ML20076L762
| ML20076L762 | |
| Person / Time | |
|---|---|
| Site: | Monticello  |
| Issue date: | 06/16/1983 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20076L760 | List: |
| References | |
| NUDOCS 8307190370 | |
| Download: ML20076L762 (19) | |
Text
a nam k
UNITED STATES
[t
'i NUCLEAR REGULATORY COMMISSION 3 r.<
]j WASH tNGTON, D. C. 20555
%.ndj
- SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION ON EXEMPTION REQUEST FROM 10 CFR PART 50, APPENDIX R FIRE PROTECTION MONTICELLO NUCLEAR GENERATING PLANT DOCKET NO. 50-263 1.0 Introduction By letter dated June 30, 1982 Northern States Power Company (licensee) requested several exemptions from Section III.G of Appendix R.
By letter dated October,28,1982 the licensee provided additional information.
The licensee requested exemptions from the requirements of Section III.G of Appendix R for the:
1.
Suppression pool Area, (Fire Zone 1F);
2.
Intake Structure Pump Room (23A);
3.
Structural Steel in Six Areas:
a)
Load Center No.1 (12A);
b)
Lube Oil Resevoir and Feedwater Pump Area (138) c)
ESF Motor Control Center Room (13C) d) RCIC Room (lC) e) Division II. Battery Room (7C) f)
Cable Spreading Room (8); and 4.
Control Room (9).
On January 12, 1983 we forwarded a draft copy of this Safety Evaluation and requested that the licensee review it for accuracy of technical content.
In a February 14, 1983 letter, the libensee agreed with the staff's con-70-clusions in denying the exemption requests for the Intake Structure l On Pump Room and three of the six areas for structural steel (especifically lO
' y items 3a, 3b, and 3c above).
For the remaining three areas for structural 00 l M steel (RCIC Room, Battery Room, cable Spreading Room), and the Control E
Room, the licensee requested a meeting with the staff.
Following the meeting and in a letter dated March 22, 1983, the licensee decided to:
l
l To 1.
withdraw the exemption request for structural steel in the RCIC Room; 2.
conforw to the requirements of the Rule-for structural steel in the Battery Room and; 3
provide alternative shutdown capahitity for the Control Room and Cable Spreading Room.
Sectfore III.G.I requires that one traint of cahTes and equipment necessary to achieve and maintain; safe shutdown be maintained. free of fire damage by one of the following means-a.
Separatfort of cahTes and equipment and associated non-safety circuits of, reduadant trains by a fire barrier having a 3-hour ratint.[
Structurat steel fbruing a part of or supporting such
~
fire barriers shalT be protected.ta provide ffre resistance equivaTent to that requfred of the barrfer, h.
Separation, of cables and equipment and-associated non-safety circuits of redundant trafns by"a horizontal distance of more
~
than 20 feet witir no intervening combustibTes or ffre hazards.
In additfort,, fire detectors and art automatic fire suppression system. shaTT be instaTied fit the fire area; or
' c.
Enclosure of cables and equipment and associated non-safety circuits of one redundant train frr a fire barrier having a I-hour rating.
En additfort,. fire detectors and an automatic ffre suporessiorr systes shaTT be instaITed. fir the fire area.
M
- Se
.e
l l
2 If these conditions are not met,Section III.G.3 requires alternative shutdowrr capability independent of the fire area of concern.
It also requires a. fixed. suppressfort systee irr the fire area of concern if it contains a Targe concentratiott of cables or other combustibles.
(
I These alternative requirements'are not deemed to be equivalent for all l
I configurations,, however,. they provide equivalent protactiort for those l
configurations fra whicht they are accepted Because it fs not possible to predict the speciffe conditions under whictr fires may occur and propagate,, the desfgn basis protective features are speciffed frr the ruTe rather than the design basis fire.
Plant specific features may require protectiort different thart the measures spectffed fru Section III.E De sucts a clase,. the Ticensee must demonstrate,. by means. of a datafTed fire hazards anaTysis,. that existing; protectforr or existing: pihi iore in: cortfunctfort witir proposed modfffcattens wfTJ prowfde a Tevet of safety equivalent ts the tactr.-
nical requirements of Section III.G of Appendfx IF.
Irr summary, Sectiort III.E is related ta fire protectiott features for ensuring; that systems and associated cfrenits used to achieve and main-tafts safe shutdower are free of fire damage. Fire protectfort configura-tions must.either meet the specific requirements of Sectiort III.G or alternative fire protection configurations must be justified by a fire hazard' analysis.
Our general criteria for accepting aTternative fire protection con-figurations are the following:-
1 The alternative assures that one trairt of equipment necessary to achieve hot shutdown froe either the control room or emergency controT stations is-free of fire damage.
1 The aTternative assures that fire damage ta at least one train of l
equipment necessary to achieve cold shutdowrr is Timited such that it cart be repaired within a reasonable time (minor repairs with l
components stored on-site).
t l
- Modifications required ta meet Sectf' ort III.G would not enhance fire protectiott safety above that provided by either existing or proposed aTternatfver Modiffcations required to meet Sectfort IIEG would be detrimentai te overeTI facfTity safety 2.0; AnaTytical Method The licensee employed art analyticaLT method ta demonstrate the inherent protectford afforded to existing; safe shutdown systems frv the contraT room The. intent of this, method was to provider common. parameters by whictr the existirig TeveT> of fire protection for the control room couTd be iudged, far order ta demonstrate that verbatie compliance with Sectforr III.E of Appendix R would not' enhance the fire protectiott for safe shutdowrr.
l l
The method can be summarized.as fo1 Tows:
- The redundant components of concern are identified.-
- Their geometry and configuratfort within the fire area are described
- The faiTure criteria is specified.
The analysis determines resultant temperatures from a constant heat flur produced by either internal cabinet fires or exposure fires to redundant cabinets in the control console The heat flux ts' generated for a limited. period of 120 seconds..
9 6
i
,m---
l l
4 l
l s
We and our consultant, Brookhaven National Laboratory have reviewed l
the analytical method. A copy of the consultant's report is attached.
l We have determined that the results of the methodology, as appifed, do not demonstrate the equivalence of the protectiort provided har safe l
l shutdown to the specific alternative set forth in Section III.G. of Appendix R.
For example:
The method does not consider the heat released to the room by secondary fires involving, in-situ. combustibles.
The method; does not consider fires of greater than 120 seconds duration.
The method; does not consider alt of the afternatives set forth fr Section III.E; f.e.
3-hour ffre barefer, I-hour fire barrier witte suppressfor systes,. twenty-feet separattom free of combus.-
tibles with autn== tic suppression and; alternate or dedicated l
shutdown capahfTfty independent of the area.
l l
The Ticensee has not: used the results of this analysis to compare tha protection provided witit that specfffed in Section III.G. The Ticensee has only stated that the accumutation of the calculated quantity of fiammahTe Tiquids in the required configuration is an unrealistic condition,. and wiTT be prevented by administrative contrais.. We do not does this to be a valid argument because there is ne positive means of preventing the accumulation of transient materials in individual piant areas.
As documented in Inspection and Enforcement Branctr Reports; recent inspections:at plants such as Davis Besse (50-346/82-03,, April 1,1982), Duane Arnold (50-331/81-25, January 11,1982), D'.C'. Cook (50-315/82-11, December 31, 1981), and Nina Mile Point (50-22d/82-09), have demonstrated that substantial quantities of hazardous substances such as iis galTon drums of wasta oil are located in even highly restricted and controlled entry areas.
-,,,. ~
, ~,
5 We hava not relied upon the results of the licensee's analysis in our evaluation.
We have evaluated each. exemption request using our standard method of review:
a)
Review the information submitted and that existing in the docket file to determine the configuratio5 of t'ha redundant components, b)
Evaluate the existing fire protection, proposed modifications, and other compensating features or mitigating factors to determine the overall level of fire prote' tion in the area of c
concern, and 4
c)
Determine if the overall level of safety is equivalent to that provided by Section III.G of Appendix R. -
~
3.0 Structural SteeT 3.1, Exemptrton Requested The licensea requests an exemption from Section III.G.2 to the extent l
it requires structural steel forming a part of or supporting a 3-hour fire rated barrier shall be~ protected to an equivalent rating.
- 3. 2 Discussion The structural steel supporting 3-hour fire rated floor / ceiling assemblies in the following areas are unprotected:
a.
Fire Zone 1C' RCIC Room b.
Fire Zone 7C Division II Battery Room c.
Fife Zone 8 Cable Spreading Room d.
Fire Zone 12A Load Center No. 1 e.
Fire Zone 138 Lube 0il Reservoir and Reactor Feedwater Pump Area f.
Fire Zone 13C ESF Motor Control Center Room
-_-,e m
e,
-g-g.,
m.-
w
6 With the exception of Fire Zone 1C the unprotected structural steel usually consists of I beams installed below the 3-hour fire rated concrete floor / ceiling assemblies they are supporting.
In Fire Zone 1C, the I beams are embedded in the concrete floor / ceiling assemblies they are supporting. Only the bottom web of the I beams could be exposed to a fire.
Fire Zone IC RCIC Room After deterinining that the steel was used for forming the concrete floor and is not considered in the building's structural analysis, the licensee has withdrawn the exemption request for structural steel in the RCIC room.
Fire Zone 7C Division II Battery Room Fire Zone 7C is located in the Control Building at elevation 928 feet.
The area contains the Division II 125-volt and 24-volt batteries and associated equipment.
The ceiling height is 10 feet 2 inches.
The in-situ combustibles comprise an equivalent fire severity of 45 minutes on the ASTM E-119 standard time temperature. curve.
Fire protection consists of an automatic Halon 1301 fire sup;:ression system, smoke detectors, manual hose stations and portable fire extinguishers.
Fire Zone 8 Cable Spreading Room Fire Zone 8 cable spreading room is located in the control building at elevation 928 directly above Fire Zone 7C.
The ceiling height is 10 feet 2 inches.
The in-situ combustibles comprise an equivalent m.
