ML20076H871
| ML20076H871 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 08/31/1983 |
| From: | Yao C FACTORY MUTUAL RESEARCH CORP. |
| To: | |
| Shared Package | |
| ML20076H863 | List: |
| References | |
| PROC-830831, NUDOCS 8309060232 | |
| Download: ML20076H871 (21) | |
Text
. -
l l
t
\
i )
i
. l
' TEST PLANS TO DEMONSTRATE HYDROGEN COMBUSTION INITIATED BY A TAYCO IGNITER IN A SPRAY ENVIRONMENT
~ .
i Submitted to:
Tennessee Valley Authority Division.of Engineering Design 400 West Summit Hill Drive .
Knoxville, Tennessee 37902 Under Contract TV61383A FMRC J.I. OJ1RS.RU 11 5 August'1983 Approved by:
W Cheng Yao, /
> ' Ym
/
Asst. Vice President & Manager Applied Research Department Factory Mutual Research 1151 Boston Provdence Tumphe Mrwood Massectusetts 02062 8309060232 830831 l PDR ADOCK 05000327 P PDR
~
FACTORY MUTUAL RESEARCH CORPORATION OJ1RS.RU ABSTRACT 5
This report describes project plans for a series of small-scale tests to demonstrate that the Tayco hydrogen igniter will effectively initiate combustion
- in the presence of a water spray representative of the spray in the upper com-partment of the Sequoyah containment. The test facility is essentially a larger version of the facility previously used by FMRC to determine the effects of water fog on hydrogen flansnability. The facility will contain redundant, in-dependent instrumentation to measure and control tes' parameters and to verify successful ignitions.
The following nine parameters have been considered in preparing preliminary test matrices: 1) water flux, 2) hydrogen concentration, 3) drop size, 4) ig- -
niter voltage, 5) water temperature, 6) gas temperature, 7) gas velocity,
- 8) water vapor concentration, and 9) drop velocity. The last four parameters will not be varied in the test matrices because 1) they do not significantly affect igniter temperature and gas mixture,ignitability, or 2) they are dependent on one of the other test parameters. Water flux and hydrogen concentration.will be varied in the ignitability test matrix, but drop size, water temperature, and igniter voltage will be varied only if preliminary screening tests indicate that they have a significant effect on igniter surface temperature. I preliminary decision tree has been constructed to show how ignitability test parameter values will be selected, assuming that only water flux, hydrogen concentration and igniter voltage need to be varied.
1
.g.s
- - - _ .---,,,,-,p. .-,y,. , _ , _.y, . . - , , , , - --
,-,,,--.g- .- - , - . - . - .
FACTORY MUTUAL Rt1EARCH CORPORATION 0J1R5.RU TABLE OF CONTENTS Title Pm Section i ABSTRACT .
1 I TEST OBJECTIVE AND APPROACH 2
II FACILITY DESIGN 2 2.1 Experimental Arrangement 5
g
~
2.2 Instrumentati5n '
6 2.3 Test Procedures 7 2.4 Data Reporting 8
III TEST MATRICES 8 3.1 Relevant Parameters 10 3.2 Igniter Surface Temperature Tests 12 3.3 Ignitability Testa 16 IV PROGRAM SCHEDULE 18 REFERENCES 0
4 e
11
FACTORY MUTUAL RESEARCH CORPORATION OJ1R5.RU
, I l 1
TEST OBJECTIVE AND APPROACH l The objective of the test plans described in this report is to demonstrate that the Tayco igniter will initiate hydrogen combustion in a water spray en-vironment similar to that in the upper compartment of the containment building in the Sequoyah Nuclear Plant. -These tests are intended to satisfy the Nuclear Regulatory Connaission (NRC) 'conditionC1) for the Tennessee Valley Authority (TVA) operating license for Sequoyah. -
The TVA approach to demonstrate igniter effectiveness is to conduct a series of lean hydrogen-air mixture combustion tests with an unshielded Tayco igniter directly exposed to water fluxes similar to those expected in Sequoyah.
