The LERF decreased from the base LERF of 2.67E-7/yr to 2.64E-7/yr. This would lead to an averted cost-risk of $5,921 utilizing the same methodology and assumptions that were utilized in the ER.

PHASE II SAMA NUMBER 17

==

Description:==

Allow cross connection of uninterruptable compressed air supply to opposite unit. (or examine lower cost altemative of providing backup air bottles or portable compressors).

The largest benefit of this SAMA would be derived by making the containment vent system more reliable. Consequently, it was judged to be adequately characterized by Phase II SAMA 13 (i.e., Passive Containment Overpressure Relief) that had previously considered the potential averted cost from eliminating all containment venting failures.

This SAMA had been shown to result in an averted cost-risk of $6,797. This is the value that is also used for Phase II SAMA 17.

The results of the reassessment including the three new Phase II SAMA analyses are provided in Table 7-3. The potential costs are consistent with those provided in Response 12.

80

Table 7-3 Revised Phase 11SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase 11 Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMA title enhancement Estimate Potential Cost Revised Disposition 1 32 Provide means for The SSMP requires room 4.85 * $11,303 >$1 M for Not cost beneficial. Current capabilities exist to alternate SSMP room cooling at extended times. . $54,820 independent utilize FPS as a backup means of providing SSMP cooling This SAMA would allow cooling room cooling. Procedural direction for performing SSMP operation late in 2 Units capabilities this action is provided inOCOP 2900-02. The HEP accidents when normal mom - $109,640 (BSSOPSSRMCLNGH-) for this action Iscurrently cooling has failed. 1.1 E-1 based on a lack of clear symptom-based direction for subsequent losses of service water following Initial use of the SSMP. However, all of the dominant cutsets that Include this HEP value result from a loss of service water Initiated event for which case, the procedural direction to utilize FPS for SSMP room cooling isvery clear. Based on a re-evaluation of the procedure, a significant reduction In the HEP value Isanticipated (for the loss of service water Initiated event) as part of the next PRA model update. This will greatly minimize the risk reduction worth associated with this HEP. No additional procedural change is required.

Hardware modifications to automate and/or provide Independent means of room cooling are Judged not to be cost beneficial, especially since the anticipated lower HEP value for aligning FPS from existing capabilifles will minimize the benefit of Implementing another redundant system.

2 36 Develop an enhanced SAMA would provide a 4.85 $9,418 $50-1OOK for The fire protection system (FPS) can already provideI drywell spray system. redundant source of water to - S procedural water to the RHR system at QCNPS, but procedures the containment to control 452 enhancements have not been developed to use it as a containment1 containment pressure, when Unith w2 engineering spray source. Assuring the viability of such a used Inconjunction with $91,354 analysis proposed change would also require engineering containment heat removal. required. analysis. However, the total ImplementatIon costs could be less than the upper bound averted cost:

estimate. Retain for best estimate cost benefit -

_ __ _ __ __ __ _ _ _ __ _ _ _ _ __ _ __ _ __ __ __ _ __ __ _ __ __ _ _ __ _ _ _ __ _ _ _ _ __ _ _ _ _ _ _ analysis (see Table 7-4).A 81

Table 7-3 Revised Phase 11SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase 11 Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMA title enhancement Estimate Potential Cost Revised DIsposition 3 94 Use fuel cells Instead SAMA would extend DC 4.85 * $4,406 >$1OOK for fuel Not cost beneficial. Either replacing batteries with of lead-acid batteries. power availability In an SO0. - $21,369 cells, or $50- fuel cells or a lower cost alternative of implementing

.2Units 1OOK for lower portable generators to prolong battery life would be 2 Units

• cost alternative more costly than the upper bound averted cost

. $42,738 of providing a estimate.

portable generator to the battery chargers and procedural Implementation /

training.

4 96 Improve 4.16-kV bus Enhance procedures to direct 4.85 * $578 $25-50K for Not cost beneficial. The upper bound averted cost cross-tie ability. 4kV bus cross-tie. If this - $2,803 procedural estimate of $2.8K Is far below the minimum procedural step already . enhancements procedural change estimate of $25K. Addbonally, exists, Investigate Installation 2 Units given the complications and concerns associated of hardware that would = 5,606 with cross-tleing buses, any related procedural perform an automatic cross- change Is probably more likely to be a higher cost tie to the opposite 4kV bus procedure change than a lower cost procedure given failure of the dedicated change.

diesel.

5 108 Create a backup This SAMA would provide a 4.85

• Negligible Not Required Not cost beneficial. Also see Response 13(c). The source for diesel redundant and diverse source = Negligible SBO DGs already Include a diverse source of diesel cooling. (Not from of cooling for the diesel generator cooling compared to EDG 1, EDG 2, and existing system) generators, which would EDG 1/2.

contribute to enhanced diesel reliability. l 6 131 Provide procedures This SAMA would allow for 4.85 * $30,171 $501 00K for ng procedural direction to bypass major q for (a)bypassing powering specific loads given * $146,329 procedural buses providing Instructlons for local start, and^

major DC buses; (b) a DC bus failure andlor the *2 U enhancements providing backup hardware capabilities for this locally starting ability to start equipment with engineering function may be cost beneficial when compared to' equipment locally that normally requires $292,658 analysis the upper bound averted cost estimate. Retain forl DC power for a control room required, plus best estimate cost benefit analysis (see Table 7-4).

start. $I00K minimum for hardware changes. _ _ _ _ X _ , _ _ _ _ _ _ _ _ _ _ _ _i?

