Accepted Manuscript

Potentials of OME/diesel blends for stationary power production - Improving

emission characteristics of a diesel CHP unit

T. Popp, R. Lechner, M. Becker, M. Hebauer, N. O'Connell, M. Brautsch

PII: S1359-4311(18)37253-3
DOI: https://doi.org/10.1016/j.applthermaleng.2019.03.015
Reference: ATE 13444

To appear in: Applied Thermal Engineering

Received Date: 26 November 2018

Revised Date: 28 January 2019
Accepted Date: 4 March 2019

Please cite this article as: T. Popp, R. Lechner, M. Becker, M. Hebauer, N. O'Connell, M. Brautsch, Potentials of
OME/diesel blends for stationary power production - Improving emission characteristics of a diesel CHP unit,
Applied Thermal Engineering (2019), doi: https://doi.org/10.1016/j.applthermaleng.2019.03.015

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Potentials of OME/diesel blends for stationary power production - Improving emission
characteristics of a diesel CHP unit
Popp, T.a, Lechner, R.a,b,*, Becker, M.a, Hebauer, M.a, O’Connell, N.b, Brautsch, M.a,b
a
Center of Excellence for Cogeneration Technologies, Technical UAS Amberg-Weiden, Kaiser-
Wilhelm-Ring 23, 92224 Amberg, Germany
b
Institute of Energy Technology at the Technical UAS Amberg-Weiden, Kaiser-Wilhelm-Ring
23a, 92224 Amberg, Germany
*Corresponding Author: r.lechner@oth-aw.de
Declarations of interest: none
Polyoxymethylene dimethyl ethers (OME) with the chemical structure H3CO[-CH2-O-]nCH3 are a promising
alternative fuel for improving emissions characteristics of internal combustion engines and can be
produced from renewable sources. Especially the PM-NOx trade-off of diesel engines is known to be
effectively mitigated by blending OME with diesel. Recent investigations concentrate on the utilization
OME fuels in mobile applications. However, these fuels could also be of interest for stationary engines
used in gensets and combined heat and power units. In the present paper we examine the use of OME2-
6/diesel blends up to 50% in a CHP unit in practical operation equipped with a heavy-duty six-cylinder
engine, applying different load conditions and engine settings, including Exhaust Gas Recirculation (EGR)
and the adjustment of injection timings. No negative effects on the performance and efficiency of the
engine were observed with the OME blends. Particulate Matter (PM) and Particle Number (PN) decreased
significantly with increasing content of OME in the blend, especially when EGR was applied. The Particle
Number Size Distribution (PNSD) showed a trend towards smaller particle sizes. NOx rose slightly in direct
comparison, but this was by far overcompensated by the reduction achieved through EGR. The results
indicate that OME fuels in combination with EGR could become a feasible option for emission reduction in
diesel gensets and CHP units due to the very low particle emissions, which mitigate the risk of EGR cooler
fouling.
Keywords: Polyoxymethylene dimethyl ether, OME, CHP, PM-NOx trade-off, particle emissions,
particle number size distribution
1. Introduction
Diesel engine powered gensets and cogeneration units are widely used for decentralized
power provision in cities, at industrial sites and construction areas. With increasing concerns
about air pollution in metropolitan areas it becomes more and more important to strictly
control exhaust gas emissions from these units. While carbon monoxide (CO) and
hydrocarbon (HC) emissions can usually be satisfactorily reduced with diesel oxidation
catalysts (DOC), the reduction of nitrous oxides (NO x) and particle emissions demands
higher efforts. In order to limit NOx-emissions, primary measures, such as exhaust gas
recirculation (EGR), Miller cycles and retarded fuel injection, or secondary measures, such
as exhaust gas aftertreatment by Selective Catalytic Reduction (SCR) are normally required.
The primary measures for NOx-reduction are targeted towards reducing peak combustion
temperatures, thus leading to reduced formation of thermal NOx, but at the same time to
inferior oxidation of the fuel jet and thus to higher particle emissions [1–4]. This characteristic
can be observed for all conventional diesel fuels and is commonly denominated as PM-NOx
trade-off. To remove particles from the exhaust gas, again secondary measures are
required, such Diesel Particle Filters (DPF).
EGR is a commonly used technology to reduce the NO x raw emissions of diesel engines
[5,6]. The gas is recirculated via a heat exchanger also known as EGR cooler [7]. One of the
main problems when applying EGR is the deposition of particles in the EGR cooler [8].
Particles in diesel exhaust gas cause a deposition on the wall of the EGR cooler which
decreases heat transfer between exhaust gas and coolant and consequently leads to a
reduction of the NOx reduction capability [5]. This problem is commonly known under the
term EGR cooler fouling [5]. There are four possible fouling mechanisms: thermophoresis,
diffusion by HC condensation, gravitational settling and Brownian diffusion
[5].Thermophoresis is the main mechanism that leads to a transport of soot particles to the
tube surface. This is due to the small size of the exhaust gas particles in a diesel engine [6].
In applications with high operating hours, such as CHP units, the use of EGR might cause
problems due to EGR cooler fouling. Usually, the EGR cooler will have a low durability in
these applications and will have to be cleaned or replaced regularly, leading to higher
maintenance costs and lower availability of the genset.
In recent years new, synthetically produced fuels have been investigated in order to
overcome the soot-NOx trade-off in diesel engines [9–13], such as polyoxymethylene
dimethyl ethers (OME) with the chemical structure H3CO[-CH2-O-]nCH3. Due to their high
oxygen content and the absence of C-C bonds, OME fuels combust nearly free from particle
emissions [9,10,14]. In addition to its favorable combustion characteristics OME can be
produced from renewable sources, starting from methanol, which can be derived from
biological processes [15–18] or gas-to-liquid processes [19]. Up to now, OME is primarily
being investigated as an alternative fuel for mobile applications in light-duty (passenger car)
[9,10,20–22] and heavy-duty (truck) engines [11,14,19,23,24]. Different chain lengths (n) of
OME molecules were tested for these applications.
Härtl et al. [11] investigated blends of reference diesel B0 with OME1a, which is OME 1
blended with additives for better viscosity, lubricity and cetane number, in a heavy-duty one
cylinder research engine. The soot-NOx trade-off observed with the reference fuel was not
existent for the alternative fuel. An increase of the EGR rate, resulting in a decrease of NOx
emissions, did not lead to higher particle number (PN) and soot mass emissions, as it was
the case with the reference fuel.
Feiling et al. [10] used neat OME1b, which is OME1 with some additives for improved fuel
properties, for investigations on a low-duty one-cylinder research engine. The authors chose
OME1 because it is already produced in large scale and easily available. Due to the lower