A
7 fire severity of 58 minutes on the ASTM E-119 standard time temperature curve.
The fire protection in the area, consists of an automatic Halon 1301 fire extinguishing system, smoke detectors, manual hose stations and portable fire extinguishers.
Fire Zone 12A Load Center No.1 Fire Zone 12A is located in the turbine building at elevation 911 feet.
The area contains 4160-volt switchgear, 480-volt load centers and askeral filled transformers.
The ceiling height is 19 feet 4 inches.
The in-situ fuel combustibles comprise an equivalent fire severity of 38 minutes on the ASTM E-119 standard time temperature curve.
Fire protection in the area consists of smoke detectors, manual hose stations and portable fire extinguish'ers.
Fire Zone 138 Lube OiT Reservoir and Reactor Feedwater Pumo Area
(
i Fire Zone 138 is located in the Turbine Building at elevation 911 feet.
The area contains the turbine lubricating oil reservoir, centrifuge and transformer pumps, service and instrument air compressors and receivers and-the reactor feedwater pumps.
The ceiling height is 19 i
feet 4 inches.
The in-situ combustibles comprise an equivalent fire severity of approximately 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> on the ASTM E-119 standard time temperature curve.
Fire protection in the area consists of an automatic deluge sprinkler system to protect the lube oil reservoir, smoke detectors, manual hose stations and portable fire extinguishers.
l Zone 13C ESF Motor Control Center Room Fire Zo'ne 13C is located in the Turbine Building at elevation 911 feet.
The area contains the ESF MCCs and associate cabling.
The ceiling height is 19 feet 4 inches.
The in-situ combustibles comprise an equivalent fire severity of 26 minutes on the ASTM E-119 standard time temperature curve.
Fire protection in the area consists of smoke detectors,. manual hose stations and portable fire extinguishers.
...m
._,,_...___m,y
8 3.3 Evaluation t
Section III.G.2 of Appendix R requires that the structural steel i
which supports or is apart of barriers separating redundant divisions to have a rating equivalent to the fire resistance of the barrier.
The protection of the structural steel is required because steel loses streneth when subjected to temperatures that may be attained in a fire.
A temperature of 1100 F is normally considered to be the critical temperature. At this temperature the yield stress in steel has decreased to about 60 percent of the'value' at room' temperature.
This is approximately the level normally used as the design working stress.
Steel has a high thermal conductivity, therefore heat is transferred away from a localized heat sourc~a rather quickly, thereby requiring a relatively long period of time to rea'ch the critical temperature. However, an exposure fire that distributes heat over a greater area may reduca this tima considerably.
Because it is not possible to predict the specific condition under which fire may occur add propagate, structural steel forming a part of a supporting, 3-hour fire rated barriers needs to be protected to provide a fire resistance equivalent to that required of the barrier.
The combustible loading, the configuration of the areas of concern and the potential for the accumulation of transient combustible materials are what is typically found in nuclear power plants.
There are no other fire protection features in these areas to compensate for the omission of the protection of the structural steel.
Fire Zones 7C and 8 are equipped with an automatic Halon 1301 fire j
suppression system.
In the event of an exposure fire involving transient or in-situ combustible materials, there will be a time l
lag between the ignition of the fire, detection and alarm, and the i
i
9 fire brigade response.
The existing configuration of structural steel in each area.provides no protection against the thermal flux of an exposure fire.
We therefore do not have reasonable assurance i
that the critical temperature of the structural steel will not be reached in this interval.
The existing protection of the structural steel in the above fire zones does not provide a level of fire protection equivalent to Section i
III.G.
Modifications such as applying a sprayed on fire proofing to
~
the stuctural steel to obtain a 3-hour fire rating would provide the requisite levels of safety.
3.4 Conclusion Based on the above evaluation, the existing protection for the structural steel in Fire Zone 7C Division II Battery Room, Fire Zone 8 Cable Spreading Room,. Fire Zone 12A Load Center No,1, Fire Zone 13B Lube Oil Reservoir and Reactor Feedwater Pump Area and Fire Zone 13 C ESF Motor. Control Center Room is not equivalent to the protection required by the technical requirements of Section III.G of Appendix R.
Therefore, the licensee's requests for exemption in the above areas should be denied.
Fire Zone 1C RCIC Room was not included in this finding, since the licensee withdrew the exemption request for this area.
4.0 Suppression Pool 4.1 Exemption Requested l
l The licensee requests an exemption from Section III.G.2 to the extent that it requires the installation of an automatic fire suppression system.
10 4.2 Discussion The area is located at 896 feet elevation of the reactor building.
The area is separated from other plant areas by 3-hour fire rated barriers.
The ceiling height is 35 feet.
Fire protection in the area consists of smoke detectors, manual hose stations and portable fire extinguishers.
The only redundant safe shutdown equipment in the area consists of in-strumentation for measuring the water temperature and level in the torus.
The redundant trains are separated by 100 feet free of intervening combustibles.
I In-situ combustibles are essentially non-existent.
All surfaces are concrete except the torus, which is steel.
All cables in the area are installed in conduit.
4.3 Evaluation This area does not have an automatic suppression system and there is.no alternate shutdown capability independent of the area.
The l
licensee justifies this alternative on the following mitigating
\\
features.
a)
All cables are routed in conduit i
b)
The in-situ combustible loading is essentially non-existent c)
Smoke detection is provided d)
The area has limited areas during normal operation
)
e)
The redundant trains are horizontally separated by 100 feet free of intervening combustibles.
l 9
__e,,.
..---.7-
..--,_v,-_
11 Because of the restricted access to this area, the probability of an exposure fire from the accumulation of transient combustibles during normal operation is low.
This combined with the 100 feet of separation between redundant trains and early warning fire detection provides reasonable assurance that one train will be maintained free of fire damage, and therefore, is acceptable.
- 4. 4-Conclusion Based on our evaluation, we conclude that the level of safety provide in the Suppression Pool Area is equivalent to the technical require-ments of Section III.G of Appendix R and therefore, the licensee's request should be granted.
5.0 Intake Structure 5.1.
Exemption Requested The Ticensee requests an exemption froar Section III.G.2 to the extent that it requires an automatic fire suppression system and no intervening combustibles between redundant trains.
5.Z Discussion The Intake Structure is separated from adjacent areas by 3-hour fire rated barriers.
The ceiling height in areas containing safe shutdown systems is 13 feet.
Safe shutdown systems in the area include 2 Residual Heat Removal (RHR) pumps, 2 Emergency Service Water Pumps and associated equipment.
One RHR and one Emergency Service Water Pump are required for safe shutdown.
The RHR and Emergency Service Water Pump trains are separated horizontally by 28 feet traversed by open horizontal cable
~
trays..
a The combustible loading in the area consists of PVC cables in open horizontal cable trays, and approximately 170 gallons of lubricating
? oil contained in the 9 oump reservoirs.
A smoke detection system is installed in the area.. Portable fire extinguishers and manual fire hose stations are also available.
The licensee justifies this alternative on the basis that (1) manual fire suppression capability, and (2) smoke detection are provided.
I 5.3 Evaluation This area does not comply with Section III.G because it does not have an automatid suppression system and twenty feet oY separatior. free of l
intervening combustibles or one hour fire rated barriers.
There is no alternative shutdown capability independent of this area.
There are generally two mechanisms by which fire damage can' occur; either an exposure fire in close proximity to the redundant equipment or an exposure fire at any point in the roos of sufficient magnitude to form a stratified layer of hot gases at the ceiling, which descends to the floor level at a rate correlated to the room volume, the burning time and fuel quantity.
In the case of a fire which produces a stratified layer of hot gases at the ceiling level, the most severe damage will occur to cables and equipment located within several feet of the ceiling.
The redundant emergency service water and RHR pump cables are installed approximately 12 feet above the floor level, and the ceiling height of the room is 13 feet.
The configur-ation does not provide reasonable protection from a descending hot gas layer.
A local exposure fire could also cause damage to the redundant cables and pumps if they are exposed to a heat flux of sufficient intensity.
This exposure is independent of room volume.
a.--,~.-n.------v.
,----,,,,,,~n,,,--,,-,w,-n
,-,.-~,--,-----m.
een..
m, e~
13 No additional protective features or mitigating features are provided to compensate for the lack of automaticasuppression specified,by Section III.G.
The combustible loading, the configuration of the room, the type r
of cable insulation, and the potentia 1 'for the accumulation of transient combustible materials are what is typically found in intake structures. There are no fire protection features in this area to compensate for the omission of an automatic suppression system.
In the event of an exposure fire involving transient combustible materials, there will be a time lag between the ignition of the fire,
~
detection and alarm, and the fire brigade response.
The existing configuration of cables, without an automatic suppression system provides no protection against the thermal flux of an exposure fire.
We,.theref4re, da not have reasonable assurance that redundant cables of both trains will not be damaged in this interval.
5.4 Conclusion The level of existing protection in this area does not provide a level of fire protection equivalent to the technical requirements of Section III.G of Appendix R, therefore the exemption should be denied.
6.0 Control Room 6.1 Exemption Requested The licensee requests an exemption from Section III.G.2 to the extent that it' requires 20 feet of separation without intervening combustibles between redundant trains and the installation of an automatic fire suppression system.
1 S
__w,._-.
14 6'. 2 Discussion The control room is separated from all other areas nf the plant by 3-hour fire rated barriers.
Fire protection is provided by foniz-ation smoke detectors with manual fire suppression provided by standpipehosestationsandpoitablefireextinguishers.
The combustible loading in this area consists of wood, paper and plastic.
Cables and components of all redundant safe shutdown trains are located in the control room.
Redundant divisions are in separate cabinets, but are separated by less than 20 feet free of intervening combustibles.
This fire area is continuously manned by operating personnel, trained in fire-fighting.
The licensee justified the exemption based on the' following considerations:
- 1., The centroT room is continuously manned by licensed-operators.
If a. fire did occur, it. would be discovered and extinguished promptly by the operators using portable extinguishers; 2.
The control room is a restricted area.
