international journal of hydrogen energy 35 (2010) 1420–1429

Available at www.sciencedirect.com

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

Conversion of a commercial spark ignition engine to run

on hydrogen: Performance comparison using hydrogen
and gasoline

C. Sopena a, P.M. Diéguez a, D. Sáinz a, J.C. Urroz a, E. Guelbenzu b, L.M. Gandı́a a,*
a
Escuela Técnica Superior de Ingenieros Industriales y de Telecomunicación, Universidad Pública de Navarra, Campus de Arrosadı́a,
E-31006 Pamplona, Spain
b
Acciona Biocombustibles S.A., Avenida Ciudad de la Innovación n 5, E-31621 Sarriguren, Navarra, Spain

article info abstract

Article history: The modifications performed to convert the spark ignition gasoline-fueled internal
Received 9 September 2009 combustion engine of a Volkswagen Polo 1.4 to run with hydrogen are described. The car is
Received in revised form representative of small vehicles widely used for both city and interurban traffic. Main
12 November 2009 changes included the inlet manifold, gas injectors, oil radiator and the electronic
Accepted 12 November 2009 management unit. Injection and ignition advance timing maps were developed for lean
Available online 14 December 2009 mixtures with values of the air to hydrogen equivalence ratio (l) between 1.6 and 3. The
established engine control parameters allowed the safe operation of the hydrogen-fueled
Keywords: engine (H2ICE) free of knock, backfire and pre-ignition as well with reasonably low NOx
Hydrogen-fueled internal emissions. The H2ICE reached best brake torque of 63 Nm at 3800 rpm and maximum brake
combustion engine power of 32 kW at 5000 rpm. In general, the brake thermal efficiency of the H2ICE is greater
Spark ignition engine than that of gasoline-fueled engine except for the H2ICE working at very lean conditions
Hydrogen fuel (l ¼ 2.5) and high speeds (above 4000 rpm). A significant effect of the spark advance on the
Injection timing NOx emissions has been found, specially for relatively rich mixtures (l < 2). Small changes
Ignition timing of spark advance with respect to the optimum value for maximum brake torque give rise to
an increase of pollutant emissions. It has been estimated that the hydrogen-fueled
Volkswagen Polo could reach a maximum speed of 140 km/h with the adapted engine.
Moreover, there is enough reserve of power for the vehicle moving on typical urban routes
and routes with slopes up to 10%.
ª 2009 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.

1. Introduction on Climate Change, during the period 1990–2004, global CO2

emissions increased by 27%, from 20,463 to 26,079 million
The transport sector is a crucial element of the current ener- tons, and the energy demand from the transport sector
getic and environmental policies because of the problems increased by 37% [1]. In this context, it is clear that stronger
associated to its almost complete dependence on petroleum legal frameworks are required, as it is the case of the
and very relevant contribution to greenhouse gases (GHG) Renewable Energy Directive recently put in place by the
emissions. Despite the measures taken to comply with the European Parliament [2]. According to this Directive each
Kyoto Protocol to the United Nations Framework Convention Member State shall ensure that the energy from renewable

* Corresponding author. Tel.: þ34 948 169 605; fax: þ34 948 169 606.
E-mail addresses: caso@unavarra.es (C. Sopena), pmde@unavarra.es (P.M. Diéguez), eguelbenzu@acciona.es (E. Guelbenzu),
lgandia@unavarra.es (L.M. Gandı́a).
0360-3199/$ – see front matter ª 2009 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2009.11.090
 international journal of hydrogen energy 35 (2010) 1420–1429 1421

Nomenclature PFI port fuel injection

MBP maximum brake power
Acronyms MBT maximum brake torque
BDC bottom dead center NOx nitrogen oxides
BTDC before top dead center SI spark ignition
DAS data acquisition system TWC three-way catalyst
ECU electronic control unit WOT wide open throttle
GHG greenhouse gases
Symbols
HC hydrocarbons
bmep brake mean effective pressure, kPa
H2FCEV hydrogen-fueled fuel cell electric vehicle
bsfceq brake specific fuel consumption of gasoline
H2ICE hydrogen-fueled internal combustion engine
equivalent, g/kWh
ICE internal combustion engine
LEL lower explosive limit Greek
PEMFC polymer electrolyte membrane fuel cell l air to fuel equivalence ratio (lambda)

