ML20136G217

Operating Experience Feedback Report.Assessment of Spent Fuel Cooling

ML20136G217
Person / Time
Issue date:	02/28/1997
From:	Ibarra J, William Jones, Lanik G, Ornstein H, Pullani S NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
To:
References
NUREG-1275, NUREG-1275-V12, NUDOCS 9703170237
Download: ML20136G217 (48)

Text

. - . .- - - _ - - - . - . - _ . - - - . . - .

1 NUREG-1275 Vol.12

+

. Operating Experience Feedback Report Assessment of Spent Fuel Cooling l

U.S. Nuclear Regulatory Commission OfTice for Analysis and Evaluation of Operational Data Prepared by J. G. Ibarra, W. R. Jones, G. E Lanik, It L Ornstein, S. V. Pullani DW3}

f[ W ,j o a llR 4822?

• 22" 1275 R PDR cianssp. ppa

_ _ _ _ _ _ _ . _ _ . _ _ _ - . _ . = _ . _ _._. __ - - - - - . . _ . . . - - - -

L 1

AVAILABILITY NOTICE Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be avaliable from one of the following sources *. ,

1. The NRC Public Document Room, 2120 L Street, NW., Lower Level, Wr.shington, DC 20555-0001

2. The Superintendent of Documents, U.S. Government Printing Office, P. O. Box 37082, Washington, DC 20402-9328

3. The National Technical Information Service, Springfield, VA 22161-0002 Although the listing that follows represents the majority of documents cited in NRC publica-tions, it is not intended to be exhaustive.

Referenced documents available for inspection cnd copying for a fee from the NRC Public Document Room include NRC correspondence and internal NRC memoranda; NRC bulletins, circulars, information notici, inspection and investigation notices; licensee event reports; vendor reports and correspondence; Commission papers; and applicant and licensee docu-

]

- ments and correspondence. i The following documents in the NUREG series are available for purchase from the Government Printing Office: formal NRC staff and contractor reports, NRC-sponsored conference pro-coedings, international agreement reports, grantee reports, and NRC booklets and bro- )

chures. Also available are regulatory guides, NRC regulations in the Code of Federal Regula-tions, and Nuclear Regulatory Commission issuances.

i Documents available from the National Technical Information Service include NUREG-series reports and technical reports prepared by other Federal agencies and reports prepared by the ,

Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission.  !

Documents available from public and special technical libraries include all open literature items, such as books, journal articles, and transactions. Federal Register notices, Federal and State legislation, and congressional reports can usually be obtained from these libraries. l Documents such as theses, dissertations, foreign reports and translations, and non-NRC con-forence proceedings are available for purchase from the organization sponsoring the publica-tion cited.

]

Single copies of NRC draft reports are available free, to the extent of supply, upon written i request to the Office of Administration, Distribution and Mail Services Section, U.S. Nuclear Regulatory Commission, Washington DC 20555-0001.

I Copies of industry codes and standards used in a substantive manner in the NRC regulatory j process are maintained at the NRC Library, Two White Flint North,11545 Rockville Pike, Rock- l ville, MD 20852-2738, for use by the public. Codes and standards are usually copyrighted j and may be purchased from the originating organization or, if they are American National Standards, from the American National Standards institute,1430 Broadway, New York, NY 10018-3308.  !

[

i

NUREG-1275 l Vol.12 4

Operating Experience Feedback Report

Assessment of Spent Fuel Cooling  ;

Manuscript Completed: February 1997 Date Published: February 1997 J. G. Ibarra, W R. Jones, G. F. Lanik, H. L Ornstein, S. V. Pullani I

l I

l l

Safety Programs Division Omce for Analysis and Evaluation of Operational Data U.S. Nuclear Regulatory Commission Washington, DC 20555-0001

.p"..s

..... /

ABSTRACT This report documents the results of an team assessed the likelihood and consequences independent assessment by a team from the of an extended loss of SFP cooling and Office for Analysis and Evaluation of suggested corrective actions, based on their Operational Data of spent-fuel-pool (SFP) findings.

cooling in operating nuclear power plants. The 1

4 iii NUREG-1275, Vol.12

CONTENTS Page i

l A B STRA CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i ii AB B REVI ATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix EXECUTIV E S UM M ARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi 1 I NTRO D U CTI O N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 1

2 S PENT F U E L COO LI N G . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 2.1 Sy ste m Desc rip t ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2.2 Loss of Spent-Fuel-Pool Coolant Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 l

2.2.1 Connected Sys tems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2.2 G ates and Seals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2.3 Pool Stmeture or Liner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.2.4 Consequences of Loss of Spent-Fuel-Pool Coolant Inventory . . . . . . . . . . . . . . . . . . 6 2.3 Loss of Spent-Fuel-Pool Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.3.1 Loss of Spent-Fuel-Pool Cooling System Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.3.2 Ineffective Spent-Fuel-Pool Heat Sink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.3.3 Consequences of Loss of Spent-Fuel-Pool Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.4 Preventmg and Responding to Spent-Fuel-Pool Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3 OPE RATING EXPE RIENCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.1 Loss of Spent-Fuel-Pool Coolant Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.1.1 Connected Systems . . . . . ............................ ............... 10 3.1.2 Gate s and Seal s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.1.3 Pool Stme ture or Liner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.1.4 Spent-Fuel-Pool Make-up Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.1.5 Impact on Safety Equipment .......................................... 13

. 3.2 S pent-Fuel-Pool Cooling . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 3.2.1 Loss of Spent-Fuel. Pool Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.2.2 Ineffective Heat Sink . . .......................................... 15 3.3 Spent-Fuel-Pool Instmmentation Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.4 Ve n tilati on Even ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.5 Review of Foreign Operating Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.6 Operating Experience Review Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4 OBSERVATIONS FROM TIIE SITE VISITS AND INTERVIEWS . . . . . . . . . . . . . . . . . . . 18 5 REGULATORY REQUIREMENTS AND GUIDANCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 y NUREG-1275, Vol.12

Contents Page 6 ENGINEERING ASSESSMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 6.1 Electrical Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...................... 23 6.2 Instru mentation Assess ment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 6.3 Heat Load Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 6.4 Radiation Assessment ............................. ....................... 27 7 RISK ASSESSMENT ............. ........................................... 28 7.1 Existing Probabilistic Risk Assesment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 7.1.1 NUREG/CR-4982, " Severe Accidents in Spent Fuel Pools in Support of Generic Safety Issue 82" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 7.1.2 NUREG-1353, " Regulatory Analysis for the Resolution of Generic Issue 82, 'Beyond Design Basis Accidents in Spent Fuel Pools'" . . . . . . . . . . . . . . 28 7.1.3 " Risk Analysis for Spent-Fuel-Pool Cooling at Susquehanna Electric Powe r S tat i on " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 7.2 Risk Assessment . . . . . . . . . . . . . . . . . . . . . . . . .................................. 29 7.2.1 Risk Assessment-Quantitative Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 7.2.2 Risk Assessment-Qualitative Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 8 FINDINGS AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 8.1 Likelihood and Consequences of Spent-Fuel-Pool Events . . . . . . . . . . . . . . . . . . . . . . . . . 32 1 8.1.1 Loss-of-Coolant-Inventory Events . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . 32 '

8.1.2 Possible Consequences of Loss-of-Coolant Inventory . . . . . . . . . . . . . . . . . . . . . . . 32 8.1.3 Need for Specific Corrective Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 8.2 Prevention of Spent-Fuel-Pool Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 8.2.1 Configuration Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 8.2.2 Plant Modifications at Multiunit Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 8.3 Response to Spent-Fuel-Pool Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 8.3.1 Operator Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 8.3.2 Procedures and Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 1

9 RE FE REN C ES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 l

FIGURES l

2.1- PWR Spent Fuel Cooling Systems ....................... ........................ 3 2.2 BWR Spent Fuel Cooling Systems ............. .................................. 3 2.3 Loss of S pent Fuel Coolin g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 3.1 Loss of Inven tory Duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 3.2 Loss of Inventory Levels . . . . . . . . . . . . . . . ....................................... 10 3.3 Loss of Cooling Duration .. . ................................................ 14 3.4 Loss of Cooling Temperatures ......... ........................................ 15 6.1 History of Full Core Offloading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 6.2 Red uced Ti ne to B oil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 NUREG-1275, Vol.12 vi

Contents Page TABLES 3.1 Spent Fuel Pool Events ......................................................... 9 3.2 Loss-of-Coolant Inventory Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.3 Loss of Cooling Events ....................................................... 14 3.4 H VAC Syste m Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.5 Events at Foreign Plants ....................................................... 17 6.1 S FP In strumentation S ummary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 6.2 S FP Heatup Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 6.3 Radiation Shielding Estinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 7.1 Near-Boiling Frequencies ........................................... .......... 30 1

vii NUREG-1275, Vol.12 I

i A

d ABBREVIATIONS AEOD Analysis and Evaluation of Operational Data (NRC Office for)

BNL Brookhaven National Laboratory BWR boiling-water reactor CDF core damage frequency CFR Code ofFederalRegulations ECCS emergency core cooling system

~EDO Execut:vo Director for C perations GDC. Genersi iksign Criterior./ Criteria HVAC heating, ventilation, and ir conditioning

-INEL Idaho National Engineering Laboratory LOCA ~1oss-of-coolant accident LOOP loss-of-offsite power NBF near-boiling frequency NRC Nuclear Regulatory Commission (U.S.)

NRR Nuclear Reactor Regulation (NRC Office of)

NSSS nuclear steam supply system PNL' Pacific Northwest Laboratory PRA probabilistic risk assessment PWR pressurized-water reactor RCS reactor coolant system RiiR residual heat removal SFP spent fuel pool

~SRP Standard Review Plan i

1 ix NUREG-1275, Vol.12

e.e EXECUTIVE

SUMMARY

As directed by the Executive Director for performed model refinements that resulted in Operations, the Office for Analysis and better estimates of near boiling frequency Evaluation of Operational Data (AEOD) (NBF). Although INEL performed no performed an independent assessment of the quantitative estimates of core damage, the ,

likelihood and consequences of an extended loss analysis provided qualitative insights for l of spent-fuel-pool (SFP) cooling. The overall identifying improvements to SFPs that may  !

conclusions are that the typical plant may need lessen the risks of events, improvements in SFP instrumentation, operator procedures and training, and configuration Findings from these assessments are as follows:

control.

From reviewing more than 12 years of The AEOD staff conducted six si visits to gain operating experience, the staff determined an understanding of each license SFP that loss of SFP coolant inventory greater physical configuration, practices, and operating than 1 foot occurred at a rate of about 1 ,

procedures. During these visits, they found event per 100 reactor years. Loss of SFP l great variation among the designs and cooling with a temperature increase greater capabilities of SFPs and systems at the nuclear than 20 *F occurred at a rate of plants on these sites. approximately 3 events per 1000 reactor years. The consequences of these actual In November 1992, Mr. Donald Prevatte and events were not severe. However, these Mr. David Lochbaum submitted a defects and events resulted in loss of several feet of SFP noncompliance report on the Susquehanna SFP coolant level, some of the events have lasted to the U.S. Nuclear Regulatory Commission longer than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The primary cause of (NRC). The AEOD staffinterviewed Mr. these events was human error.

Prevatte and Mr. Lochbaum to better understand their concerns. Their report, which has potential

• During review of existing SFP risk generic implications, provided the impetus for assessments, the staff found that after the NRC and the nuclear industry to take a correction for several problems in the closer look at SFPs. analyses, the relative risk produced by loss of spent fuel cooling is low when compared AEOD reviewed the applicable SFP regulations, with the risk of events not involving SFP.

the applicable acceptance criteria in the NRC The likelihood and consequences of loss of Standard Review Plan, and the applicable SFP cooling events are highly dependent on Regulatory Guides. Because the criteria human performance and individual plant evolved and each reactor was licensed over design features.

time, the criteria varies for evaluating these SFP designs.

• The staff determined that utilities' efforts to reduce outage duration have resulted in full The AEOD staff performed independent core off loads occurring earlier in outages, assessments of the electrical systems, This increased fuel pool heat load reduces instmmentation, heat loads, and radiation from the time available to recover from a loss-of-which they determined the typical SFP SFP-cooling event early in the outage.

configurations and potential problems.

Actions recommended by AEOD based on these Utilizing a previous Susquehanna risk analysis, assessments are as follows:

Idaho Nctional Engineering Laboratory (INEL) xi NUREG-1275, Vol.12

4 i

, Executive Summary

• The need for corrective actions at each plant should be evaluated on a plant-specific

where failures of reactor cavity seal or gate basis.

l seals, or ineffective antisiphon devices l could potentially cause loss of SFP coolant

• The need for improved procedures and

! inventory sufficient to uncover the fuel or training for control room operators to endan'ger makeup capability, should be respond to SFP loss-of inventory and SFP evaluated. loss-of-cooling events, consistent with the time frames over which events can proceed l

i

. The need for improvement to configuration and recognizing the heat load and the controls related to the SFP to prevent or possibility of loss of inventory, should be

, mitigate SFP loss-of-inventory events and evaluated on a plant-specific basis.

loss-of-cooling events should be evaluated on a plant-specific basis. - The need for improvements to

! instrumentation and power supplies to the

• The need for plant modifications at some SFP equipment to aid correct operator j multiunit sites to account for the potential response to SFP events should be evaluated

effects of SFP boiling conditions on safe on a plant-specific basis.

shutdown equipment for the operating unit,

particularly during full core off-loads, j

I i

i i

1 1

4 NUREG-1275, Vol.12 = xii

a Performed six site visits to gather 1 INTRODUCTION information on SFP physical configuration,

, practices, and procedures; and conducted Several instances occurred in necent years in nterviews with Mr. Lochbaum and which the adequacy of SFP cooling systems Mr. Prevatte to better understand their 10 were questioned. For example, Mr. David CFR Part 21 report.

Lochbaum and Mr. Donald Prevatte, former Susquehanna Steam Electric Station plant

• Reviewed applicable SFP regulations, contractors, subnu,tted a report (Ref.1) in applicable acceptance criteria in the NRC accordance with Part 21 of Title 10 of the Code Standard Review Plan (SRP) and applicable of Federal Regulations (10 CFR Part 21)

Regulatory Guides.

(Ref. 2) on the adequacy of SFP cooling at the Susquehanna plant. In addition, the agency .

