Anda di halaman 1dari 8

Int. J. Hydrogen Energy. Vol. 7, No. 9. pp. 729-736. 1982. 0360-3199/82/090729~)8 $03.

00/0
Printed in Great Britain. Pergamon Press Ltd.
~) 1982 International Association for Hydrogen Energy.

CATALYTIC COMBUSTION OF H Y D R O G E N - - I I . AN EXPERIMENTAL


INVESTIGATION OF F U N D A M E N T A L CONDITIONS FOR B U R N E R
DESIGN
M. HARUTA,Y. SOUMAand H. SANO
Government Industrial Research Institute of Osaka, Midorigaoka 1, Ikeda, Osaka 563, Japan

(Received for publication 7 December 1981)

Abstract--The performance of catalysts in different forms was investigated for the design of a catalytic combustor
with hydrogen fuel. The catalysts tested had dimensions of 150 x 150 mm, and consisted of a ceramic honeycomb
impregnated with Pt, two Ni metal foams coated with Pd powder which differed from each other in pore size,
and a ceramic foam coated with Co-Mn-Ag oxide powder. In the diffusive mode of operation, the Pd-coated Ni
foam with larger pores exhibited the highest combustion efficiency. The ceramic foam with the oxide coating also
provided smooth hydrogen combustion in the range 0.2-1.0 kcal cm-2 h -t. Combustion efficiency was improved
by increasing the amount of premixed air and totally supplied air. Spot measurements of surface temperature and
gas composition were carried out over the catalyst surface and the characteristic features of each catalyst were
compared and discussed.

INTRODUCTION EXPERIMENTAL
Previous work in this series has shown the results of The performance tests of four catalysts (dimensions
catalyst screening for low-temperature catalytic com- 150 x 150mm) were carried out using 99.99% pure
bustion of hydrogen [1]. It has been found that several hydrogen and the apparatus shown in Fig. 1. A ceramic
inexpensive oxides of 3d transition metals, as well as wool mat adapted to facilitate uniform dispersion of
platinum-group metals and their oxides, are active at hydrogen was disposed opposite the front side of the
intermediate temperatures around 150C and that pal- catalyst. Combustion air dried by silica gel was fed to
ladium metal supported on silica gel can initiate hydro- a premixing chamber concurrently through two tubes
gen combustion even at temperatures below 0C. and to a gas outlet chamber tangentially through four
The present investigation was undertaken to see what tubes. Spot temperature and spot composition of gas
catalyst structure is suitable for a catalytic burner with mixtures on the front surface of the catalysts were
hydrogen fuel. Although the use of catalytic combustion measured with nine chromel-alumel thermocouples
in heating appliances has invoked much interest [2-7], connected to a digital multi-temperature recorder (Yok-
little attention has been paid to the effect of the catalyst ogawa Electrics Works Ltd., Type 3874) and with nine
shape on the performance of a catalytic heater. microprobes for gas sampling connected to a gas chro-
Development of practical catalytic heaters requires matograph (Yanaco, Model G1800). The local spots for
the determination of optimum construction and the both temperature and gas composition measurements
knowledge of operating properties. The shape of a were numbered as shown in Fig. 1.
catalyst body is assumed to be one of the major factors
affecting the performance of a catalytic heater, because
Secondary
distribution of fuel over the catalyst body and diffusion air
of air to its surface against the bulk flow of fuel controls
the operating properties in the diffusive mode of
$ ~ cos
operation.
In most commercial and prototype catalytic space
heaters [2-5, 8], asbestos wool, ceramic wool, fibreglass
or an aluminium sheet is used to support the catalyst.
Unfortunately, information on the effect of catalyst
structure is incomplete and no type of catalyst body
other than a fibrous pad and a metal plate impregnated
with platinum-group metals seems to have been used
t t t
Primary Secondary air
for a diffusive catalytic heater. air
Therefore, four types of catalysts (all non-fibrous) Fig. 1. Schematic diagram of catalytic combustion apparatus.
were tested. Their operating properties are described A, ceramic wool mat; B, catalyst body; C, thermocouple; D,
and discussed in the present investigation. gas sampling probe.
729
730 M. HARUTA, Y. SOUMA AND H. SANO
Table 1. Catalysts used in performance tests

