Proceedings

Alexander Heintzel Hrsg.

Heavy-Duty-,
On- und
Off-HighwayMotoren 2022
Stand der Energiewende im
Heavy-Duty-Bereich

Proceedings

Ein stetig steigender Fundus an Informationen ist heute notwendig, um die immer
komplexer werdende Technik heutiger Kraftfahrzeuge zu verstehen. Funktionen,
Arbeitsweise, Komponenten und Systeme entwickeln sich rasant. In immer
schnelleren Zyklen verbreitet sich aktuelles Wissen gerade aus Konferenzen,
Tagungen und Symposien in die Fachwelt. Den raschen Zugriff auf diese
Informationen bietet diese Reihe Proceedings, die sich zur Aufgabe gestellt hat,
das zum Verständnis topaktueller Technik rund um das Automobil erforderliche
spezielle Wissen in der Systematik aus Konferenzen und Tagungen zusammen zu
stellen und als Buch in Springer.com wie auch elektronisch in Springer Link und
Springer Professional bereit zu stellen. Die Reihe wendet sich an Fahrzeug- und
Motoreningenieure sowie Studierende, die aktuelles Fachwissen im Zusammenhang mit Fragestellungen ihres Arbeitsfeldes suchen. Professoren und Dozenten
an Universitäten und Hochschulen mit Schwerpunkt Kraftfahrzeug- und Motorentechnik finden hier die Zusammenstellung von Veranstaltungen, die sie selber nicht
besuchen konnten. Gutachtern, Forschern und Entwicklungsingenieuren in der
Automobil- und Zulieferindustrie sowie Dienstleistern können die Proceedings
wertvolle Antworten auf topaktuelle Fragen geben.
Today, a steadily growing store of information is called for in order to understand the
increasingly complex technologies used in modern automobiles. Functions, modes of
operation, components and systems are rapidly evolving, while at the same time the
latest expertise is disseminated directly from conferences, congresses and symposia
to the professional world in ever-faster cycles. This series of proceedings offers rapid
access to this information, gathering the specific knowledge needed to keep up with
cutting-edge advances in automotive technologies, employing the same systematic
approach used at conferences and congresses and presenting it in print (available at
Springer.com) and electronic (at Springer Link and Springer Professional) formats.
The series addresses the needs of automotive engineers, motor design engineers and
students looking for the latest expertise in connection with key questions in their
field, while professors and instructors working in the areas of automotive and motor
design engineering will also find summaries of industry events they weren’t able
to attend. The proceedings also offer valuable answers to the topical questions that
concern assessors, researchers and developmental engineers in the automotive and
supplier industry, as well as service providers.

Alexander Heintzel
(Hrsg.)

Heavy-Duty-, On- und
Off-Highway-Motoren 2022
Stand der Energiewende im
Heavy-Duty-Bereich

Hrsg.
Alexander Heintzel
Springer Fachmedien Wiesbaden
Wiesbaden, Deutschland

ISSN 2198-7432
ISSN 2198-7440 (electronic)
Proceedings
ISBN 978-3-658-41476-4
ISBN 978-3-658-41477-1 (eBook)
https://doi.org/10.1007/978-3-658-41477-1
Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie;
detaillierte bibliografische Daten sind im Internet über https://portal.dnb.de abrufbar.
© Der/die Herausgeber bzw. der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung, die nicht
ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung des Verlags.
Das gilt insbesondere für Vervielfältigungen, Bearbeitungen, Übersetzungen, Mikroverfilmungen und die
Einspeicherung und Verarbeitung in elektronischen Systemen.
Die Wiedergabe von allgemein beschreibenden Bezeichnungen, Marken, Unternehmensnamen etc. in
diesem Werk bedeutet nicht, dass diese frei durch jedermann benutzt werden dürfen. Die Berechtigung zur
Benutzung unterliegt, auch ohne gesonderten Hinweis hierzu, den Regeln des Markenrechts. Die Rechte des
jeweiligen Zeicheninhabers sind zu beachten.
Der Verlag, die Autoren und die Herausgeber gehen davon aus, dass die Angaben und Informationen in
diesem Werk zum Zeitpunkt der Veröffentlichung vollständig und korrekt sind. Weder der Verlag noch
die Autoren oder die Herausgeber übernehmen, ausdrücklich oder implizit, Gewähr für den Inhalt des
Werkes, etwaige Fehler oder Äußerungen. Der Verlag bleibt im Hinblick auf geografische Zuordnungen und
Gebietsbezeichnungen in veröffentlichten Karten und Institutionsadressen neutral.
Planung/Lektorat: Markus Braun
Springer Vieweg ist ein Imprint der eingetragenen Gesellschaft Springer Fachmedien Wiesbaden GmbH und
ist ein Teil von Springer Nature.
Die Anschrift der Gesellschaft ist: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany

Foreword

Welcome
Sustainable climate protection can only be achieved if all the stops are pulled out
for energy converters and carriers. This can only be done with openness to technology. Especially in the heavy-duty sector, there is no way around alternative fuels and
hydrogen.
The 17th International MTZ Conference “Heavy-Duty, On- and Off-Highway
Engines” therefore offers engine developers and designers of commercial vehicles and
mobile machinery, as well as marine and stationary equipment a platform to inform
themselves about and exchange ideas on the latest developments in new power units
and alternative energy sources.
The conference in Donaueschingen will be supplemented by a guided tour of the
MAHLE GmbH plant.
I look forward to your participation in the conference.
Dr. Alexander Heintzel
On behalf of the Scientific Advisory Board
Chefredakteur ATZ | MTZ-Gruppe
Springer Nature

v

Vorwort

Herzlich willkommen
Nachhaltiger Klimaschutz lässt sich nur dann erreichen, wenn alle Register bei
Energiewandlern und -trägern gezogen werden. Dies geht nur mit Technologieoffenheit. Insbesondere im Heavy-Duty-Bereich führt an alternativen Kraftstoffen und
Wasserstoff kein Weg vorbei.
Die 17. Internationale MTZ-Fachtagung „Heavy-Duty-, On- und Off-Highway-Motoren“ bietet daher Motorenentwicklern und -konstrukteuren von Nutzfahrzeugen, mobilen Maschinen, Marine und stationären Anlagen eine Plattform, sich
über die aktuellen Entwicklungen neuer Aggregate und alternativer Energieträger zu
informieren und auszutauschen.
Ergänzt wird die Tagung in Donaueschingen durch eine Führung im Werk der
MAHLE GmbH.
Ich freue mich auf Ihre Teilnahme an der Tagung.
Dr.Alexander Heintzel
Für den Wissenschaftlichen Beirat
Chefredakteur ATZ | MTZ-Gruppe
Springer Nature

vii

Inhaltsverzeichnis

The TCG 7.8 H2 – Further Development Steps to Realize a
CO2 Free-Powertrain for NRMM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Georg Töpfer, Heiner Bülte, Benedikt Nork, and Carsten Funke

1

The Development of a 6-Cylinder Hydrogen Engine for the
Off-Highway Market. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bouzid Seba and Ulrich Weiss

13

Investigation of An Ammonia Diesel Dual-Fuel Combustion Process
on a Heavy-Duty Single Cylinder Research Engine for the Development
of Suitable Simulation Tools for Maritime Applications . . . . . . . . . . . . . . . . . . .
Till Mante, Sascha Prehn, Martin Theile, Lars Seidel, Laura Mestre,
Bert Buchholz, and Fabian Mauss

24

H2 ICE DI Multicylinder Engine Tests for Thermodynamics and Component
Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Simon Schneider, Christian Trabold, Thomas Friedrich, Florian Mayer,
Fabian Weller, and Roman Stiehl

40

Assessment of a Direct-Injection, Spark-Ignited, Hydrogen-Fuelled
Heavy-Duty Engine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
John Hughes, David Bennet, Angela Loiudice, Nicholas Coles,
Trevor Downes, Agam Saroop, Richard Penning, Lukáš Valenta,
Peter Rabanser, Jonathan Davis, Jackson Harvey-Bush,
Alvaro Concepcion Calero, Richard Osborne, Penny Atkins,
Roger Allcorn, and Nigel Fox

53

The Compact Catalytical Heater (CCH): Thermal Management
for HD EU-VII/EPA-27 with Low Impact on Existing EATS Architectures. . . .
Manuel Presti, Oswald Holz, Mathias Keck, and Dennis Sailer

76

Liebherr’s Approach to Hydrogen Fuel Injection Systems. . . . . . . . . . . . . . . . . .
Richard Pirkl, Mario D’Onofrio, Lydia Kapusta, and Dennis Herrmann
Hydrogen Dosing Systems for Large Engines: Challenges and Potentials
of Three Different Approaches. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Enrico Bärow, Michael Willmann, Andreas Kühner, and Rick Boom

95

112

ix

x

Inhaltsverzeichnis

Hydrogen Storage Technologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mathias Keck, Dirk Bessey, Frank Buehler, and Manuel Eugen Faiß
Well-to-Wheel CO2-Analysis of Different Powertrain Systems
on Representative Heavy-Duty Mission Profiles . . . . . . . . . . . . . . . . . . . . . . . . .
Nicolas Hummel, Tim Herold, and Christian Beidl
Hydrogen in the Gas Network – Challenges and Solutions for High
Performance Engines for Power Generation. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Clément Leroux, Robert Böwing, Bernadet Hochfilzer,
Alexander Zuschnig, and Manuel Behr

123

131

143

Safe and Sustainable Testing of Hydrogen Powertrains. . . . . . . . . . . . . . . . . . . .
Nicolas Weyland

158

Hybrid PEM Fuel Cell Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sönke Gößling, Matthias Bahr, and Felix Smyrek

167

Fuel Cell System Development for Heavy Duty Application. . . . . . . . . . . . . . . .
Stephan Schnorpfeil, Arne Kotowski, Hauke Sötje, and Guido Hartmann

172

Autorenverzeichnis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

185

The TCG 7.8 H2 – Further Development
Steps to Realize a CO2 Free-Powertrain
for NRMM
Georg Töpfer(*), Heiner Bülte, Benedikt Nork, and Carsten Funke
DEUTZ AG, Cologne, Germany
{georg.toepfer,heinrich.buelte,
benedikt.nork,carsten.funke}@deutz.com

Abstract.  The development of a hydrogen internal combustion engine (HICE)
represents special challenges in combustion development. The HICE is being
developed with external mixture formation and injection of hydrogen upstream
of the intake valves. The paper presents simulation results to show the
challenges for the hydrogen injection (H2-injection) system to avoid backfire
into the air tract. The achievable engine performance and emission behavior
of the PFI-engine are significantly influenced by the turbocharging technology and its ability to rise charge pressure to necessary level. The maximum
achievable engine power is optimized by simulating the complete system.
The permissible operating conditions are respected to avoid overloading the
components. In NRMM applications in particular, the high engine dynamics
required (load step) must be considered. Experience in real-world applications
enables the suitability of the HICE to be evaluated. It can also be used to
evaluate the robustness of the engine concept. For this purpose, several
demonstrator projects with different applications are being realized, which are
also briefly presented in this paper.
Keywords:  Hydrogen · Internal combustion engine · Off-Highway

1 Sustainable Drive Systems for CO2-Reduction at NRMM
Currently, most of the EU’s activities are focused on reducing CO2 emissions in
the mobility sector. However, there are also opportunities for CO2 reduction in the
relevant area of Non-Road Mobile Machinery (NRMM). According to statistics of
TREMOD report [1], Germany alone emitted 13,9 million tons of CO2 in the NRMM
sector in year 2018. Diesel engines at agricultural applications and construction
machinery together account for around 80% of CO2 emissions. Currently, most of
NRMM applications run on fossil fuels, with a statistic average lifetime of about
15 years.

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 1–12, 2023.
https://doi.org/10.1007/978-3-658-41477-1_1

2    
G. Töpfer et al.

NRMM vehicles are mostly used for business purposes, i.e., if the machine works
reliably and with a high efficiency, the machines will continue to be used. A new
acquisition means an investment. Therefore, alternative drive systems are needed in a
timely manner, which do not limit the efficiency of the machines regarding the daily
work performance (Fig. 1).

Fig. 1.  Share of pollutant emissions from Non-Road Mobile Machinery (NRMM) in Germany
2018 by sector [1]

The emitted CO2 emissions show that there is a need for action in the sector of
mobile machinery. DEUTZ follows two basic approaches for sustainable drivetrains
[2].
• Development of electric drivetrains with 48 and 360 V within E-DEUTZ strategy
with available power up to 60 kW. An operation range extension and independence
from electric charging infrastructure will be envisaged with fuel cell technology in
nearby future.
• The further development of combustion engines will head forward operation with
non-fossil gaseous and liquid fuels. The current focus is on the use of hydrogen,
which will be the main energy carrier in Europe due to the German national hydrogen strategy [3] and will therefore have very good availability in nearby future.
Thus, the question arises why both a hydrogen internal combustion engine (HICE)
and electric drive systems with fuel cell (FC) are in the portfolio. Both drivetrains
reveal dedicated advantages which can be explained by evaluating the operating
conditions of the work machines using the example of a hydraulic excavator.

The TCG 7.8 H2 – Further Development Steps …    3

Figure 2 shows a characteristic load profile of a hydraulic excavator application.
Depending on the power requirement, the vehicle hydraulics operate at specific,
controlled engine speeds at which the torque and thus the work output (moving
the boom, rotating the upper carriage) are continuously varied in a highly dynamic
manner. The engine governor has the task of maintaining the speed in the case of a
sudden load increase by the hydraulics, which demands high requirements on the
engine dynamics.

Fig. 2.  Engine map with characteristic load profile of a hydraulic excavator application

A general comparison of the two systems is shown in Fig. 3. The combustion
engine as a currently integrated drive system meets these requirements and thus sets
the expected development targets for the HICE. An FC system can also be applied.
The compact, electric drive system provides an immediately available torque over the
complete speed range. The high energy supply for dynamic operation is provided by
an energy storage system (battery).

4    
G. Töpfer et al.

Fig. 3.  Comparison of fuel cell system to hydrogen combustion engine

When designing the cooling system, it should be noted that an adjustment
should be made for both drive systems. Here, it is to be expected that in high-load
applications, the cooling capacity required for a FC is somewhat higher than for a
HICE due to low exhaust gas temperatures and enthalpy [4]. This can be compensated
by adapting a greater cooler package.
Looking on hydrogen purity, the HICE also allows hydrogen with a lower degree
of purity.
NRMM vehicle often operate in dusty environment. Also, the vehicle is not
suspended and hence the drivetrain experience severe vibrations during operation.
In summary, there are some special aspects to be considered in the NRMM
application for both drive systems. A suitable drive system must be selected
regarding efficiency, availability of energy infrastructure, acceptable operation range.
Additionally, the Total Cost of Ownership (TCO), especially operating costs and
service life, must be considered.

2 Hydrogen Internal Combustion Engine (HICE)
The TCG 7.8 H2 is a six-cylinder inline engine with a displacement of 7.8 L, a
bore of 110 mm and a stroke of 136 mm (Table 1). The gas engine is based on the
series diesel engine TCD 7.8 with further changes to obtain a hydrogen fired engine.
The hydrogen combustion engine with hydrogen PFI injection (external mixture
formation).
Table 1.  Engine specification
Cylinder
Displacement
Bore/Stroke
Injection

6 inline
7.8 L
100/136 mm
Hydrogen PFI System

The TCG 7.8 H2 – Further Development Steps …    5

To develop a gas engine version several changes were done (Fig. 4). The engine
receives a modification of the existing cylinder head to accommodate the ignition
system for a hydrogen lean combustion, which avoids a residual spark discharge due
to the low ignition energy required and thus unwanted combustion events. The spark
plugs especially suited for hydrogen engines with a low heat characteristic and special
designed electrodes are used.
The gas exchange valves and its seats are improved with wear-reducing measures,
which consider the low lubricity of hydrogen.

Fig. 4.  Principle, necessary changes of base diesel engine to obtain hydrogen combustion
engine

The compression ratio is reduced by means of a new piston bowl. The piston ring
package is adapted to minimizes oil intake and hence prevent pre-ignition due to small
oil droplets.
The intake manifold was redesigned to accommodate the injectors in front of each
inlet valve. To compensate for intake air supplanting by the high volume of injected
hydrogen, the turbocharger is modified so that high boost pressure can be provided
over a wide speed range. Boost control is software-based with an electric actuator.
The engine has an externally cooled and controlled exhaust gas recirculation (EGR).
Depending on the power density and required engine dynamics, the engine is
equipped with a conventional exhaust aftertreatment system with SCR for NOx
reduction, which is adapted for typical operation with hydrogen (higher amount of
water content in exhaust gas, suitable for temperature range).
The challenging tasks are the development of a H2-injection system, in particular
the injector and the control software for a spark ignited lean burn engine considering
low NOx emission during dynamic engine operation.

6    
G. Töpfer et al.

2.1 Hydrogen Injectors
The injectors are the key component in the development of a hydrogen engine. They
must operate with high repeatability over the entire service life in the entire engine
map and always close tightly. Since there is also only a very short time slot available
for metering into the inlet port, a conflict of objectives arises between the required
flow rate of the injectors at the point of maximum power and the metering accuracy at
the minimum quantity requirement at idle.
In the case of large injection quantities, it is necessary to resolve the conflict of
objectives between mixture homogenization and safety against backfire. While
the longest possible injection time would be advantageous for good mixing of
hydrogen and air, the formation of an ignitable mixture in the intake manifold and
the impingement of such a mixture on a potential ignition source must be avoided
(Fig. 5).

Fig. 5.  Simulation of the hydrogen concentration in the intake manifold during a PFI injection

The TCG 7.8 H2 – Further Development Steps …    7

To prevent ignitable mixture from remaining in front of the closing inlet valve,
the end of the injection must also be selected with sufficient time gap before the inlet
closes.
The location of injection into the inlet port also plays a major role. Here a
compromise must be found between the requirements for mixture formation and
safety against backfiring, too.
2.2 Charging Concepts
The choice of a suitable turbocharging concept for the TCG 7.8 H2 presents a
particular challenge, since the conflict of requirements between high specific power
and very good dynamic response must be resolved.
To keep NOx Engine Out (EO) emissions at a very low level, a hydrogen engine
must be operated with an air-fuel ratio λ ≥ 2. In combination with the low mixture
calorific value due to external mixture formation (PFI), this leads to very high charge
pressures to achieve a specific power up to 28 kW/l.
The use of a turbocharger (TC) with variable turbine geometry (VTG) improves
dynamic response with nearly comparable limits of maximum power output. For
the complete concept, the EGR system is additionally adopted with modifications.
Since high-pressure (HP)-EGR reduces the air mass flow through the compressor, the
compressor and turbine wheel can be made smaller, resulting in a lower moment of
inertia and thus an accelerated buildup of boost pressure in the event of a load jump.
In this way, the conflict between high specific power and rapid dynamic load buildup
can be resolved without two-stage turbocharging.
Due to the requirements for H2-injection and charging described above, there
are fixed, limiting operating conditions. With the aim of a series application, the
robustness of the combustion process must also be validated and the resulting
installation restrictions about pressure losses and max. permissible temperatures
must also be included in the evaluation. The simulations carried out show clear performance limits, restrictive boundary conditions are summarized (Table 2). Knocking
intensity was also calculated and considered. All simulation results were obtained
with a VTG –TC, the center of energy conversion (combustion) is constant.
The maximum achievable engine power is limited by the maximum permissible
boost pressure. Acceptable air-fuel ratio λ and the associated NOx emissions define
the maximum H2-injection quantity.

Table 2.  Installation restrictions
Max. temperature downstream compressor
Max. temperature upstream turbine
Max. TC speed

Tds. Compr. max.
Tus. Turb. max.
nTC max.

8    
G. Töpfer et al.

Figure 6 shows the simulation result for an air-fuel ratio λ1. The respective
boundary conditions are shown in the diagram. The max. temperature downstream
compressor Tds. Compr. max. is power limiting with increasing EGR rate and compressor
mass flow. The turbine inlet temperature increases with decreasing EGR rate and
thus also limits the power. This results in a range in which operation is feasible with
the given boundary conditions. At this operating point, a maximum power output of
164 kW/2,200 rpm is achievable.
For further power gain, a further increase in the H2-injection quantity at constant
maximum boost pressure is required. This results in a richer mixture with airfuel ratio λ2(λ2 < λ1). The simulation results are shown in Fig. 7. Compared to the
previous operating point, the possible operating range is smaller. A higher EGR
rate is required, and to avoid an excessively high intercooler outlet temperature,
the air mass flow rate must be limited. Under these assumptions, a rated power of
220 kW/2,200 rpm is achievable. Due to enriched air-fuel ratio λ2, a higher NOx EO
emission level should be reduced by a suitable exhaust gas aftertreatment system
(SCR).

