Relative Role of H2 and Hydrocarbon in reduction process of HC SCR system
driven by on-board plasma reformer
D. H. Lee, J.-O. Lee, K.-T. Kim, Y.-H. Song
Eco-Machinery Research Division, Korea Institute of Machinery & Materials, Daejeon, Korea
Abstract: On board reformer for hydrocarbon selective catalytic reduction (HC SCR) system
for automobile de-NOx process is suggested. Basically HC SCR catalyst is designed to utilize
HC species in engine emission, however de-NOx performance by HC species are rather lower
than that of NH3 or H2. In this study, on board plasma reformer is suggested as a possible
provider of H2 as a reductant. Suggested plasma reformer is developed for the possible installation in 3,000 cc diesel engine. It has small volume of about 100 cc and consumes less
than 100 W of electric power. The reformer can control relative composition of H2 and HC
species by changing operating parameter. In engine bench test, relative role of H2 and HC in
reduction process is investigated. Plasma reformer showed good de-NOx performance across
wide range of operating temperature. Though the operating condition of a SCR system depends much on the characteristic of catalyst used, plasma reformer can follow the need for
reductant composition dynamically. It is expected that the application of the plasma reformer
is promising.
Keywords: HC SCR, H2, on-board reformer
1. Introduction
NOx is representative hazardous emission of diesel
engine. Regulations are getting strict for NOx emission.
Euro VI demands 0.08 g/km of NOx emission for passenger cars.[1] And both advanced engine technology
and after treatment technology are required for further
reduction of NOx.
In this presentation, plasma technology for deNOx is
described. Plasma can be used as a provider of reducing
agents as an on-board reformer. Among various deNOx
technologies, hydrocarbon selective catalytic reduction
(HC SCR) system is selected in this work. Though HC
SCR has rather lower deNOx performance compared to
NH3 SCR or lean NOx trap (LNT), it can be applied with
simple configuration and low cost of installation and operation. And advanced engine technology that can level
the NOx emission much less than 100 ppm reduces the
burden of deNOx performance of after treatment system,
HC SCR can be an attractive way of deNOx strategy.[2~5]
Plasma can be easily combined as a HC SCR system,
it can be independently operated with engine condition,
and plasma reformer developed has small volume of less
than 100 cc and consumes only about 100W based on the
condition of 300cc engine. Moreover, plasma on-board
reformer can produce hydrogen that is kind of deNOx
booster in Ag based SCR catalyst.[6,7]
This work focuses the function of reformer especially
as a generator of on-board hydrogen for a enhancement
of SCR performance.
as a diesel reformer.[8] without additional vaporizer and
mixer, rotating arc reformer can produce reformate gas
that has hydrogen concentration of up to 10%. Because
HC SCR catalyst uses hydrocarbon species as reducing
agents, reformer should produce hydrocarbon species that
is good at reduction process such as C2H4, C3H6 as well as
hydrogen. Figure 1 shows relative production of selective
hydrocarbon species according to O2/C ratio. Commercial
diesel used for the experiment was simulated by surrogate
fuel of hexadecane for analysis. Though the stoichiometric O2/C ratio of partial oxidation of hexadecane is 8, inaccuracy of surrogate fuel and requirement of heat for
partial oxidation of liquid phase fuel results in incomplete
reforming process above O2/C ratio of 8 and still decomposition of heavy hydrocarbon species occurs [8]. In the
course of these reforming characteristics, hydrogen
production does not varied much according to O2/C ratio
as shown in Fig. 2
2. Plasma reforming
Rotating arc already showed its superior performance
Figure 1. Hydrocarbon production in the reformer according to the O2/C ratio.
Figure 2. The ratio of H2/HC according to the O2/C condition
3. Catalyst
Ag/Al2O3 catalyst are tested for HC SCR. The catalyst
has been widely studied as SCR catalyst and known to
show good performance especially in the case of H2 addition.[6]
HeeSung Catalysts corporation provided catalyst for
the test and catalyst used in this work has 2.5% Ag/Al2O3
with size of 5.66” × 6” (2.5 L), cell density is 400/6. Figure 3 shows basic test result of the same catalyst with test
specimen with dimension of 1” × 2” (Heesung catalyst
corp.) with S/V of 50,000/hr.
Figure 3. Basic test result of the catalyst
Figure 4. Schematic of engine bench test apparatus
4. Engine bench test
Engine bench test for the proposed SCR system has
been done. Diesel engine of 3L (EURO III) capacity is
used. Overall configuration of the test apparatus are
given in Fig. 4 NOx sensor before and after SCR catalyst records NOx conversion and AC power supply is
used for the operation of the reformer. Fuel flow of
2.7~6 ml/min are supplied to the reformer by precision
pump. Air supply is controlled by MFC to match the designed O2/C condition. Emission conditions of the engine according to the temperature in rpm of 2,000 are
listed in table 1.