Performed independent assessments of corresponded with Mr. Lochbaum and Mr.

electrical systems, instrumentation, heat Prevatte on this topic, and reviewed and loads, and radiation to better understand responded to a 10 CFR 2.206 petition from their effects on SFP cooling.

them. As a rest.!t of the issues raised with respect to SFP,, on February 10,1996, the Contracted with Idaho National Engineering Executive Director for Operations (EDO)

Laboratory (INEL) to review existing risk requested tha; the Office for Analysis and analyses and to use risk assessment Evaluation of Operational Data (AEOD) techniques to evaluate the risk oflosing SFP perform an indepe.: dent study of the likelihood cooling and coolant inventory.

of, and consequences of, an extended loss of SFP cooling (Ref. 3). On February 29,1996, In order to accomplish the stated goals in the 7-AEOD sent the EDO a plan and schedule month schedule, AEOD utilized existing agency (Ref. 4) for performing this independent study.

SFP data and analyses. AEOD worked closely with the Office of Nuclear Reactor Regulation The 10 CFR Part 21 report filed by Mr. (NRR) throughout the assessment to be able to Lochbaum and Mr. Prevatte postulated loss of use current SFP information. Specific analyses SFP cooling that resulted in boiling of the SFP, and specific sites were assessed for generic failure of the emergency core cooling system applicability to other plants.

(ECCS) and of other equipment that was due to steam releases and condensation of SFP vapors, reactor core heatup and damage, spent fuel 2 SPENT FUEL COOLING heatup and damage, and large offsite radioactivity releases. A survey of SFPs indicates that a wide variety of configurations exists. This section provides AEOD completed the independent assessment. simplified general descriptions of SFP during which the staff- configurations; the descriptions may not apply to any specific SFP but are considered to be

• Developed generic configurations typical or " generic" PWR and BWR SFPs.

delineating SFP equipment for a boiling- Since most plants were built before the NRC water reactor (BWR) and a pressurized. issued specific regulatory guidance for SFPs, water reactor (PWR) and utilized these diverse designs would be expected. For generic configurations to assess the loss of Purposes of this study and this report, loss of SFP cooling and inventory. spent fuel cooling is considered to include two subcategories: the loss of SFP coolant inventory

• Assessed more than 12 years of operational and the loss of SFP cooling. Potential problems experience for both domestic reactors and with SFP coolant inventory and SFP cooling foreign reactors with designs similar to that can lead to loss of spent fuel cooling and those in the United States. the potential consequences of loss of spent fuel 1 NUREG-1275, Vol.12

. l l

Spent Fuel Cooling cooling are discussed. Once the problems are and 2.2, SFP coolant is pumped through heat identified, possible approaches to prevention exchangers where sensible heat is transferred to and response to loss of spent fuel cooling an intermediate cooling system, which finally situations are described, rejects heat to the plant's ultimate heat sink. l The SFP cooling system takes suction from the l 2.1 System Descriptiert SFP through a skimmer or strainer at such an elevation that a water level change in the SFP ,

Figure 2.1 shows a generic PWR SFP and would cause the pumps to lose suction and l Figure 2.2 shows a generic BWR SFP. SFPs are prevent further SFP coolant inventory loss constructed of reinforced concreto, several feet through a break in the SFP cooling system i thick, with a stainless steel liner to prevent piping. The SFP cooling return lines either l leakage and maintain water quality. The pools discharge near the top of the SFP or are '

are designed to survive seismic events although arranged to distribute coolant flow around the the cooling system may not. For BWRs, the bottom of the fuel in the SFP. With a few I SFP is generally located within the reactor exceptions, SFP cooling piping, which extends I building. For PWRs, the SFP is located outside deep into the SFP, is equipped with antisiphon the containment but adjacent to it in a separate devices (usually drilled holes) to prevent loss of fuel handling building or within the auxiliary SFP coolant inventory should a system building. Typically, SFPs are abcut 40 feet misalignment or pipe break create an inadvertent  ;

deep and vary in width and length. The fuel is siphon flow path. SFP pumps, heat exchangers,  ;

stored in stainless steel racks and submerged and intermediate cooling systems are single with approximately 23 feet of water above the train or redundant, depending on the plant's SFP top of the stored fuel. The waterin the SFP of a design. Many plants have the capability to align BWR is demineralized water; whereas PWRs the residual heat removal (RHR) system to use borated water. In addition to cooling, SFP remove heat from the SFP in the event that the water inventory provides radiological shielding normal SFP cooling system is unavailable. Each for personnel in the fuel pool area and adjacent Pl ant has a nonsafety-related system that is used areas. Each plant generally has Technical to purify and clarify the SFP water. This system Specification requirements for water level and is often integrated with the normal SFP cooling reactivity of the spent fuel stored in the SFP. system. The system is typically made up of ,

filters, ion exchangers, and other supporting l Each plant has a source of high purity water to equipment. j fill the SFP, referred to in nuclear power plants as make-up. The preferred sources are usually Most plants have leak detection systems to the refueling water storage tank for PWRs and determine if leakage is occurring from the SFP the condensate storage tank for BWRs. The liner, spent fuel shipping cask pool, or from normal make-up is through a connection from other portions of the fuel pool or reactor cavity the water source to the suction of the SFP structure. The leak detection system is usually cooling system pumps or a water source. The made up of several channels that can be make-up rates amone plants have a wide range. monitored individually or are designed in such a Local valve operatious are needed to initiate way that leakage empties into drains that can be SFP make-up. Plants also have alternate monitored and returned to either sumps, liquid methods to provide make-up if normal make-up radioactive waste, or other cleanup or collection is unavailable and may include the service water systems. The SFP leak detection system can system and the fire water system. usually be isolated if necessary to attempt to reduce SFP leakage.

SFP coolant inventory is cooled by a dedicated cooling system. As shown in Figures 2.1 NUREG-1275 Vol.12 2

Spent Fuel Cooling Fuel Hanerg enng Cavey W N b@ Gate (s)

. +O Room - [e s ,

Fuel Transier g

~

& l Makow Fuel Racks non

• Purification ,

{"' & -

ase,ar EL*w*--+

_ cy,or L'"4 V

= -/ l l

c*

_ H**y w/ ',' / / s s , i ,' / ,'i;m a.e u w/ C***

'i,/ ,

1 more A=>

Figure 2.1 PWR Spent Fuel Cooling Systems 1

4 d

^"'

my j % EA ~.

[' . . . . ,

Ie7 skimmer -

k N..oinctierge Streiner Fuel Pool nefu sep[itrEor i

p s.enngei Puei nme

~

d b 3

44 K oeie6erca5 3 s-

• Purification f

5  ?

M"*T D neector Vessel b

+ sws"  :

N/

d HX F 9- ,

s l

Figure 2.2 BWR Spent Fuel Cooling Systems 3 NUREG-1275, Vol.12

Spent Fuel Cooling During refueling operations, the refueling cavity level) for each annunciator, and local conditions above the reactor is filled with water to equal usually need to be investigated to determine the the water level in the SFP, Fuel is moved from cause of SFP annunciator actuation. SFP the SFP to the reactor via transfer canals in instrumentation is discussed in more detail in BWRs or transfer tubes in PWRs. For BWRs, Section 6.2 of this report, the movable gates that separate the SFP and the transfer canal from the reactor cavity are several Some SFPs have a separate pool for spent fuel feet wide and extend approximately 24 feet shipping cask operations. Typically, these areas I down to provide an opening for fuel to be are separated from the main SFP by movable <

moved in a vertical position.' Removal and gates. At most sites, this area is open to the SFP replacement of the gates requires use of the only during cask operations. The spent fuel plant travel!ag crane because of their size and shipping cask pool usually has a drain system to weight. Thus, during refueling, a loss of water allow raising and lowering of the water level, from the refueling cavity resulting in a drop in water level would also lower the water level in Based on the SFP descriptions in Section 2.1, j the SFP. Replacement of the gates to isolate the the discussion in Section 2.2 describes potential l leak would be a major time-consuming scenarios that car. lead to loss of spent fuel j operation. cooling caused by (1) loss of SFP coolant inventory sufficient to interrupt heat transfer to For PWRs, the transfer tube provides for the cooling system or result in uncovery of the movement of the fuel in a horizontal position. fuel and (2) failure of the SFP cooling system The opening provides a much smaller flow path pumps and heat exchangers to transfer heat from from the SFP to the reactor cavity than the the pool to the ultimate heat sink. Figure 2.3 is movable gates of a BWR. Also, a gate valve at a schematic classification of the types of events the SFP end of the transfer tube can be closed that could lead to loss of spent fuel cooling.

fairly quickly to stop the flow path from the SFP to the reactor cavity.

Refueling cavity seals are installed between the e,y ,

reactor vessel flange and the bottom of the - --

reactor cavity to maintain a leak-proof volume --

l during refueling operations. Both PWRs and g* -

g ,

BWRs have drains in the refueling cavity area to allow draining of the cavity when the refueling -

is complete. Some plants have leak detection 8T*

• gg systems to monitor the cavity seal.

BWRs and PWRs have indicators for temperature, level, and radiation instrumentation c nne now in the SFP area. Analog meters are generally not provided in the control room for SFP level H m a* '

and temperature, but at certain sites, some SFP l parameters are available on the_ computer in the Figure 2.3 Loss of Spent Fuel Cooling l control room. SFP radiation values are generally available on analog meters in the control room. Control room annunciators for SFP parameters generally have more than one function (e.g., low SFP level and high SFP NUREG 1275, Vol.12 4

l l

Spent Fuel Cooling 2.2 Loss of Spent-Fuel-Pool 2.2.2 Gates and Seals Coolant Inventory A second classification of inventory loss is through movable gates or seals and, during The primary pathways for loss of SFP coolant refueling operations, the reactor cavity seal. As inventory can be broadly categorized as (1) loss shown in Figures 2.1 and 2.2, both PWRs and through connected systems, (2) leakage through BWRs have seals that keep water above the movable gates or seals, and (3) leakage through vessel in the refueling cavity during refueling.

or failure of the fuel pool or the fuel pool liner. BWRs usually require two seals to keep refueling water above the reactor vessel; in 2.2.1 Connected Systems Figure 2.2 these seals are referred to as the l refueling seal and the cavity seal. Some plants Piping connected to the SFP may include the use inflatable bladders to form a seal between SFP cooling and purification system, the spent the reactor vessel Gange and the containment fuel shipping cask pool and fuel transfer canal building (PWRs) or the drywell, and the reactor drains, and, when in communication with the building (BWRs). In some BWRs, these cavity reactor during refueling operations, reactor seals are permanent spring steel bellows that are piping systems such as the RHR system and the expected to have little susceptibility to large chemical and volume controi gstem.

leaks. Several other types of seals are used that do not rely on inflatable bladders. These Losses through connected systems could include include bolted cavity seal rings, which use both pipe breaks or leaks and configuration gaskets to seal between mating surfaces, and control problems. Piping systems that extend permanent seals, which are welded in place.

down into the SFP have the potential to siphon. These types of seals are not prone to rapidly For most designs, the loss of SFP coolant developing large leaks.  !

inventory via the SFP cooling system piping, l

whether initiated owing to a pipe break or a The refueling cavity seal and movable gate seals configuration control problem, would be limited at some plants are inflatable seals of varied by antisiphon devices. However, siphoning can designs. Depending on the physical relationship '

occur if the antisiphon devices are incorrectly of adjacent stmetures, catastrophic failure of an designed, are plugged, or otherwise fail. NRR inflatable seal could result in rapid loss of determined through a recent survey of all power inventory. However, the geometry of the reactors (Ref. 5) that some sites do not have relationship between the SFP, adjacent cavities, antisiphon devices in potential siphon paths. reactor vessel, and connecting structures must be considered in evaluating the vulnerability to During refueling operations, when a flow path loss of SFP coolant inventory caused by failed exists to the reactor vessel, inventory loss inflatable seals. Many seal failures will result in through the RHR, cl.emical and volume control only limited water levelloss because of the system, or reactor cavity drains would not be l various physical configurations. '

limited by the antisiphon devices; the same applies when the SFP is open tc the spent fuel In BWRs, the bottom of the movable gate shipping cask pool drains. For these situations separating the reactor cavity from the SFP is in many designs, the extent of the inventory loss generally above the top of the stored fuel so that is limited by internal weirs or drain path for a loss of the caiity seal, the levelin the SFP elevations, which maintain the water level above will remain above the top of the fuel. Although the top of the stored fuel in the SFP. the fuel would not b3 immediately uncovered, SFP cooling water would be lost because the SFP pumps would trip on loss of suction; and 5 NUREG-1275, V al.12

Spent Fuel Cooling the remaining SFP coolant inventory would heat 2.2.4 Consequences of Loss of Spent Fuel-up to near boiling within a few hours. Also, Pool Coolant Inventory because of the reduced water level above the fuel, high radiation fields would inhibit access For a large loss of SFP inventory, the primary to the refueling floor. Plants that have gate consequence is potential uncovery of the stored bottoms or internal weirs that limit the fuel. Given the unlikely occurrence of a large  !

draindown from a cavity seal or gate seal leak at the bottom of the SFP structure, beyond failures to a level that would continue to provide the available make-up capacity, the fuel could radiation shielding sufficient to allow operator become uncovered and heat to the point of clad l actions would be more likely to be able to damage and release of fission products. l mitigate these events. When not in refueling, Extremely high radiation fields would also l most BWRs have two gates in series at major result around the SFP area if the fuel were openings. uncovered.

i l

Where PWRs do not have interposing structures A more likely sequence would be a loss of l between the fuel transfer tube and the SFP or inventory through a gate or seal that would ,

where the gates between the SFP fuel transfer terminate when the level reached the elevation l canal are left open, a vulnerability to loss of SFP of the leak. Then, because of the decreased l

coolant inventory through the fuel transfer tube inventory of water in the SFP and the loss of l is increased. During the NRR survey suction to the SFP cooling system, the assessment, the staff found that five SFPs have remaining water in the pool would boil away fuel transfer tubes that are lower than the top of until the fuel was uncovered. Unless corrective the stored fuel without interposing structures. actions were taken, the final consequences would be similar to loss of SFP coolant 2.2.3 Pool Structure or Liner inventory described in the first paragraph of Section 2.2.4.