This is a highly conservative test arrangement, since the igniters in the Sequoyah plant are outfittad with spray shields to divert most of the water spray around the igniters. The spray flux impinging on the unshielded igniter will be the primary parameter to be varied in these tests. Spray fluxes in the tests will be compared to the Sequoyah upper compartment average flux of 0.915 gpm/ft2 corresponding to 9500 gym (from both trains of the containnent spray system) in the 10,400 ft2upper compartment cross section. If ignition occurs with water fluxes of this magnitude.or only slightly less, the igniter affactiveness will be demonstrated.
TVA confidence in this approach is based on separate effects taisting prev'iously reported (1' ) to the NRC. These tests indicated that a 0.92 gpm?ft spray flux would cool the igniter so as to reduce its surface temperature from about 1715'F dry to about 1275'F at full spray exposure. The 1275'F surface temperature obtained with the water spray is higher than the 1160*F-1200*F ig-i niter surface temperatures reported ( ) to ignite relatively dry hydrogen-air
. mixtures (water vapor concentrations less than about 15 vol% for turbulent gas mixtures). Thus, relatively dry hydrogen-air mixtures, such as might be re-leased from the ice condenser into the Sequoyah upper compartment during postu-laced accidents, should be ignitable even when the Tayco igniter is exposed to these large water fluxes.
The test facility and instrumentation planned for these igniter demonstra-tion tests are described in Section II. Test parameter values and a tentative test matrix (for screening tests) and a decision tree (for combustion tests) are presented in Section III.
1 I
, 1 .
FACTORY MUTUAL RESEAACH COOPORADON OJ1RS.RU II FACILITY DESIGN 2.1 EXPERIMENTAL ARRANGEMENT Conceptually, the test facility is a larger version of the vertical flam-mability tube previously used(3) by FMRC to determine the af'fects of water fog A schematic drawing of the test on the ignitability of hydrogen-air mixtures.
enclosure and auxiliary equipment is shown in Figure 2-1. 'The size of the test d
' enclosure is based on the constraints,that: 1) the spray envelope at the ig- .
nicer elevation roughly fill the enclosure without significant wall inter-f arance; 2) ambient air intrusion into the enclosure be negligible during testing;
- 3) enclosure walls be transparent; and 4) fabrication time and costs be minimized.
These constraints and the spray nozzle described below have led IMRC to design a steel framed Lexan enclosure 3 ft x 3 ft in cross section and 5 ft long.
A SPRACO 1116-1317 full-cone' spray nozzle has been selected for most of these tests because it produces a volume mean drop size (350 micro meters at 20 psig) close to that (370 pm) of the SPRACO 1713A hellow cone nozzle used in .
Sequoyah. The SPRACO 1116 generates a relatively uniform narrow cone spray (64* cone angle) as opposed to the bimodal spray from the 1713A hollow cone nozzle. The S-spm nozzle flow rate at 15 psig corresponds to an enclosure cross section average flux of 0.89 gpm/ft , which is close to the average Possible drop-size effects due to agglomeration in spray flux at Sequoyah.
Sequoyah will be investigated by conducting some tests with SPRACO full-cone ~
nozzle No. 4824-1213, which produces an 850 micrometer volume mean drop size at the same 8-gpm flow rate. Water spray instrumentation described in Section 2.2 will be used to determine drop size and water flux distributions in the cast enclosure.
The water supply for the nozzles will be pumped at a controlled rate from an 82-gal electric water heater. Water collected at the bottom of the test en-closure will be automatically recycled to the heater in order to allow for a relatively long test period.
