ML13304B573
| ML13304B573 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 03/31/1980 |
| From: | Randy Hall BROOKHAVEN NATIONAL LABORATORY |
| To: | Ferguson R Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8004080515 | |
| Download: ML13304B573 (17) | |
Text
REGULATORY INFORMATION DISTRIB3UTION SYSTEM (R I-D)
ACCESSION R8R:8004000515 0C.DATE: 80/03/31 NOTARIZED: NO DOCKET FACIL:50-206 San Onofre Nuclear Station, Unit 1, Soutrnern Californ 09929 50-361 San Onofre Nuclear Station, Unit 2, Southern Californ C 0036 50-362 San Onofre Nuclear Station, Unit 3, Southern Californ AUTH.NAME AUrHOR AFFILIATION HALLH.E.
Brookhaven National Laboratory RECIP.NJAME RECIPIENT AFFILIATION FERGUSONR.L.
Plant Systems branch
SUBJECT:
Forwards review summary of fire orotection covering Items 3.1.b,foam sucoression sys,3.1.7,gas suopression sys 3.2.3, turbine bldg structure.Items 3.1,1,3.1.2,3.1.5 are partially comoleted & summary will be provided accordinqly.
DISTRISUTION CODE: AOObS COPIES RECEIVED:LTR -1 ENCL SIZE:
TITLE: Fire Protection Information (After Issuance of OP.
Lic.)
,N 0 T ES:A----
RECIPIENT COPIES RECIPIN4T COPIES ID CDDE/'JAME LrTR ENCL 10 CODE/NAMIE LTTq ENCL ACTION:
05 6C 0,60-9-d 4
4 INTERqNAL:
01 4L-1 1
02 NRC P!R 1
09 I&E 2
2 12 AUXIL SYS SR 2
2 14 FLANJT SYS 9R 5
5 19 AAMBACH 1
1 20 r YURANA AR 1
1 UELD 1
0 EXTERNAL:
03 LPOR 1
1 04 NSIC I
I 22 ACRS 1.6 16 APR 10 1980 38 37 TOTAL NUMiHFR OF COPIFS EOJIRED: LTTR ENCL
-35
BROOKHAVEN NATIONAL LABORATORY ASSOCIATED UNIVERSITIES, INC.
Upton, New York 11973 Department of Nuclear Energy (516) 345-2144 March 31, 1980 Mr. Robert L. Ferguson Plant Systems Branch U.S. Nuclear Regulatory Commission Washington, D.C.
20555 RE:
San Onofre, Fire Protection Review, Items 3.1.1(5), 3.1.2(b), 3.1.5(b),
3.1.5(d), 3.1.6, 3.1.7(a), 3.1.7(b), 3.1.15(b), 3.2.3.
Dear Bob:
The items for San Onofre that we have fully reviewed are Items 3.1.6, Foam Suppression Systems, 3.1.7, Gas Suppression Systems, and 3.2.3, Turbine Building Structure. The other items referenced are partially completed.
This completes the fire protection review of San Onofre for all the items we have on hand.
Respectfully yours, Robert E. Hall, Group Leader Reactor Engineering Analysis REH:EAM:sd attachment cc.:
J. Benaroya M. Levine wo/att.
L. Derderian E. MacDougall D. Eisenhut V. Panciera W. Kato wo/att.
E. Sylvester 3004 08ij615
/1
SAN ONOFRE Fire Protection Review Item 3.1.1(5) -
Fire Detection Systems Subitem (5) under Item 3.1.1 of the San Onofre Unit SER indicated the licen see's proposal to provide additional smoke detectors in the turbine lube oil reservoir area of the turbine building.
Item 3.1.1 states:
"Early warning automatic fire detection systems will be provided in the following areas:
(5) Additional smoke detectors will be provided in the turbine lube oil re servoir area of the turbine building."
The turbine lube oil area at the north end of the turbine building contains the lube oil *reservoir, lube oil conditioner, lube oil pumps, and 15,000 gallons of lube oil.
In addition, numerous redundant safety-related cables in trays and conduit are routed through this area. The present smoke detection system is shielded from much of the cable tray area by a metal deck mezzanine, with the detectors in the area above the mezzanine and the cables in the area below it.
The additional detectors are to be installed to provide detection in the area below the mezzanine.
In their letter dated January 23, 1980, the licensee stated that four ioniza tion detectors would be installed between the cable tray runs on the underside of the mezzanine floor plating, and submitted drawings showing the detection system layout. The licensee also stated their intent to conduct in-situ tests to verify the adequacy of the detector arrangement, and to make appropriate modifications if necessary.
The drawings submitted indicate that the four detectors will be installed be tween cable trays in an area centering on column B-7.
The closest detector to the east wall of the area is 22 ft. and to the north all is 23 ft.
Based on a review of the drawings and data submitted by the licensee, we find that the proposal to provide smoke detection capability below the mezzanine in the turbine lube oil area by placing four ionization detectors as shown on the drawings is unacceptable.
NFPA 72E-1978 recommends spacing for spot-type detectors in section 4-3.5.1, which states:
"On smooth ceilings, spacing of 30 feet (9 m) may be used as a guide. In all cases, the manufacturer's recommendations shall be followed. Other.
spacing may be used depending on ceiling height, different conditions or response requirements."
