ML19295A630

Smoke Simulation Testing at Monticello Nuclear Generating Plant.

ML19295A630
Person / Time
Site:	Monticello
Issue date:	09/19/1980
From:	Martin T, Neilson C NUTECH ENGINEERS, INC.
To:
Shared Package
ML19295A628	List: ML19295A627 ML19295A630
References
30.2252.0001, NSP-52-001, NSP-52-1, NUDOCS 8010020654
Download: ML19295A630 (38)

Text

. .

NSP-52-001 September 1980 30.2252.0001 SMOKE SIMULATION TESTING AT MONTICELLO NUCLEAR GENERATING PLANT VOLUME I Prepared for:

Northern States Power Company Prepared by:

NUTE :H San Jose, California Prepared by: k. .

C. C. Neilson Reviewed by and 7 .

NUTECH Approval: - 1 T. D. Martin, P.E.

Date: _ h /f

/ /

nutech -

8010020 66*f

REVISION CONTROL SHEET

SUBJECT:

Smoke Simulation Testing at REPORT NUMBER: NSP-52-001 Monticello Nuclear Generating Plant - Volume I C. C. Neilson, Specialist C NAME/ TITLE INITIAL T. D. Martin, Project D.irector NAME/ TITLE INITIAL D. A. Gerber, Sr. Engineer DAG)

NAME/ TITLE . INITIAL NAME/ TITLE INITIAL PRE- ACCURACY CRITERIA PRE- ACCURACY CRITERIA PAGE REV PARED GECK OIECK PAGE REV PARED GECK OECK i 0 d N M/A 3-1 0 N f EO N Cfk 11 0 I 3-2 0 111 0 3-3 0 iv 0 34 0 v 0 3-5 0 1-1 0 3-6 0 l 1-2 0 4-1 0 1-3 0 5-1 0 1-4 0 5-2 0 2-1 0 5-3 0 2-2 0 5-4 0 2-3 0 5-5 0 2-4 0 5-6 0 ,

2-5 0 5-7 0 2-6 0 5-8 0 2-7 0 5-9 0 .

2-8 0 5-10 0 Y 2~9 0 6-1 0 .@My ' ,(

2-10 0 2-11 0 Y p

2-12 0  %{g(d y QEP 001.1 C nutech --

EXECUTIVE

SUMMARY

During the months of May, June, and July, 1980, NUTECH successfully conducted 45 smoke simulation tests in 30 specified fire zones at the Northern States Power Company's Monticello Nuclear Generating Plant. The objectives of the testings were (1) to demonstrate to the Nuclear Regulatory Commission (NRC) that Northern States Power Company (NSP) had satisfied NRC requirements pertaining to insitu smoke detector testing; (2) to verify, where applicable, that currently installed detectors were optimally sited for prompt detection of an incipient fire; and (3) to enhance overall plant safety by recommending effective upgrading of the existing fire protection system, based on test results.

The testing was conducted utilizi portable space heaters to simulate an incipient fire. The hoauers czeate a plume of heated air, or simulated smoke, that is sf -dlar to ti e heated air created by a fire as it is just beginnia .

To monitor movement of the simulated smoke, a very small amount of tracer gas (SF6) is continuously releastd into the plume of hot air from the heaters and serves as a tracer of the smoke movement. This SF6 will be distributed by the ventilation system just as smoke would be distributed. Detectable in extremely low NSP-52-001 11 nutech --

concentrations, such as one part per billion (ppb), the SF 6 provides a means of mapping the concentration profile of the simulated smoke in each area over time. Thus, with the knowledge of how the simulated smoke is distributed, the optimum locations for smoke detectors can be identified. This report presents the results of the testing that demonstrated that NSP has satisfied NRC requirements pertaining to insitu smoke detector testing.

