ML20041B520

Effect of Ignitor Location & Water Fogs on Hydrogen Combustion within Enclosed Compartment

ML20041B520
Person / Time
Site:	Cook
Issue date:	12/31/1981
From:	Hudson F, Shiu K, Torok R ACUREX CORP., AMERICAN ELECTRIC POWER CO., INC., DUKE POWER CO.
To:
Shared Package
ML17319B196	List: ML17319B195 ML17319B197 ML17319B198 ML17319B199 ML17334A364 ML20041B520
References
NUDOCS 8202240173
Download: ML20041B520 (62)

Text

{{#Wiki_filter:- s 1 s. e,. EFFECT OF IGNITOR LOCATION AND WATER F0GS ON HYDROGEN COMBUSTION WITHIN AN ENCLOSED COMPARTMENT PROJECT REPORT December, 1981 Prepared by: F. G. Hudson, Duke Power Company K. K. Shiu, American Electric Power R. C. Torok, Acurex Corporation J. J. Wilder, Tennessee Valley Authority Project Conducted by: Acurex Corporation 485 Clyde Avenue Mountain View, California Project Sponsors: American Electric Power Service Corp. Duke Power Company Electric Power Research Institute Tennessee Valley Authority J l 8202240173 820217 PDR ADOCK 05000315 P PDR

t s, j 1 TABLE OF CONTENTS 1 i 1.0 Introduction 2.0 Test Facility i-3.0 Test Matrix and Procedures 4.0 Test Results 5.0 Conclusions Appendix Gas Chromatography Analysis 4

f T

t i s [ i I, 4 1 i e r~- ,,m-. -c,,-.- ,,m-- -,w- ,..,,n,,,,n.a., , ann n,.. - - -,---.n.-. - - - ---- -,, -

t w, Section 1 Introduction Approximately ten hours into the accident at Three Mile Island, a hydrogen burn occurred inside the containment. Although this burn posed no real threat to the TMI containment, it did create interest in hydrogen combustion and its effects on containment structures. The operating license applications for McGuire and Sequoyah Nuclear Stations contributed to the growth of this interest into a major safety concern, especially for ice condenser containments. The individual and joint activities of the three utilities owning ice condenser stations (American Electric Power, Duke Power Company, and the Tennessee Valley Authority) are well documented in various licensing submittals or licensing proceedings and will not be repeated here. However, when the three utilities decided to install a distributed ignition system as a hydrogen mitigation system, the question of ignitor location within a compartment arose. Additionally, concurrent with the design of a distributed ignition system, several independent organizations suggested coupling a water fog system with the distributed ignition system to act as a pressure suppressant during combustion. To investigate the effect of ignitor location on hydrogen combustion within a compartment, the three utilities', in conjunction with the Electric Power Research Institute, con-tracted with Acurex Corporation to conduct a series of tests. These tests were conducted at the SRI International Explosives Test Site near Livermore, California. Although analyses had shown a pressure suppressant was not necessary for ice condenser containments, the utilities believed that investigating the effect of water fogs on hydrogen combustion could be of some potential interest to the

1, industry. Therefore, an investigation of water fog effects on hydrogen combustion was added to the Acurex project. This report presents the results of that project. I l

Section 2 Test Facility 2.1 TEST VESSEL AND MECHANICAL SYSTEMS The test vessel selected for this project has a volume of approximately 630 3 ft. Vessel dimensions are: internal diameter - 7 f t., overall height - 21 ft., and " barrel" height - 17 ft. Auxiliary mechanical systems provide the ability to:

1) inject hydrogen or a hydrogen / steam mixture into the lower portion of the test vessel, 2) supply a water spray or microfog from the upper portion of the test vessel, 3) obtain pre-test and post-test vessel atmosphere samples, and 4) provide a means of premixing the vessel atmosphere for quiescent tests. Additionally, the capability to ignite the vessel atmosphere from the top, middle or bottom of the vessel was provided. A schematic of the test vessel and its auxiliary mechanical systems is presented in Figure 2-1.

A propane-fueled boiler supplied steam for the facility. This steam served as a parameter for several tests, as well as to preheat the test vessel to the desired temperature. The steam flowrate was monitored with an annular flow sensor and a differential pressure gauge. When steam was not required as a j test parameter ( the boiler was isolated from the test vessel af ter preheating was completed. Bottled hydrogen served as the hydrogen source for the test vessel. The hydrogen flowrate was monitored with a rotameter and controlled f with a control valve. 6

s: A three horsepower electric motor and gear pump supplied water to the test vessel spray nozzles. A bypass loop was included to control the flowrate. Utilizing a closed loop spray system, i.e., recirculating the spray water, avoided the potential problems associated with accumulating large volumes of water within the vessel. For the Phase 1 tests, a single Sprayco 1713 nozzle, 15 gpm flowrate, was mounted at the top of the test vessel. A manifold containing nine Sprayco 2163-7604 pinjet nozzles was mounted at the top of the test vessel for the Phase 2 tests. Depending on the pressure drop across the nozzles, the total spray flowrate for the Phase 2 tests varied from 1.1 to 1.4 gpm. The spray manifold was constructed so as to provide an even spray distribution throughout the test vessel. An air-operated fan was mounted inside the test vessel to assure a well mixed vessel atmosphere prior to the quiescent tests. Use of an air operated fan eliminated the potential of an electrical malfunction resulting in a spurious ignition. The fan's air exhaust was vented outside the test vessel to avoid diluting the vessel atmosphere. Two 4 inch butterfly valves located at the top and near the bottom of the test vessel allowed the vessel to be purged following the completion of each test. A squirrel-cage ~ blower attached to the lower butterfly valve provided the motive force for purging the vessel. The vessel contents were vented to the atmosphere through the upper butterfly valve. The vessel was not vented until post-test samples were obtained.

i Vessel atmosphere sample taps were located near the top and near the bottom of the test vessel. A remotely operated solenoid valve isolated each of the two sample lines from the ?.est vessel. When a solenoid valve was open, the vessel atmosphere sample was pumped through a cold trap to remove water. The sample then passed through a silica gel trap to remove any remaining moisture. The sample then flowed through a gas meter into a glass sample bottle. A sample was extracted from the sample bottle by a syringe and injected into a gas chromatograph. (A detailed discussion of the gas analycis methodology is presented in Appendix A). Two ignitor assemblies supplied by Duke Power Company were mounted inside the test vessel. The ignitors were located on the vessel centerline at either the top, middle or near the bottom of the test vessel. Only two ignitor locations were occupied at one time. The top ignitor location was not used during the Phase 1 tests requiring sprays or during any of the Phase 2 tests since an ignitor assembly located at the top effectively created a significant spray / fog maldistribution within the test vessel.

2. 2 INSTRUMENTATION The test vesser was instrumented to provide the following information:

vessel atmosphere temperature, vessel wall temperature, flame front propagation, and vessel pressure. Three mil Type K thermocouples were used to measure temperatures and detect flame front propagation. Strain gauge pressure transducers and piezoelectric pressure transducers were used to measure vessel pressures. A schematic of the test vessel instrumentation is presented in Figure 2-2. 9

s, As just mentioned, 3 mil thermocouples were used to measure temperatures and detect flame front propagation, Male Type K thermocouple jacks were used as attachment points for the 3 mil thermocouple junctions to increase the robust-ness of the thermocouple. The junctions were located between.the jacks with the leads attached directly to the jacks. Vessel wall temperature thermo-couples were welded directly to the vessel wall. To determine the flame front propagation pattern within the vessel, a special electronic circuit was developed. This circuit used a high input impedance operational amplifier comparator to detect the temperature rise in the 3 mil Type K thermocouples located within the vessel. A schematic diagram of the flame front detector circuit is shown in Figure 2-3. Five circuits were used, each circuit monitoring seven thermocouples located in a vertical array within the vessel. The 5 x 7 grid was located vertically on a plane formed by the diameter and centerline of the test vessel as shown in Figure 2-2. When one of the thermocouples in the grid was exposed to the flame front, the thermocouple output voltage would rise and trigger the comparator. This, in turn, placed a signal at one input point of the digital-to-analog converter. The output from the DAC was ther, recorded. Since the input signals from the thermocouples c~ould be considered as binary bits, every voltage that was generated by the DAC corresponded to a discrete combination of hot thermocouples. Thus, by comparing the output voltage against time, the instant that the flame t front arrived at each location could be determined, and a map derived showing 1 flame front propagation. l l 4

