ML20151W652
| ML20151W652 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 09/10/1998 |
| From: | Nunn D SOUTHERN CALIFORNIA EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20151W656 | List: |
| References | |
| NUDOCS 9809160060 | |
| Download: ML20151W652 (21) | |
Text
. - . .-
pu som--u'o-d EDISON An LDISON IN7LRNATIONAL** Company
==
i September 10, 1998
]
U. S. Nuclear Regulatory Commission Document Control Desk Washington, D. C. 20555 l I
Gentlemen:
Subject:
Docket Nos. 50-361 and 50-362 Request for Exemption to 10CFR50.44,10CFR50, Appendix A, General Design Criterion 41, and 10CFR50, Appendix E, Section VI.
Proposed Technical Specification Change NPF-10/15-496 +
San Onofre Nuclear Generating Station, Units 2 and 3 (SONGS 2 & 3)
Reference:
Letter from Stephen D. Floyd (Nuclear Energy Institute) to Ashok Thadani (NRC) dated August 21,1997, in accordance with the provisions of 10 CFR 50.12, " Specific exemptions," Southern Califomia Edison (SCE) is requesting an exemption to the requirements of 10 CFR 50.44, " Standards for combustible gas control systems in light-water-cooled power reactors," Part 50, Appendix A, General Design Criterion 41, " Containment atmosphere cleanup," and Part 50, Appendix E Section VI," Emergency Response Data System."
The purpose of this exemption request is to remove requirements for hydrogen control systems from the SONGS 2 & 3 design basis. As such the consideration of hydrogen ;
generation will no longer be included in the design basis of SONGS 2 & 3. I Accordingly, amendment application number 180 for SONGS Unit 2 and amendment application number 166 for SONGS Unit 3 are enclosed, which remove the hydrogen control systems from the plant licenses and Technical Specifications.
This request is made in accordance with Task Zero of the proposed pilot program for 1 risk-informed, performance-based regulation as described the referenced letter. This
, request was discussed at a technical briefing of NRC staff on December 15,1997, and gi$jg with the Subcommittee on Probabilistic Risk Assessment (PRA) of the NRC Advisory A Committee on Reactor Safeguards on February 20,1998. l 0 Documentation supporting the exemption request is provided in Enclosure 1.
02 Enclosure 2 provides the amendment request and supporting documentation.
h
$< As documented in the Enclosures, the requested exemption will improve the safety hCY
> focus at SONGS 2 & 3 and represent a more effective and efficient method for k maintaining adequate protection of public health and safety.
.. .; o P. O. Ika 128 2 o' San Clemente. CA 92674-0128 944-% 8 1480 I?ax 949.%81440
7 Document Control Desk !
1
! If approved, the requested amendment will have a pes!!!te impact on the SONGS 2 & 3 Level 1 Probabilistic Risk Assessment (PRA) due to a reduction in operator errors by ;
permitting simplification of the Emergency Operating Instructions, thereby enabling l l operators to give priority to important safety functions immediately following postulated j plant accidents. The requested amendment will also have a positive impact on the
- Level 2 PRA and the Level 3 PRA by eliminating the potential for a stuck open l hydrogen purge valve during accidents.
i Should you have any questions, or desire additional information concerning this request for exemption and the proposed changes to the Technical Specifications, please contact me. j l
Very tr y yours,
\ ,)
A l l
l cc: E. W. Merschoff, Regional Administrator, NRC Region IV l
! J. A. Sloan, NRC Senior Resident inspector, San Onofre Units 2 & 3 J. W. Clifford, NRC Project Manager, San Onofre Units 2 and 3 i S. Y. Hsu, Department of Health Services, Radiologic Health Brarsch l I
i l
1
{
l l
l 1
l SOUTHERN CALIFORNIA EDISON
- SAN ONOFRE NUCLEAR GENERATING STATION l PROPOSED CIIANGE NUMBER NPF-10/15-496
( TO TIIE TECHNICAL SPECIFICATIONS FOR UNITS 2 AND 3 INDEX l
1 1 l ENCLOSURE 1 - Exemption Request Supporting Documentation
]
ENCLOSURE 2 - Amendment Application, Description And Supporting Documentation,
! Proposed Change PCN-496
. ATTACHMENT A - Existing Plant License and Technical Specifications, Unit 2
! l ATTACHMENT B - Existing Plant License and Technical Specifications, Unit 3 l l
ATTACHMENT C - Proposed Plant License and Technical Specifications, Unit 2 (redline and strikeout) !
l ATTACHMENT D - Proposed Plant License and Technical Specifications, Unit 3 (redline and ;
l strikeout) l ATTACH. MENT E - Proposed Plant License and Technical Specification pages, Unit 2 ATTACHMENT F - Proposed Plant License and Technical Specification pages, Unit 3 .
l f
I l
l I
i
-~
I l
{ i I
i l
l ENCLOSURE 1 EXEMPTION REQUEST SUPPORTING DOCUMENTATION 10CFR50.44 " STANDARDS FOR COMBUSTIBLE GAS CONTROL SYSTEM IN LIGHT-WATER-COOLED POWER REACTORS" 10CFR50, APPENDIX A, GENERAL DESIGN CRITERION 41 j
" CONTAINMENT ATMOSPHERE CLEANUP" j 10CFR50, APPENDIX E, SECTION VI
" EMERGENCY RESPONSE DATA SYSTEM" I
1'
. . . m . . . _ . ._ _. _ ._ - . _ _ _ . _ - . _ . . . . _ _ _ _ _ _
i i
t
1.0 INTRODUCTION
1.1 PURPOSE _;
This enclosure provides information in support of a request for exemption pursuant to Title 10 of the Code of Federal Regulations Part 50.12, " Specific Exemptions," from requirements contained j in 10CFR50.44 and 10CFR50, Appendix A, General Design Criterion 41, and Appendix E, '
Emergency Response Data System (ERDS). A technical briefing on this subject was given to
- Nuclear Regulatory Commission (NRC) staffin Rockville, Maryland on December 15,1997 and l discussed with the Subcommittee on probabilistic risk assessment (PRA) of the NRC Advisory ;
Committee for Reactor Safeguards on February 20,1998. i 1.2 REGULATORY REQUIREMENTS !