The zones were evalurted based on a four category grading system for the 30 zones tested:

1. Category A corresponded to a zone with all its detectors optimally sited. (18 zones)

2. Category B was a zone meeting the Category A guidelines for its existing detectors; however, additional detectors were recommended for optimal sitting. (5 zones)

3. Category C was reserved for zones not having a detector system installed. (3 zones)

4. Category D included zones with detectors recommended for relocation. (4 zones)

NSP-52-001 lii nutech --

TABLE OF CONTENTS VOLUME I Page EXECUTIVE

SUMMARY

ii LIST OF FIGURES v

1.0 INTRODUCTION

1-1 1.1 History 1-1 1.2 Prototype Test 1-2 1.3 Monticello 1-3 2.0 TEST METHODOLOGY 2-1 2.1 Pre-Test Activities 2-2 2.2 Test Procedures 2-7 2.3 Post-Test Activities 2-9 2.4 Data Reduction 2-10 2.5 Concentration Profiles 2-10 3.0 ANALYSIS 3-1 3.1 Analysis of Normalized Concentration 3-1 Profiles 3.2 Experimental Error 3-2 4.0 RESULTS AND CONCLUSIONS 4-1 5.0 RECOMMENDATIONS 5-1 5.1 Generic Zone Categories 5-1 5.2 Zone Specific Recommendations 5-4 6.0 RECOMMENDATIONS 6-1 NSP-52-001 iv rititeac:li --

LIST OF FIGURES Figure Title Page 2.1 Manifold Testing System 2.12 5.1 Firs Zone Classification by Category 5.10 NSP-52-001 y nutech . .

1.0 INTRODUCTION

1.1 History Following the fire at the Browns Ferry Nuclear Station on March 27, 1975, The NRC initiated a program to investigate the causes of the fire. This investigation resulted in requirements for each licencee to upgrade the fire protection program at their nuclear plants. One phase of this review concentrated on smoke detectors and the smoke detection system at each plant.

Specifically, the NRC questioned how the smoke detectors were positioned to provide early warning.

To answer this question the NRC required that licensees conduct insitu tests to verify the performance of their smoke detection system (Reference 1). This test would demonstrate that a fire would be promptly detected by the installed smoke detection system in each critical space of each nuclear plant.

In 1978, when the Safety Evaluation Reports were issued to plants, there was no acceptable testing methodology that could satisfy this requirement for insitu testing. A test using a real fire within the critical spaces of a plant was not practical, and other suggestions, such as smoke bombs or ventilation mapping, were inadequate or had disadvantages that prohibited their use.

The utilities approached the NRC in July of 1978 and explained NSP-52-001 1-1 nutech --

that they were willing to comply with this requirement, but an acceptable method was not currently available.

1.2 Prototype Test In January 1979, NUTECH proposed a testing methodology to satisfy the insitu testing requirement. The proposed testing would verify the adequacy of the smoke detection system by conducting a smoke simulation test to determine the optimum location of smoke detectors. The proposed testing was presented to the utilities and the NRC. Five utilities contracted NUTECH to conduct a prototype or demonstration test for NRC review and to provide data for evaluation of the testing methodology.

The smoke simulation prototype test was conducted on August 23 and 24, 1979 at the Yankee Rowe Nuclear Plant (Reference 2). A series of seven fire simulation tests were conducted by NUTECH in the switchgear room. The purpose of the testing wass (1) to demonstrate the validity of the testing methodology to satisfy the NRC requirements; and (2) to identify the optimum smoke detector locations for the switchgear room at Yankee Rowe.

Overall, results from the prototype test were encouraging. On commenting to an NRC official, a Brookhaven National Laboratory staff member wrote: "...I find the use of tracer gas, such as sulfur hexafluoride (SF6) in conjunction with electron-capture gas-chromatography an acceptable technique for assessing the NSP-52-002 1-2 nutech - .

/

convection flow-patterns within (or around) complex geometries (Reference 3). Furthermore, he stated, " . . .the NUTECH approach can, in principle, quantitatively determine smoke migration within areas dominated by HVAC systems." (Reference 4).

1.3 Monticello Smoke Simulation Testing began at Monticello on May 13, 1980 and concluded on July 2, 1980. During this time, 45 tests were conducted in 30 fire zones throughout the plant. A list of the zones tested appears in Appendix A. The following sections of this report describe the smoke simulation testing conducted at i Monticello. Volume I includes 2.0 TEST METHODOLOGY This section describes test preparations, test execution, and data collection.