Fig. 6.  Simulation result at rated power, n = 2,200 rpm, λ1, with installation restrictions

The TCG 7.8 H2 – Further Development Steps …    9

Fig. 7.  
Simulation result at rated power, n = 2,200 rpm, λ2(λ2 < λ1), with installation
restrictions

3 Demonstrator Application and its Use for Series
Development
In a series development, it is necessary to validate the temporal robustness of the
HICE under typical operating conditions. The combustion process shows that a PFI
hydrogen engine has a higher tendency to abnormal combustion events (backfire, preignition and knocking) than other gas engines and therefore reacts more strongly to
changes of influencing parameters. Over engine lifetime, a change in H2-injection
quantity (e.g., due to injector wear), drift of a sensor (e.g., accuracy, measured value),
or changes of operation conditions (e.g., filter clogging, higher ambient temperature)
could affect combustion.
In the case of hydrogen exposed components, the question arises of a possible
change in material properties due to hydrogen deposition and a resulting reduction
in tensile strength (embrittlement). Especially for cast materials (e.g., cylinder head)
almost no information can be found in the literature.
Blowby and its possible change due to wear of piston rings must also be
evaluated. Hydrogen enters the crankcase through blowby gases and accumulates
in the crankcase. It must be ensured that the concentration does not exceed 4% by
volume and that an ignitable mixture is present.

10    
G. Töpfer et al.

In addition to intensive functional testing further experience will be gained by
means of field tests, ideally in 4-season operation. Due to the intended short development period, it is necessary to gather experience parallel to series development
activities by means of very early applications of demonstrators. Different applications
are planned for this purpose.
In the case of applications, a distinction must be made between stationary
applications operated at constant speed and mobile applications with variable speed
and high dynamic requirements. The respective operation has an influence on the
engine control and operating strategy. Stationary operation with constant speed and
low motor dynamics can be realized in earlier stage of development in terms of the
control software.

Fig. 8.  Hydrogen genset demonstrator with TCG 7.8 H2

For this purpose, a hydrogen genset (H2-genset) was successfully realized as a
demonstrator in June 2022 with a very early development stage of the TCG 7.8 H2
(Fig. 8). The H2-genset generates an electrical power up to 170 kVA in grid-parallel
operation at a constant speed of 1,500 rpm.
In addition, the TCG 7.8 H2 is being further developed and tested for a truck
application in a project funded by the Bundesministerium für Digitales und Verkehr
(BMDV). The research project aims to demonstrate the sustainability potential of
trucks with hydrogen combustion engines in transport logistics.
NRMM applications require further demands on the operation of the hydrogen
engine due to high dynamic requirements and the harsh operating conditions. For this
reason, DEUTZ and other cooperation partners are building a hydraulic excavator as
a demonstrator in another BMBF (Bundesministerium für Bildung und Forschung)funded project.

The TCG 7.8 H2 – Further Development Steps …    11

4 Summary
DEUTZ is currently developing a Hydrogen Internal Combustion Engine (HICE) with
a displacement of 7.8 liters, the TCG 7.8 H2. It is foreseen, that this new engine type
will set up into series production in the year 2024.
Studies [4, 5] showed, that HICE reveals several advantages at NRMM
applications. In NRMM the powertrain must operate well at rough dust contaminated
environment and experience high vibration load at a lot of applications. Operation
cycles are mainly highly dynamic, and the engine load factor is quite high, depending
on the application. Mobile machinery mainly operates at stable place or with a small
vehicle velocity, hence the coolant radiator is always supported by an additional
coolant fan. These requirements could very well be fulfilled with a HICE. As with
all gaseous fuels, the challenge is to integrate sufficient tank volume into the vehicle
to realize an appropriate operation range and high working efficiency, which are the
most important duties of the mobile machinery.
In the early stage of development, a pilot project was realized within
predevelopment and the TCG 7.8 H2 engine was applied at a mobile power generator
(H2-genset). At the end of year 2021 the first start of the power generator was realized.
The TCG 7.8 achieves a mechanical power of 150 kW at 1,500 rpm. The strategic
aim of the pilot project is to demonstrate the concept in a first application and to
gather experience in operating HICE over a longer time. The H2-genset is a necessary
supplement to the energy supply of the future and allows to produce CO2-free on-site
electric energy at places without connection to the power grid, such as construction
sites.
Next step is to extend the application to the construction machinery, enlarge the
operation range (engine speed and torque) and to realize a HICE, which is capable to
drive a hydraulic drivetrain. This step enquires an increased engine operation range
(speed and power) and improved engine dynamics in transient operation. The high
engine dynamic is necessary to realize high torque demands requested from hydraulic
systems. In oncoming steps, it is foreseen to build up a 30t hydraulic excavator
together with strong development partners within a funded project.
The TCG 7.8 H2 is based on the diesel engine with same displacement. Several
changes in hardware are necessary to obtain a HICE suitable engine design. It was
decided to develop the engine with a Port-Fuel-Injection (PFI). To compensate the
lower caloric mixture value with a high boost pressure a suitable turbocharger technology was chosen. Special attention was taken to realize good dynamic response at load
steps. The engine is equipped with an EGR system to realize a high-power density
at low knocking sensitivity as well as low pumping losses at part load. Turbocharger
matching and the EGR system layout were optimized using simulation tools.
Development aim was to improve the hardware to minimize flow losses.
Due to the chosen PFI for hydrogen, measures must be taken, to avoid backfiring
(self-ignition of hydrogen-air mixture) in the inlet port. Hence, additionally
optimization steps were investigated to avoid the coincidence of ignitable mixture and
potential ignition sources.

12    
G. Töpfer et al.

Abbreviations
CO2	Carbon Dioxide
EGR	Exhaust Gas Recirculation
EO	Engine Out
FC	Fuel Cell
HICE	Hydrogen Internal Combustion Engine
HP-EGR	High-Pressure Exhaust Gas Recirculation
GHG	Greenhouse Gas
NOx	Nitrogen Oxides
NRMM	Non-Road Mobile Machinery
PEMFC	Polymer Electrolyte Membrane Fuel Cell
PFI	Port Fuel Injection
VTG	Variable Turbine Geometry

References
1. Aktualisierung der Modelle TREMOD/TREMOD-MM für die Emissionsberichterstattung
2020 (Berichtsperiode 1990–2018). Berichtsteil „TREMOD-MM,“ https://www.umweltbundesamt.de/sites/default/files/medien/1410/publikationen/2020-06-29_texte_117-2020_
tremod_mm_0.pdf
2. Schwaderlapp, M., Bülte, H., Plumpe, A.: Solutions for CO2-free powertrains for mobile
machinery. In Heavy-Duty-, On- und Off-Highway-Motoren (2021)
3. Die Nationale Wasserstoffstrategie. https://www.bmbf.de/bmbf/shareddocs/downloads/files/
die-nationale-wasserstoffstrategie.pdf?__blob=publicationFile&v=1. BMWi (2020)
4. Systemvergleich zwischen Wasserstoffverbrennungsmotor und Brennstoffzelle im schweren
Nutzfahrzeug. e-mobil BW, Studie, https://www.emobil-sw.de/fileadmin/media/e-mobilbw/
Publikationen/Studien/e-mobilBW-Studie_H2-Systemvergleich.pdf (2021)
5. Wasserstoffverbrennungsmotor als alternativer Antrieb. Metastudie, NOW GmbH, https://
www.now-gmbh.de/wp-content/uploads/2021/10/NOW_Metastudie_Wasserstoff-Verbrennungsmotor.pdf (2021)

The Development of a 6-Cylinder Hydrogen
Engine for the Off-Highway Market
Bouzid Seba(*) and Ulrich Weiss
Liebherr Machines Bulle SA, Bulle, Switzerland
{bouzid.seba,ulrich.weiss}@liebherr.com

Abstract.  Hydrogen internal combustion engine (H2-ICE) has the potenzial to
be an enabler for fast decarbonization in several sectors. In this context, Liebherr is developing a hydrogen combustion engine based on spark ignition. This
new powertrain is a part of a strategic approach toward zero emission technologies for off-highway sectors.
This technology has gained an important interest from the off-highway
sector due to several reasons such as:
•
•
•
•

Robustness of the technology regarding the duty conditions in the Offhighway sector
Cost of the power train system
Quick time to market solution
Similar installation condition compared to the current powertrain

Currently, the diesel engine finds its way in a wide range of vehicles in the
Off-highway sector and transformation of the current power train to the new
zero emission technologies (Hydrogen Engine) will require considering several
aspects:
•
•
•
•

Investigation and analysis of the current vehicle architecture capabilities
Development and adaptation of several subcomponents of the powertrain
(such as cooling system)
Modification of the storage system and the fuel path from the tank to the
engine fuel interface
Analysis and installation of the required additional elements regarding the
safety precaution

The development of the technology started with several investigation in the
combustion engine test bench; this new power train has been then introduced in
vehicle integration and used to build-up a zero emission excavator.
This paper reviews the current investigations regarding the transformation/
adaptation of an excavator powertrain (Diesel-hydraulic excavators) to
a hydrogen combustion engine powered machine: Description of the
excavator characteristics and analysis of the different characteristics of this
new powertrain. Thenceforth, the fuel and storage system is explained and
illustrated, followed by a review of several safety precaution aspect. The
© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 13–23, 2023.
https://doi.org/10.1007/978-3-658-41477-1_2

14    B. Seba und U. Weiss
paper concludes with an evaluation of the new proposed powertrain in terms
of dynamic, efficiency and emission. At the end an outlook regarding the
further application of the hydrogen combustion engine in different other type
of machines in construction and off-highway sector has been given.
Keywords:  ZE (Zero emission) · Future power trains · Hydrogen combustion
engine (H2-ICE) · Safety requirement · Hydrogen storage system · Tanks · Offroad vehicles · Excavator

1 Introduction
In the last 5 years, the development of the hydrogen combustion engine gained a
strong interest from different industries. Several entities in off-road and heavyduty applications now see that the H2-ICE will play an important role in the
decarbonisation of their vehicle. As mentioned in literature and in our previously
publication [1], the main benefits of the technology are:
• The robustness against environmental working condition
• The high degree of performance in transient operation
• The efficiency in high load operation
One of the drivers for the introduction of zero emission technologies is the forecast of
future CO2 emissions regulations for different industry sectors. Additionally, several
emission restrictions are expected in the cities, impacting the implementation and the
deployment of ZE-Technologies for specific applications (Fig. 1).

Fig. 1.  EU Regulation for On-Road/Off-Road

The off-road sector can expects the combustion engine using hydrogen to be
one of the technologies aiming toward these targets. Hydrogen has been the object
of several investigations over the years. Different applications such as spatial, automotive and other industrial applications were especially involved. A considerable
gain in experience over fuel compatibility with different materials and fuel properties
happened thanks to those investigations. Figure 2 summarizes some specific

The Development of a 6-Cylinder Hydrogen …    15

properties of hydrogen and displays a comparison with different fuels. The hydrogen fuel has a high gravimetric calorific value. However, its lower density reduces
the energy density of the mixture. This leads to lower performance especially for
some injection concepts such as H2 PFI. This characteristic has a direct impact on
the torque performance of the engine which decreases with higher air/fuel ratio. This
inconvenience is a challenging issue in both heavy-duty and off-road applications
due to packaging requirements. H2 direct injection (H2 DI) is one possible solution
to improve the mixture heating value especially if the fuel injection occurs after full
closure of the inlet valve. However, H2 DI main drawback is the emission level,
which is higher due to bad homogeneity of the air/fuel mixture.

Fig. 2.  Fuel properties and comparison

For the first demonstration of the technology, Liebherr presented an excavator
powered by a PFI combustion engine, assuring the deployment capability of the
technology for off-road sector.

2 Development Phases of the Technology
Following the selection of the combustion technology, several steps ensured the
meeting of the requirement of the final product “excavator with H2-ICE”:
•
•
•
•

Preliminary assessments and target definition
Engine design & component selections
Engine preparation and optimization
Vehicle testing

Several challenges emerge when operating in the off-road sector. Here is the summary
of the main conditions to consider:
•
•
•
•
•
•
•

Resistant to hot and cold temperatures environment
Higher transient operation
(especially notable for application such as excavators)
Operation in high humidity environments
High vibration/shock intensity
Operation at higher altitude (>3000 m)
High lifetime expectation

16    B. Seba und U. Weiss

• Operation in dusty environment (possible affect the air and fuel quality)
The combustion engine with a diesel configuration already proved its capability
to fulfill the above-mentioned conditions. The flexibility of the diesel technology
also enables one specific unit to cover different application, eliminating the need of
hybridization (Figs. 3 and 4).

Fig. 3.  Development phases

Fig. 4.  Applications with D966

3 Engine Data and Description
The engine used for excavator applications is the Liebherr hydrogen engines H96613.5 L. This engine family will find its way in various applications in the off-road
sectors. Figure 5 pictures the original diesel engine and the converted hydrogen
engine:

The Development of a 6-Cylinder Hydrogen …    17

Fig. 5.  Diesel original engine “left” and hydrogen engine “right”

Table 1 summarizes the main specifications and characteristics of the engine. The
consideration of several compression ratios and other hardware parameters suggests
future optimization of the engine. Expectations point toward improvement of the main
performance indicators such as efficiency and engine power density. The design of
fuel path components also relies on the necessary optimization. Among the points
to address is the rail pressure pulsation level or the time response of the pressure
regulation loop in transient operation.

Table 1.  Engine data of H966

Several combustion technologies underwent development at Liebherr, revealing
their potential regarding the engine performance, power density and NOx emissions.
In our previous publication [1], we showed the advantages and drawbacks of both
injection strategies (port fuel injection “PFI” and direct injection “DI”) using the
measurements performed with a 4-cylindre engine “H964”. For different reasons such
as residual pressure level and technology maturity, following activities used the PFI
technology.

18    B. Seba und U. Weiss

4 Engine Optimization for Vehicle Application
The optimization of the hydrogen combustion engine for vehicle application
considered several phases. Figure 6 summarizes the main phases of this H2-ICE
development.

Fig. 6.  Vehicle optimization phases

After the first investigation (in steady state operation), where the engine entered
the test bench for performance characterization, several steps aimed to improve the
power density by considering the following challenges:
• Backfire in certain operating conditions (higher knocking sensitivity)
• Safety measures related to hydrogen applications: Fuel system, etc …
• Lower NOx emissions by ensuring in the same time an acceptable dynamic
behavior of the engine
Additionally, the engine-vehicle optimization requires the preparation of the engine
for dynamic and transient operation. Considering the excavator is one of the most
challenging vehicles dynamically, dynamic evaluation followed these two stages:
• Load steps: The engine follows load profiles derived from real operating cycles.
As a torque increase happens, analysis of the speed drop and recovery time of the
engine speed evaluates the load step response. This is an important criterion for the
performance analysis of a vehicle.
• NRTC: NRTC cycles demonstrate the general performance of the engine in
different off-road vehicles just as transient driving cycle for mobile off-road
diesel engines do. This cycle also serves emission evaluation between different
combustion technologies.

The Development of a 6-Cylinder Hydrogen …    19

Additionally to the load step with “torque increase”, there are also some technical
challenges in the phase of torque decreases “release”. This concerns mainly the
injection path optimization. Solution to the later can emerge in the test bench phase
using for example dedicated combustion, regulation of the rail pressure and purge
procedures.

5 Vehicle Integration
The following milestones summarize the identification of the components and their
integration to the engine and to the vehicle (Fig. 7):
• Analytical investigation using component characteristics
• Use of 1D-Simulations and measurement data in order to estimate the accurate fuel
consumption
• Component identification and creation of specification data communicated to
different suppliers
• Definition of all interfaces between the tank pressure regulator and engine
• Investigation and implementation of all safety items

Fig. 7.  Development phases

Figure 8 illustrates the main components and layout of the vehicle integration. Key
components, such as the pressure regulator, the fuel path elements and the ECU, have
been the subject of several investigations. Investigations happened in the component

20    B. Seba und U. Weiss

Fig. 8.  Component of the DI hydrogen combustion engine

test bench before HiL “Hardware in Loop” testing. This was crucial to assess the
interaction between the actuators and the ECU, the electrical interfaces as well as the
time response of the components.

6 Storage System
The fuel consumption of the combustion engine in different operating point is the
reference for hydrogen amount calculation. The storage and integration of the tank
is one of the main investigated subjects in the integration of the H2-Technology. The
autonomy of the vehicle operation for a certain given packaging is dependent on
different parameter such as:
• Storage pressure/temperature
• Powertrain efficiency
• Injection strategy (which determines the minimal pressure level that is required for
engine operation)
• Tank geometry

The Development of a 6-Cylinder Hydrogen …    21

Fig. 9.  Storage system

Figure 9 shows the equivalent volume required for 4 h of the excavator operation. As
expected, diesel has the lowest required volume value (150 L) thanks to its higher
specific volumetric energy density (10 kWh/L). Hydrogen storage systems with
“700 bar” also offer an interesting compromise compared to other ZE-Technology
competitors based on electrical storage system.

7 Outlook
The hydrogen technology offers a large potenzial for decarbonisation in off-road
sectors by using direct fuel combustion. This potenzial covers the current local
emission regulations (Tank-to-Wheel) in terms of CO2 reduction as well as the global
emissions regulation by using green electricity. The main areas that still require
optimization to ensure the entire satisfaction of different aspects of industrialization
are:
• Storage system optimization
• Efficiency of the complete power train
• Infrastructure for fuel supply in working site
The extension of the technology toward different vehicle type in off-road sector
is on-going considering the many advantages of the technology presented in this
investigation.

8 Summary and Conclusion
Development of an appropriate ZE “zero emission” technology is one of the main
investigation subject in the last decade in different industries sector. The target for
CO2 emission reduction has been already introduced in heavy duty “On Road sector”
with milestones in 2025 (−15 %) and 2030 (−30 %). Furthermore, this CO2 emission
reduction target will be extended to several sectors in the future. There is additionally

22    B. Seba und U. Weiss

some CO2 regulation that are introduced in some cities in Europe. The combustion
engine has already demonstrated its robustness in term of operation in harsh condition
such as temperature, altitude, and dusty environment. These characteristics will
strength the introduction of CO2 free power train based on combustion.
One promising way for decarbonization is to use green energy to produce hydrogen and build CO2 neutral powertrain based on hydrogen combustion engines. This
type of technology could even operate with blue or other low carbon hydrogen fuel.
The way for production of green hydrogen in large scale will need several years
for the building and the installation of necessary infrastructure. In the transition
period, several type of (low carbon) hydrogen could be used with combustion
engine to ensure decarbonization even in early phase. Hydrogen combustion engine
investigation has been performed for several years and in different area/application.
Liebherr is developing hydrogen combustion engine in order to be ready for the future
CO2 emission reduction target that will be implemented in different area. One of the
Offroad vehicle which operate in very harsh condition is the excavator, furthermore
the operation with this type of vehicle is challenging in terms of dynamic.
This integration of the H2-ICE and the corresponding H2- actuators and sensors
has been performed and successfully integrated in the excavator. Several technical
challenges has been solved during this phase of integration and testing with the
vehicle. Several specific tests related to excavator operation has been successfully
realized with H2-ICE and excavator. This activity has demonstrated that operation of
the H2-ICE can fulfill the different requirements of the vehicle and generally in Offroad sector. Additional challenge in the vehicle integration phase is to find the optimal
size of the tank that cover the necessary autonomy of the vehicle by considering the
real operating cycle. In another hand, the H2-storage system has advantages regarding
the filling time, which is shorter compared to electrical storage systems, In order to
deploy this technology in large scale production, additional infrastructure are required
to ensure the transport the fuel in the working sites. Mobile tank station that can
ensure the supply of the fuel in the working site could be one of the possible technical
solutions to fulfill requirements in terms of necessary tanking frequency. The problem
and dilemma of hydrogen network filling stations number and the demand from
customers is one of the key elements for a profitable series production. The launch
of small series production of hydrogen vehicles in working sites will help to grow the
interest to the fuel stations network.
In general, the potential of the technology to operate in excavator environment has
been shown and demonstrated, the spread out of the technology to other vehicle type
in the Offroad sector is one of the next steps and outlook for future horizon.