1.2.1 Requirements of 10CFR50.44 and 10CFR50. Anoendix A. General Desian Criterion 41
. and Annendix E. ERDS -
i
- . . 10CFR50.44, " Standards for combustible gas control system in light-water-cooled power reactors," and 10CFR50, Appendix A, General Design Criterion 41," Containment atmosphere cleanup," establish requirements for controlling the amount of hydrogen inside the reactor ;
containment following a postulated Loss of Coolant Accident (LOCA). These requirements ;
~
provide specific assumptions and methods to define the amount of hydrogen generated, the rate at ;
which the hydrogen is generated, and the requirements of a combustible gas control system to control the concentration of hydrogen in the containment following a design basis LOCA to below flammability limits. 10CFR50, Appendix E section VI requires that containment hydrogen concentration be transmitted via ERDS.
As applied to San Onofre Nuclear Generating Station Units 2 and 3 (SONGS 2 & 3), the regulations require the following:
e 'a means for control of hydrogen gas that may be generated, following a postulated LOCA by:
a metal-water reaction involving the fuel cladding and the reactor coolant; a radiolytic decomposition of the reactor coolant; and, a corrosion of metals.
e 2t he hydrogen control measures must be capable of:
a measuring the hydrogen concentration in the containment and transmitting the measurement via ERDS; 8
10CFR50.44(a) aloCFR50.44(b) and 10CFR50 Appendix E section VI.2.a.(i).(4) 1
I a insuring a mixed atmosphere in the containment; and, i a controlling combustible gas concentrations in the containment following a LOCA.
l
- 'it must be shown that following a LOCA, but prior to effective operation of the combustible i gas control system, either: l m an uncontrolled hydrogen-oxygen recombination would not take place; or, a the plant could withstand the consequences of uncontrolled hydrogen-oxygen recombination without loss of safety function.
- ' hydrogen recombiners internal to the containment ' capable of controlling the greater of(l) five times the amount of hydrogen generated by metal-water reaction calculated in demonstrating compliance with 10CFR50.46(b)(3) [less than 0.01 times the hypothetical amount generated if all cladding reacted], or (2) the amount of hydrogen that would result from reaction of all clad to a depth of 0.00023 inch. 5The hydrogen must be assumed to be generated over a 2 minute interval follow'mg a LOCA. l
- 'a combustible gas control system to maintain the concentrations of combustible gases i
following a LOCA below flammability limits. Such systems may be of two types.
l a those allowing controlled release from containment such as a purge system a those that do not result in a significant release from containment such as recombiners.
7 Such a system must control hydrogen as necessary following a LOCA to assure that containment integrity is maintained. It must meet redundancy and single failure requirements.
1.2.2 Criteria for Exemotions - 10CFR50.12 Reauirements The NRC has established certain criteria which permit any interested person to request specific exemptions to its rules and regulations provided special circumstances exist. These criteria are promulgated in 10CFR50.12, " Specific Exemptions:"
- a. The Commission may, upon application by any interested person or upon its own initiative, grant exemptions from the requirements of the regulations of this part which are-3 10CFR$0.44(c)(1)
'10CFR50.44(c)(3)(ii) 8 10CFR50.44(d)(1) 6 10CFR50.44(h)(2) 7 10CFR50 Appendix A General Design Criterion 41," Containment atmosphere cleanup" 2
l l
l
l (1) Authorized by the law, will not present an undue risk to the public health and i safety, and are consistent with the common defense end security. l (2) The Commission will not consider granting an exemption unless special ;
circumstances are present.
Special circumstances are identified in 10CFR50.12(a)(2). The special circumstance most relevant to Southern Cr.lifornia Edison's request is:
(ii) Application of the regulation in the particular circumstances would not serve the underlying purpose of the rule or is not necessary to achieve the underlying purpose of the mle.
Special circumstances may also be present with respect to: )
(iv) The exemption would result in benefit to the public health and safety that compensates for any decrease in safety that may result from the grant of the exemption.
(vi) There is present any other material circumstance not considered when the regulation was adopted for which it would be in the public interest to grant an ;
exemption. '
This enclosure provides documentation in support of Southern California Edison's (SCEs) request for an exemption.
2.0 TECHNICAL DISCUSSION 2.1 OVERVIEW The containment hydrogen control system was installed in SONGS 2 & 3 in accordance with the requirements of 10CFR50.44 and 10CFR50, Appendix A, General Design Criterion (GDC) 41, to control the hydrogen concentration inside the reactor containment, following design basis Loss of Coolant Accident (LOCA) conditions, below the hydrogen flammability limit of 4 volume percent (4%). The hydrogen control system design basis is provided in the SONGS 2 & 3 Updated Final Safety Analysis Report (UFSAR) Section 6.2.5.