Finally, inventory loss could occur directly owing to SFP liner leakage or gross failure of Loss of SFP coolant inventory events for which the SFP stiucture. The impacts of a dropped corrective actions are taken before severe heavy (a load weighing more than one fuel consequences occur can potentially cause other assembly) load or a seismic event are potential Problems. Even a minor loss of SFP coolant causes of gross failure, although SFPs are inventory can lead to loss of SFP cooling designed to survive seismic events. because the lower SFP level causes loss of Radiological and structural response and suction to the SFP cooling system. Losses of makeup capability for dropped light loads (those SFP coolant inventory may produce flooding or weighing no more than a fuel assembly) are environmental problems in other areas of the bounded by analyses of a fuel handling accident. Pl ant. Ventilation and drain systems can On the other hand, dropped heavy loads have transport water and steam to other parts of the the potential to exceed the design basis of the Pl ant and affect emergency equipment. A fuel pool structure and the make-up system. significant amount of water vapor may be Thus, heavy load control programs have been generated either by direct boiling or evaporation instituted to evaluate the potential effect of a from the SFP. Various SFP equipment and dropped heavy load or to implement special ventilation configurations may allow the water controls on the design and operation of heavy vapor to accumulate on SFP cooling equipment load handling equipment, and cause it to fail, further exacerbating the loss of inventory.

Where the SFP area atmospheric water vapor can be transported to areas which house other NUREG-1275, Vol.12 6

Spent Fuel Cooling equipment important to safety, that equipment four SFPs could not be cooled by systems that may be affected. This potential problem is could be powered by onsite power sources.

important in some multiunit sites during and immediately following full core off-loads. In The likelihood of an extended loss of SFP these units, the fuel pool atmospheric water cooling caused by loss of electrical power to the vapor from the unit refueling can be transported pumps is fairly low owing to the combination of to areas housing safety equipment when the unit available on-site power, the existence of is operating at or near full power. This transport workable procedures for power restoration, and could cause equipment required for a safe the plant operations staff knowing that they need i shutdown of the operating unit to be damaged or to restore power and the time available to to fail. This issue is discussed in Section 7.2 of restore the power.

this report. Most plants have sufficient Hood protection, ventilation, and equipment For other than loss of electrical power, failure of separation to prevent this scenario. However, both SFP cooling pumps is unlikely. Except for according to the NRR survey assessment, eight situations in which a full core has been multiunit sites may be susceptible to this transferred to the SFP relatively soon after plant scenario. shutdown, a single SFP cooling pump generally provides sufficient cooling.

2.3 Loss of Spent-Fuel-Pool .

A loss of SFP coolant can result m a loss of Coolm.g cooling now when the level drops below the I suction intake of the SFP cooling pumps. Thus, Figure 2.3 also presents potential causes of loss such a loss of inventory will be accompanied by of cooling to the SFP, Cooling can be lost by a loss of SFP cooling.

loss of SFP cooling now or because of an ineffective SFP heat sink. Losses of SFP Flow can also be lost because of a blockage or cooling system now can be due to several diversion. For example, foreign material could mechanisms, including loss of electrical power clog a filter or strainer in the SFP cooling to the SFP cooling pumps, pump failure, flow system. If flow blockage were to occur during a blockage, loss of suction caused by the loss of full core off-load, implementation of a backup water level, or a diversion in the SFP cooling cooling process might be required to prevent system. Losses of heat sink can be due t adverse conditions from developing in the SFP.

operation with less than the required SFP cooling system complement or with heat loads 2.3.2 Ineffective Spent Fuel Pool Heat Sink in the SFP that exceed the capability of the SFP cooling system design. SFP cooling system heat exchangers are usually cool- by the component cooling water system 2.3.1 Loss of Spent Fuel Pool Cooling System o- . mervice water system. An ineffective SFP Flow i " sink can occur because of misalignment of

. ooling water sources, failure of the cooling All SFP cooling pumps are electncally powered.

water source, heat exchanger fouling, or i Loss of electncal power to these pumps results  !

insufficient heat exchanger capacity, among m loss of SFP cooling system flow. Loss of other causes. j electncal power can be due to losses of offsite j power or human error in electrical alignnrnts.

Current practice of full core off-loads a short Most SFP cooling system pumps can be loaded time after shutdown has greatly increased the  !

on available onsite power sources. During tne heat load in the SFP. Any degradation in the NRR survey assessment, the staff found that heat removal of the cooling system at these 7 NUREG-1275, Vol.12

Spent Fuel Cooling times could result in heating the SFP. Errors in properly tested and controlled to prevent loss. l the calculated heat load or assumption of Better seal performance could be achieved by nonconservative ultimate heat sink temperatures seal replacement at intervals consistent with could mislead operators. manufacturers recommendations or when inspection of seals shows evidence of aging, 1 2.3.3 Consequences of Loss of Spent Fuel. cracking, or tearing, i Pool Cooling The response to loss of inventory events An extended loss of SFP cooling would result in )

depends, first of all, on timely discovery of the ;

heat up and boil off of SFP coolant inventory event by the operator. The rate ofloss of SFP l and the eventual uncovery of the stored fuelin coolant inventory can vary greatly depending on the unlikely event that no corrective actions the cause; for examrie a drop in the water level were taken. This would result in high levels of from a reactor cavity sal failure can be quite radiation in the SFP area and having to prohibit rapid. The reduction in level during these personnel access to the area. Clad failure and events is usually discovered either by direct radiation release could be the final outcome. observation by operations staff in the spent fuel However, loss of cooling poses less hazard than area or by alarm actuation in the control room.

loss of inventory because loss of cooling does Reliable and accurate instruments and 1 not pose the immediate threat of uncovering the annunciators can alert the operator to a SFP fuel. No fuel damage is probable until the fuel  !

event. If the operators become quickly aware of is uncovered. a SFP event, the large volume of water in the l SFP will usually allow sufficient opportunity for During an extended loss of SFP cooling, water the operator to diagnose and correct the vapor may be generated either by direct boiling problem.

or evaporation from the SFP. Various SFP equipment and ventilation configurations may Response to loss of SFP cooling requires allow the water vapor to condense and effective instrumentation, procedures, and accumulate in locations that could affect other training. Most operating situations would allow equipment. All the potential effects that apply a relatively long time to respond to such an to the situation described in Section 2.2.4 for event. However, following a full core off-load, loss of SFP coolant inventory leading to the SFP could heat up to near boiling in a few generation of steam and water vapor being hours. Operators would attempt to restore transported to other parts of the plant apply to cooling either by correcting any problem with the extended loss of SFP cooling-the SFP cooling system or by initiating operation of backup cooling systems, if 2.4 Preventing and Responding to available.

Spent-Fuel-Pool Events .

As with prevention and response to SFP coolant No systems automatically respond to a loss of inventory ewnts, prevention and response to SFP coolant inventory or a Icss of SFP cooling. I ss of SFP cool,mg is also largely dependent on Consequently, operator actions form the basis e nfigurati n c ntr 1 and human performance.

The primary concern is to maintam electrical for preventing and responding to a loss of spent ,

fuel cooling. Power to the equipment involved in SFP cooling.

Both a gate seal and cavity seal must be correctly installed and tested in order to prevent a loss of SFP coolant inventory. And in the case of an inflatable seal, the air supply must be NUREG-1275, Vol.12 8

3 OPERATING EXPERIENCE While these events have been included in this report, they were not initially captured by the The staff reviewed operating experience with event review pr cess, primarily because some SFP loss of coolant inventory and loss of relev nt events are below the reportmg cooling, using, as the primary source of threshold required by NRC regulations.

information, licensee event reports from 1984 through early 1996 that were screened from the 3.1 Loss of Spent-Fuel-Pool Coolant Sequence Coding and Search System. In some Intyentory cases, events before 1984 were included.

Additional information sources included event About 38 events involved actualloss of SFP notifications made in accordance with 10 CFR coolant or refueling water. About 55 precursor 50.72, NRC Inspection Reports, NRC regional events occurred. Table 3.2 provides some moming reports, NRC preliminary notifications, details about loss of SFP coolant inventory and industry communications. Foreign events. Figures 3.1 and 3.2 provide an overview operating experience is discussed in Section 3.5 of the SFP loss of coolant inventory events in of this report. After reviewing more than 700 which the water level dropped and for which separate sources of information, the staff found duration times could be quantified. These about 260 events related to SFPs. Table 3.1 is a figures show that SFP losses of coolant summary of these SFP events and lists the inventory have been infrequent. However, number of events of each type under the two several events lasted more than 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and main categories: loss of SFP coolant inventory about 10 events resulted in water level decreases and loss of SFP cooling. The table shows of more than I foot before the event was numerous precursor events found during the tenninated. The low number of events found review. These precursor conditions represent with smaller level changes may be due to a lack potential losses of SFP coolant inventory or loss of reporting of such events.

of SFP cooling given the condition that occurred plus other postulated failures. Table 3.2 Loss-of-Coolant Inventory Events Table 3.1 Spent Fuel Pool Events Type of Event Actual Precursor Type of Event Actual Precursor Connected Systems 20 12 SFP Inventory M M Configuration Control 16 2 Siphoning 3 1 Connected Systems 20 12 PWR Transfer Tube 1 1 Gates and Seals 10 8 Piping 0 1 Structure or Liner 8 35 Piping Seismic Design 0 7 Gates and Seals 10 3 SFP Cooline M 22 Cavity Seals 0 6 Cooling Flow 50 20 Gate seals to 2 Heat Sink 6 2 j

Pool Structure or Liner J M  !

The operating events found during the review Liner Leaks 7 1 provide a reasonable representation of lead Drops 1 32 experience with SFPs. However, during Pool Seismic Design 0 2 discussions with operations staff, the staff leamed about a number of additional events that provide insights into problems with SFPs.

9 NUREG-1275, Vol.12

1 1

Operating Experience l NUMBER OF OCCURRENCES (in PWRs), the fuel transfer canal (in BWRs),

e and, during refueling, the reactor.

7 g Configuration Control. Sixteen loss of SFP coolant inventory events were due to  ;

5 configuration control errors. These events are  !

4 about equally distributed between BWRs and l 3 PWRs. Two recent configuration control events 8 2 2' are described here.

, uTw At Cooper Station on October 31,1995, about

<1 1To4 4 To o eTo 24 > 24 10,000 gallons of refueling water were DURATION (HRS) inadvertently lost from the refueling cavity and transferred to the plant's low-level waste system Figure 3.1 Loss ofInventory Duration (Ref. 6). At the time, the full core had been l

placed in the SFP, the reactor refueling cavity NUMBIR OF OCCURRENCES was filled with refueling water, and the

10. - -

refueling gates were open. A cable from a

" remote video camera came in contact with and el caused a submerged valve to open. The valve was part of the main steam line plug. This 6: allowed refueling water to flow to the main steam line drains. About 30 minutes after the 4: valve was opened, the SFP surge tank low-level ,

2 2 2 alarm alerted the operations staff to an ongoing j 2; ,.om loss of water. While the operations staff started  ;

'7" to add water, the make-up was not sufficient to

<a a To 12 12 To so > eo avoid tripping both SFP cooling pumps on low LEVEL DF. CREASE (INCHES) suction pressure. One SFP cooling pump was ]

restarted in about 3 minutes with no observed Figure 3.2 Loss ofIriventory Levels increase in SFP temperature. About 40 minutes later, the source of the inventory loss was Using the number of events found during this identified and the valve was closed. This event review over a period of about 12 years for which resulted in reducing the water level about 1 inch the decrease in the water level could be in the refueling cavity and SFP, a fairly slow ,

quantified, the frequency of loss of inventory drainage rate. More that 23 feet of water was i events in which loss of mcre than 1 foot still above the top of the fuelin the SFP. {

occurred can be estimated to be on the order of l

less than i event per 100 retetor years. At Millstone Unit 2 on July 6,1992, about i 10,000 gallons of SFP water was drained to the j 3.1.1 Connected Systems reactor coolant system (RCS). At the time of j the event, the unit had been shut down about 37 l The majority of losses of SFP coolant inventory days and the full core had been placed in the through connected systems was due to SFP. A loss of normal power resulted in loss of configuration control problems. These SFP cooling. During the response to the event, connected systems include the SFP cooling and the operations staff decided to align the purification system, a spent fuel shipping cask shutdown cooling system to provide cooling to ,

pool, sources of make-up, the fuel transfer tubes the SFP. However, during the alignment I l

NUREG-1275, Vol.12 10 l

Operating Experience process, a flow path was created that permitted to fill the SFP that created a siphon path when flow via a gravity drain from the SFP to the the pump was secured. In this event, about RCS. The SFP water level dropped about 21 feet 9 inches of water remained above the 14 inches. According to the information fuel.

' reported, at least 23 feet of water was above the top of the fuel because no Technical One precursor event was reported in which Specification violation was reported. A 4 *F antisiphon holes in the two SFP cooling return temperature rise occurred before the SFP lines were not present even though 0.5-inch cooling was restored (Ref. 7). holes were previously thought to exist. Also, further investigation indicated that the 0.5-inch Siphoning. Although reported operating holes would not have been adequate to stop a experience with siphons (both actual events and siphon, given postulated failures.

precursor conditions) is very sparse (three actual events), losses of SFP ccolant inventory have Pressurized-Water Reactor Transfer Tube. l occurred because of siphoning problems. One Only one actual event was found in which the event at River Bend on September 20,1987, transfer tube actually leaked while closed. In (Ref. 8) involved plugging of a single this event, the SFP end of the transfer tube was (nonredundant) vertical vent pipe acting as an open and the flange on the containment end of antisiphon device. In this event, the SFP the transfer tube leaked. AEOD was informed  ;

coolant loss was due to siphoning, but was during some site visits that minor leakage masked by the SFP low-level annunciator being through transfer tubes has occurred, in the alarm condition because of other ongoing I plant work. The event lasted about one-half One site (Oconee Units I and 2) has a fuel hour. This event was terminated when a transfer tube that has piping penetrations at a radiation alarm occurred coincident with a high level 6 feet below the top of the spent fuel in the level in the tank receiving the SFP water, SFP. This penetration is used during operation alerting personnel to the coolant loss. This of the Oconee Standby Shutdown Facility. This event resulted in a loss of between 5 and 10 feet facility has a mission time of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Water is of the SFP water level, one of the largest level taken from the SFP through the transfer tube via decreases found in the study. Further, how far the penetration and injected into the reactor the level would have fallen had no operator coolant pump seals for cooling. In this design, action occurred is not clear, continued use of SFP coolant inventory for reactor coolant pump seals could have caused In another event at San Onofre Unit 2 on June radiation doses in the SFP to reach such high 22,1981 (Ref 9), about 9000 gallons of SFP levels that make-up to the SFP would be coolant di, ted from the SFP to the reactor impossible. This problem has been corrected by cavity through the SFP purification system adding remote make-up capability to the SFPs.

because that system lacked siphon protection.