References
1. Seba, B.: Hydrogen Combustion Engine – A Suitable Concept for Decarbonization in the
Offroad Sector. ATZ live (2021)
2. Weiss, U.: The Off-Highway Sector in the Field of Tension of Future Powertrain Concepts
– What Chances Does the Internal Combustion Engine have in This Industry? Baden Baden
Engine congress, Baden Baden (2022)

The Development of a 6-Cylinder Hydrogen …    23
3. Kovacs, D., Rezaei, R., Englert, F., Hayduk, C., Delebinski, T.: High Efficiency HD Hydrogen Combustion Engines: Improvement Potentials for Future Regulations (2022)
4. Krepec, T., Tebilis, T., Kwok, C.: Fuel Control Systems for Hydrogen-Fueled Automotive
Combustion Engines – A Prognosis. Elsevier (1984)
5. Heywood, J.B. (2018). Internal Combustion Engine Fundamentals. McGraw-Hill (2018)
6. Freund, E., Lucchese, P.: Hydrogen, the Post-Oil Fuel, Edition Technip (2013)
7. Hermanns, R.T.E.: Laminar Burning Velocities of Methane Hydrogen Air Mixtures, Dissertation, Eindhoven University of Technology (2007)
8. Delneri, D.: Combustion System Optimization for Alternative Fuels, 17th Conference The
Working Process of the Internal Combustion Engine (2019)
9. White, C.M., Steeper, R.R, Lutz, A.E.: The hydrogen-fueled internal combustion engine: A
technical review. Int. J. Hydrogen Energy (2006)
10. Van der Put, D., et al.: Efficient commercial powertrains – how to achieve a 30% GHG
reduction in 2030. In: Proceedings of the FISITA 2020 world congress, Prague, 14–18
September 2020
11. Tang, Q., Liu, J., Zhan, Z., Hu, T.: Influences on combustion characteristics and performances of EGR vs. lean burn in a gasoline engine. SAE technical paper 2013-01-1125
(2013)
12. Das, M.L.: Hydrogen engine: Research and development (R&D) programmers in Indian
Institute of Technology (IIT), Delhi. Int. J. Hydrogen Energy (2002)
13. Caton, J.: A comparison of lean operation and exhaust gas recirculation: thermodynamic
reasons for the increases of efficiency. SAE Technical Paper 2013-01-0266 (2013)
14. Hitting the gas on hydrogen tech for commercial trucks, sae, 2022-05-03 MATT WOLFE

Investigation of An Ammonia Diesel DualFuel Combustion Process on a Heavy-Duty
Single Cylinder Research Engine for the
Development of Suitable Simulation Tools
for Maritime Applications
Till Mante1(*), Sascha Prehn1, Martin Theile2, Lars Seidel3,
Laura Mestre3, Bert Buchholz1, and Fabian Mauss4
1

LKV—University of Rostock, Rostock, Germany
{till.mante,sascha.prehn,
bert.buchholz}@uni-rostock.de
2 FVTR GmbH, Rostock, Germany
martin.theile@fvtr.de
3 LOGE Deutschland GmbH, Cottbus, Germany
{lars.seidel,laura.mestre}@logesoft.com
4 BTU Cottbus-Senftenberg, Cottbus, Germany
fmauss@b-tu.de

Abstract.  This paper discusses the adaption of a single cylinder research
engine for a retrofit application with an ammonia diesel dual-fuel combustion
process and the build of an ammonia fuel system. The gaseous ammonia will
be injected in the air intake pipe and the premixed ammonia air mixture will
enter the combustion chamber. The diesel injection is carried out via a highpressure common rail system. All relevant parameters can be freely adjusted
via a freely programmable control unit. With the help of experimental data
from a single cylinder research engine at the chair of piston machines and
internal combustion engines of the University of Rostock (LKV), a dualfuel combustion model based on detailed chemistry will be developed and
optimized. This model will be integrated in a full research engine model, which
ensures the best possible representation of the real engine. The combustion
model is being developed by LOGE Deutschland GmbH. The full research
engine model is developed by FVTR GmbH. The analysis of the combustion
process starts with pure diesel operating points and is successively substituted
by ammonia in the course of the measurement campaigns. Both the combustion
characteristics are relevant, as they significantly influence the resulting performance and engine operation, as well as the exhaust emissions, as the carbon
emissions can be reduced, but the nitrogen oxides and ammonia slip increase
significantly in relevance due to the ammonia. The results obtained will be
used to derive initial recommendations for action and to estimate the potential
for application in the inland waterway shipping. In addition, the d­ evelopment

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 24–39, 2023.
https://doi.org/10.1007/978-3-658-41477-1_3

Investigation of An Ammonia Diesel Dual-Fuel …    25
of the systematic simulation tools covers a broad spectrum of research
questions and aims to increase the efficiency of the necessary R&D.
Keywords:  Combustion engines with alternative fuels · Combustion processes ·
Fuel and air systems · Engine and system optimization

1 Introduction
Ammonia is considered one of the best-known chemical substances. To use this
knowledge and make a positive contribution to the goal of climate neutrality,
ammonia can also be considered as a fuel. Looking at the sum formula NH3, it
quickly becomes clear that this potential fuel does not contain any carbon atoms
and therefore does not form any carbon emissions during combustion. In general,
the use of regeneratively produced carbon-containing fuels (e.g. synthetic methane,
e-methanol, synthetic paraffinic fuels, synthetic OME, etc.) can achieve complete
climate neutrality, since a closed CO2 cycle exists during the production of the
fuel. With a time perspective beyond 2030, however, the availability of CO2 for the
synthesis of the above-mentioned fuels could become problematic, especially since
the separation of CO2 from the atmospheric air is energetically costly. Here, ammonia
as a nitrogen-based hydrogen carrier offers a potential solution. By switching from
carbon-based fuels to the nitrogen-based fuel ammonia, emissions of unburned
hydrocarbons (including methane slip) and particulates can assumedly be reduced to
virtually zero. Thus, the potential of a “zero carbon emissions” ship exists, whereby
local as well as global (e.g. CO2 emissions) environmentally harmful effects are
reduced or completely prevented. Furthermore, the production and infrastructure
is already available for a century (Haber-Bosch), e.g. in the agricultural use as
fertilizer[1]. It is not only these properties that make the substance attractive as
a power-to-X fuel. In addition to alternative combustion processes such as spark
ignition or diffusive processes, there is still the potential for application as a dualfuel combustion process in combination with conventional fuels and thus a possible
retrofit perspective for shipping. Figure 1 shows all possible combustion processes
with ammonia. The challenges in handling and using ammonia should not be
underestimated. The main problems are the high toxicity, the challenging combustion
properties and the resulting nitrogen compounds in the exhaust gas.

Fig. 1.  Combustion processes with ammonia

26    T. Mante et al.

As a retrofit variant in shipping, the combination with diesel fuel (e.g. NH3-PFI
with pilot diesel injection) offers itself. In order to fundamentally investigate such a
combustion process, the project “AmmoniakMotor”, which is funded by the BMWK
(Federal Ministry for Economic Affairs and Climate Action)/PTJ and carried out
by the project partners FVTR GmbH (process simulation), LOGE Deutschland
GmbH (reaction kinetics in cylinder combustion model) and LKV—University of
Rostock (NH3 research engine/test bench measurements), exists. The collaboration
is displayed in Fig. 2. In the selected combustion process, the ammonia is injected
in gaseous form into the intake air and brought into the combustion chamber as an
ammonia-air mixture. For ignition & combustion support, diesel fuel is injected
into the combustion chamber at high pressure via a common rail system. The LKV
is converting an existing diesel single-cylinder research engine (SCE) to dual-fuel
operation with ammonia.
The FVTR is focused on the thermodynamic analysis of the incylinder process by
using one the one hand sophisticated pressure trace analysis tools, which are capable
of handling different ammonia-diesel ratios. On the other hand, a GT-Power model of
the research engine will be created and combined with a stochastic reactor combustion
model from LOGE.
Within the project LOGE developed a detailed chemistry model for the
combustion of diesel fuel and ammonia including the formation and reduction of
emissions such as soot and nitrous species. The detailed model was reduced to
lower computational time. The combustion developed model is used to model the incylinder combustion using LOGE’s stochastic reactor model (SRM). In a later phase
of the project also a tabulated combustion modelling approach will be applied to reach
even faster CPU times. Test bench data from LKV, in particular measured emissions,
will be used in later phase of the project to improve the chemistry model.

Fig. 2.  AmmoniakMotor—workflow

Investigation of An Ammonia Diesel Dual-Fuel …    27

2 Ammonia as an Alternative Fuel
The major challenge in the use of ammonia as a fuel is its high toxicity and the
associated safety requirements. Another characteristic is the pungent odor, which
can be perceived from approx. 5 ppm in the air. Further contamination should be
prevented as far as possible by early detection but should be avoided altogether due
to the high toxicity. This requires sensitive and accurate safety technology. Affected
rooms should be equipped with gas detectors and a gas warning system and should
be permanently in operation if ammonia is present in said room. Persons should
enter this room only with a suitable gas mask and under the supervision of another
colleague, and all pipes should be free of ammonia during work. In the event of an
emergency, sufficient ventilation should be provided to ensure that the entire amount
of ammonia can be removed fast from the room. All these points make it challenging
to use the fuel on ships and require trained personnel and maximum focus [2].
Furthermore, ammonia shows disadvantages in its combustion behavior compared
to conventional fuels and thus cannot be replaced 1:1. The auto-ignition temperature
of 630 °C is significantly higher than that of diesel with values between 180–
320 °C [3], which means that a diesel-like combustion process with auto-ignition
and diffusion-controlled combustion is classified as technically complex to implement and would require very high compression ratios. The laminar flame velocity,
which is relevant as an essential parameter for premixed combustion processes, is
with 0.07 m/s [4] at standard conditions significantly lower than gasoline (approx.
0.4 m/s) or methane (approx. 0.37 m/s). This creates challenges in achieving complete
conversion of the ammonia in the combustion chamber. Although the adiabatic flame
temperature at stoichiometric conditions of approx. 1800 °C is lower than for many
other fuels (comparison diesel: 2327 °C, gasoline: 2307, methane: 1950 °C), it is
highly unlikely that global lean burn of the order of e.g. today’s natural gas engines
(lambda significantly higher than 2) can occur in premixed combustion, which means
that the flame temperature is expected to be higher in comparison. This may have a
negative impact on the generation of thermal nitrogen oxides.
Even though ammonia has the potential to be a CO2 free fuel, new challenges in
exhaust gas aftertreatment arise with nitrous oxide (N2O) and the resulting ammonia
slip. With a significantly higher greenhouse gas factor than carbon dioxide, nitrous
oxide must be viewed very critically. Initial measurements show that nitrous oxide
emissions from ammonia combustion increase simultaneously with the ammonia slip.
Another important aspect is corrosion and the interaction with other materials (for
e.g. pipes, valves, measurement systems, injectors). There are not that many possible
combinations with either liquid and gaseous ammonia and to avoid leaks after a short
time, the systems have to be thoroughly planned. Ammonia and its corrosion behavior
on different materials is not widely researched yet.

28    T. Mante et al.

3 Test Bench Setup, Fuel System & Simulation Model
Due to its high toxicity, the safe handling of ammonia is a decisive criterion for its use
as a fuel. Especially in the process of basic research, the fuel system must be designed
carefully along with sufficient the safety technology.
3.1 Test Bench Setup
At the LKV in Rostock, a SCE is being converted to dual-fuel operation with
ammonia and diesel. Figure 3 shows the test bench in its current state. A validated
diesel operation forms the basis for the investigation of the dual-fuel concept. The test
bench has a fully conditionable air system to best simulate the use of a turbocharger,
diesel fuel system (free variation of injection pressures, times, temperatures &
durations, common rail), cooling water & oil system as well as an open access
research engine control unit with all possible degrees of freedom. The recently built
ammonia fuel system is also variable in injection pressure & quantity.

1 cylinder, 4 stroke diesel
External intake air conditioning
Nominal speed
Nominal power
Bore
Stroke
Cylinder volume
Compression ratio

2300 rpm
45 kW
110 mm
136 mm
1,29l
18:1

Fig. 3.  Test bench setup

The ammonia used is stored in a gas bottle. To ensure safety when the gas bottle
is open, it is located in a permanently vented gas bottle cabinet in the test bench
room. The entire fuel system was also installed in this safety storage cabinet so that
possible leaks do not contaminate the test bench room. A gas sensor in the exhaust air
of the cabinet detects possible ammonia leaks. The safety technology also includes a
sufficient powerful exhaust air system, additional gas sensors, a gas warning control
unit, gas masks, an emergency shower and an oxygen respirator for emergencies. The
procedures for changing the gas bottle, opening and closing the bottle or shutting
down the test bench at the end of the working day are firmly regulated.

Investigation of An Ammonia Diesel Dual-Fuel …    29

3.2 Fuel System
The fuel system is shown schematically in Fig. 4. The ammonia is taken in
liquid form, processed and evaporated after measuring the mass flow with a
Coriolis measurement system. At this point, the ammonia is still liquid and can, if
necessary, be transported into the engine in this state. A pressure reducer regulates
the evaporated gaseous ammonia to the desired pressure and a proportional valve/
special gas injectors control the mass of ammonia introduced into the engine intake
pipe. Furthermore, there are several ball valves closed in the normal state, a safety
valve, several purge ports for nitrogen and a throttle for safety empty of pipes with
almost odorless release in the ambient air. The introduction of the gaseous ammonia
into the engine intake is controlled by a proportional valve. This is to be replaced
in the further course of the project by gas injectors, which realize a more realistic
application.

Fig. 4.  Ammonia fuel system

Following successful setup and commissioning, the aim in the current stage
of the project is to promote detailed knowledge about the combustion process with
successively increase of the diesel substitution rates. In addition to the internal
combustion processes, there is a focus on exhaust gas emissions, wear and lubricant
effects as well. The latter is investigated in the associated lubricant laboratory of the
LKV, which analyzes and evaluates oil samples using modern equipment and can very
quickly detect signs of wear and change of lubrication characteristics.

30    T. Mante et al.

For the further investigation, the simulation model of the project partner LOGE
Deutschland GmbH will be validated with the measurement data from the SCE. This
model is described below. First modelling results are discussed in the next section.
3.3 Simulation Model
Stochastic Reactor Model
The SRM is used in this study to model the combustion process. The SRM accounts
for turbulence and inhomogeneity effects during the combustion. In the calculation,
the combustion chamber homogeneity is represented by a statistical homogeneity and
the physical quantities are described by a probability density function (pdf) [5–9]. In
the SRM the mass in the reactor is divided stochastically into virtual packages called
particles. Each particle has its own chemical composition, temperature, and mass,
and can mix with other particles and exchange heat with the walls [7]. The transport
Eq. 1 of the pdf contains on the left side the accumulation term and various source
terms Qi (ψ). In the source terms, the piston work, convective heat transfer, chemical
reactions, direct injection and vaporization are included. Therein, φ is a vector of
random variables, ψ is its realization in the sample ψ -space and t is the time. The
term on the right side P2 describes the molecular mixing due to turbulence and is in
a non-closed form. A modified Euclidean Minimum Spanning Tree (EMST) mixing
model is used to close the P2 term [10, 11], which accounts for locality in the particle
mixing process for mixture fraction. As a result, only neighboring particles in the
mixture fraction space can mix with each other.


∂ 
∂
Fφ (ψ, t) +
Qi (ψ)Fφ (ψ, t) = P2 Fφ (ψ, t)
∂t
∂ψi

(1)

The SRM incorporates a phenomenological turbulence model to calculate the scalar
mixing time τφ [10, 12]. The change of turbulent kinetic energy k in Eq. (2) is
calculated including on the right side, the first term for density changes, the second
term for dissipation ε, the third term for squish flow, the fourth term for direct
injection and the last term for swirl flow.


2 
3
dV
k
k
2
dk
sq 
cyl

= − ·
− ε + Csq ·
·
dϕ
3 Vcyl
dt
dt

TDC<ϕ<EVO

dkinj
+ Csw ·
+ Cinj ·
dt



3
cm
 1
·
l
6·n

(2)
In that equation, ϕ is the crank angle, Vcyl is the instantaneous cylinder volume, t is
the time in seconds, ksq is the turbulent kinetic energy from the squish flow, kinj is
the turbulent kinetic energy from the direct injection, cm is the mean piston speed,
l is the integral length scale and n is the engine speed. The dissipation in Eq. (3) is
derived from the turbulent kinetic energy and length scale l , which is based on the
instantaneous cylinder volume.

k 3/2
l
Finally, the scalar mixing time is calculated by Eq. (4):
ε = Cdiss ·

(3)

Investigation of An Ammonia Diesel Dual-Fuel …    31

τφ = Cτ ·

k
ε

(4)

The factors Cdiss, Csq, Cinj, Csw and Cτ need to be calibrated for the respective engine.
The calibration was done using LOGEengine expert system with the experimental
pressure trace of the operating point with pure diesel as target. As good match
between experimental pressure trace and model prediction was reached (Fig. 8).
Major parameters for the SRM modelling are listed in the table below (Table 1):

Table 1.  SRM modelling parameters
Parameters
Number of particles
Time step size [°CA]
Number of cycles
Mixing time coefficient CΦ
Dissipation factor Cε
Particle mixing model
Heat transfer model
Turbulence model
Start of simulation
End of simulation

Value
200
0.5
5
14.29
1
Enthalpy conditioned EMST [6]
Woschni
K–ε model [6]
Intake valve closing (IVC)
Exhaust valve opening (EVO)

Chemistry model
The commercial diesel used in the experiment was modelled by a 3 component
surrogate composed of 17.2 vol% 1-Methylnapathalene to represent the aromatic
fraction, 76.1% n-Decane to represent to n-alkane fraction and 6.7 vol% methyl
decanoate to model the FAME content. A comparison of the properties of the
commercial diesel and the surrogate is shown in the table below:

Table 2.  Comparison of commercial diesel & surrogate
Property
Density at 20 °C [kg/m3]
Lower heating value [MJ/kg]
Lower heating value [MJ/l]
Mean molecular weight [g/Mol]
C/H/O ratio (atomic)
Mass% C/H/O
Cetane number
FAME % (v/v)

Commercial EN 590
835.7
42.71
38.08
192
13.7/22.3/0.09
85.94/13.33/0.74
52.9
6.7

Surrogate
786.1
42.53
33.5
140
10.26/19.32/0.12
85.35/13.37/1.28
52.9
6.7

32    T. Mante et al.

The chemistry model for the fuel oxidation and emission formation is based on
the current LOGEfuel reaction model. The model for methane and methyl decanoate
was adapted from [15, 16] and updated for better match with experiments for ignition
delay time and emission formation. A comparison of predicted and measured ignition
delay time for n-Decane is shown in Fig. 5. The sub-models for ammonia oxidation
and NOx chemistry are adapted from [14]. The detailed reaction mechanism consisted
of 1307 species and was than reduced to 491 species using the technique described in
[16, 17] to reduce computation time.

Fig. 5.  Comparison of predicted and measured ignition delay times of n-Decane/air mixtures
in shock tubes at different pressures

4 First Simulation & Measurement Results
In the course of the project, as said, the project partner LOGE will develop a
combustion model for dual-fuel combustion, based on a stochastic reactor model, with
diesel and ammonia and validate it on the basis of the measured data from the LKV
test engine in addition to published experiments for ignition delay time, flame speed
and speciation. In the following, both initial simulation and measurement results and
thus detailed validation data will be evaluated. In contrast to the simulation, there are
limits to the test bench and measurement operation, e.g. due to emission values or
cylinder peak pressures.

Investigation of An Ammonia Diesel Dual-Fuel …    33

4.1 Adding Gaseous Ammonia & Investigations on the Influence of Early
Diesel Pilot Injection on Ammonia Combustion
The first aspect to be investigated is the addition of ammonia to a diesel-only
operating point.
For the analysis, a medium load engine operating point was selected, which is
shown below in Table 3. These parameters are constant during the measurement.
Table 3.  Reference operating point “diesel only” (PI… pilot injection, MI… main injection)
Engine speed
1500 rpm

Intake air pressure
1.91 bar
λ = 2.0

Diesel rail pressure
900 bar

PI 1
20°CA BTD
0.2 ms

MI 1
14°CA BTD
1.6 ms

Figure 6 shows the cylinder pressure and heat release rate for this operating
point in pure diesel mode. Based from this, gaseous ammonia is added to the intake
air (approx. 3.7 kg/h, which corresponds to an energy share of the total combustion
of approx. 25%). This results in an increase of engine torque from 115 Nm to 145
Nm. The increased amount of fuel can also be seen in the cylinder pressure & heat
release rate. Due to ammonia slip caused by valve overlap and incomplete ammonia
combustion, NH3 emissions of 3270 ppm can be detected in the exhaust gas.
To reduce the ammonia emission, another diesel pilot injection is added at 60°CA
BTD (0.2 ms, yellow graph). This operating point is shown in Table 4.
Table 4.  
Operating point with early diesel pilot injection (also for comparison to the
simulation later on)
Engine speed
1500 rpm

Intake air
pressure
1.91 bar

Diesel rail
pressure
900 bar

Early PI

PI 1

MI 1

60°CA BTD 20°CA BTD 14°CA BTD
0.2 ms
0.2 ms
1.6 ms

The injection duration is very short to inject only a small amount of diesel within
the combustion chamber in order to get a high premixed diesel share that initialize the
ignition of the homogeneous ammonia/air mixture in the combustion chamber. Due
to the very early diesel pilot injection a higher volume of the combustion chamber
is captured with diesel ignition energy which results in a more efficient ignition and
combustion behavior of the ammonia. Compared to the engine operation without the
early pilot injection the ammonia concentration in the exhaust is significantly reduced
to 2500 ppm (−23%). For the further series of measurements, the early pilot injection
is consequently used to reduce the ammonia slip. Furthermore, a slight increase of
the diesel amount in the early pre-injection also has a positive effect on reducing NH3
emissions and is going to be investigated in the near future.