82

Table 7-3 Revised Phase 11SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase II Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMA tti enhancement Estimate Potential Cost Revised Disposition 7 216 Delete High DW This SAMA would allow the 4.85 * $718 $25-50K for Not cost beneficial. The upper bound averted cost Pressure Signal from Initiation of SDC when the $3,482 procedural estimate is far below the minimum procedure change SDC isolation drywell Is at elevated

• 2 Units enhancements estimate of $25K.

pressures.

l $6,964 8 219 Develop procedures This SAMA Increases the 4.85 * $15,505 $100-200K for Overall Implementation costs would I to control Feedwater functionality of Feedwater in - $75,199 procedural developmental work and extensive training.

flow without 125 VDC loss of DC scenarios and enhancements, However, this could be cost beneficial when -

power to prevent Increases the probability of '2 Units analysis, and compared to the:upper bound averted cost estimate.-

tripping Feedwater on successful level control. $150,398 testing Retain forbest estimate cost benefit analysIs (see High/Low level Table 7-4).

9 220 Remove Loop Select In the event that there Is no 3 Negligible Not Required Not cost beneficial. The benefit from this change is Logic break In the recirc loops and = Negligible llmited to LOCA scenarios.

there is a Loop WBinjection path failure, the Loop 'AY (Not a injection path Is precluded contributor to the from use. Removal of the Fire CDF)

LPCI Loop Select Logic or installation of a bypass switch would allow use of the WA' loop for injection in the event of a 'B' injection path failure.

10 221 Demonstrate RCIC This SAMA would Increase 4.85 * $20,309 $100-200K for Overall Implementation costs would Include operability following the operators' options for low - $98,499 procedural developmental work and extensive training.

depressurlzation pressure vessel injection. 2 U enhancements, However, thisbcould be cost beneficial when 2 Units analysis, and compared to the upper bound averted cost estimate.

= $196,998 testing Retain for best estimate cost benefit analysis (see l_ _ _ _ _ _ _ __ _ __ _ __ _ _ __ _ __ _ __ _ _ __ _ _ _ _ _ _ __ T able 7-4).

11 242 Diversify the explosive An aitemate means of 3 ' $2,390 >$1 00K / unit Not cost beneficial. Any hardware change would valve operation opening a pathway to the - $7,170 easily exceed the upper bound averted cost RPV for SBLC Injection would estimate.

improve the success (ATWS Is not a probability for reactor signiticant shutdown. contributor to the Fire CDF) _____

83

Table 7-3 Revised Phase 11SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase II Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMAtil enhancement Estimate Potential Cost Revised Disposition 12 243 Enrich Boron The increased boron 3 * $858 Not Required Not cost beneficial. Minimal benefit is obtained and concentration will reduce the = $2574 associated implementation costs would easily time required to achieve the exceed the upper bound averted cost estimate.

shutdown concentratlon. This (ATWS Is not a will provide Increased an significant Increased margin In the contributor to the accident timeline for Fire CDF) successful operator activation of SBLC.

13 265 4.d. Passive This SAMA will prevent 4.85 * $6,797 >$$100K / unit Not cost beneficial. Implementation of this SAMA Overpressure Relief catastrophic failure of the = $32965 would Involve extensive hardware changes that containment. Controlled relief would exceed the upper bound averted cost through a selected vent path estimate.

has a greater potential for reducing the release of radioactive material than through a random break.

14 279 Control containment This SAMA was derived form 4.85 * $22,150 $100-200K for Current proceduresallow c e iblty venting within a the Quad Cities Risk Insights = $107,428 procedural Implementingcontainment venting. Additionally,,-

narrow bend of document to establish a enhancements there IsWplenty of time for the Emergency Response pressure narrow pressure control band 2 Units with engineering Organizatlon to develop a strategy to supplement the that would thereby prevent = $214,856 analysis and guidance In the current procedure. However, rapid containment testing required. Implementing, testing, and establishing' a procedure depressurization when for the recommended approach may be cost :

venting is Implemented thus beneficial when compared to the upper bound avoiding adverse Impacts on averted cost estimate. Retain for best estimate cost the low pressure ECOS benefit analysis (seeTable 7W-4).C:;I m injection systems taking suction from the torus.

150) 20 Enhance procedural SAMA would reduce the 4.85 $5,75702 Aitemative Not cost beneficial. Implementation of this SAMA guidance for use of frequency of the loss of - $27,921 investigated to would Involve extensive hardware changes that cross-tied component component cooling water and provide TBCCW would exceed the upper bound averted cost cooling or service service water. Units c2ross-tie estimate.

water pumps. = $55,842 capabilities to other unit.

$100K minimum for hardware change.

84

Table 7-3 Revised Phase II SAMA Disposition (Assuming Maximum Averted Cost Risk of $500K)

Phase II Phase I Upper Bound SAMA ID SAMA ID Result of potential Averted Cost number number SAMA tMe enhancement Estimate Potential Cost Revised Disposition 16(1) 237 Bypass MSIV Isolation SAMA will afford operators 3 $5,921 I) $50-1OOK for Not cost beneficial. Implementation of this SAMA in Turbine Trip ATWS more time to perform actions. $17,763 procedural would Involve procedural and hardware changes that scenarios The discharge of a substantial enhancements would exceed the upper bound averted cost fraction of steam to the main 2 Units with engineering estimate.

condenser i.e., as opposed m $35,526 analysis to into the primary (ATWS Is not a required, plus containment) affords the significant $1OOK minimum operator more time to perform contributor to the for hardware actions (e.g., SLC injection, Fire CDF) changes to lower water level, implement depressurize RPV) than If the automatic MSIV main condenser was isolation bypass unavailable, resulting In lower capabilities.

human error probabilities 17(t) 225 Allow cross SAMA would increase the 4.85 * $6,797(3) Lower cost Implementation of hi SAMA would requiel connection of ability to vent containment - $32,965 aitemative of procedural and hardware changes. However, this uninterruptable using the hardened vent . providing backup could be cost beneficial when compared to the upper compressed air 2 Units bottles or -bound averted cost estimate. Retain for best -- I i I supply to opposite = $65,930 portable air estimate cost benefit analysis (see Table 7-4). l unit. compressors ------

estimated at

$50-1OOK for procedural enhancements, -- -----

training, and hardware modffications.