This event lasted about 5.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. The licensee Piping and Piping Seismic Design. No actual stated that this condition would be corrected by events were found during which SFP system providing siphon protection. The licensee piping actually leaked, causing a loss of SFP determined that the minimum amount of water coolant inventory., However, a variety of above the top of active fuel in the SFP would be seismic piping design problems have been about 13 feet if the operations staff failed to reported. The most prevalent type of problem respond to two alarms. involves use of the nonseismic SFP purification system for purifying the large sources of Another event at Davis Besse on February 1, _ refueling water in both BWRs and PWRs.

1982 (Ref.10), involved a temporary pump used Failure of the nonseismic SFP purification iI NUREG-1275, Vol.12

Operating Experience system while connected to the refueling water of potential failures. Most of these events source could cause loss of this source as involved design problems. Only one was due to make-up to the SFP as well as compromise these failure to maintain an adequate air supply to the sources as ECCS sources. In addition, other inflatable seal. One event involved a gasket minor piping seismic design problems were type (noninflatable) seal that leaked during the discovered and reported. Seismic analysis is draining operation following the refueling.

discussed in Section 7 of this report.

Gates. The second most prevalent type ofloss 3.1.2 Gates and Seals of SFP coolant inventory (10 events) was leaking fuel pool gates. The majority of these Large losses of SFP coolant inventory have leaks were due to failure to maintain the air occurred through SFP gate seals. Also, supply to the inflatable gate seals. In one case, potentially large losses of SFP coolant inventory the seal did not completely inflate. The could be lost through reactor cavity seals. majority of the air supply events was due to human error. Three of these events involved Refueling Cavity Seals. At least two rapidly failed or disconnected level instrumentation.

developing leaks were due to inflatable reactor Most of these events occurred at PWRs. Leaks cavity seals. In both these cases, the SFP was were generally large, involving tens of isolated from the reactor cavity by the closed thousands of gallons of water, and a decrease of j fuel transfer tube before the event. At Haddam 2 or more feet of SFP water level. The decrease Neck on August 21,1984, the seal failed and in water level rates ranged from fractions of a about 200,000 gallons of water were drained to foot per hour up to several feet per hour. These the containment building in about 20 minutes. drop rates can be dealt with and, in fact, in these At Surry Unit 1 on May 17,1988, with all the events, the operations staff responded and fuel in the SFP, the seal failed and about 25,800 restored level effectively, gallons were drained to the containment in about one-half hour. In the case of Surry, the One event, at Hatch on December 2,1986, instrument air supply to the containment was resulted in the fuel pool level dropping about isolated and a backup nitrogen supply was used 5.5 feet (Ref. I1). This event resulted from to reinflate the seal. Problems resulted in the isolating the single air supply to the transfer inflatable seal deflating enough to cause canal's six gate seals. The seals partially leakage. While in both these cases, the SFP was deflated. This deflation resulted in a path for not connected to the reactor cavity, these events SFP water to go to the gap between the two unit and an additional four events discussed in the reactor buildings and into areas of each unit's rest of Section 3.1.2 are precursors that indicate reactor building. When the source of the leak the possibility of failure of the cavity seals and was discovered, the air source was restored and consequent loss ofinventory. Review of the leak was stopped. However, the event lasted individual plant-specific geometry is required to about 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. During this time, the SFP level evaluate each plant vulnerability to this type was noted to be low and make-up was event. performed several times without attempts to determine the cause. The leak detection alarm The staff found four additional events in which was miscalibrated and a drain valve was left cavity seals failed tests before the refueling open, which defeated or impaired the ability to )

cavity was flooded or where leaks developed in detect a leak from the transfer canal gates.

l the seals following refueling. These events Subsequent corrective action included alternate '

indicate that testing of inflatable seals is supplies for alternate gate seals so that inner important in ensuring their operability. The seals were supplied from one unit and outer events further emphasize the need to be aware NUREG-1275, Vol.12 12 L. . .. .

Operating Experience seals were supplied from the other unit, potential SFP puncture events. They indicate

. establishing a degree of redundancy. that movement ofloads heavier than allowed over the SFP is continuing even though the 3.1.3 Pool Structure or Liner agency has taken steps to reduce the problem.

No events involving major SFP leakage have Pool Seismic Design. Only two conditions been reported. However, some events involved were related to seismic design problems with small leaks or potential leaks. SFPs. One condition was related to block walls in the fuel handling building that could collapse L.Iner. Seven events involved leaking from the during a seismic event. The walls were fuel pool liner. These events generally involved replaced. The other condition involved only the relatively small leak rates (less than about 50 fuel racks, which were subsequently seismically gallons per day). One event, involving small qualified.

tears in a PWR refueling cavity seal, was also reported. The events appear evenly spread out 3.1.4 Spent Fuel-Pool Make-up Capability over the review period. Thus, operating experience suggests that occurrence of SFP liner Only two events found during the operating leakage is relatively low. However, Salem experience review involved potential loss of reported (Ref.12) a PWR design problem in SFP inventory make-up capability; no actual which the SFP liner could buckle and leak at losses were found. One event involved a small temperatures above 180 *F. This site is one of accumulation of marine life in the service water the sites that apparently does not have liner pipe used for make up to the SFP Had the drainage isolation capability. Subsequent accumulation of clams gone undetected, it may licensee analysis determined that the liner would have blocked the pipe. Another Seismic Class I not fail. The NRC is currently evaluating the source was available. One event involved a 2-licensee's analy sis, minute loss of an electrical bus needed to supp!y make-up water to the SFP. Operating Load Drops. Only one event was found during experience indicates that losses of all make-up the operating experience review in which the capability are not very likely.

fuel pool liner was punctured by dropping a load into the SFP. A cose shroud bolt was dropped in 3.1.5 Impact oa Safety Equipment this event at Hatch Unit 1 on December 28, 1994. An approximate 0.7-gallon-per-minute Several events were reported that involved leak resulted between the fuel pool liner and the flooding caused by SFP overflow. These events concrete SFP stmeture. The fuel poollevel was had the potential to affect equipment in other restored and maintained with normal make-up portions of the plant. In some of the events, (Ref.13), actual flooding took place when the SFP overflowed into the ventilation system or the No other examples were found of loads actually reactor building. None of these flooding events being dropped that damaged the SFP. However, was serious. They were all caused by human more than 30 situations involved loads heavier error, than allowable that were moved or could have potentially been moved over the SFP. Less than Two conditions were reported in which 20 percent of these events involved actual problems within the SFP could potentially lead downward motion or drops of objects (usually to failure ofimportant safety equipment. The fuel assemblies)into the SFP. Although not licensee for Susquehanna Unit I submitted one judged safety significant by theraselves, these report of a potential effect on safety equipment events represent continuing precursors to that was due to boiling of the SFP on 13 NUREG-1275, Vol.12

Operating Experience November 17,1992 (Ref.14). It describes a Table 3.3 Loss of Cooling Events condition in which a loss of SFP cooling is postulated to occur subsequent to a design basis accident such as a loss-of-coolant accident Type of Event Actual Precursor (LOCA) or a loss-of-offsite power (LOOP).

The design basis accident is postulated to Cooling Flow 1Q 2Q prevent makeup to the SFP. Subsequent boiling SFP Pumps 39 8 of the SFP is postulated to create an environment that could affect safety-related Configuration Control 1 0 equipment in the reactor building. The licensee L ss of Pump Suction 4 0 Flow Blockage 1 0 event report stated that the postulated events were beyond the plant's design basis. These Smgle SFP Pump Failure 5 12 conditions were postulated by Mr. Lochbaum Heat Sink 6 2 and Mr. Prevatte in the 10 CFR Part 21 report and were addressed in a June 1995 letter from the NRC to Pennsylvania Power and Light Company (Ref.15). numbers and types of loss of SFP cooling events. Figures 3.3 and 3.4 give an overview of The second report was a licensee event from the loss of SFP cooling events for which WNP 2 issued May 28,1993 (Ref.16), which temperature increase and duration could be l describes a circumstance in which, under quantified. These figures indicate that the l operating conditions at the time of discovery losses of SFP cooling are infrequent. However, while the local manual service water valve was some events lasted for significant periods, and closed, a postulated LOCA would render four events resulted in temperature increases of emergency SFP make-up capability inoperable. more that 20 *F. The low number of events ,

Subsequent evaporation of SFP inventory and with small temperature increases may be due to tripping of SFP cooling pumps were postulated a lack of reporting of such events.

to result in SFP boiling. The evaporated and l

boiled water is postulated to condense and flood '

the ECCS pump rooms, causing failure of ECCS NUMBER OF OCCURRENCES t equipment needed to mitigate the cagoing 25 l LOCA. The LOCA is postulated to make the 22 local manual SFP make-up valve inaccessible. 20 ~

1@ 32 HRS In this postulated scenario, the normal nonsafety 1@ 30 HRS  !

make-up source is also assumed to be 15 1@ 24 HRS i unavailable. Subsequent licensee investigation 50 k I indicated that the local manual valves in the 10 service water lines for make-up to the SFP could s s I 5

be opened when required after a LOCA. 3 3.2 Spent-Fuel Pool Cooling <1 1-4 4-8 a - 24 > 24 DURATION (HRS)

Fifty-six events found during the operating Figure 3.3 Loss of Cooling Duration experience review involved actual losses of SFP cooling. There were 22 precursor events, which when coupled with additional failures or postulated events, could result in losses of SFP cooling. Table 3.3 gives a summary of the NUREG-1275, Vol.12 14 l l

1 Operating Experience NUMBER OF OCCURRENCES Operable. During these events, the second SFP l 20 cooling pump was adequate to cool the SFP.

Because these events did not result in an actual 15; I ss f SFP cooling, they are not counted in the overall total for this category. While events l with the potential for common-cause, common-10 t mode failure have been reported, none have e occurred.

j 5- p 1

, l In four events, SFP cooling was lost owing to 0 unumas! loss of SFP coolant inventory and consequent 0 0 TO 20 20 TO 40 40 TO 60 tripping of the SFP cooling pumps on loss of TEMPERATURE INCREASE (DEG F) suction. In one flow blockage event, a rubber boot blocked an SFP cooling pump strainer.

Figure 3.4 Loss of Cooling Temperatures About 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> was required to remove the blockage. Although engineered safety features actuations have resulted in losses of SFP Using the number of events found during this cooling, these losses resulted in almost no study over a period of about 12 years for which temperature increase and generally lasted for temperature and duration could be quantified, only short periods. They did not appear to the frequency ofloss of SFP cooling events in present a threat to long-term cooling.

which .i temperature increase of more than 20 *F occurred can be estimated to be on the No actual events involving insufficient cooling order of about 2 to 3 events per 1000 reactor occurred. However, several conditions were years. reported in which full core off-loads were performed with insufficient evaluation of the 3.2.1 Loss of Spent Fuel Pool Cooling heat loads or SFP cooling system during the off-load. Errors in the calculated heat load and The dominant cause of the actualloss of SFP nonconservative ultimate heat sink temperature cooling events was loss of electrical power to assumptions also occurred. This issue surfaced the SFP cooling pumps. Thirty-nine of the loss at Millstone Unit I (Ref.17). The licensee of cooling events were due to loss of power to determined that, during prior refueling outages, the SFP cooling pumps For these losses of the SFP cooling system would not have been electrical power, the time for which cooling was capable, by itself, of maintaining pool not available ranged from a few minutes with no tempercture below the 150 F design limit under accompanying temperature increase to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> certain postulated conditions, including a single with an associated temperature rise of 20 'F. active equipment failure.

Most plants have alternate sources of power available for SFP cooling pumps. No attempt 3.2.2 Ineffective Ileat Sink was made during the event review to determine if alternate power was available in each event.

The primary causes appeared to be human error The second leading cause of loss of SFP cooling and administrative problems in 22 of the 39 was loss of SFP heat exchanger cooling. Of events. The events appeared evenly distributed these 6 events, almost all were caused by human between BWRs and PWRs. error. These events lasted from some very short periods to about 13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br /> with temperature Five events involved failure of one SFP cooling increases ranging from zero to 40 *F.

pump while the second pump remained i

l 15 NUREG-1275, Vol.12

l Operating Experience example, the most prominent type of event, 3.3 Spent-Fuel-Pool

, moving fuel or other loads over the SFP with the Instrumentation Experience liVAC inoperable, is not important to SFP cooling equipment. Events related to breaches Several events involved losses of SFP coolant of buildings are events during which doors were inventory or SFP cooling, in which associated opened or panels were removed when they instrumentation was inoperable or failed before should not have been. Indication was generally or during the events. In one event, a shared received by the operations staff, and the annunciator window was illuminated because of problem was corrected relatively rapidly.

an instrumentation problem when the loss of inventory occurred. Since the window was Lower-than-required flow was not a major already illuminated, the operations staff was not problem with SFP equipment performance, alerted to the loss-of-coolant inventory event Generally, flows were near the required amount.

when it began. While relatively few of these Likewise, negative pressure problems generally mstrumentation problems occurred, they raise did not involve significant deviations from concerns about how SFP instrumentation is requirements.

treated and regarded. Section 6.2 of this report discusses SFP instrumentation. Problems with radiation monitors that actuated SFP HVAC were generally identified quickly.

3.4 Ventilation Events Repair or compensatory action was generally taken in a timely manner. Filter efficiency The staff reviewed about 59 SFP area heating, problems were generally minor.

ventilation, and air conditioning (HVAC) events. Portions of HVAC systems would be Two types of conditions involving dampers and needed if a postulated loss of spent fuel cooling HVAC heaters are potential problem areas. In with consequent boiling and fuel failure were to the case of dampers, events indicate that occur. The summary of the 59 events is in sometimes the problem is difficult to identify Table 3.4. and sometimes difficult to repair quickly.

Heaters may be required to maintain relative humidity within filtering systems. Without the Table 3.4 HVAC System Problems heaters, charcoal may lose ability to filter.

However, the staff found relatively few heater Type of Event Number problems.

Fuel hioved Over SFP / IIVAC Inop 15 oampers 12 3.5 Review of Foreign Operating Building Breaches 9 Experience Loads hioved Over SFP / IIVAC Inop 5 Inefficient Filters 5 During a review of foreign operational HVAC Radiation hionitor 4 experience, the staff found about 24 separate Unable to hiaintain Pressure 4 lleaters Inop 3 events that were related to SFPs. Table 3.5 is a Insufficient Flow 2 list of the types and numbers of events found.

Generally, these events were consistent with Total 59 U.S. experience.

