Journal of the Air & Waste Management Association

ISSN: 1096-2247 (Print) 2162-2906 (Online) Journal homepage: http://www.tandfonline.com/loi/uawm20

Control of Diesel Gaseous and Particulate

Emissions with a Tube-Type Wet Electrostatic
Precipitator

Phirun Saiyasitpanich , Tim C. Keener , Mingming Lu , Fuyan Liang & Soon-Jai

Khang

To cite this article: Phirun Saiyasitpanich , Tim C. Keener , Mingming Lu , Fuyan Liang & Soon-
Jai Khang (2008) Control of Diesel Gaseous and Particulate Emissions with a Tube-Type Wet
Electrostatic Precipitator, Journal of the Air & Waste Management Association, 58:10, 1311-1317,
DOI: 10.3155/1047-3289.58.10.1311

To link to this article: https://doi.org/10.3155/1047-3289.58.10.1311

Published online: 24 Jan 2012.

Submit your article to this journal

Article views: 283

Citing articles: 7 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=uawm20
TECHNICAL PAPER
Control of Diesel Gaseous and Particulate Emissions with a
Tube-Type Wet Electrostatic Precipitator
Phirun Saiyasitpanich, Tim C. Keener, Mingming Lu, and Fuyan Liang
Department of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, OH
Soon-Jai Khang
Department of Chemical and Material Engineering, University of Cincinnati, Cincinnati, OH
ABSTRACT
In this study, experiments were performed with a bench-
scale tube-type wet electrostatic precipitator (wESPs) to
investigate its effectiveness for the removal of mass- and
number-based diesel particulate matter (DPM), hydrocar-
bons (HCs), carbon monoxide (CO), and oxides of nitro-
gen (NOx) from diesel exhaust emissions. The concentra-
tion of ozone (O3) present in the exhaust that underwent
a nonthermal plasma treatment process inside the wESP
was also measured. A nonroad diesel generator operating
at varying load conditions was used as a stationary diesel
emission source. The DPM mass analysis was conducted
by means of isokinetic sampling and the DPM mass con-
centration was determined by a gravimetric method. An
electrical low-pressure impactor (ELPI) was used to quan-
tify the DPM number concentration. The HC compounds,
n-alkanes, and polycyclic aromatic hydrocarbons (PAHs)
were collected on a moisture-free quartz filter together
with a PUF/XAD/PUF cartridge and extracted in dichlo-
romethane with sonication. Gas chromatography (GC)/
mass spectroscopy (MS) was used to determine HC con-
centrations in the extracted solution. A calibrated gas
combustion analyzer (Testo 350) and an O3 analyzer were
IMPLICATIONS
Long-term exposure to diesel emissions are linked to in-
creasing adverse human health impacts because of the
potential association of the diesel emissions to carcinoge-
nicity. This study showed that at the optimum operating
condition, a wESP removed more than 90% of number-
and mass-based DPM from undiluted diesel exhaust. The
wESP also provided a significant removal of gas- and solid-
phase HCs (such as alkanes and PAHs) and CO. O3 con-
centration generated by corona discharge was negligible in
the wESP exhaust. The use of wESP has advantages over
filter-based technologies in terms of moderate energy con-
sumption, low maintenance, simplicity in operation, no
backpressure, and no interference with engine operation.
These advantages would be especially advantageous for
stationary diesel engines, marine diesel engines, and diesel
engines used in unique applications such as underground
mining operations. Limitations would include the require-
ment for exhaust cooling before treatment, the use and
disposal of water in the collection scheme, and the require-
ment for sufficient exhaust vertical piping for the wESP.
Volume 58 October 2008
ISSN:1047-3289 J. Air & Waste Manage. Assoc. 58:1311–1317
DOI:10.3155/1047-3289.58.10.1311
Copyright 2008 Air & Waste Management Association
used for quantifying the inlet and outlet concentrations
of CO and NOx (nitric oxide [NO] ⫹ nitrogen dioxide
[NO2]), and O3 in the diesel exhaust stream. The wESP was
capable of removing approximately 67– 86% of mass- and
number-based DPM at a 100% exhaust volumetric flow
rate generated from 0- to 75-kW engine loads. At 75-kW
engine load, increasing gas residence time from approxi-
mately 0.1 to 0.4 sec led to a significant increase of DPM
removal efficiency from approximately 67 to more than
90%. The removal of n-alkanes, 16 PAHs, and CO in the
wESP ranged from 31 to 57% and 5 to 38%, respectively.
The use of the wESP did not significantly affect NOx
concentration in diesel exhaust. The O3 concentration in
diesel exhaust was measured to be less than 1 ppm. The
main mechanisms responsible for the removal of these
pollutants from diesel exhaust are discussed.
INTRODUCTION
Diesel exhaust emissions have been regarded as a threat to
public health because of their harmful compositions, pri-
marily comprised of diesel particulate matter (DPM), ox-
ides of nitrogen (NOx), carbon monoxide (CO), and hy-
drocarbons (HCs).1 DPM and its associated HCs,
especially polycyclic aromatic hydrocarbon (PAHs) com-
pounds adsorbed on the particle surface, are of particular
interest because of their potential link to carcinogenicity.
Despite the increasingly stringent legislative standards
being based solely on DPM mass,1,2 several studies suggest
that at a given mass concentration, the nanometer-sized