3.0 ANALYSIS This section describes data recording, data reduction, and mathematical analysis.

I i

4.0 RESULTS AND CONCLUSIONS Overall results of the testing are presented. Conclusions

{ are drawn from the analysis.

t A

NSP-52-001 1-3

nutech -

2.0 TZST METHODOLOGY NUTECH's smoke simulation testing. utilizes a portable electric space heater to simulate the convective heat flow of an incipient fire. The heater creates a plume of hot air, which simulates the hot combustion products from a fire at its initial stage. A trace amount of sulphur-hexafluoride (SF6) was continuously released into the plume of hot air to serve as a tracer of the smoNe movement. During the testing, samples are taken throughout the room using 20 cm 3 syringes at predetermined times.

SF 6 is detectable in concentrations as low as one part per billion, and therefore provides a highly sensitive means of mapping smoke distribution. An SF6 gas chromatograph analyzer was used to determine the concentration of SF6 in each syringe sample. This data was then used to generate concentration profiles according to time and location of the sample. These concentration profiles appear in Appendix B for each fire zone test, and provide the basis for conclusions about smoke distribution patterns.

NSP-52-001 2-1 nutech - .

2.1 Pre-Test Accivities 2.1.1 Pre-Test Survey Prior to conducting the smoke simulation tests, the thirty designated fire zones that were to be tested at Monticello were inspected. During the inspections, each room or zone was examined and the following data was noted: (1) room height; (2) beam locations; (3) v,entilation pattern; (4) inlet and exhaust duct locations; (5) existing locations of smoke detectors; and, (6) potential fire sites. Fire zones varied in size and, with the exception of two, were selected based on the Monticello Fire Hazard Analysis (Reference 4). The two zones not previously designated as fire zones were: (1) the Injection Valve Area; and, (2) the Fuel Pool Cooling Area.

Normally, a single room determined a fire zone; however, imaginary boundaries separating two zones were not uncommon.

The dimensions of the largest fire zone tested was 134 feet by 88 feet by 25 feet.

2.1.2 Simulated Fire Locations Simulated fire locations were selected with consideration for location of flammable materials, location of potential ignition sources and apparent air circulation patterns. Simulated fires were assumed to be located in such a way as to accurately anticipate conditions during a postulated fire in any probable fire location.

NSP-52-001 2-2 nutech ..

2.1.3 Sampling Location Selection Proper identification of the smoke distribution pattern required adequate sampling within the test space. A total of 5334 SF6 samples were taken during the tests at Monticello. Depending on the height of the room, sampling locations were designated at High, Medium, Low and Bottom elevations. The letters H, M, L, and B were used in the mapping process to denote these elevations, respectively.

High samples were generally taken close to the ceiling and next to smoke detectors, if installed. The medium sample point usually varied from 6 to 10 feet below the high sample. Low samples ranged from 5 to 10 feet off the floor and bottom samples from 0-5 feet from floor level . Also, L and B samples were taken only in rooms where the ceiling was greater than 20 feet or where

unusual convection patterns might be encountered. Appendix B indicates in greater detail where and why these locations were chosen.

In each test, a sufficient numbac of 20 cm3 capacity syringes were used to collect samples for later analysis which were representative of the SF6 concentration profile in the fire zone. The sample point density varied between 70 and 280 ft2 / sample point. The number of samples taken per unit area varied because some zones required that more sampling be done NSP-52-001 2-3 nutech -

where unusual convection patterns, geometry, or disturbances might be encountered. Furthermore, many mid-level samples were taken to: (1) assist in identifying those zones in which strari-fication occurred; and, (2) confirm general smoke migratory patterns.

2.1.4 Test Equipment Test equipment was set-up in a zone during pre-test activities and consisted of:

A. Sampling Syringes: The em 3 syringes were each labeled with the proper sample location and time.

B. SF 6: Release of SF6 was accomplished by controlled release of the gas at the simulated fire location.

SF6 was selected as a tracer gas because it is stable and non-reactive up to temperatures of 800 C and is detectable in extremely low concentrations ,

such as 1.0 part per billion (ppb). A trace amount of SF6 released into the heated air stream has no efiect on the physical characteristics of the hot air plume. Because of the ease of detection, a very low flow of 1-4cm 3/ minute was used during each test. Thus, for a test period of 25 minutes, only 40 - 80 cm3 og 3p6 was typically released into the room. The basis for the particular release rate NSP-52-001 2-4 nutech -

used was that the concentration of SF6 in the space being tested would not approach the upper limit of detector accuracy during the duration of a test.