The 30 foot recommendation of NFPA 72E is used since the Underwriters Labora tories tests of smoke detectors are conducted using that spacing.
With 30 feet between detectors, the distance to any wall should be no more than 15 feet for a smooth ceiling, and the proposed installation would exceed even this cri teria, and is therefore not in agreement with the guidelines of NFPA 72E.
For the licensee to provide adequate smoke detection capability below the mez zanine in the turbine lube oil area, we recommend that a smoke detection system which complies with the guidelines of NFPA 72E 1978 be installed. This would require that the maximum detector spacing be 30 feet with no more than 15 feet 1
from any wall to a detector. Because of the congestion in the area due to cable trays, the maximum recommended spacing may need to be reduced to compen sate for the obstructions.
We recommend that the staff request the licensee to submit revised plans to the NRC for review and approval prior to implementation, and the proposed in-situ tests should be conducted to verify the adequacy of the instalation when completed (See SER item 3.2.2).
Item 3.1.2(a) Fire Water Supply Fire Pump Test Feature The first paragraph under Item 3.1.2 of the San Onofre Unit 1 SER, identified here as 3.1.2(a), indicated the licensee's proposal to provide test features for the fire pumps which meet the requirements of NFPA 20.
Item 3.1.2(a) of the SER states:
"Test features will be provided for the fire pumps which meet the require ments of NFPA 20."
Section E.2(c) of Appendix A to BTP 9.5-1 requires that the pump installation conform to NFPA 20 requirements as a minimum. Section 7-5.2.1 of NFPA 20-1978 states in part that:
"Suitable provision shall be made for relieving pressure to the pressure actuated switch, to test the operation of the controller and the-pumping unit.
(See Fig. A-7-5.2.1.)"
The referenced figure, located in the appendix of NFPA 20, indicates a pressure sensing line which is connected to the pump discharge line between the indicat ing control valve and the check valve for that pump. The provisions for reliev ing pressure to test the controller consist of a 1/2" globe valve to drain the pressure sensing line.
Additionally, Section 7-1.1.1 of NFPA 20-1978 states:
"All controllers shall be specifically listed for electric motor driven fire pump service;" and in Section 7-1.1.2 states:
"All controllers shall be completely assembled, wired, and tested by the manufacturer before shipment from the factory."
In their letter dated January 23, 1980, the licensee acknowledged the require ments of Section 7-5.2.1 of NFPA 20, and stated that the test method used at San Onofre Unit 1 complies with that requirement. In so doing, they correctly indicated that the drawing in the appendix of NFPA 20 is only a recommendation and not a requirement.
The present fire pump arrangement at San Onofre Unit 1 consists of two electric motor driven fire pumps, each with an 8" discharge line to a common 8" dis charge header. The separate discharge line from each pump is provided with a check valve and an indicating gate valve. The jockey pump is also connected to the 8" discharge header via a 1-1/2" line. There are no isolation valves in the common discharge header. The fire pump test manifold is connected to the 8" discharge line from the west fire pump, with the connection being between the pump isolation valve and the common discharge header. Backup fire pump supply is provided by the Unit 2 and 3 fire pumps through a separate connection to the underground supply loop.
The test provisions for relieving the pressure to the pressure actuated control switches for each pump consist of opening a 2" gate valve (drawings 5154097-1 2
and 5154143-1 submitted by the licensee indicate that these are 2-1/2" gate valves) on the fire pump test manifold. The present fire pump control system utilizes a 480v switchgear breaker as a controller. The switchgear is not a listed fire pump controller.
Based on a review of the present fire pump and controller design and the licen see's proposed method of relieving pressure to the pressure-actuated control switch with a 2" test manifold valve, we find San Onofre's proposal unaccept-able.
The use of a 2" valve on an 8" pump header to relieve the pressure to the pres sure-actuated control switch as a means of testing the control setpoints of the pumps does not allow for sufficient accuracy in determining the pressure set points and is not in compliance with the intent of NFPA 20. The present ar rangement would also require that both Unit 1 fire pumps and the jockey pump be taken out of service to conduct a test. In addition, the use of the 480v switchgear as a pump controller is not in compliance with NFPA 20.
In order for San Onofre to comply with the requirements of Appendix A and NFPA 20, we recommend that each fire pump be provided with a UL listed fire.pump controller which meets all of the requirements of NFPA 20, specifically Chapter 7 of NFPA 20-1978. We recommend that the licensee indicate their intent to provide listed fire pump controllers for each pump, and provide drawings indic ating the proposed pump/ controller arrangement, including pressure setpoint test arrangements, to NRC for review and acceptance prior to implementation.
Item 3.1.2(b) Fire Water Supply Isolation Valve in Yard Loop Cross Connection The second paragraph under Item 3.1.2 of the San Onofre Unit 1 SER, identified here as 3.1.2(b), indicated the licensee's proposal to provide isolation valves in the aboveground cross connection of the yard loop which is routed through the turbine building to isolate the manual and automatic water suppression sys tems from each other. SER Item 3.1.2(b) states:
"An isolation valve(s) will be provided in the aboveground cross connec tion of the yard loop which is routed through the turbine building to prevent the loss of both manual and automatic water suppression due to a single impairment."