Hydrogen-air mixtures, heated to the spray temperature, will enter the top of the test enclosure through a four-arm gas distribution manifold. Hydrogen concentrations will be in the range 4-8 vo1% and mixture flow rates will be 2
9
FACTORY MUTUAL RE SEARCH CORPORATION OJ1RS.RU
^
3o ,
- @l g
d O) i E :
gl :..@
- @ rw b@'"l '
,g ,g?\$9 ' @ @ '0 2
,illv&N
& M (M 4M l @
,' @\'s'
' "' , , i
,d 's\ lsNCa@
\ \'rl
~ '
i h@ h@ , ' ,t i
@\ \, . H \ l'/ bo @
)j
/
/ g s :
h h
~
.l
@K') n li ig M AIR '
BLOWER H2 3 g i i w ll n2 i h d iI % @
g& ll
= L" :::::::::::::::: u 9
Figure 2-1 Hydrogen Igniter Test - Preliminary Schematic 3
FACTORY Mutual Re5EARCH CORPORAnON OJ1R5.RU FIGURE 2-1 (LIST OF COMPONENTS) e
~
Descript1on_
Ites No.
3'x3'x5' Lexan Test Enclosure SPRACb 1116-1317 Full Cone Nozzle Tayco Igniter (120-130 VAC) h Rocameter
! Heat Tracing Sections h
h Hydrogen Air Mixer h Flash' Arrestors h Water Flow Meter h Filter Thermocouple with spray shield h Pressure Transducer (0-15 psig Data Instruments) h Gas Sampling Line h
h 3-Way Solenoid Valve h Oxygen Analyzer (Servomex Type A250) '
h Hydrogen Anslyzer (Backman Model 7C) h Water Pump Water Heating Tank (82 gal) h h Vent Disks
^
h Gas Sampling Pump Rubber Flaps O
II Igniter Surface Thermocouples (2) h 0-290 VAC Variable Transformer h Float Switch for Water Return Control h check Valve Items Not Shown
- PMS Drop Size Probe
- Water Flux Collection Array
- Data Recorders
- Temp Controller (Onega Model 6102)
- Power Supplies
- Signal Conditioning Amplifiers 4
- _. _ _ , , , ,, _ . , _ - _ _ . _ - . _ . . ... . _ _ _ _ _ , . ~ ~ _ _ . , , ,w__-_--_ . _
~
l i
FACTORY MUTUAL RESEARCH CORPORADON OJ1R5.RU 15-60 cfm, corresponding to average velocities in the test enclosure of l 0.03-0.11 ft/s. The gas mixture will flow over the Tayco igniter and out the
~
funnel-shaped bottom section of the enclosure a1ong with the water spray.
Detailed drawings of the enclosure and the gas and water supply systems and instrumentation are available' from MfRC.
2.2 INSTRUMENTATION The tests will be sufficiently instrumented to eliminate any doubt as to gas mixture, igniter, and spray conditions and a successful ignition. For the gas mixture, temperatures, flow rates, and compositions will be measured. Cas temperatures will be measured with chromel-alumel thermocouples in 'the test enclosure and the temperature controllers installed in the gas feed lines.
Flow rates will be measured with rotameters in the air and hydrogen lines.
- Hydrogen concentrations will be estimated by the ratio of flow rates and will be checked by a hydrogen analyzer sampling from the top of the enclosure.
Igniter surface temperature will be measured with a type K (chromel-alumal) thermocouple spot welded to the underside of the igniter coil. This is the same technique used by AECL( } for their igniter temperse;r: measurements. Two thermocouples will be used in parameter screening tests (Section 3.1), while only one thermocouple will be needed for the gas ignition tests.
Water spray measurements will include drop size, drop velocity, and water flux distributions. Drop size and velocit'y distributions will be obtained with the' Particle Measuring Systems (PMS) optical array imaging probe. The PMS' probe available at FMRC is designed to sample, count, size, and statistically analyze i
veter drops in the size range 100-1,000 pm. According to SPRACO, 90% of the
. drops from the 1713A spray nozzle are in that size range.
f Although approximate estimates of water fluxes can be inferred from the PMS probe data, a more reliable and direct measurement is planned for these tests. FMRC will fabricate an array of water collection troughs each approxi-mately the same size as the igniter. This array will be inserted into the test
, enclosure prior to running surface temperature and ignitions tests. Water tloving for a specified period of time will be collected in graduated cylinders connected to the trough array.
