ML20086G516

Technical Evaluation Rept on Individual Plant Exam Front End Analysis

ML20086G516
Person / Time
Site:	Hatch
Issue date:	06/29/1995
From:	Darby J SCIENCE & ENGINEERING ASSOCIATES, INC.
To:	NRC
Shared Package
ML20086G518	List: ML20086G516 ML20086G521 ML20086G528
References
CON-NRC-04-91-066, CON-NRC-4-91-66 SEA-92-553-026, SEA-92-553-026-A:4, SEA-92-553-26, SEA-92-553-26-A:4, NUDOCS 9507140441
Download: ML20086G516 (57)

Text

7.

7 SEA 92-553-026-A:4 June 29,1995 f

1 Hatch 1 and 2 Technical Evaluation Report on the Individual Plant Examination .

Front End Analysis

?

NRC-04-91-066, Task 26

.i i

John L. Darby, Technical Analyst ,

Willard R. Thomas, Technical Editor ,

Science and Engineering Associates, Inc.

l 1

l t

Prepared for the Nuclear Regulatory Commission 1

,~"_m., l' s

%m om,J \

e .

TABLE OF CONTENTS

1. E X E C U TIVE S U M M A R Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 B a c k a ro u n d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Lice n s e e's I P E P roce s s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Front-End Analysis ..................................... 3 1.4 Plant Imorove ments . . . . . . . . . . . . . . . . . . . .............. 5 1.5 Conclusions ........................ .............. 6

11. RESPONSE TO WORK REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . 7 l1.1 Lice n s e e's I P E P roce s s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 l1.1.1 C o mpleten es s . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 7 l1.1.2 M ethod ology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 11.1.3 M ulti-Unit Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 l1.1.4 As-Built Status . . . . . . ........................... 8 11.1.5 Licensee Participation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 11.1.6 In-House Peer Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 11.2 Accident Seauence Delineation and System Analvsis . . . . . . . . . . . . 10 l1.2.1 Initiating Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 11.2.2 Eve n t Tre e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 l1.2.3 Syst e ms Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 11.2.4 System Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 11.3 Quantitative Proce ss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

11. 3 . 1 Quantification of Accident Sequence Frequencies . . . . . . . . . 20 11.3.2 Point Estimates and Uncertainty / Sensitivity Analyses ...... 20 l1.3.3 Use of Plant Specific Data ......................... 20 l1.3.4 U s e of G e n e ric D ata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 11.3.5 Common Cause Quantification . . . . . . . . . . . . . . . . . . . . . . 22 11.4 Core Damage Seauence Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

11. 4 . 1 Dominant Core Damage Sequences . . . . . . . . . . . . . . . . . . 23 II.4.2 Vulne rabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 11.4.3 Proposed improvements and Modifications . . . . . . . . . . . . . . 34 11.5 I nt e rf a c e i s s u e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 l1.5.1 Front-End and Back-End Interfaces . . . . . . . . . . . . . . . . . . . 35 l1.5.2 H uman Factors Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . 35 ll.6 Evaluation of Decav Heat Removal and Other Safety Issues . . . . . . . 36

11. 6 . 1 Exa mination of D H R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 11.6.2 Diverse Means of DH R . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 ll.6.3 Unique Features of DH R . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 11.6.4 Other GSI/USI's Addressed in the Submittal . . . . . . . . . . . . . 37 II.7 I nte rn al Flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

11. 7 . 1 Internal Flooding Methodology . . . . . . . . . . . . . . . . . . . . . . . 37 11.7.2 Internal Flooding Results .......................... 38 i

.l

.. ;a-: 1 111. REVIEW

SUMMARY

. AND DISCUSSION OF IPE INSIGHTS'AND -

' I M P ROVE M E NTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , , , , 39

[ IV.' D ATA S U M M ARY S H E ETS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 .

I APPENDIX A. SCOPING CALCULATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . ' 47 REFERENCES ................................,,,,,,,,,,,,,,,, gg i-

.(

q

{

l l

i i

il

LIST OF TABLES I:

-21

. Table 11-1. Plant Specific Data Table 11-2. Common Cause Factors.for 2-of 2 Components 24 31 Table 11-3 Top Five Sequences for Unit 1 32

- Table 11-4 Top Five Sequences for Unit 2

' Table 11-5. Top Five Sequences Affected by IPE Update. 33 ill

LIST OF FIGURES Figure ll-1. CDF by initiating Event Group, Unit 1 25 Figure 11-2. CDF by initiating Event Group,. Unit 2 26 Figure ll-3. CDF by Accident Class, Unit 1 27 Figure Il-4. CDF by Accident Class, Unit 2 28 Figure A-1. Decay Heat for Hatch 48 Figure A-2. Integrated Decay Heat Energy for Hatch 49 Figure A-3. Vessel Level for One SRV Open with RCIC 50 Figure A-4. Vessel Lovel for Two SRVs Open with HPCI 51 4

iv

I. EXECUTIVE

SUMMARY

t This report summarizes the results of our review of the front-end portion of the Individual Plant Examination (IPE) for Hatch 1 and 2. We completed a review of the Submittal, and made observations and identified items for further discussion with the l licensee. Requests for Additional Information (RAl) were made to the licensee for these items, and information from the licensee responses to these RAls are included  :

in this report.

l.1 Backaround (

Hatch is a dual unit site located in south Georgia on the south side of the  !

Altamaha River. Both units are Bolling Water Reactor (BWR) 4 reactors with Mark l l containments. General Electric (GE) was the Nuclear Steam System Supplier (NSSS);

Southern Nuclear Company and Bechtel were the Architect Engineer (AE) for Unit 1 and Bechtel was the AE for Unit 2. The units achieved commercial operation in 1975 and 1979, respectively. Rated power for each unit is 2436 MWt and 795 MWe (net).

Similar units in operation are: Grunswick 1 and 2, Peach Bottom 2 and 3, and Vermont Yankee.

Design features at Hatch that impact the Core Damage Frequency (CDF) relative to other BWR 4 plants are as follows. ,

Swing Diesel Generator (DG) with dedicated Plant Service Water (PSW) cooling I water pump Hardened Containment Vent  ;

Ability to flood core / containment with alternate sources such as Residual Heat .

Removal Service Water (RHRSW) and firewater 2.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> battery lifetime, procedures do not instruct operators to shed DC loads.

The impact of these design features on the CDF is as follows. The swing DG tends to l lower the CDF compared to a plant with two DGs, since it provides a third DG to backup the two dedicated DGs at the unit with the accident. The hardened containment vent tends to lower the CDF since it provides a backup for energy )

removal from containment if containment cooling systems fall. The ability to use RHRSW and firewater for core cooling tends to lower the CDF since this provides for more low pressure options for core cooling in response to transient events. The 2.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> battery lifetime tends to raise CDF since it is shorter than the battery lifetime at many BWRs and this reduces the time available to recover AC power during station blackout accident scenarios.

1

l.2 Licensee's IPE Process The IPE is a continuation of a limited scope Probabilistic Risk Analysis (PRA) performed in 1985. The freeze date for the IPE model is November 1991. The model included consideration of a number of changes to be completed after the freeze date.

Section 1.1.4 of this report tabulates these changes. Some of the most notable of these changes are as follows:

(a) installation of hardened containment vent (b) removal of common PSW discharge valve in Unit 1 (c) recovery actions to initiate purge ventilation for the control room on loss of chilled water Heating Ventilation and air conditioning (HVAC)

(d) procedure changes to allow tripping of Residual Heat Removal / Core Spray (RHR/CS) pumps in Emergency Core Cooling System (ECCS) rooms to allow continued operation of 1 pump without room cooling.

Removal of the common PSW discharge valve in unit eliminates the possibility of closure of this valve disabling all service water for unit 1. Once through ventilation for the control room provides a backup method for cooling the control room. The procedure change for tripping ECCS pumps allows for minimal core cooling capability with one ECCS pump to be maintained event if room cooling to the ECCS rooms is lost.

Utility personnel were involved in much of the total IPE effort. All IPE work was directed by the utility. The major contractor for the front-end analysis was PL&G.

Three major walkdowns were performed. A systems / general information walkdown was conducted to: (a) determine the physical arrangement and spatial dependencies for specific components, (b) determine factors affecting local operator recovery actions, and (c) verify normal system configurations and designs. An internal flooding walkdown was performed to gather information relative to: component location, flood propagation pathways, and flood sources. A back-end walkdown for the level 2 PRA analysis was also performed.

PRA studies for 4 other plants were reviewed by the utility.

Numerous levels of review of the IPE were performed. Systems engineers reviewed the system description notebooks. Operations personnel reviewed the system dependency tables and proposed recovery actions. Corporate staff reviewed the Event Sequence Diagrams (ESDs), initiating events, and key analysis assumptions. An in-house Independent Review Group (IRG) reviewed the PRA. GE reviewed selected portions of the PRA, including the success criteria. l Two criteria were used to determine a front-end vulnerability:

I (a) Any accident class with a frequency that is greater than 1E-4/ year or which contributes greater than 50% to the total CDF (b) Any accident class V sequence with a frequency that is greater than 1E-5/ year or which contributes greater than 20% to the total CDF.

I 2 l

I Based on these criteria, the licensee concluded that no vulnerabilities exist. The minor corrections reported in the January 10,1994 letter do not alter this conclusion.

The Submittal does not indicate whether or not the licensee intends to maintain l J

a "living" PRA.

1.3 Front-End Analysis The methodology chosen for the Hatch IPE front-end analysis was a Level l l PRA; the large event tree /small fault tree technique was used and quantification was  :

performed with RISKMAN software.

The IPE quantified 37 initiating events: 5 Loss of Coolant Accidents (LOCAs), i 15 plant specific support system failures, and 17 generic transients. The IPE j j

developed large systemic event trees for both frontline and support systems, to model the plant response to initiating events, initiating events were quantified using plant l j

specific data and industry data for frequent events, data from previous PRAs for infrequent events, and component failure data for plant specific initiating events.

Loss of instrument air as an initiating event was modeled as being included in closure of Main Steam isolation Valves (MSIVs) as an initiating event. Loss of HVAC for the main control room was modeled as a plant-specific initiating event.

The core was assumed to be coolable if water covered at least 2/3 of the active core over the long term. Temporary uncovery of more than 2/3 of the core was l allowed during accidents such as LOCAs.

System level success criteria were developed based on design basis information and MAAP analyses.  !

l Support system dependencies were modeled in electrical and mechanical support state event trees. Tables of inter-system dependencies were generated.

The IPE primarily used plant specific data to Bayesian update generic data for  !

both hardware failures and testing / maintenance unavailabilities.

The Multiple Greek Letter (MGL) method was used to model common cause I failures. Plant specific common cause analysis was perforrned for selected j components, Common cause failures were modeled within systems. The PLG generic data base was used to quantify common cause failures, and the data were consistent with generic data used in most IPE/PRAs.

Flood scenarios were identified based on the source of flooding, flood propagation, and key equipment locations. Scenarios with little or no impact were screened from further analysis. The remaining scenarios were quantified by combining l the flooding effects with independent failures, evidently using the internal initiating event trees. Both submergence and spray related failures were addressed in the flooding analysis.

The total CDF from internal initiating events is 2.1E-5/ year for unit 1 and 2.2E-5/ year for Unit 2. The total CDF frorn internal flooding is 3.48E-8/ year for Unit 1 and 3.3E-8/ year for Unit 2. The Submittal reported core damage sequences consistent with the reporting criteria of NUREG 1335. The top 100 systemic core damage 3

sequences for each unit were reported. These top 100 sequences constituted 57%

(55%) of the total CDF for Unit 1(2).

The top five initiating event contributors to CDF for each unit are as follows:

Unit 1 Top Initiating Events: Loss of Offsite Power 26%

Loss of 600 V AC Bus C 14 %

Loss of Feedwater 9%

Loss of Station Battery A 8%

MSIV Closure 7%

Unit 2 Top initiating Events: Loss of Offsite Power 23%

Loss of 600 V AC Bus C 13%

Loss of Feedwater 12%

MSIV Closure 10%

Loss of Station Battery A 8%

The major classes of accidents contributing to the total CDF for Unit 1 (Unit 2),

and their percent contribution are as follows:

Transient with loss of high pressure injection and 32.5% (36.8%)

failure to depressurize (IA).

Loss of containment heat removal, injection 23.0% (19.5%)

lost after containment failure (11).

Station blackout with loss of high pressure injection 16.0% (14.9%)

and failure to depressurize (IB).

Small or medium LOCA with loss of high pressure injection 14.1% (16.2%)

and failure to depressurize (IllB).

Transient with loss of low pressure injection (ID). 11.6% (9.5%)

Sections 11.4 and IV of this report provide more information about the contributors to CDF, namely: dominant hardware failures, dominant human errors including recovery failures, and the CDF by each accident class.

The results (ndicate that a transient with loss of all high pressure injection and j failure to depressurize is the dominant class of accident for overall CDF. Loss of l containment heat removal and containment failure is the second most important class  :

of accident. The station blackout class of accident and small/ medium LOCAs with  !

failure to depressurize, contribute about equally to overall CDF. Transients with loss of low pressure injection also contribute somewhat to overall CDF. 1 Anticipated Transients without Scram (ATWS) classes of accidents contribute i only about 2% to the overall CDF, and unisolated LOCAs outside containment l contribute only about 1%.

In January,1994, the licensee transmitted a letter to the NRC indicating that I two minor errors were identified in the IPE. The major source of error was a software error involving the fraction of High Pressure Core injection (HPCI) and Reactor Core l

4 l l

l l

J

. . i Isolation Cooling (RCIC) failures attributed to failure to start. With these errors corrected, the overall CDF for Unit 1 increased by 8%, and the overall CDF for Unit 2 increased by 9%. The update had no significant impact on the conclusions of the l original Submittal. For Unit 1, the CDF for accident class IA increased by 16% and I the CDF for accident class ID increased by 24%. For Unit 2, the CDF for accident j class IA increased by 17% and the CDF for. accident class ID increased by 27%. l The IPE specifically addressed loss of Decay Heat Removal (DHR) only in 1 terms of loss of the final heat sink. The Submittal contains a description of the contributors to loss of DHR for the restrictive definition of DHR used in the evaluation of DHR; these are the class 11 accidents. Loss of DHR, using the restricted definition, contributes 24% and 21% to overall CDF at Unit 1 and Unit 2, respectively. Loss of  ;

600 V AC bus C is the dominant initiating event, and failure of the hardened vant is the dominant mitigating system failure.

No other Unresolved Safety Issues / Generic Safety issues (USI/GSis) were addressed in the IPE . l Plant Damage States (PDS) were not used to bin level 1 core damage sequences for subsequent level 2 analyses; the RISKMAN software directly linked the level 1 event trees to the containment event tree.

Based on our review, the following modeling assumptions used in the IPE have an impact on the overall CDF:

\

(a) no loss of adequate Net Positive Suction Head Available (NPSHA) for ECCS pumps pulling from the suppression pool before, during, and after containment venting if suppression pool cooling is lost (b) no overheating of the running pump if only one RHR/CS pump in an ECCS room is running after loss of ECCS room cooling (c) requirement for control room cooling with either chillers or once through smoke purge ventilation.