The present design of the cross connection contains no valves to isolate sec tions of the piping containing feed connections to the various automatic and/or manual suppression systems and the standpipe hose stations. A single break in this cross connection would cause the loss of all fire protection water in the turbine building.
With their letter dated January 23, 1980, the licensee provided drawings indi cating the addition of a valve in the cross connection through the turbine building.
Based on a review of the drawing submitted by the licensee, we find San Onofre's proposal to add a single valve to the turbine building cross con nection unacceptable.
The indicated valve does not adequately isolate the automatic suppression systems and standpipe hose stations in the turbine build ing.
3
We recommend that the staff request that the licensee revise their design to provide a sufficient number of valves in the yard loop cross connection through the turbine building so that a single break will not simultaneously impair any standpipe hose stations and an automatic suppression system. We recommend that the staff request revised drawings be submitted to NRC for review and approval prior to implementation.
Item 3.1.5(b) Water Suppression Systems North Turbine Building Area, Cable Water Spray System The second paragraph under Item 3.1.5 of the San Onofre Unit 1 SER, identified here as 3.1.5(b), indicated the licensee's proposal to provide a sectionalized directed water spray system to protect the cable trays in the north turbine building area. Item 3.1.5(b) states:
"A sectionalized directed water spray system will be provided to protect the large concentration of cable trays in the north turbine building area.
Actuation of the system will be provided by the use of line-type tempera ture detectors placed in the cable trays (Area 9A)."
NFPA 15-1977 sets forth certain requirements for water spray systems.
Section 4-4.1.4 of NFPA 15 states:
"Cable Trays and Cable Runs. When insulated wire and cable or nonmetallic tubing is to be protected by an automatic water spray (open nozzle) system maintained for extinguishment of fire which originates within the cable or tube (i.e., the insulation or tubing is subject to ignition and propagation of fire), the system shall be hydraulically designed to impinge water di rectly on each tray or group of cables or tubes at the rate of 0.15 gpm per square foot (6.1 1./min.m 2) on the horizontal or vertical plane contain ing the cable or tubing tray or run.
Automatic detection devices shall be sufficiently sensitive to rapidly de tect smoldering or slow-to-develop flames. When it is contemplated that spills of flammable liquids or molten materials will expose cables, nonme tallic tubing and tray supports, design of protection systems shall be in accordance with that recommended for exposure protection (see 4-4.3.3(d))."
Section 4-4-3.3(d) states:
"Cable Trays and Cable Runs.
When electrical cables or tubing in open trays or runs are to be protected by water spray from fire or spill expo sure, a basic rate of 0.3 gpm per square foot (12.2 1./min.m 2) of pro jected horizontal or vertical plane area containing the cables or tubes shall be provided. Water spray nozzles shall be arranged to supply water at this rate over and under or to the front and rear of cable or tubing runs and to the racks and supports."
Section 4-4.3.3(d) continues to say:
"Where other water spray nozzles are arranged to extinguish, control or cool exposing liquid surfaces, the water spray density may be reduced to 0.15 gpm per square foot (6.1 1./min.m 2) over the upper surface, front or back of the cable or tubing tray or run.
4
Fixed water spray systems designed for protecting cable or tubing and their supports from heat of exposure from flammable or molten liquid spills shall be automatically actuated."
In their letter dated January 23, 1980, the licensee explained that the line type heat detection system located in the cable trays and used to actuate the deluge system has a setpoint of 2800F. They also submitted drawings showing the proposed cable tray water spray deluge system.
The drawings indicate that the cable tray protection is provided by three sepa rate systems. All three systems are fed from a single 6-inch connection to the looped fire protection header in the turbine building.
The area sprinkler sys tem and the wall and column water spray system for this area are also supplied by the same 6-inch connection.
The design density was not indicated for the cable tray protection systems, and details were not provided concerning the ac tuation of the systems.
Based on the available data and the drawings of the systems, we find San Ono fre's proposed cable tray water spray systems unacceptable.
The cable tray water spray systems are not sufficiently separated from other area protection systems to assure that a single break or failure of the fire suppression system will not impair all the water protection for this area.
In order for the licensee to adequately address the concerns of the staff for this area, we recommend that the staff require the three cable tray water spray deluge systems be supplied from a connection to the turbine building fire sys tem h'eader which is independent of the wall and column spray system and the area sprinkler system. Additional isolation valves should be provided on the fire system header so that a single break will not impair more than one of the protection systems (foam, wall and column, area, cable tray) in this area.
In addition, the licensee did not indicate the design density of the proposed cable tray water spray deluge systems nor did they indicate the operation sequ ence of the systems. In accordance with the requirements of NFPA 15, we recom mend that the systems be designed to provide a water spray design density of 0.15 gpm/ sq.ft. over the upper surface or front of the cable or tubing tray or run. To provide for sectionalized operation, we recommend that each of the three cable tray deluge systems operate independently upon actuation of the line-type heat detectors placed in the trays protected by that system. We further recommend that suitable manual operating means should be provided for each system.