This restriction on access to the control room, coupled with administrative controls, would result in no significant quantities of flammable liquids being present in the control room, thus l
limiting the fire hazard.
3.
The results of an analysis featuring a fire model were presented to demonstrate that a fire involving a 2-foot by 2-foot pan of flammable liquid for a 2-minute duration external to the control console and an internal panel l
fire of 2-minute duration would not affect the ability to l
achieve safe. shutdown from the control board.
I
15 6.3 Evaluation The co...rol room is net in compliance.with Section III.G because of the absence of a complete area wide fixed fire suppression system, the lack of adequata physical separation between redundant shutdown divisions, and the lack of an alternat[ shutdown capability indepen-dent of the control room.
The control room contains the majority of the controls essential for station operation and for shutdown of the plant under all operating conditions.
Redundant systems necessary for safe shut-down are located in close proximity within the control console and, without adequate protection, would.be damaged by a singla fire of significant. magnitude. With the present design, Tf such a fire occurred, there is no capability tp achieve safa shutdown independent of the control room.
Adaknistrative co~ntrols, even if they are included.into the plant Technical Specifications, do not provide reascnable assurance that hazardous accumulations of flammable liquids and combustible materials will not be present in individual plant areas.
As documented in recent Inspection and Enforcement Branch Reports, recent insp,ections
~
at plants such as Davis Besse (50-346/82-03, April 1,.1982), Duane Arnold (50-331/81.-25, January 11,1982), O. C. Cook (50-315/81-11, December 31,'1981), and Nine Mile: Point (50 220/82-09), have demon-k strated that substantial quantities cf huardous substances, such as computer printoyt paper, are locatad in' even highly restricted and controlled entry areas.
Conse'quently, they do not preclude the need for other Nre protection design features.
~
With regard to the control roo1 being constantly manned, we do not
-have reasonable' assurance that p g apt fire discovery and fire
+ighting activities by control. room operatcrs would assure that no s
damage would be sustained'by redundant safety related cableland
\\(
v s
s h
s
f 16 equipment.
The uncertainties concerning the location of the fire, the degree of physical separation of redundant trains, fire propagation speed, the fire damageability of cable and equipment, the timeliness and effectiveness of operator actions and extinguishing efforts, prevent the prediction of damage from fire or fire suppressants.
Consequently, the continuous presence df control room operators and the availability of portable fire extinguishers by themselves, would not assure that redundant trains would be free of significant fire damage.
Only when these cond'iserations'are coupled with the provision of an alternate shutdown capability, are they considered to be sufficient justification for granting an exemption from the require-ment for a fixed fire suppression system in control rooms.
The licensee's fire analysis models both an expos'ure fire consist 1ng of flammable liquid in a metal pan located on the floor adjacent to and outside of the control console and an internal panel fire for a two minute duratiert.
This, represents only two of a large number of potential fire scenarios for a plant control room and demonstrates that, only under the postulated conditions, coul.d safe shutdown be achieved.
There is no basis to conclude that this analysis represents fires which would define a limit of concern for fire exposure in the area.
The impact of a fire located in other areas of the control room, which would involve in-situ combustibles was not considered.
The licensee assumed that a 2-foot by 2 foot pan of heptane fire burning for 2 minutes located adjacent to the control panel or an internal fire within one single control cabinet burning for two minutes are " worst case" fires.
A similar fire of longer I
than a two minute duration or outside the control console could l
cause damage to control circuits for many shutdown systems.
Therefore, there is no reasonable assurance that the plant could be safely shutdown after such a fire.
y_ _ _ _.,.. _ _ _ _.. - _ _ -.
,,_.,,,_-_,t.
.s
. =
17 l
Because of the complex nature of the fire phenomenon and the infinitely large number of variables such as ventilation, flame propagation rate, length of time until fire detection and suppression, that impact it, no generally accepted, consistent, predictor of fire effect exists for any room or area.
Consequently, the licensee's uni-directional approacfi toward providing protection from a postulated flammable liquid fire cannot provide reasonable assurance that; safe shutdown capability will not be significantly compromised.
Although the licensee has provided capability to take local control for essential systems, the control room is not electrically isolated from the emergency control stations.
We find that a. fire in the control room or in the area of any emergency cont ~rol station could affect both areas, thus resulting,in the inability to safely shut-down the p'lant.
Because the nature of the electrical panels in this arer pake protection in accordance with Section III.G.2 of Appendix R impractical, the licensee should provide an alternate shutdown system for the area in accordance with Section III.G.3 of Appendix R.
The alternate shutdown capability should be electrically isolated from the control room. so that a fire in the control room or in the area of alternate shutdown capability which destroys redundant circuits will not affect the ability ta safely shut down the plant from the other area.
With the i
alternate ::hutdown capability installed, a suppression system is not required in the area.
l G.4 Conclusion The level of existing protection in the control room does not provide l
a level of fire protection equivalent to the technical requirements of Section III.G, therefore, the exemption should be denied.
l 0
4,.,.
-~..---....,-,y, y
,,,,.v.,
-~%
18 Summary Based on our evaluation, we conclude that the licensee's request for exemption from Section II.G of Appendix R for the suppression col should be granted.
However, the licensee's request for exemptions from Section III.G of Appendix R for the following areac should be denied:
- Structural Steel in:
Fire Zone 7C Division II Battery Room Fire Zone 8 Cable Spreading Room Fire Z ne 12A Load Center No.1 o
Fire Zone 138 Lube Oil Reservoir & Feedwater Pump Fire Zone.13C ESF Motor Control Center Room (Fire Zone 1C RCIC Room has been withdrawn)
- Intake Structure
- Control Room f.ttachment:
Consultants Report dated March 7,1983 Dated: June 16, 1983 Principal Contributor:
J. Stang I
BROOKHAVEN NATIONAL t/.B
~ f.
ASSOCIATED UNIVERSmES, INC.
l; Upton. Long Istcnd. New Yo& 11973 (516) 282%
FTS 666' 7690 D:pccment cfi uctect Energy March 7, 1983-Mr. Francis J. Nolan U.S. Nuclear Regulatory Comdssort-on, b.C 205 a
n D:ar Frank:
Enclosed for your review is our respon'se ta Norttfern States Power Company' (NSP) fire modeling' and analyses, used ta justify exemption request to-10 CFR -
507 Appendix R, for their MonticeTTo facilitys
~
~
The justification anaTyses reviewed concerit the fire hazard in the control
_ room and a structural steeT exemption request for barriers If you have any questions on the enclosed or any previous reviews, please feel free to call either th,e authors or the. undersigned.
~
Yours truly,
- "r'%
John L BoccierGroup Leader Reliability and Physical Analysis JLB/srx Enc.
cc: R. Bari w/o ene R. Hall W. Kato e
e
.:g;o;;;eecne 6 5u Su/
ADOCK0500Cg3 CF
t EVALUATION OF THE ANALYTICAL FIRE MODELING BY NORTHERN STATE.P,0WER COMPANY (NSP),
IN THEIR JUNE 1952 REPORT AND.0CTOBER 1922 SUPPLEMENT
" FIRE PROTECTION AND SAFE SHUTC0WN SYSTEMS Af!ALYSIS REPORT, MONTICELLO NUCLEAR GENERATING PLANT" Charles J. Ruger and Arthur Tingle Department of Nuclear Energy Brookhaven National Laboratory Upton, N.Y.
11973 1.
INTRODUCTION,
This eport contains*our evaluation of the fire-modeling methodology em-
~
r played by Northern States Power Comp,any (NSP) in their June 1982 report, and October 1982 Supplement, " Fire Protection and Safe Shutdown Systems Analysis Report, Monticello Nuc.13ar Generating Plant." As an alternative to the re-quirements specified in Section III.G of Appendix R to 10CFR50, NSP purports to provide analyses that justify exemption from these requirements in par-ticular plant fire" areas 'T; To strengthen the justification for the exemptions requested, NSP-has in'-
cluded two fire-modeling scenarios which are reviewed here.
First, a deter-ministic, quantitative analysis was conducted which_ attempts to model the effects of a two-minute transient combustible fire. involving one gallon of__
acetone in the control room of the Monticello Plant. A description of this analysis is contained in-Supplement 1 to Attachment 5 of the June 1982 submit-tal.
Briefly, the approach taken is to make conservative assumptions sup-porting a bounding calculation of the maximum surface temperatures of both a CMC-type switch and the supporting console panel, and the effects of the same
. fire on components within the panel.
Second, NSP models the effects of a two-minute internal panel electrical fire on the components inside an adjacent panel. This is done with a one-dimensional heat transfer model which represents the bulk temperatures of the panel surfaces, internal air, and internal objects.
~
Further elaboration dealing with the consequences and likelihood of a fire in the control room, the response times of fire brigades, and an analysis supporting structural steel exemption requests is provided in a second sup-plemental document dated October,1982.
A more detailed description of the NSP approach is contained herein. The overall basis of our evaluation is to assess that (1) the method employed is technically sound; (2) the overall approach will yield realistic or conserva
- tive results; and (3) the end use of the results is valid.
Section 2 of this review describes the fire modeling procedures employed by NSP, our general thoughts on the complexity of the fire-phenomena modeling; some key items we consider as forming the foundation of our appraisal; and our overall evaluation of the approach based upon a detailed critique.
Sec-icn 3 addresses the methods employed by NSP and the results obtained as docu ented in the October 1982 supplement.
An ove'rall summary of our findings is provided in Section 5.
2.
C0:: TROL ROOM FIRE MODELING PROCESS This section summarizes and reviews the deterministic control room fire modeling contained in Supplement 1 to Attachment 5 of the June 1982 submittal by'NSP.
- 2.1
SUMMARY
OF THE NSP MODELING PROCESS The analysis attempts to model the effects of a two-minute transient com-bustible fire involving one-gallon of acetone confined within a two foot by two foot square pan. located adjacent to the lower section of a typical control panel in the control room. The approach taken in modeling the effects of the postulated fire involves an analysis of the thermal response of segments of the panel structure. Time dependent, finite difference methods are employed for discretizing the governing partial differential equations and solutions are obtained which describe the control panel surface temperature distribu-tion.