34    T. Mante et al.

Fig. 6.  Addition of ammonia and impact of an early diesel pilot

The operating point with the early diesel pilot injection and the additional
ammonia (yellow graph in Fig. 6) has been chosen for the first comparison of model
prediction with experiments. The second point used here is the same point without the
addition of ammonia (Table 4). Therefore, they only differ in the amount of premixed
NH3. The base line case fueled only by diesel injection was used for training of the
turbulence model. The second point was used to test the model prediction for addition
of premixed ammonia. The ammonia was assumed to be perfectly premixed with the
air after intake valve closes. For both cases no EGR (there is probably a small amount
of internal EGR) and dry air was assumed. Table 5 compares the measured peak
pressure, max. rate of heat release and raw emissions with the model prediction. No
tuning of the combustion model was performed to match the emissions.
For both operating points the peak pressure is well matched (see Fig. 7). Also,
the trend that additional ammonia increases the peak ROHR is matched, though a
difference in absolute numbers is seen due to discretization in the model of 0.5 °CA.
It was found that the model predicts a complete consumption of the ammonia
while in the experiments an ammonia slip of 2457 ppm was observed (about 5% of
ammonia trapped at intake valve closes). For major emissions such as O2 and H2O the
trends are also well kept while absolute numbers differ from the experiment which
will be discussed at the end of this section.
The predicted formation of NO and NO2 are remarkable close to the measured
values considering the NO model tuning was done. Model and experiment predict
a reduction of NOX emissions for the operating point with additional NH3, even
though the mean in cylinder temperature is much higher (about 200 K at exhaust
valve opening). Like the experiment the model predicts a significant amount of N2O
at exhaust valve opens. It can be seen in Fig. 7 that engine out emissions of N2O are
a result of formation and reduction process. Formation is dominant until the peak
pressure is reached, and all fuel is consumed, followed by a rapid consumption to about
20% of its peak value. The underlying chemistry will be studied in a later publication.

Investigation of An Ammonia Diesel Dual-Fuel …    35

Fig. 7.  Comparison in pressure curve and N2O emissions for simulation and experiment
(experimental N2O represented as dot)

As mentioned before these are the first modelling attempts for diesel/ammonia
dual-fuel combustion in this project and there is a discrepancy for the major species
which is higher than the expected experimental uncertainty of the measurement
equipment. We expect that the model assumption of dry air as well as no EGR play
an important role for underestimation of major species. In addition, the complete
consumption of ammonia directly leads to a higher water content. The reason for the
complete NH3 consumption in the model is currently under investigation. Based on
the now available experiments the combustion chemistry—in particular the ammonia/
hydrocarbon interaction—will be revised. We also will study in which extend the
crevice volume needs to be considered to capture the ammonia emissions (see
discussions in [13]).

Table 5.  Comparison of experimental & simulation results

pmax
ROHR
O2
NH3
NO
NO2
N2O
CO
H2O

bar
J/°CA
ppm
ppm
ppm
ppm
ppm
ppm
ppm

Diesel only + early pilot
Exp
149,91
136,46
106,088
0
1985
86
0
125
76,891

Sim
150,94
165,26
102,187
0
2504
74
0
13
64,266

Diesel only + early pilot + ammonia
Exp
Sim
165,54
167,82
190,71
234,61
60,250
57,409
2457
0
1757
2449
37
46
60
40
235
1
131,940
131,633

36    T. Mante et al.

4.2 Investigation of Substitution Potential
To give an outlook to the substitution potential, a new operating point is defined. This
differs by reducing to one early pilot-injection and one main injection of diesel. The
injection durations have also been adjusted to the operation point before. The injection
and engine parameters are listed in Table 6.
Table 6.  Starting operating point for substitution investigation
Engine speed
1500 rpm

Intake air pressure
1.91 bar

Diesel rail pressure
900 bar

Early PI
60 °CA BTD
0.4 ms

MI 1
10 °CA BTD
1.2 ms

Starting from pure diesel operation, ammonia is now injected to the intake air.
With increasing amount of ammonia, the diesel amount is being reduced for constant
total fuel energy in the combustion chamber. Figure 8 shows the burning duration and
NH3 concentration in the exhaust gas over the substitution rate of diesel. It can be
derived that an increase of ammonia amount in the combustion chamber results in an
increased ammonia slip. At the same time, the burning duration is decreasing, which
indicates an increased burning speed for higher ammonia mass in the combustion
chamber.

Fig. 8.  Diesel substitution and the effects on torque and ammonia slip

Figure 9 shows the cylinder pressure curves and corresponding rate of heat release
calculated by an in-house thermodynamic analysis tool. Compared to the pure diesel
mode a substitution rate of 10% leads to an increased ignition delay with lower heat
release rates before the main injection event. These effects are even stronger at 30%
substitution rate. This leads to the assumption, that the lower reactivity and lower
oxygen partial pressure has negative effect on the pilot injection combustion, which
should be considered when dealing with higher substitution rates by adjusting pilot
injection duration and timing.

Investigation of An Ammonia Diesel Dual-Fuel …    37

After the first pilot injection combustion, there is a second event of moderate heat
release rate right before the main injection event. This event is still under investigation
and is believed to result from unburned fuel from the pilot injection, which undergoes
a different evaporation process (maybe interaction with the piston or liner).
The influence of the ammonia on the main combustion event is rather complex.
There is a slight increase in ignition delay with increasing substitution rate, which
can be explained due to the different in-cylinder conditions due to the pilot injection
combustion. There is a slight increase in the maximum heat release rate at 30%
substitution rate. This behavior shows the complex nature of ammonia-diesel dualfuel combustion modes and underlines the need for further investigations.
The effect of the increased burning speed can also be seen in the sum heat release
in Fig. 9, where the increase of the heat release at 30% substitution rate starts the
latest, but increases faster than the points with less ammonia.

Fig. 9.  Effect of increasing substitution rates on cylinder pressure, rate of heat release and
cumulated heat release

38    T. Mante et al.

From this point with 30% energetic ammonia in the combustion, different parameters (e.g. share of diesel amount, diesel injection timing, and lambda) will be
variated to reduce ammonia slip and improve the combustion process. Then, the
energetic ammonia amount can be increased further and the potential of the dual-fuel
combustion process can be clearly seen.

5 Summary
To meet the challenge of climate change, the use of alternative fuels is essential. This
also includes existing fleets of ship with operating lives of more than 25 years. Here,
retrofit concepts can make a significant contribution to achieve the climate targets.
Using ammonia, a carbon-free energy source, requires well-designed safety technology, a reliable fuel system, and efficient combustion. These challenges are high, but
if successfully implemented, a large part of the maritime industry can thus capitalize
on them and paves the way for carbon-free shipping. The installation of the ammonia
fuel system at LKV and the initial investigations of a dual-fuel concept on a SCE have
shown that the challenges can be met and that ammonia is a viable candidate as a
future fuel. Furthermore, the combustion modelling of LOGE Deutschland GmbH
allows this novel combustion process to be investigated and developed beyond the
limits of experimental possibilities.
The greatest challenges in the use of ammonia as fuel are N2O formation,
ammonia slip and new challenges in the areas of exhaust gas aftertreatment due to
very different gas composition as well as durability & corrosion.

References
1. Grindberg et al.: Stickstoffbasierte Kraftstoffe: eine „Power-to-Fuel-to-Power“-Analyse,
vol. 128. https://onlinelibrary.wiley.com/doi/10.1002/ange.201510618 (2016)
2. AirLiquide: Sicherheitsdatenblatt Ammoniak (2019)
3. Gupta B. R.: Hydrogen Fuel Production, Transport and Storage. CRC Press, Boca Raton (2008)
4. Theile, M., Drescher, M., Reska, M., Dahms, F., Swiderski, E.: Development of a GHGneutral combustion concept exemplified by methanol. In: 7th Rostock Large Engine
Symposium, Rostock, Germany (2022)
5. Pasternak, M.: Simulation of the diesel engine combustion process using the stochastic
reactor model. BTU Cottbus, Senftenberg PhD Thesis (2016)
6. Kraft, M.: Stochastic modeling of turbulent reacting flow in chemical engineering. VDI Verlag, Düsseldorf (1998)
7. Tuner, M.: Stochastic reactor models for engine simulations. Lund: PhD Thesis (2008)
8. Pope, S.B.: Pdf methods for turbulent reactive flows. Prog. Energy Combust. Sci. 11(2),
119–192 (1985)
9. Harworth, D.: Progress in probability density function methods for turbulent reacting flows.
Prog. Energy Combust. Sci. 36(2), 168–259 (2010)
10. Franken, T., Sommerhoff, A., Willems, W., Matrisciano, A., et al.: Advanced predictive
diesel combustion simulation using turbulence model and stochastic reactor model. SAE
Technical Paper 2017-01-0516. https://doi.org/10.4271/2017-01-0516 (2017)

Investigation of An Ammonia Diesel Dual-Fuel …    39
11. Bernard, G., Scaife, M., Bhave, A., Ooi, D., et al.: Application of the SRM engine suite over
the entire load speedoperation of a U.S. EPA tier 4 capable IC engine. SAE Technical Paper
2016-01-0571. https://doi.org/10.4271/2016-01-0571 (2016)
12. Kozuch, P.: Phenomenological model for a combined nitric oxide and soot emission
calculation in DI diesel engines. PhD Thesis, Stuttgart (2004)
13. Franken, T., Matrisciano, A., Sari, R., Fogué Robles, Á., et al.: Modeling of reactivity
controlled compression ignition combustion using a stochastic reactor model coupled with
detailed chemistry. SAE Technical Paper 2021-24-0014. https://doi.org/10.4271/2021-240014 (2021)
14. Shrestha, K.P., Giri, B.R., Elbaz, A.M., Issayev, G., Roberts, W.L., Seidel, L., Maus, F.: A
detailed chemical insights into the kinetics of diethyl ether enhancing ammonia combustion
and the importance of NOx recycling mechanism. A Farooq Fuel Communications (2022)
15. Wang, X.: Kinetic mechanism of surrogates for biodiesel. Ph.D. Thesis Cottbus (2018)
16. Matrisciano, A., Seidel, L., Mauss, F.: An a priori thermodynamic data analysis based
chemical lumping method for the reduction of large and multi-component chemical kinetic
mechanisms. Int. J. Chem. Kinet. 54, 523–540 (2022)
17. Seidel, L., Netzer, C., Hilbig, M., Mauss, F., Klauer, C., Pasternak, M., Matrisciano, A.:
Systematic reduction of detailed chemical reaction mechanisms for engine applications. J.
Eng. Gas Turbines Power 139(9), 091701 (2017)

H2 ICE DI Multicylinder Engine Tests
for Thermodynamics and Component
Development
Simon Schneider1(*), Christian Trabold2, Thomas Friedrich1,
Florian Mayer1, Fabian Weller1, and Roman Stiehl1
1

MAHLE International GmbH, Stuttgart, Germany
{simon.schneider,thomas.friedrich,florian.b.mayer,
fabian.weller,roman.stiehl}@mahle.com
2 MAHLE GmbH, Stuttgart, Germany
christian.trabold@mahle.com

Abstract.  For commercial vehicles, the hydrogen internal combustion engine
(H2 ICE) is a propulsion variant that enables fast defossilisation in the HD
vehicle segment. MAHLE is intensively investigating this kind of engines
and develops components for this application. A research engine was built
and operated with a combustion system with direct injection of hydrogen (H2
ICE DI) whereas experience is used from operating the same aggregate in port
fuel injection configuration (H2 ICE PFI) in the preceding project. The results
demonstrate the potentials of the DI combustion system in terms of engine load
and efficiency. Furthermore, a direct comparison to the PFI combustion system
is done. Boundary condition is the realization on a typical commercial vehicle
engine and the use of components that were developed for a future series use
of H2 ICE engines. MAHLE deducts important insight related to requirements
for the components from these investigations and uses the opportunity to
test further subsystems for this application, e.g. in the field of crank case
ventilation.
Keywords:  Hydrogen combustion engine · Hydrogen direct injection · CO2
neutral truck

1 Hydrogen Combustion Systems for Trucks
and Configuration of the Research Engine
In order to achieve climate neutral transportation, fossil fuels have to be replaced by
renewable energy carriers. Hydrogen is a very promising candidate for applications
in which high energy density and fast refueling is required. Onboard the vehicle, it
can be used in a hydrogen fuel cell or in an internal combustion engine. Both technologies are currently investigated and developed by many companies. DAF [1], MAN
[2] and Cummins [3] are developing ICE engines and vehicles for on- and offroad
© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 40–52, 2023.
https://doi.org/10.1007/978-3-658-41477-1_4

H2 ICE DI multicylinder engine tests …    41

applications. DEUTZ has announced the SOP of their H2 ICE focused on electricity
generation in 2024 [4]. In parallel, suppliers like BOSCH are investigating the technology to understand the system and its impact on their components [5].
With the same motivation, MAHLE has converted a 13l HD Diesel engine to be
operated on hydrogen in our in-house test facilities. For this engine, a dedicated Power
Cell Unit (PCU) was developed [6] and the engine was successfully operated with Port
Fuel Injection (PFI) [7]. As a next development stage, the engine was further adapted to
be operated with Direct Injection (DI). This mitigates the risk of backfire and provides
higher air flow rates, giving the potential for higher BMEP and efficiency. However, due
to the shorter time between injection and combustion, proper mixture formation is crucial
[8]. The setup and the results of the engine test at MAHLE are presented in this paper.

2 Design of DI Research Engine and Comparison to PFI
Setup
2.1 DI Configuration and Conceptual Design Choices
The existing MAHLE H2 ICE research engine was adapted to enable DI combustion.
The cylinder head now contains an insert with central injector position and near central
spark plug position at a slight angle and a distance of approx. 15 mm to the center.
This promises a maximum of symmetry for mixture preparation and combustion. CFD
simulations for several operating points were performed to ensure sufficient mixture
quality. Spark plugs identical to those from the PFI testing were used with cold rating
and a J electrode. The configuration of the gas path was kept identical to the PFI setup.
Single runners for each cylinder are connected to a plenum. The charge motion of the

Fig. 1.  MAHLE H2 ICE research engine – view from exhaust side on valvetrain cover

42    S. Schneider et al.

engine is very low – there is no swirl and no tumble, identical to the Diesel baseline
engine. Some charge motion is generated from the fuel injection itself, depending on
injection timing. An impression of the engine is given in Fig. 1.
2.2 MAHLE Hydrogen Optimized Piston and Ring Set
The piston bowl shape was kept from the PFI testing. The top land was modified to
reduce gap volumes in this region. Low blow-by and low lube oil consumption are
main challenges for the power cell unit of hydrogen combustion engines. Thus,
the piston ring set is further improved compared to the setup presented in [7]. An
optimization loop was performed with data from the previous run and the measures
were integrated to the design. The DI ring set proved to enable even lower blow-by
flows. The configuration of the combustion chamber is shown in Fig. 2.

Fig. 2.  Combustion chamber configuration: DI with central injector position

2.3 Scavenged Crank Case Ventilation
The scavenging crank case ventilation with the MAHLE high pressure impactor
system was applied in an identical way as described in [7]. Besides keeping hydrogen concentrations under control, the scavenging ventilation also lowers the moisture
content within the crank case. This is important to avoid oil dilution and corrosion,
especially at engine operation at low loads and low ambient temperatures. The dew
point temperature at which condensation starts is shifted by 20 °C if scavenging is
applied.

H2 ICE DI multicylinder engine tests …    43

3 Testing Results and Potential of H2 ICE DI Combustion
3.1 Engine Power Compared: DI Setup vs. PFI Setup of 2021
With the research engine, extensive testing was performed. The main targets were
verification and validation of the components applied as well as the investigation of
the potential of DI combustion especially in comparison to the PFI baseline. When
operating the H2 ICE DI research test engine, two main modes of the combustion
systems can be distinguished by the choice of the injection timing (Fig. 3).

Pressure trace / bar

Injection
PFI mode

0

180

Injection
DI mode

360
540
Crank angle / deg a TDCF

Valve lift, Injection / -

• Injection during the intake stroke, so called “PFI Mode”
• Injection during the compression stroke, so called “DI Mode”

720

Fig. 3.  PFI mode and DI mode injection timing schematic

The PFI mode has as an advantageous longer time for mixture preparation. The DI
mode on the other hand enables higher volumetric efficiency, as the fuel is injected
into the cylinder after intake valve closing (IVC). This enables higher air-fuel ratios
with benefits for raw NOX emissions. A prerequisite for this mode is a sufficiently
high injection flow rate and resulting short injection duration, to complete injection
before the cylinder pressure is too high and to leave sufficient time for mixture
preparation.
The engine load achieved with DI in an initial power curve is shown in Fig. 4.
With the DI setup, a slightly lower power curve was achieved, ranging from 15.5 bar
BMEP at 1800 rpm to 18 bar BMEP at 1166 rpm. The difference to the PFI setup
was in the range of 1.0 to 2.5 bar BMEP. Limiting factors towards higher load were
different depending on the combustion phasing. Knocking prohibited higher loads
at early combustion phasing. For late combustion phasing misfires and combustion
instabilities were the limiting factors. Similar results have also been shown by Bunce
et al. [9].

44    S. Schneider et al.
25

20.5 bar
18.0 bar

20

BMEP / bar

PFI setup (2021)

18.5 bar
17.0 bar

Injectors in intake port

16.5 bar
15.5 bar

Intake stroke injection
With HP EGR

A85

15

A75

B75

C75

A50

B50

C50

DI setup (2022)

10

Injectors in DI position
Intake stroke injection &
compression stroke injection

5

A25

0
1000

B25

1200

C25

1400
1600
Speed / rpm

With HP EGR or wastegate

1800

2000

Fig. 4.  Power curve in previous project and initial power curve in current setup

The engine was operated across the map and the performance of the
thermodynamics was studied in detail. Fig. 5 shows a map of the selected results
in PFI mode. It can be seen that efficiency values >44% can be achieved with well
controlled NOX emissions.
Effective efficiency etae / %
20

NOX spec. / g/kWh
43

20

1.5

15

41

43
10

42

40

41

40

39

39
36

5
1200

1400
1600
Speed / rpm

36

1800

BMEP / bar

BMEP / bar

42

15
10

1.1

1.50

0.7

1.10
0.70
0.30

0.3
0.04

0.04

5
1200

1400
1600
Speed / rpm

1800

Fig. 5.  Engine maps for efficiency and NOX emissions in PFI mode of the DI engine

3.2 Analysis of PFI and DI Mode
Fig. 6 shows a comparison of the two modes at 1166 rpm and 10 bar BMEP. The airfuel ratio is 2.38 for the PFI mode and 2.81 for the DI mode, each being the maximum

H2 ICE DI multicylinder engine tests …    45

air mass flow of this operating point. The air-fuel ratio benefit of DI mode comes with
the disadvantage of increased pumping work (not shown in the diagram). The test was
chosen to understand the potential of each mode without controlling the air mass flow
by throttle, wastegate, or EGR. Therefore the different air-fuel ratios as results of the
different operation modes were accepted. A discussion of the advantages of throttle,
wastegate and EGR can be found in Sect. 3.3.

A50 Var MFB50 Lambda const PFI Mode
A50 Var MFB50 Lambda const DI Mode
burn duration 10-90 % (INDI)

covariance pmi avr. (INDI)

etae opt.

COV PMI - avr / %

BD1090 / °KW

34
30
26
22
18

NOX spec / g/kWh

46
etai,HD / %

3.0

misfire

1.5
0.0

indicated efficiency high press. cycle

44
42
40

misfire

4.5

0

5

10

15

MFB50 / °CA

20

25

1.2

NOX emissions engine out

0.9
0.6
0.3
0.0

0

5

10

15

20

25

MFB50 / °CA

Fig. 6.  A50 ignition timing variation at constant air-fuel ratio: DI mode (Lambda 2.81) vs. PFI
mode (Lambda 2.38). The range of optimum effective efficiency is marked

Both modes reach best efficiency at early MFB 50 timings of 4–5 °CA aTDCf.
Indicated efficiency of the DI mode with max. 46% is slightly above those of
PFI mode (45.5%) despite of the big difference in air-fuel ratio. NOX emissions
are acceptable for both modes, but better for DI mode with 0.2 g/kWh at optimum
efficiency compared to 0.75 g/kWh for PFI mode at same timing. This is mainly
caused by the leaner operation of the DI mode, which reduces NOX emissions. In

46    S. Schneider et al.

contrast, the PFI mode has shorter burn duration of approx. 20 °CA compared to DI
mode with 23 °CA which is a result of lower air-fuel ratio.
When phasing combustion towards late, combustion duration increases and
efficiency drops as expected for both modes. However, the late limit for MFB50 is
later for the DI mode. The reason for this couldn’t be perfectly clarified. Instability
increases, which is expressed in terms of the covariance of IMEP in our data, until
cyclic variations increase so that the operating point isn’t stable anymore. The early
MFB50 limit is defined by a steep increase in knock tendency for each mode. Tests
show that the knock limit of the DI mode is at a higher Lambda value than that of
PFI mode. This comes from a higher amount of inhomogeneities coming from shorter
mixing time.
Burn rates and split of losses
From Fig. 6, the two points with optimum efficiency are investigated via
thermodynamic analysis. Both are at MFB of 5° aTDCf and have an effective
efficiency from measured data of 42.3% each. Comparison of burn rates (Fig. 7)
shows typical curves with fast burn-off matching the duration known from the Indiset.