boiling point of OME1b, the fuel had to be cooled to prevent cavitation in the injection
system. With the alternative fuel, higher NOx emissions, but also significantly lower particle
number and soot mass emissions compared to the reference fuel (diesel with 7% FAME)
were observed in the engine tests. The authors did not consider EGR.
Pélerin et al. [24] used compared OME1 and OME3-6 to Paraffinic Diesel Fuel (PDF) as
reference fuel in a heavy-duty engine (MAN D26) that was externally charged. Independent
of the EGR, low particle number and soot mass could be achieved with both OME fuels. The
research group also investigated a stoichiometric operation with OME by increasing the EGR
rate up to air-to-fuel equivalence ratios of one. At this equivalence ratio, higher formaldehyde
emissions in comparison to the reference fuel were observed. At the same time also the
methane emissions were significantly higher.
Iannuzzi et al. [23] argue that OME1 (Dimethoxymethane, DMM) requires modifications to
the injection system to prevent vapor lock. This is why they prefer OMEs with longer chains
over DMM. According to the authors, higher chain lengths could lead to precipitation at low
temperatures and therefore to clogging of the filter system. Hence, OMEs with the chain
length 2-5 were identified as suitable for application in unmodified injection systems. The
research team tested diesel blended with OME fuel (chain length 2-7) in a heavy-duty one-
cylinder research engine. Using a specially developed optical light probe for three color
pyrometry, they found out, that the initial soot formation with a 10% blend was on the same
level as with the reference fuel. However, the subsequent soot oxidation is improved with the
blended fuel, which eventually results in lower soot emissions compared to conventional
diesel.
Investigations with a 7,5% OME3-5/diesel blend on a light-duty engine in a passenger car
operated on a chassis dynamometer were carried out by Pellegrini et al. [20]. They
measured the Particle Number Size Distribution (PNSD) of the exhaust gas downstream of
the Diesel Oxidation Catalyst (DOC) with a Dual Differential Mobility Particle Spectrometer
(DDMPS). For high load conditions, a shift of the PNSD towards smaller particles was
measured. The authors stated that this may be explained by a larger fuel jet penetration
which results in a better fuel atomization and therefore in an increased number and smaller
size of the fuel droplets (c.f. [25] in [20]).
Liu et al. [21] tested OME3-4/diesel blends (10 and 20%) in a low duty research-engine,
retrofitted from a four cylinder common-rail engine. In high load condition, the ignition delay
increased with the OME content of the fuel. At the same time, the combustion duration was
observed to be shorter. However, the authors speak holistically of no significant influence of
the OME fuels on the combustion process. Despite of having an oxygen content of only 10%,
a reduction of the soot emissions by 90% was achieved. The corresponding NOx emissions
were slightly higher, compared to the reference fuel (market diesel).
Li et al. [22] analyzed the use of an OME2-4/biodiesel blend (15/85) in a low-duty one-
cylinder research engine and compared their results with those obtained when using
biodiesel and market diesel. Different EGR rates and combustion phasings were
investigated. Due to a higher volatility, blending of biodiesel with OME resulted in a shorter
ignition delay, an accelerated combustion and a shorter combustion duration. These effects
led to the highest Indicated Thermal Efficiency (ITE) of the blend (compared to the reference
fuels) in all tested operating conditions. Besides the positive effect on the engine’s efficiency,
the emission characteristics could also be significantly improved by using OME. The soot-
NOx trade-off was mitigated mainly due to lower combustion temperatures and considerably
reduced soot emissions.
Liu et al. [26] tested OME3-8/diesel blends (10, 20 and 30%) in heavy duty four-cylinder
diesel engine. At full load, they observed a reduction of fuel consumption with all blends due
to a higher engine efficiency. Using OME blends resulted in a slight increase of NO x
emissions compared to diesel. Due to its high cetane number and oxygen content, the
utilization of OME resulted in lower HC emissions. In addition, a significant reduction of
smoke emissions, especially at high load conditions, was achieved with the blends.
According to the authors the intramolecular oxygen in the OME lead to a considerable
reduction of CO emissions because of improved combustion conditions. The research group
also measured the Particle Number Size Distribution (PNSD) when using the OME blends by
means of a differential mobility spectrometer. They observed a smaller mean diameter of the
PNSD with increasing OME content in the blend. Using the OME blend caused an increase
of nucleation-mode particles (10-40 nm) with a simultaneous decrease of accumulation
mode particles. They found out that the accumulation-mode particles play a major role in the
mass concentration distribution, while it is hardly influenced by nucleation-mode particles.
The existing research focuses mainly on the use of OME in mobile applications. So far, no
research results on the use of OME in stationary power production have been published.
However, OME fuels could be an interesting option in this regard, since these units are often
used in urban areas, where air pollution from particulate matter is a serious issue. Our aim
was thus to evaluate the potential of OME for emission reduction in diesel powered gensets
and Combined Heat and Power (CHP) units. We paid special attention to the Particulate
Matter (PM) emissions, the Particle Number (PN) and the Particle Number Size Distribution
(PNSD), since these are crucial with regard to EGR cooler fouling. Lowering the particle
emissions by using OME/diesel blends could help to overcome this problem, thus enabling
effective NOx control strategies with EGR also in stationary engines. In contrast to other
studies, which were carried under testbed conditions, we used a stationary turbocharged
heavy-duty diesel engine installed in a CHP unit, which is in practical operation. No hardware
modifications with respect to OME were applied to the serially produced CHP unit, but we
investigated different injection and EGR parameters in order to determine the effect on the
emission characteristics.
2. Methodology
The experiments were carried out on a CHP unit manufactured by Burkhardt Energy and
Building Technology and equipped with a MAN D26 heavy-duty diesel engine. This six-
cylinder turbocharged engine has a displacement of 12.4 liters and a common-rail injection
system with solenoid injectors. The engine was modified for dual-fuel operation with natural
gas and liquid fuel (diesel or fuel oil) by the CHP unit manufacturer. The engine is equipped
with a cooled high-pressure EGR system. No additional modifications for operation with
OME fuels were made. In the present paper, only liquid fuel operation was considered. For
detailed information about the CHP unit see Table 1.