5 'h Two types of pressure transducers were used. Bell and Howell CEC Model 1000 strain gauge pressure transducers were used for static and slow response conditions. These were powered by CEC 1-183 strain gauge signal conditioners t located within a CEC 1-080 power supply chassis. To record high frequency transient pressure pulses, rCB Piezoelectronics Model 111A24 piezoelectric pressure transducers were used. These transducers were powered by a Model 484810 power supply, t Two recording systems were used for data acquisition. A twenty eight channel FM tape recorder, EMI Model 7000C, was used to record all potentially fast response data. Frequency response of the unit was 10kHz or greater. An Autodata 9 datalogger recorded relatively slow response signals and served as a i backup to the FM tape recorder. 1 { l b O m

' FIGURE 2-1 TEST FACILITY ltECilAtlICAL SCHEMATIC 21 51 . 11 Pr.p.ne Vent $j _g 9s <...,.i, _ l#, _ y 4~ / o n,.., / = 4 h a

_gz, y

0 ........, } 4 ....,b j#y e., ks o o g......., y/ m / e ~ ~ v o+,r.n.ost = A s-x 3

s FIGURE 2-2 TEST FACILITY INSTRUMENTATION SCHEMATIC 3 ? .,t_ i ut 'l ,r Flare frtnt gIge F T Themocouplei wi Wall thPTMCOTple

• Syg 530 Strain gauge' pressure transducer PIO PED Piercelectri,C pressure transducer STG Strain gauge SG Silica gel water trap "TG' DCM Dry gas m ter G8 Glass bottle Ignitor GC Gas chroma tograph F F

.F ,F h 'h I $3 D.Q V / Pi,mo y

r ff l r r

r F r GB WT ' t, r t 'F.,* F F F l i I p 7. / p lgnitor l 0 ' rsfG g PurN ( FF f'(. F

F l

f 1

1. gnitor Gn l

>1 F F F ] l ig 6r suo ~ ~<rro l T F F f. f 1 t < j b- -C. 520 GC Injection nottle i s i I e ,e t e / gh D r ',-4

\\ 9 FIGURE 2-3 FLAME FRONT DETECTOR CfRCUIT 1 i. i

• 15V

) $30k f th Th(Pv)(0;Ptt -) n,e>yLED ,,...s...,,,,1 2.m 20 O O O 0 0 0 1 0.078d 0 0 0 0 0 1 0 0.157v 0 0 0 0 1 0 0 0.31st 4 g 0 0 0 1 0 0 0 0 627v \\ 0 0 1 0 0 0 0

1. 2'. 5 v 3290 0 1 0 0 0 0 0 2.510#

1. 3 i

1 0 0 0 0 0 0 5.020v 0 0 0 0 0 l I

5. tH e (1C 1 1 1 1 1 I I ICV g

3790 ~ fare g d 0.,3;:,.., e.t, y. A 3290y ... L, s A t F,', M 190Y A 3?93 4

i, Section 3 Test Matrix and Procedures 3.1 TEST MATRIX The test program was designed to investigate the effects of ignitor location and water fogs / sprays on hydrogen combustion. Test parameters were selected based on degraded core analyses conducted by American Electric Power, Duke Power Company, and TVA. Flowrates for steam and hydrogen were scaled by the ratio of test vessel volume to the combined lower compartment and deadended compartment volumes of an ice condenser containment. These scaled _flowrates were derived from the average steam and maximum hydrogen release rates dur-ing the hydrogen generation portion of an S2 0 accident sequence. Containment response analyses conducted by the aforementioned utilities indicated that 160 F was a maximum atmospheric temperature that would be encountered in the lower and deadended compartments at the onset of hydrogen generation. Hydrogen or hydrogen / steam mixtures were injected into the lower portion of the test vessel since many subcompartments within an ice condenser containment are accessed from the bottom. Spray nozzle flowrates were determined by the desired mean droplet diameter. The Sprayco 1713 spray nozzle, a model used in containment spray systems, was operated at a AP of 40 psi. This provided a flowrate of 15 gpm and is the l cesigned operating condition in a containment spray application. Vendor supplied information indicated that the number mean droplet diameter at this flowrate is 200p. Based on studies conducted by Factory Mutual Research

Corporation, the Sprayco 2163-7604 nozzle was selected for use in the water fog tests. Measurements taken at Factory Mutual Research Corporation1 indicated that when operated at AP's of 20 psi and 30 psi the resulting number mean droplet diameters were 11p and 8p, respectively. The total fog flowrates from the test vessel's nine nozzle manifold were 1.1 gpm and 1.4 gpm, respectively. Tests outlined in Table 3-1 were intended to investigate the effect of ignitor location on combustion. Both hydrogen and hydrogen / steam mixtures were injected to account for the potential variation in the transient conditions within an ice condenser containment. Tests with sprays were included to account for the presence of subcompartment sprays in some containment designs. As mentioned earlier, the steam and hydrogen flowrates of 2.1 and 0.035 lbm/ min were based on transient analyses. Tests 1.6 and 1.7 were conducted with the hydrogen flowrate arbitrarily increased by a factor of three. Test 1.11 was conducted i to observe the effect of a lower preheat on combustion characteristics. The I duration of tests was determined by the time required to obtain the same relative hydrogen mass injected in the test vessel as is calculated in the previously mentioned degraded core analyses. All tests were conducted with the test vessel atmosphere initially saturated. ,/ 1Zalosh, Robert G., " Water Fog Inerting of Hydrogen Air Mixtures," Factory Mutual Research Corporation, September, 1981. i l l

The water fog test matrix is presented in Table 3-2. Quiescent tests were conducted as a basis of comparison to observed results of the dynamic tests. The basis for all remaining test parameters was discussed above. 3.2 TEST PROCEDURES Two types of tests were conducted: quiescent and dynamic. A known amount of hydrogen was injected into the test vessel for the quiescent test prior to energizing the ignitor assembly. The dynamic tests consisted of injecting hydrogen or hydrogen / steam into the vessel with the ignitor assembly pre energized. For all tests, the test vessel was pre-heated to the desired temperature and the instrumentation and data acquisition system was checked and calibrated. After the completion of each test, the test vessel fan was turned on and a post-test sample obtained. Subsequently, the test vessel was purged. Test procedures varied slightly for the quiescent and dynamic tests. For the quiescent tests, the vessel fan was turned on af ter completing the pre-test activities mentioned above. A known amount of hydrogen was injected into the vessel and a pre-test sample was obtained. The vessel fan was then turned off and, if required, vessel sprays actuated. At this point, the data acquisition system and the-ignitor assembly were energized. For the dynamic tests, the ignitor was energized after completion of pre-test activities. If necessary, vessel sprays were then actuated. The data acquisition system was energized prior to the initiation of hydrogen injection. l l I 1

TABLE 3-1 Ignitor Location Test Matrix

• Ignitor Hydrogen Flow (1bm/ min)

Steam Flow (1bm/ min) Spray Flow Test Location 0.035 0.105 2.1 15 gpm 1.1 Top 'X X 1.2 Top X 1.3 Bottom X. X X t i 1.4 Bottom X X 1.5 Bottom X 1.6 Bottom X X

1. 7 Bottom X

- 1. 8 Center X X X 1.9 Center X X 1.10 Center .X 1.11 Bottom X X X { STest vessel was preheated to 160*F for all tests except 1.11. Test 1.11 was i performed with 120 F preheat. 1 i

t l' TABLE 3-2 ) Water Fog Test Matrix

• l Hydrogen Hydrogen Flow (lbm/ min)

Steam Flow (lbm/ min) Fog Nozzle Pressure (psi) Test # v/o 0.035 0.105 2.1 20 30 2.1 5.0 'l 2.2 7.5 2.3 10.7 i 2.4 10.7 X 2.5 10.7 X i 2.6

7. 5 X

2.7

7. 5 X

2. 8 X

X t r 2.9 X X X 2.10** X X 2.11 X X 2.12 X X 2.13 X X X I RTest vessel was preheated to 160 F for all tests. Ignitor located near the bottom. i CQVessel mixing fan was operating. i