The first assumption tends to reduce the CDF since it allows for core cooling to continue with containment venting. The second assumption tends to reduce the CDF since it allows for core cooling to continue without ventilation to the ECCS rooms. The third assumption tends to increase the CDF since it requires that ventilation for the control room be provided, or else the shutdown panel must be used.

l.4 Plant Imorovements No plant improvements, other than those considered in the model that were scheduled to be implemented after the freeze date, were discussed in the Submittal.

These improvements are listed in Section 11.1.4 of this report.

We believe that the most important of the improvements considered after the freeze date is the containment vent, since with the assumptions about Net Positive Suction Head (NPSH) margin in the IPE, containment venting preserves core cooling if suppression pool cooling is lost.

5 1

1.5 Conclusions  ;

We found no significant shortcomings in the process used for the Hatch 1 and 2 IPE front-end analysis. The overall quality of the Submittal is comparable to that of typical IPE/PRAs. We conclude that the Hatch 1 and 2 IPE provides an analysis of '

the CDF that is comparable to the analyses of typical IPE/PRAs, and that the dominant core damage sequences reflect the plant design. The Hatch 1 and 2 IPE is consistent with state-of-the-art IPE/PRA analyses.

Significant level-one IPE findings are as follows:

1

. core damage was quantified separately for both units, but the results are not significantly different for the two units e two plant specific initiating events contribute significantly to the overall i CDF: Loss of 600 V AC Bus C and Loss of Station Battery A

. loss of control room HVAC and loss of once through ventilation, requires ,

control of key systems from the shutdown panel to mitigate accidents

. battery lifetime is only 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> and procedures do not direct DC load shedding to preserve battery lifetime a room cooling is not required if only one Low Pressure Core Injection (LPCI) or core spray pump in a room is operating

. core cooling can be maintained without containment cooling if containment venting is successful; however, with no containment cooling and no containment venting, no credit was taken for continued core cooling after containment failure by overpressurization.

6

II. RESPONSE TO WORK REQUIREMENTS We reviewed the process used by the licensee with respect to: completeness and methodology; multi-unit effects and as-built, as-operated status; and licensee participation and peer review.

11.1 Licensee's IPE Process 1

We reviewed the process used by the licensee with respect to the requests of Generic Letter 88-20. [GL 88-20]

11.1.1 Completeness The Hatch IPE is a level 2 PRA. The Submittalis complete in terms of the overall requests of NUREG 1335. No obvious omissions were noted.

II.1.2 Methodology The front-end portion of the IPE is a level l PRA. The specific technique used for the level l PRA was a large event tree /small fault tree technique with support states, and it was clearly described in the Submittal.

The support state methodology and split fraction logic utilized in the large event trees was discussed in the Submittal. The development of component level system failure equations was summarized, and system descriptions were provided. Inter-system dependency tables were provided for both support and frontline systems. Data for quantification of the models were provided, including common cause and recovery data. The application of the technique for modeling internal flooding was described in the Submittal. The Submittal does not address uncertainty, although data used in the front-end analysis has uncertainty distributions. The Submittal does not contain the results of sensitivity studies, although an importance analysis for operator actions was performed. [lPE submittal, Page 3.4-6]

The PRA upon which the IPE is based was initiated in response to Generic  ;

Letter 88-20; however, a limited scope PRA had been previously performed in 1985. '

[lPE submittal, Section 2.4.2]

11.1.3 Multi-Unit Effects l

Hatch is a two unit site. The two units share the following systems and j structures: control room, intake structure, electrical power (portions), and control room j HVAC. The two units share a swing emergency diesel generator, DG 18. '

The Submittal does not discuss the dual unit core damage frequency due to common initiating events and/or failures in shared equipment. The dual unit CDF due I

to loss of offsite power is of special interest, since this initiating event can affect both units simultaneously, and the two units have the swing DG in common.

7

The licensee stated that the Hatch IPE does not address dual unit CDF, and GL 88-20 did not request dual unit core damage to be quantified. (IPE Responsesj The IPE did not credit aquipment to mitigate an accident at one unit that was required j to provide for shutdown cooling on the other unit. The licensee discusses the four initiating events common to both units, these being:

Loss of Offsite Power  ;

PSW Discharge Valve Closure intake Structure Plugging Loss of Main Control Room Cooling.

Loss of offsite power was conservatively modeled as dual unit loss of offsite power, although the value used for this initiating event includes single unit loss of offsite power. Neither PSW Discharge Valve Closure nor Loss of Main Control Room Cooling contribute significantly to the overall CDF. Intake Structure Plugging results in the loss of PSW and RHRSW for both units; systems available to mitigate the loss of the intake structure (eg, RCIC and SRVs) while recovery of the intake structure is attempted, are independent systems at the two units. With offsite power available, the l electrical systems at the two units are essentially independent. If offsite power is lost j to both units and is not recovered following plugging of the intake structure, core  !

damage occurs at both units.

The IPE did not quantify dual unit CDF; however, GL 88-20 does not request that the IPEs quantify dual unit CDF. Some licensees have the dual unit CDF readily available as a result of the analyses performed for the IPE, and have provided this information in response to RAl.

II.1.4 As-Built Status Three major walkdowns were performed. [lPE submittal, Section 2.4.4] A .

systems / general information walkdown was conducted to: (a) determine the physical j arrangement and spatial dependencies for specific components, (b) determine factors l

affecting local operator recovery actions, and (c) verify normal system configurations i and designs. An internal flooding walkdown was performed to gather information l relative to: component location, flood propagation pathways, and flood sources. A l

back-end walkdown for the level 2 PRA analysis was also performed.  :

Major documentation used in the IPE included: the UFSAR, technical specifications, system evaluation documents, station blackout coping study, plant procedures, drawings and diagrams, and plant reliability data. Other IPE/PRA studies reviewed were:

Oconee PRA i IREP PRA for Browns Ferry Limerick PRA Shoreham PRA.

8

Also referenced are reports from the "EPRI PRA Repository", for other plants. [lPE submittal, Section 2.4.1]

The freeze date for the IPE model was November,1991. The IPE modeled a few modifications to be installed after the freeze date, and a few changes to procedures after the freeze date, these being: [lPE submittal, Sections 2.4.3, 1.4.1, andu.2)

(a) installation of hardened containment vent (b) removal of common PSW discharge valve in Unit 1 (c) changes in procedures and modifications to ducting for the control building to allow continued operation of electrical equipment in the control building following loss of HVAC (d) recovery action to use smoke purge ventilation upon loss of control room chillers was proceduralized (e) modifications to intake structure ventilation system (f) procedure changes to allow tripping of RHR/CS pumps in ECCS rooms to allow continued operation of 1 pump without room cooling (g) procedure changes to allow cross connection of PSW cooling water to RHRSW pump motors (h) modification to allow swing chiller compressor for control room HVAC to be powered by either division of electrical power (1) modifications to meet the Station Blackout Rule (SBO) rule including:

replace station service battery chargers, and enhance procedures dealing with loss of ventilation.

11.1.5 Licensee Participation One engineer was dedicated to the overall PRA effort. Another engineer assisted in the level 2 analysis for Hatch, Farley, and Vogtle. Other engineers were i

involved during various phases of the IPE effort. Site engineers reviewed the PRA l

system description notebooks. Operations personnel reviewed the PRA systems dependency tables. The Submittal states that the total licensee effort was 12.4 l

person years, and the total contractor effort was 4.3 person-years.

The major contractor for the front-end analysis was PL&G. GE provided input into the original models and analyses, and performed reviews. Bechtel and Southern Nuclear Company (SNC) provided AE related support as required.

We could find no discussion in the Submittal of the utility plans to maintain a "living" PRA.

11.1.6 In-House Peer Review Section 5.3 of the Submittal summarizes the IPE review process. Systems engineers reviewed the system description notebooks. Operations personnel reviewed the system dependency tables and proposed recovery actions. Corporate staff 9

l l

l reviewed the Event Sequence Diagrams (ESDs), initiating events, and key analysis assumptions. An in-house independent Review Group (IRG) reviewed the PRA. GE reviewed selected portions of the PRA, including the success criteria. The review  ;

process was consistent with that used in other IPEs.

II.2 Accident Seauence Delineatio'n and Svstem Analvsis This section of the report documents our review of both the accident sequence delineation and the evaluation of system performance and system dependencies -

provided in the Submittal.

II.2.1 Initiating Events  ;

The list of initiating events includes quite a few plant specific initiating events, such as: intake structure plugging, loss of drywell cooling, and loss of main control i room cooling, in addition to the expected plant specific initiating events such as loss of .

i PSW, loss of ac buses, and loss of de buses. [lPE submittal, Section 3.1.1.4)

Initiating events were quantified using plant specific data and industry data for frequent events, data from previous PRAs for infrequent events, and component failure data for plant specific initiating events.

The list of initiating events does not consider recirculation pump seal LOCAs, since the Submittalindicates that this initiating event was removed based on a GE l comment that seal LOCAs are not large enough to be classified as LOCAs. [lPE submittal, Section 5.3] Many BWR PRA/IPEs consider seal LOCAs as an initiating event. For example, the Peach Bottom PRA included a small-small LOCA as an initiating event to model seal LOCAs, and this LOCA had a frequency a factor of 10 l higher than the small LOCA. [NUREG/CR 4550, Peach Bottom) Such studies have  :

assumed that the feedwater system can compensate for the seal LOCA, assuming the f leak is sufficiently small so that the vessel pressure does not drop to where the MSIVs isolate and cause loss of turbine-driven main feedwater, and in effect the event can be i

mitigated sirnilar to a transient with feedwater initially available. The NUREG/CR 4550 study for Peach Bottom concluded that a small-small LOCA contributed a negligible amount to the overall CDF. [NUREG/CR 4550, Peach Bottom, Table 5-3] Had the i Hatch IPE quantified seal LOCA initiating events as small-small LOCAs, the change in  ;

overall CDF would have been small.

The Submittal does not provide the equivalent break diameters for steam and water LOCAs. The licensee provided these ranges for small and medium LOCAs.

[lPE Responses) A small LOCA was defined as within the makeup capacity of RCIC and was less than 0.007 sq ft for liquid line breaks and less than 0.03 sq ft for steam line breaks. A medium LOCA was defined as requiring either high pressure injection (RCIC insufficient for medium LOCA) or depressurization for use of low pressure injection and was up to 0.3 sq ft for liquid line breaks and up to 0.2 sq ft for steam line l breaks. These size ranges are comparable to ranges used in other IPE/PRAs; however, it is noted that the upper limit for a small LOCA in a steam line is smaller ,

l 10

than the upper limit used in other IPE/PRAs. For example, NUREG/CR 4550 for  !

Peach Bottom uses an upper limit of 0.05 sq ft for a small LOCA in a steam line.  !

[NUREG/CR 4550, Peach Bottom, Page 4.3-4]

Table 3.1-3 of the Submittalindicates that one stuck open SRV is modeled as a >

medium LOCA. [lPE submittal, Section 3.1.1.3 and Table 3.1-3]. Other studies have modeled a stuck open SRV as a small LOCA and credited RCIC as sufficient to mitigate a transient with one stuck open SRV. [NUREG/CR 4550, Peach Bottom] [lPE, Browns Ferry] [lPE, Quad Cities] We performed a scoping calculation to evaluate the ability of RCIC to provide core cooling with one SRV failed open for Hatch. Appendix  !

A summarizes this evaluation. Our evaluation indicates that RCIC can meet the core l cooling success criteria stated in Section 3.0 of the Submittal, that being no raore than i 1/3 of the core uncovered over the long term. We conclude that the IPE is  ;

conservative in the treatment of a stuck open SRV as a medium LOCA.

Many IPEs distinguish between loss of offsite power at all units and at an '

individual unit, since cross-tie of power between units is modeled. Hatch has two switchyards, one at 500 KV and one at 230 KV; offsite power at shutdown is 230 KV, but an auto transformer is available to transform 500 KV to 230 KV. Thus, it possible that if the 230 KV switchyard alone is lost, offsite power from the 500 KV switchyard i;

can be supplied.

The licensee stated that the frequency assigned to the loss of offsite power i initiating event was generated based on a per-unit frequency, but that the event was modeled as loss of offsite power to both units. [lPE Responses]

The frequency assigned to the initiating event, inadvertent opening of an SRV, is 0.018/ year, which is less than the frequency used in some other studies. [lPE i submittal, Section 1.4.1 and Table 3.1.1] For example: the Peach Bottom PRA used 0.19/ year; the Browns Ferry IPE used 0.04/ year; the Cooper IPE used 0.09/ year; and the Grand Gulf PRA used 0.14/ year. [NUREG/CR 4550, Peach Bottom] [lPE, Browns Ferry] [lPE, Cooper] [NUREG/CR 4550, Grand Gulf]

The frequency of an interfacing systems LOCA,8.4E-8/ year, is low compared  ;

to values used in some other IPE/PRAs. The CDF for Class V accidents, unisolated I

LOCAs outs _ide containment, equals the sum of the initiating event frequencies for interfacing systems LOCAs and steam /feedwater breaks outside containment that are  ;

not isolated. This implies that the frequency assigned to an interfacing systems LOCA includes failure to isolate the LOCA. Also, consideration of the best estimate failure i pressure at which components exposed to greater than design pressure actually fail j can lower the frequency of the interfacing systems LOCA by a factor of 100 or more.  ;

The frequency assigned to the interfacing systems LOCA initiating event is l comparable to that used in other IPEs that have credited operator action to isolate the l LOCA and that have used best estimate analyses to quantify failure of piping exposed l to beyond design basis pressure. l Spurious actuation of ECCS is considered as part of the reactor scram initiating  ;

event. [lPE submittal, Table 3.1-2] I The initiating event for a LOCA outside containment includes failure to isolate (event ULOCA). [lPE submittal, Table 3.1-1] This includes such events as 11

steam and feedwater line breaks outside containment with closure of required isolation valves such as MSIVs and check valves. The total frequency for this event is 8.7E-8/ year.

The !PE did not model loss of Reactor Building Component Cooling Water (RBCCW) ns an initiating event. RBCCW cools the recirculation pump motors and seals, and the CRD pumps according to the UFSAR. [UFSAR, Section 9.2.2] Without RBCCW the plant would trip and a seal LOCA could eventually occur. The licensee discussed the systems cooled by RBCCW, and the impact of loss of cooling to these systems. (IPE Responses] The components important for mitigation of accidents that are cooled by RBCCW are the recirculation pump motor bearings and seal coolers and the CRD pump seals. Loss of cooling to the recirculation pumps and seals could result in some leakage, but the amount of leakage is expected to be small enough to not be considered as a small LOCA and would be well within the capability of RCIC to makeup. Loss of RBCCW will result in loss of CRD when the CRD is supplied from ,

the condensate system; however, loss of RBCCW does not directly cause loss of feedwater, and sequences caused by loss of RBCCW coupled with random loss of feedwater are not risk significant.

Loss of instrument air was assumed to be equivalent to closure of MSIVs. [lPE submittal, Section 3.1.1.8] However, the table of system dependencies indicates that air is important for the hardened vent and control building ventilation in addition to the MSIVs. [lPE submittal, Table, 3.2.3.1]

The licensee stated that loss of instrument air does not impact HPCI, RCIC, PSW, the SRVs, LPCI, RHR, or core spray, and that condensate can be still used for injection. [lPE Responses] Thus, the systems available to respond to closure of the MSIVs and loss of instrument air are essentially the same. Closure of the MSIVs has a frequency about a factor of 10 greater than loss of instrument air. For these reasons, loss of instrument air was assumed to be included in closure of the MSIVs initiating event.