particles are more deleterious than micrometer-sized par-
ticles.3,4 A possible explanation is that smaller particles
have more possibility to penetrate deeper into the respi-
ratory system when inhaled and they have more active
surface areas permissible for adsorption of toxins than
those of larger particles for a similar mass dose.4 – 6
Several techniques have been proposed for reduction
of diesel exhaust emissions such as engine modifications,
fuel additives, alternative fuels, and aftertreatment sys-
tems.7–10 Nevertheless, to comply with the more stringent
standards being imposed, the major attention has shifted
toward the development and utilization of the aftertreat-
ment systems such as diesel oxidation catalysts (DOCs)
and diesel particulate filters (DPFs). DOCs oxidize CO
(⬎90%), gas-phase HCs (60 –70%), and the less volatile
HCs adsorbed on the DPM (53–71%) to carbon dioxide
(CO2) and water in an exhaust gas stream.11–13 However,
Journal of the Air & Waste Management Association 1311
Saiyasitpanich, Keener, Lu, Liang, and Khang
at high exhaust temperatures, the unfavorable oxidation
of nitric oxide (NO) to nitrogen dioxide (NO2) and sulfur
dioxide (SO2) to sulfite (SO3) (depending on the activity
and selectivity of the catalysts used and fuel sulfur con-
tent) could take place, thus intensifying the toxicity of the
emitted exhaust and substantially increasing DPM
mass.11,12 DPFs offer a wide range of mass-based DPM
removal efficiency from 28.5 to 99% depending on filter
types and ages, engine sizes, sampling methods, and fed
fuels.12,14,15 However, this device is restricted by its com-
plexity in operation and regeneration, which in turn leads
to a high-energy consumption, fuel penalty, the possibil-
ity of exhaust back pressure, and high maintenance
costs.16 –18
Nonthermal plasma such as wet electrostatic precip-
itators (wESPs) and dielectric barrier discharge (DBD) re-
actors have also been developed for cleaning of submi-
crometer particles and other gaseous contaminants, i.e.,
gas-phase HCs, CO, and NOx in diesel exhaust.19 –22 The
ion attachment outside a plasma region under the electric
field is the main mechanism responsible for removal of
submicrometer particles from the exhaust gas. Inside the
plasma region, gas-phase HCs can be removed from the
exhaust gas primarily via chemical reaction with radicals
(OH䡠,O, etc.), ion-cluster formation, and polycondensa-
tion induced by dissociative electron attachment.22,23 CO
can be oxidized to CO2 by chemical reaction with radicals
(OH䡠,O, etc.),24 and NO can be reduced or oxidized by
radicals (OH䡠,O,O3, N䡠, H䡠, etc.) to form molecular nitro-
gen (N2), NO2, nitrous acid (HNO2) or nitric acid
(HNO3).20,21 Previous studies have also showed that a
corona discharge in air may lead to a production of NOx
at 3.6 ⫻ 1014 molecules/J of energy input.25
Ozone (O3) is a byproduct of the nonthermal plasma
reactor and its generation rate is governed by the amounts
of atomic oxygen (O) principally generated by the reac-
tion of excited nitrogen and oxygen molecules (N2* and
O2*) with the molecular oxygen (O2).26,27 Previous studies
have reported that the rate of O3 production in the neg-
ative corona discharge in dry air is 6.1 ⫻ 10⫺3 mg/sec per
Figure 1. Schematic drawing of the experimental setup.
1312 Journal of the Air & Waste Management Association
length of a charging electrode.26 However, in humid air,
water vapor significantly reduces O, which leads to re-
duced O3 production by a factor of 6.28
In this project, a bench-scale wESP has been designed
and evaluated for the removal of DPM mass and number,
n-alkanes (C10-C25), and 16 PAHs (U.S. Environmental
Protection Agency [EPA] recommended), CO, and NOx
from diesel exhaust. The n-alkanes and PAHs were sub-
jected to evaluation in this study because they are one of
the major HC compounds present in diesel exhaust and
because of the important health impact associated with
PAHs.29 The O3 concentration in diesel exhaust at the exit
of the wESP was also measured. The use of the wESP has
advantages over the filter-type units, especially for the
stationary diesel engine, in terms of moderate energy
consumption, low maintenance requirements, simplicity
in operation, no backpressure, and no interference in
diesel engine operation.
EXPERIMENTAL SETUP
Emission Source
A 5-L (4-cylinder, 80-kW maximum output at 1800 rpm)
nonroad Generac diesel generator (1992, Model SD080)
featuring direct-injection, turbocharging, and compres-
sion ignition was used as a stationary source of diesel
emissions. In contrast to diesel vehicles, diesel generators
run at a fixed rpm and voltage and its amperage output is
variable with loads. In this study, the generator utilizing
low-sulfur diesel fuel (0.05% sulfur) was operated at con-
stant loads of 0, 25, 50, and 75 kW regulated by a Merlin
100 load simulator manufactured by SIMPLX.
wESP Characteristics
Figure 1 shows the schematic diagram of the experimental
setup. The exhaust gas was cooled in an indirect water-
cooled heat exchanger before it was sent to the wESP for
treatment. The heat exchanger was able to cool the ex-
haust gas temperature from 138 –288 °C to approximately
32– 49 °C to prevent the evaporation of water and damage
to Teflon and plastic connectors inside the wESP, but
Volume 58 October 2008