Table 1: Specifications of the heavy-duty CHP unit

nominal electrical power

output 200 kW (diesel mode) / 240 kW (dual fuel mode)
nominal thermal power
output max. 240 kW
nominal fuel power input max. 571 kW
base engine MAN D 26
4-stroke diesel / dual-fuel, turbocharger with two stage
working process
intercooler
compression ratio 16:1
EGR high-pressure, cooled
displacement 12.4 liters
bore 126 mm
stroke 166 mm
rotational speed 1500 1/min
injection system common-rail direct injection with solenoid injectors
cooling water-cooled with exhaust gas heat exchanger
generator Leroy Somer LSA 47.2 M8
type Synchronous
cooling system air cooled
maximum power rating 550 kVA
voltage 400 V
frequency 50 Hz
cos φ 1

Figure 1 shows a schematic diagram of the CHP unit with the installed measurement
instruments. The engine was equipped with cylinder pressure sensors (imes CPS-01) in all
six cylinders. The indicating data was processed via an indicating system (AVL Indicom with
AVL Indicom software). The engine was equipped with a mappable Engine Control Unit
(ECU), which allowed to set injection timing, injection volume, etc.
 Figure 1: Schematic diagram of the test bed with the installed measurement instruments

The exhaust gas measurements were performed downstream of the turbocharger and
upstream of the Diesel Oxidation Catalyst (DOC). The gaseous exhaust gas components
were measured by means of a Fourier Transform Infrared (FTIR) spectroscope (Gasmet
CX4000 FTIR Gas Analyzer). The concentration of oxygen in the exhaust gas of the engine
was measured with a Paramagnetic Detector (PMD). To provide comprehensive data about
Particulate Matter (PM) and Particle Number (PN), two different measurement instruments
were operated. Firstly PM was analyzed based on a non-collecting electrostatic particle
measurement method with a Pegasor Mi2 particle sensor. This method determines electrical
charge carried by the particles in the exhaust gas and delivers a particle number as well as a
particle mass concentration in the exhaust gas based on a generic particle size distribution
and particle density. In addition to that PN and Particle Number Size Distribution (PNSD)
were determined with a Scanning Particle Mobility Sizer (SMPS) manufactured by TSI. In
Figure 2, a schematic diagram of the SMPS with all required components is depicted.
 Figure 2: Schematic diagram of the SMPS with all required components