sources in all forms of transport in 2020 is at least 10% of the drawbacks. Interesting comparisons between them can be
final consumption of energy in transport in that Member found in the above-mentioned reviews [4,5] and more specific
State. information on each technology in the papers by Lee et al. [6],
Hydrogen is considered as a clean way of powering vehi- Vandenborre and Sierens [7], Guo et al. [8], Sierens et al. [9,10],
cles for the future since neither carbon-based pollutants nor Verhelst et al. [11–13], White et al. [14], Mohammadi et al. [15],
GHG would be emitted when hydrogen produced from Kahraman et al. [16], Szwaja and Grab-Rogalinski [17], Gomes
renewable sources is used. The European Parliament has Antunes et al. [18] and Thomas [19].
established requirements for the type-approval of motor Most car manufacturers are focusing their research and
vehicles using hydrogen as fuel and for the type-approval, development efforts on the H2FCEVs. But the cost of polymer
installation and safety of hydrogen components and systems electrolyte membrane fuel cells (PEMFCs) is currently
in hydrogen-powered vehicles [3]. This regulation establishes prohibitively high and they require hydrogen of extremely
safety requirements in a ‘‘technology-neutral basis’’ since it is high purity (above 99.99%) thus making the fuel also more
taken into account that manufacturers might follow different expensive. Significant technical advances as improved
approaches to the development of these vehicles (e.g. using PEMFCs reliability and durability and price reduction by
fuel cells and electric drive motors, internal combustion a factor of 20–50 are necessary for PEMFCs to compete with
engines, hybrid configurations, different hydrogen onboard internal combustion engines (ICEs) [4]. On the other hand,
storage methods, etc.). Of course, care should be taken that some important vehicle makers like BMW, Ford, Mazda and
hydrogen is produced cost-effectively from renewable energy MAN have developed and still continue developing cars and
sources so that the overall environmental balance of intro- buses powered by H2ICEs with very good performances
ducing hydrogen as a fuel is positive. Muradov and Veziroğlu [20–25]. Particularly relevant are the achievements made by
[4] and Winter [5] have recently updated in their excellent the BMW group with the mono-fuel vehicles of the Hydrogen 7
reviews critical aspects of hydrogen energy. Muradov and series. The spark ignition (SI) engines of 6.0 l consist of 12
Veziroğlu showed that neither of the three options (decar- cylinders in V achieving a maximum brake torque (MBT) of
bonization of fossil energy, nuclear energy, and renewables) 390 Nm at 4300 rpm and maximum brake power (MBP) of
currently considered could alone provide the carbon-free 192 kW at 5100 rpm. Emissions levels are well below the Super
power required by mid-century. These authors have proposed Ultra Low Emissions Vehicles standard for nitrogen oxides
a strategy for the transition to hydrogen economy in which (NOx) and the hydrogen consumption for the highway cycle is
hydrogen produced from water using carbon-free energy only 2.1 kg per 100 km [25].
sources (nuclear, solar, wind, geothermal) would play a key Main argument in favor of H2ICEs in the short term may be
role [4]. According to Winter the cost of hydrogen from that they are based on a very well-known technology with
renewables is by the moment prohibitive since renewable more than one hundred twenty years of experience so
energies are not yet fully developed to unsubsidized market implementing H2ICEs in vehicles may be easy and fast. It will
levels. However, renewables need hydrogen as a means of also allow taking advantage of manufacturing infrastructure
energy storage in order to guarantee their contribution to the now in operation for conventional ICEs. Moreover, current
world energy trade instead of being limited to local or regional costs are considerable lower than that of PEMFCs and H2ICEs
applications [5]. can run on relatively cheap hydrogen of industrial quality.
There is no doubt that the share of hydrogen-powered Perhaps H2ICE might be a transitional technology contributing
vehicles in the total fleet will increase in the near future. to a more rapid introduction of hydrogen in the transport
However, it is not yet clear whether the hydrogen-fueled sector while H2FCVs and hybrid configurations continue
internal combustion engines (H2ICEs) or the hydrogen- developing [5,14].
powered fuel cell electric vehicles (H2FCEVs) will be the pre- In this work the steps followed to convert a commercial SI
vailing solution since both options have advantages and ICE to run on hydrogen are described. An engine from
1422 international journal of hydrogen energy 35 (2010) 1420–1429

5000 rpm, respectively. Main engine characteristics are

Table 1 – Specifications of the original engine.
summarized in Table 1. All the engine transformations and
Manufacturer Volkswagen the complete test program of the resulting H2ICE were carried
Base vehicle Polo 1.4 out at the Laboratory of Internal Combustion Engines of the
Type Spark ignition Public University of Navarra. To this end the test cell was
Fuel Gasoline 95 NO adapted to work with hydrogen by means of suitable
Cylinder line/head material Aluminium hydrogen supply and safety systems.
Number of cylinders 4 in line
Main motivation of this work was to show the technical
Bore/Stroke (mm) 76.5/75.6
viability of the engine conversion in order to obtain a H2ICE
Swept volume (cm3) 1390
Compression ratio 10.5:1 with performances well suited for using it in a Volkswagen
Valve train configuration DOHC Polo demonstration car taking care that the engine operates
Number of valves per cylinder 4 safely in all conditions and with acceptable level of NOx
Fuel injection system Port-injection emissions.
Ignition system Single spark ignition coil As it is well-known, the unique combustion characteristics
Maximum brake power 59 kW at 5000 rpm
of hydrogen (very low ignition energies, wide flammability
Maximum brake torque 132 Nm at 3800 rpm
Load control Drive by wire
limits of hydrogen–air mixtures and very high flame propa-
gation speed) are very advantageous for SI engines at low
loads when safe lean burn operation at wide open throttle
(WOT) is possible [11–14]. However, the same characteristics
Volkswagen (Polo 1.4) was selected for this study. The features
give rise to some limitations at high engine loads when the
of the experimental facilities, safety measures and the
mixture becomes richer, due to the risk of pre-ignition,
comparison of the engine performances when operated on
backfire, knock and increased NOx emissions. With the so-
gasoline and hydrogen are presented as well. This work has
called advanced hydrogen engines [14] it is intended to over-
been carried out in the framework of a R&D contract granted
come these limitations by implementing direct hydrogen
by Acciona Biocombustibles S.A. (a branch of Acciona Ener-
injection into the cylinders [15] or supercharging combined
gı́a), a company whose activities are focused in the renew-
with exhaust gas recirculation [13] solutions. In our case,
ables sector [26]. Previous works by our group dealt with
whereas the original motor block is preserved, the main
renewable hydrogen production from water electrolysis and
changes have been made in the fuel feeding system using
wind energy [27–30].
naturally aspirated gaseous port fuel injection (PFI) and in the
electronic management system. Some minor changes as in
the exhaust system have been made as well. These modifi-
2. Engine specifications and modifications cations are described in more detail in the following
performed subsections.