2 L of catalyst are tested. Considering that typical HC
SCR system requires the catalyst volume of about 2
times of that of engine volume, much higher S/V of
80,000/hr are designed. Reduction of catalyst volume is
important in view of real application. Because the room
of facility installation in underbody are very limited.
Table 1. Engine emission condition according to the
exhaust gas temperature
Temperature (oC)
170
220
270
320
370
420
NOx (ppm)
133
200
231
224
240
252
THC (ppm)
27.0
23.0
21.0
19.3
22.3
26.3
THC/NOx
0.203
0.126
0.098
0.079
0.093
0.104
Figure 5 shows compared deNOx performance according to the exhaust gas temperature for different
O2/C ratios. Interesting enough is that in the case of reformer off condition, almost no deNOx is observed even
in temperature above 400oC that is why the S/V of the
testes catalysts are too high. However, once the reformer
is operated, the performance shows totally different results. DeNOx performance in each condition shows
much higher rate up to 60%. In the case of O2/C 9.6
shows best deNOx performance in all of the temperature
conditions.
Figure 5. De-NOx rate of tested conditions where the
effect of temperature and O2/C ratio is clearly distinguishable.
5
4
50
3
40
30
2
20
1
10
0
0
0.2
0.4
0.6
50
0
0.8
6
De-NOx rate
HC / NOx
H2 / NOx
(HC + H2) / NOx
5
40
4
30
3
20
2
10
1
0
0
0.2
O2 / C ratio
0.4
0.6
HC / NOx, H2 / NOx, (HC+H2) / NOx
60
60
De-NOx rate (%)
70
De-NOx rate (%)
6
De-NOx rate
HC / NOx
H2 / NOx
(HC + H2) / NOx
HC / NOx, H2 / NOx, (HC+H2) / NOx
80
0
0.8
O2 / C ratio
(a) Temperature at 420 oC
(b) Temperature at 320 oC
Figure 6. Comparison of de-NOx behavior and variation of produced amount of reducing agents. (a) for temperature
of 420oc (b) for temperature of 320oC
60
50
DeNOx rate (%)
The results show that one of the important merits of
plasma assisted deNOx process is possible reduction of
catalyst volume. The result in this work implies that less
than a half of the catalyst may be sufficient for the desired
deNOx performance. This is important for the cost of the
system can be reduced drastically for the cost of the catalyst has biggest portion in the cost of overall system.
Figure 6 shows the result of correlation between deNOx rate and the supplied reducing agents conditions. In
the case of 420 oC, deNOx rate follows similarly with the
ratio of (HC+H2)/NOx. However, as the temperature decreases, or in the case of 320 oC, deNOx rate deviates
from the ratio of (HC+H2)/NOx as the O2/C ratio increases. This is why in the relatively low temperature
condition, oxidation of HC species is less active than in
the case of high temperature. For the H2 has still strongly
oxidative in the relatively low temperature condition,
deNOx rate in higher O2/C ratio condition shows less deactivation than in the cases of lower O2/C ratio. For the
catalyst itself has less oxidative function, the effect H2
amount can be distinguishable as the temperature decreases. If the catalyst has strong oxidative function, H2
will be oxidized before it can work as reducing agents. So
as the oxidative function becomes less strong, the dependence of deNOx performance becomes less significant.
This can be confirmed from Fig. 7, That shows the result
of deNOx performance against O2/C ratio. DeNOx rate
does not show much deviation along all the tested O2/C
conditions.
The fact that deNOx rate is less dependent on the O2/C
ratio is good for robust operation of the system. However,
it can be a challenging aspect in view of long term operation. For the oxidative function of the catalyst can be
beneficial in the removal of HC deposition. Reformer
possibly produces white smoke or heavy hydrocarbon
species as well as H2, CO and light hydrocarbon species
especially in the start up condition or cold start condition.
40
30
20
10
0
0.4
0.5
0.6
O2/C ratio
0.7
0.8
Figure 7. De-NOx rate according to the O2/C ratio, temperature of 370oC condition
Once the heavy hydrocarbon species deposits on the
catalyst surface it can reduce the active site of the catalyst
and deactivates deNOx performance. if the catalyst has
oxidative function, these depositions can be oxidized even
to a useful reducing agents even in the low temperature
condition. however, if the oxidative function of the catalyst is not so strong, the deposition results in progressive
deactivation of the catalyst. To figure out the deactivation
characteristics, aging test was done for different O2/C
ratio conditions.