4 5
i
, . . . - - , . - , --. .- _ _ - - - - - , , _ , - - - , - . - - - . . - _ . - . . - - - - ~ , - - - . . _ . _ . -- - - , - - - . .
FACTORY MUTUAL RESEARCH CORPORADON OJ1RS.RU Combustion tests will include three independent measurements to determiss whether the gas mixture is ignited:
- 1) The response of three spray-shielded 0.010 in. thermocouples (two abo detect either upward or down-the igniter and one below it) will be monitored to This was the primary technique used in the DfRC water ward flame propagation.
fog tests ( }. g 0-5 volt output)
- 2) A pressure transducer (Data Instruments, 0-15 psi.
mounted on the side wall of Ehe test enclosure will be used lto measure re. any transient overpressure associated with flame prop'agation in the enc osu Threshold values for theimocouple and pressure transducer response i to desig ignition will be determined based on checkout tests with dry hydrogen-air m Slight pressure increases should also cause deployment of the vent disks top of the enclosure.
3)
Cas will be sampled from the lower section of the enclosure and input The former is a Beckman thermal conductivity co hydrogen and oxygen analyzers.
l range, and the latter is a Servomex paramagnetic analyzer with 0-10% H2 range. A decrease in both hydrogen concentration Type OA-250 with a 0-25% 0 2 f com-and oxygen concentration is an indication of ignition and the extent o A decrease in hydrogen concentration accompanied by an increase, or bustion. ignition, negligible change, in oxygen concentration does not necessarily imply because it might be due to air entry through the bottom of the test enclosure.
Air leakage should not be an insurmountable' problem with the large gas and wa
' flow rates and the flexible plastic flaps installed on the bottom of the enc'losu Data from the thermocouples, the pressure transducer, and the gas analyzers will be digitized and recorded on a computer-based data acquisition system-Cas concentration comparisons for ig-incorporating a DEC LSI-11 computer. Auxiliary data, nitability determinations will be performed on the computer. k such as gas and water flow races,will also be stored in the computer data ban .
2.3 TEST PROCEDURES Water spray characterization tests will entail removing the igniter assembly from the test enclosure and then inserting first the FMS drop-size probe, and PMS probe data will provide drop-size then the water flux collection array.
and velocity distribution data, while the water collection array will provide i 6 .
5 i
- - - , _ - ~ . - , . . _ - , . . - . - - - _ _ _ _ _ . - _ . _ -
I FACTORY MUTUAL RESEARCH CORPORATION 0J1R5.RU water flux data at the igniter location (at the center of the test enclosure) and at several other locations in the enclosure. ,
Igniter surface temperature tests will be conducted by first establishing the specified gas and water flow rates and temperatures and then actuating the igniter and allowing the temperature to reach equilibrium. Two thermocouples will be used in order to reduce potential problems with bead detaching and/or water impingement onto the head.
Ignitability tests will also be initiated by first obtaining the desired water flow rate and temperature and then actuating the igniter. Once the ig-niter temperature is stabilized, the gas mixture will be established by first 4
flowing air and then gradually increasing the hydrogen flow rate. The gas flow will be maintained until ignition is confirmed or until the test enclosure has been completely purged with the gas mixture and the hydrogen concentration is stabilized.
2.4 DATA REPORTING Water spray drop-size distributions will be reported in the form of dis- .
tribution plots along with the standard statistical parameters such as volume mean diameter and standard deviation. Water flux data will be reported in terms of flux at the igniter, cross-section average flux, and the standard deviation among the flux measurements throughout the cross section.