Nine losses of SFP heat sink events occurred. In eight of these events, raw service water tiow Most reported IIVAC events had little impact on was I st at one plant over a period of about 1 SFP equipment related to SFP cooling. For year. The raw service water system cools the NUREG 1275, Vol.12 16

Operating Experience intermediate cooling water system, which cools actual SFP heat exchanger failures were found the SFP heat exchanger. In one event, in the review of U.S. operating experience.

component cooling water cooling was lost to the SFP heat exchanger. Two losses of SFP cooling One event involved a severe loss of SFP level were due to loss of electrical power to the caused by SFP pump seal leaks that were due to pumps. These losses occurred at one site within lack of maintenance. These leaks had existed a short period. They occurred during periods for some time. Also, the operations staff knew when a significant amount of electrical about the loss of level but had not treated it as equipment (including i of 2 diesel generators) important. The problem received little attention, was out of service for maintenance or although corrective actions should have been inspection. taken to comply with procedures. The level dropped from about 34 feet to about 18 feet Table 3.5 Events at Foreign Plants above the fuel.

Type of Event Number One event involved a loss of SFP level because compressed air was lost to the gate seal between loss of Heat Sink (8 at I site) 9 the SFP and the transfer canal. The gate Inventory / Configuration Control 3 between the SFP and the fuel transfer canal was Loss of Cooling / Electric Power (1 site) 2 closed for work on the fuel transfer machine.

Pool Liner Leakage 2 Water passed through the fuel transfer tube to Neutron Poison 2 the containment. The fuel transfer tube could Refueling Cavity Seal I not be shut because the fuel transfer machine Seal Deflation / Loss of Air I Heat Exchanger Leakage 1 could not be moved to dear the isolation valve.

Sever Inventory Loss / Pump Leakage i Tools left in the machine when the area was Fuel Assembly Dropped on Pool I vacated because of incoming water from the Water in HVAC Ducts / High SFP Level 1 SFP were blocking movement of the transfer machine. Air was reconnected to the seal but Total 24 excess air pressure caused the seal to burst, increasing the flow rate to about 26,000 gallons per hour. The operations staff was able to close off an area in the containment. This closure One of the three events involving loss of SFP limited the volume needed to be made up.

coolant level was due to configuration control. About 211,000 gallons of make-up water were This event also resulted in loss of SFP cooling needed to equalize the levels in the containment caused by tripping of the SFP cooling pump and area, the fuel transfer canal, and SFP. Adding was caused by cavitation when the level this volume of water took about 7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br />. One dropped. event involved loss of refueling water at a BWR when the rubber bellows seal between the SFP liner leakage at weld seams occurred at two drywell and the refueling cavity failed.

sites. In one case, the amount of leakage was acceptable and repair was not made. At the otter plant, weld seam repair was performed.

3.6 Operating Experience Review Findings One event involved a generic design problem with SFP heat exchangers and resulted in a leak Losses of SFP or refueling water inventory are in one SFP heat exchanger. This leak developed dominated by events involving system or SFP after 7 years of operation and led to redesign of configuration control problems that were due to all similar types of SFP heat exchangers. No human error. The second most prevalent cause of loss of SFP inventory is leaking inflatable 17 NUREG-1275, Vol.12

Operating Experience gate seals that were generally due to loss of air concern when coupled with postulated SFP i to the seals because of human error. Losses of events that could lead to radiation release.

inventory from SFP gates caused by leaking inflatable gate seals have generally been of Foreign operating experience appears to be )

greater magnitude than those caused by consistent with that from U.S. plants. Operating j configuration control problems. Loss of experience suggests that losses of make-up j inventory was due to configuration control capability are not very likely. .

problems is more easily controlled by the operations staff than leaks from gates. 4 OBSERVATIONS FROM THE However, configuration control problems seem to have taken longer to diagnose. SITE VISITS AND INTERVIEWS Pool leakage events do not appear to have caused problems with long-term losses of spent The staff conducted six site visits to gain fuel cooling. Inadvertent movement of heavier understanding of the licensees' SFP physical than allowed loads over SFPs is continuing even configurations, practices, and operating though the agency has taken steps to reduce this procedures. Site selection was a cross-sampling problem. of the industry that included BWRs and PWRs, large and small architect-engineer designs, The most prevalent type of loss of cooling shared and single pools, old and new designs, events involved loss of electrical power to the and all four nuclear steam supply system vendor SFP cooling pumps that were generally due to designs. The sites visited were North Anna, human error. The few losses of SFP cooling South Texas Project, Susquehanna, Three Mile that were due to loss of SFP heat exchanger Island, River Bend, and Calvert Cliffs. In cooling were also generally due to human error, addition to the site visits, one trip was made to Both types of events resulted in losses of about Pennsylvania Power and Light headquarters.

the same time frame and associated temperature 'wo more trips were taken to conduct rises. The events were evenly distributed interviews. Mr. Prevatte and Mr. Lochbaum between BWRs and PWRs. were interviewed to better understand their concerns as documented in the 1992 While conditions have been reported that Susquehanna 10 CFR Part 21 report, and to suggest the possibility of SFP boiling affecting apply the generic implication of those concerns other plant equipment important to safety, to the industry. The following observations are operating experienee does not provide insights from the site visits and the interviews. These into what is apparently a very complex issue. observations are a cross-sampling and representative of the nuclear power industry.

Opersing experience provides only limited insight into instrumentation problems. Several Each site visit included a tour of the SFP,its loss-of-level events have taken place while level associated equipment, the spent fuel building, instrumentation was inoperable or level and the control room to see the SFP indications.

annunciators were already actuated for other This allowed the AEOD engineers the reasons. This study captured relatively few of opportunity to see the physical arrangement of these instrumentation problems, but they the equipment in relation to other equipment and represent concerns about how SFP to the rest of the plant. The tours were instrumentation is treated and regarded. conducted by licensee personnel who were highly knowledgeable about the configuration Some ventilation events (damper problems, and equipment. In-depth discussions were heater problems) could be potential areas of conducted with the licensees on the procedures NUREG-1275, Vol.12 18

--- - - - - - - - - _ . - - . . _ . . - - ~ -. .-

l:

Site Visit Observations i

j and practices utilized for the SFP activities and rooms and other pathways for water to escape i the analyses that have been performed for the are a concern. Most of these pathways are plant

{

SFPs Discussions were held with control room specific, and it is hard to determine from '

operators, outage planning engineers, observations the path that water would take.

probabilistic risk assessment (PRA) engineers,

systems engineers, maintenance engineers, Local control panels usually have the SFP nuclear engineers, and electrical and - parameter indications and manual controls. The

. instrumentation engineers, manual controls for the valves exist in various locations.

No two SFP physical configurations were the same with respect to the locations of the SFP The control rooms have annunciators for the i pump rooms, heat exchangers, and local water level and temperature located on the main equipment control panels. Most pumps and ' control boards. Few licensees have analog

associated equipment are located below the meters for both temperature and level. If one

level of the SFP. Most SFP cooling pumps are meter exists, the preferred parameter is the SFP j provided safety-related power. Switchgear temperature. The rest of the alarms are grouped i rooms were not in the vicinity of the SFPs. into one annunciator labeled "SFP Trouble "

Very little equipment other than refueling This trouble alarm would include flows, loss of i . equipment is located in the SFP area. power, and inlet / outlet heat exchanger pressure, i Generally, no equipment important to safety that Radiation alarms are in a separate annunciator

could be damaged by the inadvertent boiling of and are part of the plant radiation monitoring i the pool is within the pool area. The pools are system. Radiation monitors do have analog

. divided into distinct areas that are used for readings in the control room, but they are specific purposes, such as cask loading. located in the back panels.

i

, Water level and temperature sensors are located Most activities related to the SFP are covered by

{ in the pools. A very visible scale generally procedures, especially the refueling activities.

denotes pool level. The water level sensor is Procedures exist for logging operator rounds aligned with a vertical plate indicator. Power to and using the specific refueling tools. ' Operators i this sensor and to the temperature sensor is use procedures for aligning the make-up paths,

' generally safety related, but the sensors and for reconnecting the SFP equipment after themselves are not safety related and do not load shed and signals of engineered safety have redundant instmments. features activation. However, reconnecting the

! SFP pumps to the electrical supply is not usually All the plants visited he:I once-through HVAC covered in the top level of emergency

systems so that SFP atmosphere is not procedures. Procedures for responding to all the  !

i recirculated to other perts of the plant. Most SFP annunciators are in the control room.

plants had the capability to isolate the SFP area. Before a refueling outage, the workers are trained in the general SFP activities and their i Several radiation monitors were in each SFP specific tasks. Additional reactor operator j area. Some of these monitors are local alarms assistance is present in the control room during set on certain radiation levels for personnel refueling outages.

evacuation. The rest of the radiation monitors

are part of the plant's radiation monitoring Operator rounds to record the SFP parameters

! system. are conducted at least once every 12-hour shift.

' Most utilities make the rounds twice per shift.

In the case of inadvertent draindowns, system With the recent trend toward shorter outages,

leaks, and overflow of the SFP, the adjacent SFP activities are now generally in the critical 1

1 j 19 NUREG-1275, Vol.12 i

"'+

T

S;te Visit Observations path of the outage schedules. Outages have Responsibility for the SFP and its systems varies been shortened from 90 days to a target of about among licensees. While all have SFP system 20 days for some utilities. These shorter engineers, responsibility does not necessarily outages have forced utilities to perform SFP reside with the system engineer. The individual activities more efficiently and to do more work in charge of the various aspects of the SFP could before the refueling outages begin. Outage reside in the Operations, System Engineers, scheduling and planning include more attention Maintenance, or Nuclear Engineering and detail to the SFP activities. organization. In some utilities, the responsibility is shared between groups. With Although utilities are doing a good job of shared arrangements, the possibility always analyzing the SFP heat loads and heat up rates, exists that, if one does not know the other's the results of these analyses are not always responsibility, issues could be dropped given to the control room operators. These inadvertently. Regardless of responsibility, results could prove to be critical in worst-case when refueling starts, the Operation staff seems refueling outage conditions (e.g., full core off- to have a very tight control of the SFP.

load and a very short outage schedule). Some of the utilities are performing risk analysis as part The newer designs have more of the better of the outage planning. features such as safety-related power, analog control room meters, more parameter indicators Some utilities have used lessons from operating in the control room, more sources of water, and experience and have done a very good job in generally better qualified equipment. However, correcting problems through better analysis, some older plants have made improvements by good operator aids, training, and procedure adding indicators or annunciators in the control revisions. Some utilities have a good system to room, and supplying safety-related power to the evaluate industry experience. SFP equipment. All of the sites visited are including the SFP system in the equipment During the site visits, the engineers identified covered by the Maintenance Rule.

events in which connected systems could have caused loss of SFP coolant inventory. Many All the plants visited had examples of good events such as draindowns are not being practices. Although not every plant used all the reported through the standard mechanisms that practices listed, some of the good practices would allow for the standard analysis of the observed in our visits include-events. Therefore, the actual frequency of draindowns is higher than is typically assigned

• Using licensed reactor operators and training in the risk analysis. Little attention is paid to them for the refueling outages.

the antisiphon devices. Very few sites performed testing or had analysis on the efficacy

• Including SFP risk in the outage planning.

of the antisiphon devices.

Having SFP system power restored in the top There is a large variation in utility practice level emergency operating procedures.

regarding full core off-loads versus fuel shuffles. One plant that had been performing

• Forming a refueling team with fonnal full core off-loads now plans to do fuel shuffles structure.

instead. Another plant that had intended to do fuel shuffles now routinely does full core

• Providing classroom and simulator training off-loads. in preparation for the outage.

NUREG-1275, Vol.12 20

Site Visit Observations ,

Producing user friendly graphs of pool heat

• electrical and instrumentation weaknesses in up rates from the analysis for use in the SFPs control room.

• potential for multiunit sites with shared pools to have an increased SFP risk

• Doing analysis beyond heat loads and heat

• a lack of awareness for SFP issues rates, such as SFP risks in outage planning.

The 10 CFR Pan 21 repon provided an impetus Having strong command and control of SFP for the NRC and the nuclear industry to take a activities. closer look at SFPs, which historically have not received much attention. In the efforts to a

Providing a second source of power for the address the 10 CFR Part 21 report concerns, SFP system. Pennsylvania Power and Light has improved the Susquehanna SFP design, modified its

• Having a mimic on the control board for the operation, improved emergency procedures, and SFP system lineup, improved operator training. A limited PRA showed that the net effect of these actions at Utilizing a system diagram before making Susquehanna would diminish the risk from SFP i SFP system alignment changes. events.

Having an effective program to learn from mternal and industry operating experience. 5 REGULATORY REQUIRE-MENTS AND GUIDANCE

• Refining the SFP risk moden used in the outage planning down to the component Regulatory criteria for the design of SFPs have level, evolved with case-by-case approval for the early plants to the existing criteria. Today, Three good design modification examples were acceptance criteria are specified by the guidance found: in the Standard Review Plan, NUREG-0800 (Ref.18); several Regulatory Guides; and the

• Adding additional SFP indication to the requirements in the General Design Criteria control room. (GDC) of Appendix A to 10 CFR Part 50.

Because of the evolution of the criteria and the Adding safety-related power to the SFP different times that reactors were licensed, the instrumentation. criteria to evaluate the SFP designs among the operating facilities varies. Generally, the newer

• Providing a dedicated HVAC system for the plant, the closer the design is to the specified refueling. SRP criteria. Final acceptability of the SFP design, as described in the applicant's safety The interviews with Mr. Prevatte c.nd Mr. analysis report, is based on cenain GDC and Lochbaum were very informative. They Regulatory Guides, and on independent provided the details of their concern that the as. calculations and staff judgement with respect to found Susquehanna SFP configuration did not system functions and component selection.

meet the licensing basis. The 10 CFR Part 21 AEOD did not attempt to review any existing report that they filed does have potential generic system against the criteria but did observe implications, including- substantial variation in the designs.

mechanisms to transport vapor to and create The SRP provides the specific criteria from the high temperatures in other parts of the plant applicable GDC and regulations and acceptable methods that can be used to meet the criteria.

21 NUREG-1275, Vol.12

1 i

Regulatory Requirements l Two sections of the SRP apply to the SFP: SRP and tsunami. Criterion 4 states that structures, Section 9.1.2, " Spent Fuel Storage," and SRP systems, and components important to safety Section 9.1.3, " Spent Fuel Pool Cooling and shall be designed to accommodate the effects of Cleanup System." SRP Section 9.1.2 covers the the environmental conditions associated with acceptance criteria for the structural aspects of normal operation, maintenance, testing, and the pool for coolant inventory, reactivity postulated accidents. Criterion 5 states that control, and the monitoring instrumentation. structures, systems, and components important SRP Section 9.1.3 covers the acceptance criteria to safety shall not be shared among nuclear for the SFP cooling system and coolant power units unless it can be shown that such temperature control. Because the AEOD study sharing will not significantly impair their ability dealt with the extended loss of SFP cooling, the to perform their safety functions, including, in AEOD study dealt more with the criteria in SRP the event of an accident in one unit, an orderly Section 9.1.3 than Section 9.1.2. shutdown and cooldown of the remaining units.