II.2.2 Event Trees

~

Each accident initiating event was included in an appropriate class of initiating events, and each class of initiating events had a corresponding event tree. The success criteria for the combinations of mitigating systems required to prevent core damage were described for each class of initiating events and were consistent with the criWria for core damage. All functions or systems important to the accident sequences were reflected on the event trees. The interface among the events in the event trees and the corresponding mitigating systems was clearly indicated. The event trees properly accounted for: time ordered response, system level dependencies, sequence specific effects on system operability- such as environmental conditions, and high level operator actions as appropriate. We conclude that the event tree models in the IPE are consistent with event trees used in typical IPE/PRAs.

Core damana was assumed when substantial parts of the core (more than a few rods) exceed 2200 F. The success criteria for core damage is uncovery of no i

1 12

more than 1/3 of the core for an extended time; during depressurization level may temporarily drop below 2/3 core height but core damage will not occur if level is quickly restored. The basis for this collapsed level success criteria for core cooling is evidently from the large LOCA situation where the long term steady state collapsed level is at 2/3 core height. The 2/3 coverage criteria is less optimistic with respect to other BWR PRAs which have assumed that long term collapsed level can be as low as a few feet above the bottom of the core. [NUREG/CR 4550, Grand Gulf)

Therefore, we conclude that the criteria used for core damage in the Hatch IPE are consistent with numerous other BWR PRA/IPEs.

The Submittal summarizes the success criteria for different classes of accidents.

[lPE submittal, Tables 3.1-3 through 3.1-6) Table 3.1-3 is the success criteria for all accidents with successful scram, including LOCAs and transients. The licensee clarified the success criteria used in the IPE for specific situations. [lPE Responses]

For pressure control, the success criteria reference Table 3.1-6 of the Submittal; however, Table 3.1-6 does not specify the number of Safety Relief Valves (SRV) required to open in response to a small LOCA if the Power Conversion System (PCS)is not available. The licensee states that for a small LOCA initiating event, it is likely that no SRVs will open since the reactor scrams in response to the LOCA but the turbine does not trip on reactor scram and the MSIVs do not close on high drywell pressure. If steam pressure remains high, the operator may not trip the turbine until the generator output is significantly reduced, and turbine bypass to the condenser is available after turbine trip. If turbine bypass is not available, pressure willincrease until one or more SRVs open. The pressure increase is expected to be much smaller that for that due to closure of MSIVs as an initiating event, since for a small LOCA reactor scram is expected to precede the interruption of steam flow. For a small LOCA with bypass unavailable, one SRV can provide sufficient relief, and this was the model used in the IPE for response to a small LOCA wi e ypass unavailable.

For high pressure inventory control, the feedwatei system is credited. The feedwater pumps at Hatch are steam driven, and are not available if the MSIVs close.

[UFSAR, Chapter 10) Thus, the feedwater system is not available for large and medium break LOCAs due to closure of the MSIVs on low pressure or low level.

[UFSAR, Section 7.3.2.2.3] Also, for the larger of the small LOCA steam breaks, feedwater may be lost following reactor trip on high drywell pressure and MSIV closure on low pressure. [NEDO 24708A) The licensee stated that the model for a small LOCA credits use of feedwater if the operator places the mode switch in the shutdown position, thereby preventing closure of MSIVs on low pressure at 850 psig. Failure of operator action to place the mode switch into the shutdown position was included in the model.

Neither condensate injection not RHRSW crosstie were credited for a large LOCA Condensate injects into a feedwater line and injected water would flow out the break and never reach the core; there is insufficient time to align and use RHRSW injection to RHR to prevent excessive clad heatup after the blowdown when the core is uncovered.

s 13 1

The licensee stated that one RHR pump and one RHRSW pump operating in one RHR loop is sufficient for all non-ATWS accident sequences. For an ATWS in which shutdown with Standby Liquid Control (SBLC) is successful, one RHR pump and one RHRSW pump in each of the two loops is required.

The model for ATWS assumes that without boration with SBLC and without the PCS, an ATWS sequence can be mitigated.and that sufficient Suppression Pool (SP) cooling can be provided with 1 RHR pump in both SP cooling loops. [lPE submittal, Figure 3.1-8 Sheet 4 of 5 and Table 3.1-4) We questioned the ability of one RHR pump in both loops to mitigate an ATWS without boration using SBLC. The licensee stated that the IPE assumed that successful level control would maintain reactor power near 8% of full power; however, recent analysis indicates that the power level may be closer to 30%. [lPE Responses] At 30% power, it is unlikely that full operation of all pumps in both loops could keep suppression pool temperature below the NPSH limits for the RHR pumps. The effect of the higher power level in rendering SP 1 cooling inadequate was estimated by calculating the increase in CDF if all ATWS events were associated with closure of the MSIVs; also, core damage was assumed if shutdown with SBLC failed. This sensitivity analysis resulted in an increase in total CDF of less than 5%. The response indicates that the success criteria used in the IPE for SP cooling following an ATWS without shutdown with SBLC are possibly incorrect, but the impact on overall CDF is small.

The Submittal states that no credit was taken for long term core cooling with only the Control Rod Drive (CRD) system, but that CRD was credited for extending the time available to recover other injection systems. [lPE submittal, Section 3.1.2)

The success criteria table for a large LOCA indicates that either 1 LPCI or 1 CS pump can successfully mitigate a large LOCA. Although this success criteria is consistent with that used in other BWR IPE/PRAs, it deviates significantly from the success criteria indicated in Table 6.3.7 of the UFSAR. For example, the UFSAR success criteria always require CS. We have seen summaries of best-estimate GE analyses for BWR 4 plants in which spray over the top 1/3 of the uncovered core is not required; [TER, Quad Cities) however, these analyses require either one CS pump or two LPCl_ pumps, not one.

The licensee stated that the success criteria for a large LOCA were based on GE's SAFER /GESTR methodology; the code and its application methodology have been approved by the NRC. [lPE Responses][NSAC-131) [NEDO-30936P, Proprietary]

This analysis indicates that with best estimate assumptions, one core spray pump or one LPCI pump can maintain PCT below 1600 F. The analysis does include delayed neutron fission energy as a heat source. The analysis does not specifically address leakage from jet pump joints, but this loss is accounted for by reducing best estimate LPCI and core spray flow rates by 10% in the analysis.

Some of the IPEs that we have reviewed for BWR 4 plants have not required I closure of the recirculation discharge valve in the intact recirculation loop in response to a large LOCA; if this valve is not closed, LPCI is lost out the break. Closure of this valve was modeled in the Hatch IPE as required to mitigate a large LOCA. [lPE i Responses) 14 I l

l

l We reviewed the event trees. The descriptions describing the events in the trees and the assumptions related to the events are complete. Both LOCAs and l transients are modeled in the same trees, which increased the difficulty of our review.

The IPE credits the RHR heat exchangers for containment heat removal. Some plants have experienced degradation of heat exchanger efficiencies, as indicated by l

testing performed in response to Generic Letter 89-13. The licensee stated that evidence indicates that the effectiveness of the RHR heat exchangers has not degraded below the design basis value. [lPE Responses] Furthermore, the IPE models assumed worst case heat sink temperatures of 95 F for service water.

Many BWR 4 plants, for example Quad Cities, have little NPSH margin during a design basis accident even with suppression pool cooling, as indicated by the UFSAR analysis for long term mitigation of a Design Basis Accident (DBA). Typically, BWR 4 plants may be adversely affected with NPSH margin, defined as NPSHA - NPSHR, if the suppression pool heats up to over about 200 F. The UFSAR for Hatch discusses l l

NPSH margin for the ECCS pumps at the peak suppression pool temperature of 209.8 F. [UFSAR, Section 6.3.2.14] The temperature was calculated with the minimal acceptable containment cooling using 1 RHR heat exchanger. {UFSAR, Figure 6.2-29]

The NPSH margin for the RHR pumps was calculated as 4 feet, assuming a conservative containment pressure of 0 psig; the NPSH margin for the CS pumps was calculated as 1 foot. [UFSAR, Section 6.3.2-14]  ;

The Submittalindicates that at SP temperature of 250 F and pressure of 10 psig, NPSH margin is an item of potential significance for the RHR and CS pumps.

[lPE submittal, Section 3.1.2.1 Part B-6] The Submittal discuses loss of NPSH margin, but concludes: )

l "For the level i PRA, the potential for core damage due to loss of injection l

during the venting process was assessed as insignificant compared to other core damage contributors for these scenarios." [lPE submittal, Section 3.1.2.1 Part B-6]

This is a major assumption in the IPE. This assumption means that complete loss of suppression pool cooling does not result in loss of ECCS from the suppression pool due to inadequate NPSH margin prior to, during, or following venting of containment at between 50 and 60 psig. Furthermore, this assumption drives an important change made to procedures as a result of the IPE, that being to turn off all but one RHR/CS pump in each ECCS room if cooling to the room is lost to prevent overheating of components in the room. This procedure can cause termination of suppression pool cooling, by turning off RHR pumps.

The Submittalindicates that it requires about 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> to heat up the SP to 250 F, implying that a long time is available for restoration of SP cooling prior to loss of NPSH margin. [lPE subm'ttal, Section 3.1.2.1] We have made conservative calculations of time to loss of NPSH margin for Quad Cities, and concluded that about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> are available. [TER, Quad Cities] Therefore, we agree with the implication that a long time is available for recovery of SP cooling prior to potential loss of core 15

cooling due to cavitation of ECCS pumps pulling from the SP. Given this long time available for recovery of SP cooling, it is not clear how much redu': tion in CDF is due to the assumption in the IPE that NPSH margin is not lost if SP cooiing is not available and the containment is vented. The IPE did not quantify the impact of the assumption that NPSH margin is not lost if the containment is vented on reducing overall CDF, and the impact of recovery of SP cooling on reducing overall CDF.

Assuming that the ECCS pumps do not cavitate prior to, during, and after containment venting, we could find no discussion on the Submittal for makeup to the SP to compensate for inventory boiled off over time.

The licensee provided information supporting the models used for core cooling with loss of containment cooling, but with successful venting. [lPE Responses) MAAP calculations were performed to evaluate the NPSH considerations for RHR and core spray pumps using the suppression pool to cool the core with loss of containment heat removal. These analyses indicated that when saturation is reached in the suppression pool, the NPSH margin for the core spray pumps and the RHR pumps in unit 2 would be maintained, but that the NPSH margin for the core spray pumps and the RHR pumps in unit 1 would be lost; for example, the RHR pumps in unit 1 would be operating about 7 feet below the required NPSHA. NPSH margin for RHR and LPCI pumps in unit 1 is lost prior to containment venting but this does not necessarily result in failure of the pumps; the IPE assumed that low pressure makeup was not lost prior to and after containment venting due to the numerous options available for low pressure makeup even if the RHR and core spray pumps did fail, and the time available to implement these options before core uncovery. The IPE assumed that makeup to the suppression pool after venting was not a limiting factor due to the time available to use numerous options for makeup.

The Submittal indicates that if SP cooling is lost and the containment is not vented, then containment failure by overpressurization results in loss of injection from all sources, including ECCS and alternate injection with such systems as RHRSW cross-tie or firewater, due to Equipment Qualification (EO) failures in the reactor building. [lPE submittal, Section 1.4.2 page 1.4.4 and Section 3.1.2.1 page 3.1-8)

However, in.the event tree descriptions, it is indicated that low pressure ECCS can remain operable after containment failure by overpressure. [lPE submittal, Figure 3.1.3 Sheet 3 of 3 Event 9) The licensee was requested to clarify this aspect of the modeling. The licensee stated that no credit for core cooling after containment failure was actually credited in the quantification, although the structure of the event sequence diagrams allows for such an option. [lPE Responses)

In summary, the IPE model assumes that the ECCS pumps will not be lost due to containment venting. The licensee stated that even if the pumps are lost due to inadequate NPSH margin, the number of options available for core cooling and the time available for implementing these options can prevent core damage even though i these options were not explicitly modeled in the IPE. The IPE did not credit any core  ;

cooling if containment cooling was lost and containment venting failed. '

)

Other IPEs for BWR 4 plants have credited use of non-safety systems for core / containment flooding, and have indicated that such credit lowers the overall CDF 16 i

1

. . l I

by as much as a factor of 5. [TER, Quad Cities][TER, Dresden] The IPE for Hatch credits the use of alternate injection systems, specifically: RHRSW, fire pumps, ECCS keep full system, SBLC, and condensate transfer pumps, for low pressure makeup if l

LPCI and CS are lost. [lPE submittal, Figure 3.1-2 Sheet 12 of 14] Operator action is l

required to use these alternate injection systems.

The IPE did credit switchover of RCIC or HPCI from the SP to the CST if containment pressure is too high for RCIC or containment temperature is too high for HPCI. [lPE submittal, Section 3.1.2.1 and Figure 3.1.3 Sheet 3 of 3] Operator action j

is required for such switchovers. '

The Submittalindicates that HPCI can mitigate a transient with two SRVs stuck open. [lPE submittal, Figure 3.1-4 Sheet 4 of 4] We performed an analysis of such a '

transient, similar to the one we performed for RCIC with one SRV open as previously j

discussed in this report. A summary of our calculation is provided in Appendix A of this report. Our results support the statement in the Submittal. l The event trees model drywell cooling and control room HVAC. Consideration I of these systems is a positive aspect of the models. Also, the event trees model the closure of the target rock SRVs at high containment pressure, thereby negating use of  ;

low pressure injection systems. This recognition and consideration of the special  !

characteristics of the target rock SRVs, namely closing at high backpressure, is commendable.

The event trees model RBCCW as required to support operation of CRD. This indicates close attention to requirements for pump cooling in the event tree models. At Hatch feedwater and condensate pumps are cooled by PSW.

The IPE does not modelloss of cooling for recirculation pump seals leading to a seal LOCA. For Hatch, which has RCIC instead of an isolation condenser, this is not a major shortcoming since all methods for core cooling involve makeup directly to the vessel.

Credit is taken for the smoke purge mode of control room ventilation, event KMCR) following loss of control room HVAC with chillers. [lPE submittal, Figure 3.1-12 Sheet 2 of 3] If event KMCR fails, then credit for transfer of control of key systems to the shutdown panelis required, event KRSDP. These recovery events require operator action. It is not clear that the shutdown panel has the capability to mitigate LOCAs, in contrast to transients. The licensee stated that equipment to mitigate a LOCA can be controlled from the shutdown panel. [lPE Responses]

11.2.3 Systems Analysis f

System descriptions are included in Section 3.2 of the Submittal. We reviewed the UFSAR to more fully understand the system capabilities and inter-dependencies.

l To document our overall review, we have included comments on the system descriptions related to where we had to use supplementary information from the UFSAR to understand the system characteristics. Our comments on the system descriptions in the Submittal are as follows.