2.2 DESIGN CONSIDERATIONS 2.2.1 System Descriotion The hydrogen control system consists of a hydrogen monneing subsystem, hydrogen recombiners, and a hydrogen purge subsystem. These are briefly described below.
3
.. __ _. __ . _ ._ _ .- . _ . _ . _ . ~ . _
Hydrogen Monitoring Subsystem The redundant hydrogen monitoring subsystem is a safety-related system that is designed to measure the hydrogen concentration inside the containment and to alert the operators in the control room of the need to activate the hydrogen recombiners or hydrogen purge system.
Following an accident, when directed by the Emergency Operating Instructions (EOls) to establish containment combustible gas control, the operators place the hydrogen monitoring system in operation and initiate its periodic calibration ' . hydrogen monitors become decalibrated due to temperature changes inside containnm. The system is only partially compensated for temprature changes.) The containment hydrogen concentration is currently transmitted via ERDS. The hydrogen monitoring subsystem is currently required by license conditions 2.C(19)i for Unit 2 and 2.C(17)d for Unit 3.
Hydrogen Recombiners Subsystem The hydrogen recombiner subsystem consists of two redundant recombiner trains which are permanently installed inside the containment. It is designed to maintain the hydrogen concentration below the lower flammability limit of 4%, consistent with the range of 4% to 6% as l specified in the Regulatory Guide 1.7. Each recombiner train consists of controls located on a control board in the main control room, a power supply cabinet located in the electrical penetration area adjacent to the containment building, and a recombiner located on the operating deck of the containment. There are no moving parts inside the recombiner. Air flows by natural convection through the unit at a rate of100 standard fWmin. Heating elements cause the hydrogen to chemically combine with atmospheric oxygen. As presently described in the SONGS 2 & 3 EOls, the hydrogen recombiners are manually started by the control room operators before hydrogen concentration reaches 3.0% indicated (3.5% actual). Figure 6.2-63B in the SONGS 2
& 3 UFSAR shows that this is postulated to occur at approximately 9 days after a design basis LOCA.
Hydrogen Purge Subsystem l
The hydrogen purge subsystem provides the capability for a controlled purge (filtered vent) of the l centainment atmosphere in order to maintain the hydrogen concentration below its flammability limit of 4% following a design basis LOCA. The hydrogen exhaust and supply trains consist of fihers, fans, fan heaters, and associated piping, valves, ductwork, dampers, instruments, and ;
controls. Supply and exhaust containment isolation valves are the only moving parts located inside the containment. The hydrogen purge subsystem is designed to purge the containment atmosphere at a rate of 50 standard it'/ min.
In the event of a design basis LOCA and failure of the hydrogen recombiners or increasing hydrogen concentration, the hydrogen purge subsystem may be utilized to control the hydrogen concentration inside containment. This is manually performed by the control room operators as 4
directed by the EOls. Since the purging of any amount of post-LOCA containment atmosphere is undesirable, the operation of the hydrogen purge subsystem would be initiated only when it has been determined that the hydrogen concentration level inside the containment is reaching about 3.5%. The Functional Recovery EOI for Containment Combustible Gas Control contains the following caution: "When the hydrogen purge unit is operated following a LOCA, then loss of Train B power will prevent closing the Containment Isolation Valve HV-9917, and dose rates may prevent closing valves manually, resulting in loss of containment integrity."
Valve HV-9917 is the inside-containment six inch hydrogen purge exhaust motor-operated valve.
The EOl directs that the hydrogen purge supply flow path of outside makeup air not be used following a LOCA because the containment would be pressurized and the supply flow path is not monitored. Thus, operability of only the hydrogen purge exhaust valves would be of concern during purging. If the air-operated outside-containment hydrogen purge exhaust containment isolation valve failed to seat and valve HV-9917 lost power to its motor operator while open, a loss of containment integrity could ensue due to high dose rates preventing operation of manual isolation valves in series with the containment isolation valves.
Ilvdrogen Mixing Hydrogen mixing within the containment is accomplished by the Containment Spray System (CSS), the Containment Emergency Fan Coolers (CEFCs), the Containment Dome Air Circulators (CDACs), and the internal structure design, which permits convective mixing and prevents hydrogen entrapment. These systems and the internal structures of the containment such as intermediate floors and other internal structures are designed to maintain a well-mixed containment atmosphere, and to prevent hydrogen pocketing.
The equipment for hydrogen mixing is not part of the hydrogen control system. The equipment for hydrogen mixing starts on automatic signals following a LOCA to remove heat and fission products from the containment atmosphere, as well as to minimize localized hydrogen buildup inside containment. This exemption request proposes no changes to the hydrogen mixing equipment.
2.2.2 Imoact of Hydrogen Control on Defense-in-Denth Design As explained below, the SONGS 2 & 3 defense-in-depth accident control design is unaffected by the proposed exemption request due to (1) existing margin in the containment design, and (2) the minimal impact of the hydrogen control system on the ability of the containment to withstand challenges due to hydrogen production following a design basis LOCA. Furthermore, (3) due to its limited capacity, the hydrogen control system has no value in defense against containment failure resulting from hydrogen buildup inside the containment following severe accidents.
5
l I
1
- 1. ]mpact of Ilydrogen Control on Containment Safety Marcin SONGS 2 & 3 employs a large, dry containment design with a design pressure of 60 psig.