In 1970, the U.S. Atomic Energy Commission, SFP requirements for fluid systems are in forerunner of NRC, developed Regulatory Appendix A, GDC 44,45, and 46. Criterion 44 Guide 1.13. " Design Objectives for Light-Water states that a system to transfer heat from Reactor Spent Fuel Storage Facilities at Nuclear- structures, systems, and components important Power Stations," to provide specific methods to safety, to an ultimate heat sink shall be acceptable to the staff for preventing loss of provided. The system safety function shall be to water from the SFP, protecting fuel from transfer the combined heat load of these mechanical damage, and providing capability structures, systems, and components under for limiting the potential offsite exposures from normal operating and accident conditions.

a significant release of radioactivity from the Criterion 45 states that the cooling water system fuel. The other applicable Regulatory Guides, shall be designed to permit appropriate periodic Regulatory Guide 1.26, " Quality Group inspection ofimportant components, such as Classification and Standards for Water , Steam , heat exchangers and piping, to ensure the and Radioactive-Waste-Containing Components integrity and capability of the system. Criterion of Nuclear Power Plants," Regulatory 46 states that the cooling water system shall be Guide 1.29, " Seismic Design Classification," designed to permit appropriate periodic pressure Regulatory Guide 1.52, " Design, Testing, and and functional testing to ensure the structural Maintenance Criteria for Engineered-Safety- and leak tight integrity of its components, the Feature Atmosphere Cleanup System Air operability and the performance of the active Filtration and Adsorption Units of Light-Water- components of the system, and the operability of Cooled Nuclear Power Plants," and Regulatory the system as a whole and under conditions as Goide 8.8, "Information Relevant to Ensurmg close to design as practical, the performance of That Occupational Radiation Exposures at the full operational sequence that brings the Nuclear Power Stations Will Be As Low As Is system into operation for reactor shutdown and Reasonably Achievable," were not developed for LOCAs, including operation of applicable specifically for the SFP but have some guidance portions of the protection system and the that applies to the SFP. transfer between normal and emergency power sources.

SFP overall design requirements are in Appendix A, GDC 2,4, and 5. Criterion 2 Fuel and radioactivity control requirements are states that structures, systems, and components in Appendix A, GDC 61,62, and 63. Criterion important to safety shall be designed to 61 states that the fuel storage and handling, withstand the effects of natural phenomena such radioactive waste, and other systems that may as earthquakes, tornadoes, hurricanes, floods, contain radioactivity shall be designed to ensure NUREG-1275, Vol.12 22

i  :

l Regulatory Requirements  ;

adequate safety under normal and postulated ; determine if the time is consistent with the

accident conditions. These systems shall be industry calculations. The radiation assessment designed with a capability to permit appropriate presented the results of utility calculations on periodic inspection and testing of components the radiation level that would exist for different important to safety, with suitable shielding for SFP waterlevels. Detailed results of these radiation protection, with appropriate assessments are discussed in Sections 6.1 containment, confinement, and filtration through 6.4 of this report.

systems, with a RHR capability having i reliability and testability that reflects the importance to safety of decay heat and other 6.1 Electrical Assessment l RHR, and to prevent significant reduction in The staff reviewed design features of spent fuel fuel storage coolant inventory under accident

. cooling systems of a representative sample of 14 conditions. Criterion 62 states that criticality in plants to understand the type of electrical power the fuel storage and handling system shall be supplies to the SFP cooling systems at these 1 prevented by physical systems or processes, '

plants. The review included representative preferably by use of geometrically safe samples of BWRs and PWRs for vendors -

configurations. Criterion 63 states that General Electric, Westinghouse, Combustion i appropriate systems shall be provided in fuel Engineering, and Babcock & Wilcox. The storage and radioactive waste systems and '

design features of electrical power supplies associated handling areas to detect conditions varied among different plant types and vendors, that may result in loss of RHR capability and and sometimes even among plants designed by ,

excessive radiation levels and to mitiate the same vendor. I appropriate safety actions. '

The SFP pumps for approximately 80 percent of '

6 ENGINEERING these plants have qualified and fully ASSESSMENTS independent Class IE power supplies. For these plants, the normal source of power is the offsite In suppoit of this study, AEOD performed grid system, and the emergency source is the engineering assessments of the electrical diesel generators. Load shedding under LOOP system, instrumentation, heat load, and radiation c nditions is initiated by undervoltage relays.

areas. The electrical assessment was to After the diesel generators have energized the understand the type of power supplies for SFP emergency buses, the emergency loads are cooling system components, such as pumps, automatically started by the load sequencer.  :

valves, and instrumentation. The The SFP pumps are not automatically started, instmmentation assessment included gathering but need to be manually started by the operator of information on the type of instrumentation after all emergency loads are started.

used to monitor the system parameters and desirable enhancements to the instmmentation. The power supplies for the SFP pumps for the The electrical and instrumentation assessments remaining plants reviewed are Non-Class IE. In were based on a review of system design data the event of a LOOP at these plants, the SFP for a sampling of plants and the results of the co ling function will be lost.

site visits.

The information in the Final Safety Analysis The heat load assessment included independent Report and in other sources was insufficient to calculations on heat up and boiling of the SFP determine the type of power supplies for the resulting from complete loss of cooling for a system valves and instrumentation,(i.e.,

typical PWR and a BWR. The calculations whether they are Class IE, qualified, and l estimated the time to reach boiling conditions to redundant).

l 23 NUREG-1275, Vol.12

I l

1 1

l Engineering Assessments The observations from the site visits were in Table 6.1 SFP Instamentation Summary (14 plants) l general agreement with those from the review l on the type of power supplies for SFP pumps. Parameter L' T 2 LD' P' F' Mo.t sites have Class IE power supplies for the SFP pumps and the system instruments, Plants monitored 14 14 11 11 3 although the instruments themselves a'e not Indicated (CR*/ Loc') 5/10 6/6 0/8 0/8 0/1 safety related.

Alarmed (CR/lec) 1I/712/33/3 5/3 2/0 6.2 Instrumentation Assessment i sei a

Temperature Review of the design features of SFP cooling system instrumentation for the same sample of l

' now l' ' Q" p_

,p 14 plants noted in Section 6.1 gave the team a Control Room general understanding of the type of instrumentation at these plants. In addition, the team visited six other plants to obtain instrument information. As in the case of .

electrical power supplies, the design features of m. For half the plants, the temperature is

[indicated or recorded in the control room; for instrumentation varied among different plant the other half, the temperature is on a local types and vendors, and sometimes even among plants designed by the same vendor. Panel. For most plants reviewed, an abnormal Notwithstanding these differences, this section temper ture is individually alarmed in the c ntrol room. For other plants, the alarm is on describes the instrumentation features that are ,

typical of the industry. the local panel that imtiates a common trouble alarm in the control room.

The results of the review are summarized in The SFP level sensor has a narrow range, Table 6.1. Each plant had some type of typically 4 feet, c vering high and low alarm instrumentation to monitor the SFP system ,

setpoints and the minimum Technical performance, although the type and extent of .

instrurnentation varied significantly among Specification lewt h control room level plants. The parameters monitored include SFP indicator provided by this sensor is good only f r this narrow range. Therefore, the control level, temperature, liner leakage, pump ,

discharge pressure, and system tiow. Indication, r m indicator cannot monitor a level below this range and becomes useless for lower level recording, or alarm functions of these parameters are provided either in the main c nditions expected in case of a gross loss of control room or on a local panel. Typically, S c 1 nt buntory ennt.

most instrumentation is on the local panel, and . .

only important parameters are monitored in the A direct indication m the control room of SFP control room. However, most local alarms IS".I".nd temperature would be desirable to mimm3Ze Perator response time for events initiate a common trouble alarm in the main control room and an operator is dispatched to inv lying rapid loss of SFP coolant inventory or investigate the cause of the trouble. I ss of SFP cooHng. The present design feature of local ind,ication with a trouble alarm in the Each plant had level and temperature e ntr 1 r m f r these parameters may prove to instrumentation. SFP level is monitored locally, be insufficient for quickly responding to such events as full core off-load heat up caused by but an abnormal level is alarmed in the control loss of inventory. Lack of direct indication in the control room will complicate diagnosis of l NUREG-1275, Vol.12 24

Engineering Assessments events. Typically an operator needs to be old, have no redundancy for the SFP dispatched to determine the cause of trouble, instrumentation.

which is time consuming. Developing trends for l SFP level and temperature can be difficult 6.3 Heat Load Assessment because the control room operators have to depend on infrequent local operator rounds The AEOD staff performed independent (typically once in 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />). The capability to calculations on the heat up and boiling of the develop trends for the parameters allows the SFP resulting from a complete loss of cooling, operator the opponunity to react more quickly to calculating the SFP heat up rate, the time for the developing problems. Therefore, a direct SFP water to reach the boiling point, and the indication of SFP level and temperature in the time for the water to boil down to the top of control room, consisting of analog readings and fuel. The calculations were done for a typical annunciators, would be a desirable safety PWR and a BWR, using simplified and l enhancement.

generally conservative assumptions, under maximum heat load conditions.

In most plants, SFP pump discharge pressure is used as an indication of adequate system flow. The major assumptions and the results of the {

Only a few plants employ direct flow calculations are summarized in Table 6.2. For measurement. In all cases, the pressure or flow the typical PWR and BWR, the calculated heat is indicated locally. An abnormal pressure or up rates were 9.3 and 15.2 *F/ hour, the times to flow would be annunciated in the control room reach the boiling point were 12 and 7.4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, for most plants and on the local panel for others. and the time to boil down to the top of the fuel were 80 and 50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br />, respectively. The The SFP liners in almost all plants reviewed difference in values for the PWR and BWR have with some form of local leakage detection. cases were mainly due to the larger volume for  ;

Abnorrnal leakage is alarmed only for a few plants, locally or in the control room. For other Table 6.2 SFP Heatup Calculations plants, an operator would periodically check the leakage detection system for any indication of abnormal leakage. PWR BWR These plants have various radiation monitors in Plant (Data Source) Surry Limerick Plant Rating, MWe 781 1055 a system separate from the SFP cooling system.

Local area monitors are provided for personal Maior Assumptions:

safety in case of a need to evacuate the area.

The other monitors are part of the station Heat Load, BTU /hr 41 E6 38 E6 radiation monitoring system. These monitors Water Volume, cubic feet 71,000 40,000 alarm in the control room through the Water Volume above fuel 46,000 26,000 j

annunciator system. In addition, the radiation Initial SFP Temp., 'F 100 100 monitors have analog meters and recording Time after Shutdown, hours 100 216 signals.

Results The newer plants have safety-related power to the SFP instrumentation, but the instruments 9.eatup Rate, *F/ hour 9.3 15.2 g int, " "

themselves are not safety related. The older

, T o , u 55 plants have neither safety-related power supph.es for the instrumentation, nor instruments that are safety-related. The plants in general, new and 25 NUREG-1275, Vol.12

Engineering Assessments the PWR (71,000 cubic feet versus 40,000 cubic 35 days in the first outage to 13 days in the l feet for the BWR). fourth refueling outage. Figure 6.2 shows the time to initiate boiling as a function of outage The calculations were based on maximum number with the refueling pool gates in and out. I expected heat load under worst-case conditions, During the first four refueling outages the such as full core off load and a back-to-back refueling pool gates were out at Nine Mile Point refueling scenario (for a dual unit site) with Unit 2. However,if maintenance work would maximum expected inventory of spent fuel from have been required on the reactor vessel or previous refueling outages. The heat load appurtenances during those time it would have assumed was 40.8 X E6 BTUs/ hour for the been necessary to have the refueling pool gates )

PWR and 37.6 X E6 BTUs/ hour for the BWR, installed, thereby leading to shorter times to as found in the Final Safety Analysis Report or spent fuel boiling. Similarly, during a visit to licensee's calculation. the South Texas plant, AEOD learned that calculations performed for the most recent Operating experience has shown that,in an refueling outage estimated that the initiation of effort to minimize refueling outage time, many boiling could begin approximately 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> after plants perform full core off-loads early in their the SFP cooling is lost. The NRR survey outages. AEOD discussions with the assessment of the South Texas plant also engineering manager of Nine Mile Point Unit 2 indicated that if a full core had to be off-loaded I provided good insight into the effect this during midcycle, boiling could begin about 2 to practice has upon reducing the time available 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> after losing SFP cooling, until boiling begins.

Figure 6.1 shows the time from reactor so shutdown until completion of the full core off- E ,, a omsour ,

load at Nine Mile Point Unit 2. As Figure 6.1 60-a oms = i indicates, the period of shutdown until ,,

completion of the off-load decreased from

• 30- l 34 2 20- -

t r.a

! 35 -

e I h 10 a4 l30 f25 24 o, _.

1 2 3 4 1 REFUEL OUTAGE NUMBER 15 is l i

I 3 3 nuun ouwm usea Figure 6.2 Reduced Time to Boil Although the estimate of time to reach the boiling point and time to boiling down to the top I

of the fuel could vary among plants, the results Figure 6.1 History of Full Core Offloading of the AEOD calculation are indicative of the typical order of magnitude estimate for U.S.

plants. These estimates are consistent with the estimates provided by several licensees.

i i l NUREG-1275, Vol.12 26

1 Engineering Assessments 6.4 Radiation Assessment m intained s I w as re s n bly achievable.

However, little mformation was available to the In addition to providing the vehicle to remove perating staff for comprehending the radiation hea: from the spent fuel, the water in the pool is fields that would be present in the vicimty of the relied upon to provide shielding for plant SFP during an accident. A comment expressed personnel. Loss of SFP coolant inventory with at several plants was that if things went bad the decreased SFP water levels can result in radiation monitors would go off and that was the excessively high radiation fields that would signal to " dear wt/

prohibit entry into the SFP area. The shielding Recogm. . zing the need to add water to the SFP effect of the water in the SFP increases exponentially with increasing water level. during an accident from the standpoints of fuel Table 6.3 shows the results of severallicensees. c ling and personnel habitability, many plants calculations that indicate the effectiveness of the do have remote " alternate" or " emergency" fuel water shielding associated with spent fuel. Pool make-up capability. Remote " alternate" or

" emergency" fuel pool make-up capability is not As shown on Table 6.3, the radiation dose level a regulatory requirement; however, it does at the surface of an exposed spent fuel bundle appear to be a matter of prudence. In a worst-was estimated to be 250,000 rem /hr. The same case scenario, aceessibility to the SFP area bundle would produce a radiation dose level of c uld become an important issue if local manual 2.5 rem /hr with a shielding of 5 feet of water. actions were necessary to connect a make-up The radiation dose level from the same bundle swrce.

would decrease to less than 20 mrem /hr with a . . . .

shielding of 8.5 feet of water. Disgussions with plant personnelindicated that the information available to them about radiation doses was limited to individual Table 6.3 Radiation Shielding Estimates bundles but did not address the entire pool.