17

The system description for offsite power does lot discuss the ability to provide offsite power from the 500 K sources if the 230 K sounes are lost; although all offsite power at shutdown is 230 K, there is an auto transformer that connects the 500 K and 230 K switchyards. The description does not point out that the normal supply of offsite power to the 1E buses is from the shutdown offsite power supply from the startup transformers, not from the plant generator auxiliary transformers as it is as some plants, which means that supply of offsite power does not have to transfer following plant trip. The description does not clarify the alignment of the swing DG,18. The UFSAR indicates that the swing DG automatically connects to the unit with the '

accident, following a design basis accident. The description does not discuss the requirements for DGs to both mitigate an accident at one unit and to maintain the other unit in shutdown. We did find a discussion in Table 3.4-3 of the Submittal, which discusses differences Lueen the two units, stating that the swing DG 1B normally aligns to Unit 1, and will stay aligned to Unit 1, unless a LOCA signal is generated for Unit 2. The system description states that the battery lifetime was conservatively taken as 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />, and that the Hatch procedures do nat direct operators to shed DC loads to preserve battery lifetime.

The licensee stated that crosstie of power between the two units' 1E power supplies was not modeled. (IPE Responses]

The system description for nitrogen supply in the drywell points that the SRVs require nitrogen, but that accumulators are provided. Over the long term, the accumulators willleak resulting in inability to maintain the SRVs open without nitrogen supply. The dependency table does indicate that loss of nitrogen to the SRVs does result in long term loss of the SRVs. (IPE submittal, Table 3.2.3.1]

The inclusion of system descriptions for control room HVAC and for drywell cooling is notable; most BWR IPE Submittals have not discussed these systems.

Section 3.3.9 of the Submittat discusses the treatment of HVAC. This is a strength of the Submittal; other IPE Submittals have provided little information on the treatment of HVAC. HVAC was assumed to be required for the following areas:

HPCI. Pump Room RHR/CS Pump Rooms (unless only one pump operating)

LPCI inverter Room (except for large LOCA)

Control Room (chillers or smoke purge)

DG Rooms.

HVAC was assumed to not be required for the following areas:

RClO Pump Room RHR/CS Pump Rooms (if operator stops all but 1 pump)

CRD Pump Room {

Battery Rooms LPCI inverter Room (for large LOCA)

Cable Spreading Room 18 I

Intake Structure (PSW and RHRSW pump rooms) 4160 V AC Switchgear Rooms.

The licensee addressed the contention that without room cooling, operation of only one RHR/ Core spray pump will not result in overheating of the pump. [lPE Responses) The licensee summarized a room heatup analysis that was performed, which concluded that the room temperature would be below Equipment Qualification (EO) limits.

The licensee stated that the heat loads in the 4160 V AC switchgear rooms were assessed to be sufficiently low so that no HVAC is required during the mission time. [lPE Responses] The licensee stated that recovery actions, such as operator actions to open doors, are not required to prevent loss of switchgear following loss of HVAC. [lPE Responses] This clarifies the statement on page 3.3-54 of the Submittal which indicates that opening doors for compensatory ventilation is possible; such actions are not required in the IPE model.

11.2.4 System Dependencies The Submittal provided tables that indicate the dependency relationships among the systems, both frontline and support important asymmetries in train-level system dependencies were indicated. The following types of dependencies were considered:

shared component, instrumentation and control, isolation, motive power, direct l 1

equipment cooling, area HVAC, operator actions, and environmental and phenomenological effects. l Table 3.2.3.1 of the Submittal summarizes dependencies among systems at Unit 1. Table 3.2.3.2 of the Submittal summarizes dependencies among systems at  !

Unit 2. It is commendable that separate dependency tables for both units were prepared and provided in the Submittal. i The dependency tables have notes to explain the dependencies. The degree of l dependence- complete, partial, delayed- is clearly indicated and described. j i

lt shuuld be noted that the dependence of recirculation pump seal cooling on RBCCW is noted in the dependency tables; [lPE submittal, note 42 page 31 of 35 for Table 3.2.3.1) but as discussed previously in this report, loss of seal cooling was not considered in the mitigation models for transient sequences. l The Submittal clearly points out the cooling requirements for: RHR pumps, CRD pumps, RBCCW pumps, and condensate /feedwater pumps. However, the system dependency tables and the system descriptions in the Submittal do not discuss pump / driver cooling for the HPCI, RCIC, or CS pumps. Based on dependency information provided in the Submittal, we infer that the HPCI, CS, and RCIC pumps are self-cooled. The UFSAR states that the RCIC pump is self cooled and implies that j the HPCI pump is self cooled. [UFSAR, Sections 5.5.6.3 and 6.3.2.2.1) The UFSAR does not discuss cooling for the CS pumps. [UFSAR, Section 6.3.2.2.3) 19

II.3 Quantitative Process This section of the report summarizes our review of the process by which the ,

IPE quantified core damage accident sequences. It also summarizes our review of the data base, including consideration given to plant specific data, in the IPE. The uncertainty and/or sensitivity analyses that were performed were also reviewed.

II.3.1 Quantification of Accident Sequence Frequencies The Hatch IPE used the large event tree /small f ault tree model for quantifying core damage. Support states were modeled with support system event trees. The event trees are systemic. Fault trees were used to develop component level failures for event tree events. [lPE submittal, Section 3.2.2) A mission time of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> was used. Quantification of linked event trees was accomplished with Riskman software.

The analysis used a sequence truncation value of 1E-12. [lPE submittal, Section 3.4.1]

11.3.2 Point Estimates and Uncertainty / Sensitivity Analyses Mean values were used for point estimate failure frequencies and probabilities.

Although the data base consists of probability of frequency distributions, no uncertainty analyses and no sensitivity analyses are described in the Submittal, aside from sensitivity analyses discussed for operator errors.

11.3.3 Use of Plant Specific Data The IPE used plant specific data to Bayesian update generic data for selected components. [lPE submittal, Section 3.3.2.1] Plant specific data for component failures were taken from plant data covering the time period January 1,1984 through December 31,1990.

The IPE used plant specific data for system unavailability for testing and maintenance to Bayesian update generic testing and maintenance data. Plant specific data for testing and maintenance were taken from plant data covering the time period January 1,1986 through December 31,1990. l We noted that different periods of time were used for plant specific data for component failures and for testing / maintenance. The licensee stated that the information for component failures back to 1984 was readily available so it was used; however, maintenance data had to be copied from microfilm, a labor-intensive task, so -l the start time was selected to be 1986, to cover an suitable time period. [lPE Responses)

Bayesian updating was not done for unavailability due to periodic surveillance testing. [IPE Responses] Similarly, Bayesian updating was not done for preventative maintenance activities scheduled at power. Bayeslan updating was done for corrective maintenance activities performed at power.

20

We performed a spot check of the plant specific data for component failures. i The results of this check are summarized in Table 11-1 of this report.  !

l Table 11-1. Plant Specific Data l l

Component and Hatch Value UM' NUREG/CR 4550  ;

Failure Mode Submittal Table 3.3-4 Value ou2)

Peach Bottom Table 4.9-1 1 Diesel Generator Fail to Start 1.3E-2/D 3.0E-3/D Diesel Generator Fall to Run 7.0E-3/H (First Hour) 2.0E-3/H 1.5E-3/H (After First Hour)

HPCI Turbine Fail to Start 2.7E-2/D 3E-2/D HPCI Turbine Fall to Run 1.6E-2/H SE-3/H LPCI Pump Fall to Start 9.3E-4/D 3E-3/D LPCI Pump Fall to Run 9.2E-6/H 3E-5/H Core Spray Pump Fall to 2.0E-3/D 3E-3/D ,

Start Core Spray Pump Fail to 3.4 E-5/H 3E-5/H Run MOV Fall to Change State 2.4E-3/D 3E-3/D (Open/Close)  !

AOV Fail to Change State 2.0E-3/D 1 E-3/D (Open/Close)

RCIC Pump Fail to Start 2.5E-2/D 3E-2/D RCIC Pump Fail to Run 4.0E-3/H SE-3/H (1) D is per demand; these values are probabilities.

(2) H is per hour; these values are frequencies.

Based on the data in Table Il-1 of this report, the plant specific component failure data are comparable to those used in typical IPE/PRAs.

The Submittal stated in Section 1.4.1 that the SRVs at Hatch have a high probability of failure, yet Table 3.3-4 of the Submittal which summarizes plant specific component failure data does not include the SRVs. Table 3.3-1 of the Submittal, I generic failure data used for components, includes the SRVs, and assigns a mean failure-to-open probability of 6.3E-2/ demand. This generic value is higher than that i used in other IPEs; Quad Cities used 3E-4/ demand and Browns Ferry used 4E-3/ demand, [TER, Quad Cities] and it may be appropriate for Hatch; however, the 21 l l

Submittal stated that failures of the target rock SRVs are particular items of significance at Hatch.

The licensee stated that plant specific data for Hatch indicates a failure probability of 0.04 for the SRVs, and the less optimistic generic data, which includes Hatch data, were used. Also, the plant has implemented a modification on both units to improve the reliability of the SRVs during.a pressure transient, consisting of a direct electrical signal sent to each SRV based on reactor pressure.

11.3.4 Use of Generic Data The generic data used for component failures are listed in Table 3.3-1 of the Submittal. This is the PL&G generic data base, since the Submittal states that the PL&G generic data base was used for generic data for component failures and for testing / maintenance. [lPE submittal, Section 3.3.1.2] This generic data is comparable to that used in other IPE/PRAs.

The IPE used generic data for recovery of offsite power. [lPE submittal, Section 3.3.3.2) The Submittal states that the probability of recovery of offsite power within 30 minutes is 0.30 and that the probability of recovery of offsite power within 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> is 0.86. The model for recovery of offsite power in the Hatch IPE is consistent with typical data for recovery of offsite power.

II.3.5 Common Cause Quantification The MGL method was used to model common cause failures. [lPE submittal. j Section 3.3.4) The methodology of NUREG/CR-4780 was used for plant specific  ;

analysis of common cause failures. A plant specific common cause analysis was  ;

performed for selected components, including: DGs, circuit breakers, RHR pumps, CS l pumps, RHRSW pumps, MOVs, and SRVs. [lPE submittal, Table 3.3-17]

Components modeled with generic common cause failure are listed in Table 3.3-18 of I the Submittal.

The licensee provided additionalinformation related to common cause failures.

[lPE Responses] The PLG data base was used to quantify common cause failures in the IPE. The common cause analysis was done assuming two room coolers in the plant-specific population; this resulted in a plant-specific beta factor. Common cause failure was actually modeled among all four coolers in the RHR/ core spray pump rooms; the plant specific beta factor was used but generic gamma and delta factors ,

were used. No common cause failures among HPCI and RCIC were considered; the l HPCI system is an order of magnitude larger in capacity than the RCIC system, and the parts between the two systems are not interchangeable. Industry experience validates the assumption that common cause failure between the two systems is not important. The independent failure probabilities for the turbine powered HPCI and RCIC systems are high compared to the independent failure probabilities for motor driven pumping systems; this reduces the importance of potential common cause 22

failures between HPCI and RCIC. Hatch has experienced no common cause failures of HPCI and RCIC.

Table 3.3-19 of the Submittallists the common cause failure data MGL factors for the components for which a plant-specific common cause analysis was performed.

We performed a spot check of this data, as summarized in Table 11-2 of this report.

Based on the data in Table 11-2 of this report, the common cause factors based on plant specific data are comparable to common cause failure data used in typical IPE/PRAs. The high common cause beta factor for SRVs is notable.

II.4 Core Damage _J;eouence Results This section of the report reviews the dominant core damage sequences reported in the Submittal. The reporting of core damage sequences- whether systemic or functional- is reviewed for consistency with the screening criteria of NUREG-1335.

The definition of vulnerability provided in the Submittal is reviewed. Vulnerabilities, enhancements, and plans for plant hardware and procedural modifications, as reported in the Submittal, are reviewed.

II.4.1 Dominant Core Damage Sequences The IPE utilized systemic event trees, and reported results consistent with the criteria for systemic based analyses delineated in NUREG-1335. [lPE submittal, Section 3.4.1] Figures 11-1 and 11-2 of this report summarize the contribution to core damage by groups of initiating events. Figures ll-3 and 11-4 summarize CDF by class of accident for the two units. The definitions of the accident classes are as follows:

lA Transient with loss of high pressure injection and failure to depressurize 11 Loss of containment heat removal, injection lost after containment failure IB Station blackout with loss of high pressure injection and failure to

. depressurize lilB Small or medium LOCA with loss of high pressure injection and failure to depressurize ID Transient with loss of low pressure injection IV ATWS with containment overpressurization followed by loss of injection V Unisolated LOCA outside containment IC ATWS with loss of injection IllC LOCA with failure of low pressure injection 23

! Table 11-2. Common Cause Factors for 2-of-2 Components Component Hatch Beta Factor Value from Source Indicated Submittal Table 3.3-19 in Footnote Diesel Generator 0.014 (fail to start) 0.04 (23. <3) 0.018 (fail to run) 0.03 (') fail to run

{0.006 for fail to start}

MOV 0.052 0.05 0) 0.09 (2). <3) 0.05 (d)

RHR Pump 0.17 (fail to start) 0.1 03. <23 0.011 (fall to run) 0.2 ( )

0.1 (*) fail to start

{0.02 for fall to run}

Safety / Relief Valve 0.28 0.1 0) 0.2 ( )

0.3 (') fail to open on pressure

{0.1 fail to open on signal}

High Head Pump Not Provided 0.2 0)- (')

Core Spray Pump 0.16 (fail to start) 0.2 ( )

0.007 (fall to run) 0.2 fail to start

{0.02 for fail to run}

Service Water Pump 0.0068 (PSW fall to start) 0.03 0)- ( )

0.061 (PSW fail to run) 0.16 (RHRSW fall to start) 0.010 (RHRSW fail to run)

Circuit Breaker 0.14 (both > 600 V and 0.2 (') for 480 V and higher

< 600 V) 0.07 (*) for less than 480 V HPCl/RCIC Turbine Pump Not Provided 0.02 fail to start t')

{0.009 for fall to run}

(1) NUREG/CR 4550 Peach Bottom, Table 4.9-1.

(2) NUREG/CR 4550 Grand Gulf, Table 4.9-29 (3) NRC IPE Review Guidance, Rev 1, November 1993 (4) PLG Generic Data in Brown Ferry IPE Submittal Table 3.3.4-10.

24

3 CDF by initiating Event Group Hatch Unit 1

-~

Spec. Electrical i Loss Offsite Power

} -,

o Generic Transients 0 -

i , ,

c Spec. Mechanical as

> -.I i LOCAs

.5 ' '

) .5 ATWS

! .t C

~

Internal Flood

-m& mm Total

x, N 1E-008 1E-007 1E-006 1E-005 0.0001 l

CDF,1/Ry yr 4 Figure ll-1. CDF by initiating Event Group, Unit 1 j 25

. _ . . _ _ _ _ _ . _ _ _ _ _ . . . _ _ . _ _ _ _ _ . . _ _ _ _ _ . - _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ m__ _ _ _ _

e e

l CDF by initiating Event Group Hatch Unit 2 i Spec. Electrical l

Generic Transients

\

l 0 Loss Offsite Power 0 - , ,

c Spec. Mechanical o -

I i l

LOCAs

, a

,c _ i .

.3 ATWS

~

• i: - -r

~

Internal Flood

-mb -

m Total

'  :  ; \,

1E-008 1E-007 1E-006 1E-005 0.0001 ,

CDF,1/Ry yr Figure 11-2. CDF by initiating Event Group, Unit 2 26

CDF by Accident Class Hatch Unit 1 IA 11

- w IB e _

m

.g! 111B 0 - , ,. ,

E ID i e i i y -

~5 IV u

4 -

V '

_ -r

' IC /

- --F j lilC ' / \ N.