H2 comb. eff.
Catalyst Catalyst Preparative Dimensions at
Catalyst material support method (l x w x d) (mm) 1.2 kcal cm -2 h -1

Ni foam No. 2? Coating of Pd


02* Pd (11-15 cells inch -1) fine powder 150 x 150 x 5 98.0
Ni foam No. 5t Coating of Pd
Ds* Pd (35-44 cells inch -1) fine powder 150 x 150 x 5 88.4
Ceramic honeycomb Pt was impregnated
T,~ Pt (14 cells inch -t) from its HCI solutions 150 x 150 x 25 87.5
Co-Mn-Ag oxide, Ceramic foamll
20: 4:1 Coating of
Go (atomic ratio) (11-15 cells inch -1) oxide powder 150 x 150 x 5 93.0

* Available from Daido Steel Co. Ltd. (Japan).


t Available from Sumitomo Electric Co. Ltd. (Japan).
~: Available from Tanaka Precious Metals Industry Ltd. (Japan).
Active oxides developed by our laboratory. Further details will be published in a subsequent paper.
II Available from Bridgestone Co. Ltd. (Japan).

The combustion efficiency of hydrogen and the equiv- on the catalyst body, amount of air supplied and oper-
alent ratio of air at the catalyst surface were determined ation duration. Figs. 2 and 3 show how the catalyst
by analysing the concentration of hydrogen, oxygen and surface temperatures on the top and bottom of the
nitrogen with a gas chromatograph. Since water was the catalyst body change with time of operation and how
only product of combustion and was absorbed by silica they are dependent on the amount of air supplied and
gel before a sampled gas was injected into the gas type of combustion, i.e. diffusive or completely pre-
chromatograph, the efficiency of hydrogen combustion mixed combustion.
was defined as: After self-starting from room temperature, the sur-
face temperature rose more rapidly on the metal foam
Hydrogen combustion efficiency (E)
catalyst D2 than on the ceramic honeycomb catalyst Th.
= (1 - hydrogen slippage) x 100% For both catalysts under diffusive operation, at least
30 minutes were required to reach the steady state and
IN2] - [ 0 2 ] x 100% (1)
[S2] - [ 0 2 ] + [H2]
where [N2], [02] and [HE] are concentrations by volume 200 . . . . , . . . . , " , , , i . . . .

% of nitrogen, oxygen and hydrogen, respectively.


The equivalent ratio of combustion air to all supplied 150 . 8
fuel was defined as: lffusive
combst.
Equivalent air ratio (X) f 100 2
3
1
= (2) 50 i
[H21 1 g8 [ 0 2 ] o. 1.0 1.2
1+1.88~-2x ._ IN2] E L

Heat inputs were inferred from the calorific value of '~ t i i


hydrogen and the gas flow rates, which were measured
by rotameters fully calibrated over the range of experi-
ments. Table 1 gives a list of the catalyst bodies exam- u Premixed !
ined, with a brief description of each and the efficiency 100 co mb st. [
of diffusive hydrogen combustion at a specific heat input i !
of 1.2 kcal cm -2 h -1.
1:o8 i 1o
RESULTS A N D DISCUSSION 0 . . . . i . . . . . . . . , . . .

50 100 150 200


Time course of catalyst surface temperature Time, rain

Surface temperatures of the catalyst bodies are Fig. 2. Change in spot temperatures of surface of catalyst D~
dependent on heat input, type of catalyst body, location with time of operation. Heat input, 0.4 kcal cm -2 h -1.
CATALYTIC COMBUSTION OF HYDROGEN--II 731

