ARTICLE IN PRESS

I N T E R N AT I O N A L J O U R N A L O F H Y D R O G E N E N E R G Y 33 (2008) 3225 – 3229

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/he

Experimental tests of blends of hydrogen and natural gas

in light-duty vehicles

Fernando Ortenzia,, Maria Chiesab, Riccardo Scarcellic, Giovanni Peded

a
Centre for Transport and Logistics, University ‘‘La Sapienza’’ of Rome, Rome, Italy
b
Environmental Physics Department, Catholic University of the Sacred Heart, Brescia, Italy
c
Dipartimento di Ingegneria Meccanica, Universitá di Roma Tor Vergata, Italy
d
ENEA, Italian National Agency for New Technologies and Environment, Rome, Italy

art i cle info ab st rac t

Article history: While the use of hydrogen in fuel cell vehicles will be the ultimate target for sustainable
Received 15 October 2007 mobility, much attention has to be devoted to a suitable ‘‘bridge technology’’, to accelerate
Accepted 22 January 2008 the process of the introduction of these new technologies and in the meantime to build an
Available online 21 March 2008 infrastructure suitable for the utilisation of hydrogen. In the short-medium term, the
atmospheric emissions could be significantly reduced by using different percentages of
Keywords:
hydrogen mixed with natural gas in internal combustion engines.
Hydrogen and natural gas blends
The paper is based on the results of an experimental test campaign carried on in ENEA
HCNG
labs, aimed at identifying the prospective of the use of blends of natural gas and hydrogen
Engine tuning
(HCNG) in existing ICE vehicles. The tested vehicle is an IVECO Daily CNG, originally fuelled
Experimental test
with natural gas and the tests had been made on the ECE15 driving cycle comparing the
emission levels of the original configuration (CNG) with the results obtained with different
blends (percentage of hydrogen in the fuel) and control strategies (stoichiometric or lean
burn).
& 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights
reserved.

1. Introduction Nevertheless, a widely diffused use of hydrogen must foresee

the existence of infrastructures not yet available; a possible
In order to reduce CO2 emissions different technological ‘‘bridge technology’’ could be represented by blends of hydrogen
solutions can be taken over: among others, an increase of the and natural gas that does not need dedicated infrastructures
global efficiency of the energetic cycles or a reduction of the (natural gas from the grid via pipeline).
carbon content in the fuels. From the beginning of the industrial
era onwards, the trends related to the consumption of different 1.1. State of the art
fuels had always underlined that the market moves towards the
cleaner ones, with decreasing C/H ratios. If we think about CH4, Experiences of application of blends of hydrogen and natural
the fossil fuel with the lowest C/H ratio, the next step could be gas in internal combustion engines (ICEs) started in 1991,1 in
represented by zero carbon fuels such as hydrogen. In this the framework of a research programme financed by DOE and
sense, the use of pure hydrogen could represent a ‘‘cultural NREL, the ‘‘Denver Hythane Project’’ [1], with good results,
shift’’ towards a great abatement of local pollutants and CO2. shown in Table 1.

Corresponding author.
E-mail address: fernando.ortenzi@uniroma1.it (F. Ortenzi).
1
In the same years, the University of Pisa and ENEA too carried on some activities.
0360-3199/$ - see front matter & 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2008.01.050
 ARTICLE IN PRESS
3226 I N T E R N AT I O N A L J O U R N A L O F H Y D R O G E N E N E R G Y 33 (2008) 3225 – 3229