The original engine was a four-cylinder naturally aspirated SI 2.1. Hydrogen feeding system
motor mounted in the Volkswagen Polo 1.4. The motor was
kindly donated by Volkswagen Navarra S.A. When run on The inlet manifold in plastic of the original engine was
gasoline MBT and MBP were 132 Nm at 3800 rpm and 59 kW at replaced by a cast manifold to prevent breakdown in case of
backfire (explosion of the air-fuel mixture in the inlet mani-
fold [11]). The gasoline injectors were substituted by hydrogen
injectors manufactured by Quantum Technologies; therefore,
it was necessary to adapt the holes where the injectors are
placed in the intake manifold. In order to prevent hydrogen
leakages due to engine vibration the injectors were firmly
fixed to the manifold by means of a support. As shown in Fig. 1
a gas accumulator was manufactured and connected to the
injectors through stainless steel tubes in order to maintain
constant the pressure at the injectors inlet. Hydrogen is fed to
the accumulator at low-pressure from the supply line through
flexible tubing to compensate for engine vibrations. The
accumulator has another entry connected to a venting line to
purge the complete circuit with nitrogen gas in case of
emergency or prolonged shutdown.

2.2. Electronic management system

Fig. 1 – Photograph of the hydrogen feeding system of the The electronic control unit (ECU) of the original engine was
H2ICE: 1) hydrogen injectors; 2) support; 3) low-pressure replaced by a programmable MoteC M 400 unit. Most of the
hydrogen accumulator; 4) hydrogen inlet; 5) purging sensors and actuators of the original engine were retained;
nitrogen inlet. however, the original lambda sensor was removed and
 international journal of hydrogen energy 35 (2010) 1420–1429 1423

Fig. 2 – Scheme of the experimental set-up: 1 & 2) hydrogen and nitrogen high-pressure regulators; 3 & 4) solenoid gas
shutoff valves; 5) low-pressure regulator; 6) safety valve; 7) manual purge valve; 8) solenoid purge valve; 9) hydrogen mass-
flow meter; 10) non-return valve; 11) air mass-flow meter; 12) air filter; 13) engine hydrogen sensor; 14) test bed cell
hydrogen sensor; 15) MoteC M 400 ECU; 16) working pressure regulator.

replaced by a wideband lambda sensor (Bosch LSU 4.9) and operated with lean hydrogen–air mixtures so, as will be
a new oil temperature sensor was mounted. It should be noted shown, the NOx formation is low due to the relatively low
that the original lambda sensor was a binary one that only combustion temperatures. It should be noted that the TWC
allowed mixture control close to stoichiometric conditions operates efficiently when stoichiometric or slightly rich
(l ¼ 1). As the H2ICE is fed with lean mixtures (l values mixtures are used due to the presence in the exhaust gases of
between 1.5 and 3), a new wideband lambda sensor is required reducing chemical species (e.g. unburned hydrogen).
for proper operation. Sensors and actuators were connected to
the MoteC unit and were calibrated either directly with the
MoteC or in some cases obtaining calibration functions 3. Test facilities
externally and then introducing them in the MoteC. On the
basis of the test program carried out in the test bed, injection The testing bed cell of the laboratory was adapted to work
and ignition timing maps were obtained as explained in with hydrogen with the greatest possible safety. A data
Section 4. acquisition system (DAS) was mounted that allowed on-line
collecting and storage of most of the engine sensors and
2.3. Other modifications actuators signals as well as that of the auxiliary equipment
installed (hydrogen and air mass-flow meters, in-cylinder
A water–oil heat exchanger was installed close to the oil filter pressure transducer, etc.). A scheme of the experimental
to maintain as constant as possible the oil temperature. The set-up is shown in Fig. 2.
crankcase venting pipe was replaced by another of larger
diameter to reduce the possibility that hydrogen accumulated 3.1. Test bed cell
in the carter and thus reducing the risk of flammable mixtures
formation in the crankcase. The exhaust was modified by The test bed cell consisted of an eddy current dynamometer
removing the three-way catalyst (TWC) because the H2ICE was AVL 80 capable of absorbing up to 80 kW of power, with a BME
1424 international journal of hydrogen energy 35 (2010) 1420–1429