After the aerosol inlet, particles with an aerodynamic diameter higher than a specified
threshold are separated from the stream by means of an impactor in order to prevent
damage to the instrument. To provide a known electrical charge distribution, a neutralizer
with a radioactive radiator is used Subsequent to the neutralizer, the aerosol is separated by
a Differential Mobility Analyzer (DMA) into different size fractions. The working principle is
similar to an electrostatic dust filter. According to their electrical mobility, the particles are
separated by applying different voltages. The hereby selected particles of the same size
fraction are channeled through a saturated butanol atmosphere in the next step. Through
condensation, the particle size is increased so that they can be counted by a Condensation
Particle Counter (CPC). Due to the required scanning time for the different size fractions, a
SMPS can only be used for measurements in stationary operation. [27]
For every stationary operating point of the CHP unit, at least five samples (135 seconds per
sample) were measured with the SMPS. For the data sets of the electrostatic particle
classifier as well as for those of the optical particle classifier the mean value graphs were
calculated. The gap between the mean graphs of the two different measurement methods
was fitted with equal weighting. As a result of this procedure, a steady Particle Number Size
Distribution (Composite Fit) was generated.
The electrical power production of the CHP unit was measured with a power analyzer (Hioki
3169-20), for the thermal power production a clamp-on ultrasonic flowmeter with
temperature inputs for heat flow measurements (Flexim FLUXUS F601) was applied. Fuel
consumption was determined by measuring the consumed fuel mass by means of scales.
For calculating the electrical efficiency of the CHP unit the following equation is applied.

(1)

Here, represents the electrical power output of the CHP unit, the consumed fuel
mass per time and Hi the lower heating value of the fuel.
The thermal efficiency is of a CHP unit is defined as the net heat power output divided by
the fuel power input:

(2)

The ratio between electric and thermal power output can be expressed by the power to heat
ratio:
(3)

For the experiments different OME/diesel blends based on defined volume concentrations
were prepared. The required masses of OME and diesel were calculated as follows:

(4)

(5)

With being the desired volume concentration of the alternative fuel in the blend in
percentage points, represents the density of the fuel and the total volume of the
mixture. The different mixtures were prepared based on these two equations by using
scales. For the experiments discussed in this paper, an OME mixture with chain lengths
from 2-6, a density of 1046 kg/m³ and a cetane number of 74.7 was used for blending. The
OME fuel was provided by ASG Analytik Gesellschaft. Detailed information about this fuel
can be found in the appendix. As base fuel for the blends, reference diesel fuel B0 is
applied. Compared to OME, B0 has a significantly lower density of 834.5 kg/m³ and lower
cetane number of 55.5. A detailed analysis can be found in the appendix. The investigated
blends are displayed in Table 2.

Table 2: Specifications of the investigated blends

φOME φDiesel ρmix Hi,mix c293K c373K

vol-% vol-% kg/m³ kWh/l J/kg/K J/kg/K
0 100 830.0 9.96 1968 2303
20 80 874.7 9.14
30 70 897.0 8.72
40 60 919.4 8.31
50 50 941.0 7.89
100 0 1053.4 5.82 1979 2152

In Table 1 the heat capacities at 293 K and 373 K for neat B0 diesel and neat OME are
given. While at low temperatures the differences in heat capacity are negligible, OME shows
a lower heat capacity at elevated temperatures, which implies that it absorbs less energy
when heated up and thus has less cooling effect when injected into the cylinder. However,
since OME has a lower heating value, more fuel has to be injected in order to introduce the
same amount of energy. Both effects compensate each other to some extent, but due to the
higher injected mass, OME blends will generally have a slightly higher cooling effect (heating
up and evaporation of the fuel). For the investigated diesel CHP unit, which runs with excess
air of up to two times the stoichiometric requirement, the temperature of the mixture will
however not be largely influenced by the differences in fuel injection.
For all the measurements, the CHP unit was operated in liquid fuel mode with a power
output of 200 kW el and a rotational speed of 1500 1/min. The thermal power output of the
investigated unit is approximately equal to the electric power output, i.e. the power to heat
ratio is near to one. The experiments were carried out under rated load since this is the
normal mode of operation for CHP units used for baseload supply. Flexible operation with
varying loads was not investigated in the present study.
The injection timing was modified in a range of 17 to 6 °CA btdc in order to identify the ideal
operating point for each fuel blend, while the EGR valve was set to alternatively fully open or
fully closed. The range of injection timings was restricted by the occurrence of knocking at
early timings and by deteriorating emission values for late injection timings. Every operating
point was kept constant over a period of 9 to 13 minutes. During this measurement time,
continuous measurements with FTIR and PMD were performed. Every 30 seconds, a value
was recorded by the emission measurement system. Mean values were calculated for every
operating point to provide sufficient representativeness. The measurement time was long
enough to enable at least five SMPS samples per operating point.