Additionally, the-ahly. sis calculates the maximum surface temperature of, a CMC-type switch due to'_arfire-induced thermal radiation field.
~
The objective of the analysis is to demonstrate a comprehensive un' der-standing of worst case conditions that may result from a one-gallon acetone fire,that is extinguished by operator action (although in Section 1, NSP states that the fire would be unmitigated). The ant ysis considers the effects of such a fire ou the survivability of control switches on panels wi~ thin the control room and the protection of cabinet internal components.
~
The approach attempts to use conservative models and assumptions to carry out a bounding calculation of the maximum surface temperatures of both the switches and their supporting console panel.
The control panel is assumed to be heated by three mechanisms:
convective heatiing'~from the fire plume, e
e thermal radiation from the hot gases and luminous soot within the flames,
e heat conduction through the benchboard panel.
In addition to these heating effects, three heat-mitigating effects are considered:
e re-radiation, post-fire free. convective heat transfer to the air above the panel, e
e conduction to a heat ' sink.
l Also considered is the effect of an internal cabinet electrical fire on the internal components of an adjacent cabinet by use of a one-dimensional heat transfer model. The time-dependent model considers a heat flux impinging l
on the panel wall and computes the bulk temperatures of the front and rear l
panel surfaces, the internal air,.and internal objects by the mechanisms of l
radiation and free convection.
The assumptions used are intended to be both bounding and conservative and include:
a)
Sufficient ventilation exists for the fuel to burn stoichiometrically.
Maximum, laboratory-extracted values of the heat release rate are used and are assumed to be achieved instantaneously.
b)
Panels and switches are assigned large values of radiation absorptiv-ity and low valves of emissivity thereby minimizing re-radiation.
~
c) Panel heat conduction is considered one dimensional; re-radiation is ignored as well as lateral cooling effects.
d)
Subsequent to ' fire extinguishment, panel cooling is provided throuch free convection to the surrounding air environment.
l The b sic fire model's used are presented in Appendices A.1 to A.5 and B.1 to B.4 of Supplement 1 to Attachment 5 of the submittal.
Included therein are data on heat release rates and descriptions of the mathematical models em-ployed for calculating tuoyant diffusion plume growth; pool-fire induced stratification; pool-fire thermal radiation; heat transfer inside a cabinet; thermal radiative. beat flur to a switch; a one-dimensional heat conduction model for the console paneliteinperature; a panel heating'model for determining boundary conditions at the console. panel edge; 'and a turbulent free convection heat transfer model, which accounts for the post-f.f re cooling of thelanel to the air above its surface after fire extinguishment.
Sections 1 to 4 of Supplement 1 to Attachment bf the submittal descri5e the analytical methods and presents the results of the control room analysis.
Tife discussions provided~in these sections, along with each of the Appendices, comprises the scope of our review. The following section describes the BNL re-view philosophy.
2.2 BASIC BNL MODELING REVIEW PHILOSOPHY For our appraisal, som'e. general thoughts are deemed warranted on the com-plexity of fire phenomena and the state of fire science with regard to enclo-sure fire development.
Computer models for enclosure fire development appear capable of predicting quantities of practical importance to fire safety, provided the model is sup-plied with the fire-initiating item's empirical rate of fire growth and the l
effect of external radiation on this rate. As a science, however, we cannot predict the initiating item's growth rate due to relatively poor understanding of basic combustion mechanisms.
Questions and doubts have even been raised regarding the ability to pre' dict the burning rate of a non-spreading, hazard-ous scale fire in terms of basic measurable fuel properties. 'However, while awaiting development of meaningful standard flammability tests and/or more' I
sound scientific predictions, realistic " standardized" fire test procedures should continue to be formulated for empirically measuring the rate develop-ment within an enclosure, and the convective and radiative heat loads to " tar-get" combustibles.
Thus, in lieu of large-scale computer codes to assess the fire hazard in an enclosure, we " define " state-of-the-art" for the purposes of
~
this evaluation as one which incorporates a unit-problem approach to seven general components of the fire consi' fred relevant in undbrstanding, at least' d
i
}
on heuristic principles and pragmatic efforts, the phenomena of fire.
The following list may be obvious, but, in the framework of this unit-problem approach, how one considers the complex heat flux and material flux interac-tions within the fire-modeling methodology forms the general basis for our appraisal.
The seven components and the various important interactions are:
t e The burning object receives radiative and convective heat from the com-busting plume and radiatfve heat from the hot ceiling layer and possib-ly the ceiling.
o The combu' sting plume (or flame) receives volatile species from the j.
burning object.
It receives air (which may be preheated and vitiated in oxygen) from the cold layer.
When the upper point of the flame ex-sands into the hot layer, overall burning may be modified.
Room geome-4 I
try, non-combusting' obstacles, and burning object location influence plume development.,
-r-The hot layer will be' influenced by natur'ai and forced ventilation, by
~
~
e i
the heat and gas combustion products produced by the flame, and by heat i
losses to the enclosure walls, ceiling, and other objects.
Also, tran-l sient combustion within the hot ceiling layer has been observed and may be considered an interaction with the flams.
Transient combustion'in the hot layer could be due to excess pyrolyzate from the burning ob-ject (both solid _ firebrands and gaseous inc.qmplete products of,combus-tion).
e The cold layer is influenced by the natural and forced ventilation, the hot layer, and obstacles within the enclosure.
The targets are heated by radiation-(and also convection for an upward e
spreading fire), c.oming from the combusting plume, the hot layer, and possibly the cei~1ing.(if the hot layer is transparent to radiation).
Ignition of a target increases the overall thermal energy content with-in the enclosure.
The enclosure geometry (ceiling and walls) is heated by convection and e
radiation from all burning objects, and the hot ceiling layer.
e The vents influence the mass ficw rate of oxidizer and the radiative i
and convective components of thermal energy loss.
Positive feedback'is a critical part of the fire growth phenomenon and.its accountability within the licensee's submittal has als.o been a factor in our evaluation.
Granted, each form of interaction has a characteristic time or l
physical dimension associated with it, whi,ch would provide a measure of its.
l relative importance.
l
. ~ -
In this connection, we agree with the tacit assumption made by the 1icensee in that for the particular.. fire scenario consid,ered, the character-istic time is such that the effects'6f the control room enclosure should not greatly affect the predominant heat-transfer mechanisms considered.
2.3
SUMMARY
EVALUATION OF THE NSP APPROACH As a concept, the overall methodology represents, in part, a technically sound and conservative technique for assessing the potential hazard presented by exposure fires to a surrogate control room panel.
Thb modeling tools 'used in performing the bounding calculation of the max-imum surface temperatures of the console and the switch and cabinet internals, consists in employing the' following unit models:
o pool-fire plume model e panel switch radiation model e one-dimensio_nal console panel heat conduction model o panel heat-up boundary. condition model a turbulent free convective cooling model thermal analysiMpf. cabinet / panel inter.nals e
The approach employed, together with the assumptions made can bh clas'si-fied as methdologically consistent for assessirft) the fire hazard potential to console components associated with limited sized transient combustible pool fires.
~
Thus, in most respects, we find the method employed to be technica.lly sound and the overall approach, if applied properly (as described subsequently) could yield more realistic results for assessing the thermal response of the panel elements under study in the fire area.
However we have some concerns
, regarding the completeness and accuracy of the approach.
These are based on the following general observations:
~
1)
The analysis' does not account for convective heating through panel ventilation ducts or other openings.
2)
The analysis does not account for the heating of the backside of the l
upper console panel by convection from the hot air within the console l
interior.
3)
No substantiation is given for the assumption that the bakelite i
switch plate can be treated as thermally thin.
4)
The form of the configuration factor used to obtain.the decay of the radiar.ive heat flux with distance from the fire is not presented; 5)
It is i.nclear how the material presented concerning front panel heat-up is related to the physical problem considered.
4 6)
Results are not presented for the post-fire coupling between the heat conduction along the panel and.the convective cooling to the air above the panel.
- 7) A clear definition of damage criteria is not given.
Other than items (1),(2), and (3) above, it is our feeling that resolution of the remaining points will not markedly deter from the conservatism em-ployed, but will provide additional credence for assessing the overall fire vulnerability of the structure under study.
The possibility of convective
, heating of elements within the control panel or indirect heating to the back-side, unexposed portions of the p.anel alluded to in (1) and (2) are in our estimation areas of concern which require more in-depth elaboration by the '
licensee.
WhetEer the themalTesponse of the Bakelite switch is dominated by the resistance in the plate or by a gas film heat transfer coefficient dictates the temperature distribution through the plate thickness. A physically thin plate having a."small"_ thermal conductivity and a "large" film heat transfer coefficient, i.e., a Biot Number greater than 0.1 manifests itself with higher surface temperatures at a given time than a plate which has associated with it a Biot. Number whicli is Tfnithan 0.1.
This is our concern under Item (3) above.
.a 2.4 DETAILED EVALUATION OF THE NSp APPROACH The basic fire models are presented in Appendices A.1 to A.5 and B.1 to B.4 of the submittal. These appendices include: data on heat release rates-and models for buoyant diffusion plumes; pool-fire induced stratification; pool-fire thermal radiation; heat transfer inside a cabinet; thermal radiation to a switch; a one-dimensional heat conduction model for the console panel temperature; a panel heat-up model for determining the boundary condition at the edge of the console panel; and a turbulent free convection model, which
~
accounts for the post-fire cooling of the panel to the air above.
Sections 1 to 4 of Supplement 1 to Attachment 5 of the submittal describe the analytical
' methods and presents the resbits of the control room analysis. These appen-dices and sections are now described further with regard to modeling, assump-tions, completeness and application of the methodology.
2.4.1 Review of Appendix A.1 l-Appendi.x A.1 of the submittal describes a basis for selecting liquid hydro-carbon heat release rates, based on the current state of knowledge in fire sciences.