Section 4 Test Results 4.1 DEFLAGRATION CHARACTERISTICS Based primarily on pressure and flame front detector data, two distinct types of deflagrations occurred. These deflagrations were termed " discrete" and " intermittent". A " discrete" deflagration was characterized by a rapid pressure and temperature rise. The duration of the burn appeared to be dependent on the fraction of the vessel volume that could support a propagating hydrogen flame. By comparison, an " intermittent" deflagration appeared as repeated burns accompanied by much slower and lower pressure / temperature rises. Observed deflagrations were further categorized as " major" or " minor". A " major" deflagration, whether discrete or intermittent, occurred throughout the test vessel. A " minor" deflagration, on the other hand, was localized in nature. Pressure and temperature histories typical of a major-discrete deflagration are presented in Figure 4-14. " CENTERLINE T1" and " CENTERLINE T4" are centerline thermocouples located near the top and bottom of the test vessel. " VESSEL PRESSURE 2" is a strain gauge pressure transducer located in the lower portion of the test vessel. As seen in the figure, the test mixture ignited approxi-mately twenty seconds into the test. The periodic disturbances in the two temperature traces were caused by the datalogger scanning these channels. The flame front detector output for the same test is presented in Figure 4-15. The five data channels correspond to the five vertical columns of transducers shown 4 -e., v

in the inset figure; each channel receiving the output of seven flame detectors. The size of each step is determined by the combination of detectors that triggered. A fullscale reading indicates that all seven detectors within a channel have triggered. Of the channel outputs presented in Figure 4-15, only channel 5 was not reading fullscale. The flame front detector data indicated l that approximately 95% of the flame front detectors triggered. This, therefore, was classified a " major" and " discrete" deflagration. Figures 4-5 and 4-6 present temperature, pressure, and flame front detector data typical of major intermittent deflagrations. The flame front detector data in Figure 4-6 indicate that all channels were triggered, signifying that a major deflagration occurred. Temperature data in Figure 4-5 shows a sharp, but not extremely large, temperature increase that remained relatively constant for several minutes before starting to slowly decay. This indicates that intermit-tent deflagrations were occurring. Note that near the end of this particular test, a major discrete burn occurred. Minor intermittent burning is demonstrated by the temperature, pressure and flame front detector data presented in "fg9res 4-21 and 4-22. Flame front detector data, Figure 4-22, indicate that few flame detectors were triggered-during this test. This characterizes a " minor" deflagration. The temperature and pressure data in Figure 4-21 show the small and gradual increases characteristic of intermittent burning. Figure 4-3 also provides temperature and pressure histories indicating minor intermittent deflagrations. Note that near the end of this test, several minor discrete deflagrations occurred.

4.2 INSTRUMENTATION UNCERTAINTY Test data was primarily test vessel pressure / temperature histories and flame front detector output. Additionally, vessel atmosphere constituents were deter-mined via a gas chromatograph. To properly evaluate the test data, it was necessary to know the errors associated with the instrumentation used. Sources of error in the thermocouple data were: thermocouple material and junction uncertainty, thermocouple amplifier error, test facility offset errors due to electrical ground loops, tape recorder input and playback error, analog to digital conversion errors and plotter inaccuracy. Standard Type K thermo-couple error estimates were 12.2 C from 0 to 278 C and 3/4% above 278 C, ANSI Standard C96.1. The low temperature range error corresponded to approximately 13% when peak temperatures were around 100 C. The estimated gain error for each of the three signal amplifiers was 1%. Errors associated with the analog to digital conversion and plotter inaccuracies were considered negligible. Therefore, using the root-mean-square method, the total random error for low temperature cases was approximately, h [32 + 12 +12 +123 = 3.5% and for high temperature cases, 4 [(3/4)2 +12, 12 + 12 3 = 1. 9%. 9 j

Pressure data were obtained from two strain gauge pressure transducers mounted at the top and near the bottom of the vessel. Although data from the piezoelectric pressure transducers were recorded, the observed pressure transients did not warrant use of the piezoelectric instead of the strain gauge pressure transducers. Sources of error in the pressure data were: transducer error, the transducer amplifier error, and the same tape recorder amplifier errors encountered in the thermocouple data. The manufacturer's estimated error for the transducer was.25% of full range. This corresponded to 0.5% of the signal resulting from a large aP, i.e., a major deflagration, and 2.5% for a small AP. Using the root-mean-square method, the total random error for high AP cases was approximately, h [( )2 + 12 + 12 + 123 = 1.8% and for low AP cases, 12 + 12 + 12]h = 3. 0%. [2.52 4 f Calibration and test gas chromatograms were analyzed using the peak height method. Water vapor corrections were made to convert dry gas sample analyses l to actual test'%essel conditions. Uncertainty estimates for the resulting gas constituent volume percentages were based on the repeatability of sample analyses. Normalized mean peak heights were calculated from the analyses of each test run. The ratio of the largest deviation from the mean to the mean for each test run was used as an estimate 1 i

of the gas analysis uncertainty. The mean estinated uncertainties were approximately 10% for the hydrogen analysis and 20% for the oxygen analysis. Problems encountered with the gas chromatograph's thermal conductivity detector were suspected as being a primary cause of scatter in the data. Discussions with the manufacturer indicated that filament oxidation was hampering performance. 4.3 IGNITOR LOCATION TEST SERIES Tests were conducted by varying the ignitor location in three test environments: hydrogen injection, hydrogen / steam injection, and hydrogen / steam injection with sprays. Two additional tests were conducted with the hydrogen flowrate arbitrarily increased by a factor of three. The final test of this series was conducted with a reduced vessel pre-heat. A summary of the results obtained from this test series is presented in Tables 4-1 and 4-2. Figures 4-2, 4-4, and 4-10 provide the pressure histories from ignitor location tests without steam or spray (tests 1.2, 1.5, and 1.10). At lean hydrogen con-centrations, flame propagation is only upwards. With this in mind, it was anti-cipated that while localized burning was occurring in the vicinity of the top ignitor, the hydrogen concentration would increase throughout the remainder ..c.. of the vessel. When the flammabil.'ty limit for downward propagation was reached, a major discrete deflagration would occur. This appeared to be the sequence of events in test 1.2 with minor intermittent deflagrations beginning at 300 seconds followed by a major discrete deflagration at 580 seconds. The maximum pressure rise was expected to be smaller for the center ignitor location (test 1.10) than the top location because of the increased vessel volume that would be

'i. ~ exposed to upward propagating flames at lean concentrations. Table 4-1 shows that the pressure rise was lower by approximately a factor of three. Minor intermittent deflagrations began around 220 seconds and continued throughout the test. The lowest ignitor location was expected to produce an even milder pressure rise since a substantial portion of the vessel would be exposed to upward propagating flames. However, as Figure 4-4 shows, that was not the Apparently, the relative locations of the injection port and the lowest case. ignitor location precluded the ignitor from ignitirg hydrogen early in the test. The major discrete deflagration that occurred at 450 seconds irdicated that the injection flow apparently bypassed the ignitor until most of the vessel contained a flammable mixture. The resulting deflagration produced a higher pressure rise than that attained by the top ignitor for two apparent reasons:

1) The top ignition was preceeded by localized deflagrations; thus reducing the mass of hydrogen within the vessel; 2) Flames propagate slower downward than upward; thus allowing more time for heat transfer.

Figures 4-1, 4-3, and 4-9 show that results from the tests with hydrogen / steam injection (tests 1.1, 1.4, 1.9) were similar to those obtained from the hydrogen injection tests. It was anticipated that the hydrogen / steam injection tests would yield milder pressure increases. This was due to steam impeding the combustion process as well as acting as a diluent; thus reducing the flame propagation velocity. This would result in increased heat transfer and decreased temperatures / pressures. Additionally, adding steam increased the injection velocity from approximately 0.7 ft./sec. to 5.7 ft./sec. This was believed to increase mixing within the vessel and thus allow deflagrations to occur at leaner - - ~

+ hydrogen concentrations. Table 4-1 indicates that the peak pressures at the three ignitor locations were reduced with steam added to the injection flow. The most dramatic change was with the bottom ignitor~ location. Apparently, the increased mixing provided by the steam flow allowed the lowest ignitor to function as discussed previously. The top ignitor provided the largest pressure rise, with the center and bottom ignitors being approximately a factor of three less. The addition of a water spray was expected to create some amount of turbulence within the vessel that would enhance mixing and allow combustion to occur at leaner hydrogen concentrations. It was also anticipated that a water spray would act as a dispersed heat sink; thus further reducing temperatures and pressures. Table 4-1 shows that the addition of water sprey, tests 1.3 and 1.8, did reduce the maximum pressure. The bottom ignitor yielded only a very slight pressure rise with no corresponding flame front detector activity.