1E-008 1E-007 1E-006 1E-005 j

4 CDF,1/Ry yr 4

Figure 11-3. CDF by Accident Class, Unit 1 i

1 27 4

l

~

l CDF by Accident Class Hatch Unit 2 l -

\ ~

l IA 1

l Il l

l lilB 1 m -

$ IB O _

E ID i ,

'5 IV N ~

V i

_- r-IC illC >

- / \ \ I 1E-008 1E-U07 1E-U06 1E-U05 CDF,1/Ry yr Figure 11-4. CDF by Accident Class, Unit 2 28 I

I

it is notable that the Submittal reports results for both units, even though the results are similar for the two units. The Submittal discusses the risk significant differences between the two units. [lPE submittal, Table 3.4-3) Important differences between the units are as follows:

Swing DG 18 aligns to Unit 1 unless a LOCA signal is generated for Unit 2, in which case DG 1B aligns to power Unit 2.

I Initiating event frequencies were slightly different between the two units for the following initiating events: MSIV closure, loss of feedwater, reactor trip, turbine trip, and loss of condenser vacuum.

Drywell cooling more redundant at Unit 2.

Control room cooling for both units is supplied by Unit 1 systems.

Table 3.4-4 of the Submittallists CDF by initiating event. The top five contributors to CDF for each unit, and the percent contribution to overall CDF, are as follows:

Unit 1 Top Initiating Events: Loss of Offsite Power 26 %

Loss of 600 V ac Bus C 14%

Loss of Feedwater 9%

Loss of Station Battery A 8%

MSIV Closure 7%

Unit 2 Top initiating Events: Loss of Offsite Power 23%

Loss of 600 V ac Bus C 13%

Loss of Feedwater 12%

MSIV Closure 10%

Loss of Station Battery A 8%

Table 3.4-5 of the Submittal summarizes initiating events contributing to each class of accident. Table 3.4-6 summarizes dominant hardware, human error, and recovery failure contributors to CDF. The ranking in Table 3.4-6 is based on the percentage of the total CDF involving the given event. Dominant hardware failures include:

HPCI falls RCIC fails Offsite Power not recovered Condensate fails DG C falls DG A fails Hardened Vent fails 29

~- , . - - _ - _ _ . -__

~

DG 1B Fails Instrument Air fails Pressure Relief inadequate 600 V DC Bus fails  ;

PSW failures  !

RHRSW failures. i Dominant human errors and recovery failures include:  ;

Failure to depressurize Failure to recover HPCI Failure to recover DGs  :

Failure to align for containment heat removal i Failure to recover containment heat removal i Failure to initiate purge mode ventilation for control room cooling.

Tables 3.4-10 and 11 provide the top 100 core damage sequences for units 1 l and 2, respectively. The top 5 core damage sequences for the two units are summarized in Table 11-3 and 11-4 of this report. i The Submittal states that the CDF from internal initiating events for Unit 1 is  !

2.1E-5/ year, and the CDF from internal flooding is 3.4E-8/ year. The Submittal states  ;

that the CDF from internalinitiating events for Unit 2 is 2.2E-5/ year, and the CDF from j internal flooding is 3.3E-8/ year. In January,1994, the licensee transmitted a letter to  !

the NRC indicating that two minor errors were identified in the IPE. [lPE,1/94 Letter) l With these errors corrected, the overall CDF for Unit 1 increased by 8%, and the overall CDF for Unit 2 increased by 9%. The update had no significant impact on the conclusions of the original IPE. For Unit 1, the CDF for accident class IA increased by 16% and the CDF for accident class ID increased by 24%. For Unit 2, the CDF for accident class IA increased by 17% and the CDF for accident class ID increased by 27%.

The licensee provided the specific sequences affected by the update to the IPE.

[lPE Responses] The update corrected a software error involving the calculation of the fraction of HPCI and RCIC failures that were attributed to failure to start. Start failures were modeled as recoverable; other failures in these systems were modeled as non-recoverable. The top 5 sequences affected by the update are summarized in Table 11-5 of this report.

30

~_ _ .,_ __ _

Table 11-3. Top Five Sequences for Unit 1 Initiating Event Mitigating System Failures Sequence Frequency 1/ year Loss of 600 V ac Loss of 600 V ac Bus 1D; all DC 7.7E-7 Bus 1C power is lost; all high and low pressure injection is lost; SRVs operate only in safety mode; boil off inventory at high pressure; no recovery of buses credited Loss of Feedwater Failure of HPCI and RCIC; Operators 7.5E-7 inhibit ADS and do not depressurize; no injection available at high pressure Dual Unit Loss of Dedicated DGs 1 A and 1C fail; swing 6.6E-7 Offsite Power DG 1B fails; HPCI lost due to no room cooling; RCIC works; control room cooling lost but RCIC control transferred to remote shutdown panel; RCIC lost at 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> after DC lost; No recovery of offsite power or DGs Loss of Plant Failure of containment Vent; no 5.0E-7 Service Water containment cooling; containment fails by overpressurization and causes failure of inlection systems in reactor building due to EQ Loss of DC Panel RCIC falls due to IE; Inadequate 4.6 E-7

~

number of SRVs open to relieve pressure resulting in medium LOCA; HPCI fails; Operators fail to Depressurize 31

o .

Table 11-4. Top Five Sequences for Unit 2 Initiating Event Mitigating System Failures Sequence Frequency 1/ year Loss of Feedwater Failure of HPCI and. RCIC; Operators 1.1 E-6 inhibit ADS and do not depressurize; no injection available at high pressure Loss of 600 V ac Bus Loss of 600 V ac Bus 1D; all DC 7.7E-7 1C power is lost; all high and low pressure injection is lost; SRVs operate only in safety mode; boil off inventory at high pressure; no recovery of buses credited Dual Unit Loss of Dedicated DGs 1 A and 1C fail; swing 6.6 E-7 Offsite Power DG 1B fails; HPCI lost due to no room j cooling; RCIC works; RCIC lost at 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> after DC lost; No recovery of I offsite power or DGs i

MSIV Closure Inadequate number of SRVs open to 5.5E-7 l relieve pressure resulting in medium j LOCA; HPr' fails and RCIC i inadequate wr medium LOCA; Operators fail to depressurize Loss of Plant Service Failure of containment Vent; no 5.0E-7 Water containment cooling; containment fails  !

by overpressurization and causes failure of injection systems in reactor building due to EQ 32

O e I

l Table 11-5. Top Five Sequences Affected by IPE Update  ;

Old Frequency New Frequency Sequence )

Sequence l 1/ year 1/ year Frequency Percent Change Loss of Feedwater; 7.5E-7 1.05 E-6 40%

Unrecovered Failure of HPCI and RCIC; Operator Failure to Depressurize Loss of Station Battery A; 3.97 E-7 8.01 E-7 102 %

Unrecovered Failure of HPCl; Successful Depressurization; Failure of Low Pressure Permissive for Core Spray Loss of Station Battery A; 2.01 E-7 4.03 E-7 100%

Unrecovered Failure of HPCl; Operator Failure to Depressurize Loss of Offsite Power; 1.18E-7 1.62E-7 37 %

Unrecovered Failure of HPCI and RCIC; Operator Failure to Depressurize 1.25E-7 1.52E-7 22 %

Closure of PSW Valve 1P41-F303A; Unrecovered Failure of RCIC; Operator Failure to Depressurize II.4.2 Vulnerabilities We reviewed the definition of vulnerability as provided in the Submittal, and the vulnerabilities,if any, identified and discussed in the Submittal.

Section 3.4.2 of the Submittal discussed front-end vulnerabilities. Two criteria were used to determine a front-end vulnerability:

(a) Any accident class with a frequency that is greater than 1E-4/ year or which contributes greater than 50% to the total CDF (b) Any accident class V sequence with a frequency that is greater than 1E-5/ year or which contributes greater than 20% to the total CDF.

33

1 l

)

l l

Based on these criteria, the licensee concluded that no vulnerabilities exist. The I corrections reported in the January 10,1994 letter do not alter this conclusion.

11.4.3 Proposed improvements and Modifications l PRA analysis for Hatch continued over four years prior to completion of the IPE l Submittal. As a result, items of significance were identified and changes to address  !

these items were scheduled. Also, plant changes driven by other regulatory l considerations were scheduled for implementation. The modifications scheduled for i completion past the IPE freeze date that were credited in the IPE are as follows: [lPE 1 submittal, Section 6.2]

(a) installation of hardened containment vent (b) removal of common PSW discharge valve in Unit 1  ;

(c) changes in procedures and modifications to ducting for the control j building to allow continued operation of electrical equipment in the control '

building following loss of HVAC l (d) recovery action to use smoke purge ventilation upon loss of control room l chillers was proceduralized 1 (e) modifications to intake structure ventilation system l (f) procedure changes to allow tripping of RHR/CS pumps in ECCS rooms to allow continued operation of 1 pump without room cooling (g) procedure changes to allow cross connection of PSW cooling water to RHRSW pump motors l (h) modification to allow swing chiller compressor for control room HVAC to j be powered by either division of electrical power (1) modifications to meet the SBO rule including: replace station service battery chargers, and enhance procedures dealing with loss of .

ventilation. l It is not clear which of these plant changes were a direct result of PRA activities.

The Submittal states that the cumulative effect of these changes lowers the overall CDF by about a factor of 10.

We believe that the most important of these potentialimprovements is the installation of the hardened containment vent, since based on the assumptions used in the IPE, if suppression pool cooling is lost the containment can be vented and core cooling using recirculation from the suppression pool maintained.

11.5 Interface issues This section of the report summarizes our review of the interfaces between the front-end and back- end analyses, and the interfaces between the front-end and human factors analyses. The focus of the review was on significant interfaces that affect the ability to prevent core damage.

l 34

II.5.1 Front-End and Back-End Interfaces The IPE assumes that if suppression pool cooling is lost, core cooling with ECCS in recirculation from the suppression pool can be maintained before, during, and after containment venting.

The IPE assumes that if the containment falls by overpressurization, all injection to the vesselis lost as a result of EQ considerations in the reactor building.

The IPE credits operator action to switchover RCIC or HPCI from the SP to the CST over the long term if SP cooling is lost. Also, operator action to institute core / containment flooding with non-SP sources such as RHRSW cross tie and fire water are credited.

The IPE does not address the impact of containment isolation on the ability to cool the recirculation pump seals; without seal cooling, a transient can evolve into a small LOCA. It is possible that the supply of cooling water to the recirculation pumps is isolated on containment isolation; however, the Sections of the UFSAR available to us did not address this. The impact of a seal LOCA is not of major item of significance at Hatch, since all systems for core cooling involve injection to the vessel.

The RISKMAN software allows direct linking of the Containment Event Tree (CET) to the level 1 event trees; thus, it is not required to bin level 1 core damage sequences into PDSs for level 2 analysis. [lPE submittal, Section 3.1.5] Core damage sequences were binned into accident classes for presentation and interpretation of results. No PDSs for the front-end core damage sequences were produced.

l1.5.2 Human Factors Interfaces Based on our front-end review, we noted the following operator actions for possible consideration in the review of the human factors aspects of the IPE:

manual initiation of depressurization (note: procedures direct manual inhibition of automatic ADS for ATWS mitigation considerations [lPE submittal, page 3.1-13]) .

manualinitiation of SP cooling manualinitiation of containment venting manual initiation of smoke purge ventilation for the main control room manual tripping of all but one RHR/CS pump in ECCS room (s) on loss of ventilation to the room (s) manual switchover of RCIC and HPCI from the SP to the CST long term when SP cooling is lost 35

manual alignment of core / containment flooding systems such as RHRSW crosstie and firewater, and subsequent control over containment water level.

11.6 Evaluation of Decav Heat Removal and Other Safety Issues This section of the report summarizes our review of the evaluation of Decay Heat Removal (DHR) provided in the Submittal. Other GSl/USI's, if they were addressed in the Submittal, were also reviewed.

11. 6 . 1 Examination of DHR Although the IPE evaluated all aspects of decay heat removal, the evaluation of DHR in Section 3.4.3 of the Submittal is restricted to the final heat sink options: RHR, PCS, or containment venting.

The Submittal contains a description of the contributors to loss of DHR for the restrictive definition of DHR used in the evaluation of DHR; these are the class 11 accidents. Loss of DHR, using the restricted definition, contributes 24% and 21% to overall CDF at Unit 1 and Unit 2, respectively. Loss of 600 V AC bus C is the dominant initiating event, and failure of the hardened vent is the dominant mitigating system failure, ll.6.2 Diverse Means of DHR The IPE evaluated the diverse means for DHR and for core cooling. Cooling options evaluated included: main condenser /feedwater, high and low pressure ECCS systems with containment cooling or containment venting, and core / containment flooding with alternate injection systems such as RHRSW or firewater.

The use of containment venting as a backup to suppression pool cooling is an important aspect of DHR modeled in the IPE.

11.6.3 Uniqu.e Features of DHR The unique features at Hatch that directly impact the availability to provide DHR are as follows:

Swing DG with dedicated PSW cooling water pump Hardened Containment Vent Ability to flood core / containment with alternate sources such as RHRSW and firewater 2.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> battery lifetime, procedures do not instruct operators to shed DC loads.

36 1

. e The impact of these design features on CDF is discussed in Section ill of this report.

Other notable features of the Hatch design are as fcilows:

Offsite power for 1E buses not off auxiliary transformers, but off startup transformers thus not requiring switchover on plant trip Existence of 31E buses, not 2 DC loads diversified between station and DG batteries No dependence of instrument air on PSW, dedicated cooling water provided Ability to cross tie instrument air between units Passive nitrogen supply for SRVs LPCI loop selection logic eliminated thus removing major LPCI single failures All offsite power provided by 230 K, but two switchyards present at 500 K and 230 K interconnected by auto transformer.

11.6.4 Other GSI/USI's Addres. sed in the Submittal No GSI/USI's other than DHR were specifically addressed for resolution in the Hatch IPE Submittal.

11.7 Internal Floodino We reviewed the treatment of internal flooding in the IPE. The review focused on both the. methodology and the results of the analysis of internal flooding.

11. 7 . 1 Internal Flooding Methodology Flood scenarios were identified based on the source of flooding, flood propagation, and key equipment locations. Scenarios wi the flooding effects with independent failures, evidently using the internal initiating event trees. Both submergence and spray related failures were addressed in the flooding analysis.

37 e

11.7.2 Internal Flooding Results Flooding analyses were performed for both units. For Unit 1, two flood scenarios were retained as significant. The first scenario involves a breach in the fire i protection system inside the reactor building that fails the CS and RHR systems, and causes degradation in the HPC! and RCIC auto start instrumentation. This scenario has a CDF of 1.9E-8/ year. The second scenario involves a breach in the fire protectia, system inside the reactor building that fails the CS and RHR systems and one loop of drywell spray, and causes degradation in the HPCI and RCIC auto start i instrumentation. This scenario has a CDF of 1.5E-8/ year. For Unit 2, two flood scenarios were retained as significant. These two scenarios are similar to the two scenarios retained for Unit 1, involving breaks in the fire protection system inside the reactor building and flooding induced loss of CS and RHR. The CDFs from these two scenarios for Unit 2 are 1.8E-8/ year and 1.5E-8/ year.

l l

l l

38 i

l l

l l

111. REVIEW

SUMMARY

AND DISCUSSION OF IPE INSIGHTS AND IMPROVEMENTS This section of the report provides an overall evaluation of the quality of the IPE 1 based on this review. Strengths and shortcomings of the IPE are summarized. Also, l l

the dominant contributors to CDF are summarized, with insights as to the effect of plant design and modeling assumptions on the overall CDF and the types of scenarios '

that dominate CDF. Improvements planned as a result of the IPE are summarized.