In completely premixed combustion, at any ratio of


100 air, the spot temperatures were always higher at the top
than at the bottom. The temperature difference between
~ f u s i v e 1 the top and bottom of the catalyst body was considerably
5O A,_ i i i combst.-]
larger for the catalyst D2, which was composed of
? ~ 1.0 i 1.2i 1.3 } three-dimensionally distributed macropores. Over this
3 catalyst, the mixed gas could flow upward along with
:::::::::::::::::::::::::
0
temperature gradient and burn at the top. This may
o.
E1001- Premixed. ! ! 2 lead to a further rise in temperature at the top. On the
m ~ comDst. I
other hand, vertical flow of the gas mixture was inhibited
over the honeycomb-type catalyst by the two-dimen-
sionally arranged channels. Absence of vertical flow
1.0 1.2 1.4 0.8 could account for the relatively small temperature dif-
ference between the top and bottom of the catalyst.
0 50 100 150 200 250 On the ceramic honeycomb, surface temperature con-
Time, min tinued to change with time for a long time, especially
in premixed combustion. However, as shown in Fig. 3,
Fig. 3. Change in spot temperatures of surface of catalyst Th in experimental runs carried out after the run at a larger
with time of operation. Heat input, 0.4 kcal cm-z h -t. heat input or larger air ratio, surface temperature
reached a steady state in a short time. The combustion
the spot temperature was higher at the vertically lower efficiencies shown in Figs. 6 and 7 were obtained at the
position of catalyst (No. 8) than at the higher position end of each run. They varied only slightly with time
(No. 2) with air below the stoichiometric ratio. With after 30 min although surface temperature was still
air over the stoichiometric amount, this temperature changing to some extent.
distribution was reversed. Fig. 4 shows the spot surface temperature on the
An increase in secondary air from an equivalent ratio centre of the catalysts in diffusive combustion as a
of 0.8 to 1.0 may cause a drastic change in the distri- function of operation duration. Since a sufficient
bution of hydrogen and combustion air over the catalyst amount of air with an equivalent ratio of about 1.5 was
surface. At a low air ratio, it was probable that the supplied, the surface temperature rose rapidly on
distribution of air was localized on the bottom of the initiation of combustion and reached nearly constant
catalyst body due to its specific gravity being remarkably values within 20 min, except for the ceramic honeycomb
larger than that of hydrogen. Therefore, combustion catalyst. The steady-state temperatures changed with
of hydrogen may take place predominantly over the heat input but were almost independent of the type of
lower part of the catalyst. When a stoichiometric catalysts. By taking the time variation shown in Fig. 4
amount of air was supplied, it could also be distributed into consideration, the following spot measurements of
to the upper part of the catalyst. As hydrogen was catalyst temperature and gas composition were per-
assisted to flow upward through the catalyst body by formed after 1 h operation.
an air stream, combustion at the top predominated.

L r , i Effect of total amount of air supplied


kcat/cm2-h Figs. 5 and 6 show the effect of the amount of air
..... . . . . . . . .
supplied to the chambers on the combustion efficiency
of hydrogen. The dotted line ( . . . . . ) shows the the-
0.8 k~t/cm2-h

,00
oretically obtainable combustion efficiency. It was
found that, at a heat input as low as 0.4 kcal cm -2 h -1,
0++ [+rYII _ . .. _..._.... ..-.- .- -.-...- .- -...- .- -.".

_ 'U
i
0.16 kcal/cm2-h ~. 90 -,~

0 - - l I I I I

0 20 40 60 80 70
Time, min
0.8 1.0 1.2 1.4
Fig. 4. Change in spot surface temperature on the centre of Totat air ratio to fuel
catalyst in diffusive combustion as a function of operation
duration. Secondary air supplied: ;t2 = 1.5; except for catalyst Fig. 5. Effect of equivalent ratio of supplied air on combustion
D5 at 0.8kcalcm-2h-] where ;t2 = 2.0. Catalysts: O,O = Ds; efficiency of hydrogen. O, m: Premixed combustion; O, r-l:
i7,1 = Dz; A = Go; O = T,. diffusive combustion. Heat input, 0.4 kcal cm-2 h-L
732 M. HARUTA, Y. SOUMA AND H. SANO
.... p .3-- ~.-'~- accompanied by high catalyst temperature yielded a
100
/! ~';,~" high combustion efficiency of hydrogen.
In the heat input range above 0.8 kcal cm-Eh -1, pre-
o. ao mixed combustion was advantageous over diffusive
I' dl/ combustion, as typically shown in Fig. 6. It is worth
i #
~ 7o noting that the nickel metal foam support with larger
// / ~-~D5 pore size offered a higher combustion efficiency than
60 / the one with smaller pore size in diffusive combustion.
/
Fig. 7 shows detailed features of the operating proper-
50 t" , i , , i . . . . i

