Content uploaded by Peter Koch
Author content
All content in this area was uploaded by Peter Koch on Jul 02, 2021
Content may be subject to copyright.
10. Dessauer Gasmotoren-Konferenz 2017
1
Recent CNG bus engine developments and comparison
of on-road emissions characteristics of a conventional
Euro VI and a hybrid city bus
Aktuelle CNG Busmotoren-Entwicklungen und ein Vergleich der
Abgasemissionen eines konventionellen Euro VI und eines hybriden
Busantriebskonzepts durch Messungen im realen Fahrbetrieb
Dipl.-Ing. Lars M. Nerheim*, Dr.-Ing. P. Koch, C.E. Harald Moen, B.Sc. Roger Aamot,
M.Sc. Ørjan Høyvik
Bergen University College (HiB), Bergen, Norway
Abstracts
In the beginning of this paper, the change in CNG bus engine operating principles from the
earlier lean burn concepts to the latest 1 without or with EGR is summarized. The
conclusion is that further improvements, particularly with an eye to reduced CO2-emisisons,
are now leading to hybrid CNG plus battery drive-train concepts.
Hence, a study has been carried out to investigate the environmental performance of such
a new CNG serial hybrid bus compared to a conventional Euro VI bus drive-train by means
of in-use emission measurements. Both vehicles were equipped with 6-cylinder MAN EURO
VI gas engines, the serial hybrid having a 206 kW CNG engine charging a 196 kW battery
pack and electric drive motors, and the standard gas bus a 228 kW engine driving through
an automatic transmission. The test program consisted of on-road measurements of NOx,
CO and THC emission measurements as well as characteristic engine and bus operation
parameters like load, AFR, speed and CO2 output for three different load scenarios. The
first finding is the much more stable operation at stoichiometric conditions of the serial hybrid
bus even in city traffic. This gives improved performance of the TWC with correspondingly
reduced emissions. In transient or stepwise load changes on the other hand, it can be
concluded that the overall thermal management and control strategy for the implementation
of the combustion engine has still some room for optimisation. Also, the battery power
contribution and hence any emission improvements, becomes less significant with
increasing passenger loading.
In diesem Beitrag wird zuerst der neulich erfolgte Übergang von Otto-mager Gasmotoren-
Konzepten zum 1 – Verfahren ohne oder mit EGR bei Euro VI Gasbusmotoren
zusammengefasst. Es wird gefolgert, dass weitere Verbesserungen insbesondere
bezüglich der CO2-Emissionen zu Gas-Hybrid Antriebs-Kombinationen führen werden.
Darüber hinaus wird über Vergleichsmessungen zwischen einem mit konventionellen Euro
VI CNG Antrieb und einem seriellen CNC-Hybrid angetriebenen Stadtbus berichtet.
10. Dessauer Gasmotoren-Konferenz 2017
2
Auffallend war in diesem zuerst der selbst im Stadtverkehr weitaus gleichmäßigere Betrieb
des Hybrid-Antriebes, was zur verbesserten Funktion des TWC- Katalysators führte und
damit zu deutlich reduzierten Abgasemissionen. Dagegen konnte die Betriebstemperatur
des TWC-Katalysators im Hybridbus nicht immer aufrechterhalten werden. Bis der
Katalysator wieder auf Betriebstemperatur kam wurden kurzzeitig erhöhte Emissionen
beobachtet. Daraus kann geschlossen werden, dass das «thermal Management» dieses
Hybrid-Triebwerkes und deren Kontrollstrategie noch verbesserungsbedürftig war. Auch
wurde festgestellt, dass mit steigender Passagieranzahl der Beitrag des Batterieantriebes
relativ abgenommen hat und sich somit die Verbesserungen im Vergleich zum
konventionellen Antrieb ausgeglichen haben.
Introduction
The introduction of the Otto lean burn combustion concept in the late 80s / early 90s put
spark-ignited (SI) gas-fueled engines back into the limelight across most industries and
markets. This engine type has since become very popular as industrial engines in the oil-
and gas industry and above all in cogeneration applications, and they still maintain their
market lead over diesel and heavy fuel engines in these markets (1).
Even though gas may not necessarily be the ideal transportation fuel, lean burn gas engines
rather quickly also managed to become established engines for city buses and similar
vehicle applications for fixed routes of limited range. The advantage in terms of low NOx
and minimal PM emissions combined with good efficiency, durability and high specific
power made the lean burn CNG engines very attractive with many OEMs, often developed
from their existing Diesel engines. There were of course individual design differences, which
indirectly said something about the efforts and commitment to the CNG-market by their
makers. Generally, most engines offered considerable improvements in terms of emission
characteristics and noise, which were the main drivers for their development.