indicated the concentration of hydrogen in the room. The

8
safety system was designed in such a way that supplying
7 hydrogen is not possible if the sensors are not in operation. If
the second alarm level is reached, the controller will auto-
6
matically stop the hydrogen supply and the line would be
5 purged with nitrogen. The operator can also make manually
ms 4 a nitrogen purge, if necessary. The test cell ventilation system
is equipped with an extractor for the engine exhaust as well as
3
a blower allowing the complete air renewal of the cell in less
2 than two minutes. All the extracted gases are sent out of the
1 building. Similar safety measures have been adopted in
related works [21,25].
0
1400
2200 100
3000 60 3.2. Data acquisition system
3800 40 % Load
rev/min 4400 20
4800
0 The data acquisition system (DAS) consisted of a National
Instruments Ni-CompacDAQ with three modules measuring
Fig. 3 – Representative injection map showing the duration
up to eight channels. It allowed collecting, showing real-time
of the injection pulse (ms) as a function of the engine load
and storing in a computer with a LabView program signals
(%) and speed (rpm) for l [ 2.
from the main engine sensors and actuators, as well as
additional signals like brake torque, engine speed, hydrogen
and air flow-rates, knock sensor, NOx and O2 in the exhaust,
300 control system allowing performing tests using as control etc. The engine was also equipped with a Kistler measuring
parameters the torque, the accelerator position and the spark plug with integrated cylinder pressure sensor and
engine speed. The engine was mounted on the test bench a Kistler crank angle encoder.
which included a Bosch analyzer to determine CO, CO2,
hydrocarbons (HC) and O2 and a Horiba 730 analyzer to
determine NOx in the exhaust gases.
The gas supply system consisted of two lines, one for 4. Engine management parameters
hydrogen of industrial quality and the other for nitrogen.
Nitrogen is used to purge the hydrogen line prior to start-up The engine control was done by means of a MoteC M 400
when the H2ICE is not used for long time or in the even of an electronic control unit in which injection and ignition timing
emergency; such an incident has never occurred on the maps were introduced. MoteC receives signals from engine
modified engine during the test period. Both lines are fed by sensors and gives the output signals to the actuators, essen-
gas cylinders of 50 l at 200 bar (Air Liquide). Control of the tially, width injection pulse and ignition timing. Management
hydrogen feeding system is governed by a programmable included also idle control, water and air temperature correc-
controller performing the functions of hydrogen supply, tions, maximum speed engine control and air to fuel equiva-
nitrogen purge and emergency stop. lence ratio (l) control. To this end, the narrow original lambda
The safety equipment consisted of two hydrogen sensors sensor was replaced by a wide lambda sensor Bosch LSU 4.9
TQ122 (electrochemical type), capable of detecting hydrogen that was calibrated to be used with hydrogen fuel.
concentrations well below the lower explosive level (LEL) of Injection and ignition timing maps introduced in the ECU
hydrogen/air mixtures. One is located just above the engine were obtained from the results of tests conducted in the
feeding system, and the other on the test bed cell ceiling close engine test bed. Engine speeds varying between idling and
to the hydrogen supply line. These sensors are connected to 5000 rpm and loads between unloaded and fully loaded were
a control unit TQ4000 with two adjustable alarm levels, which considered.
Spark advance

30
20
10
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
rev/min

Spark advance for lambda 1.6 and WOT spark advance for lambda 2 and WOT

spark advance for lambda 2.5 and WOT Spark advance for lambda 1.6 and 50% throttled
Spark advance for lambda 2 and 50% throttled Spark advance for lambda 2.5 and 50% throttled

Fig. 4 – Ignition advance in degrees BTDC as a function of the engine speed (rpm) for l values of 1.6, 2 and 2.5 at WOT and
50% throttled engine.
 international journal of hydrogen energy 35 (2010) 1420–1429 1425

70
60
50
16
40

kW
14 30
20
12 10
Spark advance

0
10 0 1000 2000 3000 4000 5000
rev/min
8 Brake power for gasoline fuelled engine
Brake power for H2ICE for lambda 1.6
6
4800 Fig. 7 – Brake power (kW) at WOT versus engine speed
4
3800 (rpm) for the gasoline-fueled engine and the H2ICE with
2 l [ 1.6.
2200 rev/min
0
100 60 1000
40 20 0 inlet air thus reducing the possibility of backfire. By choosing
Load %
this point, it was achieved for all loads and engine speeds that
Fig. 5 – Representative ignition map showing the spark the injection finished before the inlet valve closed and in most
advance in degrees BTDC as a function of the engine load cases even before the piston reached bottom dead center
(%) and speed (rpm) for l [ 2. (BDC) during the intake stroke. A representative injection map
can be seen in Fig. 3 where the duration of the injection pulse
is given as a function of the engine speed and load.