This type of PWR containment is believed to be the least susceptible to damage from a hydrogen burn. The hydrogen burn during the 1979 event at Three Mile Island 2 (TMI-2) with '
a hydrogen concentration of about 8.1% resulted in a containment peak pressure of about 28 psig, well below the containment design pressure of 60 psig (NS AC-22,1981). The SONGS 2 & 3 containment with a similar design has sufficient safety margin against hydrogen burn i following design basis and severe accidents without use of the hydrogen control system.
Following a design basis LOCA without operation of the hydrogen control system, the hydrogen concentration would realistically remain below the flammability limit of 4% (see Section 2, below), and hence the containment integrity would not be challenged, for at least i 30 days. The containment peak pressure will remain below the SONGS 2 & 3 containment design pressure of 60 psig during this time (see Figure 6.2-5 in SONGS 2 & 3 UFSAR). A <
time period of 30 days would be sufticient for determining whether action would be required. I Beyond 30 days, hydrogen would accumulate inside the containment and could reach the j flammability limit. However, containment failure due to hydrogen combustion is unlikely, based on the results of the SONGS 2 & 3 Individual Plant Examination (IPE) study. The SONGS 2 & 3 IPE concluded that for the worst case accident sequence with respect to hydrogen combustion, it is unlikely that enough hydrogen would accumulate to produce a hydrogen burn that could challenge the containment ultimate pressure capacity. The IPE further states that none of the accident sequences addressed in the SONGS 2 & 3 IPE could realistically threaten containment due to hydrogen combustion.
Both the nuclear industy and the NRC conducted numerous analyses and tests follow'mg the event at TMI-2 in 1979 to determine the containment capability of pressurized water reactor plants with large, dy containments. For example, NUREG/CR-5662 (1991) reports the computed containment peak pressure due to global hydrogen burn based on a 75% fuel cladding metal-water reaction (MWR) (which can be expected to occur during severe accidents) for a group of pressurized water reactor plants with large, dry containments, similar to the SONGS 2 & 3 containments. The reported containment peak pressure values are all within the plants' estimated containment capacities. Therefore, the NRC-sponsored study concludes that it seems unlikely that containment integrity would be threatened by a hydrogen burn from a 75% MWR in the containments examined. The 75% MWR estimate was intended to be representative of a range of core melt accidents. It should be noted that the TMI-2 accident involved about 45% MWR which resulted in a hydrogen concentration of about 8.1% (NUREG/CR-4330, Volume 3,1987). The NRC concluded that the large, dry containments could withstand the containment pressure following severe accidents and there was no need to backfit these containments with " glow plug" igniters or to inert the containment atmospheres.
6
L
! J l
A detailed plant-specific containment integrity analysis'for SONGS 2 & 3 indicates that the j L ' containment leak pressure is about 99 psig at a 95% confidence level, and the containment L - rupture pressure is approximately 139 psig at a 95% confidence level (SONGS 2 & 3 IPE, l 1993). ' Hence, a safety margin exists for containment integrity at higher hydrogen l reacentration levels beyond 30 days following a design basis LOCA, without the use of a j hydrogen control system.
With respect to equipment survivability, NUREG/CR-5662 states:
- j " Equipment survivability depends on the specific plant design and on the containment
!~
environment during a specific accident. The large-scale Nevada test site experiments ;
demonstrated that various types of plant equipment are capable of operating successfully j when subjected to the severe thermal environments associated with large-volume hydrogen burns. ;
"The recent analytical and experimental study performed at Sandia National Laboratories -
showed that the simulated equipment can withstand a LOCA and single burn resulting from a 75% MWR in a large, dry containment. However, the multiple burn due.to the operation ofignition systems could pose a serious threat to safety-related equipment located in the source compartment."
i o
lt should be noted that the SONGS 2 & 3 containments do not have " glow plug" igniters.
This reduces the potential for multiple burns. During the TMI-2 accident, containment was not breached and damage inside containment was essentially limited to plastics and other low melting point materials such as telephone cases and the crane operator's seat (NUREG/CR- l 4330, Volume 3,1987). !
i Summarv of Safety Marvin Imnact l
For pressurized water reactor plants with large, dry containments a safety margin remains for l containment rupture from hydrogen burn or detonation at higher hydrogen concentration I levels during severe accidents or beyond 30 days following a design basis LOCA, without using any hydrogen control system. Additionally, the NRC has determined that pressurized '
water reactor plants with large, dry containments can withstand the containment pressure following severe accidents and there was no need to backfit these containments with " glow plug" igniters or to inert the containment atmospheres.
- 2. Imnact of Hydrogen Control on Design Basis Accidents The containment hydrogen control system is provided in accordance with the requirements of 10CFR50.44 and 10CFR50, Appendix A, GDC 41 to control the concentration of hydrogen
, which may be released into the reactor containment following postulated design basis
. accidents. The containment hydrogen control system is designed to ensure that the hydrogen I
t 1 7 l
3- . - - - .- e .,, , , _. . _,----,.-.- -, -
l l
l l concentration is maintained below the flammability limit of 4% following a design basis l LOCA. Additionally, the SONGS 2 & 3 safety-related containment systems (CSS, CEFCs, l and CDACs) and the internal containment structural design provide excellent hydrogen mixing .
l capability inside the containment that would prevent hydrogen pocketing following a I postulated design basis LOCA. ]
The hydrogen control system design basis is provided in SONGS 2 & 3 UFSAR, Section !
6.2.5. UFSAR Figure 6.2-63B shows predicted hydrogen concentration versus days after the occurrence of a design basis LOCA for various sources of hydrogen, as well as the total hydrogen concentration. The figure also shows the rate of hydrogen removal by the hydrogen recombiners or the hydrogen purge system when initiated manually by the control room operators. The air flow volume of each hydrogen recombiner train is 100 standard fl'/ min.