Ignitability test data from the three ignitability determination measure-ment's (Section 2.2) will be compared. If the three methods all agree, the t:est results will b( obrious. If the three methods produce ambiguous data, the test result may or may not be classified as ignitable depending on instrument re-sponse during the dry check-out tests.
~
7
,-,,-.n- - - - . --,,,-,.e, - - -
, - - - - - - _, .w, , - - , - - - , - - - - - - - - - - , , , a nr. . -- - - - ., , . - . , - - , -
PACTony MUTUAL RESEARCH CORPORATION OJ115.RU TII TEST MATRICES 3.1 RELEVANT PARAMETERS Nine different test parameters have been considered in developing test matrices. These parameters are: 1) water flux, 2) hydrogen concentration,
- 3) drop size, 4) igniter voltage, 5) water temperature, 6) gas temperature,
- 7) gas flow rate, 8) water vapor concent. ration, and 9) drop velocity. Our
. approach has been to reduce this list to a more manageable number by eliminating variables which .cma be shown either 1) to have only an insignificant effect on ,
hydrogen mixture ignitability and igniter surface temperature, or 2) to be dependent on one of the other test parameters.
Some of the parameters can be eliminated from considerations of their ex-pacted effects during hypothesized burns in the Sequoyah upper compartment.
Cas flow rate; i.e., gas velocity in the test enclosure, is one such parameter.
Gaa velocities anticipated from the outlet of the upper plenum of the ice con-denser in Sequoyah are expected.co be about 1 ft/s (D. Renfro, TVA personal ,
communicatica June 1983). In the absence of water spray, smaller gas velocities would be expected at igniter locations on the upper compartment walls. Since spray droplet velocities are much larger (terminal velocities on the order of 10 ft/s for the SPRACO 1713A nozzle droplets), water spray flow rate and drop size will actually determine entrained air flow rates which in turn will dic-tate.the effective turbulent gas velocity in the vicinity of the igniter.
Therefore, the nominal gas velocity through the test enclosure (on the order of 0.1 ft/s) will be irrelevant and need not be included in the test matrix.
Cas temperature is also expected to have only a minor effect on ignic-ability and ignites surface temperature. Temperatura effect data for dry mix-ture dovr. ward propagation limits are shown in Figure 3-1. The expected tem-perature range (between burns) for Sequoyah upper compartment hypothesized i accident conditions, as simulated by CLASIX calculations, is 120-130*F
! (41*-55'C); i.e.,on the order of the estimated water spray temperatures.
Over this narrow range, Figure 3-1 indicates a negligible change in the lower hydrogen flammability limit. In fact, gas temperatures would have to vary by about 60*C (108'F) to produce a 1 vol% change in the hydrogen lower limit l
8 i
l >
1 I
L
. l- l
~
FACTORY MUTUAL RESEARCH CORPOR AllON OJ1RS. RU i i I '
l l_ l
- T l\
(l I
.!' \
'/ l
/ ' I I am
\l
.) , f; i c.$
.\ t/ \' I
?,f I
\ \ \ Y \ \ l cc .j, \ . , ,1 l
$2" 3 / I I1 I ei
z 3 f, ' i Y il 1 1 1*
\ f -
i - ( / ) l l
\ +
/- l l
*j , 1 11 iI 75 80 85 5 10 i 70 HYDROGEN, PERCENT Figure 3-1 Influence of Temperature on Downward Propagation Limits of Flaasnability of Hydrogen in Air (from ref 5) 9 I
j
3 FACTOny MUTUAL RESEAacH ConPORATION OJ1R'.RU 5 concentration. Since expected mas' temperature variations are 1) much smaller than 60*C, and 2) determined by water spray temperatures, there is no need to include gas temperature variations in the test ma'trix. Water vapor concentration is another parameter that can be bounded from accident scenario considerations. If the maximum spray / gas temperature is 130*F, the corresponding maximum water vapor saturation concentration (for a one atmos-phere total pressure) is about 15%. The effect of steam concentration on hy-drogen ignitability limits can assessed from the AECL data shown in Figure 3-2.