additionally to take advantage of the effects of flux force
condensation on particle growth and agglomeration.30 A
tubular wESP made of carbon steel measuring 0.914 m in
length and 0.178 m in internal diameter and was
equipped with a high-voltage, stainless-steel electrode
(0.559 m long and 0.25 mm in diameter), made taut by a
weight. The charging electrode was energized with a dc
voltage of negative polarity and controlled within a range
of 0 –70 kV by a Glassman high-voltage transformer
(WK125P5). The collecting electrode was continuously
irrigated by means of a water film to avoid a breakdown of
the electric field. The water washing flow rate was main-
tained between 2.27 and 3.18 L/min. The contaminated
water was filtered before it was recirculated into the wESP.
By controlling the diesel exhaust flow rate entering the
wESP, the gas residence time was varied from approxi-
mately 0.1 to 0.4 sec, corresponding to specific collection
area (SCA) of approximately 0.038 to 0.15 m2/m3/min,
which is significantly lower than that used in industrial-
ized ESPs (0.25–2.1 m2/m3/min).31
DPM Mass and Number Concentration
Measurements
DPM mass analysis was done by isokinetic sampling using
the EPA standard Method 5. DPM mass concentration was
determined by the gravimetric method. An electrical low-
pressure impactor (ELPI; Dekati) was used to quantify
DPM number size distributions every second in the raw
and treated diesel exhaust. Oiled sintered metallic sub-
strates were utilized at each stage to reduce particle
bounce and overloading.32 Before entering the ELPI, the
treated or untreated exhaust gas was diluted inside a mini-
dilution chamber using dry, filtered ambient air so as to
obtain appropriate particulate number concentrations for
reliable ELPI measurement. The minidilution chamber is a
0.3 m long stainless-steel tube. The gas residence time
inside the dilution chamber was approximately 2.5 sec.
The dilution ratio could be varied between 0 and 100 by
controlling the exhaust flow rate according to particulate
concentrations under different engine loads. A portable,
digital gas combustion analyzer (TESTO 350) was used to
measure NOx as an indicator of the dilution ratio.
Determinations of n-Alkanes and PAHs
A gas chromatography (GC)/mass spectrometry (MS), Var-
ian, CP-3800 GC, Saturn 2200 MS) equipped with a CP-
8400 automatic sampler was used to identify and quantify
particle- and gas-phase HCs (n-alkanes: C10-C25 and 16
PAHs recommended by EPA). The column used for com-
positional analysis was a CP-Sil 8 CB low bleed column
(30 m ⫻ 0.25 mm ⫻ 0.25 ␮m,), which is the equivalent of
DB5-MS often used for these types of compounds. Selec-
tive ion search was also used to enhance the determina-
tion of the PAHs. A high-volume air sampler was used to
pull partial flow (300 L/min) of diesel exhaust generated
at idle mode (0-kW engine load) through a moisture-free
quartz filter and a glass cartridge packed with XAD-2 resin
trapped in between polyurethane plugs (PUF). Sample
preparation involves the extraction of the quartz filter,
PUF, and XAD in high-performance liquid chromatogra-
phy grade dichloromethane (DCM) (Fisher Scientific)
Volume 58 October 2008
Saiyasitpanich, Keener, Lu, Liang, and Khang
with sonication. The detailed procedures of DPM extrac-
tion and instrumental analysis have been published else-
where.33 The extracted solution was concentrated by ni-
trogen purging to approximately 1 mL for GC/MS
analysis.
Determinations of CO, NOx and O3
Concentrations
A calibrated digital gas combustion analyzer (TESTO 350)
and the O3 analyzer (Dasibi, Model 1003-AH) were used to
measure the inlet and outlet concentrations of CO and
NOx, and O3 in diesel exhaust, respectively. The samples
were taken during four engine modes (0, 25, 50, 75 kW)
with varying gas residence time inside the wESP (⬃0.1–
0.3 sec). Different exhaust flows were fed into the wESP to
observe the effect of gas residence time or SCA on CO
removal. The wESP was energized at 70 kV in all tests, thus
generating a constant corona current of 3 mA.
Determination of wESP Removal Efficiency
The DPM mass and number, n-alkanes (C10-C25), 16
PAHs, CO, and NOx removal efficiencies by a wESP were
computed using the following equation. The inlet and
outlet concentrations were measured at the exit of the
wESP, with the ESP turned off and on, respectively.
removal ⫺ efficiency
⫽ 冉 Inlet concentration ⫺ Outlet concentration
Inlet concentration 冊 ⫻ 100
RESULTS AND DISCUSSION
Removal of DPM Mass and Number
Concentrations
Figure 2 presents the reduction of DPM mass and number
concentrations as a function of gas residence time at
varying engine loads. For a measurement of the DPM
number concentration, the temperatures of the diluted
gas were kept in the range of approximately 4 –7 °C, rela-
tive humidity (RH) between 64 –79%, and dilution ratios
of 7–32, respectively. The wESP was operated at a constant
voltage of 70 kV, corresponding to a corona current of 3
mA in all tests. Varying flow rates of exhaust gas (slip
stream) were fed into the wESP to create different gas
residence times of 0.1 and 0.16 sec for tests at 0, 25, and
50 kW, and 0.1– 0.4 sec for tests at 75 kW, For tests at 0,
25, and 50 kW this corresponded to a ratio of the exhaust
fed into the wESP to the total exhaust emitted from the
engine to values of 100 and 77%, respectively. For tests at
75 kW, this corresponded to ratios between 25 and 100%.
The DPM mass concentrations in untreated diesel
exhaust ranged from approximately 2.8 to 15.9 mg/N䡠m3
for 0- to 75-kW engine loads. The DPM number concen-
trations were higher at the lower engine load conditions
but they were in the magnitude of approximately 108/cm3
with particle diameter modes between 20 and 40 nm
(measured particle size range of 10 –1000 nm).34 The
higher DPM number concentration measured at the lower
load is believed to be primarily because of dilution, cool-
ing, and the amount of unburned fuel and lubricating oil
Journal of the Air & Waste Management Association 1313
Saiyasitpanich, Keener, Lu, Liang, and Khang