The hydrogen recombiners are manually started by the control room operators before hydrogen concentration reaches 3.5%, which occurs approximately 9 days after a design basis LOCA. The operating range of the hydrogen recombiners is between 1% and 3.5%. The control room operators would shut off the hydrogen recombiners when the hydrogen 4 I
concentration reaches about 3.5%, as directed by the EOls, to prevent hydrogen ignition by the heater elements. The hydrogen purge evstem is designed to purge the containment atmosphere at a rate of 50 standard ft'/ min. In the event of a design basis LOCA together with failure of the hydrogen recombiners, upon hydrogen concentration reaching 3.5%, the EOls permit use of the hydrogen purge system to control hydrogen concentration. The hydrogen purge system is manually initiated by the control room operators in accordance with I the EOls.
UFSAR Figure 6.2-63B shows hydrogen concentrations of approximately 1.7% after the first !
day and 3.5 % at about 9 days after a postulated design basis LOCA. Using the current assumptions, the hydrogen concentration would reach the flammability limit of 4% at about i 13.5 days and 5.3% at about 30 days following a design basis LOCA, given that neither the ;
hydrogen recombiners nor the hydrogen purge system is started by the control room operators.
Hydrogen production calculations in the UFSAR for a design basis LOCA are based on a number of conservative assumptions. For example, the design basis analysis considers a factor of five increase for hydrogen generation by the metal-water reaction (i.e., affecting the Zr-Water Reaction curve in Figure 6.2-63 A) over the maximum amount calculated in accordance with 10CFR50.46 (UFSAR Section 6.2.5.3.A.2). The UFSAR also assumes that 1% of the l fission product solids in the core mix with the LOCA water, causing radiolytic hydrogen production (i.e., affecting the Radiolysis curve in Figure 6.2-63A). Assuming 1% fission product solids in LOCA water is more appropriate for severe accidents than for design basis accidents.
l Also, the UFSAR over predicts the hydrogen concentration level resulting from the radiolysis l by 20% as stated in the UFSAR source document (Stdard Review Plan Section 6.2.5, 8
I l
l Appendix A, p. 6.2.5-13,1981). Another conservative assumption in the UFSAR analysis is ,
the use of 0.5 molecule /100 ev for the hydrogen yield during radiolysis. The published data l from ORNL (ORNL-NSIC-23,1968) support a net yield of 0.3 molecule /100 ev for hydrogen l under conditions similar to those of the containment sump (UFSAR, Section 6.2.5.3.A.1). l The Zr-water generation curve is flat at a constant hydrogen volume of about 1%, and is the dominant hydrogen production method in the short term (i.e., the first five days post-LOCA),
whereas, the radiolysis curve steadily increases from zero to about 3% over a period of 30 days post-LOCA. Radiolysis is therefore the dominant hydrogen production method in the long term (i.e., more than five days post-LOCA).
In order to gain an understanding of hydrogen production under more realistic conditions than those assumed in the UFSAR, Southern California Edison (SCE) utilized the hydrogen ,
generation rates described in UFSAR Sections 6.2.5.3. A.1 (radiolysis) and 6.2.5.3.A.2 (Zr- J water reaction) with the following modifications: 1
conservatism specified in Standard Review Plan Section 6.2.5 in the isotope energy production rates;
- 2. for the amount of hydrogen generated by radiolysis in the sump, SCE modified the hydrogen yield of the reaction from the 0.5 molecule /100 ev in UFSAR Section 6.2.5.3.A.1 to the value of 0.3 molecule /100 ev in ORNL-NSIC-23 to obtain a hydrogen l
yield more closely corresponding to the yield under conditim expected in the sump; and, l 1
1
- 3. for the amount of hydrogen generated by the Zr-water reaction prescribed by 1 I
10CFR50.44(d)(1), SCE used the 0.23 mil cladding penetration depth criterion rather than the factor of five increase used in UFSAR Section 6.2.5.3. A.2; Hydrogen generation from all other sources was taken directly from UFS AR section 6.2.5.3 with no change.
Under the more realistic assumptions, the hydrogen concentration will be less than 4% at 30 l days following a design basis LOCA without operation of any hydrogen control system.
As a sensitivity analysis, for a higher radiolytic hydrogen yield in the sump of 0.4 molecule /100 cv (i.e., mid-way between the values based on ORNL data and the UFSAR), the evaluation shows that the hydrogen concentration will still not exceed 4% at 30 days post-LOCA without operation of any hydrogen control system.
Therefore, realistically, the hydrogen concentration would not reach the flammability limit of 4%, and hence there would be no challenge to containment integrity, for at least 30 days post-LOCA, even without operation of the hydrogen recombiners or the hydrogen purge system.
i l
i
l The containment peak pressure will remain below the SONGS 2 & 3 containment design pressure of 60 psig during this time (see Figure 6.2-5 in SONGS 2 & 3 UFSAR). A time l
period of 30 days would be suflicient for determining whether action would be required. '
Beyond 30 days, hydrogen would accumulate inside the containment and could reach the i flammability limit. However, containment failure due to hydrogen combustion is unlikely, ;
based on the results of the SONGS 2 & 3 Individual Plant Examination (IPE) study. The !