Values of the heat of combustion, vaporization rate, and heat re-i lease rate, are given for acetone, lubricating oil and heptane. The assump-l tion that ventilation is always sufficient to provide ideal fuel-oxygen ratios l
leads to the use of a conservative upper bound for the heat release rate.
Also, conservative asymptotic values (large scale fires) for steady-state mass loss rate per unit area are used, i.e., the.f. ire is assumed to reach steady--
state conditions immediately. The use of laboratory-scale generated, actual l
heat of combustion data by Tewarson is also conservative since the most ef-ficient combustion achievable in the laboratory is employed in the analysis.
l 6-O
1 2.4.2 Review of Accendix A.2 Appendix A.2 of the submittal is based on the correlation of Newman and Hill l for the convective and radiative heat flux in the stratified ceiling hot gas layer developed by a pool fire within an enclosure.
The heat flux is related to the room's dimensions, the target height above the floor, the' fuel's flammability parameters, a.nd the room ventilation rate.
The submittal's a,nalysis of small control room fires did not necessitate-consideration of room enclosure effects. Therefore, the stratification model was not used, but is reviewed below for completeness.
This correlation should be adequate for evaluating the heat flux due'tio pool exposure fires.
Howayer, it should be pointed out that one conclusion reached from the data in Reference 1 and carried over into the correlation, namely that horizontal ~ heat flux variations are mi'nimal, is not in agreement with some other authors 2-5 In these references,' data 4 and theory
,4 2
show that, for radial distances from the fire plume axis greater than 20% of the ceiling height, -the heat flux decreases with radial distance to the -1/3 power.
However, in re-examination of Figure 7 of Reference 4, the heat flux appears to have a radial _Aependency to the -1.25 power.
This is shown in Figure 1 provided herein.v To further check this difference, we utilized the
~
heat transfer coefficient parameter, hc. presented by Veldman, et al (Re.
l' ference 10) in their Figure 14.
This shows a radial dependency foF'this para-meter to the -0.6 ~ power which, when applied to the -2/3 power correlation i
presented by Alpert in Reference 2 for the maximum plume temperature diff.er-ence, aT, yields in concert a radial power law dependency of approximatebr
(-1.27), which is in close agreement with the -1.25 power indicated in Fig.1.
s These works consider a quiescent enclosure while Newman and Hill. include forced ventilation in most of their tests.
However, since Newman and Hill's heat flux data for no ventilation fall in the center of You and Faeth's data 4 for radial distances ' closer than 20% of the ceiling height (no radial dependence), the neglect of the decrease in heat flux with radial distance by.
Nawman and Hill should yield a conservative result.
This also tends to show no benefit to. horizontal cable separation for radial distances closer than 20%
of the ceiling, height.
On the other hand, References 3 and 5 show that if the exposure fire is near a wall or in a corner, the ceiling temperatures increase as if the fire
~
heat release rate is increased by a factor of 2 and 4 respectively.
There-l fore, care must be taken in applying the Newman and Hill correlation for ex-posure fires in the vicinity of walls or corners so that non-conservative re-
~
sults are not obtained.
j The submittal docs not* use the Newman and Hill correlation exactly as. pre-
=sented in. Reference 1.
Instead, a modified form as given on page A.2-4 is used.
Apparently, this was done to extend the correlation at ventilation' rates greater than those for which measurements were taken in Reference 1.
This fact, coupled with the unrealistic cooling behavior of the original Newman and Hill correlation at _ higher ventilation rates as shown in Figure A 2-2, leads to the need -for the modified correlation, which continues the r
data trend to higher values of ventilation.
This modified correlation is core conservative than tne original.
Sin.ce the labeling of Figure A.2-2 is some-wha: confusing, it is replotted as Fiiure 2 (attached)'wlth the modified f
correlation on page A.2-4 included.
The correlation is not valid if secondary I
fires occur, or if excess pyrolyzates Durn in the stratified layer.
2.4.3 Review of Appendix A.3
~
Appendix A.3 of the submittal describes a turbulent, buoyant diffusion plure model which is essentially the classical Morton-Taylor model.
The ex-peri ents of Stavrianidis6 are considered along with his correlations for
~
criticEl height, (height above which plume correlations are valid),4and virtaal' source height. 'The heat flux correlations of You and Faeth for the stagnation region (r/H < 0.2) and the ceiling jet are also presented.
The cor-relations are for Rayleigirnumbers of 109 to 1014, whereas the fires dis-cussedln the submitta1 have Rayleigh numbers of about 1017 There should be some defense of this extension.
These represent state-of-the-art correlations for hydrocarbon pool-fire pl umes. However", th'ere ar'e several errors, most likely typographical, which should be corrected.
First, the exponent of the factor Fa in the buoyancy expressions on pages A.-34 and A.3-3 should be. 2/3 rather than 1/3.
A review of You and Faeth's work yi-el'ds the following connents concerning the heat flux correlation on pages A.3-8 and A.3-9 of the submittal.
The Greek srymbol i ap-I-
pearing in the Rayleigh number is defined as the~ kinematic viscosity, not the radial velocity.
The heat flux correlation appearing on the bottom of pace A.3-9 is valid in the ceiling jet, outside the stignation region (r/H > Q 2) for free-flame height to ceiling-height ratios up to 2.5, as evidenced by the Jata in Figure 7 of Ref.erence 4.
The radial dependence in the correlation-m should be to the -1.25 power as explained in the review of Appendix A.'2.
Also, the stagnation region (r/H < 0.2) correlation is valid for. free flame height to the ceiling height ratios less than 1.5, not greater as stated, a...
2.4.4 Review of Appendix A.4 The radiant heaitiansfer from a high-temperature, turbulent, buoyant dif-i fusion plume is discussed in Appendix A.4 of the submittal.
A classical ap-proach based on the Stefan-Boltzmann law is used.
A uniform gaseous tempera-ture of 1255' K is assumed based on the work of Stavrianidis6 It-is not clear which correlation for flame height is used, although Stavrianidis has a correlation for hydrocarbon which is consistent with-data.
However, passing mention of Steward's7 work is all that is found in this. Appendix.
Effective-values for. gaseous and soot emissivities are used, with a value of 0.1 being taken for soot.
An expression for the gaseous emissivity, which is dependent on :ne gaseous temperature, the partial pressure of CO2 (a combustion procuct), and the mean beam length is presented.
These classical expressions and assumptions are acceptable as the present state of knowledge in radianti heat transfer.
~
However, there is some confusion about the definition of mean beam length
~
on pages A.4-5 and A.'4-7, where it is' defined as a fraction ~of the electrical cable diameter.
The mean beam length cannot be.a function of the target I
receiving the radiation, but must be a geometric property of the. flame prod;cing the radiation.
Hottel and Sarofim8 have shown that the average r
1 -
mean beam length for a target at the flane boundary (very conservative) is we11 approximated by La = 3.5V /Af f
where Vf 1s the flame volume, and Af the flame bounding area.
Less conser '
vatively for tar;e:s far removed from the flame, a somewhat better approxima-tion 9 for L is 0.9 times the ratio of the effective flame volume to the m
flame area projected on a vertical plane.
It is not clear if this expression was used in the de ermination of the needed gaseous emissivity in the calcula-tion of radiant heat transfer, or whether a value of 0.2 was used as in Appen-
~
dix B. l' of the submittal. Also, calculations for a cylindrical flama, using the above mean beam length, give approximately the same heat flux results -as the expression on page A.4-7, with D equal to the fire diameter.
Therefore, the use of cable-diameter -tn the submittal may only be a documentation error.
A typd'g?aphical error does exist on page A.4-6, where both the factors 0.131 i
and 0.940 should be raised to the 0.412 power.
Also in need of clarification is the nature of the configuration factor used to obtai' the fraction of the heat flux delivered to a target point by n
the assumed radiant right, cylinder.
The equation on page A.4-7 contains this factor but no mention is gade as to what values,are used or from where they are obtained.
~
2.4.5 Review of A:oendix A.5 This appendix contains a brief discussion of {_one-dimensional heat tf}ns-fer model for conputing temperatures of objects inside a cabinet or panel. We feel that NSP sh uld discuss the limitations of the model.
For exampl.e the back wall in Figure A.5-1 appears to be exposed to a constant ambient tempera-ture during the fire.
This may not be valid in general.
There are so e typographical errors.
Part of a heat radiation term is
' ~~ ~ missing in Eqn. 1 on page A.5-2; the numbers at the bottom of page A.5-3 are for the product of density and heat capacity (oC in the model and not c in 3
the text); and the u. nits.in parenthesis should be BTU /in -R for both steel and air.
2.4.6 Review of Arsendix B.1 This appendix contains a model for thin-wall temperature response.
Under
(
the thin wall condition, there are small te'mperature gradients through the plate thickness and heat received diffuses instantaneously with little re-sistance through the material thickness.
This simplifies ~the mathematics of heat conduction and also possibly affords analytical treatment of more complex systems.
As a practical measure, a plate is considered thermally thin if the temperature difference across its thickness at a given instant is less than some prescribed value.
However, the thin-wall approximation may possibly not-be valid for Bakelite since its thermal conductivity is much less than that for steel.
Since Bakelite is a poor conductor of heat, the concern here is that the switch plate surface temperature may reach much higher temperatures for the,same ' input energy than that obtained by assuming the heat to be more easily diffused :nroughout the thickness.
_g-t
-s
-m.
i 1
Thin wall approximation notwithstanding, the model calculates the response of a thin plate exposed to a radiant heat input while re-radiating to a con-stant sink environment.
The equatidh-is solved by a co'amonly used fourth-l order Runge-Kutta method.
However,theabsorptivityconstantigmissingfrom the equation and the density of Bakelite should be 0.046 lbm/in, not 0.46 as stated.
We compared the temperatures computed for the switch at the fire extinguishment time of 120 seconds in in Table 3-3 with those plotted from an analytical solution to a similar problem found on page 3-36 of Rohsenow and Hartnettll(Fig. 3 attached).
We took the incident heat flux, q" in the I
~
figure 7 to be that given as QRAD at various distances from the plate edge as given i'n Table 3-3 and u' sed an irmissivity of 0.1.