3. Results and Discussion

Power output and efficiency
Although the heating value of OME is lower than of diesel and thus the volumetric energy
content of the OME/diesel fuel blends is lower, no significant negative effects on the power
output of the engine were observed in the experiments. This is due to the fact that the tested
engine is operated with a reduced power rating of about 50% of the maximum power output,
as it is generally the case in CHP applications to ensure long term durability (as opposed to
emergency or standby power, where the engine operates on full load). Thus, the lower
heating value was compensated by a higher amount of fuel injected, since the hydraulic
limits of the injection system were not reached. Additionally the lower volumetric energy
content of OME/diesel blends might to some extent be compensated by the better
combustion characteristics (higher cetane number). In other applications than CHP, where
the engine is pushed to its limits, e.g. emergency power provision or mobile applications, the
peak power output might suffer when using OME. As the electrical efficiency is crucial for the
economic feasibility of CHP units, it should be extensively studied.
In Figure 3 the electrical efficiency of the investigated unit for the different injection timings
and different blends with and without the utilization of EGR is presented. Additionally the
exhaust gas temperatures (average of all six cylinders) are given. The thermal efficiency and
the power to heat ratio are displayed in Figure 4.
The results show that the OME fuel, despite having different combustion properties, does not
significantly influence the electrical efficiency of the CHP unit. Even when applying a 50/50
blend, the efficiency does not deteriorate compared to neat diesel. The relative deviations
with regard to neat B0 are between -0.86 % and +0.53 %, which is in the range of the
achievable measurement uncertainties under the given test conditions in practical operation.
The effects of the injection timing on the electrical efficiency of the CHP unit is more
pronounced. An early injection timing slightly increases the efficiency in the considered
range, while at the same time the average exhaust gas temperatures decrease.
The use of EGR hardly influences the electrical efficiency. However, it is interesting that the
average exhaust gas temperatures are higher with EGR, which is beneficial for the thermal
power output.
The test conditions for determination of the thermal efficiency could not be kept perfectly
constant during the experiments due to the heat sink, which is a building heating system in
practical operation. Therefore, the results presented in Figure 4 are not as accurate as those
for the electrical efficiency and the deviations between the single measurements should thus
not be over-interpreted. However, some general trends can be observed: the thermal
efficiency slightly decreases with early injection timings, which correlates well with the
simultaneous increase in electrical efficiency. The use of EGR results in higher exhaust
temperatures and thus an increased heat flux in the exhaust gas heat exchanger, which
increases the thermal power output. The power to heat ratios, which are in the range of 1.01
to 1.12, indicate that the CHP unit was optimized for electrical efficiency rather than thermal
efficiency. The measured peak total efficiency (electrical + thermal output) is in the range of
78.3% (without EGR) to 82.5% (with EGR).
As a general result from the efficiency measurements, it can be concluded that EGR is
beneficial for the total efficiency of the investigated diesel CHP without sacrificing electrical
efficiency.

B0 OME20 OME30 OME40 OME50

44.0 500
43.0
450

avg. exhaust gas temperature (°C)

42.0
electrical efficiency (%)

41.0 400

40.0
350
39.0
38.0
300

37.0 250
36.0
200
35.0
34.0 150
8 °CA btdc 8 °CA btdc 10 °CA btdc 10 °CA btdc 12 °CA btdc 12 °CA btdc
w/o EGR w/ EGR w/o EGR w/ EGR w/o EGR w/ EGR

Figure 3: Electrical efficiency (bars) and average exhaust gas temperatures (diamonds) of the
CHP unit for different injection timings and blends with and without the utilization of EGR
 B0 OME20 OME30 OME40 OME50
44.0 1.20
43.0 1.10
42.0 1.00

power to heat ratio (-)

thermal efficiency (%)

41.0 0.90
40.0 0.80
39.0 0.70
38.0 0.60
37.0 0.50
36.0 0.40
35.0 0.30
34.0 0.20
8 °CA btdc 8 °CA btdc 10 °CA btdc 10 °CA btdc 12 °CA btdc 12 °CA btdc
w/o EGR w/ EGR w/o EGR w/ EGR w/o EGR w/ EGR

Figure 4: Thermal efficiency (bars) and power to heat ratio (dots) of the CHP unit for different
injection timings and blends with and without the utilization of EGR

Exhaust gas emissions

Besides the consideration of the electrical efficiency, a comprehensive analysis of the
exhaust gas emissions was conducted. Figure 5 depicts the measured NOx and PM
emissions for all of the considered operating points. It can be clearly seen that with the use
of OME a significant reduction of particulate matter is achieved. Within the different data
series, various injection timings were analyzed both with and without EGR, resulting in
different NOx values. The highlighted data points in Figure 5 represents operation with neat
diesel B0 at an injection timing of 12 °Crank Angle before top dead center (°CA btdc). At this
injection timing, the CHP unit achieves a high electrical efficiency (up to 41.7%) while at the
same time NOx and PM values are still in an acceptable range. In the following analysis, the
injection timing at 12 °CA btdc will be picked out to illustrate the effect of OME admixture and
application of EGR on the emission characteristics.
 Figure 5: PM-NOx trade-off with different OME blends

Starting with reference fuel B0, NOx emissions are significantly reduced by approx. 560 ppm
when EGR is applied (see Figure 6). At the same time, the emission of PM rises
considerably from 0.24 mg/m³ up to 1.59 mg/m³, as expected (PM-NOx trade-off). By
blending the diesel with 20% OME, an appreciable reduction of particulate matter by about
1.0 mg/m³ can be observed. At the same time NOx emissions rise only slightly. With higher
OME contents within the blend, NOx remains almost constant, but further substantial PM
reduction down to values below those of neat B0 diesel without EGR is achieved.
 Figure 6: NOx and PM emissions for different blends with an injection timing of 12 °CA btdc