We recommend that the staff request that San Onofre submit a commitment to the above design criteria and revised drawings to NRC for review and approval prior to implementation.
Item 3.1.5(d) Water Suppression System North Turbine Building Area, Wet Pipe Sprinkler System The fourth paragraph under Item 3.1.5.of the San Onofre Unit 1 SER, identified here as 3.1.5(d), indicated the licensee's proposal to provide-an automatic area sprinkler system in fire zone 9A. SER Item 3.1.5(d) states:
"A fuse link wet-pipe area sprinkler system will be provided for the large concentration of combustibles in the north turbine area (Area 9A)."
5
In their letter dated January 23, 1980, the licensee stated that an automatic sprinkler system designed to provide a minimum density of 0.3 gpm/sq.ft. for fire area 9A will be provided. Drawings provided with their letter indicated the sprinkler layout above and below the mezzanine. The drawings also indicate that the area sprinkler system is supplied by a connection to the turbine buil ding fire system header which also supplies the cable tray water spray deluge system and the.wall water spray system, and that the connection to the fire protection header is not adequately isolated from the supply connection for the automatic foam system for the lube oil reservoir and conditioner.
Based upon a review of the data submitted by the licensee, we find the proposed area sprinkler system for fire area 9A unacceptable.
In order for the licensee to meet their commitment to provide adequate automatic fire suppression in fire area 9A, we recommend that the licensee revise the design of the wet pipe sprinkler system to indicate a separate supply connection to the fire system header. We recommend that sufficient valves be installed in the fire system looped header to assure that a single break will not impair operation of the wet pipe area sprinkler system and any other automatic suppression systems for area 9A simultaneously.
Further, we recommend that the staff request that re vised drawings be submitted to NRC for review and approval prior to implementa tion.
Item 3.1.6 Foam Suppression System Lube Oil Reservoir and Conditioner Item 3.1.6 of the San Onofre Unit 1 SER indicates the licensee's proposal to provide an automatic foam suppression system in place of the present manual de luge system for the lube oil reservoir and conditioner. SER Item 3.1.6 states:
"The deluge system for the lube oil reservoir and conditioner will be modi fied to provide an automatic foam suppression system."
Design guides for foam-water spray systems are given in NFPA 16-1974.
Section 4022 of NFPA 16 states:
"The design discharge rates for water or air foam solution shall provide densities of not less than 0.16 gallons per minute per square foot of pro tected area."
Section 4023 specifies the discharge duration:
"The foam discharge shall continue for a period of 10 minutes at the design rate specified in 4022.
Where the system has been designed to have deliv ery rate higher than specified in the foregoing, proportionate reduction in the discharge period may be made."
Regarding the total quantity of foam concentrate, Section 2053 states:
"There shall be a readily available supply of air foam concentrate suffi cient to meet the design requirements of the system to put the system back in service after operation. This supply may be in separate tanks or com partments, in drums or cans on the premises, or available from an outside source within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />."
6
In their letter dated January 23, 1980, the licensee stated that the design water foam density for the lube oil area foam system will be 0.16 gpm/sq.ft.
The drawings of the proposed foam system included schematics and scaled draw ings of the system, including connections to the turbine building header and the foam storage tank connections. The licensee did not indicate the foam discharge duration, the type of foam concentrate used, or the type of propor tioning system used. The tank piping arrangement shown on the drawings does not indicate adequate valving arrangements in accordance with NFPA 16 or vari ous manufacturer's recommendations. In addition, a backup foam supply to com ply with Section 2053 of NFPA 16 is not provided.
The system shown on the drawings is provided with two connections to the tur bine building fire protection header. The primary connection to the turbine building fire protection header, which includes the foam proportioning system, is not isolated from the supply connection for the lube oil area sprinkler sys tem, the lube oil area cable tray deluge system, and the lube oil area wall and steel spray system. The second connection to the turbine building fire protec tion header is isolated from the primary connection and the other lube oil area protection systems. The drawings indicate a normally closed deluge valve in this connection. No foam concentrate is provided with this connection, and it feeds directly to the discharge nozzles, bypassing the foam supply.
Based on a review of the drawings submitted by the licensee, we find the licensee's proposed automatic foam system for the lube oil area unacceptable.
In order for the licensee to adequately meet the concerns of the staff, we recommend that the foam system supply connection to the turbine building fire protection header be isolated from the other lube oil reservoir area protection systems by providing approved OS&Y valves in 'the header.
We further recommend that sufficient valves be installed so that a single break in the header will not impair the foam system and any other protection system for the lube oil area simultaneously.
In addition, we recommend that the system comply with the provision of NFPA 16; and the licensee provide adequate information to verify such compliance, inclu ding foam discharge duration, a backup foam supply, and adequate valving in the foam supply and tank connections.
We recommend that an explanation of the operation and purpose of the second supply connection to the fire protection header, and details of the detection system used to actuate the foam system be provided.
We further recommend that sufficient data be provided to demonstrate that the detection system adequately covers the area protected by the foam sys tem.
Finally, we recommend that the staff request that revised drawings and the re quired information be submitted to NRC for review and approval prior to implementation.