ties of catalyst D5 in diffusive combustion at


0.5 1.0 1.5
Total equiv, ratio of air/fuel 1.2 kcal cm -2 h -1. At this heat input, overall combustion
efficiency was not as high, however, it increased with
Fig. 6. Effect of equivalent ratio of supplied air on combustion an increase in the amount of secondary air supplied.
efficiency of hydrogen. O, I : Premixed combustion; C), D: The low overall efficiency was mainly due to the spot
diffusive combustion. Heat input, 0.8 kcal-z h -1. combustion efficiency at the upper position of the
catalyst where the entrainment of air was limited to a
stoichiometric amount. Variation of spot catalyst tem-
diffusive combustion yielded a higher combustion effi- perature with the amount of secondary air showed a
ciency than premixed combustion over the nickel foam characteristic feature: that temperature at the bottom
catalyst (02) and the ceramic honeycomb catalyst (Th). (No. 9) dropped with an increase in secondary air while
The above unexpected results can be interpreted by it appreciably rose at the top (No. 1). At the bottom,
taking into consideration the observed difference in the spot temperature tended to fall due to dilution by
catalyst temperature shown in Figs. 2 and 3. Premixed an excessive amount of air diffusing from the ambience.
combustion at a low heat input lowers the catalyst On the other hand, at the top, the amount of air
temperature due to rapid removal of heat released by entrained at the catalyst surface was only slightly
combustion, and therefore results in lower combustion increased with increasing secondary air. However, the
efficiency. On the other hand, the combustion efficiency slight increase in the entrained air markedly improved
in diffusive combustion appeared to be largely affected the spot combustion efficiency and raised the spot sur-
by the extent of mixing hydrogen with air over the face temperature.
catalyst surface. Since the flow rate of hydrogen passing
through the catalyst body was moderate enough at a
Effect of premixing hydrogen with air
low heat input for the ambient air to approach the
catalyst surface, the resulting long residence time Since the flow of air to the catalyst surface from the
ambience appeared to control the performance of a
diffusive catalytic burner, partial premixing with air was
:: 100 expected to be effective, especially at a high heat input,
for the improvement of combustion efficiency. Fig. 8
"6 90 shows the results of such experiments for the catalyst
Th. The overall combustion efficiency increased appre-
ciably with an increase in the amount of premixed air.
;e 70 However, the surface temperature of the catalyst
i i i

~200
~1oo ' " ; . . 7 " -

9s
(,-) 150 i : ,
90~ ' I , I ' I
o

LO
"' (15 ~ 9
1.0 1.5 2.0 2.5
Equiv..ratio of secondary air 0 0.2 0.4 0.6
Equiv. ratio of primary air
Fig. 7. Effect of equivalent ratio of secondary air on operat-
ing properties of catalyst D5 in diffusive combustion. O, Fig. 8. Effect o premixing air on operating properties of
Overall combustion efficiency of hydrogen. Heat input, catalyst Th. Heat input, 1.2 kcal cm -2 h -i. Equivalent ratio of
1.2 kcal cm-2 h -1. totally supplied air, Z = 2.0.
CATALYTIC COMBUSTION OF HYDROGEN--II 733
: ~ I00 appreciable fall of the overall combustion efficiency. In
"a contrast to methane, as hydrogen is a readily oxidizable
..El
E 90 fuel, the convective cooling did not cause a fall in the
8 5 overall combustion efficiency.
80 t I ' I , I
Figs. 9 and 10 show the results of partially premixed
400 combustion over the catalysts D2 and Go, respectively.
The spot combustion efficiency was increased most
~'300 appreciably at the upper position, and therefore the
o overall efficiency was significantly improved. Partial
premixing was most effective on the oxide catalyst Go
"~
tJ
200 ....______~-------E ~ - "
for the improvement of overall efficiency. The efficiency
, I , I , I
was improved from 88 to 98% by premixing air to a
40% stoichiometric amount.
o 3