C. Heater: One portable space heater was used to generate the hot air plume of the simulated incipient fire.

The heater was rated at 12.5 amps and generated 1500 watts, or 5125 BTU /hr.

The heater was selected to emit energy at a rate approximating that of an incipient fire, or fire at its initial stage. For comparison of fire size, a 1-pound block of cellulosic material burning at a slow rate of 1 lb/hr, has an energy rate of 9,000 BTU /hr. Burning one quart of flammable liquid over a one-hour period will release about 6,000 BTU's in that hour. The output of the heater was suffi-ciently large to generate a plume of hot air buoyant enough to raise the simulated smoke (hot air with SF6) off the floor so that it could be distributed by the ventilation system.

D. Manifold System - To facilitate testing, a manifold sampling system (illustrated in Figure 2.1) was NSP-52-001 2-5 nutech -

utilized. The system was an improvement on the testing process used in the prototype test because it reduced man-power requirements.

Two manifolds were constructed, each receiving 15 sample lines. Sample lines were either 25 or 50 foot long polyethylene tubing and were utilized to sample various remote locations throughout the test space. These sample lines were attached to the ceiling with magnets and then directed to a sampling manifold where the SF6 sample was taken.

A 20cfm vacuum pump was used to maintain flow in the lines and ensure that there was no significant delay time in the sample lines. Discharge from the vacuum pump was directed to a non-affected space. A tee connection in each of the sampling tubes allowed sampling at the manifolds.

Thus, one operator was able to take a total of 15 samples per minute where previously it might have taken 15 men to do the same task individually.

E. Reach Rods: When the zone size was too small to warrant the use of a manifold, samples were taken with reach rods and/or by hand. Reach rods are NSP-52-001 2-6 nutech .

5 feet in length and have an attachment at one end where a syringe can be inserted and the sample taken.

F. SF 6 Counting Equipment: The SF6 Gas Chromatograph detectors were located in a room separate from the test room to facilitate counting and to avoid contamination of the test rooms by the exhaust from the SF6 detectors as samples were counted. The SF6 detectors used at Monticello had a usable range of 1 ppb to 3000 ppb SF 6*

2.1.5 Sample Collection Method After sampling points were selected, the sampling ends of the polyethylene tubing were put in place. The high sampling ends were held in place by a magnet attached to the ceiling and a string attached the tubing to the magnet. Lower height tubing was attached at appropriate heights by tape to the ceiling tubing.

2.2 Test Procedures 2.2.1 Introduction Each test lasted approximately 30 minutes and consisted of taking samples at the designated times after the simulated fire had been activated. The samples were collected and analyzed afterwards on an SF6 Gas Chromatograph analy=er (detector).

NSP-52-001 2-7 nutech .

2.2.2 Start-Up The execution of the test was simple after all the equipment and personnel were in place. The space heater was energized and then both the syringe pump (starting the injection of SF6) and the vacuum pump were started. (The syringe pump could release the SF 6 into the heater at flows as low as 0.5cm3 per minute.)

The syringe pump method of injection represented a marked improvement over the flowmeter and gas cylinder arrangement used in the prototype test conducted at Yankee Rowe due to its superior metering accuracy. Also, before each test, " pre-test" air samples were taken in the zone to verify the background SF6 concentration was less than 5-10 ppb.

2.2.3 Hand-Sampling Method (Small Zones)

At pre-determined times after initiation of the test, a signal was given and individuals took 20cm 3 samples at designated locations. These grab samples were all taken within one minute of the signal and were then set aside and measured for SF 6 content upon completion of each test.

2.2.4 Manifold Sampling Method (Large Zones)

In larger zones where the manifold sampling method was utilized, at least two individuals were present to conduct each test. Like the hand-sampling method, a signal was given at pre-determined NSP-52-001 2-8 nutech

times during the test and 20cm3 samples were taken. However, the sample-taker at the manifold took 15 samples within one minute.

He then removed the full syringes and installed a new set of empty syringes for the next sample time.

2.2.5 Sample Times The sample times were chosen so as to collect adequate data during the initial moments of the test to determine movement and yet samples had to be taken for a long enough period to identify when steady-state was reached. The dispensing of SF 6, r smoke, is influenced by the strongest gradient during the initial stage of a fire. Therefore, the first sample was usually taken one minute after the start of the test. Samples were taken about every 3 minutes for the next 10 minutes and then every 5 minutes until test completion. Long term dispersion and steady-state conditions were normally reached during the 30 minute testing duration.