3.1.7(a) Gas Suppression System Total Flooding Halon 1301 for 4160 Volt Switchgear Room The first paragraph under Item 3.1.7 of the San Onofre Unit 1 SER, identified here as 3.1.7(a), indicated the licensee's proposal to provide an automatic to tal flooding Halon 1301 suppression system for the 4160v switchgear room.
Item 3.1.7(a) states:
7
"An automatic total flooding Halon 1301 gas suppression system will be pro vided for the 4160 volt switchgear room. The system will be designed to permit a second application of the suppression agent."
NFPA 12A-1977 provides design guidelines for total flooding Halon 1301 systems.
Halon 1301 requirements for fires in solid materials are discussed in Section 2-4:
- 2-4 Halon 1301 Requirements for Fires in Solid Materials.
- 2-4.1 General.
Flammable solids may be classed as those which do not develop deep-seated fires, and those which do. Materials which do not become deep-seated undergo surface combustion only and may be treated much as a flammable liquid fire. Most materials which develop deep seated fires do so after exposure to flaming combustion for a certain length of time which varies with the material.
In others, the fire may begin as deep-seated through internal ignition, such as spontaneous heating.
- 2-4.2 Solid Surface Fires. Almost all flammable solids begin burning on the surface. In many materials, such as unfilled plastics (without filler materials), surface combustion is the only type that occurs.
These fires are readily extinguished with low concentrations, (e.g., 5 percent) of Halon 1301. Although glowing embers may remain at the sur face of the fuel following extinguishment of flames, these embers will be completely extinguished within a short time (e.g., 10 minutes) pro vided the Halon 1301 concentration is maintained around the fuel for this time (called "soaking" time).
- 2-4.3 Deep-Seated Fires.
- 2-4.3.1 Halon 1301, like other halogenated hydrocarbons, chemically in hibits the propagation of flame. However, although the presence of Halon 1301 in the vicinity of a deep-seated fire will extinguish the flame, thereby greatly reducing the rate of burning, the quantity of agent required for complete extinction of all embers is difficult to as sess. It depends on the nature of the fuel, its state of communication, its distribution within the enclosure, the time during which it has been burning, the ratio of the area of the burning surface to the volume of the enclosure, and the degree of ventilation in the enclosure. It is usually difficult or impractical to maintain an adequate concentration for a sufficient time to ensure the complete extinction of a deep-seated fire (see Appendix A-2-4).
- 2-4.3.2 Where the solid material is in such a form that a deep-seated fire can be established before a flame extinguishing concentration has been achieved, provision shall be made to the satisfaction of the au thority having jurisdiction for means to effect complete extinguishment of the fire (see Appendix A-2-4).
Appendix A, Item A-2-4 discusses various design concentrations and soak times and their effectiveness in extinguishing fires. Since we are concerned about deep-seated fires in the cables in the switchgear room, the following excerpts from Section A-2-4 provide guidance:
8
"Deep-seated fires usually require much higher concentrations than 10 per cent and much longer soaking times than 10 min." and "Underwriters Labora tories wood crib fires (lA) and stacks of wood pallets have been readily extinguished with less than 5 percent Halon 1301 maintained for less than 10 minutes following discharge. In these tests, a 10-minute preburn was allowed. Charcoal, the ultimate product of a wood fire, required over 30 minutes for complete extinguishment in a 5 percent Halon 1301 concentra tion.
In charcoal fires, higher agent concentrations were found to reduce the soaking times. At a 10 percent concentration, a 20 minute soaking time was required, and at 20 percent, the soaking time was reduced below 15 min utes.
Another important variable is the-fuel configuration. While wood cribs and pallets are easily extinguished with 5 percent Halon 1301, vertical wood panels closely spaced and parallel require about 25 percent concentrations for 30 to 40 minutes for extinguishment. Fires in boxes of excelsior and in piles of shredded paper also required about 20 percent Halon 1301 for extinguishment. In these situations, heat tends to be retained in the fuel array, rather than being dissipated to the surroundings. Radiation is an important mechanism for heat removal from smoldering fires."
Discharge time requirements are set forth in Section 2-6.2.2:
"The agent discharge shall be substantially completed in a nominal 10 sec onds or a shorter time if practicable, unless a longer discharge time is specifically permitted by the authority having jurisdiction. This period shall be measured as the interval between the first appearance of liquid at the nozzle and the time when the discharge becomes predominantly gas eous. This point is distinguished by a marked change in both the sound and the appearance of the discharge."
In their letter dated January 23, 1980, the licensee indicated that the Halon 1301 system for the 4160v switchgear room will be designed to provide a 6 per cent concentration for a soak time of 10 minutes, and that the discharge time will be 10 seconds.
They stated that the use of a 6 percent concentration is based on the fact that the fire hazard is cable insulation, and there is little possibility of a deep-seated fire occurring.
The schematic drawing of the Halon 1301 system submitted with the licensee's letter indicated 5 primary and 5 reserve cylinders of Halon 1301, thus provid ing for a second application of the suppression agent.