However, post-test atmosphere analysis indicated that combustion had occurred.

These deflagrations must have been very localized near the ignitor and apparently relied upon spray induced turbulence for a continual supply of lean hydrogen l mixtures. Figure 4-8 shows that the center ignitor provided a series of minor discrete burns. The pressure rise was slightly higher than obtained from the bottom ignitor:' Test vessel design did not allow a spray test to be conducted with the upper ignitor location. l Two tests were conducted with the hydrogen flowrate arbitrarily increased by a factor of three. One test was conducted with hydrogen injection, test 1.7, and i the second with hydrogen / steam injection, test 1.6. The bottom ignitor was used i

,~ for both tests. Comparing tests 1.7 with 1.5 and 1.6 with 1.4 shows that the transients were similar to their low flow counterparts, with the exception of ignition occurring earlier in the transient. In test 1.7, ignition occurred slightly earlier than the 150 second ignition expected from a higher flow rate (see Figure 4.7). It is possible that the increased injection velocity had a slight effect on mixing within the vessel. This would allow an ignitable mixture to reach the ignitor earlier. This would also explain the slightly lower pressure rise from test 1.7 since less hydrogen would be present within the vessel at ignition. Note that the high flowrate pressure rise was 85% of the low flowrate pressure rise and that the high flowrate ignition time was 85% of the anticipated 150 second ignition time. Adding steam to the high hydrogen flowrate, test 1.6, yielded deflagrations similar to the low flowrate counterpart, test 1.4, but a pressure rise essentially identical to that obtained from test

1. 7.

Figure 4-5 shows that ignition occurred at approximately 100 seconds, one third of the 300 second ignition time for test 1.4. The ensuing intermittent deflagrations were more severe in test 1.6 because the higher hydrogen flowrate apparently resulted in a higher energy release rate. Why these intermittent deflagrations were not followed by repeated discrete deflagrations as seen in test 1.4 is uncertain. One possible explanation is that with vessel atmosphere temperatures in excess of 400 F for over one third the duration of test 1.6, a ..r_ ~ fraction of the water collected at the tank bottom from vessel pre-heating was vaporized during the intermittent burning. This could have caused the defla-grations to be very localized, similar to those obtained in test 1.3.

Thus, hydrogen could have built up in the vessel while the steam was slowly condensing

.until an ignitable mixture was once again obtained. The result would be a lull in flame front detector activity followed by a major discrete deflagration.

=

Figures 4-5 and 4-6 show such characteristics. Some credence is lent to this possibility by noting that the post-test water concentration from test 1.6 was SC% larger than that obtained from test 1.4. One test was conducted, 1.11, with the vessel pre-heat reduced to 120 F from 160 F. The results obtained were very similar to those obtained from 1.3, an identical test with a 160 F vessel pre-heat. A very slight pressure rise occurred with no corresponding flame front detector activity observed. This indicated that the burn, as in test 1.3, was very localized. 4.4 WATER F0G TEST SERIES Tests were conducted to investigate the effects of a water fog on hydrogen combustion. The fog nozzle, Sprayco Model 2163-7604, created different fog characteristics depending on the pressure drop across the nozzle. Tests were conducted with two different fogs. Based on data obtained from Factory Mutual Research Corporation, a 20 psi AP yielded a fog with a numbe mean droplet diameter of 11p, while a 30 psi aP yielded a number mean droplet diameter of 8p. Two types of tests were conducted: quiescent and dynamic. The dynamic tests were conducted with and without steam. All tests utilized the bottom ignitor. A summary of the results obtained from this series of test is presented in Tables 4-3 and 4-4. l l i

1 4.4.1 QUIESCENT TEST SERIES 4 To provide a baseline of information for evaluating the dynamic fog tests, a series of quiescent tests were conducted. Nominal hydrogen concentrations of 5, 7.5, and 10% were selected. Table 4-4 indicates that the completeness of I combustion for those tests without fogs (tests 2.1, 2.2, and 2.3) was approximately 30%, 90%, and 99%, respectively. This data agrees reasonably well with published data. The temperature and pressure histories of these three tests are presented in Figures 4-11, 4-12, and 4-13. Further tests with 5% hydrogen.were not con-ducted. i Repeating these tests with fogs present, a significant decrease in pressure rise was anticipated. Due to.the 'arge surface area present within a fog, the fog was expected to act as a dispersed heat sink; resulting in reduced temperatures and pressures. However, as Table 4-3 indicates for the 7.5% hydrogen tests, 2.6 and 2.7, the observed pressure rises were slightly higher. Table 4-4 shows that the completeness of combustion increased from approximately 90 to greater than 99%. This indicated that these particular fogs acted very much like' sprays for lean hydrogen concentrations. The turbulence created by the fog flow apparently enhanced the completeness of combustion; thus increasing the pressure rise. The heat sink effect of the fogs was evident from the slightly lower temperatures (compare Figure 4-12 with Figures 4-17 and 4-18). The fogs had no apparent effect on the peak pressure rise in the 10% tests, 2.4 and 2.5. 9

4.4.2 OYNAMIC TEST SERIES Hydrogen injection tests 2.8 and 2.12 were identical to test 1.5 except that fogs were included. Results from 2.8 and 2.12 indicated only minor intermittent defiagrations (see Figure 4-23). The observed pressure rise was an order of magnitude lower than that observed in test 1.5. It would be reasonable to assume that a great deal of the pressure reduction was due to fog flow induced mixing, which allowed a flammable mixture to reach the ignitor earlier. Note that the effect of adding steam to test 1.5 (test 1.4) was minor intermittent burning early in the transient with a pressure rise of 2.3 psi; the same effect observed in tests 2.8 and 2.12 (see Figure 4-3). Figure 4-19 shows a pressure history typical of that obtained from hydrogen / steam injection tests with fog present, tests 2.9 and 2.13. The observed pressure rises were similar to that obtained in test 1.4, hydrogen / steam injection with no spray. In both cases, with and without fog present, ignition occurred at approximately 300 seconds. However, in test 1.4 the result was minor intermit-tent deflagrations eventually becoming minor discrete deflagrations. Tests 2.9 and 2.13 provided minor discrete burns immediately upon ignition. As a result, 1 it appears that more hydrogen was consumed in the fog tests than in the non- . s._ fog tests. This would also appear to indicate that a major contribution from the generation of fog in this test series was to provide uniform mixing within the vessel. .One test was conducted with the hydrogen flowrate arbitrarily increased by a factor of three in the presence of a fog, test 2.11. Figures 4-21 and 4-22 show

the pressure history and flame front detector activity obtained from this test. This test provided an indication of the heat sink effect of a fog as well as its effect as a source of turbulence. Note on Figure 4-21 that ignition occurred approximately 20 seconds into the test, while in test 1.7, an identical test without fog, Figure 4-7 shows that ignition did not occur until 130 seconds into the test. This difference in ignition time could be accounted for by the mixing created by the fog. The heat sink effect was apparently demonstrated since the peak temperature of test 2.11 remained around 200 F, while the peak temperature in test 1.7 hovered around 400 F. In test 2.11, fog produced turbulence apparently prevented a major discrete burn by inducing intermittent deflagrations early, thus precluding the relatively high hydrogen concentrations needed for a major discrete deflagration. Additionally, the fog acted as a heat sink during the resultant intermittent deflagrations; thus minimizing pressure and temperature increases. Comparing test 2.11 with test 2.12, an identical test with the lower hydrogen flowrate, the deflagration character-istics were similar with the high flowrate test having an earlier ignition. Finally, data on the effect of fan induced turbulence was obtained when the mixing fan was accidently actuated prior to test 2.10. The result was that ignition occurred earlier (see Figure 4-20) than in a similar test without the fan operating (see Figure 4-23). Other than the ignition time, the results of both tests were relatively similar; minor intermittent deflagrations with very slight pressure rises.