Notes to Table 7-3 (1)This is a new Phase II SAMA identifier that was not Included in the ER.

(2) Detailed development of the PRA model changes made for this Phase II SAMA investigation are provided prior to the table.

(3) This SAMA is conservatively estimated as providing the same benefit as Phase II SAMA 13 (with vent failure modes set to zero).

85

Response 7(c) - continued:

"[Provide]an assessment of the impact on the Phase 2 evaluation if rsk reduction estimates are increased to account for uncertainties in the risk assessment and the additionalbenefits associatedwith external events (as applicable). Please consider the uncertainties due to both the averted cost-risk and the cost of implementation to determine changes in the net value for these SAMAs."

As can be seen in Table 7-3, six of the Phase 11SAMAs could be categorized as cost beneficial when compared to the upper bound averted cost estimate. It should be noted, however, that there are many factors to consider when looking at the benefits of the SAMA candidates. Plant specific implementation of SAMA candidates may be complicated by space limitations, outage costs, regulatory requirements, and other considerations. These factors tend to result in underestimation of the costs.

Additionally, the specific PSA analyses that were performed in addressing specific SAMA candidates were done optimistically. That is, the potential cost-benefit was derived from a case that maximized the CDF reduction that would result from implementation of the SAMA. Both of these factors would, in effect, offset the uncertainties associated with the CDF estimates.

A factor of 2 is therefore judged as a reasonable value to use as the best estimate averted cost risk to account for uncertainties and potential impacts from external events.

While the reported fire CDF in the IPEEE is more than a factor of two higher than the current internal events CDF, there have been several plant changes that have occurred to address the insights obtained from the external events analysis (as detailed in Response 6(a) and Response 6(b)), and it is judged that it is not appropriate at this time to directly compare internal events CDF values with external events CDF values'.

Consequently, while it is agreed that the averted costs could be more than the actual implementation costs in this case if the implementation costs are compared to the upper bound averted cost estimates, when compared to best estimate averted cost estimates, none of the SAMAs end up as being cost beneficial. The best estimate is obtained by using a factor of 2 on the unadjusted internal events averted cost (to account for uncertainties and external events, but not both simultaneously). The results of this additional screening are illustrated in Table 7-4.

3 Attachment A provides an assessment of the use of quantitative risk estimates from Fire PRAs, and why it isjudged that the calculated CDF values should not be directly compared at this time.

86

Table 7-4 Best Estimate Phase 11SAMA Disposition of Remaining SAMA Candidates Phase II Phase I SAMA ID SAMA ID Result of potential Best Estimate number number SAMA tMe enhancement Averted Cost Potential Cost Best Estimate Disposition 2 36 Develop an enhanced SAMA would provide a 2.0 * $9,418 $50-l OOK for Not cost beneficial. The fire protection system (FPS) drywell spray system. redundant source of water to = $18,836 procedural can already provide water to the RHR system at the containment to control enhancements QCNPS, but procedures have not been developed to containment pressure, when 2 Units with engineering use It as a containment spray source. Assuring the used in conjunction with = $37,672 analysis viability of such a proposed change would also containment heat removal. required. require engineering analysis. The overall Implementation costs are estimated to be higher than the best estimate averted cost.

6 131 Provide procedures This SAMA would allow for 2.0 * $30,171 $50-1 00K for Not cost beneficial. Preparing procedural direction for (a) bypassing powering specific loads given = $60,342 procedural to bypass major DC buses, providing Instructions for major DC buses; (b) a DC bus failure and/or the ' enhancements local start and providing backup hardware locally starting ability to start equipment 2 Uns with engineering capabilities for this function would lead to overall equipment ocally that nommally requires = $120,684 analysis implementation costs that are higher than the best DC power for a control room required, plus estimate averted costl start. $lOOK minimum for hardware changes.

8 219 Develop procedures This SAMA increases the 2.0' $15,505 $100-200K for Not cost beneficial. Costs would Include to control Feedwater functionality of Feedwater In -$31,010 procedural developmental work and extensive training. This flow without 125 VDC loss of DC scenarios and enhancements, would lead to overall Implementation costs that are power to prevent Increases the probability of 2 Units analysis, and higher than the best estimate averted cost.

tripping Feedwater on successful level control. - $62,020 testing HighiLow level 10 221 Demonstrate RCIC This SAMA would increase 2.0' $20,309 $100-200K for Not cost beneficial. Costs would Include operability following the operators' options for low - $40,618 procedural developmental work and extensive training. This depressurization pressure vessel Injection. enhancements, would lead to overall Implementation costs that are 2 Units analysis, and higher than the best estimate averted cost.