Normalized Burn Rate / 1/°CA

0.05

PFI mode, λ = 2.38, BD1090 = 19.7°CA
DI mode, λ = 2.81, BD1090 = 23.0°CA

0.04

N = 1166 rpm
BMEP = 10 bar
MFB50 = 5°CA

0.03

0.02

0.01

0.00
-30

0

30

60

90

Crank Angle / °CA

Fig. 7.  Burn rates from optimum points of A50 ignition timing variation

For closer comparison of the operating points, a split of losses within the pressure
trace analysis is performed (Fig. 8). DI mode has a slight advantage for real gas
composition as only air is inducted. PFI mode has an advantage from the shorter burn
duration from lower lambda. In contrast, DI mode has a similarly big advantage from

H2 ICE DI multicylinder engine tests …    47

better fluid properties of the leaner gas. Finally, PFI mode has an advantage as its gas
exchange work is lower due to the turbocharger configuration. In sum, the split of
losses confirms that both operating points have identical efficiency.

DIFF of losses / %

DI mode
benefit

Lambda difference:
higher

1.0%
0.5%

Sum

Real Gas Exchange

Ideal Gas Exchange

Compression

Expansion

Heat Transfer

Real Fluid Properties

Friction

Overall
εeff diff.
-0.16%

Gas exchange work:
impact of lower
air mass

Real Burn Rate

Incomplete Combustion

Real Composition

-1.0%

Lambda impact:
burn duration
3 CA shorter

Combustion Phasing

-0.5%

PFI mode
benefit

0.0%

Fig. 8.  Split of losses difference of optimum points of A50 ignition timing variation

Comparison of the advantages of the two modes
To better demonstrate the effects of both combustion modes, Fig. 9 shows a variation
of start of injection (SOI) for typical timings of PFI mode and DI mode at operating
point 1166 rpm, 15 bar BMEP at const MFB50 of 11 °CA.
The COV PMI plot shows that both modes can be calibrated for stable operation.
In PFI mode, SOI can be chosen rather soon after intake valve opening. Best performance for this mode is reached with SOI in first third of the intake stroke. Injecting
later, with SOI later than 300 °CA bTDCf, leads to stratification effects indicated by
a strong NOX emission increase. Possible injections timings in DI mode start with
intake valve closing and can be retarded until the cylinder pressure increases too much
to inject properly. With the switch to DI mode Lambda increases by around 0.4 points
(not shown). For DI mode with early SOI, the combustion is slowed down due to the
higher Lambda compared to PFI mode. The later the injection is done, the more the
combustion is sped up, until the burn duration reaches levels below those of the PFI
mode. This can be explained with turbulence from the injection. The later it takes
place, the more turbulence is being preserved until the combustion starts and is able to
support the flame propagation.

48    S. Schneider et al.
A75 SOI Var PFI mode
A75 SOI Var DI mode
burn duration 10-90 % (INDI)

3.0
COV PMI - avr / %

BD1090 / °KW

22
20
18

covariance pmi avr. (INDI)

2.5
2.0
1.5
1.0

NOX spec / g/kWh

etai,HD / %

indicated efficiency high pressure cycle

45

44
360

300

180

120

Start of inj. / °CA bTDCf

NOX emissions engine out

5
4
3
2
1
0
360

300

180

120

Start of inj. / °CA bTDCf

Fig. 9.  A75 SOI var at const. MFB50 DI mode vs. PFI mode

This effect can also be seen in the combustion efficiency of DI mode. It starts at an
identical level as in PFI mode despite higher Lambda and rises by 1% point (etai,HD
= 45.5% compared to 44.5%) with phasing SOI late. This sort of directly influencing
combustion via the injection is only applicable in DI mode. DI mode improves NOX
emissions down to levels of 0.4–0.5 g/kWh at the same time in this test. Both modes
have strengths that can be summed up as shown in Fig. 10.
PFI mode

DI mode

Better homogenization from
longer mixing time

Higher volumetric efficiency
enables leaner lambda
Charge motion benefit
from fuel injection
Option for combustion improvement at perfect choice
of timing and pressure

Lower gas exchange work
Robust calibration - wide
range of timings and injection
pressures

Fig. 10.  Benefits of each mode of the DI combustion system

H2 ICE DI multicylinder engine tests …    49

Comparison of DI operation to PFI Setup performance
Compared to PFI operation, the DI operation in PFI mode is rather similar in terms
of characteristics and performance. This is confirmed at an EGR sweep at operating
point A75, Fig. 11. The Covariance of PMI is lowest for PFI as it is the benchmark
for perfect mixing of the cylinder charge and slightly higher for DI setup in PFI mode.
Boost pressure and exhaust gas pressure are very similar but slightly higher for DI
setup in PFI mode. Flow losses at the intake valves are a little lower as only air is
transported and so the pressure ratio of the EGR shifts – and the EGR flow slightly
increases as well. Burn duration is identical at low EGR and follows the same trend,
with slightly longer duration for DI setup PFI mode at high EGR as the EGR rates are
slightly higher as described before. Overall, the DI setup in PFI mode is very similar
in characteristics of the combustion compared to the PFI setup.

PFI Setup A75 HP EGR var.
DI Setup PFI mode A75 HP EGR var.
3
COV PMI - avr / %

BD1090 / °KW

burn duration 10-90 % (INDI)

22
20
18

Increasing EGR rate

1

1300
P31 / mbar

P_LL_5 / mbar

1800

1700

1600
2.2
air-fuel ratio / -

2

0

air pressure
upstream cyl. 5 (rel.)

2.0

covariance pmi avr. (INDI)

2.4

exhaust gas pressure
upstream turbine cyl. 4-6 (rel.)

1200
1100
1000

2.0

2.2
air-fuel ratio / -

Fig. 11.  A75 EGR variation PFI setup compared to DI setup in PFI mode

2.4

50    S. Schneider et al.

3.3 Lambda Control
As described before, a measure is needed to control the air-fuel ratio to a certain
value across the engine map. At the research engine, three of those were investigated:
cooled high pressure exhaust gas recirculation (HP EGR), turbocharger wastegate
(WG) and use of a throttle flap. Below, the comparison of HP EGR and wastegate as
the two most favorable options is discussed in detail.
HP EGR vs. wastegate air-fuel ratio control
In Fig. 12, testing results are shown for a test comparing HP EGR and wastegate
operation at the 1166 rpm, 10 bar BMEP operating point discussed before. Both
methods of air-fuel ratio control are operated in DI mode at start of injection of
180 °CA bTDCf and a const. MFB50 timing of 8 °CA aTDCf.
As can be seen, combustion efficiency is best without wastegate or EGR. However,
both measures reduce pumping work with increasing setting, which leads to best
overall efficiency at mid setting of EGR and at moderate settings of the wastegate.
The DI mode wastegate variation reaches knock limit at approx. Lambda 2.2. Near
this limit, there is a steep increase in NOX emissions, so that only a small application
window can be utilized.
The DI mode high pressure EGR variation expands lambda range down to Lambda
<1.7. The EGR operation shows 1 to 5 °CA slower burn duration compared to WG
but nevertheless has higher combustion efficiency etai,HD. In numbers, this is 44.8%
compared to 44.1% at Lambda 2.35. The NOX emissions are well controlled for the
HP EGR variant and stay near 1 g/kWh even at the most extreme EGR setting. This
results in a wider calibration window for the EGR variant at Lambda 2.0–2.4 with
NOX values of 0.55–0.2 g/kWh. Stability is good for both variants over the sweep.
Summing up, HP EGR shows clear advantages over wastegate as means of airfuel ratio control. It allows for a wider applicable Lambda range at lower NOX levels,
while even providing a small fuel consumption advantage.
Influence of wall heat loss in DI combustion
The data in Fig. 12 shows a trend which appears counter-intuitive at first sight: As
expected, the burn duration decreased with the air-fuel ratio. However, this is not
translated into a benefit in efficiency. In contrast, the EGR operating points that
burn slower have better efficiency. This can be explained by two effects. Indicators
like energy balance and component temperatures show that an increase in wall heat
loss could occur under these conditions. This would reduce the benefit of the faster
combustion. However, a more detailed analysis is necessary to confirm this effect.
Another additional explanation is the influence of the EGR. It keeps the gas mass
flow high, even at low Lambdas, due to the recirculation of inert exhaust gas. This
keeps exhaust gas temperatures moderate and avoids high component temperatures at
low air mass flow conditions as with the wastegate. This classical EGR effect is an
important benefit for this combustion system at high engine loads. It makes EGR the
preferred method to control the air-fuel ratio, in interaction with the turbocharger.

H2 ICE DI multicylinder engine tests …    51

A50 Var WG DI Mode
A50 Var EGR DI Mode
covariance pmi avr. (INDI)

COV PMI - avr / %

burn duration 10-90 % (INDI)

BD1090 / °KW

24
21
18

etai,HD / %

46

1.5
1.2

indicated efficiency high
pressure cycle

NOX emissions engine out

NOX spec / g/kWh

15

1.8

45
44
43

3
2
1
0

etae / %

43

pmi LW avr. (INDI)

Baseline

42
41
40
1.7

Increasing EGR and WG
2.1
2.5
air-fuel ratio / -

2.9

pmi LW - avr / bar

brake efficiency

-0.1
-0.2
-0.3
1.7

2.1
2.5
air-fuel ratio / -

2.9

Fig. 12.  A50 wastegate var DI mode vs. EGR var DI mode at const. MFB50 timing

4 Summary and Conclusion
The H2 DI combustion engine shows comparable performance to the PFI engine.
Both modes of the DI combustion system – “PFI mode” and “DI mode” – reach
similar performance in terms of efficiency and emissions in the stationary testing of
this project. The combustion system is sensitive on proper adjustment of the preferred
Lambda range, raising the requirement of a variable element at the turbocharger. The
second critical item for the combustion is quality of homogenization that has to be
ensured in each mode.

52    S. Schneider et al.

In order to achieve good dynamics in transient operation, the higher Lambda of
the DI mode and the ability to use and switch between both modes seem essential, as
explained e.g. in [2].
Limiting to reaching higher load is combustion anomalies and increase of
covariance which narrows the application window of calibration parameters. On the
lower limit of Lambda, knocking is the limit. HP EGR showed to be stabilizing here
and helps to widen the range. On the lean side, inhomogeneity leads to limited range
compared to perfectly mixed cylinder charge as seen in PFI setup operation.
To overcome this, a certain amount of charge motion is beneficial to enable more
stable operation. As the engine investigated in this work does not have any significant
charge motion, it is expected that higher loads can be reached with other engines. An
additional enabler for better homogenization in DI mode is the optimization of the
injection flow pattern, which promises advances in terms of efficiency and load.
Optimizations on the ignition side, e.g. pre-chamber ignition like MAHLE Jet
Ignition, widen the lean burn area to enable a good calibration over the whole engine
map. As demonstrated by MAHLE Powertrain, this is a strong measure to increase the
combustion stability in a lean-burn H2 ICE [9].
Acknowledgements. We would like to thank Robert BOSCH GmbH for providing the injectors
for this testing campaign.

References
1. DAF Trucks N. V.: Hydrogen. DAF Trucks. (n. d.). https://www.daf.com/en/about-daf/
sustainability/alternative-fuels-and-drivelines/hydrogen. Accessed 5 Oct 2022.
2. Walter, L., Sommermann, A., Hyna, D., et al.: The H2 combustion engine – the forerunner of
a zero emissions future. In: 42nd International Vienna Motor Symposium, Vienna (2021)
3. Cummins Inc.: Cummins to reveal zero-carbon H2-ICE concept truck at IAA expo
powered by the B6.7H hydrogen engine. [press release]. https://www.cummins.com/news/
releases/2022/09/13/cummins-reveal-zero-carbon-h2-ice-concept-truck-iaa-expo-poweredb67h (2022, 13. September)
4. DEUTZ AG.: Der Wasserstoffmotor von DEUTZ ist reif für den Markt. [Press release].
https://www.deutz.com/media/pressemitteilungen/der-wasserstoffmotor-von-deutz-ist-reiffuer-den-markt (2021, August 21)
5. Kufferath, A., Schünemann, E., Krüger, M., et al.: H2 ICE powertrains for future on-road
mobility. In: 42nd International Vienna Motor Symposium, Vienna (2021)
6. Marlok, H., Trabold, C., Puck, A.: Engine Component Development for H2 Combustion
Engines, Heavy-Duty, On- and Off-Highway Engines. Rostock (2021)
7. Rieger, D., Mayer, F., Weller, F., Schneider, S., Stiehl, R.: Experimental investigation of a
hydrogen powered heavy-duty truck engine. In: Internationaler Motorenkongress, BadenBaden (2022)
8. Kapus, P., Raser, B., Arnberger, A., et al.: High efficiency hydrogen internal combustion
engine – carbon free powertrain for passenger car hybrids and commercial vehicle. In: 43rd
International Vienna Motor Symposium, Vienna (2022)
9. Bunce, M., Peters, N., et al.: Jet ignition as an enabling technology for stable, highly dilute
hydrogen combustion in off-road and heavy duty engines. In: Internationaler Motorenkongress, Springer (2021)

Assessment of a Direct-Injection,
Spark-Ignited, Hydrogen-Fuelled
Heavy-Duty Engine
John Hughes1(*), David Bennet1, Angela Loiudice1, Nicholas Coles1,
Trevor Downes1, Agam Saroop1, Richard Penning1, Lukáš Valenta1,
Peter Rabanser1, Jonathan Davis1, Jackson Harvey-Bush1,
Alvaro Concepcion Calero1, Richard Osborne1, Penny Atkins2,
Roger Allcorn2, and Nigel Fox2
1

Ricardo Shoreham Technical Centre, Ricardo UK Ltd, Shoreham by Sea, UK
{John.Hughes,David.Bennet,Angela.Loiudice,
Nicholas.Coles,Trevor.Downes,
Agam.Saroop,Richard.Penning,
Lukas.Valenta,Peter.Rabanser,
Jon.Davis,Jackson.Harvey-Bush}@ricardo.com
2 Advanced Engineering Centre, University of Brighton,
Brighton, UK
P.A.Atkins@brighton.ac.uk

Abstract.  Hydrogen-fuelled internal combustion engines (ICE) offer a zerocarbon fuel option for many ICE applications. As part of a global interest
to characterise and study the behavior of hydrogen fuel in existing ICE
applications Ricardo is collaborating with the University of Brighton to test
hydrogen fuel in a Ricardo designed Proteus single cylinder engine. The engine
is representative of a 13 litre Euro VI heavy duty (HD) production application
converted to run on hydrogen fuel with minimal changes. The engine is fitted
with a 35-bar direct injection (DI), hydrogen injector which gives improved
flexibility for injection strategies and greatly reduces the presence of hydrogen in the intake system compared to a PFI system. Steady-state testing was
carried out at an array of speed load points covering a large part of a typical
heavy duty (HD) drive cycle area. An extract of the test results are shared
and discussed in this paper. Lambda (λ) sweeps show the system is capable
of running out to values exceeding λ = 5.0, exhaust gas recirculation (EGR)
sweeps show over 40% EGR can be tolerated at given lambda conditions.
Abnormal combustion events present sizeable challenges at lower lambdas due
to very large knocking pressures and pre ignition risks. The lambda-threshold
where the majority of these events are observed increases with speed and load
thus narrowing the initial operating range of the engine prior to introducing
mitigating measures like cooled EGR. The basic impact of lambda, EGR,
injection and ignition timing sweeps are presented in this paper and show how

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 53–75, 2023.
https://doi.org/10.1007/978-3-658-41477-1_5

54    J. Hughes et al.
the system responds to the corresponding changes in specific heat capacity,
mixture preparation, and combustion phasing.
Keywords:  Hydrogen Engine · Hydrogen Combustion · Hydrogen Emissions ·
EGR · Proteus · ICE

1 Introduction
Ricardo and the University of Brighton are collaborating to study the effects of
running a typical mainstream HD engine on pure hydrogen. Hydrogen is increasingly
envisaged as a future energy carrier in the transition to a net zero carbon economy.
While using fuel cells to extract energy from hydrogen is seen by some as a long-term
goal, hydrogen fueled combustion engines provide definite advantages in the short to
medium term, as they offer lower cost solutions building on robust and proven technologies. They are also tolerant to fuel contamination and small particles, as well as
being easier to manage thermally.
In this research work Ricardo decided to investigate a Direct Injection (DI)
hydrogen system. This approach provides greater flexibility in injection strategies,
improves volumetric efficiency, allows lower boost pressures to be used, and reduces
the presence of hydrogen in the intake manifold, minimising the risk of backfire
compared to port fuelled injection (PFI) system. The work was started in a single
cylinder Proteus research engine, before being moved to a multi-cylinder engine. This
paper will focus on the results from the Proteus single cylinder engine.
Ricardo has a long history of designing state of the art single cylinder engines
(SCE) for research and development purposes and these are in use across the globe
both for commercial and academic research alike. Ricardo’s SCEs are available in
four families, namely: “Hydra” for light duty applications, “Proteus” for heavy duty
applications, “Atlas” for large engine applications and the largest called “Prometheus”
for power generation applications etc. A Proteus SCE was used for this study and
the chosen engine configuration has a modified Scania OC13 unit head and a Volvo
D13 bottom end, making the engine representative of a typical Euro VI production
application.

2 Key Engine Features and Test Facility Overview
As mentioned earlier testing was carried out on a Ricardo Proteus single cylinder
research engine at the University of Brighton. Table 1 gives an overview of how the
engine incorporates components representative of a Euro VI heavy-duty engine and
the conversion to run on hydrogen. A Scania unit head was modified to incorporate a
side DI H2 injector and a Kistler pressure transducer as per Fig. 1.

Assessment of a Direct-Injection, Spark-Ignited …    55
Table 1.  Engine Specification and Instrumentation
Item
Piston and Bottom End
Cylinder Head

Ignition System
Fuel Injection System
EGR
High-Speed Measurement Equipment

Low-Speed Measurement Equipment

Hydrogen Proteus
• Volvo D13 piston with modified bowl to reduce
compression ratio to 12.6:1
• Scania OC13 CNG unit head
• Modified for direct injection, in cylinder pressure
measurement and metal temperatures
• Custom spark-ignition system using AEM smart coil
• BorgWarner prototype H2 injector (35 bar injection
pressure)
• Cooled EGR
• A&D Combustion Analysis Software (CAS) data
acquisition system
• I ntake and exhaust manifold and in-cylinder dynamic
pressure measurements using Kistler 6025 pressure
transducers
• Ignition and fuel injection timing via current clamps
•E
 ngine and test cell system temperatures and
pressures
•F
 uel mass flow rate using a Bronkhorst F-113AC
mass flow meter
• Air mass flow via a T Mass sensor and critical flow
nozzles
• I njection and spark timing, throttle and EGR valve
position data from National instruments control
system
• HORIBA HyEVO H2 analyser
• Two HORIBA MEXA 7000 exhaust gas analysers
monitoring intake and exhaust composition

Fig. 1.  Instrumented Scania unit head with H2 DI injector fitted

56    J. Hughes et al.

A CFD study was carried out to assess the impact on the unit head cooling circuit
to help define the optimal positioning of the DI injector as indicated in Fig. 2. Further
instrumentation was added to the cylinder head to measure an array of temperatures
at key locations close to the flame face and known critical locations close to the valve
bridges, spark plug and the injector itself.
Inlet – Inlet
bridge

Inlet – Exhaust
bridge

Injector

EX2
EX1
IN2

Sparkplug – Exhaust
bridge

IN1

Injector
drilling
Exhaust – Exhaust
bridge

Outer bore

Fig. 2.  Coolant analysis of OC13 head water jacket to assess cooling risk due to injector
installation. Head instrumented with thermocouples to log the thermal loads during testing,
shown on the right

Further CFD studies were carried out to assess the injector location in relation
to mixture preparation and homogeneity across the intake and compression strokes.
Various combustion chamber configurations as well as injector locations were
compared to assess the impact on mixture preparation prior to initial hardware
selection. Figure 3 shows some of the analysis carried out to optimise the final hydrogen injector location for the Scania unit head.

Fig. 3.  3D Computational fluid dynamics of direct hydrogen injection carried out with
VECTIS software to optimize mixture preparation and homogeneity

Assessment of a Direct-Injection, Spark-Ignited …    57

A Volvo D13 production piston was modified to reduce the compression ratio
from 16.2:1 to 12.6:1 as per Fig. 4.

Fig. 4.  Original Volvo D13 piston on the left and modified D13 piston for H2 Proteus SCE test
application is shown on the right

The engine was installed in the Sir Harry Ricardo Laboratories at the University of
Brighton as per Fig. 5 and additional instrumentation as per Table 1.

Fig. 5.  University of brighton single cylinder hydrogen engine test cell

58    J. Hughes et al.

3 Test Results
The initial test programme covered a range of steady-state speed and load operating
conditions representative of the expected operating range for HD applications as per
Fig. 6. Testing included lambda, EGR, injection and ignition timing sweeps at these
key points. Initial boundary conditions including fixed gross indicated mean effective
pressure (GIMEP) and an initial target 50% mass fraction burned (MFB) angle of
8º crank angle after top dead center firing (ºCA ATDCF). An extract of some of the
results of the testing is present-ed here.