This appears to result from a mind-set in which water the operating staff envisions a spent fuel Plant Depth Bundles Rem / hr accident to be one that involves the handling of Susquehanna 0 inches single 250.000 a single bundle as a very credible event, whereas 5 inches multiple 100.000 they may not envision a major loss of coolant 5 feet single 2.5 inventory from the SFP as a credible event. At 8.5 feet single .02 several plants the operators did not have ready Oconee I foot multiple 900.000 access to information about the radiation doses from the SFP (versus an individual bundle) as a TMI1 6.5 feet single .007 function of SFP water level even though engineering had already performed such North Anna 7 feet single <.05 d @ ionL The plant staffs were all familiar with the SFP water-level requirements, however their Each of the plants visited had radiation detectors responses were rmxed when it came to in the SFP areas with control room and local addressing the minimum amount of water monitors and alarms. Discussions with plant needed for shielding to enable habitabihty staff indicated that the personnel were well during a SFP accident. Operators mdicated that trained and very knowledgeable of plant policies water shielding requirements to allow access to and regulatory aspects of radiation such as the refueling and spent fuel areas were between radiation control and health physics for ensuring 4.5 to 10 feet.

that the exposure of personnel to radiation is 27 NUREG-1275, Vol.12

Frequency estimates for loss of SFP coolant 7 RISK ASSESSMENT inventory initiated by loss of coolmg, nussiles, and pneumatic seal failure were very low, Over the years, the SFP has not received the risk However, the frequency estimates for loss of assessment attention that the reactor has because C f ant l inventory caused by structural early analyses put the risk of a SFP accident an f ilure in a seisnue event and heavy load drops order of magnitude below a reactor event. ,

were quite uncertain. In the case of seismic Therefore, the analyses done for the SFP were events both the seisnue hazard and structural limited. However, in recent years several issues ,

frag lities contribute to the uncertainty range.

have required that certain aspects of the SFP be ,

For heavy load drops, human error probabilities, studied further. INEL was contracted to review structural damage potentials, and recovery the previous SFP risk assessments and to utilize

" " "S were the primary sources of the useful insights to assess the current risk of .

uncenannties.

SFP accidents. In addition to those risk insights, INEL utilized the AEOD operating experience BNL assessed the conditions that could lead to review, engineering analyses, site visits, and site failure f the spent fuel Zircaloy cladding if the interviews in assessing the likelihood of SFP ,

cladding ruptured or a self-sustammg oxidation events.

reaction occurred, and they estimated the SFP fission product inventory and the releases and 7.1 Ex,st,i m g Probabilistic Risk consequences for the various cladding failure Assessment scenarios. Possible preventive or mitigative measures were qualitatively evaluated. The The INEL study included the review of three uncertainties in the risk estimate for a pool fire previous risk assessments that were relevant to are large, and BNL identified areas where SFPs: (1) NUREG/CR-4982, " Severe Accidents additional evaluations are needed to reduce j in Spent Fuel Pools in Support of Generic uncertainty.

Safety Issue 82," (Ref.19); (2) NUREG-1353,

" Regulatory Analysis for the Resolution of 7.1.2 NUREG 1353," Regulatory Analysis Generic Issue 82,'Beyond Design Basis for the Resolution of Generie Issue 82, Accidents in Spent Fuel Pools'" (Ref. 20); and is Accidentsin

'Be],o" gp t is' (3) " Risk Analysis for Spent-Fuel-Pool Cooling at Susquehanna Electric Power Station," NUREG-1353 was a value/ impact and cost.

(Ref. 21). benefit evaluation performed in 1989 that concluded that there were no cost-effective 7.1.1 NUREG/CR 4982," Severe Accidents options to mitigate the risk beyond the licensing in Spent Fuel Pools in Support of basis for SFPs. Previousl3 , WASH-1400 Generic Safety Issue 82" (Ref. 22) considered SFP risks to be an order of m gnitude below reactor risks. In 1989, the NUREG/CR-4982 documented an assessment agen y determined that SFPs required re-performed in 1987 by Brookhaven National l; Laboratory (BNL) of the likelihood and ?xamin ti n because more fuel was being stored consequences of a severe accident in the SFP. '".the pool than ongmally mtended and new seisnue concems had arisen for the eastern plant  ;

The authors concluded that the risk estimates sites. More fuel m, the SFP mereased the nsk of are quite uncertain and could potentially, under the worst-case assumptions, be significant. The fire PmPagation owing to the zircaloy cladding.

A zircaloy cladding fire can occur at 1650 *F j assessment identified potential mechanisms and and such temperatures could be reached if the 1 conditions for failure of spent fuel cooling and subsequent release of fission products.

SFP lost all cooling and inventory. From this ev lu tion, PWRs were deternuned to be four Millstone Unit I and Ginna, two older designs, were the plants evaluated for the assessment. times more susceptible to cladding fire than BWRs because of the configuration of their NUREG-1275, Vol.12 28

Risk Assessment storage racl:s. Risk from beyond design basis NBF reduction of about a factor of four with a earthquakes to the SFP was no greater than commensurate reduction of risk of about a factor damage to the reactor core from safe shutdown of four.

earthquakes.

The results of the PNL study were integrated 7.1.3 " Risk Analysis for Spent Fuel Pool into NRR's Safety Evaluation, "Susquehanna Cooling at Susquehanna Electric Steam Electric Station, Units I and 2, Safety Power Station" Evaluation Regarding Spent-Fuel-Pool Cooling Issues." The PNL analysis was used to augment In October 1994, Battelle Pacific Northwest the deterministic analysis of the Susquehanna Laboratory (PNL) prepared a draft report, " Risk plant. From their deterministic analysis, NRR Analysis for Spent-Fuel-Pool Cooling at found that " systems used to cool the spent fuel Susquehanna Electric Power Station," for NRC's storage pool are adequate to prevent Risk Applications Branch of NRR. The report unacceptable challenges to safety-related presented the results of PNL's analysis ofloss of systems needed to protect the health and safety SFP cooling events at the Susquehanna nuclear of the public during design basis accidents." On power plant, including estimates of the the basis of the PNL analysis, NRR indicated likelihood for loss of SFP cooling, the NBF, and that " loss of SFP cooling events represented a order-of-magnitude estimates of core damage low safety significance challenge to the plant frequency (CDF) attributed to SFP heat-up [Susquehanna] at the time the issue [Part 21 events. report] was brought to the staff's attention."

The PNL analyses addressed design basis Although large uncertainties may be associated accidents that would cause mechanistic failure with the absolute values and specific numerical of the nonsafety-related SFP cooling system. results of the PNL analyses, much insight can be The accident scenario postulated by gained from the PNL analyses of the Mr. Lochbaum and Mr. Prevatte in their 10 CFR Susquehanna station. For example, the PNL Part 21 report, an RCS LOCA, would result in analysis shows that the most significant risk de-energizing SFP power and could also induce reduction could be achieved from three hydrodynamic loading of systems and strategies:

equipment associated with SFP cooling. In addition to addressing RCS LOCA, NRR had (1) installing SFP level and temperature PNL analyze other initiating events: instrumentation in the control room, earthquakes, LOOP, and Gooding. The PNL analysis did not find major SFP coolant (2) enhancing SFP normal and off-normal inventory losses from configuration control, procedures and training staff to be gates, and seals to be credible events. Proficient, and The results of the analyses indicated that the (3) cross-tying SFPs.

risk from SFP events was low compared to reactor events that did not account for any risk 7.2 Risk Assessment contribution from the SFP. The PNL study showed that for the Susquehanna plant, the AEOD obtained technical assistance in the area i largest contributors to SFP risk emanated from of risk assessment from INEL. INEL reviewed extended LOOP and LOCA events. The the PNL Susquehanna PRA, assessed the improvements that were made at the adequacy of the risk analysis, and addressed the Susquehanna station in response to the issues adequacy and reasonableness of the assumptions raised by the 10 CFR Part 21 report resulted in a made. INEL extracted insights from tF PNL 29 NUREG-1275, Vol.12

Risk Assessment Susquehanna PRA and the other relevant PRAs found the NBF for the Susquehanna plant, after in industry to assist in generically assessing the implementing the 10 CFR Part 21 likelihood ofloss of SFP cooling. Information improvements, to be SE-5/ year, which is from the AEOD reviews of operating approximately twice that found by PNL.

experience, interviews, site visits, and independent SFP analyses was used to refine the Table 7.1 Near-Boiling Frequencies PNL PRA model. This study provided quantitative estimates of the NBF and INEL PNL qualitative discussions about the risk oflosses of SFP cooling. The following sections provide Total NBF 5 E-5 2 E-5 the results and the insights obtained from these INEL efforts (Ref. 23). LOOP 3 E-5 i E-5 7.2.1 Risk Assessment-Quantitative Inventory Losses 2 E-5 1 E-6 Results INEL corrected modeling problems identified in the PNL study. The event and fault trees were '

The dominant event initiators were LOOP and refined to more accurately describe current SFP inventory losses, including configuration Susquehanna plant operations. To refine the control errors and seal failures. Because of event trees, the INEL staff visited Pennsylvania limited time and resources,INEL did not Power and Light engineering offices and the include a quantitative estimate of the CDF.

Susquehanna station. The event and fault trees Also, given the limited data available for were quantified, using recent operating development of estimates of event frequencies experience data supplied by AEOD. INEL also and the limited resources available for model refined and updated the data and models that development, these estimates need further PNL had used to account for human refinement before they can be used as a basis for performance.

regulatory actions.

In some cases, the modifications and improvements resulted in increases in the NBF 7.2.2 Risk Assessment -Qualitative Results in the SFP, which in turn would result in The SFP PRAs done by PNL and INEL were increased estimates of risk. Correcting the specifically for the Susquehanna plant. Many initiating event frequencies for station blackout, features of the design and operation of LOCA, reismic events, configuration control Susquehanna are unique, consequently the errors, and seal failures would tend to increase results of the PNL and INEL analyses cannot be the NBF. Counterbalancing this,INEL applied directly to other plants. Nonetheless, identified possible sources of conservatism in certain qualitative insights that have been the PNL study. Chief among them were the learned from those studies may have generic estimates of human performance associated with applications. Forexample:

recovery and mitigation.

(1) Effect of defueled unit upon operating INFL performed the aforementioned unit. The analyses showed that for a dual-refinements, including modifications of the unit BWR, a problem with SFP cooling at a initiating event frequencies, using AEOD's shutdown unit could affect the adjacent operational event database, to cover a full operating unit. The accident scenario spectrum ofloss of SFP inventory events, postulated in Mr. Lochbaum's and Mr.

including catastrophic seal failure. The results Prevatte's 10 CFR Part 21 report was found of their analysis are shown in Table 7.1. INEL NUREG-1275, Vol.12 30

Risk Assessment to be a credible event, but less likely than system pumps are not on the emergency I other events. busses. The original accident scenario raised in the 10 CFR Part 21 report did not (2) Uncertainties of core damage f.equency appear to be a significant contributor to ,

estimates. The task of estimat!ng the NBF NBF. INEL found that the dominant I appears to be amenable to the use of PRA contributors to NBF were LOOP and SFP techniques. However, the tas c of inventory loss.

estimating CDF is subject to 3 ery large uncertainties. PNL and INEL both (5) Deviation from the modeled plant design.

acknowledged that the methodology used Risk estimet" from the SFP for the for this task provided only " order of Susquehanna plant may be affected by magnitude estimates." changes planned for fu ure refueling outages, which may represent major (3) Effect of Mr. Lochbaum's and Mr. deviations from the models used by PNL Prevatte's 10 CFR Part 21 Report, and INEL. Some of those anticipated Comparison of the analyses that were done changes are-for the Susquehanna plant as it existed at the time of the 10 CFR Part 21 report and

• operation without the SFP cross-tied for after corrective actions were taken revealed the future dry-cask storage operations a reduction of refueling outage from 55 days that the improvements made in the areas of instrumentation, accident response to 35 days procedures, operator training, and a partial core off-loads ta'ing place earlier in shutdown operations reduced the estimated the outage NBF.

(6) Operating experience. INEL found that Improvements in instrumentation consisted of SFP inventory lo'sses such as draindowns providing reliable SFP level and temperature or pneumatic seal failures may be monitoring instruments in the control room. important contributors to NBF at the Susquehanna plant. In previous PRAs such Improvements in operations and accident ' events were either not modeled or their response procedures involved the following: occurrence frequency was assumed to be very low; once every 10,000 reactor years.

• ventilation system isolation a

installation of drains in the standby gas 8 FINDINGS AND treatment system

• utilization of the RHR system of the CONCLUSIONS operating unit to cool the SFP verification that removal of cask storage pit The findings and conclusions presented in this gates results in effective heat transfer section are based on review of operating events between the SFPs and interpretations of the available risk analyses. The conclusions are stated, followed (4) Dominant accident sequences. For the by indented paragraphs giving the findings on Susquehanna plant, PNL found that the which these conclusions are based. These accident sequences that were the largest findings and conclusions are grouped under the contributors to NBF were extended LOOP headings of (1) likelihood and consequences of and LOCA. The extended LOOP is a SFP events, (2) prevention of SFP events, and dominant contributor because at the (3) response to SFP events.

Susquehanna station the SFP cooling 31 NUREO-1275, Vol.12

Findings and Conclusions 8.1 Likelihood and Consequences about 25,800 gallons were lost in about 30 l

nunutes.

of Spent-Fuel Pool Events l . Several losses of SFP cooling continued 8.1.1 Loss of Coolant Inventory Events for more than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />; one had a maximum temperature increase of 50 *F From review of more than 12 years of operating to a final temperature of 140 *F. There experience, the staff determined that loss of SFP were no reported approaches to boiling coolant inventory greater than I foot has during the experience review period.

occurred at a rate of about i event per 100 reactor years. Loss of SFP cooling with a .

