ML17319B195

Combustion Behavior Study of Glow Plug Ignitor in Hydrogen- Air-Stream Mixtures

ML17319B195
Person / Time
Site:	Cook
Issue date:	12/31/1981
From:	Hudson F, Shiu K, Wilder J AMERICAN ELECTRIC POWER CO., INC., DUKE POWER CO., TENNESSEE VALLEY AUTHORITY
To:
Shared Package
ML17319B196	List: ML17319B195 ML17319B197 ML17319B198 ML17319B199 ML17334A364 ML20041B520
References
NUDOCS 8202240175
Download: ML17319B195 (26)

Text

l~ ~

P

~

COMBUSTION BEHAVIOR STUDY OF GLOW PLUG IGNITOR IN HYDROGEN-AIR-STEAM MIXTURES INTERIM PROJECT REPORT December 1981 Prepared by:

K. K. Shiu, American Electric Power F.

G. Hudson, Duke Power.

Company J. J.Wilder, Tennessee Valley Authority Prospect conducted by:

Whiteshell Nuclear Research Establishment Prospect Sponsors:

American Electric Power Service Corporation Atomic Energy of Canada Limited Duke Power Company Electric Power Research Institute Ontario Hydro

. Tennessee Valley Authority 8202240175 8202i7 PDR ADOCK 05000315 P.

~

~

g~.*

r wW). i f~'4g

) ~,~$ >~ed'a g-'

~

~

A Cl TABLE OF CONTENTS 1.0 Introduction 2.0 Experimental Setup 3.0 Instrumentation

3. 1 Pressure Measurements 3.2 Flame Arrival Detection

3. 3 Chemical Analysis 4.0 Experimental Results r

~

w ~

~*., + i ~

~ ~

~

~

p w

~

~

'i ah

~ e~'I

1.0 iNTRODUCTION One of the research efforts undertaken at Mhiteshell Nuclear Research Establishment pertains to investigating the effectiveness of the glow plug igniter in a more detailed and comprehensive manner and of other potential.igniter types.

This report presents the GM AC Yodel 7G glow plug igniter test results observed to date.

2.0 EXPERIMENTAL SETUP A 17-litre quasi-spherical vessel with a pressure rating of 600 lb/in2 (4 MPa) was used in this study.

Figure 1

shows schematically the vessel and the components used in these experiments.

The vessel has a

pair of 3.9-in.

(100 mm) diameter viewports on a horizontal axis for flame visualization and photography.

There are two 3/4-in. (1.9 cm) pipes welded to one of the convex walls.

One of these is used for gas injection and sampling.

A branch from this pipe is connected to a strain. gauge (Data Instruments, Incorporated Model RS-101) for measur'ing the static pressure before and after ignition tests.

The other 3/4-inch pipe is not used in these experiments and was capped.

The other convex wall has a 600 lb/in (4 MPa) safety rupture disk.

2.

Figure 2 shows a schematic of the gas,supply system.

All lines are standard 1/4-in.

(6 mm) stainless steel tubing.

Steam is provided from distilled water in a 100 ml.flask heated by a hot water bath.

The vessel and gas pi.ping are electrically. trace heated to 176-212 F

0 (80-100 C) to prevent steam condensation inside.

Eight

~

~

chromel-alumel (type K) thermocouples attached to the outside vessel and pipe surfaces are used to monitor this temperature.

3.0 INSTRUMENTATION The instrumentation is shown in Figure 1.

Specific components described below are numbered in the figure for clarity.

An ionization gap probe (No. =1) coated with sodium bicarbonate (NaHCO ) is supported by a 1/8 in. (3 mm) steel rod (No. 2) at about 1.58 in.

(4 cm) below the upper wall for detecting flame arrival.

The support,rod is mounted on the bottom flange.

e A similar rod (No. 2A) screwed to a strong magnet (No. 7) is used to support the following:

a.010 in. sheathed K-type thermocouple (No.

3) to measure gas temperature, an ionization gap probe (No.

4) to determine ignition and the Rl AC Model 7G glow plug (No. 5).

The ion probe and thermocouple are located approximately

.32 in.

(8 mm) above the glow plug and at an angle of 45 away from the central axis.

Two type K thermocouples (No.

3A) have been spot welded to the bottom surface of the glow plug to determine its temperature history.

As seen from Figure 3, this assembly is bound with high temperature tape and inserted into the vessel such that the glow plug is horizontal, about

.39 in.

(1 cm) above the lower edge of the viewport to permit flame photography.

A small fan (No. 6) with aluminum blades 2.76-in. diameter (70 mm) mounted on a magnet (No. 7) is available for forced convective flow experiments.

All of the electrical wires (No. 8) penetrate the flange

via Conax fittings (No. 9).

A Kistler 603B1 piezoelectric transducer (No.

10) is flush-mounted on the flange to measure the pressure transients.'

slow response 1/16" type K thermocouple (not shown in figure 1) located near the lower vessel wall level is used to measure initial and final gas temperatures.