. In particular, water vapor concentrations less than about 15% do not appreciably affect the hydrogen ignitability limic- for turbul'ent gas mixtures. By the same token, igniter surface temperature should also be insensitive to these water vapor concentrations because the thermal conductivity of water vapor at 130*F is very close to that of air. Therefore, water vapor concentration variability can also be eliminated as a test parameter; its effective value in the tests will be close to the saturation concentration at the test temperature.
Drop velocity effects are also believed to be relatively insignificant compared to the other experimental parameters. The drop velocity anticipated to impinge on the igniter in the test enclosure should be only slightly higher than the terminal velocity . Thus, it should be a reasonable, or slightly con-servative, simulation of the droplet impingement velocity for the Sequoyah ig-nicer. Drop velocities in the test enclosure will be measured with the PMS drop size probe, but drop velocities cannot be varied independently of water flux and drop size.
'Of the remaining five parameters, only one (hydrogen concentration), refers to the inherent ignitability of the gas mixture itself. The other four parame-ters (water flux, drop size, water temperature, and igniter voltage) are only important insofar as they may significantly affect igniter surface temperature.
Therefore, their importance can be ascertained frem noncombustion tests in which only the igniter temperature is measured. A tentative cast matrix for these screening tests is described in Sectfon 3.2. 3.2 IGNITER SURFACE TEMPERATURE TESTS These tests are designed to determine the influence of the following j parameter ranges on the unshielded igniter surface temperature. l
- l The initial drop velocity for the SPRACO 1116 nozzle operated at 8 gal / min
. is estimated to be only a few feet per second higher than the terminal velocity.
10
=
,5v--.-- v --. ,__.-,e, --.-.*y, - - - -
FACTORY MUTUAL RESE A RCH CORPOR A flON OJ1R5.RU 0 50
; 4 a 6 .t c 6e
.' Ita8Eth4LLT Isp CAS . OPEAAT1sC Istm 1Gulftp 9C31710s (Inng1710u, MULTact e o O stane sa vests e 5 0 static is vetts y 7 9 runsatyr la wors Pe0805tB LIMITS F0e stagge p ITUEES
- - . - . PSCPOSt3 List!TS FOR *L9taatET 90 27112.5
\.
i' 1-r -
*r a a er +6 I IL i
e ** i a e s-1 E / - g n o +..- >.
. . ,e ,
>w, , , /,...,. "7 s - ,, ,. .- .
t
./ -
n ' to n As*
. E 3g* 'A C**
.d o
E
" m- /
- 5 0" ** -
au '5" n n
- 3) a y Y,. Mr. -
E 5 13n.n,,,,. - .3 y Yn y / , py ** - 23 4 " . .
--T, * """~ .
' - &n ~
I t t i f O 10 20 30 40 50 60 Silms C0%Ctsf aaf tch 8 2 87 90LUPE ) Figure 3-2 De ignitien Limits of Hydrogen / Air / Steam Mixtures Using a GM AC Model No. 7 n ermal Glow Plug Located at the Center of a 17-Liter Quasi-Spherical Vessel (from ref. 4) 11
PACTORY MUTUAL sESEARCH CORPORATON OJ1R5.RU 2 0 - 1.8 gpa/ft Water Flux Igniter Voltage 120 - 130 volts . Drop Size 380 - 850 um Water Temperature Ambient (70-75) - 130*F TVA data reported in References 1?and 2 are already available for the effects of water flux and igniter voltage. Water flux has 2 been shown to be a critical factor. For example, a water flux of 0.92 gym /ft reduced the igniter surface tamperature from about 17007F in dry air to about.1275'F with one specific spray nozzle. Igniter voltage variations produced a less dramatic change: The test matrix shown in from about 1760*F at 120v to about 1880*F at 13Ov. Table 3-1 for the planned tests includes only a few confirmatory tests (Tests 2 and 3) with water flux and igniter voltage variations. Drop-size effects will be investigated in the screening tests by replacing SPRACO nozzle 1116, which is to be used in the first four tests, with SPRACO nozzle 4824. Operation of SPRACO nozzle 4824 at 30 psig (Test 5 in Table 3-1) should produce the same water flux as nozzle 1116 at 15 psig, (Test 2), but the volume average drop size should increase from 385 micrometers to about 850 micrometers. The FMRC PMS probe will be used to measure drop sizes duringn se two tests, and the corresponding pair of tests (Tests 6 and'7, these with a water flux of 1.8 gpm/f t ) . If igniter temperatures in either pair of tests differ by more than 50*F, both nozzles will be included in the ignitability test matrix. -
- Water temperature effects will be determined by comparing the results of Tests 2 and 4, which bave water temperatures of 75'F and 130*F, respectively.