TABLE 4-1 Summary of Test Results: Ignitor Location Test Series Max. AP Test # Test Characteristics Ignitor Location (psi) Deflagration Characteristics 1.1 Low H, steam Top 13 Minor, major intermittent 2 ' 1. 2 Low H Top 20 Minor intermittent, major discrete 2 l 1.3 Low H, steam, spray Bottom 1 Minor intermittent 2 1.4 Low H, steam Bottom 4.5 Minor intermittent, minor discrete 2 1.5 Low H Bottom 28 Major discrete, minor intermittent 1 2

1. 6 High H, steam Bottom 24 Major intermittent, major discrete l

2 1.7 High H Bottom 23.5 Major discrete, minor intermittent 2

1. 8 Low H, steam, spray Center 2.7 Minor discrete 2

1. 9 Low H, steam Center 4

Minor intermittent 2 1.10 Low H Center 6 Minor intermittent 2 1.11 Same as 1.3, lower preheat Bottom 1 Minor intermittent

TABLE 4-2 Test Vessel Atmosphere Constituents: Ignitor Location Test Series Test # Post-Test H (v/o) H 0(v/o) 0 (V/0) z 2 2 1.1 11.0 17.6 8.9 1.2 2.6 23.8 15.1

1. 3 6.5 19.8 6.I 1.4 7.9 21.0 5.7

1. 5 2.1 38.2 12.2

1. 6 10.9 32.1 6.5

1. 7 12.7 37.6

1. 7 1.8

7. 9 30.1 7.2

1. 9 3.6 46.1 5.2 1.10 0.4 36.3 5.7 1.11 3.5 27.8 2.6 l

~ l i

TABLE 4-3 Summary of Test Results: Water Fog Test Series Max. AP Test # Test Characteristics (psi) Deflagration Characteristics 2.1 Quiescent, 5 v/o H 8 Minor discrete 2 2.2 Quiescent, 7.5 v/o H 36 Major discrete 2 2.3 Quiescent, 10.7 v/o H 48 Major discrete 2 2.4 Quiescent, 10.7 v/o H, fog 30 47 Major discrete 2 2.5 Quiescent, 10.7 v/o H, fog 20 50 Major discrete 2 2.6 Quiescent, 7.5 v/o H, fog 20 40 Major discrete 2 2.7 Quiescent, 7.5 v/o H, fog 20 39 Major discrete 2 2.8 Dynamic, low H, fog 20 2 Minor intermittent 2 2.9 Dynamic, low H:, fog 20, steam 5 Minor discrete 2.10 Dynamic, low H, fog 30, fan 1 Minor intermittent 2 2.11 Dynamic, high H, fog 30

2. 9 Minor intermittent 2

2.12 Dynamic, low H, fog 30 1 Minor intermittent 2 2.13 Dynamic, low H, steam, fog 30

1. 6 Minor discrete 2

I e' f t

2 l... TABLE 4-4 i Test Vessel Atmosphere Constituents: Water Fog Test Series Test # Pre-Test Post Test l H (V/0) H O (v/o) H (V/0) H 0(v/o) 0 (V/0) 2 2 2 2 2 2.1 4.7 25.6 3.4 33.2 11.5 i 2.2 7.8 26.1

0. 9 43.2 10.0 2.3 10.2 24.9 (0.1 40.4 10.9 2.4 9.7 23.8 (0.1 44.9 8.7 2.5 10.2 25.4 (0.1 40.4 9.2 2.6 7.2 32.9 (0.1 47.6 8.4 2.7 7.6 29.4 (0.1 41.6 11.4

2. 8 3.3 30.1 6.8 2.9 6.2 45.5 5.1 2.10 2.8 49.2 2.7 2.11-13.5 35.1 (0.1 2.12 (0.1 37.0 9.2 2.13 3.7 45.4 3.7 I

I I r

y _y..__ e a 9 4 (DISd) 380SS3Md O O O o o o C o v M N C o 1 i f f f g e o O l t S W L e L (- o -g C -O 47 O -o C3 7-G o... m CE H -o-om D e CW w L g l -o E l c.-. b o -O C O -O y I v Cd o D ~8 m (M ~ Ed o e -o 4 3 Cd -g m m Cd o i i i O O O o o O C C C v O N (van) aanssraa

I PI' ESSURE (KPA) TEh1PERATURE (C) to a o e o 5 -o o o o o o o o o Q O O Q Q Q Q Q Q Q Q i i i Q O__. i n m m m 2 m g g-m g-m r x a r m o-Z o-Z Q m Q m m m g u C u o-o- 4 O N O M o-

• l o-O O

e, v. 0-o- O l O e -4 j

g, -

r-K g, - e e 9 7 M 0;g c-1

• A n

wa ~, .,c 8 - m 8-a s v o e o-0- O O to to o-0- Q Q b o o-o- a O Q = o-o- Q Q n E o" o~ O O t n 0 -, 1 i 2 1 o i Q Q o a e o O ) o O O O O O O g g g g Q O PRESSURE (PSIG) TEhiPERATURE (F)

• e

PRESSURE (KPA) TEMPERATURE (C) N O 4 O N w N e O V O Q Q Q Q Q Q Q Q O O O O O O O-g m m enw 2 -o-o m o-e r mx a e

0 s-o m

o Z in m w a e 0-C u N o-o }

• 0 o

m a o-1 o o-6 O i e 0-n o 0-t o _) a 0 M s e

• e

? o.

• o Ya-n m
• O i

--e.-. m rn O g

== e n j) g - f CD M >\\ m 8-b n o _r oo-o O-y ~7 9 s-s- t g-8- Q Ll. -g- =o-o Q l L l 1 l' n 2 o-Q 8, i m r O s s o o a r-a g g E 5 O Q PRESSURE (PSIG) TEMPERATURE (F) I

4 (a) annivasanal (otsa) aanssana s s e e o o o o o o e o m ~.-. c3 v 5 o Q V O N e 1 f f f o I i f 1 f 1

• oo n

O o e c ,. o u e oo oo e o e e e e 1 ,l I l o o -o -o

l

l-

' 1, o -o -o o g. o i c: n n x u 4 *. u Er 8W 8 5. - m -u cW l_. v -3W -r 'l4 d k o W w i, -o4 o. ll C - o,, l s C- >- e s o l-. o 2}- -oc -o-o -o e Y -o w v m o e o H+ s a g m W m 'z o W o -o x -o N n N a e a W e s -o N o -o Z m cn W W> o i i i

o i

i i o e o e o o e o e o o o o o o o o o o o o Q c w N y n N (3) annivaaawal (van) aanssaaa 8

l lllliI llI' \\ eD wM3H<mNa" g ^O~M%v $5mmW,r.{ 0 0 0 0 0 0 6 2 0 0 0 0 0 0 0 0 1 1 8 4 6 5 4 3 2 1 0 0 0 0 0 0 N s. 0 { , k l t t ,0 L' 0 0 0 8 5 0 4 3_ ,0 0 0 g ,0 7 7 r e 0 ,0 0 6 ,0 6 I 0 )C ,O Q ,0 5 5E 0 6 S C t 4 c' ( 1 E ( RT 7 E US 'r. GE 0M a IT 0s F ,0 I ,0 i 4 T 4 r' 0 ,0 ,0 0 3 3 1 i' E 2 0 R k =, 0 0 ,2 U ,0 T S 2 S E E N R I P L 0 R ,0 0 1 ,0 E L 1 T E N SS E E C I i0 V 0 0 0 0 0 0 o 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 6 4 2 1 4 3 2 1 ^b W ADH<%41~ r m<A' w D$u..L A I

s n FIGURE 4-6 FLAf1E FRONT ACTIVITY TEST 1.6 1}lN.'t.Nkllh. kI.l.M [ !h! / hk 'LijL lillh ;illllliii t 3 ' d 'l lIl lL [l If O W illE L 1 i .g.4...! [.... l l.. ll .l e g i. l g g i i .4.s. g .g ).l

5.. I a ll..

s l .I. .g g.. l-3 I i .{ ^ f L,. $j$ .j .l-l .L . s l.. ,11 e.. l.. t i Li i..f.. L. ]. . j .J..L. .s!_e_ ..L .......LL.. .1 .__f. I t J.,... /. ; ~ p. - ~ m. 3 \\ .,l

• l.]

h. Id '. _I . s. ^ j ~ j~.,kg'~ ~' ~ .4 ~. y.. .1 , _ _,,., i i...i J~.t. ~.'. ~ .- l - j.~ ~~ C'Th"'ll' E'[,, 4 ."- F.l" la.~~~'l.'. l 4.' i ' h'. ' ' l _ sq .~' l -l. --. h' ~' ~ ~ ' [l 'l' l h J p e.. I23865 a Jj 4 1.- .1 3.,.a.. 1. p i .. i...k.. // 3 g \\ p.1; Li .L j.. . l I.. F F F F F .l. a. 6 . sl.... AL. .i 4 i J. .l j . j.. _;.;. '.L.1.'.;dL,1. 4 ..i... .f . _2 k. 1.. a:.. _1 ;. : .I _. a. a a. ..,._t . 2_ u Aa 3. a m. = ' l;tIyL:.I.l ..:.. T.T.lT.g..tt.J.l..EI.'.la.r.F....-. Li.l u i i r J.:r. t' r F F r F .L p Li t ;..! 4 H"I'.fl.b." +. . F FF F F-n u. p .. # i;, -4 i. L t.. a 4.. Ll. 4..L L d. l 4' f -. l lt. .I.'.,.'. e..M' ll 11 4 ' l. I a l..- 4 4... l D WiflEL'3 C "H"'M W m.1 e 4 + + ' ~ " '.'. g { l.. ). p..g .. l g.e. j ). .l ,' [ J :1. - ~,. l. l 6 ,, g. .g. )..