Fig. 6.  Initial target steady state key points to characterize hydrogen combustion over an array
of boundary conditions including lambda, EGR, ignition and injection timing

3.1 Impact of Lambda on Engine-Out Emissions
To determine the sensitivity of the system to air/fuel ratio, lambda sweeps were
carried out at an array of key points. Data shown in Fig. 7 is for 900 rpm, 0% EGR,
over four loads: 1) 3.5 bar 2) 9.0 bar 3) 12.0 bar and 4) 14.0 bar GIMEP.
A lean lambda limit of λ = 5.5 was observed at 3.5 bar GIMEP and data at λ = 1.0
was possible to log at both 3.5 bar GIMEP and 9 bar GIMEP. The NOx trend shows
a progressive increase in NOx with load as in-cylinder temperature increases. The
values at very low lambdas are also influenced by retarding of the ignition timing to
help with abnormal combustion.

Assessment of a Direct-Injection, Spark-Ignited …    59

Peak NOx emissions of 6600 ppm and 7000 ppm were observed at 3.5 bar GIMEP
λ = 1.3 and 9.0 bar GIMEP λ = 1.4 respectively. NOx values then drop rapidly as
lambda is in-creased and combustion temperatures cool to give NOx values below
25 ppm at all conditions above λ = 2.5. Further testing at higher speeds showed
NOx formation to be more dominated by load than speed and, as expected, was also
influenced by EGR and relative combustion phasing as ignition timing was optimised
further.
The presence of hydrocarbons (HC), carbon monoxide (CO) and carbon dioxide
(CO2) in the exhaust emissions are due to the presence of oil in the system. The
absolute values of each are unique to this experimental system at that time.
The increases at lower lambdas and lower loads are due to the increased throttling
of the system which increases the propensity of ingress of the oil into the combustion
chamber. This highlights an additional benefit of maximising the lean combustion
potential of hydrogen by limiting the intake throttling amount required. This reduces
the risk of oil entering the combustion chamber via rings, stem seals etc.

Fig. 7.  Lambda sweep engine out emissions at 900 rpm, 0% EGR, 3.5 bar, 9.0 bar, 12.0 bar
and 14.0 bar GIMEP

60    J. Hughes et al.

3.2 Impact of Lambda on Combustion
Figure 8 shows, similar dataset to Fig. 7 and expands on the impact load and lambda have on combustion. For all the datasets the 50% MFB angle was held at 8 ºCA
ATDCF except for the lambdas where knock was preset. Combustion stability
is relatively good throughout the majority of the lambda sweeps with coefficient
of variance of GIMEP (CoV of GIMEP) <3% throughout the test suite. At lower
lambdas the COV of GIMEP starts to rise as abnormal combustion starts to influence
the system. However, even out to very lean values of λ = 4.5, the combustion stability
is good and hence COV of GIMEP is still very low (e.g. <3) with lower normalized
value of GIMEP (LNV GIMEP) only dropping below 90% at λ > 4.5.
From λ = 1.5 to λ = 4.0 for any given lambda as load increases there is only a
small change in combustion duration, even from very low load. This will potentially
help allow for a wide lambda range to be more easily calibrated for in vehicle
applications and help with maintaining correct combustion phasing during transient
fluctuation in lambda. As lambda exceeds λ = 4.0 there is a progressive slowing of
combustion relative to load and from then on an increased spread in ignition delay and
10–90% MFB is observed.

Fig. 8.  Lambda sweep combustion and pressure characteristics comparison: 900 rpm, 0%
EGR, 3.5 bar, 9.0 bar, 12.0 bar and 14.0 bar GIMEP. Angle of 50% MFB is constant at 8 ºCA
ATDCF

Assessment of a Direct-Injection, Spark-Ignited …    61

The relationship between knock/abnormal combustion events with load and
lambda is very evident in the dataset presented. As load increases the lambda limit
that can be achieved before knock is observed increases. At 3.5 bar GIMEP λ < 2.0
resulted in abnormal combustion, whereas at 14 bar GIMEP lambda values λ < 3.0
resulted in knock and abnormal combustion events occurring.
Thermal efficiency can be observed in terms of gross indicated thermal efficiency
(GITE). GIMEP in bar is converted to power and compared to fuel power. Typically,
the peak GITE was observed at increasing lambdas with increasing load. This
however is not the final optimal thermal efficiency as the ignition timing is not fully
optimised at this point.
As expected, peak cylinder pressure (Pmax) increases with increasing lambda as
well as load. The slope of increased Pmax with lambda rises with increasing load.
The aver-age rate of pressure rise over 300 cycles increases with load and increases
further at lower lambdas as the burn rate accelerates. Overall, as load increases there
is a narrowing of the lambda range that can be utilized. At lower lambdas knock and
abnormal combustion are the limitation. At the lean end there is a deterioration in
combustion stability which is part of an ongoing study.
3.3 Impact of EGR on Combustion and Emissions
Figure 9 gives an example of the impact EGR has on combustion and emissions at the
same load and lambda condition across three speeds (900 rpm, 1400 rpm, 1800 rpm).
Across the three sweeps a large EGR rate (>45%) was tolerated without misfire.
This high EGR rate considerably increases the specific heat capacity of the system
hence lowering combustion temperatures and reducing the rate of NOx formation.
The added EGR also allows for the engine to be de-throttled as intake manifold
pressure crosses barometric pressure at approximately 40% EGR. Hydrogen’s ability
to tolerate a large amount of EGR presents more options to help reduce pumping
losses or even remove the need for a throttle in the first place.
Adding EGR as expected slows combustion and rate of pressure rise but leads
to increased cylinder pressures due to the increased trapped mass in the system. At
high-er loads the ability to tolerate lean combustion or large amounts of EGR may be
superseded by cylinder pressure limitations especially in retrofit applications.
The downside to tolerating large quantities of EGR is there is however a
noticeable increase in hydrogen slip once the EGR rate exceeds 25%. This is believed
to be due to the expected increasing flame quench distances from the walls due to the
lower combustion temperatures. This increase in the unburned boundary layer, despite
it getting progressively more dilute with EGR, leaves increasingly more unburned
hydrogen in the combustion chamber at the end of combustion.

62    J. Hughes et al.

Fig. 9.  EGR sweep λ = 2.0 3.5 bar GIMEP, 900 rpm, 1400 rpm and 1800 rpm comparison

3.4 Ignition Timing Optimisation
The ability to optimise the ignition timing for fuel economy is limited by the
sensitivity of the system to abnormal combustion and the necessary trade-off
with NOx formation rate. Figure 10 shows the benefits of leaner lambdas both in
maximizing gross indicated specific fuel consumption (GISFC) and reducing NOx.
At λ = 2.0 as ignition timing is advanced the presence of abnormal combustion
prevented the system reaching the optimal GISFC.
NOx formation increases as ignition timing is advanced due to higher in-cylinder
temperatures; hence advancing ignition timing at λ = 2.0 also has a much greater
impact on increasing NOx emission than at the leaner conditions. At λ = 2.5 and
λ = 3.0 the slower 10–90% MFB also means the rate of pressure rise progressively
reduces resulting in a much cooler combustion system and hence a lower NOx
formation rate. As Ignition timing is advanced at λ = 2.5 and λ = 3.0, although
combustion gets faster and rate of pressure rise increases, the rate of NOx increase is
much lower.

Assessment of a Direct-Injection, Spark-Ignited …    63

Fig. 10.  Ignition timing sweep at 1400 rpm 3.5 bar GIMEP, λ = 2.0, 2.5 and 3.0

The optimum GISFC for these sweeps is found between 2 ºCA and 5.5°CA
ATDCF which is relatively advanced compared to traditional spark ignited fuels
like gasoline etc. Other sweeps show similar tendencies and were also influenced
by relative speed, load, EGR rate, etc. Figure 11 demonstrates that at higher speeds
and loads (1800 rpm 14 bar GIMEP) relatively advanced ignition timing can be
tolerated com-pared to traditional λ = 1.0 gasoline systems, which have more retarded
combustion phasing due to knock.
Figure 11 also shows how the system can be progressively further optimised by
in-creasing lambda. As lambda is increased combustion phasing can be advanced
further before abnormal combustion occurs. At the lean condition of λ = 3.0 the
optimum fuel economy potenzial of the system can potentially be reached; however
large knock events are present.

64    J. Hughes et al.

Fig. 11.  Ignition timing sweep 1800 rpm 14 bar GIMEP 0% EGR λ = 2.5, 3.0 and 3.5.
Example of abnormal combustion preventing optimal ignition timing even at very lean
conditions

Figure 12 shows ignition timing optimisation with and without EGR. The addition
of EGR delays the onset of knock and once knock is encountered the amplitude of
knock is suppressed. EGR is very powerful at reducing the rate of NOx formation but
there is a trade-off with fuel economy if ignition timing cannot be fully optimised.

Fig. 12.  Ignition timing sweep showing 900 rpm 9 bar GIMEP. As EGR is added NOx
emissions are greatly reduced and knock is suppressed to the point where ignition timing can be
optimised to clearly define minimum GISFC

3.5 Injection Timing Optimisation
A range of injection timing sweeps or start of injection sweeps (SOI sweeps) were
carried out to study how hydrogen-air mixture homogeneity is influenced by injection
timing and the effect this would have on combustion and emissions. Four sweeps were

Assessment of a Direct-Injection, Spark-Ignited …    65

selected from the Ricardo and University of Brighton investigation to show in this
paper to demonstrate some of the impacts captured during this study.
The optimising range for the SOI study is initially bookended by requirement to start
injection after the exhaust valve is closed to prevent hydrogen slip, and to end injection
before cylinder pressure reaches approximately 20 bar compression pressure to avoid
any risk of damaging the 35-bar injector from intake pressure waves. For each data point
a constant GIMEP was held and the 50% MFB angle of 8 ºCA ATDCF was maintained.
Figure 13 outlines some of the findings. As the injection timing is swept from the most
advanced setting of 300 ºCA BTDCF to 180 ºCA BTDCF a progressive increase in NOx
formation is observed. This would indicate increased temperatures within the combustion
chamber and is further backed-up by a reduction in 10–90% MFB angles, indicating faster
combustion as well as increases in rate of pressure rise and maxi-mum cylinder pressure.
Combustion stability is also influenced by the injection timing. As SOI is retarded
to <240 ºCA the COV GIMEP starts to increase and the LNV GIMEP decreases. Also
of note is the fact that at 1800 rpm 9 bar GIMEP λ = 2.5 the system is very close to
its knock threshold. As injection timing is retarded this becomes more evident and increasing knocking pressures were logged.
Overall, the data suggests mixture homogeneity has not fully stabilized even at
very advanced injection timings and feeds directly into further studies by the Ricardo
CFD team.

Fig. 13.  Injection timing sweeps at 900 rpm and 1800 rpm, λ = 2.5, λ = 3.0

66    J. Hughes et al.

3.6 Thermal Impacts
Previously a Ricardo CFD study was referenced which assessed the impact of retro
fitting an injector into the Scania unit head. To allow for further development of
Ricardo hydrogen CFD capability with hydrogen an array of thermocouples were
fitted to the cylinder head.
The thermocouples were positioned to provide a detailed overview of the
temperature profile during all the various sweeps. This data can help build a picture
of the boundary conditions to be expected from hydrogen combustion for further
modeling by the Ricardo design and analysis teams.
Figure 14 shows an example of maps at λ = 2.5 with 0% EGR showing the
temperature profiles over an array of steady-state speed and load points. This data
helps in highlighting for example how successfully the area near the injector has been
cooled. It also is combined with other datasets to understand the thermal impacts of
burning hydrogen in an ICE application.

Fig. 14.  Cylinder head thermocouple data at λ = 2.5 with 0% EGR. Data can be mapped to
show temperature profile in the unit head across all speed and load points for given lambdas,
EGR levels etc.

In addition to metal temperatures, the exhaust gas temperature data provides a
vital overview what the boost and aftertreatment systems can expect to experience
on a hydrogen engine. The absolute temperature of the engine out exhaust gas from
single cylinder engines is not fully representative of the absolute temperatures that

Assessment of a Direct-Injection, Spark-Ignited …    67

would be obtained on multi cylinder engine. The long delay between each exhaust
gas stroke, different heat transfer pathways and exhaust gas pulsations etc. greatly
influence the temperate measurement. However, the relative trends and sensitivities to
different boundary conditions (lambda, EGR, ignition timing etc.) can be captured and
used as inputs to models. The data also gives an overview of conditions best suited to
enhancing transient behavior or maintaining required aftertreatment temperatures.
Figure 15 gives a small overview of how the exhaust gas temperatures change
in relation to lambda, speed and load. These can be cross referenced with the NOx
emissions maps also generated to help understand the challenges facing boosting and
aftertreatment systems (ATS).
As can be seen from the data there is a noticeable drop in exhaust gas temperature
as lambda is increased. This provides challenges for transient strategies and can be cross
referenced with compressor and turbine options to better understand the tradeoffs to be
expected with keeping NOx emissions low while maintaining transient expectations.

Fig. 15.  Example of engine-out exhaust gas temperature data mapped against speed and load
at different boundary conditions. An array of sensitivity studies provide reference data for 1-D,
CFD, FE, 0-D etc. studies. The thermal data shows the impact numerous boundary conditions
have on the relative engine out exhaust gas temperature

68    J. Hughes et al.

3.7 Abnormal Combustion
The ignition timing sweeps seen earlier show that end-gas knock can be observed
when advancing ignition and hence by retarding ignition timing avoided. Also, the
lambda sweeps showed that at a fixed combustion phasing, as lambda is reduced an
asymptote is reached where end-gas knock occurs; this asymptote getting leaner as
load is increased. However, in addition to these system characteristics where abnormal
combustion occurrence is relatively predictable there are also occurrences of extreme
abnormal combustion events that are much more stochastic.
Figure 16 on the left shows an example of a single large knock event of 145 bar
peak-to-peak (pk-to-pk) followed by a single preignition (PI) event after which the
system returned to normal combustion cycles. This is one example of a large number
of similar events seen at many different test conditions. On the right is example of
multiple events of PI over 100 consecutive cycles. Similarly, very high pk-to-pk
knock pressures of 210 bar were logged and again the engine had relatively normal
combustion before and after.

Fig. 16.  Examples of the stochastic nature of more extreme abnormal combustion events are
shown above. Left shows large single knock event followed by single PI event before system
returns to normal combustion. Right shows PI with large knock event followed by multiple
lesser PI events before system recovers to normal combustion

Figure 17 on the left shows an example where multiple knocking PI events
get progressively more advanced before a single very early PI event occurred. The
following cycles returned to the previous PI combustion characteristics. In this case

Assessment of a Direct-Injection, Spark-Ignited …    69

the engine showed no signs of returning to normal combustion and was manually shut
down to prevent damage.
To the right in Fig. 17 is an example of a very early PI event which occurs
during the intake stroke. As the intake valve is open the pressure wave registered on
the intake Kistler transducer and is shown below. The following cycle then misfired
believed to be due to excessive amount of residuals then present in the intake system.

Fig. 17.  On the left is a condition where PI progressively advanced with each cycle, followed
by a very early PI event. On the right (top plot) is an example of single PI during intake stroke
with pressure wave registering on the intake Kistler (below plot). The following cycle misfired

Only a small number of events are discussed in this paper out of a very substantial
number that were captured during testing. The root cause of for all events is not
necessarily the same and are part of an ongoing investigation at Ricardo. The presence
of hot spots could explain some occurrences where PI gets progressively worse
however the more random stochastic events are more difficult to explain as the hot
spot would need to appear and then disappear over a very small number of cycles.
Further investigation into the impact the ingress of oil, abnormalities in the ignition
system, compression temperatures etc. have on combustion is ongoing at Ricardo.

4 Ricardo Simulation
The empirical data gathered is used initially to validate Ricardo models, from 1-D to
3-D CFD, of the Proteus combustion system in order to close the loop on analysing
the findings and to help ensure valid conclusions are drawn. These models then allow
Ricardo to adjust the boundary conditions to explore new scenarios in parallel with
the test programme.
4.1 1-D Performance Simulation
A single cylinder 1-D performance model of the tested engine was built in the 1-D
software WAVE and validated against all the gathered test data as indicated in Fig. 18.

Fig. 18.  Ricardo 1-D WAVE model of Proteus SCE H2 engine

70    J. Hughes et al.

Assessment of a Direct-Injection, Spark-Ignited …    71

This 1-D model is used for the development of a predictive Hydrogen Combustion
Duration Sub-model (HCDS). HCDS is a semi empirical sub-model based on the
com-bination of laminar and turbulent flame speed calculation and empirical test data.
HCDS reflects in-cylinder conditions such as lambda, pressure, amount of internal
and external residuals etc. The sub-model also reacts to the change of bore, stroke and
intake port design. HCDS is under continuous development following the progress of
the presented testbed programme to support the future development of engines with
hydrogen combustion system.
4.2 3D CFD Simulation
The data from the Ricardo H2 SCE also provides an opportunity to validate the
updated combustion models for lean hydrogen combustion in VECTIS CFD code
includes the thermodiffusive instability effects and the general methodology for the
prediction of hydrogen combustion. As such, the initial concept level in-cylinder CFD
simulations of the Ricardo H2 SCE test engine, shown previously, were revisited and
refined to correspond to the acquired test data. Two main enhancements were made to
the modelling; the H2 injection was validated for the final injector configuration and
the combustion simulations employed the aforementioned enhanced models for lean
hydrogen combustion.
A 3-point lambda swing was considered at 1400 rpm and 9 bar GIMEP with
nominal lambdas of λ = 2.5, λ = 3 and λ = 3.5. A relatively early start of injection
timing was used during testing to try an ensure as close to a homogeneous mixture as
possible at spark timing. Figure 19 shows an example of the spray match for the H2
injector used in the SCE along with the evolution of mass weighted H2 distribution
probability during compression for the lambda 2.5 keypoint (EQR 0.4)

Fig. 19.  Example spray match for BW CNG/CHG 6.2 injector used in Ricardo H2 SCE and mixture uniformity at TDC from updated in-cylinder
injection simulation

72    J. Hughes et al.

Assessment of a Direct-Injection, Spark-Ignited …    73

With trapped conditions at spark timing now matched to the measured test
points the combustion prediction can now be more fairly assessed. Both the standard
combustion approach and the enhanced lean hydrogen combustion model were
evaluated. The two models were tuned to match the test data at the mid, λ = 3.0, test
point of the three-point test swing before being run at the two other test points without
further tuning. Figure 20 shows the response of the models to a change in lambda.
The match achieved when tuning the models at the mid-point is very good, with
both models lying on the test pressure curve. As the operating lambda is changed the
standard model tends to undershoot the measured pressure at the richer condition and
overshoot at the leaner condition. With the new thermodiffusive model, the response is
significantly improved.

Fig. 20.  Predicted in-cylinder pressure compared to test for 3-point lambda sweep

With additional data from the SCE the combustion models are being validated
across a wider range of operating conditions, with and without EGR, and for
abnormal combustion events such as pre-ignition and knock.

5 Conclusions
The internal combustion engine will remain relevant as a chemical-to-kinetic energy
converter in heavy-duty, off-highway machines and in marine applications for some
time to come. This paper gives a small overview of some of the characteristics of
hydrogen combustion captured by Ricardo and the University of Brighton. The
hydrogen combustion lean limit extending to >λ = 4.0 in some conditions allows
the engine out NOx emissions to be reduced to <10 ppm. Similarly, the engine has
a high tolerance for EGR with levels of >40% possible without misfire, giving alternate combinations to keep engine-out emissions low, reducing the compromise with
specific power, exhaust gas temperatures, aftertreatment specifications etc. However,
the potential for extreme abnormal combustion events to occur needs to be better
understood when optimising the combustion parameters.
Retrofitting an existing combustion system with hydrogen fuel injection equipment (FIE) presents additional challenges as there are restrictions on where the

74    J. Hughes et al.

injector can be positioned etc. The use of CFD tools are essential to understand
the trade-offs available. Specific empirical tests including injection timing sweeps
etc. help show how the homogeneity can be further improved and further iterations
of CFD can help fine-tune the combustion system. Ricardo validated 1-D models
help better understand the combustion process and the thermodynamics behind the
sensitivity studies carried out. Further structural analysis, cooling circuit analysis
etc. is then possible by combining empirical data and model outputs to maximize the
understanding of burning hydrogen in an internal combustion engine.
The Ricardo Proteus SCE has shown there are very robust ways to study the impact of running an engine configuration on hydrogen fuel prior to further studies
on multicylinder applications. This study shows the possibility of combining direct
injection technology with existing swirling combustion chambers. This aligns well
with the expected desired strategy to create direct-injection hydrogen fuelled heavyduty engines via minimum modification to existing diesel and CNG products.
The learning from this study is also being transferred to 13l multi cylinder engine
currently installed in Ricardo Shoreham Technical Centre in UK.
Acknowledgements. The authors would like to thank the University of Brighton for the use
of their facilities and equipment as well as their collaboration and support in carrying out this
extensive study. Finally, the authors would like to thank the Ricardo UK Test Operations team
for their support and the wider Ricardo technical teams for their input and expertise in assisting
with this study and paper.