100 Removal of n-Alkanes and 16 PAHs

Figure 3, a and b, present the total concentrations of
n-alkanes and 16 PAHs in diesel exhaust (gas phase) and
in DPM (particle phase) taken at the idling mode (0 kW,
90 100% flow rate) with wESP off and on. The untreated
concentrations of these compounds were approximately
an order of magnitude higher in the gas phase than in the
Removal Efficiency, %

particle phase. With the wESP energized at 60 kV corre-

80
sponding to 2 mA, the results show a significant reduction
of these compounds in both phases. The removal efficien-
Generator Powers
cies of n-alkanes and 16 PAHs are 46.5 and 57% in the
70
particle phase, respectively, and 46 and 31% in the gas
0 kW-M
0 kW-N
phase, as shown in Figure 3c. The removal of these com-
25 kW-M pounds in the particle phase is mainly due to the capa-
25 kW-N
50 kW-M
bility of the wESP to efficiently remove mass- and num-
60 50 kW-N ber-based DPM as presented in the preceding section.
75 kW-M
75 kW-N

(a)
120
50

0 0.1 0.2 0.3 0.4 100

Concentration, ng/Nm 3
Gas Residence Time, s
80 Inlet Concentration
Figure 2. wESP total collection efficiency vs. gas residence time Outlet Concentration
measured for four engine loads (wESP at 70 kV, 3 mA, negative 60
polarity, dc voltage, M ⫽ mass concentration, n ⫽ number
concentration). 40

20
in the exhaust,30 consequently promoting the particle
nucleation process.35,36 0
The results indicate that the removal efficiencies of Alkanes x 1000 16PAHs x 100
the DPM mass and number concentrations were strongly
influenced by both gas residence time and engine loads. (b)
500000
At 100% exhaust flow rate with a gas residence time of 450000
approximately 0.1 sec, the removal efficiency of the DPM
Concentration, ng/Nm 3

400000
mass increased from 69 to 83%, and the removal effi-
350000
ciency of the DPM number concentration increased from
300000
67 to 86% when the engine load was reduced from 75 to 250000
Inlet Concentration
0 kW. The effect of decreasing engine power in turn led to 200000
Outlet Concentration
a production of lower DPM mass and number concentra- 150000
tions, higher organic carbon (OC) and elemental carbon
100000
(EC) fractions, and higher gas-phase HC concentra- 50000
tions37–39 were more pronounced than that of the resi- 0
dence time at a given condition. It is believed that a Alkanes x 10 16PAHs
reduction of DPM mass and number loading to be treated
inside the wESP and a change in OC/EC fraction may (c)
attenuate the effect of corona current suppression and 100

allow the particles to impart charge more efficiently. It is 90

80
Removal Efficiency, %

also believed that the wESP may efficiently remove the

unburned fuel and lubricating oil from the diesel ex- 70

haust,40 thus suppressing the formation of submicrometer 60

Particle Phase
particles (gas-to-particle conversion) inside the minidilu- 50
Gas Phase
tion chamber during the number concentration measure- 40

ment, thereby leading to higher DPM number removal 30

efficiency. At increasing gas residence times or, in other 20

words, increasing wESP SCA, the removal efficiency of the 10

wESP on the DPM number and mass concentrations in- 0

creased regardless of the engine load. The wESP was capa-
ble of removing approximately 69 –92% and 67–96% at 75
kW, approximately 71– 82% and 74 – 81% at 50 kW, ap-
proximately 82– 88% and 83– 88% at 25 kW, and approx-
imately 83–96% and 86 –90% at 0 kW on the basis of total
DPM mass and number concentrations, respectively.34,41
Alkanes 16PAHs
Figure 3. n-Alkanes (C10-C25) and 16 PAH removal in the wESP:
(a) particle-phase concentration, (b) gas-phase concentration, and
(c) gas- and particle-phase removal efficiencies (wESP at 60 kV, 2
mA, negative polarity, dc voltage, 0-kW engine load, 100% exhaust
flow rate).