Table 1 – Denver Hythane Project hydrogen content. Therefore, a natural gas engine, when
fuelled with HCNG, shows a lower power output, while
NMHC CO NOx maintaining its better efficiency.
(g/mile) (g/mile) (g/mile) To restore a good value of power output, especially for lean
burn mixtures (for lambda ¼ 1:4 the engine loses 50% of its
Gasoline 0.59 14.1 2.2
power), a good solution could be represented by a turbo-
ULEV 0.04 1.7 0.2
charged engine with an higher charging pressure.
Natural gas 0.01 2.96 0.9
Hythane 0.01 0.7 0.2 CO2 emissions had been reduced not only as a result of the
substitution of CNG by hydrogen but thanks to the special
properties of hydrogen as a combustion stimulant that can
produce a leverage factor greater than 1. An obvious benefit of
the leverage effect is that a CO2 reduction is possible even if
During the last 15 years many experiences had been the hydrogen used is produced by natural gas without any
conducted all over the world. All the experiments had mainly ‘‘sequestration’’ of CO2.
examined the reduction of the emissions of NOx with respect
to different air/fuel ratios (l values) and percentages of
hydrogen by volume. All the experiments had shown that 3. ENEA experimental tests
the blends of hydrogen and natural gas reduce the exhaust
emissions of both regulated pollutants and CO2 and increase The light-duty commercial vehicle under test was an Euro III
the efficiency of a spark ignition engine. (with three-way catalytic converter) Daily 2.8 CNG manufac-
During the last years, a number of fleet tests had been carried tured by IVECO, Fig. 1 (belonging to the fleet of ASM SpA of
on. The recent Hythanes (24.8% vol. Hydrogen, Frank Lynch, Brescia), and it had been mainly modified in the control
Hydrogen Components Inc., HCI) [2], bus demonstration project system (ECU) for the tests. The cycle adopted for the
at Sunline transit in California used a 7% hydrogen by energy characterisation had been the urban part of NEDC (ECE 15)
formula and the NOx emissions were reduced by 50. At a and the value for the vehicle mass had been set to 3500 kg,
European level, the most significative example of application of higher than the set value for the homologation.
blends of hydrogen and natural gas in ICEs is given by the tests Therefore, the results are not directly comparable with OEM
still ongoing in Malmo (Sweden) on urban buses [3]. data, but surely are more significative considering the guide
The experimental results are available for blends with a cycle of the ASM SpA vehicles that could use these blends in
hydrogen content of 8% by volume for their use on real driving the future (for example, the waste collecting vehicles).
cycles, while the data relative to a blend of 25% by volume are In order to modify the engine maps in real time during
available just for the engine tests. Since the first target related engine-tuning phase, an EPROM emulator (MET16) had been
to the use of blends is the reduction of atmospheric emissions used. The final flashing of the EPROM had been obtained
of urban pollutants, the tests tried to optimise the combustion thanks to the EMP21, an EPROM programmer.
to obtain the maximum reduction of total hydrocarbons (HC), The main objective of the engine-tuning phase had been
nitric oxides (NOx ) and carbon monoxide (CO). the reduction of the CO2 production without increasing the
emissions. Two different blends, characterised by hydrogen
percentages of 10% and 15%, had been tested: the HCNG10
2. Methane–hydrogen blends and the HCNG15.
The main parameters investigated were lambda, spark
When used in an ICE, even the addition of a small amount of advance angles and enrichment during transients.
hydrogen to natural gas (5–30% by volume, that means The NOx emissions had been considered as exhaust
1:5210% by energy) leads to many advantages, because of parameter which had not to be overcome in case of
some particular physical and chemical properties of the two stoichiometric set-up. Actually, hydrogen addition implies a
fuels [4]. higher laminar combustion speed and this causes an increase
Methane has a slow flame speed while hydrogen has a of combustion temperature and therefore higher NOx emis-
flame speed about eight times higher; therefore, when the sions. On the contrary, CO and HC emissions are always
equivalence ratio (lambda) is much higher than for the lower, thanks both to the lower quantity of carbon and to the
stoichiometric condition, the combustion of methane is not improved combustion process.
as stable as with HCNG. For lean burn mixtures, also HC monitoring had been a
As a consequence of the addition of hydrogen to natural gas decisive parameter that had been taken into consideration.
an overall better combustion had been verified, even in a wide Actually, also HC emissions can raise due to laminar
range of operating conditions (lambda, compression ratio, combustion speed decrease, with no complete oxidation of
etc.), finding the following main benefits: HC. Moreover, the fuel oxidation is delayed too by higher gas
cooling during expansion.
 a higher efficiency,
 lower CO2 production and emissions. 3.1. Measurement system