=$81,236 testing 87

Table 7-4 Best Estimate Phase 11SAMA Disposition of Remaining SAMA Candidates Phase 11 Phase I SAMA ID SAMA ID Result of potential Best Estimate number number SAMA titie enhancement Averted Cost Potential Cost Best Estimate Disposition 14 279 Control containment This SAMA was derived form 2.0 ' $22,150 $1O0-200K for Not cost beneficial. Current procedures allow venting within a the Quad Cities Risk Insights - $44,300 procedural considerable flexibility In Implementing containment narrow band of document to establish a enhancements venting, and there is plenty of time for the pressure narrow pressure control band with engineering Emergency Response Organization to develop a that would thereby prevent analysis and strategy to supplement the guidance In the current rapid containment testing required. procedure. Additionally, Implementing, testing, and depressurization when estabilshing a procedure for the recommended venting Is Implemented thus approach would lead to overall Implementation costs avoiding adverse Impacts on that exceed the best estimate averted cost.

the low pressure ECCS Injection systems taking suction from the torus.

17() 225 Allow cross SAMA would Increase the 2.0 $6,797(3) Lower cost Not cost beneficial. Implementation of this SAMA connection of ability to vent containment - $13,594 altemative of would require procedural and hardware changes.

uninterruptable using the hardened vent. . providing backup This would lead to overall Implementation costs that compressed air 2Uns bottles or exceed the best estimate averted cost.

supply to opposite $27,188 portable air unit. compressors estimated at

$50- OOK for procedural enhancements, training, and hardware modiflcations.

88

RAI8 For certain SAMAs considered in the ER, there may be lower cost alternatives that could achieve much of the rsk reduction. As one example, Phase 2 SAMA #3 evaluated the use of fuel cells instead of lead-acid batteries, but lower cost alternatives, such as adding a diesel-driven battery charger, were not explored. Please confirm that low cost alternatives to Phase 2 SAMAs were considered, and provide a brief discussion of these alternatives.

Response 8 Lower cost alternatives were considered in both the initial Phase I screening all the way through to the final revised Phase II screening. Examples included a portable generator to provide prolonged battery capacity (see Table 7-3, Phase II SAMA 3), and backup bottles or portable compressors for supplementing instrument air capabilities (see Table 7-3, Phase II SAMA 17). Several additional lower cost alternatives were also explored in the form of potential procedural changes (see Table 7-3, Phase II SAMAs 2, 4, 6, 7, 8, 10, and 14). While many of these may only involve procedural changes in concept, a more thorough investigation leads to the finding that more costs would actually be incurred when considering that the procedure changes may also require engineering analysis, experimentation, and extensive training. (See also Response 12.) As such, none of the remaining SAMAs (including lower cost alternatives) were determined to be cost beneficial.

89

RAI9 During the review of the EPU application, the staff noted several areas where the PSA should be modified to reflect modifications to the plant or changes in success paths.

These include: a plant modification to install a recirculation pump runback control circuit; a plant modification to trip the condensate/booster pump D in the event of a LOCA to prevent an overload condition from occurring; a change in success criteria for reactor pressure vessel (RPV) depressurization in a transient without a stuck open relief valve (two valves under EPU conditions); a change in success crtera for RPV overpressure protection in ATWS sequences (12 of 13 valves under EPU conditions).

Confirm if these model changes, as wel as others, have been incorporated in the PSA used for the SAMA analysis. For those not incorporated, provide an assessment of the impact that the model change would have on the SAMA analysis.

Response 9 The model was revised to include all appropriate EPU changes:

• The purpose of the recirc. pump runback control circuit is to prevent the reactor trip frequency from increasing due to EPU. The recirc. pump runback is needed because there no longer are "spare' condensate pumps or feedwater pumps. Due to this modification, the transient initiating event frequency is not expected to change. However, effects on the plant can only be incorporated in the PRA after some plant experience via the next periodic update of initiating event frequencies.

The potential risk impact of the recirc. runback modification was addressed in a response to a NRC RAI to support the EPU application

[Reference 9-1]. The response to the RAI addressed both 1) the failure of the recirc. runback to operate as designed, and 2) spurious recirc.

runback. The RAI judged that the incorporation of the recirc. runback modification would result in a negligible risk increase.

• The circuit to trip condensate/condensate booster pump MD" on a LOCA signal is expected to be very reliable. The risk impact of the condensate/condensate booster pump SD" trip logic was also addressed in Reference 9-1. The risk impact was calculated to be 1.7E-10/yr. Due to the minor contribution to CDF, this failure mode was not explicitly included in the PRA model.

• The success criterion for RPV depressurization is reflected in the revised transient without SORV model.

• The success criterion for ATWS overpressure protection is reflected in the revised ATWS model.

• The higher decay heat load due to power uprate reduces the time available for certain operator actions. This has been reflected in revised HEP's for those actions.

90

REFERENCE

[9-1] Letter from K.A. Ainger, Exelon Generation Company, to U.S. NRC, 'Additional Risk Information Supporting the License Amendment Request to Permit Uprated Power Operation at Dresden Nuclear Power Station and Quad Cities Nuclear Power Station", RS-01-1 68, August 14, 2001.

91

RAI 10

During the review of the EPU application, the staff noted that there is potentially a new means of inducing a LOOP initiating event under EPU conditions. The end result could be an overduty condition on the unit auxiliary or reserve auxiliary transformer. Given this new condition, please provide an evaluation of the costs and benefits associated with the replacement of the affected transformer with a higher capacity transformer.

Response 10 The risk impact of the induced LOOP initiating event was addressed in a response to a NRC RAI to support the EPU application [Reference 10-1]. Information from the response to the RAI is summarized below.

BACKGROUND During normal operation the station loads are distributed between the Unit Auxiliary Transformer (UAT) and the Reserve Auxiliary Transformer (RAT). Normally, the loads for two non-essential 4kV buses are aligned to the UAT and the loads for the other two non-essential 4kV buses are aligned to the RAT. If either the UAT or RAT become unavailable during normal operation without a reactor scram, the increased loads for the EPU configuration may result in an overload condition for the remaining transformer's bus duct connection to the 4kV buses.