Glossary
ATDCF

After top dead centre firing

GISFC

Gross indicated fuel consumption

ATS

Aftertreatment systems

GITE

Gross indicated thermal
efficiency

CA

Crank angle

HC

Hydrocarbon

CFD

Computational fluid dynamics

HCD

Hydrogen Combustion Duration
Sub-model

CHG

Compressed hydrogen gas

ICE

Internal Combustion Engine

CNG

Compressed natural gas

LNV

Lower normalised value

CO

Carbon Monoxide

MFB

Mass fraction burned

CO2

Carbon dioxide

NOx

Nitrogen oxides

COL

Carbon monoxide low range

PI

Preignition

COV

Coefficient of variation

PPM

parts per million

EGR

Exhaust Gas Recirculation

SCE

Single cylinder engine

EQR

Equivalence ratio

TDC

Top dead centre

GIMEP

Gross indicated mean effective
pressure

Assessment of a Direct-Injection, Spark-Ignited …    75

References
1. Walter, L., et al.: The H2 Combustion Engine – The Forerunner of a Zero Emissions Future,
Vienna Motor Symposium
2. Boberic, A., M. Sc., et al.: Measures to achieve high specific power with a heavy-duty Ha
internal combustion engine: A numerical and experimental analysis. 31st Aachen Colloquium
Sustainable Mobility 2022. Aachen, October 10th to 12th (2022)
3. Maio, G., et al.: Retrofitting a diesel baseline to a fully Ha spark ignition engine by
combining experiments, OD/1D, and 3D CFD simulations: 31st Aachen Colloquium
Sustainable Mobility 2022. Aachen, October 10th to 12th (2022)
4. Chi, Y., Dr., et al.: Hydrogen Internal Combustion Engine: Zero- Impact Emission Technology for Sustainable Mobility: 31st Aachen Colloquium Sustainable Mobility 2022. Aachen,
October 10th to 12th (2022)
5. Lazzaro, M., Catapano, F., Sementa, P.: Experimental characterization of methane direct
injection from an outward-opening poppet-valve injector. SAE Technical Paper 2019-240135 (2019). https://doi.org/10.4271/2019-24-0135
6. Atkins, P. A., Dr., et al.: Examining trade-offs between NOx emissions and hydrogen slip for
hydrogen combustion engines, THIESEL 2022 Conference on Thermo- and Fluid Dynamics
of Clean Propulsion Powerplants. 13th–16th Sept 2022

The Compact Catalytical Heater (CCH):
Thermal Management for HD EU-VII/
EPA-27 with Low Impact on Existing EATS
Architectures
Manuel Presti1(*), Oswald Holz1, Mathias Keck2, and Dennis Sailer2
1

2

Vitesco Technologies Emitec GmbH, Lohmar, Germany
manuel.presti@vitesco.com
BIN Boysen Innovationszentrum Nagold GmbH & Co. KG,
Nagold, Germany
mathias.keck@bin.boysen-online.de

Abstract.  Currently the OEMs of commercial vehicles (using an internal
combustion engine) are preparing for the new emission legislation EU-VII und
EPA-27. The challenge is an improved NOx-reduction at cold-start and low
load conditions as well as reduced CO2. In the meantime, also in the Off-Road
sector (NRMM) the first discussions started on next steps. The existing exhaust
aftertreatment systems are typically showing very high NOx-conversion rates
when operated at the appropriate temperature. OEMs are investigating a variety
of possible solutions to increase the temperature level of the SCR catalysts for
the critical operation points: cold start and low load. “Pure” electrical heating
has quite a high CO2-disadvantage in the case that the electrical power is
coming from alternator and battery. The high power-demand for higher exhaust
mass-flow is another challenge for the electrical system. Fuel dosing into the
exhaust system is a state-of-the-art process for DPF regeneration and shows
a high efficiency in energy release. Combining electrical heating with fuel
dosing results in an efficient heating with moderate electrical power demand.
“Engine-independent” heating of the exhaust system allows in turn the engine
to operate in the “CO2-bestpoint” and thus reduce the fuel consumption. The
innovative compact catalytical heater as an add-on component upstream of
a well proven CV-muffler allows to begin fuel dosing very early in the coldstart phase, generating high amount of energy for the fast heat-up of all
components (including catalysts) within the muffler. It is also intended to
operate for keeping the exhaust system warm in low load operation. The goal
is to continue using already existing and well-established exhaust components
(muffler) to reduce R&D efforts & tooling costs. This paper describes the
development of the system including simulation as well as test results on a
dynamic heavy-duty engine test bench resulting an improvement of 11%-points
in the FTP cold start and in the Low Load Cycle (LLC) a NOx conversion of
99% with active heating.

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 76–94, 2023.
https://doi.org/10.1007/978-3-658-41477-1_6

The Compact Catalytical Heater (CCH): Thermal Management …    77
Keywords:  Low NOx · Exhaust thermal management · Electrically heated
catalyst · Catalytical heater

1 Introduction
In order to comply with the coming reduced emission limits (USA & Europe)
combined with reduced fuel consumption, the exhaust aftertreatment system
(EATS) needs to be “fully operational all the time” when the engine is running.
Engines with improved fuel efficiency tend to have lower exhaust temperatures,
which further complicates catalyst operation in low load condition. Cold start of the
engine is expected to become a major challenge and will require a fast heat-up of all
components in the EATS. Typical EU-VI trucks have the most tailpipe nitrogen oxides
(NOx) emissions in the first 10% of the engine-work [g/kWh], while Stage V non-road
mobile machinery (NRMM) operate in some RDE cases within extended idle mode.
The tailpipe NOx-concentration has peaks at a power lower than 10% or higher than
75% of the nominal power [1, 2].
Several studies have been conducted on thermal management of heavy-duty
exhaust systems in order to get the catalyst-temperature into the window of best
performance [3–7]. The current paper describes the development of an advanced
catalytical heater capable to provide a high amount of heat-energy during cold start
and low load operation utilizing only a limited electrical energy (max 4.3 kW).
The compact catalytical heater (CCH) can provide the heat-energy mostly
independent from the engine operation point allowing the engine to be driven in
a fuel-efficient (or hybrid supported) mode. The goal is to continue using already
existing and well-established exhaust components (muffler) to reduce R&D efforts &
tooling costs.

2 Current State of Technology
Engine developers are continuously increasing the efficiency of combustion engines
and thus reducing fuel consumption and CO2 emissions. This improvement leads to
lower exhaust gas temperatures and in some cases decreases the efficiency of exhaust
aftertreatment systems. “Passive” measures – such as a close coupled position of
the catalysts or insulation of the tubing – can support. In our investigations with a
close coupled SCR-catalyst (including a second, “main” SCR -system: dual SCR)
an “active” heating was still necessary [3]. The prevalent active heating is currently
done with engine measures. For DPF regeneration in addition fuel is injected into the
exhaust and combusted on the oxidation catalyst (DOC: Diesel Oxidation Catalyst).
Electrical heating and external burners are being investigated as “additional active
heating”.

78    M. Presti et al.

By using engine-based measures to raise the temperature in the exhaust system,
one part of the introduced energy is dissipated via the engine cooling system and does
not contribute to the heating of the exhaust system components. External burners have
the disadvantage that, in addition to the exhaust gas, the combustion air has also to be
heated, which increases the overall fuel consumption. Efficiency of “pure” electrical
heating is in the range of ~40–60% when the energy is used directly from alternator and battery. Total electrical heating power is limited by the installed electrical
configuration on board.
Combining of electrical heating with fuel injection promises an improved overall
heating efficiency and having a fast-acting controllability. For this the electrical power
is used only to ensure the oxidation catalyst light-off for the catalytic “combustion” of
fuel. Heat release of the fuel is expected to have an efficiency close to 100% – directly
in the exhaust system. With an appropriate configuration and actual operation up to
40–50 kW heat energy can be provided for the heat-up of the exhaust system.
In previous studies at an engine test bench [4] the hydrocarbon (HC) conversion
with additional limited electrical heating power and fuel injection has been analyzed.
For a critically low load point the heat release of the injected HC on the DOC is 50%
higher than the reference without additional heating.
The small full flow EHC (Fig. 1, conf. A, before decoupling element) in position
1 has higher thermal loss along the exhaust path than small EHC (same type, conf. B)
in position 3 and EHC (conf. C) in front of the DOC. The configuration A and B can
be beneficial in comparison to configuration C as a possible slip of the injected fuel
is neglected, based on the mixing and evaporation length. Figure 2 shows an example
of HC conversion for configuration C with varying temperature upstream the DOC.
The given electrical power (4 kW) can not generate same conversion levels at a higher
mass flow compared to the lower flow rates.

Fig. 1.  Modular test build-up without bypass design to validate the HC-conversion with
electrical heating catalyst EHC in configuration A, B or C [4]. (©Boysen)

The Compact Catalytical Heater (CCH): Thermal Management …    79

Fig. 2.  HC-conversion map for electrical heating catalyst EHC in front of DOC (at position C,
s. build-up) at modular test build-up. (©Boysen)

An increasing exhaust gas temperature is beneficial for the urea preparation
as limitation towards the reduction of NOx. Alternate options to enhance the urea
preparation are the integration of active heating measures in the urea mixing path
(EHC, heating pipe) or using passive measures like mixing elements, Hydrolysis
wash-coat on the evaporation surface and pipe in pipe configuration. Especially
passive measures enhance the generation and distribution of Ammonia (NH3) for the
SCR-catalyst, but do not affect the SCR-Light-Off significantly (Fig. 3).

Fig. 3.  Measures for enhancing urea preparation as limitation of the SCR-catalyst function.
(©Boysen)

80    M. Presti et al.

3 Development and Validation of a Compact Catalytical
Heater (CCH)
For regeneration of particulate filters – in the heavy-duty sector – is common to
inject fuel directly into the exhaust system and convert it into heat at the oxidation
catalyst. To do this, the catalyst must have reached its operating temperature, which
is not ensured in all engine operating points, so that heating measures are required
to “support”. With electrical energy, the temperature of the oxidation catalyst can be
raised by means of an EHC (EHC = Electrically Heated Catalyst) to a level that the
introduced fuel is completely converted. The required electrical energy depends on
the exhaust-gas mass flow and its temperature [4]. During the cold start, a fast heatup of the exhaust system is required: i.e. to supply as much heat energy as quickly as
possible into the exhaust system. In this case the advantage of fuel dosing is evident:
with the (limited) electrical power only the HC-Light-Off of the DOC is ensured – the
dosed fuel then provides the high heat-energy input.
For cold start operation and system configurations in which a high electrical
power is not feasible or desired, it is possible to bring a partial gas flow with a lower
electrical power faster to the “ignition temperature” of the oxidation catalyst. With
the addition of fuel, a large amount of energy can be introduced not only to heat the
partial flow to an elevated temperature level (e.g. 600 °C) but also the entire system
much faster to the desired level (e.g. 220 °C of the SCR-catalyst).
Initial results of a first-generation heating module with partial flow EHC &
fuel dosing, employing an EHC with a volume of only 1.65 l (Ø 174.6 mm) and a
maximum electrical power of 4.3 kW, were described in [3]. Based on the findings
of the first PoC (Proof of Concept) configuration, a new compact system has been
developed with reduced heat capacity: CCH (Compact Catalytical Heater).
This paper describes the development and validation of the new system with 50%
reduced length, weight hence heat capacity in a close to industrialization status.
3.1 Concept
The task of the new system is to improve NOx-emission-reduction in the cold start or
low-load range without causing disadvantageous back pressure in the nominal load
case.

The Compact Catalytical Heater (CCH): Thermal Management …    81

This is achieved by means of controllable mass flow management through a
flap. Electrical heating power and flap position are adjusted to each other based on
the operation condition. The Boysen-2/2 way-flap is controlled by a brushless direct
current motor of a Sonceboz actuator with a very short closing time <150 ms and
high precision in closing position and torque control. Injection of hydrocarbons takes
place as a function of the exhaust gas mass flow, -temperature, surface temperature
of the EHC and the desired temperature level. It is essential to obtain a high mixing
quality of the injected HC onto the EHC within one universal module/component. It
is important to have an evenly distributed ratio between heat flow through the EHC,
exhaust gas (oxygen) and the concentration of the fluid being metered in. This can be
achieved by a uniform local “lambda”.
Bypassing the EHC (the first DOC) also reduces aging effects on it.

Fig. 4.  Design of the compact catalytical heater (CCH)

With the compact design of the arrangement (Fig. 4), installation space, (thermal)
mass, and surface area can be reduced to achieve a fast light-off characteristic of the
first DOC included in CCH and the second (main) DOC in the muffler box. By a large
cross-section of the main path inlet towards the merging chamber, the overall pressure
drop can be decreased (Fig. 5).

82    M. Presti et al.

Fig. 5.  Compact catalytical heater (CCH) basic design. (©Boysen)

3.2 Operating Conditions
The simulation load points are related to the engine load points of the Vitesco Technologies Emitec test engine (Table 1). Within the nominal load point (NLP) the back
pressure increase of the CCH-system should be about 15 mbar. As typical load point
in the FTP cycle of the reference heating system [3] TLP1 indicate the susceptibility
to spray drift and asymmetric flow. For the idle phase TLP2 indicate a boundary case
without the susceptibilities of TLP1.
Table 1.  Load points for simulation designing
Load Point
NLP
TLP1
TLP2

Massflow
in kg/h
2100
800
120

Inlet Temperature
in °C
500
220
120

Flap Variation
Open
45°/Closed
Closed

HC-Dosing
in g/s
–
0.75
0.12

Measurements of the hot gas test rig include a mapping with steady load points
and tests on engine dyno beyond TLP1 transient cycles.

The Compact Catalytical Heater (CCH): Thermal Management …    83

3.3 General Model Features and Simulation Procedure

Fig. 6.  Conceptional design of elements in the mixing chamber. (©Boysen)

Pre-analysis on a hot-gas test rig and a verification via CFD has shown that through
the perforated structure PE-Design® of the EMICAT® a recirculation of gas via an
inlet swirl structure can appear, when the swirl-number is very high. The flow
simulation of the heating disc and support catalyst have specific linear axial viscous –,
quadratically axial, linear viscous and quadratic radial internal resistance. The radial
heat and mass transfer are supported through the perforated structure of the EHC and
the swirl flow of the concept (Fig. 6).
The development of the CCH-design is based on the tool Star-CCM+ from
Siemens (Version: 14.04.013).
• Meshing settings:
– Prism layer (PL) Thickness: 5 mm/PL Stretch: 1.3/PL Quantity: 10
• Per simulation (load case) 2.7 million cells
• Physics:
– Realizable K-Epsilon Two-Layer (Turbulence Model).
– All y+ Walltreatment (near wall consideration)
– Ideal gas
• Thermal boundary conditions:
– Interior component walls adiabatic
– Exterior walls T = 25 °C/alpha = 20 W/(m2K)
– No radiation model
The simulation results in Fig. 7 indicate the function of the mixing chamber elements
as the flow is stabilized to avoid spray deflection. The swirl is homogenized through
the resistance of the decreasing cross section, the deflection losses of the swirl cone,
throttle sleeve, perforated sleeve and the relative velocity between the injection hydrocarbon droplets towards the gas flow.

84    M. Presti et al.

The axial velocity profile based on typical mono-swirl flow has a minimum in the
center, which characteristic is more pronounced with higher mass flow through the
bypass path. The comparison of a closed flap to a 45° opened flap for TLP1 indicates
the effect (Fig. 7). The mass flow percentage of the bypass within opening of the flap
can decrease to about half for TLP1 and about one quarter for the nominal load point
(NLP).

Fig. 7.  Analysis of the mass flow, velocity and static pressure of compact catalytical heater
(CCH) #510 model for partial load points. (©Boysen)

Fig. 8.  Analysis of the mass flow, velocity and static pressure of compact catalytical heater
(CCH) model #510 for nominal load point. (©Boysen)

The Compact Catalytical Heater (CCH): Thermal Management …    85

At the nominal load point the flow acceleration from the outlet chamber
downstream to the pipe can cause a pressure loss (Fig. 8). For further reduction of
the back pressure the flow axis of the downstream path of the CCH-system is lower
positioned in the height (Fig. 9). A much higher back pressure for bypass-path as for
the main path is advancing high HC-conversion within low exhaust mass flow through
the EHC (s. chapter “Current state of technology”, Sect. 2). The total back pressure at
TLP1 with a closed flap of the prototype is 57 mbar (Fig. 10).

Fig. 9.  Design of the compact catalytical heater (CCH) model #510 compared to model #511
(concept for prototype) with two different outlet cones. (©Boysen)

Fig. 10.  Analysis of the flow characteristic both concept for the prototype and the modular
ready for production prototype of the compact catalytical heater (CCH) with TLP1 and closed
flap. (©Boysen)

86    M. Presti et al.

For developing a control strategy, the modular design of the prototype for the hot
gas test rig is analyzed via CFD. Through small angle of cone downstream of the hotgas burner and perforated plates the inlet flow in the cross-section towards CCH is
homogenous.
3.4 Control Strategy
Within the total mass flow ṁtot , and the flap position ϕ the partial mass flow ṁBypass,%
towards the EHC can be determined.

ṁBypass,% = f(ṁtot , ϕ)

(1)

Based on the hot gas test rig measurements a control loop, which is implemented in
MATLAB Simulink, is developed for OEM-applications. The model includes the
electrical heat-up phase and the chemical heat-up phase. The chemical heat-up phase
is based on a full conversion of HC. For development of a model the conversion rate
of the EHC has to be experimentally determined. The possible dosing amount is
depending on the ratio of the converted HC between the EHC and the second main
DOC.
3.5 Validation on the Hot Gas Test Rig
The same insulated CCH-prototype is used for the validation on the hot gas test rig
and the engine dyno. For analysis of the pressure drop, temperature distribution,
-boost and the mass flow distribution, additional elements upstream and downstream
were used for solid comparison between testing and simulation. To compare back
pressure with directed flow, two perforated plates are used: configuration 1.
The comparison between simulation and measured pressure shows a very low
difference (total backpressure = dptot; static backpressure = dps), Fig. 11. It has to be
considered that the flap has an internal and an external leakage, which can reduce the
back pressure slightly.

The Compact Catalytical Heater (CCH): Thermal Management …    87

Fig. 11.  Configuration 1 of the modular test-up of the compact catalytical heater (CCH) on the
hot-gas test-rig without electrical heating. (©Boysen)

In order to analyze the the influence of a certain swirl of the inlet flow a
configuration 2 has been installed with a 107 mm outer diameter mixing element
with 16 blades and specific outlet angles of 45°. The flow characteristic and mass distribution of the CCH towards the two paths depend on the swirl flow characteristic of
the exhaust gas downstream of the turbocharger.
The inlet flow influences the kind of swirl in the HC-mixing path. Within TLP1 and
4 kW electrical power, the temperature distribution in the cross-section at position 2
(Fig. 13) of EHC differs between configuration 1 and 2. A 100% closed flap (= “flap
position 100%”) affects for configuration 1 a higher flow velocity for negative y-coordinates
and lower temperatures in the measurement than for configuration 2 (Fig. 12).

Fig. 12.  Temperature distribution at position 2 at TLP1 at the modular hot-gas test-up

88    M. Presti et al.

Differences of the surface- and gas-temperatures for configurations 1 and 2 indicate
an influence of the inlet swirl, which can be based on a different mass flow distribution
between both CCH-paths and inside the EHC-path (s. Figs. 14 and 15). Generally, the
correlation of temperature increase through EHC is similar. If the flap is closed, an
increase of the mass flow and, as a consequence, lower temperatures are measured.

Fig. 13.  Modular test-up of the compact catalytical heater (CCH) for the hot-gas test-rig.
(©Boysen)

As forecast to the ideal starting time of HC-injection it has to be considered that
for the light-off the surface temperatures should be higher than 270 °C, but to avoid
aging stay below 600 °C (including the exothermal reaction by the injected fuel).
The isolines of the transition between the green and yellow in Fig. 14 for position 1
indicates the opening angle of the flap at a specific total mass flow as a good condition
for start of HC-injection.

Fig. 14.  Mean surface-temperatures at pos. 1 for configuration 1 (left, rectified CCH inlet flow)
and configuration 2 (right, swirl flow at CCH inlet) with ambient inlet temperature and 4 kW
heating power of EHC. (©Boysen)

The Compact Catalytical Heater (CCH): Thermal Management …    89

Within warm exhaust gas and especially within additional HC-injection the
isolines would be drifting to higher mass flow. Downstream of the bypass-path the gas
temperatures the temperatures have a similar tendency of the relation to the mass flow
and the flap position (Fig. 15).

Fig. 15.  Mean gas-temperatures at pos. 2 for configuration 1 (left, rectified CCH inlet flow)
and configuration 2 (right, swirl flow at CCH inlet) with ambient inlet temperature and 4 kW
heating power of EHC

Within the CCH the total heat flux impact to the exhaust gas is almost independent
of the partial bypass-mass flow based on the conservation of energy (Fig. 16).

Fig. 16.  Mean gas-temperatures at pos. 4 for configuration 1 (left, rectified CCH inlet flow)
and configuration 2 (right, swirl flow at CCH inlet) with ambient inlet temperature and 4 kW
heating power of EHC. (©Boysen)

Based on thermodynamic principles and simplifications (cp = const.) it is possible
to determine the bypass mass flow through temperature increase in dependency of the
total mass flow and the flap position on the hot gas test bench (s. Fig. 17).