Figure 7 shows the relative change of the emission values of NO x, CO, PN and PM for the
various blends in relation to the values generated with the reference diesel for an ignition
timing at 12 °CA btdc and an open EGR valve. As already mentioned, NOx emissions
increase slightly (approx. 5%) by applying OME. The carbon monoxide emissions that
typically derive from locally rich regions in the combustion chamber can be significantly
reduced with the OME. Besides the aforementioned reduction of PM emissions, also PN
emissions (measured with the SMPS) decrease appreciably with increasing concentration of
OME in the blend. The observed reduction of PM emissions achieved by blending diesel fuel
with OME confirms the findings of other authors [24],[23],[21]. This effect may be the result
of an enhanced soot oxidation rate as described by Iannuzzi et al. [23]. In addition, OME, in
contrast to diesel, does not contain Polyaromatic Hydrocarbons (PAHs), which are known as
soot precursor substances. The absence of PAHs may thus also contribute to the observed
reduction of PM.
The reduction of particle emissions goes along with a slight increase of NO x emissions,
which was also observed by Feiling et al. [10] (neat OME1b in low-duty one cylinder
research engine) and Liu et al. [21] (10 and 20% OME3-4 in low duty four cylinder). This
indicates a higher peak combustion temperature, which is backed up by the results of the
combustion analysis presented later.
 Figure 7: Relative change of emission values achieved by blending with OME

Analysis of particle number size distribution

Apart from the integral values Particle Number and Particulate Matter, it is also important to take
a closer look on the particle size distribution. In the conducted experiments, comprehensive data
about the PNSD at different operating points was collected. Again, the results for an injection
timing of 12 °CA btdc with open EGR valve will be picked out to illustrate the general trends (c.f.
Figure 8). With increasing OME content, not only the absolute number of particles (area below
the graphs) decreases considerably, but also the PNSD moves to smaller particle diameters.
Generally the emitted particles are in the ultra-fine range, with a size distribution peaking well
below 100 nm.
These results agree well with those obtained by Pellegrini et al. [20], who observed the total
volume concentration of particles to decrease up to 33% and the PNSD to be unimodal. At high
load they measured a shift of PNSD to smaller particle sizes (OME3-5 blend with 7.5 vol% vs.
neat diesel in a light duty diesel engine). The smaller particle size observed by Pellegrini as well
as in our experiments could be explained by a larger fuel jet penetration, causing a better fuel
atomization, resulting in an increased number and smaller size of the fuel droplets, which
ultimately results in a smaller size of the particles [25].
Figure 8: Particle Number Size Distribution for the operating point with an injection timing at 12 °CA
btdc and an active EGR for various blends

In contrast to that Liu et al. [26] measured a bimodal PNSD when using different OME blends
(10, 20 & 30 vol% OME3-7) in a diesel engine, which could not be reproduced in our
experiments. The different shape of the PNSD could possibly be explained by a different sample
preparation of Liu et al. compared to our experiments. For instance, a longer distance and
therefore a higher temperature loss between exhaust manifold and sampling point can cause an
agglomeration of the particles, which changes the shape of the PNSD.

Combustion analysis
For all considered operating points a cylinder pressure analysis was performed. In order to
explain the general trends, the operating point with an injection timing at 12 °CA btdc was picked
out again. The following Figure 9 shows the results of the cylinder pressure analysis for an
injection timing of 12 °CA btdc when applying different blends.
As the diagram shows, the implementation of EGR causes a significant decrease of the cylinder
pressure (c.f. curve for 0% OME) due to a lower boost pressure. When OME is added to the
reference fuel, the peak pressure drops slightly. This reduces the strain on the combustion
engine. Additionally, the pressure curve is somewhat retarded with increasing OME content.
Figure 9: Cylinder Pressure, Heat Release Rate, Cylinder Temperature and Cumulative Heat
Release Rate for an injection timing of 12 °CA btdc with various blends

From the measured pressure traces, the Heat Release Rate can be derived to enable a
deeper understanding of the combustion process. Figure 9 depicts the Heat Release Rate for
the operating point at 12 °CA btdc for various blends.
From the trace of the graphs in Figure 9 it can be concluded that with increasing OME
content the maximum rate of heat release decreases. At the same time, the combustion
duration is slightly prolonged. This is also evident in the cumulative heat release rate. As
evident in Figure 3, the prolonged combustion duration has no negative effect on the CHP
unit’s electrical efficiency.
Based on the heat release rates, the corresponding cylinder temperatures could be derived
(see Figure 9). It can be concluded, that the maximum cylinder temperature rises with
increased OME content. The higher cylinder temperatures with increased OME content are a
by-product of the slightly retarded and prolonged combustion. These results back up the
slight increase in NOX-emissions with increasing OME content, as seen in Figure 7.
Regression Analysis
In order to validate the significance of the obtained results we conducted a statistical analysis
using Cornerstone, a software program for Design of Experiments (DoE). We used a
regression analysis in order to identify the effects of the different varied factors (injection
timing, EGR, OME content in the blend) on the target values (electrical efficiency, PM, NO x
emissions). The key results of this analysis are shown in Figure 10. Each of the nine subplots
depicts the relation between one influencing factor (x-axis) and one target value (y-axis) for
one certain operating point. The curves for different operating points vary slightly but the
general trends are similar. In this case the curves for a calculated optimum operating point,
which is described afterwards, are depicted. The upper and lower lines show the confidence
interval of the curves for each case:

 The NOx emissions are reduced by applying EGR (subplot 1) and increase with early
injection timings (subplot 2). The OME content on the contrary has only a small effect
on NOx-emissions (subplot 3).
 The particulate matter rises when applying EGR, as shown in subplot 4. The injection
timing has hardly an influence (subplot 5) on PM emissions, while the OME content
has a negative, quadratic effect on PM.
 The electrical efficiency is only slightly affected by the EGR rate (subplot 7). With
regard to injection timing an optimum operating point at 13 °CA btdc could be
identified (subplot 8). Admixture of OME had virtually no influence on the electrical
efficiency, thus it was eliminated from the regression model (subplot 9 left blank).
By regression analysis it is also possible to determine optimum parameter sets. In our study,
a good trade-off regarding particle and NO x emissions as well as a high electrical efficiency
and as low as possible OME content (in order to save costs) was identified with 20% OME
blend, an injection timing of 13 °CA btdc and open EGR (best settings identified by
regression analysis). This parameter set is depicted by the dashed lines in the subplots of
Figure 10. It has to be noted that other, maybe even more favorable optimum parameter
sets, could exist outside the investigated range. Further investigations will thus have to be
conducted in the future in order to validate the optimum settings.
 1 2 3

NOx

4 5 6
Particulate
Matter

7 8 9
Electrical
Efficiency

EGR Injection timing ( CA btdc) OME content

Figure 10: Predicted response of the regression model at the optimum point
4. Conclusion
The results of our study clearly show the potential of OME fuels for mitigating the PM-NOx
trade-off in diesel combustion. Due to the very low particle emissions achieved with OME
blends, the fouling normally observed with neat diesel could possibly be mitigated. Even
when using a blend with just 20% OME, PM decreased by more than 60%. Consequently,
the probability of fouling layer growth in the EGR cooler decreases considerably. In our
experiments, we observed a higher total efficiency of the CHP unit when applying EGR,
without sacrificing electrical efficiency. In combination with OME fuels, EGR could thus
become a feasible option for reducing NOx emissions in gensets and CHP units, without
sacrificing efficiency, durability and maintenance cost.
By applying higher EGR rates, further NOx reduction could be achieved, which would enable
a minimization of the required amount of aftertreatment. This would however require
modifications of the standard CHP engine design. Further experimental investigations should
also concentrate on finding optimized parameter sets of OME-blending, EGR, injection
timing and SCR aftertreatment, in order to meet the emission standards with the most cost-
effective solution taking into account engine efficiency, fuel cost and cost for aftertreatment.