Recent CNG bus engine developments
CNG bus engines have followed the lean burn strategy well up to Euro V emission
standards. Good transient response («drivability») while staying within the emission limits
was always the main challenge of such lean burn bus engine developments. Well- and less
well-developed engines have shown biggest differences in this area. Critical design features
have traditionally been the turbocharger arrangement, the gas admission, positioning of the
(necessary) throttle flap and the engine control software which should tie it all together as
fast and accurately as possible in order to follow the given test cycles.
Fig. 1a shows schematically a typical concept of a lean burn CNG bus engine for Euro V /
EEV emission standards.
10. Dessauer Gasmotoren-Konferenz 2017
3
Figure 1: Typical CNG bus engine concepts of recent years.
However, with the introduction of the very stringent Euro VI emission standards from 2014
on (see Table 2) the lean burn concept had reached its development limit in CNG bus
engines, as illustrated in Fig. 2 (as shown in ref. 2). Hence, a fundamental change followed
across the CNG bus engine industry, away from lean burn over to Lambda 1 concepts,
without or with EGR as in Fig. 1b. The way forward has been to fall back on the superior
emission reduction characteristics of the “three-way-catalyst” cleaning system, TWC, and
which also has the capability to more or less eliminate the dreaded «methane slip» from the
exhaust.
Hybrids
In certain countries like Norway, electrically driven vehicles and hybrids have also attracted
much attention lately as a further development step, because of their potentially lower fuel
consumption and hence even lower emissions than conventional Euro VI vehicles. Hybrids
hold particularly high potential for city buses with many transient load changes. Following
this philosophy, the Hordaland County Transport Authorities “Skyss” acquired two serial-
hybrid CNG-fueled city buses for evaluation on bus routes in the city of Bergen in 2015. The
principal arrangement of this serial-hybrid drive train is shown schematically in Fig 4, and
illustration and technical data are given in Fig. 3 and Table 1.
10. Dessauer Gasmotoren-Konferenz 2017
4
Figure 2: Technology road maps for low emission truck and bus engine developments (from
ref. 2). The vertical axes on the right side of the two diagrams show emission of NOx (left)
and PM (right) in g/kWh.
The measurement project
Skyss and HOG Energi approached Bergen University College about carrying out
comparative emission measurements between their newly acquired CNG Hybrid buses and
their latest conventional Euro VI CNG buses. For the comparison, measured in-use street
emissions and performance characteristics during operation along a given bus route in
Bergen should be the basis.
The CNG hybrid bus in question was a new type of serial hybrid 4-axle bus made by Van
Hool, equipped with a 6.9 l MAN E0836 CNG engine and solely electrical drive to the
wheels. The reference bus was a 3-axle MAN Lion’s City with a 12.8 l MAN E2876 CNG
engine with mechanical-hydraulic automatic transmission. Both engines were operating
according to the “ƛ 1 + TWC” principle.
Figure 3: Standard (left) and new serial hybrid gas bus (right).
10. Dessauer Gasmotoren-Konferenz 2017
5
Figure 4: Schematic arrangement of a serial-hybrid drive-train.
General bus data
Bus type:
MAN Lion’s City GL
VAN HOOL ExquiCity 24
Length
18.75 m
23,82 m
Width
2.5 m
2,55 m
Height
3.37 m
3,4 m
Dead weight
Load capacity
18313 kg
11687 kg
24665 kg
11835 kg
Number of passengers
148
(51 sitting, 97 standing)
151
(54 sitting, 97 standing)
Engine data
Type
MAN E2876
MAN E0836
Configuration
6 cylinder, in-line
6 cylinder, in-line
Displacement
12.8 l
6.9 l
Output
220 kW at 2000 rpm
206 kW at 2200 rpm
(max. 192 kWel)
Battery storage: 36 kWh
Torque
1250 Nm
(1000 – 1700 rpm)
1000 Nm
(1000 – 1900 rpm)
BMEP
10.7 bar at 220 kW
12.3 bar at 1250 Nm
16.4 bar at 206 kW
18.3 bar at 1000 Nm
Table 1: Overview of some bus data (9, 10).