4.1. Injection timing

4.2. Ignition timing
Injection timing has a strong influence on the engine perfor-
mance at low loads and speeds; this parameter can be The ignition advance maps were developed using the ignition
adjusted also to avoid backfire [11]. The strategy followed for timing allowing maximum brake torque (MBT) with the lowest
drawing up injection maps was working with l constant in the NOx emissions. A very conservative approach was adopted,
1.6–3 range at all loads varying the torque through the throttle. namely, retarding the ignition advance to values which are far
Values of l lower than 3 have been selected in this work in from producing knock (autoignition of the mixture ahead of
order to avoid incomplete hydrogen combustion, especially at the flame front originated from the spark [14]).
high loads and low engine speeds. On the other hand, when In contrast with the case of gasoline-fueled engines [31],
using air to fuel equivalence ratios richer than 1.6 there is the we have found that for the H2ICE the optimal spark advance
possibility of knock in some cases at full load and high engine for MBT depended essentially on l except for low loads and
speeds. speeds. This can be clearly seen in Fig. 4 where the ignition
The start of hydrogen injection was chosen at the same advance in crank angle degrees before top dead center (BTDC)
point of the cycle in all cases. That point corresponded to the is given as a function of the engine speed for l values of 1.6, 2
moment when the exhaust valve closed in order to ensure and 2.5 at WOT and 50% throttled engine. It is also shown that
that there is no leakage of hydrogen in the exhaust manifold there is almost no difference between throttled or unthrottled
while the exhaust valve has been somewhat cooled with the operation (low load influence) at engine speeds above about
1800 rpm and l greater than 2. The combustion is faster with

1200 325

1000 300
Bsfc (g/kWh)
bmep (kPa)

800 275

600 250

400 225

200 200
1000 2000 3000 4000 5000
0
rev/min
1000 2000 3000 4000 5000
rev/min Bsfc with gasoline Bsfceq with lambda 1.6
bmep with gasoline bmep with lambda 1.6 Bsfceq with lambda 2 Bsfceq with lambda 2.5
bmep with lambda 2 bmep with lambda 2.5
Fig. 8 – Brake specific fuel consumtion of gasoline
Fig. 6 – Brake mean effective pressure (kPa) at WOT versus equivalent (g/kWh) at WOT versus engine speed (rpm) for l
engine speed (rpm) for l values of 1.6, 2 and 2.5. Values of values of 1.6, 2 and 2.5. Values of fuel consumption for the
bmep for the gasoline-fueled engine are also included. gasoline-fueled engine are also included.
1426 international journal of hydrogen energy 35 (2010) 1420–1429

600 Fig. 8 with the brake specific consumption of gasoline. The

results show that the H2ICE has better brake thermal effi-
500
ciencies than the gasoline-fueled engine, especially at low and
400 medium speeds. This could be due to the fact that hydrogen
NOx (ppm)

combustion is faster and closer to a constant volume process,

300 and then with a more efficient thermodynamic cycle than that
of gasoline combustion. This is a remarkable result taking into
200
account that l values used with the H2ICE are higher than for
100 the gasoline-fueled engine which runs on stoichiometric
(l ¼ 1) or slightly rich mixtures. In the case of the H2ICE, it can
0
1000 2000 3000 4000 5000 be seen that the lowest bsfceq corresponds to the richest
rev/min mixture considered (l ¼ 1.6) which is then the most brake
thermal efficient. The leanest mixture (l ¼ 2.5) is the lowest
NOx for lambda 1.6 NOx for lambda 2 NOx for lambda 2.5
efficient except for very low engine speeds. This could be
Fig. 9 – NOx emissions (ppm) at WOT versus engine speed attributed to a lower efficiency of very lean mixtures
(rpm) for l values of 1.6, 2 and 2.5. combustion because of the higher proportion of unburned
hydrogen and slower combustion for a given engine speed
[6,9,22].
rich mixtures (lower values of l) thus requiring less spark
advance. A representative ignition map can be seen in Fig. 5, λ = 1.6
where the ignition timing is given in crank angle degrees
450
BTDC as a function of the engine speed and load. A
400
350
NOx (ppm)

300
5. Results: engine performance
250
200
The performance of the H2ICE has been evaluated in terms of
150
the brake mean effective pressure, brake power, brake specific
100
fuel consumption, and pollutants emissions. These parame-
50
ters were compared with those provided by the original
0
gasoline-fueled engine. It has been also estimated the
performance that could achieve a Volkswagen Polo powered 1800
2600
by the H2ICE. 3400
4200
rev/min 5000
5.1. Brake mean effective pressure, brake power and
brake specific fuel consumption Ignition advance for MBT-5º Ignition advance for MBT
Ignition advance for MBT+5º
Brake mean effective pressure (bmep) of the H2ICE as a func-
tion of the engine speed for l values of 1.6, 2 and 2.5 at WOT λ = 1.6
are compared in Fig. 6 with the bmep of the gasoline-fueled B
engine. The bmep obtained with the H2ICE are within the 104
% bmep respect bmep for

values that could be expected. This is because for l values of

102
1.6, 2 and 2.5, theoretical bmep to obtain the same brake
thermal efficiency are 50, 42 and 35%, respectively, of the
100
MBT