SONGS 2 & 3 IPE concluded that for the worst case accident sequence with respect to hydrogen combustion, it is unlikely that enough hydrogen would accumulate to produce a hydrogen burn that could challenge the containment ultimate pressure capacity. The IPE further states that none of the accident sequences addressed in the SONGS 2 & 3 IPE could realistically threaten containment due to hydrogen combustion.
There is also no potential for containment integrity to be challenged due to hydrogen pocketing, based on SONGS 2 & 3 containment internal stmetural design (generous vent paths) and availability of safety-related containment systems that provide sufTicient hydrogen mixing (CSS, CEFCs, and CDACs). The results of a study for several PWR plants with large ;
dry containments indicated that, depending on the contaimnent volume and fan capacity, a ,
mixing of the total containment air volume by fans alone would take only 10 to 30 minutes for j the PWRs examined (NUREG-CR-5662, Section 2.3). The time required to process one i containment volume for SONGS was shown to be 25 minutes.
Summary of Design Basis Accident Impact The hydrogen control system is designed to maintain the hydrogen concentration level below its flammability limit during design basis accidents. However, without operation of the hydrogen control system, the hydrogen concentration would realistically be expected to remain below the lower flammability limit of 4%, and hence the containment integrity would j not be challenged, for at least 30 days following a design basis accident. Beyond 30 days, l hydrogen concentration may reach the flammability limit. However, containment failure due l to hydrogen combustion is unlikely based on the results of the SONGS 2 & 3 IPE study. l
- 3. Impact of Hydrogen Control on Severe Accidents For severe accidents, i.e., those beyond the design basis, containment hydrogen concentrations in the range of 10% over short periods of time are possible, as demonstrated at the TMI-2 accident in 1979. The hydrogen control system is designed to maintain the hydrogen concentration level below the flammability limit of 4% during design basis accidents that result in small amounts of hydrogen produced slowly over long periods of time--i.e., many days.
For severe accidents during which containment hydrogen concentration will rapidly rise to above the 4% level, the present hydrogen control system is undersized, and hence would provide no benefit to hydrogen concentration control and containment performance. An NRC-sponsored study (NUREG/CR-5567,1990) corroborates this point by stating that the 10
hydrogen control systems are designed to accommodate hydrogen accumulation for design basis events (oxidation of 5% Zircalloy surrounding the active fuel). These systems are not designed for the hydrogen generation that might accompany a core meltdown. Consequently, the hydrogen control system was determined to be ineffective in mitigating hydrogen in the SONGS 2 & 3 IPE study. Subsequent to the TMI-2 accident, improvements in equipment, operator training, and procedures make it extremely unlikely that a severe core damaging event comparable to TMI-2 would occur at SONGS 2 & 3. Nevertheless, the SONGS 2 & 3 IPE worst case scenario was based on a hypothetical hydrogen concentration of 11.5%,
compared to an actual hydrogen concentration of about 8.1% during the TMI-2 hydrogen burn.
The hydrogen recombiners are ineffective at processing hydrogen at the higher rates expected to be generated during severe core damage accidents. The SONGS 2 & 3 hydrogen recombiners have a 100 standard ft'/ min capacity each, and would be placed in operation when the hydrogen concentration is in the range of 1% to 3.5%. The EOls instmet the control room operators to turn ofTthe hydrogen recombiners when the hydrogen concentration reaches about 3.5% to prevent hydrogen ignition. For hydrogen concentrations close to the flammability limit of 4%, the EOls recommend not using the hydrogen recombiners.
The hydrogen purge subsystem is also unlikely to be used after a severe core damage accident due to the concern about the capability of the hydrogen purge exhaust inside-containment ;
isolation valve to secure the path afler venting, given a high containment pressure situation.
This potential inability to secure the purge pathway due to loss of Train B electrical power could result in an open release path to the environment, a situation the EOls caution against.
Moreover, the hydrogen purge system has a limited capacity of 50 standard fl S/ min. '
The usefulness of the hydrogen monitoring subsystem is also limited during severe accidents.
The maximum range of the hydrogen monitoring system is 10%, whereas in severe core damage accidents the hydrogen concentration is likely to exceed 10%. An alternate method !
for determining hydrogen concentration, should it become necessary, is containment atmosphere samples taken from the Post Accident Sampling System, as recommended by the Combustion Engineering Owners Group Accident Management Guidelines planned for adoption at SONGS 2 & 3.
Summary of Severe Accident Impert The usefulness of the hydrogen control system is limited to design basis accidents. The system is undersized for severe accidents, and hence provides no benefit for these accidents.
2.2.3 Conclusion The proposed exemption does not affect the SONGS 2 & 3 defense-in-depth design due to (1) 11
existing margin in the containment design, and (2) the minimal impact of the hydrogen control system on the ability of the containment to withstand challenges due to hydrogen production following a design basis LOCA. Furthermore, (3) due to its limited capacity, the hydrogen control system provides no benefit for severe accidents.
2.2.4 Additional Considerations Risk Reduction Due To Instruction Simnlification In a postulated LOCA, the SONGS 2 & 3 EOls direct the control room operators to monitor and control the hydrogen concentration inside the containment after they have carried out the steps to maintain and control the higher priority critical safety functions such as reactivity, RCS inventory, RCS pressure, and core heat removal. The key operator actions in controlling the hydrogen concentration are to place the hydrogen recombiners or hydrogen purge system in operation.
These two actions involve many procedural steps and require coordination between the control room operators and other work groups, such as Instrumentation and Control (to perform periodic hydrogen monitor calibrations), Chemistry, and Health Physics (before placing the hydrogen purge in operation).