., l.

3. . J l.. .. h .g 18.- ..l :. I ' 1. t.. 4 e.

a. a.

a .i . J.l. 4 e f:. .g. m '8 l.,.I. l.: 1.*p JJ.'.... a L.L.. J ..J Al J,J ,..l.,..,.,., ] - .). . ~.. F F F.L F l l1 ..,), .I . l.1 .g.: 1. - e a. i l. s. i . a g.... ;.. g .. 3.. - L, I .i -g 1l .l ) .......l. .4 l ljf I l II-16 i.* '.*.4 --* - -* 4 ~ 6: 4* g g -b -51 f.Mai .'ilM.'l1}I.$dll] ' ! 'i ~~I' l

• b '

- I k I' I I I I I I ....l..i......, l.. ...l. _ [. p.-p.-, - . l. ) l.: k.. ,s .I' .,I

d.,

L, t. u..,.. ..i 4. .. i L.1 .p . t..

i.

o i )- i .t l.1 L . L.. ..l..l a..p.L 4. .1. l. i.L. . J. l. 4 r y r r r .i. ..l j .....v..~.......~. O .. -. ~. -... .., i y. , o ~.. 3 .,.,,.. g..I. i 6 . ~. -, e F F i F - [ g a i !,t f. .e.t. .i. O W iiEL 5 l l lp.,.i.l " / sl M., aj,j,l,j,',MudliIiiiijiii.jiiilIi(Uin g t

e I

...! 1 ....i

,,iiiiiiiiiiiiiii. m ii

..i. i i 3 N e r s-

PRESSURE (KPA) TEMPERATURE (C) ~ u -e o o o ~ m o s O o o o o o o o o o o o o o o o 1 o i i t o-g m m m m 2 m H s_ r m s_ m t m C C 2 ( w [ m 8_ 8_ _3 d m o.s m 3 .3 u a 8-8- 4. H. o_ c* l n* o. m m A 2 s BR ] 8_ @ 8- _o o P. v ~s I 0~ g~ o l a N~ N~ b $f i i t i e e } g. i r i g-i i i i i o 5 8 8 8 8 8 8 ^ o e 8 8 o o PRESSURE (PSIG) TEMPERATURE (F) g e

e 4: e (DISd) 3110SS311d O C C N e m D O c f f f f o I w !l o u o I i L) 8 a jr 4 O 6 _o i: O Ji i 'l t O -o 1 O ' I \\ o k -o 1 O CD s CO ^Q Ed e-- .o Es oW -) ww C Lu - >= k Q o w ~ 4 o -O C N v C:.1 o D l, 8 m in l Cd Z e -o N 1 Cd m 1 D kJ t \\ i i i i i ro o C o C O C O N N N b O N (VdM) SilGSS311d e s- + ,-,.r.---m, - -y, ,.-w,. o _,__mv.,- v.

4 (orsa) aanssana OJ m C O e f I t t ' O e .ON e* O O O C -O Q C -O C3 m a ~ l @ W (. 8W w

c. v:

Dem M F-W = c-b O -O 0 O ( -O M T p 1 W c:: 4 O -O D M Cn in W e Z a. J O N -O in ~ (n W -o i i i i i i O D C C O C C C OJ O N O OJ (vas) aanssana

_2 m_ .J - -. +. e 6 e e i e e 4 4 (oIsa) aanssaaa O C C N C O c s t t t -t,- O i m.* i. 6 O .O 1 N E i _OC l F r F O 1 O F O e O -C O e ,e -O ' + cc I * ^ O y 73 ' ~ i ~ _g w n cn w L v cc l-t DD CW rg O i 5 ~ / C -O C O -O A, w L a: _C i O l L U w i l D k-m O x -O N A 1 -8 w m m Cd> -O ) i i i 6 I O C C C C 4 O D N O N c N (vax) aanssana

1 ^ - W'a oH<xN1hs 3 Mm d5$m. % 0 0 0 0 0 0 6 2 0 0 0 5 0 1 1 8 4 2 1 5 0 0 0 0 0 1 1 i l-i 9 0 9 0 m, = =

== _y

=

0 = 8 0 ,8 y ,0 7 0 i ,7 1 1 ,w 1 ,0 4 6 0 2 i ,6 E RT )C ) US C GE E I T F F S N', S ,0 ( 0( g I* 5 ,5 E E M M I L T T N T g ,0 ,0 4 4 ,g g 0 3 1

30 E

R h 4T U 0 ,2 S 0 E S ,2 N E R I P L R ,0 L ,0 E 1 1 E T S N S E E C V 0 0 0 0 0 0 0 0 0 5 0 5 0 5 0 0 0 0 0 0 0 8 6 4 2 5 2 0 7 5 2 1 1 1 1 ^ wxDH<%w1hs ^ LE N a$w,. c.

FlGURE 4-12 I TEST 2.2 f . e CENTERLINE T1 1000 -- m b m 800 - - 1600 b g M g i D 600 - - 1200 D D l D '" ~ dW\\Qtf4A ,,f" y 200- -400 N .m.....-- 7N 0-N i 0 10 20 30 40 TIME (SEC) CENTERLINE T4 1000 m m 800 - - 1600 b g Z q( % w. w g-5 o soo- - '2 g 400 - js -800 w j A w p 200 - m a,.. - __ _. - a W -400 7 0-w H 0 10 20 30 40 TIME (SEC) VESSEL PRESSURE 1 m _gg m a M 300 - -50 v CD -40 m Z 200 - p -30 N 2 M I h y 100 - -20 cn 4 - 10 W 0, b ~ -o 0 10 20 30 40 TIME (SEC)

g F1GURE 4-13 TEST 2.3 CENTERLINE T2 1000 - - mS 7 p ~ 800 - -1600 g .t y Z D -g200 D 600 - sy D w n h 400 - - 03 W 'r,% W a - N!I' F' ' T E.8"0I ?, '. c 200- .T - 400 s, t ) qq W L 12 e d, u s. ro%; h W h 0-- i i i 0 10 20 30 40 TIME (SEC) ' CENTERLINE T4 1000 mb m 800 - -1600 b g .u, y {t 'l _geco D 600 - H i Sh,// D 1 p a: F Z 400 - 9 (' - 800 N W 7, 200 - 7., y,yp -400 k;g b,.,i.d w,y ;;?.T.'t. u bg;, h 7 W .r h W 0-b i i i 0 10 20 30 40 TIME (SEC) VESSEL PRESSURE 2 _go - 50 Z 300 - ~ v M -40 6 200 - [' D -30 Cn D g 100 - - :0 m 1 ] g .] - 10 g 0, n ~ O 0 10 20 30 40 TIME (SEC)

FIGURE 4-14 i' TEST 2.4 CENTERLINE T1 1000 -- m m 800 - -1600 b y b 0 600 - -1200 H D [H 400 - i -800 t f %, e .y 200 - Wp -400 } N O,' ' O 10 20 30 40 TIME (SEC) 1 CENTERLINE T4 1000 mO v N 800 - 1 -1600 i g v O 600 - - 1200 H 'g D p N 400 - -800 y y 200 - -400 C J d w = 0; H i 0 10 20 30 40 TIME (SEC) VESSEL PRESSURE 2 g_ ^ _ gg m a O - 50 Z 300 - v CM -40 N Z 200 - D - 30 N 100 - N u) 4 _d - 10 0 1 . _o 0 10 20 30 40 TD(E (SEC)

.r. FIGURE 4-15 FLAME FRONT ACTIVITY TEST 2.4 N. !ll iilli i i i

l..

t l l!ti.I i iI8 CHAfif1EL 1 1 l I I I; l. 7 t l 123f5 l- ) i L //i \\\\ CHAfillEL 2 r r r r r r I. L. .,..a r r r r r _4...[,.__._. , t,. ... l..l.. l.. Il i, .i ~ . L..,_. i I,.".. ~ + ~I + ~ ~ CilAfif1EL 3 m ...i... .i r r r r r 17 - ]: I -Tc 7r i -!,i . hI r r r r r CilA!1f1EL fl - 'LA I i I r r r r r t. i i e i.._ l i _.1 ...j. O, ) ,i, _,,, ...l.,

,....c, r.,,....