Combined with the properties of Bakelite given on page B.1-1, we found that the temperatures obtained from Fig. 3 agree-closely with those in the submittal.
We also i
^
veriffs' the other cal.culations in Appendix B.1, although it is not clear how d
the emissivity for gas is computed.
Note however that both analyses assume the " thin wall" approximation.
However, 'using a~ class 1 cal. configuration factor we could not verify the variation of QRAD with distance from the fire in Table 3-3.
Since no expres-sion for the conf.igurat4f factor was submitte.d.this question remains unanswered.
i Some brief comments are now made on the bgrn-time calculation. First, an l
inconsistent value of mass loss rate of 50 g/m sec for acetong is used in-stead of the more commonly accepted asymptotic valtre of 40 g/m sec,which_'Is stated in Section 2 and Appendix A.1 of the submittal. Use of the latter will give a 20% longer burn time.
Secondly, the burn time has been made academic by the assumption in Sec-tion 2 that the scenario for the transient exposure fire credits operator intervention after two minutes. This contradicts the stated objective in
- - Section 1 of analyzing an unmitigated transient combustible fire.
l 2.4.7 Review of Apoendix B.2 Appendix B.2 contains a' finite-difference algorithm for solving the one-dimensional heat conduction equation with a source term.
The problem under investigation is the heating of a flat steel plate. console panel by two mechanisms: 1) a radiative flux incident on the top of the plate, which varies with distance from the fire, and 2) conduction of energy deposited at the edge of the console as result of direct flame impingement to the console front face.
Thus, a two-dimensional conduction problem in a plate with boundary condi-tions consisting of an applied spacially dependent radiative flux on the upper surface, an insulated lower surface, and imposed temperatures at both ends has been reduced to a one-dimensional problem with imposed temperatures'at both ends.
The plate has been assumed thermally. thin such that the radiative heat-flux incident on the surface can be considered evenly distributed throughout
~
the thickness and thus can be represented as an internal heat-source term in
.the one~ dimensional conduction equation.
For a. steel plate of the given thick-i ness this should be acceptable.
S.
,r
--r yg g -,. -
e g,
w
.w y-9-.
w y
s-
= + ~-
p.
.g "i--*--.mp mm--t-+-
t v
er
However, one mechanism for heating the upper console panel has not been considered.
The fire is located adjacent to the console face.
Therefore, even if there are no ventilation slolif exposed to the f' re, the air tempera-i ture inside the panel will increase due to heat conduction through the front panel.
This hot air will supply heat input to the bottom of the upper panel by convection.
Admittedly, this would be a conplex mechanism to include, involving model-ing of the internal cabinet components, and it is not clear that the model in Appendix A.5 is applicable.
If the lower front panel is closed off with a non-conducting shield, the under surface heating of the upper panel should be insigi ficant.
The radiative flux distribution is not submitted.
However, hand calcula-tions using the 'radiativ'e ' flux source term in the finite-difference equation at the' bottom of page B.2-3 indicates that the distribution is similar to that given for the switch in Table 3-3.
Since the switch is assumed vertical and the console is almost horizontal the configuration factors should be diferent.
However, the yalues used are conservative since the radiative flux to the panel surface can be expected to be smaller than that to the switch due to its orientation.
' ~ *-
Also, the expression for'Cn in the analytical solution to the steady-state heat conduction equation on page B.2-1 is iilco{2, pag (probably typo -
rect graphical). The correct solution is given by Kreith e 90.
2.4.8 Review of Appendix B.3 We believe the intent of this appendix was to provide a front edge temperature boundary condition for the heat conduction problem posed in Appendix B.2.
The console front panel is heated by flame impingement and the heat is conducted to the front edge of the upper console panel.
~ ' ~
However, the first page of Appendix B.3 concerns itself with post-fire convective cooling of a plate to air at a uniform temperature.
This is repeated as the first page of Appendix B.4 and is discussed in the next section.
The'second page of Appendix B.3 appears to describe the heatup of a plate initially at a temperature T, which is then exposed to a gas at a o
temperature of 200*F.
(Note that a minus sign is nissing from the argument of the exponential term.)
It is unclear what this problem has to do with the panel heatup due to a fire with a gas temperature of 1763*F as mentioned in Appendix B.l.
It appears that some errors were made in transcribing ~ this appendix and the content should either be explained or presented again.
2.4.9 Review of Appendix B.4 This appendix describes the models employed to assess the post-fire cool-ing of the panel console.
Basically there are two mecharsisms that should be accounted for in this phase of the overall analysis, viz.,
I (1) thermal energy redistribution along the plate as a result of a non--
)
uniform lengthwise temperature distribution while the fire was in l
progress,.
(2) free convective heat transfer to the surrounding medium.
A hybrid analytical / numerical approach is employed by the licensee.
For purposes of describing their approach consider for the moment a segment, tx, -
of the plate.. This sub-element of the overall plate (or panel) is likened as a computational cell in the licensee's finite-difference analyses which we have discussed previously.
The internal thermal energy of this segment changes as a result of (a) heat being conducted along the plate since con-tiguous' elements a're at 'different temperatu'res and (b) concomitantly, heat is also being convected away due to an energy exchange with the environment ~.
If the boundary conditions that must be imposed on the two plate segments which F
make ythe front and.back edge of the plate, i.e., at x = 0 and x = L do not contain a heat source mechanism, the process (a) is a manifestation of plate temperature redistribution while process (b) is the inherent cooling mecha-1-
nism.
If process (b) did not exist and the ends were considered, say insula-ted, then thr'ough pr6 cess la) the temperature will approach a uniform steady state value.
But at each time instant, the overall thermal energy in the l
plate will be a constant -v)1ue since there is 90 mechanism for loss.
The approach taken by the licensee basically decgyples these two; processes i
as described below.
The approach, taken from Kreith" relates the convec-tive heat transfer coefficient to the Grashof number, which is taken as a function of the instanteous surface temperature.
The cool down process if_
assumed to have an exponential form.
(The Grashof number of page B.4-1 is high by a factor of 100.
This is probably a typogcaphical error.)
For obtaining the panel edge temperature during cooldown, the heat trans-fer coefficient is apparently evaluated at the panel edge temperature at the.
time of fire extinguishment. This assumption of constant heat transfer i
-- coefficient somewhat non-conservatively accelerates the-panel edge cooling since the temperature difference between the panel and the air has its maximum' value at. fire extinguishment.
Also, there is'some confusion about the temperature of, the eiivirsnment since the temperature on page B.4-1 is 200*F, while the equations on page B.4-2 use 70*F.
The input data listed at the top of Table 3-2 also shows a temperature of 70*F for the post-fire air, although j
no post-fire results are shown.
Away from the panel edge the situation is more complex. Apparently in' this case the heat transfer coefficient has been assumed' quasi-steady.
That is, for each conditional time interval, at, the heat transfer coefficient is evaluated at the last known nodal temperature at t.ime, t, and assumed constant l
for the integration over a t.
The incremental temperature, AT, thus obtained is then added to the temperature calculated during the previous time step to obtain the temperature at time t + at. This is an acceptable procedure and' will yield-accurate results if the time steps are taken small enough.
L l
..~.
~
~. - -,
1 Ho0ever, the apparent intent of the submittal is to superimpose this cooling solution on the heat conduction solution obtained from the partial differential equation in Appendix 3.2." That is, after fi're extinguishment, the partial differential equation on page B.2-3 is solved now with the source term, representing the radistive flux from the fire, set to zero and with the i
post-fire edge boundary conditions obtained by the analysis in Appendix B 4.'
By definition the initial tenperature distribution along the length of-the i
plate is-given as the values calculated at the time of fire extinguishment.
Considering conduction alone, this results in merely a temperature redistribu-tion along the plate, since no additional ~ heat is being added.
At each time i
~
step, the aT change at each spacial node, due to cooling, is then subtracted from the redistributed tenperature achieved, by using the above technique..
Conceptually, a more acceptable procedure would have been to include the turbulent free convectioh rooling mechanism as a boundary condition for the transi'sitt heat-conduction equation.
However, since the results as tabulated on page 3-6 are not continued after fire extinguishment, we are unable to determine if the procedure used gives rise to a physically realizable situa-tion..
2.4.10 Review of Sections 1 to 4
)
- ?;
Sections 1 to 4 of the-June 1982 submittal. give an overview of the
(-
modeling process employed. Such a summary was presented in Sectionr2.1 of this report and an evaluation of the details of the modeling is contained in the review of the appendices earlier in this section.
Only a few comments will be made here regarding 1) modeling assumptio51, 2) panel ventilatiort_
openings, 3) conduction and convective cooling model coupling, 4) damage cri-teria, and 5). effects of internal panel fires on. adjacent cabinet. internal-components.
The assumptions used can be considered to be conservative with respect to
. _ _ the bounding calculation of 'the analysis.
The use of laboratory values for i
heat release rate, which are assumed instantaneously achieved,-and'the assump,
tion of sufficient oxygen availability for most efficient burning should result in the most intense f. lame at the console. The high values of radiation absorptivity ensure that a ' conservative portion of the heat released from the fire is transmitted to the panel and switch.
Cooling is also conservatively minimized by using low values of emissivity,. limiting lateral heat conduction and allowing post-fire free convective cooling to surrounding air.
The analysis does not mention ventilation slots or other openings in the
~
panels.. Consideration of such openings is necessary because.the vents could draw in the flame or hot combustion products, thus subjecting the console interior to much higher heat fluxes than the analysis calculates.
In the October 1982 supplemental submittal, it is proposed that some panel backs will be closed off by a barrier. This.would be adequate to prevent flames entering, but it should be noted that this barrier must also prevent heat-conduction through its thickness so that the upper panel surface is not heated by convection from belos as discussed in Section 2.4.7 of this report.
Of course the permanent closing of the cabinets must not jeopardize the control panel unper normal operation.
The technique used to include the post-fire free convective cooling from the hot panel with the process of one-dimensional transient heat conduction has not been validated in.the submittal.