Although the two thermocouples, which are spot welded to the bottom surface of the glow plug, are similar, they do show a somewhat different response

time, as shown in Figures 0 and 5.

The higher of the two glow plug temperatures is consistently used for glow plug surface temperature evaluations.

The thermocouple for gas temperature is clearly too slow to properly determine the peak temperature.

However it does provide clear indication of ignition.

Maximum gas temperature is more reliably obtained from the application of the ideal gas law based on the more accurate pressure measurement of the strain gauge pressure tranducers.

3.

1 Pressure Measurements A Data Instrument, Incorporated, Model 101-25 strain gauge transducer is used to measure the initial partial pressures and the final pressure.

This transducer is connected to a Data Instrument Model (RS-100) digital readout system calibrated to better than+.02 lb/in (0.133 kPa).

2

~

~

v

~ >>r

~

p lt ('b't

~\\y

" l v

e

V

~

~ r P

/

h%

vvlp

'hV ls I tv

\\

Since the strain gauge transducer will record up to 25 lb/in absolute 2

(170 kPa), it also serves as added confirmation of pressure history measurement.

For fast pressure transients, a Kistler 603B1 piezoelectric pressure transducer, flush-mounted on the flange and connected through a Kistler 504E charge amplifier to. the recorder, is used for pressure measur ements.

3.2 Flame Arrival Detection The two ionization gap probes, used to detect flame arrival, are connected directly to the recorder.

Although in principle they can be used for flame speed determination, the slow recorder speed limited their use to confirming combustion.

)

The gap probes are operated at 200V dc.

Each probe has a single RC coupling circuit.

Since hydrogen flames produce few ions, the probes are coated with sodium bicarbonate to provide detectable signals.

1 The central probe tends to pick up 60 Hz noise more readily than the upper probe as a result of its close location to the 60 Hz power source for the glow plug.

However, the combustion wave signals were still strong

enough, even for lean mixtures,to determine flame arr ival.

~

~

~ ~ 4 II

~ ~ l

'E J1 ~

s r / wv r (

~ )

~

) 8, et t

rw os

'ql ar l Sr

~1' r a

~ e 9

'4 r

3.3 Chemical Anal sis Partial pressure data have been used as the principle method in determining initial concentrations due to its reliability and consistency.

Also, when available, the mass spectrometer has been 7

used to measure gas concentrations before and after ignition.

Like ij C

I I

most commercial systems, this device does not measure steam concentration.

Nonetheless, the extent of combustion can be readily evaluated by comparing the hydrogen/nitrogen ratios before and after ignition.

4.0 EXPERIMENTAL RESULTS Host tests were per formed with quiescent hydrogen-air-steam mixtures using a 14V ac supply to the glow plug.

A limited number of tests have also been done with a 12V ac supply and with the fan on.

Typical measurements made in each test are presented in Figure 6.

They include the two-time traces of the ionization probes, glow plug surface temperature, and the pressure data.

The extent of reaction can be estimated from three different measurements:

'(a) mass 'spectrometry analyses before and after

reaction, (b) static pressure measurements before and after ignition, and (c) peak pressure measurements.

From the mass spectrometer measurements, the extent of reaction is defined as follows:

1 - (H2/N2) final extent of reaction

=

(H2/N2) initial r <<l,, ',,', !~A~),

~'," ' i, w + ~r',

~

~

i wri P,

~ <<(,g\\'

g ~'h ;(1P<<

I;.('>I>gg/ <<st(g>z>](f'l(%IT/'Vw',5t ~) ~i'L Q>P Ig>"

V 'ba

~;~

r ~ late

e

The estimate of extent of reaction from static pressure is based on the fact that there are two moles of gas after ccmbustion for every three moles of fuel and oxidizer reacted.

The following formula for the extent of reaction can be derived:

2(P.

t

/P

)

extent of reaction

=

,2 initial final Where H is the volume faction of hydrogen prior to combustion.

Because the differences in pressure are small,. the calculated results are very sensitive to the measured pressures.

Figure 7 summarizes the results of the experiments.

Ignition criterion is defined as detection of flame arrival by the upper ionization probe.

For marginal ignition, the pressure and temperature rise were small and often barely recorded.

For a pressure rise less 2,

than 1.8 lb/in '12 kPa) but with a detection of flame arrival by the upper ioni ation, ignition is defined as marginal. It can be seen that'he ignition region Omit is similar to flammability limit curves.

Figure 7 depicts the ignition Umits of hydrogen air mixtures in various steam concentrations.

Ma)ority of the data points are for static mixtures; turbulent mixture tests are still being conducted.

The range of steam concentration investigated varies from 0 to about

60 percent.

At about 55-percent steam concentration, only marginal ignition was observed for a hydrogen mixture of 20 percent.

The initial pressure for these tests was essentially atmospheric, and the maximum pressure observed was about 54.7 1b/in g (372 kPa).

The surface temperature of the ignitor at which ignition occurred was r ecorded for each test.

The results are presented in Figure 8.

Xt ia obvious that as the steam concentration increases, so does the ignition surface temperature.

The maximum temperature observed was about 1560 F '(850 C).