Table 2 shows the Euro VI emission standards for heavy-duty diesel and gas engines for
transient testing and based upon the publically available data from MAN (3).
10. Dessauer Gasmotoren-Konferenz 2017
6
Table 2: Euro VI emission standards for HD diesel and gas engines - transient testing.
However, emission and fuel consumption data determined using standardized duty cycles
can differ quite considerably from actual real-world applications (4-6) Therefore, in-use
measurements can contribute not only to an improved accuracy of emission models like
e.g. EPA MOVES (7) but also indicate the problem areas of standardized test cycles and
procedures.
Experimental set-up
The main objectives of the test project was quantification and analysis of the environmental
performance of the two CNG bus types. Thus, the buses required modifications to get
access to parameters required to calculate target values, as information neither from the
ECU nor the OBD was possible. Unfortunately, no special equipment for on-road emission
monitoring was available (compare with ref. 6), so mobile laboratory-type analysers had to
be used. These were packaged and specially protected in order to avoid disturbances and
malfunctions during the test runs. Full sets of calibration gases, gas conditioning as well as
a separate power supply were carried along in the buses during measurements.
Measurement equipment
Important information like fuel consumption and exhaust mass flow were calculated from
indirect parameters by assuming stoichiometric combustion. Other logged data included the
speed and distance measured with an odometer, the position and altitude with a GPS device
as well as the ambient pressure, temperature and humidity. The emission measurements
were carried out in the raw exhaust with the exception of THC, which was routed through a
cooler unit. The entire set up was kept as close as possible to the guidance given by ISO
8178-1:2006. Table 3 shows a summary of the collected parameters and the measurement
equipment used.
Category
Parameter
Principle
Equipment
Emissions
NOx
CLD
Horiba PG-350
CO
NDIR
(CO₂)
NDIR
THC
HFID
J.U.M. HFID 3-200
Engine
data
(bef. TWC)
ZrO2 type
Horiba Mexa-110λ
Intake air mass
Pressure difference
Merian Z50MC2-4F
Table 3: Overview of the measurement equipment.
Test object information
Three different operating conditions (low, partial and high load) were tested three times
each for each bus type, altogether 18 measurement runs, whereby the passenger load
scenarios were simulated by the use of sandbags (~30 kg in average). An average
CO
in
g/kWh
NMHC
in
g/kWh
NOx
in
g/kWh
CH4
in
g/kWh
PM
in
g/kWh
PN
in
#/kWh
NH3
in ppm
EURO VI
limit
4
0.16
0.46
0.5
0.01
6·1011
10
10. Dessauer Gasmotoren-Konferenz 2017
7
passenger weight of 68 kg (8) was assumed for the calculation of the corresponding
passenger volume. More details of the individual tests are given in Table 4.
MAN Lion’s City GL
Van Hool EquiCity 24
Load
scenario
added
weight
# of
passengers
% of
capacity
added
weight
# of
passengers
% of
capacity
High
5525
+ 500
~89
~60
5835 + 580
~94
~63
Part
2652
+ 500
~46
~31
2963 + 580
~52
~35
Low
72
+ 580
~10
~7
0 + 500
~7
~5
Table 4: Overview of the different loading scenarios.
The test runs were carried out throughout the day in two testing campaigns of two and three
days for the standard and hybrid bus, respectively. The ambient temperature was in all
cases around 6°C and the relative humidity at roughly 90%, representing typical weather
conditions during this time of the year. The test buses were following a regularly scheduled
urban bus route (“bus line no. 2”) to simulate a typical operational pattern (even though
actual passengers boarding was not carried out). The length of one trip was ca 7.2 km and
had 21 official stops. One measurement-run covered a round-trip with start and stop being
at the bus service depot at the end of run from the city centre up to the turning point (see
also Fig. 5).
The average speed during the test runs with the standard gas bus was between 16.8 –
17.8 km/h, approx. 1 km/h higher than the average for the hybrid bus trips (15.8 –
16.4 km/h). This may have to do with the lower power-to-weight ratio of the hybrid bus.
The test fuel used for all measurements was natural gas consisting of ~95% methane, 3%
ethane and the rest being higher hydrocarbons (C3 to C5). Stoichiometric air requirement
was calculated to Lmin = 16.7 kgair/kggas, and additional characteristics for the fuel gas was
MN ~ 85 and a LHV ~ 49 MJ/kg.