gasoline-fueled engine bmep because less air enters the

cylinders due to the low density of hydrogen [25]. Maximum 98
bmep of the H2ICE is achieved at the same engine speed than
with the gasoline engine. This indicates that the engine speed 96
at which the H2ICE reaches the maximum volumetric effi-
ciency is similar to the value when it is run on gasoline. Of 94

course, lower bmep are obtained when leaner (higher l) 1800

2600
hydrogen–air mixtures are fed. 3400
4200
As concerns the brake power, the results are shown in 5000
rev/min
Fig. 7. The maximum H2ICE brake power is limited because the
maximum engine speed has been limited as well (5000 rpm). Ignition advance for MBT-5º Ignition advance for MBT
The H2ICE reached a MBP of 32 kW at WOT with l ¼ 1.6. When
Ignition advance for MBT+5
the engine was fueled with gasoline MBP was 59 kW at
5000 rpm. Fig. 10 – Effect of varying 58 the ignition advance for MBT on
Brake specific fuel consumption of gasoline equivalent the NOx emissions (A) and brake mean effective pressure
(bsfceq) for l values of 1.6, 2 and 2.5 at WOT are compared in (bmep) change with respect to MBT bmep (B) at l [ 1.6.
 international journal of hydrogen energy 35 (2010) 1420–1429 1427

λ = 1.87 50–75 ppm, are produced. For l ¼ 1.6 the NOx emissions varied
A between about 350 and 550 ppm depending on the engine
400
speed. Although significant, these values are well below the
350 1000–2500 ppm range, typical of the gasoline-fueled engines
300 [31]. Nevertheless, it should be noted that maximum NOx
NOx (ppm)

250 emissions of hydrogen-fueled engines of the order of

200 10000 ppm usually take place with l ¼ 1.3. This is mainly why
150 in spite of the lower power output, lean burn operation of
100 H2ICEs is recommended [6,14,22].
We have also found that NOx emissions were very sensitive
50
to the spark advance, specially for relatively rich mixtures
0
(l < 2). Small changes of spark advance with respect to the
1800
2400 optimum value for MBT give rise to an increase of nitrogen
3000 oxides emissions. This is illustrated in Figs. 10, 11 and 12,
3600
4200 where the effect of varying 5 the ignition advance for MBT on
rev/min 4800
the NOx emissions and brake mean effective pressure (bmep)
are shown for l values of 1.6, 1.87 and 2, respectively. On
Ignition advance MBT-5º Ignition advance for MBT
Ignition advance MBT+5º

λ=2
λ = 1.87 A
B
80
102
70
%bmep respect bmep for MBT

60
NOx (ppm) 50
100 40
30
20
98 10
0
1800
96 2600
3400
1800 4200
2400 rev/min 5000
3000 3600
4200
4800
rev/min Ignition advance MBT-5 Ignition advance for MBT
Ignition advance MBT+5
Ignition advance MBT-5º Ignition advance for MBT
Ignition advance MBT+5º
λ=2
Fig. 11 – Effect of the variation of the ignition advance for
B
104
MBT (MBTL58, MBTD58) on the NOx emissions (A) and
% bmep respect bmep for

brake mean effective pressure (bmep) change with respect 102

to MBT bmep (B) at l [ 1.87.
100
MBT

98
5.2. Pollutant emissions
96
It has been found that CO and unburned hydrocarbons (HC)
emissions due to lubricating oil combustion are extremely 94

low. The most important pollutants in the exhaust were 1800

2600
nitrogen oxides (NOx). Unburned hydrogen emissions were 3400
4200
negligible for l values lower than 3. 5000
rev/min
NOx emissions for l values of 1.6, 2 and 2.5 at WOT are
compared in Fig. 9. It should be noted that the TWC of the Ignition advance MBT-5º Ignition advance for MBT
original engine was removed and NOx emissions for the Ignition advance MBT+5º
gasoline-fueled engine are not available. As expected, NOx
emissions increase when the mixture becomes richer because Fig. 12 – Effect of the variation of the ignition advance for
the maximum combustion temperature increases as l MBT (MBTL58, MBTD58) on the NOx emissions (A) and
decreases. It has been found that for air to fuel equivalence brake mean effective pressure (bmep) change with respect
ratios higher than 1.8, low NOx emissions of the order of to MBT bmep (B) at l [ 2.
1428 international journal of hydrogen energy 35 (2010) 1420–1429

40 mechanical modifications included the inlet manifold, low-

35 pressure hydrogen accumulator, gas injectors and oil radiator.
30 The cost of the new components is similar to that of the items
25 replaced. The original electronic control unit was replaced by
kW