.~ 2 Effect of heat input


Fig. 11 shows the dependence of the combustion
1 j efficiency of hydrogen on heat input in diffusive oper-
"' 1
ation for the four catalysts. The efficiency decreased
'o12 'oi 'o.'6 gradually with increasing heat input over the Pt-impreg-
Eqv, ratio of premixed air nated ceramic honeycomb (Th), while it fell sharply at
a heat input of 1.2 kcal cm-2h -~ over the Pd-coated Ni
Fig. 9. Effect of premixing air on operating properties of
catalyst 02. O, Overall combustion, efficiency of hydrogen. metal foam which had a smaller porosity (Ds). It is
Heat input, 1.6 kcal cm -2 h -1. Equivalent ratio of totally sup- worth noting that the Ni metal foam support with larger
plied air, 3. = 2.3. pores (D2) offered a higher efficiency than the one with
smaller pores (Ds) at high heat ratings. The structure
dropped when the equivalent ratio of premixed air was and the size of pores may have a predominant role in
increased from 0.4 to 0.6. This implies that with an determining the combustion efficiency at high ratings.
increase in premixed air, the gas mixture passes through A different feature was observed for the ceramic
the catalyst body at a higher rate and removes the heat foam coated with oxide catalyst (Go): the efficiency was
released at the catalyst body more rapidly. a parabolic function of heat input and therefore the
In the catalytic combustion of methane [2], the pre- usable limit of heat input has to be imposed on both
mixing effect was largely offset by convective cooling the upper and lower sides. The oxide catalyst developed
due to an increase in rate of flow of gas mixture, and in our laboratory was much more active than any other
addition of the air to 50% stoichiometry caused an conventional oxide catalyst, however, it was still not
sufficiently active at lower heat ratings where temper-
ature remained at around 50C on the most of the
le,dO0,
catalyst body.
"6 According to the present European standards for
-~ 90, LPG heaters, which requires fuel slippage of less than
8 4%, the adjustable range of heat input for hydrogen is
~ 80, from 0 to 1.4 kcal cm-Eh -1 with the nickel foam coated
300 i l i I , I ,
with Pd fine powder (DE).
U
Max ....
o

d250'

E 200" r

~6
o
E
9o \
O

IM

0 ' 0'.2 ' 0:4 ' 016 ' 7Co o:s i.o i.s 2.0
Eqv. ratio of premixed air Heat input, kcal/cm2-h
Fig. 10. Effect of premixing air on operating properties of Fig. 11. Dependence of combustion efficiency of hydrogen on
catalyst Go. 0 , Overall combustion efficiency of hydrogen. heat input in diffusive combustion. Equivalent ratio of
Heat input, 1.6 kcal cm-2 hr -1. Equivalent ratio of totally sup- secondary air, 3.2= 1.8 -+ 0.2.
plied air, 3. = 1.73.
734 M. HARUTA, Y. SOUMA AND H. SANO
Difference in temperature distribution ooer catalyst Measurements of spot ratio and spot combustion
surface efficiency
Temperature at nine spots on the catalyst surface was Spot measurements of the concentrations of hydro-
continuously measured. In Fig. 12 the steady-state tem- gen, oxygen and nitrogen were made on nine spots over
perature at the three representative spots is shown as the catalyst surface, and air ratio and combustion effi-
a function of heat input. Significant differences were ciency at each spot were determined. Fig. 13 shows the
observed in the temperature distribution between cer- spot equivalent ratios of diffusive air as a function of
amic honeycomb, metal foam and ceramic foam sup- heat input. The equivalent air ratio on the top of each
ports. The ceramic honeycomb (Th) had the most uni- catalyst had a declining tendency with an increase in
form temperature distribution, while Ni metal foam heat input. Especially, over the catalyst D5 (with small
with larger porosity (D2) had considerably non-uniform pores) the local equivalent ratio of diffusive air at the
distribution. Although catalyst Dz offered the highest top was below a stoichiometric amount even at mod-
combustion efficiency, this non-uniform temperature erate heat ratings. Considering the local spot air ratio
distribution would impose a limit on operating at higher in connection with combustion efficiency shown in Fig.
heat ratings because of flame ignition at a local hot spot. 11, it can be suggested that catalysts with higher spot
The temperature difference between the upper and air ratios, such as D2 and Go could offer higher com-
lower spot became more remarkable with increasing bustion efficiency in diffusive combustion.
heat input over the catalyst D2, while it became less Fig. 14 schematically represents the distribution of
appreciable over the other three catalysts. Over the the catalyst surface temperature, the equivalent ratio
catalysts Go and Ds, the temperature at the upper spot of entrained air, and the efficiency of hydrogen com-
was exceeded by the temperature at the lower spot at bustion over the four catalysts. Surface temperature
a high heat input. This indicates the change in the local was of uniform distribution and mostly between 150C
heat rating with an increase in hydrogen feed rate. The and 200C on catalysts Th and Go at a heat input of
upward flow of hydrogen and air across catalysts with 1.2kcalcm-2h -1. As described above, the catalyst
relatively small pores was restricted, as observed in Fig. structure has a great influence on the temperature dis-
14. The lack of entrained air may reduce the combustion tribution, however, it can also be pointed out that
efficiency. Therefore, an increase in overall heat input ceramic supports are advantageous over metal supports
did not necessarily bring about higher spot heat rating in providing a uniform temperature distribution. The
at the top of the catalysts. On the other hand, with surface temperature was markedly non-uniform on
appropriately large pores of three-dimensional distri- catalyst D2. Temperature at the top was over 300C,
bution (e.g. catalyst D2) an increase in overall heat while it was as low as ca l l0C at the bottom. The
input led to higher heat rating at the top of the catalyst temperature distribution over catalyst D5 presented a
body due to the ease of upward flow of air through clear contrast to that of the other three catalysts.
pores. Because of the lack of air flowing into the catalyst body