1314 Journal of the Air & Waste Management Association Volume 58 October 2008
 Saiyasitpanich, Keener, Lu, Liang, and Khang

(a)
concentration decreased from approximately 250 to 25
300
ppm when the engine load increased from 0 to 75 kW.
The explanation is that increasing loads on diesel engines
250 (lean burn engine) leads to an increase in the combustion
CO concentration, ppm

a) Concentration

200
temperature, which combined with the high level of ex-
No wESP
wESP-T1
cess oxygen (⬃12%) at these loads results in lower CO
150 emissions when compared with low load conditions.37,39
wESP-T2
100 wESP-T3 With the wESP on, the CO concentration decreased and
the CO removal efficiency ranged from 5 to 38%, depend-
50 ing on the CO initial concentration and gas residence
0
time inside the wESP as depicted in Figure 4b. The re-
0 25 50 75 moval efficiency approached the maximum and tended
Engine Load, kW to be stable when the gas residence time was greater than
(b) approximately 0.3 sec. The removal of CO in corona dis-
charge is proposed to be by the reaction with OH䡠 and the
40
reaction products are CO2 and H2O.24 However, the re-
sultant CO2 conversion was not measurable in this study
b) Removal Efficiency because CO2 concentration in diesel exhaust (1– 6%) is
much larger than the small increment by converted CO.
30
wESP On-70 kV/3mA
Figure 5 presents the concentrations of NOx in the
diesel exhaust. These results do not show significant re-
CO Removal Efficiency, %

20
10
0
0 0.1 0.2 0.3
Residence Time, second
Figure 4. CO removal in the wESP: (a) inlet and outlet CO con-
centrations, and (b) CO removal efficiency (wESP at 70kV, 3mA,
negative polarity, dc voltage, exhaust flow rate: T1 ⫽ 100%, T2 ⫽
77%, T3 ⫽ ⬃40%).
Moreover, the removal of the gas-phase compounds may
attenuate the condensation and adsorption effects of
these compounds on the DPM surface during sampling,
thus enhancing the particle-phase removal efficiency.
Many mechanisms have been reported for the removal of
volatile and semi-volatile organic compounds such as
75 kW
50 kW
25 kW
duction or increases for the conditions tested. The main
0 kW mechanisms proposed for the oxidation of NO in the
corona discharge are the reactions of NO with O3 and O to
form NO2.43,44 However, O3 is believed to be the domi-
nant species because the lifetime of O is very short and
the reaction rate constant is very small.44 In the presence
of water vapor, the OH radicals produced from water
vapor significantly deplete O, which leads to much
reduced O3 production in the corona discharge pro-
cess,28 therefore appreciably suppressing the oxidation
of NO to NO2. The water content in the exhaust gas
measured in this study ranged from approximately 2 to
9%. There is also evidence supporting the formation of
0.4 0.5
NO in the corona discharge system (2.82 ⫾ 0.17 ⫻ 1016
and 2.93 ⫾ 0.17 ⫻ 1016 molecules/J at 40 and 90% RH,
respectively);25 however, at this rate the generation of
NO in the wESP used in this study is less than 1 ppm
and can be neglected. NO2 can also react with the OH
radical to form HNO3. However, the reaction rate con-
stant of NO2 with OH䡠 is approximately 3 orders of mag-
nitude smaller than that of O with OH䡠.43,45 This could
explain why there were no measurable changes in the
NOx concentrations in this study.
1600

benzenes, p-dichlorobenzene, n-decane, n-undecane, etc., 1400

NOx Concentration, ppm

in air in plasma discharge processes. The main mecha- 1200

nism for removal of aromatic compounds appears to be a 1000 No wESP
chain polycondensation on the anode surface (collecting 800
wESP-T1
electrode) induced by dissociative electron attachment.22 wESP-T2
600
The removal of n-alkane compounds are primarily via the wESP-T3
400
formation of negative-ion clusters, the chemical reaction
with OH䡠, and also a chain polycondensation.22,23,40 – 42 200

The results obtained here also support the assumptions 0

given in the preceding section. 0 25 50 75
Engine Load, ppm

Removal of CO and NOx Concentrations Figure 5. NOx removal in the wESP: (wESP at 70kV, 3mA, nega-
Figure 4a presents the CO concentration in diesel exhaust tive polarity, dc voltage, exhaust flow rate: T1 ⫽ 100%, T2 ⫽ 77%,
gas with and without wESP operation. The original CO T3 ⫽ ⬃40%).