While the operating experience review temperature increase greater than 20 *F has results are believed to be reasonably occurred at a rate of approximately 3 events per representative, discussions with 1000 reactor years. The consequences of thece operations staff revealed a number of actual events have not been severe. However, additional events that did not reach the some events have resulted in loss of several feet reporting threshold required by NRC of SFP coolant level and have exceeded 24 regulations and, therefore, were not hours. The primary cause of these events has initially captured by the study's event been human error. review process.

. Two loss of SFP coolant inventory events occurred m, which SFP level decrease 8.1.2 Possible Consequences of Loss-of-Coolant Inventory exceeded 5 feet. These events were terminated by operator action when Review of existing SFP risk assessments approximately 20 feet of coolant remained showed that after correction for several above the stored fuel. Without operator problems in the analyses, the relative risk from actions, the inventory loss could have loss of spent fuel cooling is low in comparison continued until the SFP level had dropped with the risk of events not involving the SFP.

to near the top of the stored fuel resultirg The likelihood and consequences ofloss of SFP in radiation fields that would have cooling events are highly dependent on human prevented access to the SFP area. The performance and individual plant design events with the largest level of decrease features, involved unavailable or inaccurate instrument readings. Ten other loss of . Risk assessment identified loss of offsite inventory events resulted in level power and loss of SFP coolant inventory decreases between I and 5 feet. Operator were the major contributors to near-response to one of the largest losses of boiling frequency. LOOP was a major SFP coolant inventory events (loss of 5.5 contributor largely because the analysis feet of water level in SFP) was deficient was based on the Susquehanna plant because several opportunities to diagnose where the SFP cooling system is not and correct the problem were missed connected to emergency power.

when make-up coolant was added to the system without evaluating the cause of the . Human performance is the most important need for make-up. Two precursor events factor for both loss of spent fuel cooling involved cavity seals that precipitated event initiators and recovery actions.

l rapidly developing leaks. In one case, Problems with configuration control I about 200,000 gallons of water were lost caused most of the SFP events. Lack of l in about 20 minutes. In the second case, automatic functions for detection and recovery from SFP events places full NUREG-1275, Vol.12 32

Findings and Conclusions reliance on operator actions. The results plant-specific design features and human of risk assessments involving operator performance. Plant-specific design actions are sensitive to the level of features that impact the near-boiling administrative controls, instrumentation, frequency include pneumatic reactor  ;

procedures, and training provided to aid cavity seals and gate seals and SFP l operator performance. geometry that might result in draindown to near or below the top of the stored fuel.

• The impact ofinstrumentation, procedures, and training is dependent

, 8.2 Prevention of Spent-Fuel-Pool l upon plant-specific design features. The NRR survey of SFPs identified a wide Events range of plant design features and specific limitations at existing plants. Plants 8.2.1 Configuration Control which have identified limitations relating to configuration control, instrumentation, The need for improvements to configuration procedures, and training could reduce the c ntrols related to the SFP to prevent or mitigate risk of SFP events by relatively modest SFP loss-of-inventory events and loss-of-cooling improvements in these areas. In fact, the events should be evaluated on a plant-specific basis, modest improvements to instrumentation and operations made by Susquehanna resulted in reduced risk.

• Operating experience shows that the most frequent cause of loss of inventory and Assessment of operating experience determined I ss of cooling is ineffective configuration that licensee efforts to reduce outage duration c ntrol. Mistaken valve alignments have have resulted in full core off-loads occurring diverted water from the SFP and have earlier in outages. This increased fuel pool heat is lated the air supply to pneumatic seals.

load reduces the time available to recover from Mistaken electrical alignments have a loss-of-SFP-cooling event early in the outage. resulted in loss of power to SFP system pumps and other equipment.

8.1.3 Need for Specific Corrective Actions 8.2.2 Plant Modifications at Multiunit Sites The need for specific corrective actions should be evaluated for those plants where failures of h need for plant modifications at some reactor cavity seal or gate seals, or ineffective multiumt sites to account for the potential effects of SFP boiling conditions on safe antisiphon devices could potentially cause loss of SFP coolant inventory sufficient to uncover shutdown equipment for the operating unit, Particularly during full core off-1oads, should be the fuel or endanger makeup capability. ,

evaluated on a plant-specific basis.

Risk assessment identified loss-of-SFP-

• The Susquehanna 10 CFR Part 21 report coolant inventory was a major contributor brought to light the potential problem that, to NBF, and review of operating when two units have a common pool, the experience and the site visits identified refueling of one unit when SFP coolm, g is that problems with configuration control, I st could impact the operating umt.

seals, and antisiphon devices were Specifically, there is the need to assess the contributors to large losses of inventory.

potential for mechanisms to transport Vapor to create high temperature in other

• Risk assessment identified the near-Parts of the plant that have critical plant boiling frequency is sensitive to individual equipment. The NRR survey identified 33 NUREG-1275, Vol.12

seven sites besides Susquehanna that have revealed many ways that water could be shared pools. Since the scenario involves provided to the pool that had not formerly the potential for many things to go wrong been described and for which procedures I and because each configuration is do not exist.

different, these seven sites need additional assessment and evaluation. 8.3.2 Procedures and Instrumentation 8.3 Response to Spent-Fuel-Pool The need to improve instrumentation and power I Events supplies to the SFP equipment to aid correct operator response to SFP events chould be l 8.3.1 Operator Response evalu ted on a plant-specific basis.

The need for improved procedures and training Instrumentation available to the operators for control room operators to respond to SFP regarding the SFP parameters can be very loss-of-inventory and SFP loss-of-cooling limited. A single annunciator may be the events, consistent with the time frames over nly indication of SFP trouble. Some which events can proceed and recognizing the Pl ants have SFP level or temperature heat load and the possibility of loss ofinventory indication readouts on control room back should be evaluated on a plant-specific basis, p nels. Allindications of the SFP parameters could easily be lost in a reactor Refueling outages are getting shorter. accident because not all of these Control room operators at some plants are instruments have safety-related power.

not aware that early transfer of the entire Plant operators make rounds to the SFP core from the reactor to the SFP during a 1 cation, but the time between successive refueling outage results in significant heat visits may be too long to adequately loads in the SFP and the potential for near- develop trends of the SFP data and stop a boiling conditions within 5 to 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> if developing problem before it becomes a cooling to the SFP is lost. Current operator serious event. In several events, SFP training and procedures do not typically emling was lost because power to the SFP include this information, or if the pumps was lost. Most power supplies to information is included it is not easy to the SFP pumps are safety related. For units I interpret. that do not have safety-related power for the pumps, there is a need to assess All licensees have, to some degree, work whether having such power during accident scheduling, training, and procedures that c nditions would assist them in reacting deal with the SFP activities during a faster to an SFP event.

refueling outage and during normal plant operations. However, the effectiveness of 9 REFERENCES these efforts was not apparent at all the plants visited. Of the licensees that had (1) 1. Lochbaum, D.A., and Prevatte, D.C., Letter a formal training structure consisting of to Martin, T., U.S. Nuclear Regulatory i classroom lectures for the workers Commission, "Susquehanna Steam Electric involved in the refueling activities, (2) a Station, Docket No. 50-387, License No.

schedule program that incorporated the NPF-14,10 CFR Part 21 Report of i SFP risks, and (3) detailed procedures for Substantial Safety Hazard," November 27, '

all the activities, the engineers and 1992, operators had knowledge and awareness of relevant SFP issues. Regarding backup 2. U.S. Code of Federal Regulations, Title 10, sources for SFP coolant inventory and SFP " Energy," U.S. Government Printing Office, cooling, discussions with the licensees Washington, D.C., revised periodically.

l NUREG-1275, Vol.12 34

- - .. . . -. . .~. .-- - . - - . - - . . . - - . .

References

3. Taylor, J.M., U.S. Nuclear Regulatory 13. U.S. Nuclear Regulatory Commission, Commission, Memorandum to Jordan, E.L., Morning Repon 11-94-0112, December 29,

" Study of Spent Fuel Pool Issues," 1994.

February 20,1996.

14. Pennsylvania Power & Light Co.,

4. Jordan, E.L., U.S. Nuclear Regulatory Susquehanna Unit 1, Licensee Event Commission, Memorandum to Taylor, J.M., Report 50-387/92-016, " Voluntary Report-

" Plan and Schedule for Spent Fuel Pool Spent Fuel Pools," November 17,1992.

Cooling Study," February 29,1996.

15. Stolz, J.F., U.S. Nuclear Regulatory

5. Taylor, J.M., U.S. Nuclear Regulatory Commission, Letter to Byram, R.G.,

Commission, Memorandum to the Pennsylvania Power and Light Company, Commission, " Resolution of Spent Fuel "Susquehanna Steam Electric Station, Pool Action Plan Issues," July 26,1996. Units I and 2, Safety Evaluation Regarding Loss of Spent Fuel Pool Cooling Issues

6. U.S. Nuclear Regulatory Commission, (TAC No M85337)," June 19,1995.

Inspection Report 50-298/95-014, December 18,1995. 16. Washington Public Power Supply System, Washington Nuclear Plant Unit 2, Licensee

7. Northeast Nuclear Energy Company, Event Report 50-397/93-018, " Spent Fuel Millstone Unit 2, Licensee Event Pool Makeup Not Adequate to Mitigate Report 50-336/92-012, " Partial Loss of Accident Conditions," May 28,1993.

Normal Power (LNP)," January 17,1993.

17. Northeast Nuclear Energy Company,

8. U.S. Nuclear Regulatory Commission, Millstone Unit 1, Licensee Event Information Notice 88-065," Inadvertent Report 50-245/93-011-02, " Spent Fuel Pool Drainages of Spent Fuel Pools," August 18, Cooling Capacity," July 25,1996.

1989.

L

18. U.S. Nuclear Regulatory Commission,

9. Southern California Edison Co., San Onofre " Standard Review Plan," NUREG-0800, Unit 2, Licensee Event Report 50-361/88- revised periodically.

017-01, " Spent Fuel Pool Drainage Due to the Failure to Implement Updated Safety 19. U.S. Nuclear Regulatory Commission, Analysis (FSAR) Commitments," January 2, " Severe Accidents in Spent Fuel Pools in 1990. Support of Generic Issue 82," NUREG/CR-4982, July 1987.

10. Toledo Edison Co., Davis Besse, Licensee Event Report 50-346/82-007, March 3, 20. U.S. Nuclear Regulatory Commission, 1982. " Regulatory Analysis for the Resolution of Generic Issue 82, Beyond Design Basis

11. U.S. Nuclear Regulatory Commission, Accidents in Spent Fuel Pools,"

Augmented Inspection Team NUREG-1353, April 1989.

Report 50-321/86-41 and 50-366/86-41, January 8,1987. 21. Battelle Pacific Northwest Laboratory, Draft Report under NRC Contract

12. Salem Unit I and Unit 2, Event Notification 30528, May 22,1996.

35 NUREG-1275, Vol.12 1'

d Ref:rences DE-AC06-76RLO 1830," Risk Analysis for Nuclear Power Plants," WASH-1400, J

Spent Fuel Pool Cooling at Susquehanna October 1975.

Electric Power Station," October 1994.

23. Idaho National Engineering Laboratory,

22. U.S. Nuclear Regulatory Commission, "An " Loss of Spent Fuel Pool Cooling PRA:

Assessment of Risks in U.S. Commercial Model and Results," INEL #6/0334, September 1996.

l I

NUREG-1275, Vol.12 36

seicecan ses ummuotaaRRomutATonyoomessema 1.REpowT mAmER

, ami la me=sermac msvet.suse.,aw

~

72 345UOGRAPHic DATA SHEET NUREG-1275 Vol.12 2.Tma AmosusTma

, Operating En 's,ce Feedbeck Report l 3. DATE REPORT PueUSSD Assessment of Spent Fuel Cooling IdONm l YEAR l February 1997 I

4.PoloR ORANT MASER

s. AuTNoR(s) e. TYPE oPReponT J. G. Ibarre, W. R. Jones, G. F. Lenik, H. L Omstein, S. V. Pulleni TN ,

7.Peleoocov5NEp punmaeowed I l  !

, s. PERPoRneNo oRomperATIoM . NAAdE AND AOoRESS pMc, powde omdest oshe edtught U S Meer Auguhkry Ph eW memy adeau; #emeuser p=am=== aw==mvses==>

4

! Safety Programs Division

Office for Analysis and Evolustion of Operational Data l U.S. Nucieer Reguietory Commission l W

shington, DC 20666-001

, 9. SPoNeofeNG oRGAfGEATIoM NAAdE AND ADORES 8 7wic, W h m atmet #emesser. pawn Mac obeweg one er Amsms ut e numer m msy e-aw means emesseJ Some as 8. above I

i d

to.SLAIPLSENTARY NOTES i

1, i 11. ANTRACTmeamus eaume This report is an assessment of the likehhood and consequences of loss of spent fuel pool cochng in the nuclear power industry A genenc pressurtzed water reactor spent fuel pool confl0uration is developed, and a generic boiling water reactor spent fuel pool configuration is developed. Over twelve years of date is reviewed and mesessed Six sites visits were conducted to gather specific information on spent fuel pool ysical configuration, hcensee practices, and licensee procedures. The regulations on spent fuel pools were reviewed ndependent engineering assessments on the spent fuel pool system were performed on the electrical system, instrumentshon, heat loads, and radiation. An assessment on the risk of loss of fuel coohng was performed The overall cxmelusions are that the typical plant may need invsev&de in spent fuel '

instrumentation, operator procedures and training, and configuration control l 12.MEywoRoseEscserToms puet e.aram metse ==tmessen- mmamm,mesapet; n avaunun tramerr unlimited n escusuncLAmemCATION refuel outage ("'" M near boiling frequency unclassM (rme m,mq unclassified is.NuneER OF PAGEs 1s. PleCE NRc Fosted SIS (MBl This tonn mes sensemusmay peduced by Ehe Fedoni Forms, he

I 1

)

i i

1 i

i 1

i

}

i J

i i l 1

I i ,

! l l

J Printed on recycled j Paper 4

I i

i

Federal Recycling Program J

emmu, g - - ~ ~

UPERATING STATES NUC W REG ISSION d=re, WASHINGTON. DC USNRC

%T No. oc

'"T@&~

1 laNICV1xtivG1 TS #VCS 1C0555130531DUPLICATIC4 CC-GACM US N FOIA g

20555

'TV CC T>S i POR-NUREG op pu cr?

$' A SNIA6TCN l

' \._l _ . _ _ _