These hydrogen control activities could potentially distract operators during the extremely busy period following an accident, and could therefore have a negative impact on the higher priority critical operator actions. As discussed paviously, these hydrogen control activities are of minimal to no benefit in mitigating accidents.
An exemption from the requirements for a hydrogen control system will eliminate the need for EOI steps for hydrogen control and hence simplify the EOls. This will have a positive impact on public health risk by reducing the probability of operator error during potential accidents and hence reduce the core damage frequency. An exemption will allow the operators to address hydrogen controlissues as part of the proposed Accident Management Guidelines, which cover operator actions at long time frames following accidents. j An exemption from the requirements for a hydrogen control system will also eliminate the need for EOI steps to initiate a hydrogen purge of the containment. This would result in a lower l probability of a failed-open containment purge valve. Consequently, the offsite doses would be l reduced due to the reduction of the probability of a failed-open containment purge valve.
Summarv of Risk Reduction Due To Instruction Simplification The changes described in this exemption request result in a " risk positive" change. Exemption from the requirements for a hydrogen control system, and, as such, the consideration of hydrogen I generation within the SONGS 2 & 3 design basis, will eliminate the need for EOI steps for hydrogen control and hence simplify the EOls, resulting in lower operator error probabilities.
l Elimination of the EOI steps to initiate the containment hydrogen purge will result in a lower l
12 1
l
i :
! I probability of a failed-open containment purge valve, resulting in lower large early release l probabilities.
l l 2.3 INDUSTRY EXPERIENCE l !
l The regulatoy requirements for containment hydrogen control systems were based on knowledge that existed before the TMI-2 event in March 1979. Following TMI-2, the nuclear industry and I the NRC initiated extensive analysis and testing to increase the scope of knowledge concerning hydrogen generation and hydrogen control following severe accidents. This new knowledge 1 l invalidated many of the assumptions and methods in the regulations. Based on the new
- knowledge, it became clear that hydrogen control systems designed for design basis LOCA conditions were not adequate to maintain the hydrogen concentration below the flammability limit ;
of 4 volume percent in severe accidents. Following TMI-2, the nuclear industry performed ;
extensive analysis and testing which indicated that for large, dry containments, the containment j would withstand the burn oflarge amounts of hydrogen generated in severe accidents. Therefore, 1 the required hydrogen control systems were determined to be unnecessary for design basis LOCA conditions, and ineffective for severe accidents (i.e., significantly beyond design basis accidents).
1 In addition, the Nuclear Regulatoy Commission conducted analyses with respect to backfitting l the installation of" glow plug" igniters to replace the hydrogen recombiners in nuclear units with l large, dry containments. The NRC determined that the requirement for " glow plug" igniters ,
could not be justified for nuclear units with large, dy containments according to the provision of l 10 CFR 50.109. This was because large, dry containments have a greater ability to accommodate 1 the large quantity of hydrogen associated with a degraded core accident than the smaller containments.' To date, the nuclear units with large, dry containments rely exclusively on the containment structure to withstand any postulated uncontrolled burn of hydrogen gas generated in severe accidents.
2.4 CONCLUSION
The hydrogen control system is of no benefit in severe accidents. Elimination of the present requirements in the EOls for the initiation of the hydrogen monitoring subsystem, the hydrogen recombiners, and the hydrogen purge subsystem will result in a " risk positive" change.
Under realistic assumptions following the design basis LOCA, the hydrogen recombiners and the hydrogen purge subsystem are not needed for control of hydrogen concentrations inside the reactor containment during the first 30 days following the design basis LOCA. Consequently, the hydrogen monitoring subsystem provides no information needed for accident mitigation.
l
'NUREG/CR-5662,1991 13 1
l 3.0 EXEMPTION CRITERIA OF 10CFR50.12 3.1 OVERVIEW The present compliance with 10CFR50.44,10CFR50, Appendix A, General Design Criterion 41, and 10CFR50 Appendix E, ERDS ("the rule") at SONGS 2 & 3 does not serve the underlying purpose of the rule and is not usefulin achieving the underlying purpose of the rule. The underlying purpose of the rule was to provide assurance that the containment would not fait due to combustible gas accumulation and ignition in accident situations where fission products were present in the containment. The reliance on the design basis LOCA conditions as described in the rule was ineffective in achieving this result.
The TMI-2 accident produced hydrogen in quanthies far exceeding the assumptions in 10CFR50.44, and, even though an uncontrolled hydrogen burn did occur, the containment did not fail.
Probabilistic risk assessments (PRA) quantify the probabilities and consequences of similar accidents. In the PRAs performed for the SONGS 2 & 3 IPE and IPE for External Events (IPEEE), the hydrogen control system was determined to be ineffective in addressing hydrogen concentrations in severe accidents. The hydrogen control system diverts operator attention from more important actions and creates the potential for uncontrolled containment release through stuck open purge valves.
As described below, the requested exemption to the requirements of 10CFR50.44,10CFR50, Appendix A, General Design Criterion 41, and 10CFR50 Appendix E, ERDS satisfies the requirements of 10CFR50.12. The purpose of this exemption is to remove the requirements for hydrogen control systems from the songs 2 & 3 DESIGN BASIS. As such the consideration of hydrogen generation will no longer be included in the design basis of SONGS 2 & 3.
3.2 NO UNDUE RISK Section (a) (D IThere is no rmdue risk to thepublic health and safervi As stated earlier, eliminating the hydrogen control requirements does not affect the SONGS 2 & 3 defense-in-depth and safety margin due to the fact that the realistic hydrogen concentration will remain below the flammability limit of 4% and hence containment integrity will not be realistically challenged for at least 30 days following a design basis LOCA, without operation of the hydrogen control system. Beyond 30 days, hydrogen concentration may reach the flammability limit.