Volume 58 October 2008 Journal of the Air & Waste Management Association 1315
Saiyasitpanich, Keener, Lu, Liang, and Khang
Table 1. Ozone concentration in the diesel exhaust.
Engine Power Exhaust Flow Rate WESP Voltage (kV),
(kW) (m3/hr)act Current (mA)
0 366 70,3
75 123 (optimal flow rate) 70,3
75 385 70,3
75 497 70,3
O3 Concentration in the Diesel Exhaust
Table 1 presents O3 concentrations measured at 0- and
75-kW engine loads. The wESP operated at 70 kV and 3
mA in all tests. The dilution ratio was fixed at 2.9 and 2.7
at 0- and 75-kW engine loads, respectively. NOx was used
as a measuring index of a dilution ratio. The results indi-
cate that the O3 concentration in diesel exhaust undergo-
ing the nonthermal plasma treatment process inside the
wESP is inconsequential (⬍1 ppm). At the optimal flow
rate (123 [m3/hr]actual ⫽ residence time of ⬃0.4 sec), at
which the wESP achieved more than 90% removal of
mass- and number-based DPM, the O3 concentration was
approximately 0.5 ppm. As discussed before, the O3 is
formed primarily by the O-O2 reaction in the plasma
zone. Several parameters such as water vapor, gas flow
rate, and energy density applied in the wESP affect the O3
formation.28,46,47 At atmospheric pressure, the kinetic
processes govern the O3 production regardless of gas flow
velocity.47 The presence of water vapor leads to the pro-
duction of OH radicals, which react with O and restrict
the O3 formation.28 At the energy density (corona power/
exhaust volumetric flow rate) of 0.1 W䡠hr/L in air, the O3
concentration was reported to be approximately 200 ppm
at a gas residence time of approximately 0.1 sec.47 In this
study, the energy density was calculated to be approxi-
mately 4 ⫻ 10⫺4 W䡠hr/L (corona power ⫽ 210 W, exhaust
flow rate ⫽ 500 [m3/hr]actual, residence time ⫽ ⬃0.1 sec).
It is thus assumed that the O3 concentration in the diesel
exhaust gas that underwent treatment in the wESP was
very low.
CONCLUSIONS
Removal of diesel emissions using a tubular wESP was
investigated as a function of corona power, engine loads,
and gas residence times. The results indicated that the
total mass- and number-based DPM removal was strongly
influenced by engine loads and gas residence times. The
wESP operating at 70 kV and 3 mA was able to remove
approximately 67– 86% of DPM at 100% exhaust volu-
metric flow rate generated from a diesel generator run-
ning from 0- to 75-kW engine loads. As engine loads
decreased from 75 to 0 kW, the DPM removal increased
significantly by 16% on the average. Reducing engine
loads leads to consequential changes in DPM mass con-
centration, OC/EC fraction, HC emissions, and a slight
increase in gas residence time. This may favor the particle-
charging mechanism inside the wESP and thus enhance
the removal efficiency. The DPM removal efficiency also
increased significantly with increasing gas residence times
(increasing SCAs) inside the wESP regardless of the engine
load. At a 75-kW engine load, increasing the gas residence
1316 Journal of the Air & Waste Management Association
Initial NOx Diluted NOx Dilution Dilution-Corrected Standard Deviation
(ppm) (ppm) Ratio O3 (ppm) (%)
231 80 2.9 0.602 13
1350 500 2.7 0.496 6
1350 500 2.7 0.398 8
1350 500 2.7 0.411 11
time from approximately 0.1 to 0.4 sec led to a significant
increase of DPM removal efficiency from approximately
67 to more than 90%.
The removal of n-alkanes and 16 PAHs in the wESP
ranged from 31–57%. The main mechanisms responsible
for the removal of these compounds were a chain poly-
condensation on the anode surface (collecting electrode)
induced by dissociative electron attachment, negative ion
clusters, and a chemical reaction with OH䡠.
The CO concentration in diesel exhaust decreased
when the wESP was turned on and the CO removal effi-
ciency ranged from 5 to 38% depending on the CO initial
concentration and gas residence time inside the wESP.
The maximum removal efficiency can be obtained when
the gas residence time was greater than approximately 0.3
sec. The removal mechanism proposed may be by the
reaction with OH radical, which produces CO2 and H2O
as the final products.
The wESP did not affect the NOx concentration sig-
nificantly as observed in this study. The possible explana-
tion is that the presence of OH radical led to a significant
depletion of O, thus suppressing O3 formation. The O3
concentration in diesel exhaust that underwent the wESP
treatment was measured to be below 1 ppm, which was
inadequate to oxidize NO to NO2.
ACKNOWLEDGMENTS
The authors gratefully acknowledge financial support
from the National Institute for Occupational Safety and
Health.
REFERENCES
1. Lloyd, A.C.; Cackette, T.A. Diesel Engines: Environmental Impact and
Control; J. Air & Waste Manage. Assoc. 2001, 51, 809-847.
2. Walsh, M.P. Global Trends in Diesel Emissions Regulations. Society of
Automotive Engineers (SAE) Technical Paper 010183; SAE: Warren-
dale, PA 2001.
3. Donaldson, K.; Li, X.Y.; MacNee, W. Ultrafine (Nanometer) Particle
Mediated Lung Injury; J. Aerosol Sci. 1998, 29, 553-560.
4. Nygaard, U.C.; Samuelsen, M.; Aase, A.; Løvik, M. The Capacity of
Particles to Increase Allergic Sensitization is Predicted by Particle Num-
ber and Surface Area, Not by Particle Mass; Toxicol. Sci. 2004, 82,
515-524.
5. Dockery, D.W.; Pope, C.A.; Xu, X.; Spengler, J.D.; Ware, J.H.; Fay, M.E.;
Ferris, B.G.; Seizer, F.E. An Association between Air Pollution and
Mortality in Six U.S. Cities; N. Engl. J. Med. 1993, 329, 1753-1759.
6. Pope, C.A.; Thun, M.J.; Namboodriri, M.M.; Dockery, D.W.; Evans,
J.S.; Speizer, F.E.; Heath, C.W. Particulate Air Pollution as a Predictor of
Mortality in a Prospective Study of U.S. Adults; Am. J. Respir. Crit. Care
Med. 1995, 151, 669-674.
7. Christensen, H.; Rak, Z.S. A Novel Diesel Particulate Converter; Catal.
Today 2002, 75, 451-457.
8. Lepperhoff, G.; Lüders, H.; Barthe, P.; Lemaire J. Quasi-Continuous
Particle Trap Regeneration by Certain Cerium Additives; Society of
Automotive Engineers (SAE) Technical Paper No. 950369; SAE: War-
rendale, PA, 1995.
Volume 58 October 2008