400 I i I I I

Th D5 D2 Go
u

t 300
ul
I1 l

200

lO0

p-
I I L i I I I I I i I I
l 0 l 0 l ]
Heat input, kcal/cm2.cat.,h
Fig. 12. Dependence of spot temperature of catalyst surface on heat input in diffusive combustion.
Equivalent ratio of supplied air, 22 = 1 . 8 -+ 0.2.
CATALYTIC COMBUSTION OF HYDROGEN--II 735

3 Th D5 D2/~ Go

9
~ 2 ~ '

~A

0 , I A / ~ I , , I ,

0 l 0 l 0 l 0 l
Heat input, kcal/cm2ocat ,h
Fig. 13. Dependence of spot air ratio over the catalyst surface on heat input in diffusive combustion.
II, Equivalent ratio of supplied air, 2~2= 1.8 -+ 0.2.

and the difficulty of upward flow of air through the Th and D2, however, the temperature distribution was
catalyst, the bottom of the catalyst was warmer than in marked contrast to each other It can be assumed
the top. from the uniform distribution of surface temperature
Spot equivalent ratio of entrained air and spot com- that hydrogen was distributed uniformly over catalyst
bustion efficiency decreased with vertical height. Defi- Th. The markedly high temperature at the top of catalyst
ciency in the entrained air considerably reduced the D2 indicated the localization of hydrogen brought about
combustion efficiency at the top of the catalysts except by the upward flow through large pores, of three-
for D2. A similar feature, that spot combustion effi- dimensional distribution.
ciency varied only slightly, was observed on catalysts It can be concluded that sufficient entrainment of air

Fmll3ooo c < , ~ 250-300 C,

Illll I III I11

82.6 88.4 77 5 85.3 80.4 97.7 96.1 97.8 85. 86. l 90 4


(I.06) (l.13) (0.94) (1.03) (0.92) (1.04) (1.04) (1.12) (0.96) (0.95) (0.99)

87.6 91.5 99.0 94.3 97.9 88.9 99.5 98.7 99.7 98.4 98.6 98,9
(1.47) (1.48) (1.42) (I.22) (1.27) (l.21) (1.78) (I 73) (2.04) (I.94) (1.87) (I.92)

91.2 88.5 87,1 98.2 88.6 99.3 98.6 98.5 98.2 99.2 98.6
(1.26) (2.02) (2.05) (l 3 6 ) ( l .32) (I51) (2.23) (2.74) (1.29) (1.89) (1.87)

(a) (b) (c) {d)


Fig. 14. Distribution of spot temperature, combustion efficiency and equivalent air ratio over catalyst
surface in diffusive combustion. Heat input, 1.2 kcal cm -2 h -1. Equivalent ratio of secondary air, A2=
1.8 -+ 0.2. (a) Catalyst Th, ).2 = 1.74, E = 83.7 - 87.3%. (b) Catalyst Ds, ).2 = 1.62, E = 81.2%. (c)
Catalyst D2, ~,2 = 1.96, E = 98.0%. (d) Catalyst Go, ;t: --- 1.94, E = 93.0%.
736 M. HARUTA, Y. SOUMA AND H. SANO