2.3 Post-Test Activities 2.3.1 Sample Analysis Method Two SF6 Gas Chromatograph Analyzers (hereafter referred to as detectors) were used to analyze the 5334 samples taken. Each sample required approximately 45 seconds to be analyzed in the SF6 detector. By using two chromatographs simultaneously, the time necessary to analyze a large number of samples (i.e., a two-NSP-52-001 2-9 nutech - .

manifold test) was shortened.

Analyzing samples on two detectors necessitates that a correla-tion between the two be made so that a reading from one is directly comparable to a reading from the other. Thus, a system of " cross-checking" approximately every sixth sample was instituted so that a ratio between detector readings could be determined. The cross-check system involved injecting half of every sixth sample into each detector and recording both readings. This is discussed in Appendix C.

2.4 Data Reduction For each test the highest SF6 concentration was used as the denominator to normalize (100 basis points) each data point to the maximum concentration. Thus, the numbers used in the data analysis section of Appendix B are normalized for each test.

The normalization was done to provide a clearer picture of smoke movement. Sample concentrations varied from 1.0 ppb to 3000 ppb (linear response range of the detector) and normalizing to the peak concentration makes the distribution easier to visualize.

The normalized values are the percentage of peak value that occurred during she test.

NSP-52-001 2-10 nutech -

2.5 Concentration Profiles The normalized concentrations for each test were then plotted on six sequential time-frame general arrangement drawings, cor-responding to the zones tested. Each of the six sequential time-frames contained identical drawings of the zone, and represented the consecutive sample times tested. Individual concentration profiles for the 45 tests are contained in Appendix B.

NSP-52-001 2-11 nutech .

M*GNET m SAMPLE SYRINGE

/

TEE SAMPLING CONNECTION

// k:n I

) MANIFOLD u

't n. -

s

+

- VACUUM PUMP EXHAUST TO UNAFFECTED SPACE ND o

- SAMPLE LINE (1 0F 30)

SAMPLE LINE (1 0F 30 - SUSPENDED BY MAGNETS AT VARIOUS HEIGHTS)

SAMPLE

,T. SYRINGE l TEE 3 i r

EXHAUST TO k O UNAFFECTED I

SPACE VACUUM

i. _ _ ._. - - - - J PUMP MANIFOLD MANIFOLD TESTING SYSTEM Figure 2.1 nutech -

NSP-52-001 2.12

3.0 ANALYSIS The normalized concentration profiles were the primary analytical tool used to describe the evolution of a smoke concentration profile in a zone. These profiles assisted in identifying trends, locations of peak concentrations, and steady state conditions. When two or more tests were conducted in the same room, the tests were compared for consistency, reasonableness and reliability. Finally, the test results were examined for possible experimental error and the data was checked via a quality assurance program.

3.1 Analysis of Normalized Concentration Profiles Both qualitative and quantitative analyses were performed using the profiles for each test. Noting locations of peak concentrations for each sample time helped identify smoke dispersion and migration patterns within a zone. Furthermore, where zone geometry was simple (e.g., rectangular or triangular as in zones lA, 1B, 1C, and lE) and smoke distribution was relatively homogeneous , an average concentration for all sample points at a given height was computed. The averages were then compared to identify vertical distribution patterns or concentration layers.

Additionally, each zone's geometry was studied for the effects of beams, concrete overhangs, arrangement of equipment, air intakes NSP-52-001 3-1 nutech .

and exhausts, and other factors that might affect smoke movement. Qualitative analysis primarily was used to support conclusions drawn from geometry effects.

3.2 Experimental Error Four principal categories were identified in which experimental error was introduced into the testing process. However, their overall effects on.the testing were minimal. Experimental error included:

1) Time-lag during multiple sample taking by one individual;

2) Correlation discrepancy between two detectors;

3) Aging of SF6 in syringes;

4) Test methodology 3.2.1 Time-Lag During multiple sample taking by one individual (i.e., manifold sample-taker), a time-lag exists between the first sample taken and the last. For example, the signal to commence sample taking is given at time T. However, the sample-taker requires approxi-mately two seconds per syringe to take a sample. Thus, at time (T), the first sample is taken, at time (T+2 seconds) , the second , and so forth. A manifold sample-taker requires 30 seconds to take all his samples. The experimental error is introduced because samples taken at slightly different times do' NSP-52-001 3-2 nutech -

not absolutely reflect the smoke pattern at a specific instant in time.