Based on a review of the information supplied by the licensee, we find that the licensee's proposed total flooding Halon 1301 suppression system for the 4160v switchgear room is unacceptable. The licensee, in setting their design concentration and soak time, has not acknowledged the possibility of deep seated fires occurring in the cable. Paragraph 2-1.1.3 of NFPA 12A, which the licensee references as restricting the maximum Halon concentration to 7 percent for evaluation times in excess of 1 minute, refers only to "normally occupied areas."
Section 2-1.1.3 states:
"Halon 1301 total flooding systems shall not be used in concentrations greater than 10 percent in normally occupied areas. For the purposes of 9
this standard, a "normally occupied" area is defined as an area intended for occupancy. Areas which may contain 10 percent Halon 1301 shall be evacuated immediately upon discharge of the agent. Where egress cannot be accomplished within one minute, Halon 1301 total flooding systems shall not be used in normally occupied areas in concentrations greater than 7 per cent."
Based on this definition, the 4160v switchgear room is not a "normally occupied area," and as such is not restricted by Section 2-1.1.3 of NFPA 12A. The re strictive provisions of Section 2-1.1.4 would apply:
- 2-1.1.4 Halon 1301 total flooding systems utilizing concentrations greater than 10 percent but not exceeding 15 percent may be used in areas not normally occupied, provided egress can be accomplished within 30 seconds. Where egress cannot be accomplished within 30 seconds or concentrations greater.than 15 percent must be used, provisions shall be made to prevent inhalation by personnel."
For the licensee to meet their commitment to provide an effective total flood ing Halon 1301 system for the 4160v switchgear room, we recommend that the licensee revise the design criteria of the proposed system to provide for a 7 percent concentration for a 20 minute soak time. We recommend that the dis charge time of ten seconds and capability for a second application remain unchanged, except that additional quantities of Halon 1301 will be required for both the primary and reserve systems to provide for the increased density and soak time. We recommend that either adequate calculations be provided or in-situ tests be conducted to verify that the required concentration can be maintained for the required soak time.
The second paragraph under Item 3.1.7 of the San Onofre Unit 1 SER, identified here as 3.1.7(b), indicated the licensee's proposal to provide an automatic to tal flooding Halon 1301 suppression system for the 480v switchgear room. Item 3.1.7(b) states:
"An automatic total flooding Halon 1301 gas suppression system will be pro vided for the 480 volt switchgear room."
NFPA 12A-1977 provides design guidelines for total flooding Halon 1301 systems.
Halon 1301 requirements for fires in solid materials are discussed in Section 2-4:
- 2-4 Halon 1301 Requirements for Fires in Solid Materials.
- 2-4.1 General.
Flammable solids may be classed as those which do not develop deep-seated fires, and those which do. Materials which do not become deep-seated undergo surface combustion only and may be treated much as a flammable liquid fire. Most materials which develop deep seated fires do so after exposure to flaming combustion for a certain length of time which varies with the material.
In others, the fire may begin as deep-seated through internal ignition, such as spontaneous heating.
10
- 2-4.2 Solid Surface Fires. Almost all flammable solids begin burning on the surface. In many materials, such as unfilled plastics (without filler materials), surface combustion is the only type that occurs.
These fires are readily extinguished with low concentrations, (e.g., 5 percent) of Halon 1301. Although glowing embers may remain at the sur face of the fuel following extinguishment of flames, these embers will be completely extinguished within a short time (e.g., 10 minutes) pro vided the Halon 1301 concentration is maintained around the fuel for this time (called "soaking" time).
- 2-4.3 Deep-Seated Fires.
- 2-4.3.1 Halon 1301, like other halogenated hydrocarbons, chemically inhibits the propagation of flame. However, although the presence of Halon 1301 in the vicinity of a deep-seated fire will extinguish the flame, thereby greatly reducing the rate of burning, the quantity of agent required for complete extinction of all embers is difficult to assess. It depends on the nature of the fuel, its state of communica tion, its distribution within the enclosure, the time during which it has been burning, the ratio of the area of the burning surface to the volume of the enclosure, and the degree of ventilation in the enclosure.
It is usually difficult or impractical to maintain an adequate concen tration for a sufficient time to ensure the complete extinction of a deep-seated fire (see.Appendix A-2-4).
- 2-4.3.2 Where the solid material is in such a form that a deep-seated fire can be established before a flame extinguishing concentration has been achieved, provision shall be made to the satisfaction of the au thority having jurisdiction for means to effect complete extinguishment of the fire (see Appendix A-2-4).
Appendix A, Item A-2-4 discusses various design concentrations and soak times and their effectiveness in extinguishing fires. Since we are concerned about deep-seated fires in the cables in the switchgear room, the following excerpts from Section A-2-4 provide guidance.
"Deep-seated fires usually require much higher concentrations than 10 per cent and much longer soaking times than 10 min." and "Underwriters Labora tories wood crib fires (lA) and stacks of wood pallets have been readily extinguished with less than 5 percent Halon 1301 maintained for less than 10 minutes following discharge. In these tests, a 10-minute preburn was allowed. Charcoal, the ultimate product of a wood fire, required over 30 minutes for complete extinguishment in a 5 percent Halon 1301 concentra tion. In charcoal fires, higher agent concentrations were found to reduce the soaking times.
At a 10 percent concentration, a 20 minute soaking time was required, and at 20 percent, the soaking time was reduced below 15 min utes.