I g.6 ! ' ' u. x ,i ', l e i ! ! 'l' i l.., l 4 I i1 i : ,.o CHAfit!EL 5 ' '. l e. i .I !,!,r.l .i.6i. 4 .... {, ,3 e. I e llg. I I il ..li'i !!!!JJI i i e i l., 0 20 11 0 60 80 100 3

FIGURE 4-16 TEST 2.5 CENTERLINE T1 1000 -- ~b m -1600 b 800 - g g O 600 - - 1:10 0 H D H< m 400- -800 W g W -y 200 - %j%- 4 00 7 W H _A W 0b H 10 20 30 40 TIME (SEC) CENTERLINE T4 1000-mU m v N -1600 v 800 - g g D 600 - -1200 H D H d 400 2 -800 y y 200 - -400 7 N y 0; h 0 10 20 30 40 TIME (SEC) VESSEL PRESSURE 2 _ gg n n 0 CL 50 ~ Z 300 - v -40 0 200 - - 30 'd a 5 N s,. - 20 m y 100 - ( 's ,A c j - 10 d S 0, E -0 0 10 20 30 40 TIME (SEC)

FIGURE 4-17 TEST 2.6 I CENTERLINE T1 1000 - I mO v m C: 800 - -1600 v g b g .3 600 - -1200 H = 0 H 400 - ( -800 .L y w r-ce 200 -

• A c

4 -400 W /4 e , h _ _ :',' 4 % }'. H y LQ n... w L H i 0 10 20 30 40 TIME (SEC) CENTERLINE T4 1000-mU C v 800 - -1600 O g Cd 600 - H -1200 ^* 400 - - 800 y 200 - ? d -400 0, l 'D, ~ ,1 c 1 F F 0 10 20 30 40 TIME (SEC) t VESSEL PRESSURE 2 400 - m -60 n g C - 50 Z 300 - - M v -40 O N Z 200 - [' o -30 ] 100 - a - 20 m W c.* N,,, - 10 W w ~- m, 0, T i -0 0 10 20 30 40 TIME (SEC) 6

FIGURE 4-18 TEST 2.7 CENTERLINE T1 1000 m 800 - ~' N D 600 - -1200 { t: 400- -800 4.* W L w p 200 - O '/-400 C. l

4

.W __-w'__ 'N/ L h y 0 -- H 0 10 20 30 40 TIME (SEC) CENTERLINE T4 1000 - 800 - -1600 O g Cl* D 600 - -1200 H C f, 400 - -a00 s 2 200 - - 400 O w jy e H 0, r U 10 20 30 40 TIME (SEC) VESSEL PRESSURN'2 400 - -60 ^ m - 50 2

4 300 -

(n -40 w x 200 - D - 30 W in D 100 - ~ I %~ - 10 Wx 0' i e i -0 0 10 20 30' 40 TIME (SEC)

=.. l l I. j t l (oIsa) aanssana O O N m e c o e n I f I I O 7 O i .O C1 e l b O -3 1 l c .O l O ~ l (J O -O -O O m e O, i cn v

• O r.

N - o @- p ce r D vi OW

• H W

u-iO- ~ = s. 4 '9, 'r O ~ O. y )q CL O D ~8 M M C g -O CL J C:] -8 M M [:3 i .-O O D C C C a O C Cl O C= c CJ ~ ~ J f (vaM) aanssana J c I t ww-= .--,,,,g, e-m- --,y n-r-,,m ++,-y

s s e e m w w w w b w / P-m w O-I mm w w w w N s b b b b h b b g \\ w w 1 f e e. !l . b }Y} f f -.! '.1, ~Il ,t .'f g p>I ,I 'r 't r . - 1. r.- .. {--{ r . + - i r, -.--.{. --..---[ .-. [,. . h...-. 6, ,{,

t. ;-

y-7.-e-r,-.--. ,. g._ -t ry. p" 77 .,. L, ,g, t r- . Tit t --lI

• tt-Ir

-t r -- -- ' F'. f - f yl:l -- -l.. ' F -- l [t l r+ 5 l t

• rt,

-{.. t, --f r r--t

{

- F. - .i 'j t. rr.,-. -1 4.- [ r -. e f- --, -) g ,.},

. [ r,.

.r., 5- --.L L -- tr Ft- ,k-- L .. I.'. l5[ f'p t r - --.r J,i< Tr{ r 7*.. pt-} } ;_ t +r g i r-tr-

g. ;-,f Fg

- 1,7 . Q.i[. .t 1 3 ffI: t .i

j p, t ---;-

.t...-- y 17 -.r _,t .-.7 r 4- [.tr-c _,) o , -~4,. r r-rr-- at ..r,_ . _} ~ r- . nf-e ( l6 -+L j( . }..p{ . t vq ,3 e ...fiI ,g, N s;. + r*- 'r- -A i. +r, r'; rr--rI I I't 'rt: Y, ,r-..l:_- f t - -I ..i g wr-e=, I jij .it-e n o v3 t t -,r.. - r

I J-J' I 1

[ --- { -f r. 4[ 7, rt 7 - r e x w- [ ~ b. t-r f - u. -e r 7 -4 IIY f' f.7.-.h: y 1 ~ff 4 41 ..p .F L. h1 :h ['O t m I f -g gi,dxt

t, 44 F4n gj.4 j

r I ~ m n m.-..r[ f -.r..., ..-} t. t [. {...f. ~. l i ~ 1 L i

t. t.

T i~ I FI! !,y I h l l-lib ll[E ~ -~ .r. j[. .E, i - j . +. .., 4 T 1 r' [.. L.[ .h

• 4,-.

Ie t y[ [ - - ( -. . [. - [ e-,} i -l r t .tIt l I --. 1 t [. i rrff ri. r. 'h-.- ' *-*, r !5 m , !,. { [ f , - - f-4r' 'r -' rn[-.rf .t g f a { ,+t 4 i ' t4 . g a ' r-r r n,-r t +-- r "T l- - ~ ' - ~ s 7 r: ,e7-- .5 L { 7_tr 5

. ur +:r m

-p u 7 ..r r . l. ...:s.pi t 4.- - .r rr,. ll I r l tlt1 r-1 1 - t ilf_.,r,. ;l{ f ,-{- 'l t gg6 t. I ':n I r, t l [

c. ' <..l

[+ 1 rr i ' t r. .r u.. s J J J J W W W W W Z Z 2'" Z Z t = = Z Z = I I b I._ 4

4 O a (a) annivuaawal (olsa) aanssaml O O O O O i o ao o n ej OJ c o o O c C C N 6 g t__O t t e o 1 1 e 1 0e Ti-u- te,?, e e -C L CO <e O -e 1, D c 4:-- u 5 7 8 a.. 8 e _3N G-fr 4 ~. h4 _8 O n N cM 85-4 ).- d 6-ae . f m am v y; N b e2 ,[ o! zy -o-Q'r h 4 -o-F b 7 -~ m. It 4'c, e -o e n -o s n 44 I' a4 N C:) t'~ o -o p o V t,~; -o m ~ N rLE W l ~34 I4 Z i= +7 a: c H

rt

- 8 CL -8 '(e- ~ ,_] N N eZ m W Wu m. W .c ,- c } --,- e r-o e o c c c C O d O O si C D 02 O N D N d Cl O s1 (a) annivasanal (vdN) aaassa:Ia 1 I e

.. ~. FIGURE 4-22 FLAME FRONT ACTIVITY TEST 2.11 s. ) .l.f .. '_ f. ) f 3.. a.

4. f,..

4 I JJ 'I d..}. .,.h... 4... j.,. g. j > /,/ \\, \\ j :F.. ..,l ,._I a,.l g 4 6. 4.,. .l I e. e. I g, [. / 44 y. .i t. 4.. '.I j._... I. 1'

1l.i.I] 1:l- ' [ ".t.11:.