20 a programmable ECU MoteC M 400 whose cost is more than

15 twice that of the original ECU. However, it was chosen due to
10
its great flexibility to perform the test program that allowed
5
obtaining the injection and ignition timing maps of the
0
30 50 70 90 110 130 150 modified engine. In the event of using the H2ICE in
km/h a commercial vehicle the ECU would be similar to that of the
Brake power (kW) required original engine.
Brake power (kW) delivered by hidrogen engine with lambda 1.6 Injection timing maps have been developed for air to
hydrogen equivalence ratios between 1.6 and 3, from unloa-
Fig. 13 – Brake power (kW) required by the Volkswagen
ded engine to full load and engine speeds between idling and
Polo and brake power (kW) delivered by the H2ICE (l [ 1.6)
5000 rpm. Spark ignition advance maps have been obtained
versus vehicle speed (km/h).
choosing advance values getting MBT compatible with low
NOx emissions. It has been found that if l < 2 the main factor
influencing the ignition advance is the mixture richness.
changing the spark advance þ5 it can be seen in Fig. 10 A that The performance of the H2ICE has been evaluated in
the NOx emissions at l ¼ 1.6 almost double whereas the bmep a suitably adapted test bed cell and compared with the
only increases 1–2% (Fig. 10B). On the other hand, a change of performance of the gasoline-fueled engine. The results are
5 reduces the nitrogen oxides emissions but the brake mean within the expected values: the H2ICE is capable of providing
effective pressure decreases markedly, specially at high a brake torque of 63 Nm at 3800 rpm and MBP of 32 kW at
engine speeds, above 3000 rpm (Fig. 10B). As the mixture 5000 rpm. The brake thermal efficiency of the H2ICE is greater
becomes leaner, the influence of the spark advance on the than that of gasoline-fueled engine except for l > 1.8. The
NOx emissions decreases. This is clearly seen in Fig. 12 A brake thermal efficiency decreases as l decreases from 3 to
where at l ¼ 2 the nitrogen oxides emissions change with the 1.6. The H2ICE performance is suitable to power the Volks-
spark advance only at engine speeds above 3400 rpm; anyway, wagen Polo 1.4. The vehicle would be capable of reaching
in this case, the NOx concentrationa are low, below 70 ppm. In a maximum speed of 140 km/h in horizontal and straight line
contrast, the negative effect of changing 5 the advance for trajectory. Moreover, there is enough reserve of power for the
MBT on the brake mean effective pressure is maintained vehicle moving on an urban route and roads with slopes up
under leaner conditions, as can be seen in Figs. 11B and 12B for to 10%.
l ¼ 1.87 and 2, respectively. Further work will be focused in supercharging the H2ICE
with a mechanical compressor to increase the brake torque
and power output. It is also being considered to introduce the
5.3. Expected performance of the hydrogen-powered
control of the different engine loads in a similar way as in
vehicle
compression-ignition engines, that is, working at WOT and
changing the air to fuel equivalence ratio. The pumping losses
The brake power required to move the Volkswagen Polo car in
will be decreased with this strategy and it is expected that the
horizontal and straight line trajectory with constant speed is
H2ICE brake thermal efficiency will increase.
compared in Fig. 13 with the brake power delivered by the
H2ICE working with l ¼ 1.6 and the gearbox ratio 1.0 [20,32]. It
can be seen that the hydrogen-fueled engine gives power
enough for the vehicle reaches a maximum speed of 140 km/h Acknowledgements
in the above-mentioned conditions. This speed is suitable for
driving safely a vehicle of these characteristics. Moreover, We gratefully acknowledge Acciona Biocombustibles S.A. for
there is enough reserve of power for the vehicle moving its financial support under R&D contract to the Public
smoothly on an urban route because the speed limit in these University of Navarra OTRI 2006 13 118 (CENIT project:
routes is usually established at 50 km/h. On the other hand, SPHERA) and Volkswagen Navarra S.A. for the Volkswagen
for roads with slopes of 5 and 10% and using the original Polo 1.4 engine donation. LMG and PMD also acknowledge
gearbox, which was not designed for the H2ICE, the car would financial support by Ministry of Science and Innovation of the
be able to achieve maximum speeds of 105 km/h (third speed Spanish Government (MAT2006-12386-C05-04).
transmission) and 70 km/h (second speed transmission),
respectively.
references

6. Conclusions and future work [1] European Environment Agency. Climate for a transport
change. TERM 2007: indicators tracking transport and
The conversion of the SI gasoline-fueled ICE of a Volkswagen environment in the European Union. Report No 1/2008.
Polo 1.4 car to run on hydrogen has been carried out. Main Copenhagen: EEA; 2008.
 international journal of hydrogen energy 35 (2010) 1420–1429 1429