The error introduced was not significant because the time delay from first to last sample was neglible in comparison to total test time (30 seconds vs. 30 minutes).

3.2.2 Correlation Discrepancy In Appendix C, SF6 analyzer correlation, a method was given for

, correlating the data from one analyzer with the data from the other. Specifically, a linear regression analysis was performed on " cross-checked" samples to determine the relationship between the two sets of analyzer data. The regression analysis was performed for each of the 21 tests where two analyzers were used. The results of the 21 " cross-checked" tests indicate two sources of experimental error.

(1) Insufficient sample size to be statistically relevant: In some tests, only 3 syringe samples were cross-checked and correlated using linear regression. Statistically, two data points are enough to determine a straight line, but are not enough to determine with sufficient confidence the relationship between two sets of data.

NSP-52-001 3-3 nutech -

(2) Fluctuation of Detector Correlation: The linear equation describing the relationship between the analyzers varied from test to test. Nevertheless, the two errors do not significantly impact the test results for the following reasons :

(a) 14 of the 21 cross-checked tests had 10 or more data points; 1 test relied on 3 data points, and 6 tests relied on 4 to 9 data points.

(b) The correlation factor (r) in all cases was greater than .92 which was satisfactory.

(c) Variance in the linear regression equation has little effect on the individual test results. Each test had its own data evaluated in the linear equation specific to that test. The r-factor in all cases was greater than .92 and indicated the linear equation used in each test was a " good" one.

(d) The normalized concentration profiles for these tests were evidenced by a high degree of reasonableness, consistency, and reliability.

NSP-52-001 3-4 nutech -

(3) Accuracy of SF6 Detectors: SF6 Gas Chromatographs utilize a separation column to segregate the SF6 from other gases present in the sample. The detectors are not calibrated to an absolute source since the column characteristics are not adjustable. This inability to calibrate the two detectors, together with the accuracy of the detectors, could contribute to the error. However, since all SF6 readings are relative, the absolute value of SF6 is not important.

3.2.3 Aging of SFg The aging of SF6 inside syringes (after the samples were taken and before they were analyzed on the detectors) was another possible source of experimental error. This test involved analyzing part of a sample, then reanalyzing the remainder of the sample two hours later. A two hour interval was selected because this is the longest time between sampling and analysis of smoke simulation test samples. No noticeable decay of sample concentration was observed during any of these checks.

3.2.4 Test Methodology Occasionally during a test , disturbances were noted which might introduce error into the controlled experiment. For example, the test conducted in the Battery Room was interrupted by a plant NSP-52-001 3-5 nutech -

worker entering through the closed access door. Hence, the normal ventilation pattern was disturbed. These disturbances are recorded in Appendix B in the fire zone description under

" Comments".

Test results that could have been affected by disturbances to the normal air flow do not show any significant deviation from other test results in the same space.

NSP-52-001 3-6 nutech .

4.0 RESULTS AND CONCLUSIONS The Smoke Simulation Testing was successful in accomplishing its two primary objectives. First, it verified in T3 of 30 zones that present detector locations were adequate. Second, the testing provided empirical results from which recommendations were made for upgrading 11 of the existing fire protection systems.

In general, the results were consistent, reasonable, and reliable. In some zones more than one test was conducted; and in most cases remarkably similar results were obtained. In many tests, the smoke pattern could be intuitively explained without a concentration profile; however, there was always agreement between test results and engineering judgement.

Smoke Simulation Testing proved that the tracer gas technique for the study of smoke movement is feasible. Many of the problems experienced during the Prototype Testing at Yankee Rowe were corrected before the Monticello testing began. In particular, the SF6 s urce as well as the heater output were made constant.

Manpower requirements were reduced significantly with the introduction of the manifold system.

NSP-52-001 4-1 nutech -

5.0 RECOMMENDATIONS 5.1 Generic Zone Categories The recommendations concerning smoke detector locations at Monticello are based on a four-category grouping system. The four categories are:

Category A - Existing smoke detector locations are adequate.