Because of hydrogen chemical and physical properties [5], Two sets of instruments had been used: the first one for the
HCNG has a lower LHV per Nm3 than NG, depending on the measurement of the in-cylinder pressure and the other one
 ARTICLE IN PRESS
I N T E R N AT I O N A L J O U R N A L O F H Y D R O G E N E N E R G Y 33 (2008) 3225 – 3229 3227

Fig. 1 – APICOM roller bench, the emissions analyzer and the IVECO Daily tested.

for measurement of CO, CO2, HC, NOx emissions and fuel Notwithstanding the above reported spark ignition time
consumption. correction, engine performances with methane–hydrogen
blends remained not acceptable during ECE driving cycle in
3.1.1. Cylinder pressure terms of emissions, due to the too high NOx emission values,
A single cylinder head had been equipped with a piezoelectric compared to pure methane.
pressure transducer. Signals had been processed by an A more detailed examination of the engine behaviour
amplifier while the angular position had been measured with during transients shows that fuel enrichment values (as
an inductive crank-angle calculator module for on-line mapped in ECU) had been too low, therefore actual l reaches
indicating measurements. Preliminary testing with pure values comprised between 1.1 and 1.2. As a result, NOx
methane (1500 rpm at low loads) was carried out for the emissions increased too much. For this reason, a map
engine model validation. correction had been adopted concerning the acceleration
phases.
3.1.2. Emissions
Emissions had been measured using a HORIBA OBS-1300 3.3. Lean mixtures
integrated system. A flowmeter (pitot type) mounted on the
sampling probe permits to calculate the exhaust flow rate and For a lean burn blend, research of better l (we wanted to reach
exhaust pressure and temperature. a value of 100 ppm, the same of pure methane) had been
limited from l ¼ 1 to l ¼ 1:45. Furthermore, the increase of l
3.1.3. Fuel consumption values had caused very important power losses, as shown in
Pure methane of certified composition and certified mixtures Fig. 3.
had been used for the tests. To assure the requested precision, Therefore, l ¼ 1:45 had been the maximum value ini-
the cylinders had been weighed before and after drive tests tially fixed (this value substantially reduces NOx ) and a series
after a consistent time ( about 20 ECE15 cycles). of tests had been produced changing the spark ignition
advance to optimise the control strategy in order to increase
3.2. Stoichiometric mixture the performances. Unfortunately, NOx had grown in an
exponential way, while power gain had not been significative.
Without any modification of the injection control map, the Therefore, it had been decided that it is more convenient to
NOx emissions increase as a consequence of the increasing adopt a lower l value without changing the spark advance
combustion speed. instead of setting a high l together with optimal advance
A spark advance reduction of only 3 degrees (which means timing.
a little retard compared to the case of pure methane) brings to Values (for l) greater than 1.45 could be interesting
a large decrease of NOx emissions, without torque reduction to analyse for their better efficiency and consumption,
(Fig. 2). but the engine will need substantial modifications such as a
 ARTICLE IN PRESS
3228 I N T E R N AT I O N A L J O U R N A L O F H Y D R O G E N E N E R G Y 33 (2008) 3225 – 3229

900 650

800 NOx 640

Torque
700 630

620
600

Torque Nm
610
NOx ppm
500
600
400
590
300
580
200 570
100 560

0 550
-6 -4 -2 0 2 4
Advance variation

Fig. 2 – NOx and torque outputs for k ¼ 1 at different ignition advance at 1500 rpm (15% H2 ).

600
2000
550
Nox
Torque 500
1500

Torque Nm
NOX ppm

450

1000
400

350
500
300

0 250
1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 1.4 1.45
Lambda

Fig. 3 – NOx and torque in function of k for lean mixtures at 1500 rpm (15% H2 ).