[2] Directive 2009/28/EC of the European Parliament and of the [17] Szwaja S, Grab-Rogalinski K. Hydrogen combustion in
Council of 23 April 2009 on the promotion of the use of a compression ignition diesel engine. Int J Hydrogen Energy
energy from renewable sources. Official Journal of the 2009;34:4413–21.
European Union 5.6.2009:L140/167–62. [18] Gomes Antunes JM, Mikalsen R, Roskilly AP. An experimental
[3] Regulation (EC) No 79/2009 of the European Parliament and study of a direct injection compression ignition hydrogen
of the Council of 14 January 2009 on type-approval of engine. Int J Hydrogen Energy 2009;34:6516–22.
hydrogen-powered motor vehicles. Official Journal of the [19] Thomas CE. Fuel cell and battery electric vehicles compared.
European Union 4.2.2009:L35/32–46. Int J Hydrogen Energy 2009;34:6005–20.
[4] Muradov NZ, Veziroğlu TN. ‘‘Green’’ path from [20] Stockhausen WF, Natkin RJ, Kabat DM, Reams L, Tang X,
fossil-based to hydrogen economy: an overview of Szwabowski SJ, et al. Ford P2000 hydrogen engine design and
carbon-neutral technologies. Int J Hydrogen Energy 2008; vehicle development program. SAE Paper 2002. 2002-01-0240.
33:6804–39. [21] Natkin RJ, Tang X, Whipple KM, Stockhausen WF. Ford
[5] Winter C-J. Hydrogen energy-abundant, efficient, clean: hydrogen engine laboratory testing facility. SAE Paper 2002.
a debate over the energy-system-of-change. Int J Hydrogen 2002-01-0241.
Energy 2009;34:S1–52. [22] Tang X, Kabat DM, Natkin RJ, Stockhausen WF. Ford P2000
[6] Lee SJ, Yi HS, Kim ES. Combustion characteristics of intake hydrogen engine dynamometer development. SAE Paper
port injection type hydrogen fueled engine. Int J Hydrogen 2002. 2002-01-0242.
Energy 1995;20:317–22. [23] Szwabowski SJ, Hashemi S, Stockhausen WF, Natkin RJ,
[7] Vandenborre H, Sierens R. Greenbus: a hydrogen fuelled city Reams L, Kabat DM, et al. Ford hydrogen engine powered
bus. Int J Hydrogen Energy 1996;21:521–4. P2000 vehicle. SAE Paper 2002. 2002-01-0243.
[8] Guo LS, Lu HB, Li JD. A hydrogen injection system with [24] Kiesgen G, Klüting M, Bock C, Fischer H. The new 12-cylinder
solenoid valves for a four-cylinder hydrogen-fuelled engine. hydrogen engine in the 7 series: the H2ICE age has begun.
Int J Hydrogen Energy 1999;24:377–82. SAE Paper 2006. 2006-01-0431.
[9] Sierens R, Verhelst S. Experimental study of a hydrogen- [25] Wallner T, Lohse-Busch H, Gurski S, Duoba M, Thiel W,
fuelled engine. J Eng Gas Turbines Power 2001;123:211–6. Martin D, et al. Fuel economy and emissions evaluation of
[10] Sierens R, Verhelst S, Verstraeten S. An overview of BMW hydrogen 7 mono-fuel demosntration vehicles. Int J
hydrogen fuelled internal combustion engines. Istanbul, Hydrogen Energy 2008;33:7607–18.
Turkey: International Hydrogen Energy Congress and [26] http://www.acciona-energia.com.
Exhibition (IHEC-2007); 13–15 July 2007. [27] Gandı́a LM, Oroz R, Ursúa A, Sanchis P, Diéguez PM.
[11] Verhelst S, Sierens R. Aspects concerning the optimisation of Renewable hydrogen production: performance of an alkaline
a hydrogen fueled engine. Int J Hydrogen Energy 2001;26: water electrolyzer working under Emulated wind conditions.
981–5. Energy Fuels 2007;21:1699–706.
[12] Verhelst S, Sierens R. Hydrogen engine-specific properties. [28] Guelbenzu E, Sanchis P, Ursúa A, Marroyo L, Gandı́a LM,
Int J Hydrogen Energy 2001;26:987–90. Diéguez PM. Electrical/electronic device for measuring and
[13] Verhelst S, Maesschalck P, Rombaut N, Sierens R. Increasing emulating wind power systems; 2007. International patent
the power output of hydrogen internal combustion engines application WO2007/057493, Applicant: Acciona
by means of supercharging and exhaust gas recirculation. Int Biocombustibles S.A.
J Hydrogen Energy 2009;34:4406–12. [29] Diéguez PM, Ursúa A, Sanchis P, Sopena C, Guelbenzu E,
[14] White CM, Steeper RR, Lutz AE. The hydrogen-fueled internal Gandı́a LM. Thermal performance of a commercial alkaline
combustion engine: a technical review. Int J Hydrogen water electrolyzer: experimental study and mathematical
Energy 2006;31:1292–305. modeling. Int J Hydrogen Energy 2008;33:7338–54.
[15] Mohammadi A, Shioji M, Nakai Y, Ishikura W, Tabo E. [30] Ursúa A, Marroyo L, Gubı́a E, Gandı́a LM, Diéguez PM,
Performance and combustion characteristics of a direct Sanchis P. Influence of the power supply on the energy
injection SI hydrogen engine. Int J Hydrogen Energy 2007;32: efficiency of an alkaline water electrolyser. Int J Hydrogen
296–304. Energy 2009;34:3221–33.
[16] Kahraman E, Cihangir Ozcanlı S, Ozerdem B. An [31] Heywood JB. Internal combustion engine fundamentals. New
experimental study on performance and emission York: McGraw Hill; 1989.
characteristics of a hydrogen fuelled spark ignition engine. [32] Gillespie TD. Fundamentals of vehicle dynamics. SAE
Int J Hydrogen Energy 2007;32:2066–72. Publications; 1992.