Areas where licensee efforts could improve safety are summarized. Areas where the IPE process could be improved are summarized.

The Submittalis complete in terms of the information requested by Generic Letter 88-20 and NUREG 1335. The large event tree, support state methodology was used to perform a level i PRA. Independent reviews of the IPE were completed.

Plant walkdowns and the use of plant specific documentation were used to assure that the IPE modeled the as-built plant.

Major strengths of the IPE are as follows. The set of plant specific initiating events appears to be complete. The IPE modeled the impact of loss of control room cooling and of loss of drywell cooling. The IPE quantified CDF at both units and discussed the differences between the two units.

No major shortcomings of the IPE were identified.

The IPE is comparable in depth to typical IPE/PRAs.

The total CDF from internalinitiating events is 2.1E-5/ year for unit 1 and 2.2E-5/ year for Unit 2. The total CDF from internal flooding is 3.48E-8/ year for Unit 1 and 3.3E-8/ year for Unit 2. In January,1994, the licensee transmitted a letter to the NRC indicating that two minor errors were identified in the IPE. [lPE,1/94 Letter] With these errors corrected, the overall CDF for Unit 1 increased by 8%, and the overall CDF for Unit 2 increased by 9%.

Section l1.4 of this report summarizes the contributions to CDF by: initiating event, accident class, and dominant hardware and human error failures. Section l1.1.4 of this report lists the plant modifications scheduled for completion after the freeze date that we.re credited in the IPE models.

The following design features impact the CDF:

Swing DG with dedicated PSW cooling water pump Hardened Containment Vent Ability to flood core / containment with alternate sources such as RHRSW and  ;

firewater 2.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> battery lifetime, procedures do not instruct operators to shed DC loads.

l 1

39 l

\

.- i ,

'The impact of these design features on the CDF is as follows. The swing DG tends to lower the CDF compared to a plant with two DGs, since it provides a third DG to 1 backup the two dedicated DGs at the unit with the accident. The hardened containment vent tends to lower the CDF since it provides a backup for energy removal from containment if containment cooling systema fail. The ability to use RHRSW and firewater for core cooling tends to lower the CDF since this provides for more low pressure options for core cooling in response to transient events. The 2.5 ,

hour battery lifetime tends to raise CDF since it is shorter than the battery lifetime at many BWRs and this reduces the time available to recover AC power during station ,

blackout accident scenarios.

Based on our review, the following modeling assumptions have an impact on the overall CDF: l (a) no loss of adequate NPSHA for ECCS pumps pulling from the suppression pool before, during, and after containment venting if ,

suppression pool cooling is lost <

(b) no overheating of the running pump if only one RHR/CS pump in an .

ECCS room is running after loss of ECCS room cooling l (c) requirement for control room cooling with either chillers or once through .'

smoke purge ventilation.

The first assumption tends to reduce the CDF since it allows for core cooling to '

continue with containment venting. The second assumption tends to reduce the CDF since it allows for core cooling to continue without ventilation to the'ECCS rooms. The ,

third assumption tends to increase the CDF since it requires that ventilation for the -

control room be provided, or else the shutdown panel must be used.

No improvements other than those credited for implementation after the freeze l date are planned as a result of the IPE.  !

We believe that the most important of the improvements considered after the ,

freeze date is the containment vent, since with the assumptions about NPSH margin in the IPE, containment venting provides a backup for suppression pool cooling. ,

The IPE provides a model of the CDF for Hatch 1 and 2 that is consistent with the models developed in typical IPE/PRAs. The IPE identifies the dominant contributors to CDF consistent with the design characteristics of the units.

Signdicant findings on the front-end portion of the IPE are as follows:

. core damage was quantified separately for both units, but the results are not significantly different for the two units

. two plant specific initiating events contribute significantly to the overall CDF: Loss of 600 V AC Bus C and Loss of Station Battery A

. loss of control room HVAC and loss of once through ventilation, requires ,

control of key systems from the shutdown panel to mitigate accidents

= battery lifetime is only 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> and procedures do not direct DC load shedding to preserve battery lifetime 1

40

~*

..;. n

, _ .g E

~

~

. - room cooling is not required if 'only one LPCI or core spray pump ln a:  !

room is operating l

. core cooling can be maintained without containment cooling if  :

containment venting is successful; however, with no containment' cooling and no containment venting, no credit was taken for continued core l cooling after containment failure by overpressurization. j ;

[

-i l

l 3

i t

l 41 1

1 IV. DATA

SUMMARY

SHEETS This section of the report provides a summar; of information from our review.

Overall CDE The total CDF from internal initiating events is 2.1E-5/ year for Unit 1 and 2.2E-5/ year for Unit 2. The total CDF from internal flooding is 3.4E-8/ year for Unit 1 and 3.3E-8/ year for Unit 2. (A January,1994 letter stated that minor errors in the IPE had been corrected resulting in an increase in total CDF of 8% for Unit 1 and 9% for Unit 2.)

Dominant initiatina Events Contributina to CDF initiating events contributing more than 1% to the total CDF for each unit are as follows:

Unit 1 Top initiating Events: Loss of Offsite Power 26%

Loss of 600 V AC Bus C 14%

Loss of Feedwater 9%

Loss of Station Battery A 8%

MSIV Closure 7%

DC Panel Deenergizes 6%

Loss of Plant Service Water 5%

Turbine Trip 5%

Reactor Scram 3%

Loss of Control Room Cooling 3%

Loss of 600 V AC Bus D 2%

inadvertent Open SRV 2%

ATWS with MSIV Closure 1%

Unit 2 Top Initiating Events: Loss of Offsite Power 23%

Loss of 600 V ac Bus C 13%

Loss of Feedwater 12%

MSIV Closure 10%

Loss of Station Battery A 8%

DC Panel Deenergizes 6%

Loss of Plant Service Water 5%

Turbine Trip 4%

Reactor Scram 3%

Loss of Control Room Cooling 2%

inadvertent Open SRV 2%

AW/S with MSIV Closure 2%

Loss of 600 V AC Bus D 2%.

42

j Ji +- ,  !

t Dominant Hardware Failures and Ooerator Errors Contributina to CDF Dominant hardware failures contributing to CDF include:

i HPCI fails  ;

-.RCIC fails .

Offsite Power not recovered q Condensate fails l DG C fails DG A falls Hardened Vent falls )-

DG 1B Falls Instrument Air fails Pressure Relief inadequate 600 V DC Bus fails PSW failures i RHRSW failures.

Dominant human errors and recovery failures contributing to CDF include: l Failure to depressurize i Failure to recover HPCI Failure to recover DGs )

1 Failure to align for containment heat removal i

Failure to recover containment heat removal Failure to initiate purge mode ventilation for control room cooling, f

Dominant Accident Classes Contributina to CDF  :

\

Major classes of accidents contributing to the total CDF for Unit 1, and their percent contribution are as follows:

l 1

IA 32.5 %

Il 23.0%

lB 16.0%

lilB .14.1 %

ID 11.6 %

IV 1.4%

V 0.8%

IC 0.5% l lilC 0.2%  ;

Major classes of accidents contributing to the total CDF for Unit 2, and their percent contribution are as follows:

i I

43

. c I

IA 36.8%

ll 19.5%

lilB 16.2%

IB 14.9%

ID 9.5%

IV 1.6% -

V 0.8%

IC 0.6%

lilC 0.2%

The definitions of the accident classes are as follows:

lA Transient with loss of high pressure injection and failure to depressurize il Loss of containment heat removal, injection lost after containment failure IB Station blackout with loss of high pressure injection and failure to depressurize IllB Small or medium LOCA with loss of high pressure injection and failure to depressurize ID Transient with loss of low pressure injection IV A'MS with containment overpressurization followed by loss of injection V Unisolated LOCA outside containment IC ATWS with loss of injection IllC LOCA with failure of low pressure injection Deslan Characteristics Imoortant for CDF The following design features impact the CDF:

Swing DG with dedicated PSW cooling water pump

. Hardened Containment Vent.

Ability to flood core / containment with alternate sources such as RHRSW and firewater 2.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> battery lifetime, procedures do not instruct operators to shed DC loads.

The impact of these design features on the CDF is as follows. The swing DG tends to lower the CDF compared to a plant with two DGs, since it provides a third DG to backup the two dedicated DGs at the unit with the accident. The hardened containment vent tends to lower the CDF since it provides a backup for energy removal from containment if containment cooling systems fall. The ability to use RHRSW and firewater for core cooling tends to lower the CDF since this provides for 44

more low pressure options for core cooling in response to transient events. The 2.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> battery lifetime tends to raise CDF since it is shorter than the battery lifetime at many BWRs and this reduces the time available to recover AC power during station blackout accident scenarios.

i Modifications The modifications scheduled for completion past the IPE freeze date that were credited in the IPE are as follows:

(a) installation of hardened containment vent (b) removal of common PSW discharge valve in Unit 1 (c) changes in procedures and modifications to ducting for the control '

building to allow continued operation of electrical equipment in the control building following loss of HVAC (d) recovery action to use smoke purge ventilation upon loss of control room chillers was proceduralized (e) modifications to intake structure ventilation system (f) procedure changes to allow tripping of RHR/CS pumps in ECCS rooms to allow continued operation of 1 pump without room cooling (g) procedure changes to allow cross connection of PSW cooling water to RHRSW pump motors (h) modification to allow swing chiller compressor for control room HVAC to be powered by either division of electrical power (1) modifications to meet the SBO rule including: replace station service l

battery chargers, and enhance procedures dealing with loss of

• ventilation.

i Other USl/GSis Addressed No USI/GSis are specifically addressed for resolution in the Gubmittal other than DHR.

1 Slanificant PRA Findinas I Significant findings on the front-end portion of the IPE are as follows:

l

= core damage was quantified separately for both units, but the results are not significantly different for the two units a two plant specific initiating events contribute significantly to the overall CDF: Loss of 600 V AC Bus C and Loss of Station Battery A

= loss of control room HVAC and loss of once through ventilation, requires control of key systems from the shutdown panel to mitigate accidents 45

?

• battery lifetime is only 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> and procedures do not direct DC load shedding to preserve battery lifetime

. room cooling is not required if only one LPCI or core spray pump in a room is operating a core cooling can be maintained without containment cooling if containment venting is successful; no credit was taken for core cooling without containment cooling and without containment venting.

i 46

. . l APPENDIX A. SCOPING CALCULATIONS This appendix provides a brief summary of scoping calculations that we performed to check the validity of certain assumptions used in the IPE. The purpose of these calculations was to assist in review of the IPE and to focus requests for additional information from the licensee. .

Figures A-1 and A-2 show the decay heat and integrated decay heat energy for Hatch, respectively.

We used our in-house SEA computer code for BWRs to model RCIC and HPCI makeup with various numbers of SRVs failed open. This code was developed and used in support of the Grand Gulf shutdown PRA, and has been used in numerous l

other reviews that we have performed for BWR IPEs. The code calculates measured and actual water levels for transients and small LOCAs with makeup, Moody flow out water or steam openings, and time dependent decay heat. The code is written in I Borland Pascal for Windows with objects and uses steam tables as an inherited object.

The following data was used in model for Hatch. Each SRV relieves about 877,000 lbm/hr at 1100 psig. [UFSAR, Table 5.2-4) Using the Moody model for '

blowdown, this indicates an equivalent valve area of about 0.1 sq ft which is larger than the maximum small LOCA steam line size used in the Hatch IPE. RCIC injection was taken as 400 gpm, automatically initiated at -47 inches measured level with respect to instrument zero. [lPE submittal, Section 3.2.1.18] [UFSAR, Figure 5.4-2]

HPCI injection was taken as 4250 gpm, automatically initiated at -47 inches measured level with respect to instrument zero. [lPE submittal, Section 3.2.1.17] [UFSAR, Figure 5.4-2) Normal level was taken as +36 inches from instrument zero, and the top of the core is at -158 inches from instrument zero. [UFSAR, Figure 5.4-2] The initial l

inventory of water was calculated as 4.6E5 lbm, and the vessel volume was taken as 1.63E4 cu ft. [UFSAR, Figure 5.1-1)

The Browns Ferry and Quad Cities IPEs credited RCIC for mitigation with one open SRV, but different plants have different size SRVs and different capacity RCIC makeup. For example, the SRVs at Quad Cities each relieve 560,000 lbm/hr at 1150 psia, and the SRVs at Browns Ferry are similar to those at Hatch in capacity but the RCIC at Browns Ferry can provide 600 gpm makeup. We calculated that the core will not uncover at Browns Ferry or at Quad Cities with RCIC and one open SRV.

Figure A-3 of this report shows pressure and levelin the vessel for a transient with one stuck open SRV. This calculation indicates that the top of the core is uncovered by less than 10 inches for about 1200 seconds.

The equivalent size of two open SRVs is about 0.2 sg ft in size. Figure A-4 provides the results of our assessment assuming HPCI makeup of 4250 gpm and auto-initiation at -47 inches measured. [lPE submittal, Section 3.2.1.17] [UFSAR, Figure 5.4-2) Our results indicate that HPCI can easily accommodate 2 open SRVs.

47 l

p.

4- .s O

O Om 4.z. ; . 4 -w o. t . .. . .,.4, , .. . n. .. ..,. z...+..k..n...

,;. .4.

.e. r....,.m.

..e . n, t. . . . .: . . . ..n

. . 4, +.. ...... , w . ...c.  ;, ... i.....t.

.). .b d 4. . ..... ~. ;.U.

. . ....a....,~+ r.,. . . .. . i, ~o...... . .. , .: ...

. . m. w. o, .

. ..n....n....,~... .

c w -..+ k. . w +

.yw ~ . . .z .~.

~ . . .

>.s....  ::: ...,.

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

., ... . y 7 W3 .~g......- 4: -..-y*hm w.M ;... w.. ... .n.:.m.. .u. .... .M..~ n. ...r ; N.. .,@.e b,t.e . .

o 4, %... . . .

n.

m W Q;-.

t m p p.3.m..p . .m{. . . s .u. ..y;.%.m. pn.. 4,+. . .1.

...g.. t..p..

4 wg . >. 3.t.m. a:M-

+c. . ... 6 W M+ .u.

. . . .v. .. .. ., .. ., .u.+. v. . ..e. ..w.. ,. :.. .q

. . u.f . a.a 6 .n. .m.a. n.. , n . {.3: 4.sa..Nr. .,;p.

~ .=.a.... . . . m.. n. .

m c ... .. -

A,c&. m.P;D J.O. (p...t

~ ~4w -i ..  %,I.7.

w .m.1~.4;

: W m. :t .t e ,.

4..:

s.

u e .r1+ 4. .. . ..r~ o, . . < . .

yd,e . , 4. 4. xCM. y .* OM.b MO.  % %;e.o.;i.c4;;

o 4. f. T v. Y..:L;m e,- ~ < ; yh . n,;g3.- ww+w.3..w: . ' .w 7 U. .M,.. !n. .Q'.11.