The scenario of concern is a loss of the UAT or RAT due to transformer failure, failure of protective relaying (e.g., false fast transfer signal), or spurious opening of multiple circuit breakers [see note (1)], causing a fast transfer of all running loads to the other transformer. Under these conditions, certain bus duct segments are overloaded, requiring operator action within one hour to reduce load to within the bus duct rating.

This action will be procedurally directed. The one hour time frame for load reduction was determined based on an Exelon Generation Company (EGC), LLC evaluation of a General Electric Company study on short term overload conditions for the bus ducts.

The simplifying assumption is made that failure to take this action would lead to a loss of offsite power (LOOP). In reality, overload of the bus duct results in heating above the allowable temperature limits if ambient temperature is at the design value. No deterministic evaluation has been conducted to determine if overheating will result in complete failure of the bus duct, thereby causing a LOOP.

RESULTS The induced LOOP initiating event is calculated to result in a 6E-9/yr increase in the Quad Cities Level 1 CDF. The risk evaluation accounts for the estimated frequency of (1) Spurious opening of an individual circuit breaker to an individual 4kV bus would cause a fast transfer of the individual 4kV bus loads to the alternate transformer. However, based on the estimated EPU loads, the transfer of loads for a single 4kV bus (i.e., loads from three 4kV buses on a single transformer) would not place the transformer bus ducts in an overload condition.

92

the transformer overduty condition and failure of the plant or operating staff to mitigate the event.

CONCLUSIONS FOR SAMA Based on the minor risk impact, the costs associated with the replacement of the affected transformer or associated electrical equipment (e.g., 4kV bus duct connections) is judged not to be warranted.

Additional details of the risk calculation can be found in Reference [10-1].

REFERENCE

[10-1] Letter from T. W. Simpkin (Exelon Generation Company) to U. S. NRC, uAdditional Information Supporting the License Amendment Request to Permit Uprated Power Operation, Dresden Nuclear Power Station and Quad Cities Nuclear Power Station," RS-01 -200, dated September 19, 2000.

93

RAI 11

In the original IPE (1993), the CDF was dominated by a dual-unit LOOP (contributing 56% to the internal events CDF). The Fussell-Vesely importance measure indicated that the most significant hardware contributors toward total CDF are the failures of the diesel generators (DGs), and the quantitative importance of emergency AC power sources is influenced significantly by the dependency of the plant on electrically-driven systems for long-term decay heat removal. In the modified IPE submittal (August 1996), the contribution for dual-unit LOOP remained unchanged. In the updated IPE (December 1996), the contribution to CDF has dropped to 33% (after two station blackout (SBO) DGs were added), however, the contribution to CDF remains significant.

SAMAs that involve adding a DG, adding batteries, and the like were evaluated by QCNPS but eliminated on the basis that the plant already has five DGs, spare batteries, and the other SAMAs are too costly. Other than these improvements, please describe what measures or evaluations have been performed at QCNPS to reduce the risk from single- and dual-unit LOOP. Include a discussion of how the new SBO DGs are modeled in the current PSA including key assumptions.

Response 11 The CDF in the 2002 Quad Cities PRA Update is the same as the 1996 Updated IPE.

This agreement in the total value is coincidental, given the number of model changes that have occurred since the Updated IPE. However, the dual-unit LOOP contribution is now 17% of the CDF instead of 33%. The single-unit LOOP contribution is now 2%

instead of 22%. The combined contribution is now 19% instead of 55%.

The update that followed the 1996 Updated IPE, the 1999 Upgrade, was a major change, and it involved a conversion to the single-top fault tree methodology from the support state methodology previously used. Because of this, it is difficult to compare the model results directly. However, the changes that most likely contributed to the reduction in importance of offsite power are the following:

• The single-top fault tree better represents dependencies on support systems. For example, common-cause failure modes between diesel-generators in the support-state model required complicated conditional probability calculations between dependent event tree nodes. Within the single-top fault tree, dependencies are modeled explicitly using a linked fault tree approach. The dependencies of frontline equipment on support systems are more clear and precise. In addition, although credit for SBO diesel-generators was included in the Updated IPE, the single-top fault tree better represents the multitude of alignments of those diesels, as well as the multitude of bus alignments between units possible at Quad Cities.

• The diesel-generator mission time was reduced from 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> to 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, consistent with the method for Peach Bottom in NUREG-4550.

• Data for loss of offsite power, loss of offsite power recovery, and plant equipment reliability and availability were updated in 1999 and, again, in 94

2002. Industry loss of offsite power performance has improved. And the plant-specific experience with diesel-generators, with breakers, and with turbine-driven pumps has significantly improved.

• Exelon revised the common-cause factors based on NUREG/CR-5497 and NUREG/CR-5485.

• Exelon completely revised the Human Reliability Analysis, using industry standard methods, the latest plant procedures, operator interviews, and simulator observations.

The combined effect of all of these changes has resulted in considerably reduced importance of loss of offsite power. In addition, iRappears that the updated IPE, while giving credit for the SBODG's, perhaps did not give sufficient credit.

Each unit SBODG can be aligned to either electrical division of either unit. In fact, since it is larger than an EDG, one SBODG can handle the shutdown loads of both units. This flexibility and operator actions based on the very detailed operating procedures for the SBODG's are reflected in the model. While the SBODG's are of the same manufacture as the EDG's, they are of larger size, are tandem machines, and have updated control systems. The model includes common-cause failure of all five diesel-generators, but the factor used is smaller than if the five diesel-generators had been identical.