As with drop size, if these two tests result in igniter temperatures that dif-for by more than 50*F, water temperature effects will be included in the ig-nicability test matrix. 3.3 IGNITABILITY TESTS Since ignitability demonstration test conditions depend in part upon the i results of the igniter surface temperature casts, a firm test matrix cannot be l presented for the ignitability tests. TVA instructions for ignitability test conditions are to choose a conservative set of initial parameter values, and then to make changes as dictated by the results of the preceding tests. The 12 . I 1
FACTORY MUTUAL RESEARCH CO2PORATION OJ1R5.RU TADLE 3-1 IGNITER SURFACE TENPERATURE PRELIMINARY TEST MATRIX SPRACO Water Temp Igniter WaterF}ux Voltage
. Test Nor.zle , (gpm/ft ) (*F) 1116 0 .- 120 ' 1 '
1116 0.9 75* 120 2 1116 0.9 75 130 3 1116 0.9 130 120 4 4824 0.9 75 120 5 4824 1.8 75 120 6 1116 1.8 75 120 7 Initial tests will be run with water at near ambient temperatures, expected to be 65-75'F. . O A
'9 i
I l 1 I 13 s' I
FACTORY MUfuAL atSEARCH CORPORATION OJ1RS.RU rationale is to merely demonstrate ignitability under realistic-to-conservative h test conditions without obtaining a complete set of ignitability data over r. e entire range of experimental parameters. In keeping with this test plan rationale, FMRC and TVA have prepared the preliminary decision tree shown in Figure 3-3 to specify test condition con-singencies. Three parameters; water flux, hydrogen concentration, and ig-We assume,'for purposes nicer voltage, have been specified for each test. of this report, that the other two parameters: drop size and water camperature, .' have been shown to have neg"ligible effec'ts in the' igniter temperature tests. Nominal values for these two parameters are 120*F and the drop size generated If the ig-by the SPRACO 1116 nozzle to obtain the specified water fluxes. a new niter temperature tests show that these parameters are indeed important, decision tree will be constructed. The first test will be conducted with 6 vo1% hydrogen, water flux of 0.9 gpm/ft , and an igniter voltage of 120V. Figure 3-3 shows that, if an ignit-ability test results in ignition, the following test will be conducted with a larger water flux. If the test results in no ignition, either the water flux is decreased, the hydrogen concentration is increased to 8 vol%, or the igniter ' voltage is increased from 120v to 130v. After the indicated tests have been conducted, FMRC and TVA representatives will discuss possible additional test conditions. The intent is to terminate testing when ignition has been verified (perhaps by repeating a successful ignition test) at a water flux comparable to, or preferably greater than, that expected to impinge on the spray-shielded , igniters in Sequoyah. 14 e
4 FACTCRY MUTUAL RE SE ARCH CORPORAilON OJ1RS. RU
=