Figure 8 shows the aging test results. Increase of O2/C
ratio produces oxidative environment in the reforming
process and results in more oxidative cracking of heavy
hydrocarbon species. Lee et al reported that in the plasma
reforming of diesel fuel, increase of O2/C ratio results in
the decrease of heavy hydrocarbon species rather than
oxidation of hydrogen.[8] and the effect of increased O2/C
ratio is reflected in the result of aging test. Even from the
result of the O2/C ratio of 9.6 shows slight decrease in the
deNOx rate. From the result, O2/C ratio above 12 is required to guarantee the long term operation of the reformer. However, the problem of aging can be solved by
changing the operation mode of the reformer. If the O2/C
ratio increases much above the partial oxidation condition,
the reformer turn into a plasma burner without change of
geometry or any configuration of the system.
One the reformer is operated as a burner, elevated temperature of the exhaust gas can be enough for the oxidation of the deposited hydrocarbon species on the catalyst
surface. The procedure is the same with that of DPF regeneration. Plasma burner for the DPF regeneration has
been commercialized after field test of more than
100,000km [9] and the plasma burner concept can be
successfully applied to the reformer.
80
70
O2 / C = 6.4
O2 / C = 9.6
O2 / C = 12.8
De-NOx rate (%)
60
50
40
has been done. Ag/Al2O3. SCR catalyst is tested. Plasma
on board reformer is used for the generation of reducing
agents for the SCR system. Rotating arc reactor is used
for the reforming or diesel fuel. Plasma assisted HC SCR
system shows much enhanced deNOx rate even in much
higher S/V condition. Merits of the system can be listed
as follows. 1) plasma reformer is best suitable for the on
board application. Plasma reactor itself has volume of less
than 100cc. and the power supply developed for the
plasma generation already went through all the test procedures of electronic device for automobile application. 2)
the reformer can be operated as a plasma burner for possible regeneration of the catalyst without any modification
or change of the sysem. Only the change of the reactant
condition (O2/C ratio) can switch the mode of operation.
And the plasma burner has been verified for the on board
device and commercialized. 3) plasma assisted SCR system can reduce the catalyst volume drastically. S/V of the
system can be modulated to be much high by adopting the
plasma reformer. Reduction of the catalyst volume means
reduction of the cost of the system. Cost for the plasma
reformer and power supply can be less than the amount of
the cost of reduction. 4) though the proposed system has
many merits listed above, the fuel penalty of the system is
rather low. Above 50% of deNOx rate can be obtained
with fuel penalty of less than 3%.
All of the results strongly supports possibility of commercialization of the proposed system.
30
Acknowledgement
This work was supported by Basic Research Program
of the Korea Institute of Machinery & Materials and Near
Zero Emission Vehicle Center research program (KEIT)
20
10
0
0
500 1000 1500 2000 2500 3000 3500
Time (second)
Figure 8. Comparison of aging test in each O2/C condition.
370 oC
The overall fuel penalty of the system is estimated to be
3%. The fuel penalty is defined to be the ratio of fuel
consumption of reformer to that of engine. The fuel penalty can be still be reduced by optimization of the reformer that controls production of reformate to be only
necessary amount of reducing agents. Considering that
other deNOx technology such as Lean NOx Trap (LNT)
generally reports fuel penalty of above 5%, the proposed
system in this work can be estimated to have good fuel
penalty.
All of the results presented in this presentation verifies
that the plasma assisted HC SCR system can be a strong
candidate of future deNOx technology.
4. Conclusions
Plasma assisted HC SCR system is proposed and the
engine bench test to verify the feasibility of the system
References
[1] Skalska, K, Miller, JS, Ledakowicz, S. Sci Total
Environ;408, 19 (2010) 3976–3989.
[2] Krocher, O. Studies in Surface Science and Catalysis;171 (2007) 261–289.
[3] Alkemade, UG, Schumann, B. Solid State Ionics,177
(2006) 2291–2296.
[4] Castoldi, L, Lietti, L, Nova, I, Matarrese, R, Forztti, P,
Vindigni, F, Morandi, S, Prinetto, F, Ghiotti, G.
Chemical Engineering Journal, 161(3) (2010)
416–423.
[5] Nova, I, Lietti, L, Forzatti, P. Catal Today 136(1~2)
(2008) 146–155.
[6] Breen JP, Bruch R. Top Catal, 39(1–2) (2006) 53–8.
[7] Kannisto H, Ingelsten HH, Skoglundh M. Top. Catal,
52 (2009) 1817–20.
[8] Lee DH, Kim K-T, Cha MS, Song Y-H. Int J Hydrogen Energy, 35 (2010) 4668–75.
[9] Lee DH, Kim K-T, Cha MS, Lee JO, Song Y-H, Cho
H, et al.(2009) SAE Paper 09SFL-0065