9. Shi, X.; Pang, X.; Mu, Y.; He, H.; Shuai, S.; Wang, J.; Chen, H.; Li, R.
Emission Reduction Potential of Using Ethanol-Biodiesel-Diesel Fuel
Blend on a Heavy-Duty Diesel Engine; Atmos. Environ. 2006, 40, 2567-
2574.
10. Walker, A.P. Controlling Particulate Emissions from Diesel Vehicles;
Top. Catal. 2004, 28, 165-170.
11. Bagley, S., Gratz, L. Evaluation of Biodiesel Fuel and Oxidation Catalysts
in an Underground Metal Mine, Part 3: Biological and Chemical Charac-
terization; Diesel Emissions Evaluation Program Technical Support;
1998; available at http://www.deep.org (accessed 2008).
12. Carder, D.K. Thesis, West Virginia University, Department of Mechan-
ical and Aerospace Engineering, 1999.
13. Pataky, G.M.; Baumgard, K.J.; Gratz, L.D.; Bagley, S.T.; Leddy, D.G.;
Johnson, J.H. Effects of an Oxidation Catalytic Converter on Regu-
lated and Unregulated Diesel Emissions; Society of Automotive Engi-
neers (SAE) Technical Paper No. 940243; SAE: Warrendale, PA, 1994.
14. Baz-Dresch, J.J.; Bickel, K.L.; Watts, W.F. Evaluation of Catalyzed Diesel
Particulate Filters in an Underground Metal Mine; RI 9478; U.S. Depart-
ment of Interior, Bureau of Mines: Minneapolis, MN, 1993.
15. Larsen, C.A.; Ledevis, Y.A.; Shimato, K. Filtration Assessment and
Thermal Effects of Aerodynamic Regeneration in Silicon Carbide and
Cordierite Particulate Filters. Society of Automotive Engineers (SAE)
Technical Paper 010466; SAE: Warrendale, PA, 1999.
16. Biamino, S.; Fino, P.; Fino, D.; Russo, N.; Badini, C. Catalyzed Traps for
Diesel Soot Abatement: In Situ Processing and Deposition of Perov-
skite Catalyst; Appl. Catal. B Environ. 2005, 61, 297-305.
17. Nguyen Huu Nhon, Y.; Mohamed Magan, H.; Petit, C. Catalytic Diesel
Particulate Filter Evaluation of Parameters for Laboratory Studies; Appl.
Catal. B Environ. 2004, 49, 127-133.
18. Stratakis, G.A.; Psarianos, D.L.; Stamatelos, A.M. Experimental Inves-
tigation of the Pressure Drop in Porous Ceramic Diesel Particulate
Filters; J. Auto. Engineer. 2002, 216, 773-784.
19. Altman, R.; Buckley, W.; Ray, I. Wet Electrostatic Precipitation: Dem-
onstrating Promise for Fine Particulate Control; Power Engineer. 2001,
105, 37-44.
20. Yao, S.; Okumuto, M.; Yashima, T. Diesel Particulate Matter and NOx
Removals Using a Pulsed Corona Surface Discharge; AIChE J. 2004, 50,
715-721.
21. Chae, J.-O. Non-Thermal Plasma for Diesel Exhaust Treatment; J.
Electrostats. 2003, 57, 251-262.
22. Tamon, H.; Imanaka, H.; Sano, N.; Okazaki, M. Removal of Aromatic
Compounds in Gas by Electron Attachment; Ind. Eng. Chem. Res.
1998, 37, 2270-2774.
23. Sano, N.; Nagamoto, T.; Tamon, H.; Suzuki, T.; Okazaki, M. Removal
of Acetaldehyde and Skatole in Gas by a Corona-Discharge Reactor;
Ind. Eng. Chem. Res. 1997, 36, 3783-3791.
24. Rudolph, R.; Francke, K.-P.; Miessner, H. Determination of OH Radical
Concentration in Dielectric Barrier Discharges by CO Oxidation. Pre-
sented at the 8th International Symposium on High Pressure Low
Temperature Plasma Chemistry, Pühajärve, Estonia, 2002; pp 323-326.
25. Rehbein, N.; Cooray, V. NOx Production in Spark and Corona Dis-
charges; J. Electrostats. 2001, 51–52, 333-339.
26. Chen, J.; Davidson, J.H. Ozone Production in the Negative DC Co-
rona: the Dependence of Discharge Polarity; Plasma Chem. Plasma
Proc. 2003, 23, 501-518.
27. Pignolet, P.; Hadj-Ziane, S.; Held, B.; Peyrous, R.; Benas, J.M.; Coste, C.
Ozone Generation by Point to Plane Discharge; J Phys. D Appl. Phys.
1990, 23, 1069-1072.
28. Ono, R.; Oda, T. Dynamics of Ozone and OH Radicals Generated by
Pulsed Corona Discharge in Humid-Air Flow Reactor Measured by
Laser Spectroscopy; J. Appl. Phys. 2003, 10, 5876-5882.
29. Schauer, J.; Kleeman, M.; Cass G. Characterization and Control of Or-
ganic Compounds Emitted from Air Pollution Sources; California Air Re-
source Board: Sacramento, CA, 1998; pp 93-329.