Another important variable is the fuel configuration. While wood cribs and pallets are easily extinguished with 5 percent Halon 1301, vertical wood panels closely spaced and parallel require about 25 percent concentrations for 30 to 40 minutes for extinguishment. Fires in boxes of excelsior and 11
in piles of shredded paper also required about 20 percent Halon 1301 for extinguishment. In these situations, heat tends to be retained in the fuel array, rather than being dissipated to the surroundings. Radiation is an important mechanism for heat removal from smoldering fires."
Discharge time requirements are set forth in Section 2-6.2.2:
"The agent discharge shall be substantially completed in a nominal 10 sec onds or a shorter time if practicable, unless a longer discharge time is specifically permitted by the authority having jurisdiction. This period shall be measured as the interval between the first appearance of liquid at the nozzle and the time when the discharge becomes predominantly gaseous.
This point is distinguished by a marked change in both the sound and the appearance of the discharge."
In their letter dated January 23, 1980, the licensee indicated that the Halon 1301 system for the 480v switchgear room will be designed to provide a 6 percent concentration for a soak time of 10 minutes, and that the discharge time will be 10 seconds. They stated that the use of a 6 percent concentration is based on the fact that the fire hazard is cable insulation, and there is little possibility of a deep-seated fire occurring.
A drawing of the Halon 1301 system for the 480v switchgear room was not submit ted with the licensee's letter.
Based on a review of the information supplied by the licensee, we find that the licensee's proposed total flooding Halon 1301 suppression system for the 480v switchgear room is unacceptable. The licensee, in setting their design concen tration and soak time, has not acknowledged the possibility of deep-seated fires occurring in the cable. Paragraph 2-1.1.3 of NFPA 12A, which the licen see references as restricting the maximum Halon concentration to 7 percent for evacuation times in excess of 1 minute, refers only to "normally occupied areas."
Section 2-1.1.3 states:
"Halon 1301 total flooding systems shall not be used in concentrations greater than 10 percent in normally occupied areas. For the purposes of this standard, a "normally occupied" area is defined as an area intended for occupancy. Areas which may contain 10 percent Halon 1301 shall be evacuated immediately upon discharge of the agent. Where egress cannot be accomplished within one minute, Halon 1301 total flooding systems shall not be used in normally occupied areas in concentrations greater than 7 per cent."
Appendix A of NFPA 12A further defines a "normally occupied area" in Section A-2-1.1.3:
"For the purposes of this standard, a normally occupied area is defined as an area which is intended for occupancy. Spaces which are occasionally visited by personnel, such as transformer bays, switch-houses, pump rooms, vaults, engine test stands, records centers, magnetic tape storage areas, cable trays and tunnels, microwave relay stations, flammable liquid storage areas, enclosed energy systems, etc., are examples of areas which are con sidered to be not normally occupied.
12
Based on this definition, the 480v switchgear room is not a "normally occupied area," and as such is not restricted by Section 2-1.1.3 of NFPA 12A. The re strictive provisions of Section 2-1.1.4 would apply:
- 2-1.1.4 Halon 1301 total flooding systems utilizing concentrations greater than 10 percent but not exceeding 15 percent may be used in areas not normally occupied, provided egress can be accomplished within 30 seconds.
Where egress cannot be accomplished within 30 seconds or concentrations greater than 15 percent must be used, provisions shall be made to prevent inhalation by personnel."
For the licensee to meet their commitment to provide an effective total flood ing Halon 1301 system for the 480v switchgear room, we recommend that the licensee, revise the design criteria of the proposed system to provide for a 7 percent concentration for a 20 minute soak time. We recommend that the dis charge time of ten seconds remain unchanged, except that additional quantities of Halon 1301 will be required for both the primary and reserve systems to pro vide for the increased density and soak time. We further recommend that either adequate calculations be provided or in-situ tests be conducted to verify that the required concentration can be maintained for the required soak time.
The eighth paragraph under Item 3.1.15 of the San Onofre Unit 1 SER, identified here as 3.1.15(h), indicated the licensee's proposal to either replace the oil in the station service transformers with a silicone base oil or to provide an automatic deluge system to protect transformers 2 and 3. SER Item 3.1.15(h) states:
"The oil in the station service transformer will be replaced with a sili cone base oil or an automatic deluge system will be provided to protect transformers 2 and 3."
We feel that the station service transformers present an unacceptable exposure fire hazard to safety-related cables and systems. An unmitigated fire at the transformers could result in damage to electrical cables and the fire water pumps in the area. Present protection consists only of yard hydrants and hose stations, portable fire extinguishers, and one wheeled unit fire extinguisher.
In their letter dated January 23, 1980, the licensee indicated that an auto matic deluge system will be provided to protect the station service trans formers 2 and 3. They stated that the system will comply with NFPA 15, and will be designed to provide a water density of 0.30 gpm/sq.ft.
A schematic drawing of the proposed system was also provided by the licensee with their letter. The diagram indicates a single row of spray nozzles on each side of the pair of transformers, but does not indicate the detection and actu ation means for the system.