This was discussed in detail in i
Section 2.4.9 where Appendix B.4 was reviewed and is briefly sunmarized here i
for completeness.
Essentially, Appendix 3.4 develops an independent exponen-tial cooling solution which results in an incremental temperature change for a given time. increment. This tenperature increment is then superimposed on a i
solution of the one-dimensional heat conduction problem of Appendix B.2.
As applied, the results this superimposition of solutions cannot be verified as to physical correctness since no post-fire results are presented.
a Another comment relates to the damage criterion applied to the switch l
and control-pnel internal components.
The submittal uses the peak temper ~ature as a criterion to co.nclude t.haj control-panel component availability is main-tained.---A damage criterion based on absorbed energy would be more desirable since damage must be thought of in terms of time-at-temperature rather than as 4
just the maximum temperature the component attains. Admittedly, it is ex-tremely difficult to define an gsorbed energy damage criterion such as
) for a component such as a switch. However, used for electrical cables (Lee the lack of such a criterion makes the evaluation of the results of an analysis, as contained in tile subm_i,t,tal, uncertain.
~
i The internal panel fire analysis postulates the instantaneous combustian of electrical cables and polymeric material withi.n a single panel and assumes achievement of an incident heat flux of 75 kW/m2 on the side metal walls separating the individual panels for a duration of 2 minutes.
The tempera.ture of a switch body, located 21/8 inches from the hot wall of the panel con-taining fire is calculated using the "HOTB0X" model described 'in Appendix A.S.
We question the application of the one-dimensional " HOTBOX" model in relation to Appendix R exemptions since it gives only a general understanding of thermal lag and panel fire resistance, and there is no clear discussion of
- - boundary conditions.
Also, there could be a loss of functionability due to thernal stresses induced by the highly transient thermal environment, and additional heating through ventilation ducts or other openings.
3.
REVIEW 0F THE OCTOBER 1982 SUPPLEltENT This section summarizes and reviews the October 1982 Supplement "ftonti-cello Nuclear Generating Plant, Clarification of Information Provided in Support of Request for Relief from Requirements of 10 CFR part 50, Appendix R, l
Section III.G."
This supplement requests exemption for six barriers based upon an analysis of structural steel.
For the control room, the Supplement provides inrormation on the response of fire brigades, frequency of control-room fires, and further discussion of external and internal panel fires.
Exemptions are also requested for the Suppression Pool Area and the Intake.
Structure, but no analysis is submitted and these requests are not reviewed.
4
i 3.1 RE7IEW 0F STRUCTURAL STEEL EXEMPTION REQUEST
. of the October 1982 sdbnittal contains infDrmation supporting an exemption frca 3-hour fire rating of structural steel supporting a 3-hour barrier separating redundant safe shutdown systems.
NSP has identified six barriers that employ exposed structural steel and separate redundant cafe shutdown systems.
NSP states that quantitative l
analysis indicates that these barriers have fire protection equivalent to Appendix R criteria.
All barriers are ceiling interfaces.
Five barriers have steel I-beams below the concrete and one has I-beam embedded in concrete.
Each barrier separates two fire zines located above each other. The lower zone fird is the only reldvant fire since thf steel is exposed on the ceiling side of the barrier.
The fire damage criteria is defined as a thermal limit for the weakening of structural members. Steel is damaged in fires through the loss of yield strength and rigidity. By design, stresses are normally limited to 60% of the yield strength.
From a NFPA handbook one finds that the yield strength of A36 steel is 60% of the ro'bm temperature design value at 1100*F. Therefore, NSP uses a steel temperature of_1000*F as the definition of failure.
NSP also believes that they-are centervative since the dead loads.are only a small. part 4
of the loads carried by steel members.
~
NSP assumes that the steel beam thermal lag approaches zero in thermal equilibrium with the environment and that the beam is thermally isolated except for re-radiation at 95% emissivity. Computations are made for the -
minimum heat and energy fluxes necessary to cause failure. These fluxes are determined through a fire modeling process that relates the fuel source and _
~~
th'e target steel. The proprietary fire model is referenced, but not presented for review.
However, NSP has not considered all of the effects of fire on structural
~~ lteel. For example, when a load is applied to a material it will undergo an instantaneous deformation and when the load is maintained there is also a 8
deformation that incre.ases.with time. This time-dependent deformation is generally known'.as creep'. At normal stresses and at temperatures representa-tive of fires, the rate of deformation caused by creep can be substantial.
Also not considered is thermal expansion of the beam which can cause failure l
due to excessive comprehensive stresses on restraints and supports and the bending of the beam due_to temperature gradients.
l.
The fire model analysis. computes the quantity of transient combustible,
. heptane, rdquired to raise the steel temperature near 1000*F and shows that this quantity is in excess of a fire hazard of concern involving 5 gallons of
' liquid hydrocarbon.
However, we note that for fire zones with heavy fixed j
combustible loadings the exe'ption request relies heavily on early detection i
and rapid suppression.
Fires in these zones are capable of developing into full room involvement and therefore the credibility of dependence on detection and suppression must be thoroughly investigated.
9 O -
3.2 REYIEW OF CONTROL ROOM EXEMPTION REQUEST
. of the October 1982 slibnittal contains information similar and additional to the control room fire analysis presented in the June 1982 sub-mittal.
Section 3 discusses the overall methodology, criteria, and results of the analysis.
Appendices A to D contain a statistical analysis of manual sup-pression time, a model for a fire in front of the control panel, an internal panel fire model, and the frequency-of control room fires.
These sections and appendices are somewhat sketchy and do not provide very
, many details or results. However, they are briefly reviewed in the following.
~
3.2.1 Review of Section 3 The control room fire-analysis attempts to model the effects of two scenarios:
1.
a fire internal to a single panel which threatens equipment in an adjacent panel; 2.
transient combustible fires adjacent to the front and back of the main control panels and.-]ocated between redundant components.
~_
In both cases it is assumed that the fire is suppressed at two mi.nutes by
-operator response.
In the case of the external, transient combustible fire, heptane is considered as the combustible.
In the June 1982 submittal, acetone was considered and suppression at two minutes was (Ihconservative due to the-approximately two and one-half minute unmitigated burn time of a 2 ft. by 2 ft., one-gallon acetone fire. However, the higher mass loss rate of heptane.
results in an unmitigated burn time of about one and one-half minutes for the same fire size and fuel volume. Therefore, the use of a two-minute burn time for the present heptane fire assumes more than one gallon of fuel and is
.therefore more conservative.
The transient combustible fire is a heptane pool fire in a 2 ft. by 2 ft.
pan. Two redundant switches located 14 inches apart were modeled to detemine if a front panel fire coul'd.. disable both switches. The time dependent thermal
~
profiles on the steel panel are calculated using the finite element model briefly discussed in Appendix B.
The model considers the effects of heat con-
~
duction through the plate, radiation to and from the plate, and convective cooling to the air flowing over the panel and to the internal panel air, as-semed to be at 122*F. The resulting temperature profiles are not provided.
NSP states that the dominant heat transfer mechanism is radiation to the hand-le surface which could result in some minor defomation at the edges of the phenolic material.
However, NSP does not provide a description of the methodology used in computing the temperature response of the switch due to radiation and we are unable to verify this conclusion.
No modeling analysis is presented for the rear panei heptane fire.
Rather, NSP proposes to protect one division.of the safe shutdown system from damage by installation of steel panel backs. Of course, the permanent closing of the panel backs must not jeopardize the control panel under normal opera-t, ion. This consideration of fire or fire product, ingestion into the interior of the panel alleviates somewhat our concern on panel openings in Section 2.
The' internal panel fire analysis and results are the same as in the orig-inal submittal. Hence, our comments o,n the overall appli.cability of the one-dimensional model are unchanged.
However, we note that NSP has identified one '
panel that does not have a re.dundant safe shutdown train back-up. To elimi-nate this situation, NSP proposes to place one or more S/RVs in a separate panel.
This modification is not included in the summary of the Control Room Proposed Modifications in Section 4.2 of the supplement.
3.2.2 Review of Appendix A This, appendix develops a distribution for the anticipated response time for manual suppression of a contr.pl room fire. The results show a mean re-sponse time of about one minute with an upper limit response time of two minutes given to 95% confidence.
A Baiesian analysis..was performed which represents an acceptable tech-nique. However, the only expert opinions utilized were obtained by question-ing licensed operators of the Montic'ello plant. More comprehensive data than this limited sample is, requi, red to substantiate the results obtained.
3.2.3 Review of Appendix B
_ 7j.
The front panel fire model assumes. that the, target switches are heated by two processes:
~
1.
direct radiation from the fire onto the switches and the panel surface surrounding the switches-
~
~
~
_. 2. conduction along_the panel to the switches...
Also considered are two cooling mechanisms:
1.
radiation from the panel and switches to the room environment and panel internals, 2.
forced convection from the panel surface to cool air pulled along the exterior surface ~ by the fire draft and natural convection to the interior of the panel.
The steel panel temperatures are computed using a time dependent finite ele-ment model. This is apparently a different model than presented in Appendix B.2 of the June 1982 submittal in that it is two-dimensional.
From the information given, it is unclear how the surface heat transfer mechanisms;
. radiation, 're-radiation, and convective cooling, are treated.
Is the problem treated as three-dimensional with these mechanisms as boundary conditions, or is the plate considered thin resulting in a two-dimensional problem with the surface heat transfer terms uniformly distributed through the thickness as source / sink terms, or are separate solutions obtained for each mechanism and then superimposed on each otherr Since no details or results are presented, further comment is not made.
~
e.
The method used to compute the temperature response of th@ switch due to radjation is not described.
Also, the steel properties used in the analysis are confusing since the units of condiictivity should be W/m-K and heat ca acity should be either.per unit mass or given as volumetric heat capacity.
3.2.4 Review of Accendix C Appendix C discusses the one-dimensional internal panel heat transfer nedel. This is the same model presented in Appendix A.5 of the June submittal.
and has been reviewed in Section 2.4.5.