An exemplary depiction of the driving cycle (in this case the test runs 7, 8 & 9 from the partly
loaded standard gas bus) as well as the elevation profile of the test route is given in Fig. 5.
The latter has been taken from the official documentation of the bus line no. 2 and shows
the start and stop as well as the direction of each test run marked with the green flag. The
test runs are considerably longer than the typical driving cycles used in emission legislation.
Hence, a broader spectrum of speeds and driving situations, which is typical for urban public
transport operation, could be investigated and analysed.
10. Dessauer Gasmotoren-Konferenz 2017
8
Figure 5: Top: Exemplary speed profile of some of the test runs;
Bottom: Altitude profile (AMSL) of the route.
Practical approach and experimental challenges
Some impressions from the experimental set-up are presented in Fig. 6.
The raw exhaust emission measurements were split into the dry measurement of NOx, CO,
CO2 and O2 via a heated sample line and gas conditioner unit (Madur PGD-100) and the
THC sampling which consisted of a filter, an unheated sampling line and a water quench
before being routed to the analyser unit (J.U.M. HFID 3-200). It should be noted that due to
vibration issues with the FID burner, continuous measurement of the THC emissions was
not always possible.
10. Dessauer Gasmotoren-Konferenz 2017
9
Figure 6: Intake air mass flow w DP-cell (1), Interior with sandbags (2),
Lambda measurement before and after TWC (3), Emission sampling after TWC (4).
Fig. 7 shows the set-up of the analyser units in the bus. Some improvements to this had to
be made during the tests because mainly of vibration issues. Less problems occurred when
measuring in the hybrid bus, as its ride was somewhat smoother especially with respect to
the acceleration behaviour.
Figure 7: Analyser arrangements in the bus.
Numerical data corrections
Although the test equipment used was of good laboratory type and quality, it was a
combined set-up not intended for on-road measurements with its many transient engine
load changes, accelerations, and vibrations influences from traffic. This resulted in some
operational challenges, which needed to be handled in real-time as well as inaccuracies,
which first appeared during the following data analysis. Various “data cleaning” procedures
were tried out, and finally the same procedure was used for all data sets. It was finally
concluded that the relative accuracy between all test runs was quite good. A typical
“correcting” measure undertaken was, for example, that of the time lag during transients of
the slow signals (like NOx, UHC etc.) which were “time-corrected” afterwards based on the
immediate signals (like throttle position, airflow, etc.). In addition, any negative emission
concentration values were omitted in the results.
10. Dessauer Gasmotoren-Konferenz 2017
10
Test results
Emission characteristics
In Fig. 8, a comparison of the NOx (left) and CO (right) emissions between the standard S
(dotted), and the serial hybrid H (full line) is presented vs. number of passengers. The
graphs represent trend curves through discrete values of the average figures from three
test runs.
The significantly lower NOx-values of the Hybrid over the Standard Euro VI bus become
very apparent, particularly at low passenger loading. Both engines use the same type and
size of Three-Way-Cat (TWC), giving larger specific catalyst volume for the smaller hybrid
bus engine. Provided this is up at operating temperature, it is very effective because the
gas residence time is longer, resulting in improved reaction conditions for the conversion of
NOx, CO and THC.
Further, at low passenger loading the battery power contribution makes up a significant part
of the total power requirement of the bus. This consequently keeps the gas engine load
rather low. The combined effect of these two factors is thought to be the reason for this
large difference in emissions at low passenger loading.
Much the same effect from the larger specific catalyst volume can also be seen in the lower
CO emissions of the Hybrid drive unit, Fig. 8 (right).
Figure 8: Average NOx and CO emissions vs passenger loading for the two bus types.
Fig. 9 gives a comparison between the measured CO2 and THC emissions from the two
buses, again as function of passenger loading, and being the average of three test runs. In
the CO2-diagram, the 2020 EU target values for light commercial- resp. passenger cars are
also plotted for comparison reasons. There is virtually no difference in the CO2-emissions
from the two bus types at higher loadings, because the battery power then becomes less
significant in light of the higher total mass of the hybrid bus. The measured THC-values
were generally very low and estimated to consist of ~95% CH4.
10. Dessauer Gasmotoren-Konferenz 2017
11
Figure 9: CO2 and THC emissions for the two bus types vs passenger loading.
Operational characteristics
An interesting comparison of the real-time emissions from both buses during actual driving
is shown in the next two figures below. In Fig. 10 the curves at the left show NOx emissions
and driven speed for the serial hybrid, while the ones at the right represents the standard
gas bus; both measured over a given distance of in-city driving. Hence, warm operating
conditions for both the engines and the TWCs can be assumed.