~+ W .v e

e.

w+n .x . a,.,.Q . c .9, , ., m m,w s y m ,.. w+ ,. m+-

c., y.~4,p. .

a-n

.s.

c r.. . %, ,

Q n . ?*.hm.m n--n..v. . y- g.,an y.-., a .m. =Q.

. .t -< w -

r . . <...s .i - un.

i w+.s; .y g &c u ne4,. .+. - . ..s M,o. .n.m Af e '.. . p....- Q

.e..hm gt? M.k. o . ..E. M.+.:.2';.W. .. .... ' d 6, m

... ..V..e. .r. M... ,.9..M. --. . .M. p<.~ww s .

W..W ..

q..pN. n".c.n.p t w+ rM 4. o.u .7, .u ..o....

y .4 .. .n t .. ~.

~-.M. y m..y;p..p* ..

. . .m

.;.: 4.'GW., . pf. .w .

..W.:,i... .4 < a e.f.;..a

4;W

+, , . . , . sn m v- ,  : wa s sev e e..y.ns. < . . ~ . ..x.

. . .'.'e +. .~ .. . V. . .

w,n. . . . .a. .a.f.aw,M.w,w: . p. +',.  :.n . p.". .. T .. . W.uG.%4.%Q.'... ,:. .a.m.a. . . y .b. . , s..m.w ,m

.e a ..

a n.m. nd.e e. .s.r. .o.L.~Q

..%.,. v... . x. . x m. .c.w

. y. .. n. .'*4 J: .u -

6 6 N N.n' .*(

., p.

i lI@ DM._ . ,J

. ..S, . . .O. .

g

. . '. l... .. o,a>* $.4. [E.J. .f. .'.E.i 6,.O $.# W,..i,f.U; .b.

.n-1.........  ; @g ,$n . .1.  %,3.,.M, r, . N. . 6 .$. .#N.J.,.. .. A. ...N..u.I,N..'.' bew.

a$mmi -

r.. w = .

qy;.

v.

Q: n.. @ ,o

$ - . .M. .;. .

.m, m.A.

- .1 / .. o

'tq *rhp @':MT7W't : px' ' .W *

. 4 y . . m.q m.-.M..

. . p . .h. k ..A. ..p.. ..a g

Q@ .%?Qp.J W /18".g%Qg"0.MM QD "" i,1;ci.*.1UM e$ c.f n%

.. d I 2. .. . ,'-

s ,,

m-D I g u P H"W hnp P*"Om - y n.7.a o ay a, . - - .. ~M. . w  %.c*

.  : > P.. M h"~

.hDh.s,.e,~  :+yy . a". v ~. y w ms,s<.+*

% mW QQ

+.o.,4..> . m'

.a f tL 3._.. W . A- P:w x y* .-

p> .

Y.c.e, .O p.n M.m.

.u.. a.

%, t p m ~. ' _.r Q g

.g % , MM 4s ., dW:

f .

@b6 a. c/W y W 4 *) p[* m,p:m t hm j~%yMp Dm i

• %e s. @*&@n.;.6 W: n".......  % u +

y

.. w . o.... O. g W t

0.,.v. .b.G. n '4 M@, .. 6 w g . . A%.;. .. . .ud.;,. . *.%.:

- ..e...y

..u .v.

g y,p.v. u m@;~...y.v.W

.. . a m x y.m..u.

y;.n.>.~.. w :..; w.~... .a,:.t.m.g..  :. x m O a o.m.u . . . . w.w. . ,.p. wWn.w=p.r. V y,p p.p.. n.@,.m> r.m :m, r . w.v.w.. n ,e . . .m o u.,w? : yQ >

u .,

w p p.p

. ~. .;. s .n. % . .m e s v

~

g p W.au. .. .

w ..e.

W..~

t.

.... .c. .e. y. . ., .u . . . . .

L. 3.4 .. .

M

..y.

m..

~ .

-(n

.. ..+. . . t..

~w.

m .

do..g D.~.. .

O O P.*..

. . . ,n . < g n

. On0,d6 . ~ m .

4 *.a.0J.[.2.'.'.*.,24'fM4&'.!'.k

~.a ;

U .L' 7.MC f. .L';.* '.,. .. / '".*.d~.%

.. .g.iW..W gg %N'a F O W MWCt%r&^hW4!TO.:MMu};.U!.R .m  ; * ' JCM.jf.CC6 g th;.M @L,:e.n...%6..,w..

an. v m 4.v

,S.g;2:

w. 4.w , Aa w . F,w. .c. t.v .

te. w*,n.y n M. p .m.., .t. 4%. %q. .,*w.w.

tr.e. ..w - .1.. a g}

.g,,o

.mm . + w, umq..

. .,e. a. w. <, . . .e. ..4..

.r a u m. , -. . . . w.

. ..h. go.9,a r.<.f4 . .. . Aw.4.-

4.s,a .a r. m ,.

m... ,....

,,- q.

(U g

=

-a.- ..#.w- p. ~. 3,y h .;4 % x,.a:y W. .k. . t.c.v . .%.g.... ..o ~o

' p.* 'yA'dy,m~+.

w .a %.

e v %..

s

. :, A*r..b.y4 (  % i*M

• g .o w 4

. ), ;i s}: :n~sk

.+s%  %.U& W ~ " ^Ihm +~ O eggM.g,

MMj#*fnMe% 9 M'. .s

.:h<*.?. m ,.. . X L- nid

.%W.f*$w#my. W,+k.w's' d Q 'a-

y. A14

.,,Lm..%y. m. ,prt..,.:

m.

--m ,.; u r. H....'.y

- ~9 9. y .s 4.,, 4 Q .cn

, . 4.,.. w .....u. w .u.s... _

wa .o. . '.u .m ys.m 4 +'. . .r. .<. v ue..s.

' a sg' . a ~Ms.M s D' .s'.a.t... ye W h.% ....h.q re mz M.tn.e.h.*.;.?  :. .. W .c.

M p.e * /.'J.E. w..:$.s&.

c .w. m .. s w .v . .., u 1,%. ~n;v. .,O. . %.~. .. a:

c. .u. n r . .n* , Q.G.,...

i 4~.a A. s . ,. ...... .

m w wva*W134:%.cas Re.,.kv . ma .gu.r w+. v.

~;.9 W '.M

.. 4 whweM;%. *r M m%,)p.,er>%+.  ;. L paq.L... .

p. ;

.m. -

...w, n. p.~. q.p.. .x n.

• d b.l. 52 M... 6 4) $NM. wL.N ,*. *. 9 .D-el A. .

y ;upm.w. ..*O., A re. . y ~ , W. ..

.".'$ M M NE*.4.4 . h ..EIm.% . k .E. . , p.F.*M..a t

.c.u4% m ;.* .n..b.,.h. p..w .W .w+y nd p+.y c .p . (s v.g r. L.:.O. .u.v .,.-

.} a paA.. ., .msp m.m p  ;. . .y vss e ....s; r.~ 9 giWp.;

s .w-.y , . ..

....y ,t u .m s 3 ='k W : W N;s r. s, w %w w,3,,. w% .g m.r: .i.; ?;n,\

. # .w.hm.. 4tw e., .c. .. . wv.d.t A - r .WM.hoo o ,m= e "..h .... q

.. w . w Mn...M ..

w u,..~..3. *y. c..*.r. .,

.%.,  % t< .;g 4sy,....,,..

.g .

5%. .1.* y* $ g%g,, e 6 go mig w.

. .p 4. Y 9 g q 9 4

• 4P*.

4gk '.sig. .py '.' 4. -....a*.a'..Nd.ge '

> a Wwe (a # (M,i L4.Eg

• 4 4

'4 9.v'N

..~.,.M.[.. .. w .w2.. .'.h.N.. 's]f% [W< f,. 4. 'hd .+.m.,.

9. N%.g I( Ag s og..

L.

. . ($_g .'$er.E.

.s x .h,. h,. O 'A..g,.

r , w,r.y .#m,. o,g .~%er ,. svm  %

. . . . . m.g ,.m. _~ r y.,.J r. . .# .,y4.,.. w *.pe go w.-

m. .J>.u

. ex .-. iy..%

. om 3m

@ O C" O O O O O. O.

+ O O W O e v" v-cas/nts 'Jemod

s a A 4 --- -

T W

.. s......; y,.

'. 14 .';..e n 'T',

e

...J" '

c) 2 : 1.,, :.r n -n , -

.+

,y. .n. . .

. . . . , ,. s. . .s.

. . : a. s; : g . ..........n...~.,.....

.. .  : . qs . . <

w .

g

%.. . &.;<x;qg. :'o z-e;.. . ..m.

llA - 3,. .

Q) w 2

4. v. , .m ,

,.~ o q

6 ,-

• m:

.A. . .g,*

, y*:

=88

.e

..3..

g 0,% . . . , . ...... . . ............ .. . ... .,.. .. ...._ .

)y ,% ,T * % ii ,E' ' *", > - l .s '.;

s., ,* ': b

- _ . ' N n~ . a* . > m..,.u...

e , . .

..,.n,. .v, n . .. -s

/......,.,; ,.

m .

.c lM (D, n

. -<rs w-- - ..-

~ o

. ,.c. .~ m,f.,a.v,~,:(os +e ~., .- .a. . ,.- -

. , m.. . ,

. a w.. , g.-

.. . m w ,s.

,o x, .n. 4 trg . ?.w:E.u;. .L c . ,,je. . w . .,. -....n., . . .

.. T y Q, . . . . ...+... L . n.... ..... . e.... ... m .. . . . . . . . . .....

,.s

.l

.,*w %,,4 ..\ ..$w:..,.",+,,..M.~.

n . ,l

-.

e .7, .3.; 3 7

_n. -

(O _'o

,.4 .

,o.. .r.s ...

y .y .

CQ ,, r #,

w.. ,.m.w , N.. o  %.c.

.s., . s

,. s..m. ~.c ..m w y) >.

-.<+<....c

.e. i - . . , . . a.v. . . ,

...~,

w . m m

r. ~ . . *t3 6o.

k . .. w.

~w. .. ..s , ~.s. m.

4 ..

s. ,

_a v w::s.o .M.. _ p, - y . . . . , m,.ao. , qC m u. -4 n , . . ..*- 5 u n*

, . e w ~

. . . . . . . . , . ..... , c. . .M.. . . .,~ .: .x., m. .. .  : x. .a. . .  :

. . . . . .c.W> . . . .

gO c

[g') g,) w

@ ' .3 +w:

. 4, .t, y. .: % . . - wa2 m:. 4 <

m.. .#.  :,.. aa. . :. * :

s .

L.g Q

~

g g r q,.x:w:,

w w ,.::r./

m m w..

v.m.:n. .&. <

a.:W,:,*. yw.p..

O 5 y; e

,.a, w%.m . o p : .:nm. . ., a. m,. .

u.: a,. ,y,.. ,sv . 4 . s

,  : 4 . m . % a ,. t x Q ,,,,,,

,# cw . y .y ,z

...3..4....,..q...#,,,.... -. s. c. .,.n....,..

., -..,...%....y.m...%....,.3.....

..... e >.

.s..

_ . ..,c .m .a. w p n.. . ,

m U C N. .,. - e E_ OoO

i vd.N, M h.YP+ , ., .. s , . .I h I,k,y.>l.[n Q gg .. .

V. .M.~..aL 3.ww 2. .$.ib* W."M7T[:$OM,% W ,, @:. # %..!T, F= d.x so a -

w.... ,,6.

WM. . :9 _..

Q 2

%A

• 1 g., .% 9 ', m 5; - .f< < x . . .U.-.4 ~ .3 re y ,. ; -.4'. . "O

.m.a, . f. . ... .#p . .- '.&. ..y.2; *, c g k , e. . .c..a...w..; n.... .. ..r.'a_m. ,

.r. . 1, .. t.s,.4.~. s. . ..y

. .. . .~..s.

. . . e

- . m.a...

< , . ac...: . #.a. . t . .. . . .

. . . . . m

~ a : .;y f.:, a :.

w:.-wh,;7:. .-

~. .. . > u w. e  : a u- M k- 0)

.s - :.x@a n , cw%.%: n P(.;h- O

.:m %. +- 9: W. %.

Q gf) QW,.af.;w:n:y: 6 p:Ws; t M

'S,h****

A. w s*' f.ckt f NA %,*, ,1.-e o n) ..

,d:a.

.y -*+i 3.,;&g. .$(* %c

" 4

. H,-* O

. Ap u. n d; 5,-

C Q q ,,

f. g:w' v > sb)%'*..u'n**..qs.w,

=

.; e . q% . s.,5;,.

~.h. k,z,y v

,. ~.

,A. q g ie n g;,.  %:: , n : g,,.

• !~N  ::. :m:.p, T ..

_ = = W=

..m.......~4..ii.#.. .. .

a a~.

9 e.,.p.e.4 s CI fp .i .c.., a.~ ,4 j,

... ...... ..g L *

. .. 4.,,

r p. y v

...y .n..,,,..~c...

p...s .

.y - $

W*

. a * ,a f* N 4. ; L* p., ..Q 4 f,#4%.% ( f .. g . ;e 3.

N Io % Ge f'W

$ .p x ,. y9 My %&.

,z. w ~Ly.:a:.c'h.s c c a.[4

.:p&.:  %'yq.r .: m p v .p g. .m. . qe.. .g w . . r.. .., u..;

n ...h . n -, .

.s s. s. s

.. +

p;.g ...-:e+

$y ..q o:..,;.,;p p

>. g.3:.3,x g w w n,a ,') g.ww:q% p . u.w e,

(.) ..c %n,,,,e%0.,.,g... o u, m,

,.u,. s...v 4 .~ . q 9., m. . % .v. w A n u

y Q .,

~3 Q g %y:rgyR;*psy*+ajyy&. a w ~he: p'ty m.d-~-L-ne~~: p:

nr.~~~~ v' : ;ym ^ .*W 0) 1 q.,

%,,,e p~;p%,-d

~.% ~' K y.

w.:t..j"qtn:~.W;; m.J Q 9,y G 14 is

..~.4...9w..  % '. Wi l % at.L % :

4 r

.c _a w us.~.r _. ai'fN .,cg

r  ; y p:

..5.%..

-w.y_:. m. p 7.,_: , _r.m9_Y_-~ ...

_m.ww. ~. u.: ..f.~. .,m_.t.

• 4l.*n%

, , .r+ {g'.,Qf. ,v. . +e y ff **  ?

n .w F,wllF,'..'g ,, ,. i e,fc r :I . erv, . n , Q, Qyy.j 0*,, , "%; Q g

a) a) o> o> m m o

m o

m o

oo o o o o o o o o o o o o o

+ + + + + + + +

W W W W W W W W

m. e. N. r m to T N r e r n1g '46aeu3

l l

I i

I l

1 Measured Level from instrument Zero, inches  ;

O O O O O O gr3 O,. st.)

, m Lf) O a e e M gc n w.

i

• 4* M. e %vt '; v. .;. , 4. . . .

W O

l j

fawm%a.c. t:g ,, ;4N k .. .

h:w

%. wWxyg% x ;x f;p4w  %. *4 mw;a- %w?%:~.&.y s.'" C;kjgom':s . %w. .v

'\ ~ i

. p t M l

%'. ne

.;pp:y.t.n.

c4.. %. g .h . ..e@w., M.NQc . , e m .- ,

e3...a..

u.;y...aa.n..?< . ' ...a l 9...,....W  ?. .s .: . m.. .J..