Preliminary evaluation of the experimental data indicates that results obtained:

(a) are consistent with glow plug data from Fenwal and (b) are consistent with other hydrogen-air>>steam mixture data with other ignition sources.

E71344.01

<<W ~

~ ~,

~y <

1raiyr" ~ .'i,'ts/Pty~t 9T<"'l ~gP;b4~'.'~'tr'i%I" '4 "~+'~'I 5>> '5't.">>' "J' QI3 Qz I / 4 Q2A Qe IO cm ELECTRICAL WIRES Qo THE SCHEMATICS OF THE INSTRUMENTATION FOR TilE IGNITIOIQ EFFICIENCY TEST USING A GM AC BO. 7 GLOW PLUG IN A l7-LITRE VESSEL. FIGURE 1 VESSEL steam strain gauge sample mercury vaccum guage air vacuum pufllp GAS SUPPLY SYSTEM FIGURE 2 i A i CQ~$,.~~ gee ~ ~ %s THE GM AC No.7 THERMAL GLOV/ PLUG EQUIPPED V/ITH DETECTORS FOR TESTING THE IGNITION EFFECTIVENESS AND RELIABILITY. FIGURE 3 ~ I

T

~ ' I'>> ~ ~ ~ I (( I ~ ( ~." ~ I ~ Qc "?I'i i"j " ~ \\ ~ I t ~ ~ (( il ~ (..I <<" '::il'.:i: ".::I.".".. ~ at I ~.' ~ 1 3 'C 0 l ~ ( ~ ' ~ 1' 4 ( ~ ::I(:: tooo ,l... ~ ~ 'I ~ ~ I ~ ~ ~ ~ ( ~ ~ ~j..: ~ ~ ~ \\ fi~ ~ (l(l..(( ~ 1I I I aca I?ii,'il ~ I

~ I

~ s.(.I I>> ~ (( ~ a (r ~ ~:I 't ~ ~ l'l?,'iiii:iii! ~ ~ t ~ ?ii:I i i::.I'll;;!i:; ~ ~ I ~ ~ ~ ~ :: I ~ ~I: C ( ( ~ Vl 1 II !?iliii' ~ ( ~ ~ ~ ~ ~ ~ I ~ ~ ~ ~ ~ ~ I I ~ I I:~ ~

.::ij i:?I FIGURE 4

I::Ji.:i:jill: h CFNTI;AL10'ZPTION PI.OBL= J I. Jj L!3 C3O C ~0 P 2~ UPPEfi IONIZATIONPI.QQE 0 C) GLO("J PLUG SUHI=ACE TEMPERATURE I O 0 hJ 0 5 t J< d Q 2-0 STPAIN GAGE -e7.2 I;Pa ( ABSOLUTE) PIEZOELECTB IC PRESSURE TRANSDUCER 0 I 1 I 27 2G 25 24 23, 22 TII'AE 'AI=tEH TUHNING ON THE POV/EA / SECONDS FIGURE 6 OBSERVED TPANSIENTS IN TlIE IGNITER EFFICIENCY TEST. A GM AC MO. 7 CLO'VI PI VG OPERATED AT I r VAC IN A 126% HYDROGEN/ 2Q.5%- STEP,M Vi(gq ) URE IN A 17 I ITIVE ~ ~ ~ V ~ =. V ~i ~ ~50 I do IGNITED MARGINALLY NO GAS OPERATING IGNITED IGNITION CONDITION VOLTAGE 6 8 0 STATIC I 2 VGLTS g STATIC IO VOLTS V v '7 TURBULENT I4 VOLTS a 15 O 0-LQ 1 0 cf ~ IO OO 7, 37,38 14 28 PROPOSED LIMITS FOR STATIC MIXTURES 5T PROPOSED LIMITS FOR TURSULENT8I MIXTURES 4F 8 5 d4 o d28 a) 2 I Q 4s'2, 49,38 CQSd 9 13 QQ~ Qg s7 s 20 29 C v 11 ~ ~34'LI C9O 31A Ci 32,39 30 22 V 24 V 17 27 IO 0 20 30 00 50 6 STEAM CONCENTRATION (%- BY VOLUiVIE) FIGURE 7 THE IGNITION LIMITS OF HYDROGEN/AIR/STEAM MIXTURES USING A GM AC t%0DEL NO. 7 THERMAL GLOW PLUG LOCATED AT THE CENTRE OF A l7-LITRE QUASI SPHERICAL VESSEL. 1 ~ 900 f600 800 O O X. ~ox. ISdO 0 CL'l 700 O I 600 14V quiescent marginal comb. with fons + 12V IO 20 30 40 50 STEA M CONC EN TRATION, v/o FIGURE 8 /~d /34) /280 Q lpga Plug le ptra+uvc a+ Za ~ oe ~+e Co cc + ~ o Attachment No. 5 to AEP:NRC:00500G Donald C. Cook Nuclear Plant Unit Nos. 1 and 2 Additional Information on Hydrogen Mitigation and Control Repor't on Hydrogen Combustion Phenomena Studies