95

RAI 12

In Section 4.20.5 of the ER, Exelon states that a preliminary cost estimate was prepared for each of the remaining candidates (surviving the initial screening). In Section 4.20.6, it is stated that a more detailed implementation cost assessment is made only ff the benefit is close to the estimated implementation cost. However, no implementation costs were provided for any of the Phase 2 SAMAs. Please provide the estimated implementation costs (preliminary cost estimates) for the 14 Phase 2 SAMAs, so that the staff can readily determine ff any of these SAMAs are potentially cost-beneficial when considering the impact of external events and uncertainties. In addition, indicate what minimal costs were assumed for procedure changes, and what minimal costs were assumed for hardware changes.

Response 12 For all of the Phase 2 SAMAs evaluated in Section 4.20.5 of the ER, none of them had a benefit that was close to the potential implementation cost. Therefore, no detailed costs were required. As a supplement to the original SAMA evaluation, Exelon has developed the following estimated implementation costs for use in Response 7(c).

These costs have been estimated based on existing SAMA evaluations and have addressed the following cost elements:

• Procedural changes

• Engineering evaluations

• Hardware modifications

• Testing required to support procedural changes and engineering evaluations The following references have been used to assign an appropriate cost to these elements.

REFERENCES

[12-1] NUREG-1437, 'Generic Environmental Impact Statement for License Renewal of Nuclear Plants, Oconee Nuclear Station", Supplement 2, U.S. Nuclear Regulatory Commission, Washington, D.C., December 1999.

[12-2] Peach Bottom SAMA Evaluation and RAI Responses

[12-3] HB Robinson SAMA Evaluation and RAI Response

[12-4] VC Summer SAMA Evaluation and RAI Response

[12-5] GE Nuclear Energy, "Technical Support Document for the ABWR," 25A5680, Rev. 1, November 1994.

.96

PROCEDURAL CHANGES Procedure development and modification requires preparation by a System Engineer, technical review and validation, oversight review, and a variety of additional plant reviews prior to release. In addition, plant staff will need to be trained prior to implementation. A few examples of other procedure change estimates are provided below.

• ABWR [12-5] indicates that improvements to existing maintenance procedures would cost approximately $300K.

• PB [12-2] describes a procedural modification to allow for cross-tie of CCW at an estimated implementation cost of $50K.

For the Quad Cities SAMA analyses, a range for procedural changes is estimated to cost from $25K to $50K. The lower estimate is judged to be more appropriate for changes to existing procedures, and the upper estimate is judged to be more appropriate for the development of new procedures.

ENGINEERING EVALUATIONS In support of procedural and hardware modifications, an engineering evaluation will be required. For a procedural modification, the engineering requirements could easily double the cost of the change. This would increase the procedural change cost to an estimated range of $50K to $100K.

HARDWARE MODIFICATIONS The following provides examples from previous SAMA evaluations.

• PB [12-2] evaluated alternate methods to provide cooling to the RHR pumps at an estimated implementation cost of $250K.

• PB [12-2] also estimated a cost of $1600K to replace all 8 station batteries.

• Numerous hardware changes were evaluated for the ABWR [12-5] at a cost range from $1000K to $6000K.

• Hardware modifications were evaluated for Oconee [12-1] including automatic refill systems for the refueling water storage tank, automatic switchover of HPI to the spent fuel pool, and others ranging from $1000K to $5000K.

97

For the Quad Cities SAMA analysis, several hardware modifications have been evaluated and range in cost from $100K to over $1000K. A minimum of $100K is used to account for engineering analysis, purchase, and maintenance of any proposed hardware modification.

TESTING Similar to engineering costs to support a procedural change, testing of a plant system to establish operating limits is estimated to double the cost of the procedural change. An example of this would be for a proposed SAMA to justify the operation of RCIC at low RPV pressures. Procedural changes in addition to potential testing costs could increase the overall implementation cost to a range of $100K to $200K.

SUMMARY

OF IMPLEMENTATION COST Based on a review of previous SAMA evaluations and an evaluation of expected implementation costs at Quad Cities, Table 12-1 provides the estimated costs for each potential element of the proposed SAMA implementation. Depending on the individual elements involved with each proposed SAMA, these estimates are then used to determine the total implementation cost with the remaining Phase II SAMAs as described in Response 7(c).

Table 12-1 Estimated Implementation Costs Type of Change Estimated Cost Range Procedural only $25K-$50K Procedural change with engineering required $50K-$1 00K Procedural change with engineering and testing $I00K-$200K required Hardware modification $100K to > $1000K 98

RAI 13

For the Phase 2 SAMAs, the following information is needed to better understand the modification and/or the modeling assumptions:

a. Phase 2 SAMA 1: The benefit of this SAMA is said to be a decrease in the CDF which applies primarily to loss of decay heat removal and late SBO scenarios. One of the proposed improvements is a procedure for opening the safe shutdown makeup pump (SSMP) doors and using portable fans for SSMP room cooling. It is unclear how this improvement would work under SBO conditions. Please clarify if this improvement is only meant to work for loss of decay heat removal scenarios, and how it might work under SBO conditions.

b. In the IPE, one of the unique features identified at QCNPS is the ability to cross-tie between units in emergency buses 14-1 and 24-1. Phase 2 SAMA 4 evaluates the development of procedures to allow the following cross-ties to be performed:

Bus 14-1 to Bus 24-1 from EDG 1 Bus 24-1 to Bus 14-1 from EDG 2 EDG 1/2 to Buses 13-1 and 23-1 Explain why procedures have not already been developed for a cross-tie (Bus 14-1 to 24-1) that has been acknowledged in the IPE. Clarify whether this capability currently exists and is credited in the current PSA. If it is credited, please provide the key assumptions regarding this action (e.g., timing and operator non-procedural capability/knowledge) and the human error rate and its basis.

c. Phase 2 SAMA 5: The following statement is made in Section 4.20.6.5 of the ER, 'An additional EDG cooling source may be cost beneficial for Quad Cities." However, the analysis indicates that there is no benefit (averted risk). Explain why there is no benefit, and also explain why it was believed that such an improvement would be cost beneficial when there is no benefit.

d. For several Phase 2 SAMAs (6, 10, and 14), it appears that a majority of the effort would be in writing/revising procedures and training, and engineering, work. Given the additional benefit of these SAMAs in external events and the impact of uncertainties, the benefit of these SAMAs could be substantially higher than assumed in the ER. Explain why these SAMAs would not be cost beneficial when the benefits associated with external events, and the impact of uncertainties are considered.

99

Response 13(a):

"Phase 2 SAMA 1: The benefit of this SAMA is said to be a decrease in the CDF which applies primarly to loss of decay heat removal and late SBO scenarios. One of the proposed improvements is a procedure for opening the safe shutdown makeup pump (SSMP) doors and using portable fans for SSMP room cooling. It is unclear how this improvement would work under SBO conditions. Please clarify if this improvement is only meant to work for loss of decay heat removal scenarios, and how it might work under S80 conditions."

Approximately 95% of the potential benefit of this SAMA was determined to be from Class 11loss of containment heat removal scenarios and 5% was determined to be from Class IBL scenarios. The Class IBL late SBO characterization is based on the dominant cutsets for that sequence that do indeed include SBO-like conditions.

However, the cutsets that are removed from that same sequence (that lead to about 5%

of the noted CDF reduction) are actually better characterized as Class 11scenarios as well since they don't involve an actual SBO condition, just a LOOP initiated event with other combinations of system failures. Removing the SSMP room cooling dependency decreases the Class 11 frequency because the primary cooling source for the SSMP is from Service Water. (The existing backup SSMP room cooling source is from Fire Protection.) Removing the room cooling dependency reduces many of the Loss of SW cutsets that lead to the Class 11or Class IBL loss of decay heat removal sequences.

The proposed procedure for opening the SSMP room doors and using portable fans or SSMP room cooling was provided as an example potential option for removing the dependency. The benefit of this proposed enhancement would only occur in loss of decay heat removal scenarios, and would not be beneficial in true SBO scenarios since SSMP would also be unavailable. As described above, the benefit derived in the Phase 11 analysis is actually limited to loss of decay heat removal scenarios (some of which could occur from a LOOP/DLOOP initiated event with other combinations of system failures).

Response 13(b):

"In the IPE, one of the unique features identified at QCNPS is the ability to cross-tie between units in emergency buses 14-1 and 24-1. Phase 2 SAMA 4 evaluates the development of procedures to allow the following cross-ties to be performed:

Bus 14-1 to Bus 24-1 from EDG 1 Bus 24-1 to Bus 14-1 from EDG 2 EDG 1/2 to Buses 13-1 and 23-1 Explain why procedures have not already been developed for a cross-tie (Bus 14-1 to 24-1) that has been acknowledged in the IPE. Clarify whether this capability currently 100

exists and is credited in the current PSA. If it is credited, please provide the key assumptions regarding this action (e.g., timing and operator non-procedural capability/knowledge) and the human error rate and its basis. "

The intent of the SAMA investigation was to determine if improved reliability of existing cross-tie actions and/or expanded cross-tie capabilities would be cost beneficial. The Phase II SAMA analysis looked at improvements to three specific HEP values utilized in the PRA model to estimate the potential benefit for this SAMA. Two of the HEP values are based on existing cross-tie procedures, and the third event is based on non-procedural capability. Each of these HEPs are described below.

1. Current procedures exist for cross-tieing Bus 14-1 to 24-1 and Bus 13-1 to 23-1. These are dictated by Quad procedures (QCOA 6100-03 and QCOA 6100-08). This is represented by the HEP event BACOPXTIEBUS-H-- in the PRA model with an estimated time to perform the action of 10 minutes and an available time window of 40 minutes based on the limiting case of an SBO with early HPCI and RCIC failures. The HEP value of 1.1 E-2 was derived based on EPRI's cause based methodology supplemented with ASEP estimates for short time frame events such as this one, and using THERP for the execution error. The Phase II SAMA analysis included a reduction in the HEP value for this event from its base PRA value of 1.1E-2 to 1.1E-4.

2. Current procedures also exist for aligning the swing diesel (i.e., EDG 1/2) to Unit 1 or Unit 2, as applicable. This action is dictated by Quad procedure QCOA 6100-03, and is represented in the PRA model by HEP event BDGOPDG1/2ALGH--. The HEP value of 5.5E-4 was derived based on EPRI's cause based methodology supplemented with ASEP estimates for short time frame events such as this one.

The estimated time to perform the action is 10 minutes (JPM LP-003-l) with 40 minutes used as the available time window for the limiting case of an SBO with early HPCI and RCIC failures. The Phase II SAMA analysis included a reduction in the HEP value for this event from its base PRA value of 5.5E-4 to 5.5E-6.