- 3.2.5 Review of Appendix D This appendix attempts to assess the relative fire hazard in control rooms with rgpect to other plant-areas.
In a report by R.W. Hockenbury and M.L.
Yeater
- a model for the estimation of the time-dependent fire occurrence rate for the control room has been used. This model, which is based on a non-hemogenous poisson process with a Weibull distribution for failure rate represents a total population of fire data accumulated from 17 nuclear power plants. The applicaticm of'this.model to the control room of the Monticello.
plant under the assumptions that the plant is 11 years old and 2% of t occur in the control roomy-bas. predicted a fire frequency of 2.5 x 10 ge fires per year for the occurrence of~ fire in the control io'on of this plant for the twelfth year of operation. By judging this estimation, it has been concluded that fire occurrences in the control room are of much smaller relative hazard than the other general plant areas.
We, as a reviewers, cannot approve of the assesTment that the fire hazard in the control, room area,is less hazardous than othe_r areas because the, frequency in this area is much lower without giving consideration to the consequences of each evont.
In addition, the fire frequency calculated in Appendix 0 as 2.5 x 10-4 per year for the eleven-year old plant seems to be
.in error by a factor of 10. Our calculations using an expregsion such as aquation B.3.2 of Reference 14 give a frequency of 2.5 x 10-4 per year.
~
~
4
SUMMARY
For the control room fire analysis, based upon our engineering judgement alene, we are of the impression that for the particular fire scenarios censidered, the candidate plate switch will survive the impressed thermal environment and that an internal cabinet fire will not destroy the functionability of the internal components of an adjacent cabinet.
However, the modeling employed by the licensee for predicting the thermal l
response of the control console to an exposure fire leaves us with some doubts
~
between what actually should occur as a result of the fire and what is i
j predicted.
l One would have felt more assured of the overall approach had the licensee addressed more fully the thermal environmen't'within the console interior; had the licensee performed comparative calculations assuming the Bakelite switch is modeled as a " thick wall" as well as a " thin wall"; and had the " cool down" -
analysis b::cn continu:d beyond fire extinguishmant.
Also, nonadhsrence to stated' assumptions of unmitigated transient combustible fire duration and post-fire convective cooling to an elevated temperature erwironment, leads to a less conservativo analysis than originally intended.
The internal panel fire analysis has covered the major elements of the problem. However, we feel that applicability of the " HOTBOX" model to internal cabinet fires should be more fully discussed, particularly in regard to three-dimensional effects and' boundary conditions.
Also, it would be advantageous if there was some experimental evidence to support the modeling.
We are of the impression that the proposed modifications to be employed by NSP in the control room of their Moaticello plant as discussed in the October supplement will in concert lead to a more fire safe area.
Ilowever, based upon the evidence provided and.the results presented, quantification of the degree of added fire protection, af forded is too uncertain.
The structural steel analysis in the October supplement would be more complete if the effects of thermal expansion on stress at restraints and supports had been considered along with the effects of creep at elevated temperatures.
.~.
5.
ACKNOWLEDGEMENT c,
~
The authors wish to express their appreciation to Dr. John Boccio for his suggestions and discussions relating to the fire-modeling methodology employed by NSP in their fire-hazards analysis of the Monticello facility.
~
w 6.
REFERENCES
- 1. Newman, J.S. and Hill, J.P., " Assessment of Exposure Fire Hazards to Cable Trays", EPRI-Np-1675, Electric Power Research Institute, Palo Alto, Ca.,
January 1981.
8
- 2. Alpert, R.L., " Calculation of Response Time of Ceiling-Mounted Fire De-tectors", Fire Technology, Vol. 8,1972, pp.181-195.
- 3. Alpert, R.L., " Turbulent Ceiling det Induced by large-Scale Fires", Com-bustion Science and Technology, Vol. 11, 1975, pp. 197-213.
- 4. You, H.Z. and Faeth, G.M., " Ceiling Heat Transfer During Fire Plume and Fire impingement", Fire and Materials, Vol. 3, No. 3,1979, pp.140-147.
- 5. Alpert, R.L. and Ward, E.J., " Evaluating Unsprinklered Fire Hazards", FMRC J.l. No. 01836.20, facto'ry Mutual Research Corporation, Norwood, Ma.,
August 1982.
- 6. Stavrianidis, P., "The Behavior of Plumes Above Pool Fires", a Thesis presented to the Faculty of the Department of Mechanical Engineering of Northeastern University, Boston, Ma., August 1980.
- 7. Steward, F.R., " Prediction of the Height of Turbulent Diffusion Buoyant Flames", Combustion and Science Tdehnology, Vol. 2,1970, pp. 203-212.
- 8. Hottel, H.C. and Sarofim, A.F., " Radiative Transfer", McGraw-Hill Book Company, New York, 1967.
- 9. Orloff,.L., " Lip Effects in Poo.1 Fires", Paper No. 33, Canadian Section of.
the Combustion Institute, Kingston, Ontario, May 1979.
- 10. Veldman, C.F., et al, "An Experimental Investigation of the Heat Transfer from-a Buoyant Gas Plume to a Horizontal Ceiling, Part 1: Unobstructed Ceiling", NBS-GCR-77-97, June 1975.
- 11. Rohsenow, W.f4., and Har-tnett, J.P., " Handbook of Heat Transfer," Mc-Graw-Hilt-New York,.1973.
- 12. Kreith, F., " Principles of Heat Transfer," Intext Press, Inc., New York, 1973.
- 13. Lee, J.L., "A Study of Damag'eability of Electrical Cables in Simulated Fire Environments," EPgJ NP-1767, Electric Power Research Institute, Palo
~
Alto, Ca., March 1981.~_
. 14. Hockenbury, R.W. and Yeater, M.L., "D'evelopment and Testing of a'ikodel for Fire Potential in Nuclear Power Plants," NUREG/CR-1819, U.S. Nuclear Re-gulatory Commission, Washir.gton, D.C., November 1980.
en.- g g.
0
~,
's G
e a
h um W
e.
e.
e
~
~
1 1
g
,W mary./tr = 5.5 tio"
. **i Baedab4 Alperii%/.Q. f 4
0.1-g
).: 9,y(g<
' SymEco. r-.
(N./nl
\\1 3
\\
+-
- 0.39
+
O.60' 1
O.m
/s O.02
.N O.tS O.0:
c6
'FW Be s tO*C
~
1-23 ?$-tO *0-rCO t."a-2000 e
ue m 0.C01 O
O.1 3
r/M 30
- - ~ -
t
' Figure 1.
Heat flux as a function of position for plume impingement on an unconfined ceiling.
(From Figure 7 of Reference 4) e l
. \\ '-
I t.O
, i j
. i i.i.. '
O,Methenol (0.GI m dio. pen).Q h
- 0.0311 'i ' ', 3 e
I 4.: 4 ee 4 g
}
i t
i l I i I4 1
(
O Me'henol (1.22 m cia. pen) O h
- C.DC H ' i
?
1 I
I I ! IIliI I
. a Meth:not (I.74 m dio. pen) Oh = 0.70 H lI V Hep ene (0.61 m dio. con Oh = 0.5CH 97
{l lll' l
Ch
- 0. CH O'6I t
I l I iei;i I rii i
i i
- I-ii CORREbATiON p
%i !I1!il i ili M OS/F/ dis V:
- 6M.,M/F 4/
9 E
7 N -'
In! l l l **
C e,qRd4 /97 /pA/! ! !
///4 4_
m r..%. - t t.
11\\
o
---m.
I e
=
5 e< @ ~ ' --
.~
N;. ' -
. w Ms : \\!
- 3-i
. -- -, m,._, A.. TN; L L%w:/.%l A$
ii7r
~
...i.
o.s 4
i
.w t
i i
i ii a
x.
, s i. i.,w
. y i
i 4itii l
x i ve i t y, :
o
'!Ifi*
I ll 4
l N
\\
if ;, 8 taJ i
- 1 t
t\\s a./..i..
-i t\\_
- u
-s;
- --t'f g g.I t
! \\ g ),. g g 1
2o i
i.x i \\ ;..:i ii;
.i I
i y.\\1 : i
- a. i I
- I;
.i i
1:
1 i i\\ : \\!. i l i
L i
i i
- )
! III - :! il!
- 3
.I i
i
_1-1 ! l il ! !I, t
l I
'I i
! 1
. !I l'-
i
'l ! !
t-l-
f a -
!.i ]
sl.i l
s s
i iv..:
l
~ 0.001 0.01 SCALED FORCED VE!JT!LATIC;1 RATC.
Yi
= - - _
. H S/2 anl/2.s )
(
Figure 2.
Scaled heat flux versus scaled forced v.entilation rate.
(From Figure 1 of P.eference 1) e 9
1 t
I
-22
e' 4
I l
3-36 Handbook of Heat Transfer 50 e._....
.m _ w
.Sf:$2 f?p
. aco E
,gttd :
~
y Mim ae'*L#6 -
!)[ffp.
..AyjpW.a's*.340%;[
=r-T... ->nemtg7ffffffm wiyyggMn
...u ~"'*4twiliJV f// f fy Fs,
2 2 m
- ~;c 5
~~~'~ ~~:MWsMMnM,tN(Mthl641
- tWAEMN#4#illMilullM11/llVMVc=uMit l
"- 1 7 3 0 fr d.t is / VT.n1.i N jf U /// H / Uf f fs 4 ; r % t h f GiEi:L..A1El$i?lf'LLnMitGilly/1/UN/#///KM!,f ff1
- m=
M! Nil!NAnMHM/H1///
ff/xWkm-mi
,}
WNL VWMHA?hW/// //###XW '^T et L
/AMMiAWAl#MH###/MM//fP
~
~
',;== Mb.17AWMfffA%%ff/f/X&DF DL:-
E7xkMFff///frAV/////Mi:
e/
/
vere; es.......,,::
,o 0000 Goos og oj
.. v.e.. -
<=>
s 3
Figure 3
-~
e e
e S
e 9
.3 9
6 es I.
.