Even if the actual driving conditions are not fully equal, there was a significant difference
between the two drive-trains, with the serial hybrid showing an almost ideal steady operation
in spite of all the transient speed changes. This of course both gives lower engine-out
emissions in the first place and a more steady gas flow through the TWC. This makes a
very big difference in the actual bus emissions. Fig. 11 shows the same comparison for the
CO2-emissions.
Hybrid
Standard
Figure 10: Momentary NOx and vehicle speed recordings in city traffic conditions.
However, even if the serial hybrid drive-train has shown very stable low-emission operation
in busy city traffic, its rather large TWC and the engine running at low load at times, f. i.
during down-hill conditions (regenerative braking), has made the catalyst cool down at
times. Subsequent sudden load increases have then produced rather high emissions for a
10. Dessauer Gasmotoren-Konferenz 2017
12
while until the TWC had heated up again (see Fig. 12 below). This was found to be a
considerable drawback with this engine-catalyst combination.
This smaller and quite steadily running engine could probably have done better with a
somewhat smaller TWC, which also would have responded quicker after low-load periods.
This effect was also registered in the larger conventional drive train, but it was much less
pronounced.
Hybrid
Standard
Figure 11: Momentary CO2 and speed recordings in city traffic conditions.
Figure 12: Catalyst heat-up: Transient start-up emissions after an engine cool-down
(Hybrid drivetrain)
Summary
CNG bus engines have changed operating concept to stoichiometric with three-way
catalytic converters, with or without EGR to satisfy the stringent Euro VI emission limits.
This concept is claimed to give the best compromise between good drivability and low
emissions. Further improvements might be achieved by going to gas-electric hybrid drive-
train combinations.
This measurement project has proven the point that considerable lower on-road exhaust
emissions than what the current strict Euro VI regulations require are achievable by going
over to a CNG-battery hybrid drive-train. The major advantage lies in the considerable
smoother operating condition of the CNG engine, which both gives more even engine-out
10. Dessauer Gasmotoren-Konferenz 2017
13
emissions as well as improved efficiency of the catalytic converter. The thermal
management of the tested CNG engine + TWC was found not to be optimal, but no attempts
were made to modify this or the strategy of the hybrid bus controller, which was kept as
given by the bus manufacturer.
The project has further shown that the biggest improvements naturally come under
load/speed conditions where the battery power contributes the most to total power
requirements, i.e. under light load conditions and moderate speed (in-city conditions).
This also goes for the possible efficiency gains (= reduced CO2 emissions). Under higher
load and speed (commuter haul) conditions, the battery contribution is relatively minor and
still its weight must be carried along so the advantages of the hybrid powertrain diminish. In
addition, a serial-hybrid without a plug-in possibility and its considerable electrical losses
might not represent the optimum configuration here.
Acknowledgements
The authors want to thank their colleagues at the Bergen University College / IMM for their
very good assistance in carrying out this rather comprehensive project and for preparing
this paper.
Further, they also want to thank Skyss and HOG Energi for making the CNG buses available
and supporting them in the project.
References
1 L. M. Nerheim:
“Modern Gas engines; recent development status”
Plenary speech, 7th Dessau Gas Engine Conference, 2011
2 F. Posada:
“CNG Bus emissions roadmap, from Euro III to Euro VI”
ICCT Report, 2009
3 R. Staimer, MAN:
“Comparing drive-train technologies”
NGVA Brussels, 09.07.2014
4 J. May et al.:
“A comparison of light-duty vehicle emissions over different test cycles and in
real-driving conditions”
FISITA 2014, Maastricht, NL.
5 M. Ferguson:
“Mind the gap! Why official car fuel economy figures don’t match up to reality”
Transport & Environment, Brussels, 2013
6 Ricardo plc:
“Route cause”.
In: Ricardo Quarterly Review no 3, 2014.
7 EPA OTAQ Modelling & Inventories
http://www.epa.gov.otaq/models.htm
10. Dessauer Gasmotoren-Konferenz 2017
14
8 Directive 97/27/EC:
“Masses and dimensions of certain categories of motor vehicles and their trailers
and Amendment of Directive 70/156/EEC”. 1997.
9 Vehicle registration document, MAN Lion’s City GL
10 Vehicle registration document, Van Hool ExquiCity 24