30. Ning, Z.; Cheung, C.S.; Liu, S.X. Experimental Investigation of the
Effect of Exhaust Gas Cooling on Diesel Particulate; J. Aerosol Sci.
2004, 35, 333-345.
31. Cooper, D.C.; Alley, F.C. Air Pollution Control: a Design Approach, 3rd
ed.; Waveland: Prospect Heights, IL, 2002.
32. Gulijk, C.V.; Marijnissen, J.C.M.; Makkee, M.; Moulijn, J.A. Oil-Soaked
Sintered Impactors for the ELPI in Diesel Particulate Measurements; J.
Aerosol. Sci. 2003, 34, 635-640.
Volume 58 October 2008
Saiyasitpanich, Keener, Lu, Liang, and Khang
33. Liang, F.; Lu, M.; Keener, T.C.; Liu, Z. The Organic Composition of
Diesel Particulate Matter, Diesel Fuel, and Engine Oil of a Non-Road
Diesel Generator; J. Environ. Monit. 2005, 7, 983-988.
34. Saiyasitpanich, P.; Keener, T.C.; Khang, S.-J.; Lu, M.; Evans, D.E. Col-
lection of Ultrafine Diesel Particulate Matter (DPM) in Cylindrical
Single-Stage Wet Electrostatic Precipitators; Environ. Sci. Technol.
2006, 40, 7890-7895.
35. Kittelson, D.B. Engines and Nanoparticles: a Review; J. Aerosol Sci.
1998, 28, 575-588.
36. Kim, D.; Gautam, M.; Gera, D. Parametric Studies on Formation of
Diesel Particulate Matter via Nucleation and Coagulation Modes; J.
Aerosol. Sci. 2002, 33, 1609-1621.
37. Saiyasitpanich, P.; Keener, T.C.; Khang S.-J.; Lu, M.; Liang, F. The
Effect of Diesel Fuel Sulfur Content on Particulate Matter Emissions for
a Nonroad Diesel Generator; J. Air & Waste Manage. Assoc. 2005, 55,
993-998.
38. Liu, Z.; Lu, M.; Eileen, B.M.; Keener, T.C.; Khang, S.-J.; Liang, F.
Variations of the Particulate Carbon Distribution from a Nonroad
Diesel Generator; Environ. Sci. Technol. 2005, 39, 7840-7844.
39. Shi, J.P.; Harrison, R.M.; Brear, F. Particle Size Distribution from a
Modern Heavy Duty Diesel Engine; Sci. Total Environ. 1999, 235,
305-317.
40. Mok, S.Y.; Nam, C.M.; Cho, M.H.; Nam, I.-S. Decomposition of Vola-
tile Organic Compounds and Nitric Oxide by Nonthermal Plasma
Discharge Processes; IEEE Trans. Plasma Sci. 2002, 30, 408-416.
41. Saiyasitpanich, P.; Keener, T.C.; Khang, S.-J.; Lu, M.; Evans, D.E. Re-
moval of Diesel Particulate Matter (DPM) in Tubular Wet Electrostatic
Precipitators; J. Electrostats. 2007, 65, 618-624.
42. Nolting, F.; Behnke, W.; Zetzsch, C. A Smog Chamber for Studies of
the Reactions of Terpenes and Alkanes with Ozone and OH; J. Atmos.
Chem. 1988, 6, 47-59.
43. Dong, L.; Wang, P.; Yang, J. The Role of H2O in Nitric Oxide Removal
by Wet-Type Corona Discharge Plasma Reactor; Jpn. J. Appl. Phys.
2006, 45, 4222-4223.
44. Mok, Y.S.; Nam, I.-S. Removal of Nitric Oxide in a Pulsed Corona
Discharge Reactor; Chem. Eng. Technol. 1999, 22, 527-532.
45. Robertson, R.; Smith, G.P. Temperature Dependence of O and OH at
136 –377 K Using Ozone Photolysis; J. Phys. Chem. A 2006, 110,
6673-6679.
46. Pignolet, P.; Hadj-Ziane, S.; Held, B.; Peyrous, R.; Benas, J.M.; Coste, C.
Ozone Generation by Point to Plane Corona Discharge; J. Phys. D:
Appl. Phys. 1990, 23, 1069-1072.
47. Peyrous, R.; Monge, C.; Held, B. Ozone Generation in Air and N2⫹O2
Mixtures in Wire-to-Cylinder Reactors—an Optimization Attempt; J.
Phys. 1999, 49, 289-299.
About the Authors
Dr. Phirun Saitasitpanich was a student in the Department
of Civil and Environmental Engineering when this research
was conducted. He is currently working with the Thai gov-
ernment for the Industrial Development Authority in
Bangkok, Thailand. Drs. Tim Keener and Mingming Lu are
faculty members in the Department of Civil and Environ-
mental Engineering, and Dr. Soon-Jai Khang is in the De-
partment of Chemical and Materials Engineering at the
University of Cincinnati. Please address correspondence
to: Tim Keener, University of Cincinnati, Department of Civil
and Environmental Engineering, 765 Baldwin Hall, P.O. Box
210071, Cincinnati, OH 45211-0071; phone: ⫹1-513-556-
3676; fax: ⫹1-513-556-2599; e-mail: tim.keener@uc.edu.
Journal of the Air & Waste Management Association 1317