TEST 1 TEST 2 TEST 3 TEST 4 TEST 5 TEST 6 STOP , 4 = 1.8 STOP H2 = 6. V = 120
* = 1.8 STCP H2 = 6. Y = 130 6= 1.8 6= 1.57 H2 = 8..V = 130 * = 1.35 "2 = 6.' V = 130
**9 -
** 1.12 H2 = 8. V = 130 TCP H2
- O'Y
- I2C H 2
- 8. V = 130 STOP -
e = 1.8 STOP
*= 1.8 e= 1,57 H2 = 6. V = 130 H2 = 8. V = 130 += 1.35 Hg = 8. V = 130 e = .9 N e= 1.12 2 = d. V = 130 H2 = 6. V = 130 WP "2 = d. V = 130 STOP
*= 1.8 e= 1.57 LEGEND e = .9 e= 1.35 130 TOP H2 = 8. V = 130 H2=:.V
+ = water flux. gpm/ft 2 H2 = s, y = 130 H2 ' = " 8g7 V = 130 H2 = hydrogen vo1% ST0p *= 1.12 V = igniter voltage *= .45 STOP H2 = 8. V = 130 H2 = 8. V ,= 130 Top Branch = Ignition H2 = 8. V = 130 4= .22 Bottom Branch = no ignition Hg = 8. V = 130 Figure 3-3 Preliminary Decision Tree for Ignitability Tests l
l 15
-- - - - .- = - _ . - _ - _ .
FACTORY MUTUAL R65EARCH CORPORATION OJ1RS.RU IV PROGRAM SCHEDULE Test preparations have proceeded to the point that engineering drawings / schematics have been completed and all equipment and/or components are avail- , able at FMRC for test facility fabrication. As indicated in Figure 4-1, it is anticipated that test facility prepara-cion and checkout can be completed within two months, and that testing can be
- complaced one month later. This schedule would allow a project final report to be written, revised to incorporate TVA comments, and submitted to the NRC by the end of 1983.
d l l l l 16 . O w.
~
5 FACTORY MUTUAL RESEARCH CORPORATION FIGURE 4-1 PROPOSED SCHEDULE Time (months) J A S O N D 3 J
- 1. Test Planning 1.1 Confirm Test Approach um
. 1.2 Specify Equipment and "" , . Instruments 1.3 Prepara Test Matrix manus -
1.4 Prepare Preliminary ,,,,, Approva:. of Test Plars Report TTA-NRC
- 2. Test Facility Preparation ,
b 2.1 Fabricate Vessel imummu 2.2 Install and Calibrate Instruments 2.3 Check-Out Test - musum
- 3. Testing 4 Final Report 4.1 Prepare Draft Reports
====
4.2 TVA & NRC Review of Draft Report 4.3 Revise Report and
~
gf
^ answer questions
- Oral Presentatiott at NEC e
17 4
=
FACTORY MUTUAL RESEARCH CORPORATION OJ1RS.RU REFERENCES _
- 1. TVA Response to NRC Request for Additional Testing Performed on Tayco Igniter used in the Sequoyah Nuclear Plant, January 31, 1983.
- 2. Renfro, D. , Smith, L. , Thomp' son, L, and Clever, R. , " Development and Testing of Hydrogen Ignition Devices," presented at the Second International Workshop on the Impact of Hydrogen on Water Reactor Safety Albuquerque, Optober 1982..
- 3. Bajpai, S.N. and Zalosh, R.G. , "The Ef fect of Water Fogs on the Deliberate Ignition of Hydrogen " EPRI NP-2637, RP1932-1, November 1982.
- 4. Tansa, H. , et al. , " Ignition Ef fectiveness of Thermal Heating Devices In Lean Hydrogen-Air-Steam Mixtures," AECL WNRE Drr.ft Report, January 1983.
- 3. Coward, H.F. and Jones, G.W. (1952), " Limits of Flammability of ;ases and
Vapors," Bureau of Mines, Bulletin 503, 1952. , e L I 18
' " ' ' - " ' " - - __ __ _ _ _ _ _}}