Acknowledgements
The authors would like to thank the Bavarian Ministry of Science and the Arts as well as the
Bavarian Ministry of Economic Affairs, Regional Development and Energy for support of the
work carried out at the Center of Excellence for Cogeneration Technologies and the Institute
of Energy Technology at the Technical University of Applied Sciences Amberg-Weiden. We
would also like to thank the R&D team at Burkhardt Energy and Building Technology for
technical support of the engine experiments.
References
[1] Rajesh Kumar B, Saravanan S, Sethuramasamyraja B, Rana D. Screening oxygenates for
favorable NOx/smoke trade-off in a DI diesel engine using multi response optimization. Fuel
2017;199:670–83.
[2] Sindhu R, Amba Prasad Rao G, Madhu Murthy K. Effective reduction of NOx emissions
from diesel engine using split injections. Alexandria Engineering Journal 2018;57(3):1379–
92.
[3] Kumar JTS, Sharma TK, Murthy KM, Rao GAP. Effect of reformed EGR on the performance
and emissions of a diesel engine: A numerical study. Alexandria Engineering Journal
2018;57(2):517–25.
[4] Jothithirumal B, Jamesgunasekaran E. Combined Impact of Biodiesel and Exhaust Gas
Recirculation on NOx Emissions in Di Diesel Engines. Procedia Engineering 2012;38:1457–
66.
[5] Hong KS, Lee KS, Song S, Chun KM, Chung D, Min S. Parametric study on particle size
and SOF effects on EGR cooler fouling. Atmospheric Environment 2011;45(32):5677–83.
[6] Abd-Elhady MS, Malayeri MR. Asymptotic characteristics of particulate deposit formation in
exhaust gas recirculation (EGR) coolers. Applied Thermal Engineering 2013;60(1-2):96–
104.
[7] Abd-Elhady MS, Zornek T, Malayeri MR, Balestrino S, Szymkowicz PG, Müller-Steinhagen
H. Influence of gas velocity on particulate fouling of exhaust gas recirculation coolers.
International Journal of Heat and Mass Transfer 2011;54(4):838–46.
[8] Woo S, Park S, Shon J, Lee K. An experimental study on removal of exhaust gas
recirculation cooler fouling using washer fluid steam. International Journal of Heat and Mass
Transfer 2017;108:301–8.
[9] Omari A, Heuser B, Pischinger S. Potential of oxymethylenether-diesel blends for ultra-low
emission engines. Fuel 2017;209:232–7.
[10] Feiling A, Münz M, Beidl C. Potenzial des synthetischen Kraftstoffs OME1b für den
rußfreien Dieselmotor. ATZ Extra 2016;21(S11):16–21.
[11] Härtl M, Seidenspinner P, Jacob E, Wachtmeister G. Oxygenate screening on a heavy-duty
diesel engine and emission characteristics of highly oxygenated oxymethylene ether fuel.
Fuel 2015;153:328–35.
[12] Härtl M, Seidenspinner P, Wachtmeister G, Jacob E. Synthetischer Dieselkraftstoff OME1:
Lösungsansatz für den Zielkonflikt NO x-/Partikel-Emissionen. MTZ 2014;75(07-08):68–73.
[13] Pellegrini L, Marchionna M, Patrini R, Beatrice C, Del Giacomo N, Guido C. Combustion
Behaviour and Emission Performance of Neat and Blended Polyoxymethylene Dimethyl
Ethers in a Light-Duty Diesel Engine. In: SAE International400 Commonwealth Drive,
Warrendale, PA, United States; 2012.
[14] Härtl M, Gaukel K, Pélerin D, Wachtmeister G. Oxymethylenether als potenziell CO2 -
neutraler Kraftstoff für saubere Dieselmotoren: Teil 1: Motorenuntersuchungen. MTZ
2017;78(02):52–9.
[15] Burger J, Siegert M, Ströfer E, Hasse H. Poly(oxymethylene) dimethyl ethers as
components of tailored diesel fuel: Properties, synthesis and purification concepts. Fuel
2010;89(11):3315–9.
[16] Schmitz N, Burger J, Ströfer E, Hasse H. From methanol to the oxygenated diesel fuel
poly(oxymethylene) dimethyl ether: An assessment of the production costs. Fuel
2016;185:67–72.
[17] Burger J, Ströfer E, Hasse H. Production process for diesel fuel components
poly(oxymethylene) dimethyl ethers from methane-based products by hierarchical
optimization with varying model depth. Chemical Engineering Research and Design
2013;91(12):2648–62.
[18] ZHAO Q, WANG H, QIN Z-f, WU Z-w, WU J-b, FAN W-b et al. Synthesis of
polyoxymethylene dimethyl ethers from methanol and trioxymethylene with molecular
sieves as catalysts. Journal of Fuel Chemistry and Technology 2011;39(12):918–23.
[19] Lumpp B, Rothe D, Pastötter C, Lämmermann R, Jacob E. OXYMETHYLENETHER ALS
DIESELKRAFTSTOFFZUSÄTZE DER ZUKUNFT. MTZ Motortech Z 2011;72(3):198–203.
[20] Pellegrini L, Patrini R, Marchionna M. Effect of POMDME Blend on PAH Emissions and
Particulate Size Distribution from an In-Use Light-Duty Diesel Engine. In: SAE
International400 Commonwealth Drive, Warrendale, PA, United States; 2014.
[21] Liu H, Wang Z, Wang J, He X, Zheng Y, Tang Q et al. Performance, combustion and
emission characteristics of a diesel engine fueled with polyoxymethylene dimethyl ethers
(PODE3-4)/ diesel blends. Energy 2015;88:793–800.
[22] Li B, Li Y, Liu H, Liu F, Wang Z, Wang J. Combustion and emission characteristics of diesel
engine fueled with biodiesel/PODE blends. Applied Energy 2017;206:425–31.
[23] Iannuzzi SE, Barro C, Boulouchos K, Burger J. POMDME-diesel blends: Evaluation of
performance and exhaust emissions in a single cylinder heavy-duty diesel engine. Fuel
2017;203:57–67.
[24] Pélerin D, Gaukel K, Härtl M, Wachtmeister G. Recent results of the sootless Diesel fuel
oxymethylene ether. In: Liebl J, Beidl C, editors. Internationaler Motorenkongress 2017.
Wiesbaden: Springer Fachmedien Wiesbaden; 2017, p. 439–456.
[25] Tsolakis A. Effects on Particle Size Distribution from the Diesel Engine Operating on RME-
Biodiesel with EGR. Energy Fuels 2006;20(4):1418–24.
[26] Liu J, Sun P, Huang H, Meng J, Yao X. Experimental investigation on performance,
combustion and emission characteristics of a common-rail diesel engine fueled with
polyoxymethylene dimethyl ethers-diesel blends. Applied Energy 2017;202:527–36.
[27] Schröder T, Lenz V, von Sonntag, Justus, Ulbricht T, Heidecke P et al.
Messmethodensammlung Feinstaub: Methodenvorschlag zur Feinstauberfassung an
Feuerungsanlagen für biogene Brennstoffe. Schriftenreihe des BMU-Förderprogramms
"Energetische Biomassenutzung;8:69–71; 2014; Available from: https://www.energetische-
biomassenutzung.de/fileadmin/media/6_Publikationen/08_MMS_Feinstaub_web.pdf.
Highlights

-NOx trade-off