No changes are recommended.

Category B - Existing smoke detector locations are adequate.

However, additional detectors are recommended.

I Category C - No smoke detectors are installed in this zone.

Locations for detectors are recommended.

Category D - Existing smoke detector locations are inadequate.

NUTECH recommends relocation or addition of detectors.

Selection of the optimum smoke detector locations is largely an engineering judgmental decision; however, a few basic guidelines were applied. The guidelines for the classification process were:

NSP-52-001 5-1 nutech -

5.1.1 Category A High concentrations of SF6 (sm ke) were registered near the detectors. Furthermore, in areas where no smoke detectors were nearby, lower concentrations were registered. In each case, the addition of another detector would not significantly enhance olant <sfety.

5.1.2 Category B The detectors located in a B zone were optimally sited according to the guidelines established for Category A. However, addi-tional detectors were recommended in locations registering high concentrations of SF 6 , but lacking a detector. Any localized accumulation of smoke not registered at nearby detector locations established the need for additional smoke detectors. Therefore, the addition of one or more detectors would significantly enhanca the safety of the current system.

5.1.3 Category C Three zones did not have an installea smoke detector system. In each of these areas, guidelines described under Cateogory A were used to determine optimal smoke detector locations. Specifi-cally, consistent and rapid increases of SF6 concentrations relative to other sample points confirmed that a particular location (e.g., 10 feet below the ceiling in the northwest corner) was optimal for prompt detection of an incipient fire.

Room geometry and layout were particularly important in the NSP-52-001 5-2 nutech - .

decision to locate detectors. Clearly, the addition of a detector in any one of these zones would enhance overall plant safety through decreased fire detection response time.

5.1.4 Category D In some zones, conditions warranted the relocation of currently installed detectors to more optimal locations. For example, in zones where the ambient temperature exceeded 100 F, SF6 (sm ke) failed to reach the ceiling. A stratified layer of gas resulted. Detectors on the ceiling would not be optimal in terms of response time. Thus, the recommendation was to relocate the detectors at some specified distance below their current location. The alternative locations significantly enhanced the plant safety because response time would be lessened.

Of the 30 zones tested, 19 were Category A, four were Category B, three were Category C, and four were Category D. (See Figure 5 .1. ) Thus, 19 zones had adequately sited detectors while another 8 either required additional detec.cors or a change from present detector location. A brief description of each zone follows ind' esting the appropriate Appendix page number and category.

NSP-52-001 5-3 nutech -

ZONE 2A - TIP Drive Room - Appendix B-37 Category A: The present locations of the 2 smoke detectors are adequate.

ZONE 2B - CRD Hydraulic Control Area - Appendix B-41 Category A: The present locations of the eleven smoke detectors are adequate.

ZONE 2C - CRD Hydraulic Control Unit and HVAC Areas - Appendix B-49 Category A: The present locations of the twelve smoke detectors are adequate.

ZONE 3B - MCC & Standby Liquid Control System Area - Appendix B-61 Category A: The present locations of the three smoke detectors are adequate.

ZONE 3C - MCC Area - Appendix B-71 Category A: The present locations of the six smoke detectors are adequate.

ZONE 3D - Cooling Water Pump and Chiller Area - Appendix B-79 Category D: The addition of one smoke detector six feet below the ceiling and directly above the cooling water pumps would enhance the fire protection system. Furthermore, the recommended NSP-52-001 5-5 nutech --

locations for the four smoke detectors already installed would be six to ten feet below their present locations.

ZONE 4A - Equipment Hatch Area - Appendix B-86 Category A: The present locations of the six smoke detectors are adequate.

ZONE 4D - Standby Gas Treatment System Room - Appendix B-97 Category A: The present locations of the 3 smoke detectors are adequate.

ZONE 5A - Laydown and Decontamination Area - Appendix B-103 Category A: The present locations of the eight smoke detectors are adequate.

ZONE SB - Contaminated Equipment Storage Area - Appendix B-111 Category A: The present locations of the four smoke detectors are adequate.

ZONE 6 - Refueling Floor - Appendix B-115 Category A: The present locations of the six smoke detectors are adequate.

ZONE 7B - Battery Room - Appendix B-120 Category A: The present location of the one smoke detector is adequate.

NSP-52-001 5-6 nutech 1.