different compression ratio, the addition of a turbo- Furthermore, HC emissions increase for the lower combus-
charger, etc. tion quality due to lean mixture.
The advantages in terms of consumption are reported in
Fig. 4: moving towards leaner mixtures and higher percen-
4. Results tages of hydrogen give better results.
In order to take into account the well-to-tank consumption,
For the stoichiometric and using the optimised maps, related to the hydrogen production by fossil fuels (we do not
the emissions levels are even lower than the original CNG consider the more favourable case of hydrogen production by
ones, especially for NOx, while for CO and HC there are renewables), let us consider the minimal (2.16%) hydrogen
improvements caused by a better combustion quality (for HC) content, in mass, for a HCNG15 mixture.
and a less carbon presence in the fuel (for CO). However, We know, from the ‘‘Well-to-Wheel analysis of future
with the HCNG10 the vehicle presents the lowest emission automotive fuels and power-trains in the European context’’,
levels. EUCAR/JRC/CONCAWE, 2006, that the energy consumption
For lean burn mixtures, the emissions present two different related to hydrogen production from natural gas (EU mix) is
behaviours: the CO emissions are always lower for blends about 16% (efficiency: 84%).
with 10% and 15% of H2 by volume. Nevertheless, concerning Therefore, HCNG15 production is a mere 0.5% less efficient,
the NOx emissions, the values are lower than those ones from an energetic point of view, than natural gas. If the
using pure CNG, but not as good as stoichiometric values. average engine efficiency is about 20% in urban use, this value
 ARTICLE IN PRESS
I N T E R N AT I O N A L J O U R N A L O F H Y D R O G E N E N E R G Y 33 (2008) 3225 – 3229 3229

12

10.05
10

8
% 7.19

6 5.43

2
1.10

0
HCNG10 λ=1 HCNG10 λ=1.4 HCNG15 λ=1 HCNG15 λ=1.4

Fig. 4 – Tank-to-wheel analysis for energy saving.

implies a 2.5% efficiency loss from well-to-wheel, that is more the reduction of fuel consumption, pollutants and CO2, even
that counterbalanced by the energy saving values measured with hydrogen produced by fossil fuels!
during tests (from 5% with a stoichiometric blend up to 10% Future experimental developments would foresee the
for lean mixtures). optimisation of the emissions of both urban pollutants and
CO2 along with a reduction of fuel consumption, investigating
the following much wider set of engine variables:
5. Conclusions
 a wider spectrum of l values;
Optimum condition for enhancement of the pollutants and  spark advance;
greenhouse gas emissions can be found using different  percentage of H2 in the blends;
approaches, i.e. with the use of lean blends and with the  compression ratio;
use of stoichiometric blends.  intake charging pressure.
Actually, these two approaches have been adopted by
VOLVO (whose engines are mounted on the urban buses of
the ‘‘Malmo Hythane project’’) and by IVECO. R E F E R E N C E S
In our case, dealing with an IVECO engine, designed in
order to work in stoichiometric conditions, the adoption of a
‘‘lean combustion’’ strategy had brought us to unsatisfactory [1] Linch F. Hythane: a bridge to an ultraclean renewable
results: actually, since no change in the engine hardware had hydrogen energy system. In: Atti del Workshop IEA in
been made, as for example the compression ratio enhance- Denver; 1991.
[2] Larsen JF, Wallace JS. Comparison of emissions and efficiency
ment and/or the engine turbocharging, with respect to the
of a turbocharged lean-burn natural gas and hythane-fuelled
vehicle’s behaviour, a reduction of the engine specific power engine. ASME J Eng gas Turbines Power 1997:119.
had been verified, along with the reduction of the power due [3] TunestÅl P, Christensen M, Einewall P, Andersson T, Johansson
to the inferior energy content in volume (11% for the blend B, Jönsson O. Hydrogen addition for improved lean burn
with a 15% hydrogen content by volume). capability of slow and fast burning natural gas combustion
Therefore, starting from the existent motorisations, the chambers. SAE 2002-01-2686;2002.
[4] Karner D, Francfort J. Freedom car and vehicle technologies
strategies that have to be adopted for the modifications
program—advanced vehicle testing activity. Arizona Public
should be coherent with the base constructor philosophy.
Service, Alternative fuel (hydrogen) pilot plant. US DOE; 2003.
Nevertheless, for both the combustion strategies, our [5] Swain M.R, Yusuf M, Dulger Z, Swain MN. The effects of
vehicle succeeded in doing the urban cycle and the engine hydrogen addition on natural gas engine operation. SAE Paper
worked regularly obtaining encouraging results in terms of 932775; 2003.