.:.! : Ll

• 1

.l:

4. _..

.l.). H3_ .. _ q..J ..4 !g.7 y ,9. .g _,y4 4

1. 4

.F{ 13{!.Q.jpl.J ' ! I'JEL.M L j.{- r r r r r m p:E42:. f:.l':.F 2TJ.'_1El: 7. If : i8+ '

n..

? a n.. -1"1:.t.1.t"t:.11. _b:_4,i J. l

'. f 1,:l: ..tl.1, -l 1 [- l t.. L. ii t CllAfillEL 3 1

. u n.. ~' t,

{. :;bm.

r r r r r

...t;

..t: -t -. j. .{.g f g ..l..j. g. ]_.. [ .l.. ., ]. j.. ......__.;.__.=_q.,_....7..._, g g g g y .i. 7_ _.i I I , I_' III .I _w CHANNEL 11 +- - -+ innuancisumaciinam n i b!:d:h:--.~.:F3:]Jl l r r r r r It - + I c 1. ~ 1. rc Tr.:r.1.a 1 P l,,,, 6 I t i N:p:.:rrro Eng.p u a

ra

. i y a y E:IRifr!MfhP~.1 !:i lT L O, CllANNEL 5 p.. .t n,. Y:.T..&:h._.llf,..i..:..:.::..Yi..'......N ..j.j., 9r r r h

s / .S>- a,- w m-- w w - O-m,m = N

• i l

i j.i -f.}p MM4D f .'.I

• --} ',t p.-

[ ' t g .. p .t-1 I i; r .1 :.- : _:. lr 1 t t r: - 1 _7:-

4. 7.

,g .[ - [. ,j [r 't .p .__-..g ,L -.7 -} __-.t_ 3- . 4 } r' . u. . p.-. .r ,f r[1 If F II - ft ri - - -.[ r 'il.7 77., { ....7_. =;. [r 4.p _{_(( .- -{ ,1, } r.- -t r 'I- _.-..{ --.4[, y .'.~. ].f _7_ f f l.-- 7 }t i. t r t tt

7. [.

. q., _..p. r.- -.I -I . -t- -} r r;t* ... { j.7_. ._4--._ __7.,; [. -l.}}. }( -., y. 7 t. _,. y . J Ii .L g cn _ L,,.. pi r } - _} - r+; F }- rt .... e- ' l. r t. .r--**- -r p. L. tr b .,t t..! l' t : - . 4.1 j. 4 ;- :..t N-N L.4.1.1 f y '-- }- . [. I,t t -. ~.. k i r ' p l i, '- [fff 6[p -}t f m r, - 1 - 'I-fr

-t '

-g'r-u t-[ L-.'Ik-r wH h.'. l.t '~ ,T 1 I'-- tr' Fr - ' r t - } -'" 6 e=g 'I f _,r.F r }... pttl. %[T,.v' [' t ' ' l [', =i o m lt IFgt,/ t t r - t .L- ~. 7. %._ 5 .,b[Jl.[1 f --' -'-{_+L.K t,. r IE

[_.

77 3 ,,.7-Mli (('] ll ? II' W E . :. t i]g-r' i Fr t ' r-- i tt't d ' e ri. t, r I3 .t'. :thr y, I t id r-i, 7 7, Il +!t !i '!. $r.,.. .. a 1,p . 4r 'phh --H- ., i ;-.tf .I sp w.}p F- _2-r 4- !*i [ 4 4. '.r . 1. m l '.irr -

i r Y r-*r rrr r-it'!

rr p ' t L, l

t. : r tr. ;_-[ i t,

,t! ' rr.. t,t_t_ r . I.t ... i n.p., .p4.t{.0 I. 4 r,e 4_.e t .E.I l rd, t I .p. s _n, 3 } H-y ,}P t .i I. L.,r- - t-ria 4. 4 i a-f tt!+' U . i-I c-1- j ':: (( r r-

t-t.

er-r . [-f =. : T, i r t r; r . e }- t t rr 1 t fIfn I . n. er r~ .t r !.r. ? r F r.' - tr-- r.1-t.r_u#-r.errfr. ---. 1 tT.r 4 L, r' - -{I - D r r 'rtj r-r-- t-t p,_r. rvr.r , 74

,.I.

-l -a i. ,,t,t ' r r.*- t. .p .tr7 r.g.r i_...e.r. 9 Frer tre

-i i. 7 t.r p

, r.i.r_i..I I rr_r+S.,r tii t y.r - 77, s,, 7 ~ F -i -t 4 . h '. M ., },.CS 7 [I ( I .4[,k_.., . r.t f r f-ri p .i n _n [ r , j.. 4 L -.p .t y ..r ~, t'- I 'o.. .E.r n {, f. ,1 j f. ,...r.r .1 ! 1 l s i L _. l CN U J J J _J LJJ LLJ LLJ LLJ Z l2'" Z Z Z Z Z Tr.- = = = = U U U U 9

4-Section 5 Conclusions 1. Location of an ignitor within the test vessel does affect the characteristics of hydrogen deflagrations. Lowering the ignitor location produces milder pressures during hydrogen combustion. This appears to be a result of increasing the fraction of the vessel volume exposed to upward propagating flames in lean hydrogen concentrations. The addition of steam and water sprays also reduced the pressure rise. However, the lower ignitor locations still produced milder pressures than the top location. 2. Fogs were thought to reduce the pressure rise resulting from hydrogen combustion. This was the case for dynamic tests, but not for quiescent tests. Water fogs apparently enhance the rate of combustion. Thus, heat transfer is not as significant in quiescent tests causing the deflagration to be more like an adiabatic deflagration. For dynamic tests, water fogs promote mixing and allow ignition to occur earlier, resulting in lower energy release rates. .A_ 4 I l l

.e APPENDIX A l I I Gas Chromatography Analysis Test vessel s.mples were obtained through nonheated k inch stainless steel probes located near the top and bottom of the vessel (See Figure 2-1.). Each probe was connected to the inlet of a Thomas diaphragm pump. Vessel isolation was provided by solenoid valves. A inch stainless steel line connected the discharge of each pump to a stainless steel condenser coil submerged in an ice bath. Tarred silica gel columns were located at the outlet of the condenser coil to remove any remaining moisture. inch stainless steel tubing carried the sample from the silica cel columns, through a dry gas meter, and into a 250 mil glass bulb. Thermocouples monitored the inlet and outlet pressures of the gas meter. Solenoid valves isolated the sample bulb. When the sample was collected, the sample lines were purged. Sample analysis was conducted by using a Carle Model 8700 gas chromatograph. Instrument specifications are presented in Table A-1. This unit was equipped with a thermal conductivity device. The output was recorded with a Linear . a-Instruments Model 252 dual pin recorder. All samples and calibration standards were analyzed using repeat injections. A ten foot by 1/8 inch 0. D. stainless steel column packed with Molecular Seive 13x, 80/100 mesh operating at 195 F, was used to separate the component gases. The gas chromatograph operating . conditions are presented in Table A-2. Calibration standards consisted of several known concentrations of hydrogen (0.595%, 5.12%,13.19%,18.27%,

4 0 balance nitrogen) and oxygen (5.0%, 18.1%, balance nitrogen). These standards were analyzed at the beginning and end of each sampling day. ,v'~ l e

TABLE A-1 GAS CHROMAT0 GRAPHY SPECIFICATI0f45 Sensitivity Variable, two level detector sensitivity switch Attenuator i 1 Eleven step binary type, 1024 to 1 Detectors Dual chamber, 100 1 volume with 8-10K, 0.013 dia. matched thermistors Power Requirements 115V, 60 Hz Temperature Control Ambient to 200 C G _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - _ _ _ - - - - - - - - - - - - - - - - - - - - - ~ ~' -

f.. 1 /- f a TABLE A-2 i GAS CHROMATOGRAPHY OPERATING CONDIT0NS Argon Carrier Gas Flow 20 ml/ min @ 25 psi Oven Temperature 90*C Column 10' SS tubing with molecular sieve 13x, 80/100 mesh Detector Temperature Low Position Attenuation Variable 4 i (- l l l i t

,o m Attachment No. 4 to AI:P :t'RC: 00500G Donald C. Cook Nuclear Plant Unit I;o s. 1 anii 2 Additional Inforr.mtion on Ilydrogen Mitigation and Control Poport on Igniter Performance Studic', W __--._J}}