W...

1

'.A.. a, O < . . . . .

y

,ycs m

.y- w: .ct ar.m. .e:ww

, O n y, .: w u ~ w. i. -s...

y i

m. p . .a.y.. .. .,m.m. . m..m2 W Qd.W.% l% &.

Q . o W,. ,

J.  ; t ;J%;GW L <

h -

w w e ~ac,e. m ; n:,., f, e x; q -

c' O

%.n.,  ; w ys,% :A:q.n%.[.

N-q bm.,4 m ..

s n yp 3 e.m :w ;p - M

[N ,

mew W n. WW p Wyv.eWj h..v. M. . n.n.M;ww;skn

.. r. .

s%

s

&. . . .m.. .... '.. . A...%... .r... a, u o*

W.w , ;%cw;.c...w ue. q e - .w nw wc *ys..O Lp.

.....w s. a.a s

vv. .... . o g

OWg m .p a, 1.e<p:.:s:.mu:.%,

e 6f%%. w2 y.71.W.l , g,' glj,',

e, .wcW:k:yn z% n :u mw. w. ;'.g.7-4.g. a  : ~

f.d " -

~

.- N 3w.~,:qu3.,7 : .

O w ;.N, -

wp. w6. %w,o.,.at.r

1. r4 k+,^.hsc ,y*aw L ' . ,.-ut,;a* a;'w,:y.y~.

C' . . w:

..,.4 W:a.b,g~

a. . d..-QE.

.,M~ ~T. v,ui

n %.b,,.4 Q un wn.

1. . +,y m's W t

,b. 'i v. ..Q ^ K;l $

y

@ c i

O=m; W

s$msf M .,

.i.7t M; ,a, p. gc ;#.sd,:th:,n.t. *a..w. ..,....a. . . ., , W.. . .W .. .;.w.. . . .r..p e.a. a3 3 W m.n o f r .,.. . , , y ,%n... ,...%.. mevv

.w s Q su. 4,......- .

a p: '

wn s e

@ O

,,W u sonu ., m ;v 9

p:%us.:ue s W:'.;;r L g;a%.c.~...,

w n. v.m n, e- 14 J

@w. ~. yMW +W Y .A O %e%d?;*Rt. xyM&gS1%m m"p . mci eO.:-- W '>

apq w ^:

V C

\

Q t.: .

l

%, 4MWdM;.f+.: f@ MiR+ W M

%s w

.u -,.h m.m.

p z.m.3 .. nv.

mar . . .:p n%,M.p'f.,%(v?

. r.

.p gw./ m. ; '%m';M.-

m)a$#,,k.w%

.Q>;;

n.

Mc in

.A. s o.~ ..,..,.

0 t iu. :s .n. . . .m. .

a . a:. . . ., a. , .m . W ...,.y .. . Q , I "F'nr ma $.gv S m pp .g y . y, p <> sw Mikn.q.:v.yh. OO Q$DMdbb l h* h ~ 6-( @% $ N N hM N'.

o. .. Wm MW3%:5 y.N.:M.t:12. Wr ..o #e"4.6;;,tauc

C.yr e g ,Q)

S -

m 3 Lgah@4.I c%  %.tq

.}Y@y~n#j%

h-. , mwg W

%. . N - h. m sm"4 M.v. up p. -

< em M * >

,_ i

%@ca,.y- ky W =a +p W

%:#.. p .W,, d. _pr .h-. Wg

[

g @ M.. f@

. h.. %..en* b..:: AN :.m.M..W.m%.i .o.

g* W",

.a

&r-.h.9 .v Q;

W,r.

• • n 5 . ;d h . ,

c Q)

,,,j

%< g M.,P .. S -u'" 'G ww& '/ *r4?c Q n @W'@. u f:w:pfO at{u$p@ ,'.,f,% , M

t.,';y@&p c5

• nW5 q $vt cI t N;rgW.te a

• Q @h #:hX S hpg;& Ki W if"ikp.y*fn R f , ,.h? m+.

.L & 8 Lc>

Q E.  % ,.C;M*2 W %gE}M 1 n. f m m .u.,v O m&n m c)

>g - %pwWx.~,4%w.d.e.4...

w m m%m....d

e. ~ x .s .o. ,.r w

..w.. .. . . y , ,4. . p%

6m,s.;.g!ugw~%y:

za.m. .~ n..r.c. A..,.j n.., n h,i.m. ,x.. ,.

,n2. . . - .

lJJ Ly v &ww.vw.w: p#

ch:n.. .w.f&m.W3+w%w,s,4t;m.

n-- n m a a&w:a.:", M -M I Q &. m w ,v.w.a. g a? ww: . -

g_a s m,

,c mmx-x_,p m' os s-

. .m ,.&_p%ga.C_

s fy - c Q

% +4, n, y

&.: ,s, a.M, &m %m . ua s me..y. <

ny Mo. .2 m v-=

0 g M yr1tmh. W>. m .~.w q qnsy W W rs n Q 3%%.o4Q 4 M @w...&~~n m

oum.

a_)

y w&vwe:w.:

s gm .u : pw2 &

.C h @e.w@@N$y; f e $f M & ,.7.w~n.1..,m..y u ,v , d. w.s g .

.~-v nyM (( N l @..w 4.

I

%M s, p ..w .n

.vw ,..,. J. - m

.e.'-&,; w#

Q n.y.m:.

e' egyg

~ m: wg w a M.,w..k mnv mi#. Ee. ;u. .ww

% a*a .n #A y a . m(+ 4 ,a"m"t,~.

8 '+=**^

m. ..,.q.,

80'd g m s m. 3._37.,w q c

• qw m. -.p .... s.... ..

- ...a e<

Eqif

• • t..ag m**pzp.fM.* ne swkg *sm ~..cm. a..-..usu n)."ne 0 .p74.'

A.w(*. 4 s 46F wy.- 'f"" l

@ u -

',%[,3lw'ag%:

t' "

Wr GM hMiG$m Aw m.4 p5}"%s:n

% '. .W< . . 9 1c .e 4,,. m< I y e..g , %.4. nm.e

.d.ag . om:

ye . a. n%., ._,qqwp,m,q,g,_ , y,q g g o

.u e , ,w . y.

g O O O O O C O \

O O O O O CD to 4 N O. i P"  !

B.ISd 'OJnSSOJd \

l l

1

\

l 1

1 i

l

r .

Measured Level from Instrument Zero, inches o o o o o e o o o e o e

,- o e e a o u. T o. o. ,O ,

.n % O

-i ;, ..:

s+~ : ...o i..i..r. .ia

_x - .. .> . 2 p: . i ,- m ay*. w . n . . n_s - W M F; v. . . r. .

. . . n- ..

. . A4% 3. , .

+ , . / :

m m$ & x,M:.0^;'p

$ . K,:ged. N.., a. w .w ..m.... i. , O

.. 4, W w a : w e.'p) c.% , ~ .

+ e- m n s O .c

.w.w w:+

w y ..u my, r:nw.a se-7: #.* d e wsw m to .. .:As=. .:.n,.m.

. - - e: . -

m m . m g na vg: a,m g, ;m m;w. um C v.

n ~p :z. . . .

e o C N . m. s,. e+ : -

w s,. ww .

w.-..,.e_:n e.e.r m..rm.Om o_

. . . .. m, . u . . . . aa. .. . .. . u .. . . . . u .n . ; . . . .a . #......

.~ O r O *w*, %wasmmrys%o#.2..%.~n..:.wsmrpmeie to..

g- .c

.e OM c .y#v, , p.w3. a. c.

., %. ,.m, w . w. v .

m

.& pe!si ,: .WW%mWW*. . . w4 m ww ..m..- ,.a.w

m.n.,.

.y

.N m9

w. yn . c n W y.

m.

3 ,

+-e C 0+- .$  %: :,J .

? T,, m PO g,

. . ... . ..e..c.. ww

. .m. .a.y ... . .<. . . m. . .evy n%. .,. ++ m.. .c. Q; .,,p.-

.. .~.. c

• en a .

. e . r . w_. M., . .* e& uAg:p - p1.; ?. . s.

o..,...~

.......%w. ': . ...s b,;

s:: <

m c3 a.

w~t .W.:. M M,

< EO u@ . ' s, ns . e

• r .

O

-S,"

q%;e. QQ MW.:q w up ,. .:+  :, LJ.M ;z attu.n e.eg% .. r:c @: v t w; . ,3 . 2 7:a ..:@s :s % m

,~

r w

se ee . .v py axyy e.s.:m,. r.w. . g... ..+.s.

,, < .: p*

o4 1 7 wmn.e. @ s s

s. .. g , . -

Q U) l> .

b ...-.~.s...e...a,.~...a.a. .u. .. c . ,o . e +

c ...... &.E.L

~ . .m. IlI" es.- g+ ,mg.

.6c ; w, .nh ./q sa ;ms ea m n a:- O .,~ O Ts W

.~ . .

m n..mx c c::s w,.c,s.wsu

~ n> .n. 7. uv. . v,q . .. Q.- O C O K *Q

"' Q X.)R:v r*%

n% W ' %w  %-

v' M ~:r':b U+# c.m~a en:;~sm:"~ ~ "'+"0 ..u,A  : ';;

• D m.,L.....

4

• 1**= s e.se.--'k O S O m (,) p

.15 MWN.

C 6 '6': " #W-

"W*y@*'M"'"F'N W'MM,s"W

" n.? i~ M re: ,w -~

a We N

kh,e

@ M g b%A.. m y% -

-1.

ah4t,s:. W..%W..,W.m p .J

g. b"~

g.qdh.

.. s.

"b+~[b~"N!bh.m.

..d.

a

. p, ys .- -.e..

t ':5 e-w.g4 w w :-Fr g;..s.:#Qk

.t3g, m

x -

g e

1. =  :>

0Q c.4 .. .n y 3.u._4..  %:cM 40.D;Sv p ?

9MM y

s. c4.%Q p .a . d... . W ... WqqQ% q.m %,W:e,yo. e,q. q.c O .C. .O

.u, .... &._ %.o.m &..:. . . .. g . J ,-

.{ C FMM W a ,. gp _ to

3. m.W~.A. sg % a m.. h .cA &h w %: ; e T:e a.4rn w...

@  ?::, s+Lm.m: o e

n a# u. m

> g m.,.;.q u,::

s n e Wp.am a

t.ewg

.w.n: Sa H9 m m% p rh,y

v. .?v,~

, m..au..b..y Ahn

.4p.ra.n s~L. a&.

p un :n ., e.w; s ..y;#,v.6,..L

%.,.4;;;,:.;w.g+n

. e O .&m%a:.a m

O c5. ;

na n w v ..s Wu .. ......u.....

y . . . . .i.sm...i.p.

. ....u=. g ll>

a,. ..

~wwv;we:p w4..m. Su w w a. m. m r- p:,-

t 4.,m. . gr tw* m w<

m.. Ms. m A m-ty . y.= .

Q pgW W

1 epu c a

t r p;. a g W" ,. ggly %e w-

~ T N ,C,, un_,p+e.Q: .# w...,J. Q.. % %,, y& m. w.e.v' bpt,g <C$u.,

o .m., m q, y'.

. : . .y,u.mxm. ,qp.9.c. .x4 %y,h~..:uy, 2..

n

e. ..y x.ma m.

.r.a .u . .pm 3., . . . w. c. . .....r .g.u. u. 9..

a. Mr19

.. nJ O O w .. o. y n..p.,y m m...t &. ,r vO

. . u. w T. .v h ;sym, .

ew we.pf .; p, n.q..r. 4 ,:n s u g.m-.ca:p ._

e M o>s

.s. e.w

,g tw , w v-m pt.e . u.4 . w mas m9a

• w ,%:.,e. ,w: N.y.n- g %

am tus w

,.*..+... p.g...mm::d g;,..a,...-..h t.%,rl4 4 .

L1 ,

N D. -'f.y&. kw.>+.=

gg,,p w.

sf a; A4 ** p % ys *4

• '* fhs.e. , , %3 e'

• s e4, _ y

• v

, a.p_ a .%l9 x.p ipn.

y % p)*5'OQ;s1..&yyg m.p.&

y'y,y'+.'

. Ar .

Q atm*

y. ' - .w* .n, p ^t}g' a.f f. w

.s w e . . ,..

b,y. p.v.g.x..,t as s %p

m. ,y gas n vy aa.

g.m.

.g... .w ,~.*4 .

Q# ._ eTs u . g. . .,.. . . ~..,u. . - O

=== .e.se o.%.e u a .,

$ h,,.. b.s. d , . aa yya y

• g g* !;v.84.$pu;.. . K,, g

.e -

e. .v u e.. ump

$"{ ' f%p,e k ' c ., a > T.', h4.

n.s_(:g g W .g+M. ..* ,,4,,p

[J D N %jN/prVt>.she M"

' .rMWh[*

wc . 1A .pe# . sgg g W % ; m % ./b v NIN s .,3y' 47 N PN,4. i-J 7- dlhMMI41*4d+ NMWPhi i h iM.$'.9 O O O O O O O O O O O O O O O O O O O O O- 03 CD b- (O LO M' C9 N v-

'~

elsd 'eJnsseJd

REFERENCES

[GL 88-20) " Individual Plant Examination For Severe Accident j Vulnerabilities - 10 CFR 50.54 (f)", Generic Letter 88.20, '

U.S. Nuclear Regulatory Commission, November 23,1988 )

[NUREG-1335) " Individual Plant Examination Submittal Guidance",

NUREG-1335, U. S. Nuclear Regulatory Commission, i August,1989

[lPE) Hatch IPE Submittal (Revised) December 11,1992 i [lPE Responses) License Responses to Questions from Review of the i Submittal, Letter from J.T. Beckham, Jr., Georgia Power, to j NRC, HL-4690, October 7,1994. j

[UFSAR) Updated Final Safety Analysis Report for Hatch Unit 2

[ Tech Specs) Technical Specifications for Hatch

[NUREG/CR 4550, NUREG/CR- 4550, Vol 6, Rev 1, Part 1, Analysis of Core  !

l Grand Gulf) Damage Frequency: Grand Gulf, Unit 1 Internal Events

[NUREG/CR 4550, NUREG/CR- 4550, Vol 4, Rev 1, Part 1, Analysis of Peach Bottom] Core Damage Frequency: Peach Bottom, Unit 2 Internal Events

[NEDO 24708A) AdditionalInformation Required for NRC Staff Generic Report on BWRs, August 1979, Rev 1 December 1980

[lPE, Browns Ferry) IPE Submittal for Browns Ferry, TVA

[lPE, Cooper) IPE Submittal for Cooper, NPPD

[TER, Quad Cities) Technical Evaluation Report of Quad Cities IPE Submittal

[TER, Dresden) Technical Evaluation Report for Dresden IPE Submittal

[lPE,1/94 Letter] Letter from J. T. Beckham, GPC, to NRC, "Edwin 1. Hatch Nuclear Power Plant, Revisions to Original IPE," HL-3491, January 10,1994.

l l

l 52

.y .,.

HATCH UNITS 1 AND 2 INDIVIDUAL PLANT EXAMINATION TECHNICAL EVALUATION REPORT (BACK-END)

Enclosure 3 j l