ZONE 8 - Cable Spreading Room - Appendix B-123 Category A: The present locations of the eignt smoke detectors are adequate. However, the relocation of one of the eignt smoke detectors to a position near the air supply would provide additional support for the fire protection system.

ZONE 12A - Turbine Building Load Center No. 1 - Appendix B-130 Category B: The addition of one smoke detector in the southeast corner of the zone would enhance plant safety.

ZONE 13C - ESF Motor Control Center - Appendix B-135 Category A: The present locations of the two smoke' detectors are adequate.

ZONE 14A - Turbine Building Load Center No. 2 - Appendix B-139 Category A: The present location of the three smoke detectors are adequate.

ZONE 16 - Turbine Building Corridor - Appendix B-143 Category A: The present locations of the four smoke detectors are adequate.

ZONE 17 - Turbine Building Corridor - Appendix B-147 Category A: The present locations of the four smoke detectors are adequate.

NSP-52-001 5-7 nutech - .

ZONE 19A - Water Treatment Area - Appendix B-151 Category B: The addition of two smoke detectors located near the turbine operating floor opening would enhance plant safety.

ZONE 19B - ESF Motor Control Center - Appendix B-156 Category A: The present locations of the two smoke detectors are adequate.

ZONE 19C - Pipe and Cable Tray Penetration Area - Appendix B-160 Category A: The present location of the one smoke detector is adequate.

ZONE 23A - Intake Structure Pump Room - Appendix B-164 Category B: An additional detector is recommended for the east wing of the zone.

ZONE N.A. - Fuel Pool Cooling Area - Appendix B-173 Category C: Two schemes for locating detectors in this :one are offered: (1) one detector located on the ceiling near the hatchway (middle of room); or (2) two detectors, one at the north end of the room, on the ceiling and the other located as in scheme #1.

NSP-52-001 5-8 NUTECH

ZONE N.A. - Injection Valve Area - Appendix B-180 Category C: A two detector scheme is recommended: one ceiling level smoke detector positioned on the west side, and the other smoke detector located near the torus access.

NSP-52-001 5-9 nutech .

FIGURE 5-1 FIRE ZONE CLASSIFICATION BY CATEGORY SMOKE DETECTORS No. Currently Additional No. No. Recommended Installed Recommended For Relocation CATEGORY A Zone 1-E 3 0 0 Zone 2-A 2 0 0 Zone 2-B 11 0 0 Zone 2-C 12 0 0 Zone 3-B 3 0 0 one 3-C 6 0 0 Zone 4-A 5 0 0 Zone 4-B 6 0 0 Zone 4-D 3 0 0 Zone 5-A 8 0 0 Zone 5-B 4 0 0 Zone 6 6 0 0 Zone 7-B 1 0 0 Zone 13-C 2 0 0 Zone 14-A 3 0 0 Zone 16 4 0 0 Zone 17 4 0 0 Cone 19B 2 0 0 Cone 19-C 1 0 0 CATEGORY B Zone 8 8 0 1 Zone 12-A 3 1 0 Zone 19-A 4 2 0 Zone 23-A 3 1 0 CATEGORY C Injection Valve Area 0 2 0 Fuel Pool Cooling Area 0 1-2 0 Zone 1-F (Torus Area) 0 8 0 CATEGORY D

one 1-A 2 0 2 Jane 1-B 2 0 2 Zone 1-C 2 0 2 Zone 3-D 4 1 4 NSP-52-001 5-10 NUTECH

1. =

6.0 REFERENCES

1. Fire Protection Action Plan, NUREG 0298, Vol 1, No.

4, pp. 1-38, December 4, 1978.

2. NUTECH Report SPP-01-001, Smoke Simulation Prototype Test Conducted at Yankee Rowe Nuclear Plant, October 3, 1979.

3. Letter to Mr. Robert L. Ferguson, Plant Systems Branch, USNRC, Washington, D.C. 20555, dated November 3L, 1979 from Mr. John Boccio, Reactor Engineering snalysis, Brookhaven National Laboratories, Tipton, New York.

4. Fire Hazards Analysis, Northern States Power Company Monticello Nuclear Generating Plant: Docket No.

50-263, License No. DPR-22; prepared by Bechtel Power Corporation, March 11, 1977.

NSP-52-001 6-1 nutech -