100 : : It is presumable that the adjustable heat input could


oo
..e* 9 4 r
f , , ,
be extended to much larger values than 1.6 kcal cm -2
h -1, because shorter residence time, resulting from an
increase in heat input, could be compensated by a
300 temperature rise of the catalyst body. Moreover, as
clearly observed in Fig. 16, temperature distribution
was more uniform in premixed combustion than in
diffusive combustion. Therefore, completely premixed
~ loo combustion can be sustained over a wide range of heat
rating without ignition at a local hot spot.
o o'.s 1'.o l'.s zo
Heat input, kcal/cm2-h
CONCLUSION
Fig. 15. Effect of heat input on operating properties of catalyst The present experimental study has focussed on the
Ds in perfectly premixed combustion. @, Overall combustion effect of fundamental conditions of hydrogen fuelled
efficiency. Equivalent ratio of primary air, ).1 = 1.5 -+ 0.15.
catalytic burners on their operating properties. In dif-
fusive combustion, Pd-power coated Ni foam with rela-
tO the catalyst surface is a prerequisite for high efficiency tively large pores could offer the highest combustion
in diffusive combustion, however, uniform temperature efficiency, however, it suffered from considerable
distribution and accordingly uniform dispersion of non-uniform distribution of surface temperature. It was
hydrogen are also indispensable for safe combustion at demonstrated that a ceramic foam coated with C o -
high heat ratings. Further work should be done to M n - A g oxide, developed in our laboratory was prac-
develop a catalyst which satisfies these two tically utilisable for catalytic appliances operating on
requirements. hydrogen fuel, although it required a little preheating
for initiating combustion. Premixing of air to a 40%
stoichiometric amount with hydrogen was effective in
Completely premixed combustion
improving combustion efficiency. In completely pre-
Although the diffusive mode of operation is advan- mixed operation, combustion efficiency higher than
tageous over the completely premixed mode for the use 99% was obtained in the heat input range up to
of space heaters as house-hold items (i.e. in safety from 1.6 kcal cm -2 hr -1.
the danger of backfiash and simplicity of burner con-
struction), completely premixed combustion seems to Acknowledgements--We wish to thank Mr. S. Masumoto of
find industrial application in drying and heating pro- Daido Steel Co. Ltd., Japan for providing the nickel foam
cesses and road-heating systems. Since there was no catalysts coated with Pd powder. Thanks are also due to
need for the flow of air against the bulk flow of fuel in Bridgestone Tire Co. Ltd., Japan for supplying the ceramic
foam support.
completely premixed combustion, the catalyst body
with smaller pores (Ds) was chosen because of the
availability of a longer residence time of combustible REFERENCES
gas mixture. As shown in Fig. 15, the overall combustion 1. M. Haruta and H. Sano, Catalytic combustion of
efficiency had a slightly decreasing tendency, however; hydrogen--I. Its role in hydrogen utilization system and
it was over 99% in the range of experiments up to screening of catalyst materials. Int. J. Hydrogen Energy a,
1.6 kcal cm-2 h-l, which corresponds to a space velocity 601 (1981).
of combustible gas mixture of 5.3 x 103 h -1. 2. S. W. Radcliff and R. G. Hickman, Diffusive catalytic
combustor. J. Inst. Fuel 208 (1975).
< 150 mm 3. H. Sadamori, Current catalytic combustion techniques for
heating, Nenryo Kyokaishi (J. Fuel Assoc. Japan) 58, 422
(1979).
4. S. Pavan and U. Capetti, Panels for catalytic combustion
of natural gas: utilizations and prospects. 14th World Gas
Conf., Toronto, 1979.
5. D. G. Johnson, W. J. Jasionoswki, J. B. Pangborn and
M. I. Scott, Development of ventless appliances and a
catalytic ignition system. Inst. Gas Tech. Symposium, New
Orleans, Louisiana, 26-30 March 1979.
6. J. B. Pangborn, Catalytic combustion of hydrogen in model
appliances. 15th Intersoc. Energy Conv. Engng. Conf.,
Seattle, Washington, 18-22 August 1980.
Diffusive combustion Premixed
combustion 7. A. Melvin, The diffusive combustion of hydrogen. Proc.
3rd World Hydrogen Energy Conf. Tokyo, 23--26June 1980,
Fig. 16. Temperature distribution over catalyst surface in two 1177 (1980).
different types of combustion. Catalyst D2; heat input, 8. Notice d'utilisation de radiateurs BUTA THERM'X
0.8 kcal cm -2 h-l; equivalent ratio of supplied air, ;t = 1.44. (France).