However, containment failure due to hydrogen combustion is unlikely based on the results of the SONGS 2 & 3 IPE study. Furthermore, the usefulness of the hydrogen control system is limited to design basis accidents that result in small amounts of hydrogen produced slowly over long periods of time (many days), and the system has no benefit for severe accidents.
l l
14 l
A detailed SONGS 2 & 3-specific containment integrity analysis indicates that the containment leak pressure is about 99 psig at a 95% confidence level, and the containment nipture pressure is i about 139 psig at a 95% confidence level. Hence, a safety margin exists for containment integrity at higher hydrogen concentration levels beyond 30 days following a design basis LOCA, without l using any hydrogen control system.
Additionally, eliminating the hydrogen control requirements for the SONGS 2 & 3 large, dry containment has a positive impact on the risk to the public by reducing the potential operator l error probabilities due to simplifying the EOls (eliminating their hydrogen control steps). l Elimination of requirements for containment hydrogen purge will reduce the likelihood of a failed- l open containment purge valve, resulting in lower release risks (i.e., large early release frequency). ,
The risks associated with offsite health effects would be reduced due to the reduction in the l probability of a failed-open containment purge valve. {
3.3 UNDERLYING PURPOSE OF THE RULE NOT SERVED Section (a)(2) (ii) IAvolication of the reenlation in the particular circumstances wouhlnot serve l
the underivine vurpose of the rule and is not necessary to achieve the underivine vurpose of the i Gdf.l The underlying purpose of the rule was to reduce the probability of failure of the containment during accidents and thus prevent fission products from the reactor core from being released l through the containment during accidents. Application of the rule at SONGS 2 & 3 has resulted in equipment and procedures that have no impact on the probability of failure of the containment under conditions where fission products from the reactor core exist in the containment. In the SONGS 2 & 3 IPE and IPEEE, the hydrogen control systems were determined to be ineffective in controlling hydrogen concentrations in severe accidents. The hydrogen control systems divert operator attention from more important actions and create the potential for containment bypass j through stuck open purge valves.
3.4 BENEFIT TO PUBLIC HEALTH AND SAFETY l Section (a)(2) (iv) I7'here is a benent to the public health and safervi Implementation of the exemption from the hydrogen control requirements would achieve a benefit to the public health and safety. In addition to the direct positive impact on the public health and safety by reducing the public risk (see Section (a)(1) above), there is also an indirect safety benefit to the public. The indirect benefit comes from eliminating unnecessary requirements from the SONGS 2 & 3 Technical Specifications and EOls. The recent NRC statement on compliance versus safety': " Requirements that are duplicative, unnecessary, or unnecessarily burdensome can actually have a negative safety impact," is a recognition of the indirect safety benefit of the
'NRC Inspection Manual Part 9900, " Operations - Safety and Compliance," issued 09/09/97 i
15
1 proposed exemption.
i 3.5 MATERIAL CIRCUMSTANCES NOT CONSIDERED Section (a)(2)(vi) [There are present material circumstances not considered when the reerdatinu fi.e. 10CFRSO.44) was adopted 1 Experience and information obtained over time provide a better perspective about hydrogen !
generation and the impact of hydrogen burning on containment integrity and safety equipment j during accidents. Two important material circumstances are (a) the effects and (b) the risks of i hydrogen generation. )
l
- a. Effects ofhydroeen eeneration l
Traditionally, technical and regulatory evaluation perspectives have held that a hydrogen burn is to be avoided due to the uncertainties ofcontainment failure. The TMI-2 accident in March 1979 provided an important benchmark for the effects of a hydrogen burn on safety equipment and contairunent integrity. TMI-2, which involved about a 45% core cladding-water reaction, resulting in about 8.1% hydrogen concentration, produced no containment breach and minimal damage to equipment (NUREG/CR-4330, Vol. 3,1987). The containment peak pressure was about 28 psig, well below the containment design pressure of 60 psig. Containment damage was essentially limited to plastics and other low melting point materials such as telephone cases and the crane operator's seat. The TMI-2 hydrogen burn !
thus provides actual experience which establishes a significantly higher threshold for containment damage than was thought to be available when the regulations were promulgated,
- b. Risks ofhvdrocen veneration ,
Many PRA evaluations (e.g., plant-specific IPEs) and tools (e.g., MAAP code) have been developed which provide a better insight about the risks of hydrogen generation and burning during severe accidents than were available when the regulations were promulgated. The SONGS 2 & 3 IPE study concluded that the containment failure due to hydrogen combustion during severe accidents is unlikely.
3.6 CONCLUSION
As discussed above, this exemption request is in compliance with 10CFR50.12, specifically, with applicable Sections (a)(1) and (a)(2)(ii). The discussion has demonstrated (1) that granting the exemption will not present an undue risk to public health and safety, and (2) that application of the rule in the particular circumstance would not serve the underlying purpose of the rule and is
' not necessary to achieve the underlying purpose of the rule. Additionally, special circumstances may also exist with respect to (3) Section (a)(2)(iv) and (4) Section (a)(2)(vi).
16 l
l
i ENCLOSURE 2 AMENDMENT APPLICATION, DESCRIPTION AND SUPPORTING DOCUMENTATION l OF PROPOSED CIIANGE NPF-10/15-496 !
i 1
,