3. Another potential option that exists at the site is to align EDG 2 to the Unit 1 buses (or EDG 1 to the Unit 2 buses). Since this action is currently not proceduralized, it is only included in the model with a relatively high failure rate of 0.9 based on engineering judgment. A reduction to the value of this HEP event (BACOP-UlU2EDGH-) to 9E-3 was also made as part of the Phase II SAMA analysis.

A factor of 100 reduction was made on three HEP values in the Phase II SAMA analysis to determine if improved reliability of existing cross-tie actions and/or expanded cross-101

tie capabilities would be cost beneficial. The averted cost risk of less than $1K indicated that such changes would not be cost beneficial.

Response 13(c):

Phase2 SAMA 5: The following statement is made in Section 4.20.6.5 of the ER, "An additional EDG cooling source may be cost beneficial for Quad Cities." However, the analysis indicates that there is no benefit (averted risk). Explain why there is no benefit, and also explain why it was believed that such an improvement would be cost beneficial when there is no benefit.'

Section 4.20.6.5 of the ER only included the statement referenced above as a prelude to the Phase II analysis to introduce the potential benefit. The ER would have been clearer if the second paragraph of Section 4.20.6.5 was not included. Based on the Phase II analysis, the potential change was determined not to be cost beneficial. The negligible benefit results from the fact that the DGCW system supports EDG 1, EDG 2, and the swing diesel, EDG 1/2, but the two Unit SBO DGs are air-cooled via a separate ventilation system that does not require DGCW. Hence, the diversification that would potentially be provided by an alternate DGCW system is already implemented at Quad with the SBO DGs.

Response 13(d):

"For several Phase 2 SAMAs (6, 10, and 14), it appears that a majority of the effort would be in writing/revising procedures and training, and engineering work. Given the additionalbenefit of these SAMAs in external events and the impact of uncertainties, the benefit of these SAMAs could be substantiallyhigher than assumed in the ER. Explain why these SAMAs would not be cost beneficial when the benefits associated with external events, and the impact of uncertaintiesare considered.-

See the revised disposition provided in Response 7(c) that includes the potential benefits for all of the Phase II SAMAs when external events and uncertainties are also considered.

102

ATTACHMENT A FIRE PRA AND USE OF QUANTITATIVE RISK ESTIMATES Overview The following summarizes the fire PRA topics where quantification of the associated figure of merit, CDF, may introduce different levels of modeling uncertainty than the internal events PRA.

The uncertainties generally reflect the following:

• lack of adequate data for initiating events

• lack of realistic fire modeling capabilities including mitigation

• lack of ability to track all cables (e.g., BOP cables)

• uncertainty in crew response, especially for control room fires, and their modeling

• limited peer reviews that examine the need for realism instead of conservatism In many cases, analysts choose to address these uncertainties by incorporating margin into the analysis (i.e., conservative assumptions).

Elements of Fire PRA Fire PRAs are useful tools to identify design or procedural items that could be clear areas of focus for improving the safety of the plant. Fire PRAs use a structure and quantification technique similar to that used in the internal events PRA.

Since less attention historically has been paid to fire PRAs, conservative modeling is common in a number of areas of the fire analysis to provide a "bounding" methodology for fires. This concept is contrary to the base internal events PRA which has had more analytical development and is judged to be closer to a realistic assessment (i.e., not conservative) of the plant.

There are a number of fire PRA topics involving technical inputs, data, and modeling that prevent the effective comparison of the calculated core damage frequency figure of merit between the internal events PRA and the fire PRA. These areas are identified as follows:

Initiating Events: The frequency of fires and their severity are generally conservatively overestimated. A revised NRC fire events database indicates the trend toward lower frequency and less severe fires.

This trend reflects the improved housekeeping, reduction in transient fire hazards, and other improved fire protection steps at utilities.

103

System Response: Fire protection measures such a sprinklers, C02, fire brigades may be given minimal (conservative) credit in their ability to limit the spread of a fire.

Cable routings are typically characterized conservatively because of the lack of data regarding the routing of cables or the lack of the analytic modeling to represent the different routings. This leads to limited credit for balance of plant systems that are extremely important in CDF mitigation.

Sequences: Sequences may subsume a number of fire scenarios to reduce the analytic burden. The subsuming of initiators and sequences is done to envelope those sequences included. This causes additional conservatism.

Fire Modeling: Fire damage and fire spread are conservatively characterized. Fire modeling presents bounding approaches regarding the fire immediate effects (e.g., all cables in a tray are always failed for a cable tray fire) and fire propagation.

HRA: There is little industry experience with crew actions under conditions of the types of fires modeled in fire PRAs. This has led to conservative characterization of crew actions in fire PRAs.

Because the CDF is strongly correlated with crew actions, this conservatism has a profound influence on the calculated fire PRA results.

Level of Detail: The fire PRAs may have reduced level of detail in the mitigation of the initiating event and consequential system damage.

Quality of Model: The peer review process for fire PRAs is less well developed than for internal events PRAs. For example, no industry standard, such as NEI 00-02, exists for the structured peer review of a fire PRA.

This may lead to less assurance of the realism of the model.

Conclusion The fire PRA may be subject to more modeling uncertainty than the internal events PRA evaluations. While the fire PRA is generally self-consistent within its calculational framework, the fire PRA does not compare well with internal events PRAs because of the number of conservatisms that have been included in the fire PRA process.

Therefore, the use of the fire PRA figure of merit as a reflection of CDF may be inappropriate. Any use of fire PRA results and insights should consider areas where the "state of the art" in fire PRAs is less evolved than other PRA topics.

Relative modeling uncertainty is expected to narrow substantially in the future as more experience is gained in the development and implementation of methods and techniques for modeling fire accident progression and the underlying data.

104