Section 4-4.3.4(a) of NFPA 15-1977 states:
"Transformer protection shall contemplate essentially complete impingement on all exterior surfaces, except underneath surfaces which-in lieu thereof may be protected by horizontal projection. The water shall be applied at a rate not less than 0.25 gpm per square foot (10.2 1./min.m 2) of projected 13
area of rectangular prism envelope for the transformer and its appurte nances and not less than 0.15 gpm per square foot (6.1 1./min.m2 ) on the expected nonabsorbing ground surface area of exposure. Additional applica tion is needed for special configurations, conservator tanks, pumps, etc.
Spaces greater than twelve inches (305 mm) in width between radiators, etc., shall be.individually protected."
The licensee has indicated they will provide a design water density of 0.30 gpm/ sq.ft., but have not indicated the area of application for this density.
The licensee has not indicated compliance with the requirement that complete water impingement on all exterior surfaces be provided, and the piping and noz zle arrangement shown on the schematic drawing does not appear to provide suf ficient nozzles to meet the water spray requirement.
Based on a review of the data provided by the licensee, we find the licensee's proposed water spray deluge system for the station service transformers 2 and 3 unacceptable. In order for the licensee to meet their commitment to provide a water spray deluge system for the station service transformers 2 and 3, we re cmmend that drawings be submitted to demonstrate that sufficient nozzles are provided to deliver the 0.3 gpm/sq.ft. water spray density over all surfaces of the transformers as required in Section 4-4.3.4(a) of NFPA 15, including 0.15 gpm/sq.ft. on all nonabsorbing ground surface area of exposure. In addition, we recommend that details (including drawings) be provided to indicate the de tection and actuation means for the system.
Item 3.2.3 Turbine Building Structure Item 3.2.3 of the San Onofre Unit 1 SER requires the licensee to evaluate the effect of a fire involving the hydrogen seal oil system on the structural in tegrity of the turbine building if there was no fire suppression system.
Item
3.2.3 states
"The licensee is evaluating the effect of a fire involving the hydrogen seal oil system on the structural steel of the turbine building without the benefit of a fire suppression system."
The licensee was also required to provide a report to NRC on the methods used to perform the analysis.
The hydrogen seal oil unit is located under the turbine deck in the south end of the turbine building. Exposed steel structural members in the area would be directly exposed by a fire involving the hydrogen seal oil from a leak either at the seal oil unit or at a point along the piping. An unmitigated fire could cause structural damage and possible collapse, with resultant damage to safety related circuits and equipment in or adjacent to the.turbine building.
In their letter dated December 26, 1979, the licensee presented a summary of their evaluation of the effects of fires involving the hydrogen seal oil system on the structural steel of the turbine building.
Included in their letter was a list of assumptions used for th e analysis, a brief description of the methods used in the analysis, and their conclusions.
The calculations and details of the heat transfer model the licensee used were not included with the letter.
The licensee concluded that a fire involving 12,300 gallons of seal oil burning 14
for a duration in excess of 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> would create peak temperatures in the structural steel of 604 0 F, and thus would not affect the structural integrity of the steel.
We feel that the licensee's conclusions that such a fire would not affect the structural steel are not acceptable. Several of the assumptions used are incorrect or are arbitrary in nature, and changing those assumptions could drastically affect the results of the analysis. A few examples are listed be low.
In assumption #4, the licensee indicated that they considered only radiative heat transfer and neglected convective heat transfer. Convective heat transfer would have a significant effect on the room temperature and thus on the steel columns.
In assumption #9, air flow is assumed to be 2 ft/sec from the column toward the fire. There is no basis for such an assumption. While the air flow near the floor may well be in the direction assumed, the same reverse flow of air must necessarily be assumed at some other point to compensate, since the total vol ume in the fire area is not changing. This, in conjunction with convective heat transfer, will tend to heat the upper portions of the steel columns at a much faster rate than the lower sections.
In assumption #10, a constant air temperature at each column is assumed.
Such an assumption neglects the general heating of the surrounding air from the fire.
In assumption #14, the licensee states that conduction of heat away from the column is to the slab only.
No attempt is made to limit that conduction due to the direct heating of the slab by the fire.
In general, it appears as though the analysis performed by the licensee consi ders a point source of heat and its effect on a single point along a steel beam.
The only heat input to the beam is from radiation from the point source of heat, and credit for heat loss is considered to the air and other surround ing surfaces. The basic facts that the fire will be three dimensional, and that all other surfaces (the slab) and mediums (air) will be heated simultane ously have been disregarded. This heating of all elements in the area will affect all surfaces at varying rates, and to consider a simple heat transfer solution similar to that presented as a definitive answer to the concerns ex pressed by the staff is not acceptable.
In order for the licensee to adequately address the concerns of the staff re garding the large fire hazard presented by the hydrogen seal oil system, we recommend that an automatic sprinkler system be installed throughout fire zone 90 to provide total area coverage. We recommend that the system be designed to provide a minimum density of 0.3 gpm/ft2 for the area, and the water supply connection be capable of being isolated from the supply for the automatic sup pression system which protects the hydrogen seal oil unit.
We further recom mend that the staff request plans, including design criteria, be submitted to NRC for review and approval prior to implementation.
15