Q̇ = ṁ

ˆ

T2

cp (T)dT
T1

(2)

90    M. Presti et al.

ṁ =

Pel
(T2 − T1 ) · cp

(3)

Fig. 17.  Calculated bypass mass flow based on measurements with configuration 2 (swirl flow
at CCH inlet), ambient inlet temperature and 4 kW heating power of EHC. (©Boysen)

To gain a desired temperature of the EHC it is necessary for a feed forward control
strategy of electrical heating and dosing HC to consider the current bypass-mass flow.
3.6 Validation on Engine Dyno
The CCH-system is validated in combination with a modified exhaust aftertreatment
system of an EU-VI truck: an EHC with hydrolysis coating (H-EHC) is installed in
the AdBlue® mixing path upstream of the SCR catalysts (Fig. 18). The inlet of the
muffler includes a perforated baffle, which enhances the temperature homogenization
and the HC-uniformity. The urea fluid is injected inside the center of a swirl flow
for an optimized urea mixing. The integrated EHC has a 9 mm heating disc and
165x40 mm support catalyst with 300/600 LS-PE® structure, which enhance the
mixing of generated urea (CH4N2O, liquid/gas), Isocyanic acid (HNCO, gas) and
ammonia (NH3, gas).

The Compact Catalytical Heater (CCH): Thermal Management …    91

Fig. 18.  
Modified exhaust aftertreatment system of an EU-VI truck with an integrated
electrically heated catalyst (EHC) downstream of the urea injection module combined with
CCH. (©Boysen)

Tests are conducted on a 15 l – US-2010 engine. Among others the following
three configurations are examined:
1. the modified EU-VI system without active heating as a reference (not shown on
Fig. 18),
2. the modified EU-VI system with heating of the “internal” Hydrolysis-EHC
(Fig. 18 – upper config.),
3. the modified EU-VI system with heating of the upstream CCH (Fig. 18 – lower config.).
For configuration 1 & 2 the flap of CCH is all the time fully open to reduce the impact
on backpressure and temperature loss. When the Hydrolysis-EHC is heated inside
the exhaust-box, the urea release temperature in the control module is lowered, as the
temperature sensor is located upstream the heating device. The maximum electrical
energy for configuration 2 and 3 is 4.3 kW. This allows to compare the impact of
heating downstream the DOC+DPF heat capacity, to the temperature increase in front
of the muffler by CCH combined with exothermal HC conversion in the DOC.
Figure 19 shows the temperature profile of an FTP cold start, upstream and
downstream the SCR catalysts and the urea dosing. With active heating the urea
dosing can be released earlier in the cycle, resulting in an improved NOx conversion
(Fig. 20). Electrical heating of the hydrolysis EHC with 4.3 kW (downstream the
DPF) causes a fast temperature-increase in front of the SCR catalysts – while heating
with CCH (upstream of the main DOC) shows a delayed heat-up in the first 220
seconds due to the heat capacity of the DOC and DPF. Nevertheless, urea dosing can
start earlier for the heated CCH as the temperature at the dosing position (upstream
the hydrolysis catalyst) is higher. With CCH the electrical energy [in kWh] could be
reduced by 12% compared to the heating of the hydrolysis-EHC.

92    M. Presti et al.

Fig. 19.  Temperature up- and down-stream of the SCR catalysts & urea dosing in FTP cold
start. (© Vitesco Technologies GmbH)

Even a larger improvement by active heating can be seen in the Low Load Cycle
(LLC). The electrical energy for CCH to keep the SCR system within the optimal
temperature window is in this cycle also reduced by 12% compared to hydrolysisEHC, while the NOx-conversion improved from 92.8 to 99.3% (Fig. 20).

Fig. 20.  NOx conversion rates in FTP and low load cycle. (© Vitesco Technologies GmbH)

Here it is evident, that 4.3 kW max power is not sufficient for the engine/massflow combination if used at the hydrolysis-EHC position. Conversely with CCH
(using the same electrical power) a higher temperature level can be maintained in the
SCR catalysts (Fig. 21) resulting in lower NOx tailpipe emissions. (Fig. 20).

The Compact Catalytical Heater (CCH): Thermal Management …    93

Fig. 21.  Temperature up- and down-stream of the SCR catalysts & urea dosing in low load
cycle. (© Vitesco Technologies GmbH)

4 Summary
Based on the results presented and information from literature, it will be essential to
provide CO2-optimized heating measures for the exhaust system to reach the future
limit values as well on-road as also off-road in Europe and in the USA.
One possibility is to run the engine in a “fuel consumption optimized map”. The
required heating measures are implemented directly in the exhaust system with high
efficiency and low heat losses – largely independent of engine operation. This also
allows a higher flexibility for hybrid operation.
Electrical energy for exhaust heating on board is limited. A significant increase in
heat-energy output can be achieved by catalytic conversion of fuel on the electrically
heated oxidation catalyst directly in the exhaust system. The electrical energy is used
“only” to ensure the “ignition” of the catalyst. A bypass allows to reduce the electrical
power demand for the light-off. Electricity from recuperation reduces the add-on fuel
consumption.
The presented compact catalytical heater is intended to be installed upstream
of today’s muffler boxes within the typically long inlet tubing. The additional
backpressure of the system can be considered small, as the system is only operated
when heating is required. During high load operation – the exhaust aftertreatment
components are typically within their operation window – no heating is required, and
the flap is open creating almost no additional flow restriction and means no additional
fuel consumption.

94    M. Presti et al.

Thus, already proven exhaust systems or components can be reused. This reduces
development and testing effort as well as tooling costs for new configurations
and opens some flexibility for different markets. In addition, the CCH enables a
great degree of flexibility for the integration of the system in various powertrain
configurations (hybrid) and vehicle architectures (On-, Off-Road).
Decisive for future solutions will be the overall system and operating costs (“total
cost of ownership”; TCO).

References
1. Rodríguez, F., Badshah, H.: Real world NOx performance of Euro VI-D trucks and
recommendations for Euro VII. International council on clean transportation (2021)
2. Desouza, C.D., Marsh, D.J., Beevers, S.D., Molden, N., Green, D.C.: Real-world emissions
from non-road mobile machinery in London. Atmospheric Environment (2020)
3. Holz, O., Presti, M.: Thermal Management for Low NOx Applications in Heavy Duty
Vehicles, ATZ heavyduty worldwide 02/2021 (pages 36 ... 39)
4. Brück, R., Presti, M., Keck, M., Dengler, J., Faiß, M.: Thermal Management on Demand; the
Exhaust Aftertreatment Solution for Future Heavy Duty Application, ATZ Live international
8th engine congress (2021)
5. Holz, O., Presti, M., Mosch, T., Dachsel, J., Rodatz, P., Fink, F., Nienhoff, M.: Heavy Duty
Exhaust Systems for Low NOx Application: Thermal Management and Control, ATZ Live
Heavy-Duty-, On- und Off-Highway-Motoren (2020)
6. Brück, R., Presti, M., Holz, O., Geisselmann, A., Scheuer, A.: Der Weg zur Erreichung der
“CARB post 2023” Emissionsgesetzgebung für Nutzfahrzeuge. 38th International Vienna
Motor Symposium, Vienna (2017)
7. Presti, M., Scheeder, A., Holz, O., Brück, R.: Motornahe Abgasnachbehandlung im Nutzfahrzeug: eine Lösung für CARB 2020 NOx? 8th Emission Control Conference, Dresden
(2016)

Liebherr’s Approach to Hydrogen Fuel
Injection Systems
Richard Pirkl1, Mario D’Onofrio2(*), Lydia Kapusta1,
and Dennis Herrmann1
1

Liebherr-Components Deggendorf GmbH, Deggendorf, Germany
{richard.pirkl,Lydia.Kapusta,
dennis.herrmann}@liebherr.com
2 Liebherr Machines Bulle SA, Bulle, Switzerland
Mario.Donofrio@liebherr.com

Abstract.  Liebherr develops hydrogen fuel injection system solutions to
be used in on- and off-highway hydrogen combustions engines. Heavy-duty
off-highway applications have partly different requirements compared to onhighway applications. Robustness against dust, dirt & vibrations and other
harsh environmental conditions must be given. Additionally higher peak
power demand and more dynamic load cycles increase the requirements on
the transient performance of the engine. To meet these requirements, Liebherr has developed a complete hydrogen injection system that includes all
the components needed for pressure regulation and fuel dosing. Throughout
the development, real load cycles of heavy-duty mobile machinery have been
considered to properly design the system and its components. This paper will
provide detailed insights on the layout, design and functionality of the hydrogen injection system. In particular, the dynamic pressure regulation by means
of a gas metering valve is shown and how this approach enables diesel-like
transient engine behavior. Furthermore, first test results on the system performance are provided. Additionally an overview on the common platform
approach to the Liebherr portfolio of hydrogen fuel injectors for port fuel
injection (PFI) and direct injection (DI) incl. actual test results will be given.
Liebherr’s approach to hydrogen fuel injection systems are not only limited
to heavy duty commercial engines. Also components for large engines are
considered in the overall platform approach.
Keywords:  Hydrogen combustion engine · Hydrogen injection system · Low
pressure direct injection · Port fuel injection · Alternative fuels

1 Introduction
Reducing greenhouse gas emissions and thus limiting global warming below the 2 °C
target is one of the key challenges of the 21st century. The importance of this goal is
also reflected by the high amount of programs and regulations towards greenhouse
© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 95–111, 2023.
https://doi.org/10.1007/978-3-658-41477-1_7

96    R. Pirkl et al.

gas reductions that are currently proposed or becoming effective. The European “Fit
for 55”, the “European Green Deal” & the Californian “Advanced Clean Truck” are
mentioned here as representative for the ongoing global activities.
The ongoing political activities influence the development activities of OEMs all
other the world. Development activities are shifted more and more towards alternative
powertrains and alternative fuels like hydrogen, methanol & ammonia for example.
Currently many public announcements on all kind of a zero CO2 powertrain
solutions from the hydrogen fuel cell to the hydrogen combustion engine can be
found. Especially in an environment that is shaped by a dynamic change process
regarding today’s powertrain concepts the hydrogen combustion engine can offer
some advantages over the fuel cell. Existing production facilities and basic engine
components can further be used. Packaging on existing equipment is relatively easy
possible and the hydrogen combustion engine allows for lower requirements on the
H2 purity and air cleanliness. Mobile construction equipment and other off-highway vehicles and machinery in particular can profit from these advantages of the
hydrogen combustion engine. Taking additionally into account the high variety of
different off-highway machines it becomes obviously that an integration of fuel cells
will not be possible in the short to mid-term into all of these machines. Thus the
hydrogen combustion engine cannot only be seen substitutional to the fuel cell but
complementary in a technology open approach [1].
The hydrogen injection system is one of the performance determining and at the
same time most challenging component for a hydrogen combustion engine. Liebherr
develops different hydrogen fuel injection systems and components for port fuel and
direct injection for commercial heavy-duty and large engines.

2 Hydrogen Fuel Injection System
2.1 Concepts for Hydrogen Fuel Injection
Main fuel injection concepts for hydrogen are multi or single point port fuel injection
(PFI) or direct injection (DI). Below Table 1 summarizes some key characteristics for
the main combustion concepts in combination with hydrogen.
The theoretical specific power level of the H2 internal combustion engine (ICE)
should be comparable high, based on the high mixture calorific values. Nevertheless,
all the hydrogen engine concepts that have been published during the last years could
not fulfill the theoretical expectations. One major reason for this is that all these
concepts where based on PFI due to a lack of components for DI [2]. In the meantime
PFI concepts have improved in terms of performance, but there is still a significant
gap to the level of a diesel engine [3].
As shown in Table 1 recently published engine tests lead to the result that the
efficiency level of a spark ignited engine with pre-mixed combustion is for both
injection concepts (PFI and DI) on the same level [4]. Intuitive one would expect
that the DI injection lead to a higher efficiency level compared to the PFI, thanks to
a better cylinder filling combined with optimized combustion timing. These benefits
are again compensated by higher wall heat losses and gas ex-change work. Finally,

Liebherr’s Approach to Hydrogen Fuel Injection Systems    97

measured efficiency levels for PFI and DI are today on the same level, but the DI
injectors used for the published test results have not been fully optimized in respect
of spray and mixture formation [2, 4]. It still can be expected to achieve an efficiency
benefit for DI in comparison to PFI. Even so, the results on the efficiency with H2 DI
today have not revealed their full potenzial, one major benefit of DI over PFI is out
of discussion. The maximum achievable specific power density of H2 DI engines is
higher than for H2 PFI engines applying the same charge air pressure level, since air
displacement effects by the hydrogen injected into the intake air manifold lead to a
lower mixture calorific heating value for PFI.
Table 1.  Hydrogen combustion key characteristics (hydrogen values based on [5])
Positive/spark ignited pre-mixed combustion

H2 injection pressure
Mixture calorific
heating value
Powerdensity
Efficiency
Risk of backfire
H2 storage and tank
utilization

Efforts for fuel
injection system

PFI
≤15 bar
3–3.2 MJ/m3

Low pressure DI
≤60 bar
4.2–4.5 MJ/m3

−20–−30%
compared to diesel
~42%
high
All gaseous H2 storage
tanks (350–700 bar)
Thanks to low
injection pressures
good usage of the tank
volume
But injection pressure
levels may be too high
for liquid H2 (LH2)
tanks

−10–0%
compared to diesel
~42%
low
All gaseous H2 storage
tanks (350–700 bar)
Higher injection
pressure levels lead to
less efficient usage of
the tank volume
Lower driving/
working range possible
compared to PFI
Use of LH2 tanks only
with cryogenic pump

Low
(based on existing
natural PFI technology)

Compression ignited
diffusive combustion
High pressure DI
250–350 bar
> 4.5 MJ/m3
Comparable to diesel

> 45%
No risk of backfire
Gaseous storage
is possible, but
compression will be
required very early.
Due to this gaseous
storage might be
inefficient for high
pressure DI
LH2 storage will
require a cryogenic
pump, but offers the
possibility for more
efficient compression
Highest demand on H2
storage
Medium
High
(new injector concepts (new injector concept
and multi-level H2
and H2 compression
pressure control)
required)

Highest efficiency levels and best transient performance can be enabled by means of
high pressure DI in combination with a diesel-like diffusive combustion concept. For
high pressure hydrogen fuel injection one major additional challenge arises. This is
the supply of hydrogen at comparable high pressure levels up to 350 bar. In case of

98    R. Pirkl et al.

compressed gaseous hydrogen storage (even at 700 bar tank pressure), a compressor
is required to use a reasonable part of the hydrogen tank capacity. It is common
knowledge that gaseous compression is very inefficient requiring large compressors
and thus from today’s point of view no option for mobile applications. Liquid hydrogen storage and compression with a cryogenic pump will be favorable in terms of
compression efficiency but cost and availability of the required tank system and pump
is likely to be an issue in the near future. Due to the complexity of hydrogen storage
and compression high pressure DI is not the primarily focus of Liebherr.
Liebherr is developing components and system solutions for PFI and DI. The
system approach shown in the following is valid for PFI and DI systems although the
following pictures and graphics will mainly show the DI system as example.
2.2 Introduction: Main Requirements on the Hydrogen Fuel Injection
System
The overarching goal for the H2 ICE is to offer the same performance and driving
characteristics as a diesel engine, at the same displacement. The typical rated output of a 13-litre 6-cylinder engine is 400 kW (on-highway) and can be up to 450 kW
(off-highway). A static flow of 11–12 g/s is required to achieve the specific output
of 75 kW/cyl. Injection quantities ranging from approx. 2.5 mg during idling up to
at least 110 mg under full load must be possible. These basic requirements are valid
for both injection concepts, PFI and DI. In light of these very basic requirements,
it becomes clear that injection pressure modulation will be required to fulfill the
minimum and maximum injection quantity requirements.
Depending on the application profile, demanding transient requirements must also
be met. The transient behavior is important for both on- and off-highway applications.
Off-highway applications such as wheel loaders have a more dynamic load profile than
heavy trucks, for example. This is reflected in more frequent changes between low/noload and full-load operation. Lower injection pressures (at ~10 bar for DI and ~4 bar for
PFI) are required to achieve the minimum injection quantities for engine idling, while
full load operation requires higher injection pressures (up to 30 or even 60 bar for DI and
max 15 bar for PFI). Highly dynamic pressure control is required to meet the transient
requirements and to achieve a load assumption comparable to that of diesel solutions.
Table 2 summarizes the main performance requirements on the hydrogen injection
system.
Table 2.  Hydrogen injection system: comparison in term of flow and pressure requirements
System type
Injection pressure
Static flow rate
Injection duration
Max. quantity range
Min. quantity range
Pressure control

DI
PFI
10–30/60 bar
3–15bar
min. 11 g/s up to 15g/s
up to 12 g/s
≤7.5ms
≤10ms
~100–110 mg/stroke
~2.5 mg/stroke
highly dynamic, enabling fast pressure reduction in foot-off mode

Liebherr’s Approach to Hydrogen Fuel Injection Systems    99

Especially in DI applications relatively short injection duration is crucial in order
to achieve the required performance levels. So, it is important to understand how flow
conditions, cylinder backpressure and injected quantity are linked to each other and
why supersonic or supercritical flow conditions help to achieve the injection quantity
targets.
The relationship between the state of the system that is expressed by pressure
(Pup) and temperature (Tup) and quantity introduced into the combustion chamber can
be expressed via Eq. 1, which gives the mass flow rate of a compressible ideal gas
through an orifice:

Pup
dm = ACq Cm 
Tup

(1)

where Pup, Tup are temperature and pressure upstream the injector metering area A
(for sake of simplicity assumed corresponding to the value sensed inside the hydrogen
system rail), Cq the flow coefficient, Cm the flow parameter.
The flow coefficient Cq is less than one and is used to include friction losses and
a phenomenon known as vena contracta. Flow condition and gas type are accounted
in Eq. 1 via the flow parameter Cm. The goal of bringing the performance provided
by an H2 DI engine up to those typical for a Diesel engine requires that the system is
capable of guaranteeing high flow rates.
The higher the injection pressure, the higher is the quantity of hydrogen which is
possible to inject in a given time as the flow increases proportionally to the upstream
pressure.
If the hydrogen flow reaches supersonic (speed of sound) velocity at the metering
valve it is in so called choked condition. This means the gas velocity is from now
on limited to speed of sound and the injected mass will not decrease in case of
downstream pressure variation (cylinder pressure) for a fixed upstream pressure.
To ensure a sonic flow condition at injector nozzle the ratio between downstream
pressure and the injection pressure must smaller than the critical pressure (Eq. 2).

Pcombustion chamber
=
pinjection



2
γ +1



γ
γ −1

(2)

For hydrogen with a heat capacity ratio γ in the range of [1.406 @ 20 °C ÷ 1.399
@ 100 °C] critical pressure ratio varies around ~0.52. This means that the injection
pressure must be about twice the backpressure in the cylinder.
Based on this physical requirement, Fig. 1 shows an example of the possible DI
injection windows under sonic flow conditions for 30 and 60 bar rail pressure from
inlet valve closure. At 60 bar rail pressure injection the maximum cylinder pressure
within the barriers for sonic flow conditions is 30 bar. At 30 bar rail pressure the
maximum cylinder pressure to stay within sonic flow conditions is 15 bar. Higher
rail pressure thus enables a significantly longer injection window and thus potentially
higher power per cylinder.

100    R. Pirkl et al.

Fig. 1.  Example of injection windows at different injection pressure levels

2.3 Real-World Off-Road Engine Transient Requirements
The application load profile is determined directly from measurements on real machines
and converted into the H2 fuel requirement. Since the wheel loader has one of the most
dynamic load profiles of mobile construction machinery, it was chosen as the reference.

Fig. 2.  Mission profile measurements for a Liebherr wheel loader

Liebherr’s Approach to Hydrogen Fuel Injection Systems    101

The mission load profile is derived from measurements on today’s diesel
powered machines. Relevant parameters are recorded at high sampling rate and later
on converted in H2 system fueling requirements. Figure 2 shows mission profile
information recorded for a wheel loader.
Figure 3 shows how the machine mission profile has been converted in H2 fueling
quantity requirements.
In order to achieve the same level of drivability for the H2 system as for diesel
systems, the dynamic levels in terms of torque and power must be equivalent. In terms
of the demands on the H2 system, this leads to high gradients for the regulation of fuel
quantities, especially during the transition from idling to full load, and thus also to
very dynamic, transient demands on the system pressure.

Fig. 3.  Mission profile conversion into hydrogen fueling requirements

The wheel loader is run in full load for only ~5% of its operating time. Most of
the time it runs at low load or idles. Another major challenge is the very frequent load
jumps between full load and idling and the associated reduction in system pressure.
These operating conditions differ greatly from a railway application profile (high load
without frequent load changes), for example.
2.4 System Design & Layout
The structure of the H2 system is determined by the vehicle layout and installation
space. The goal of the Liebherr H2 system is to allow for fast and accurate pressure
and flow control, regardless of the position of the fuel tank, engine, vehicle size and
layout. The Liebherr design is compatible with two-stage pressure control as shown in
Fig. 4 which, based on simulations and experimental results, offers clear advantages
in terms of dynamic pressure control.

102    R. Pirkl et al.

Fig. 4.  Proposed system layout: 2 stages regulation

In stage 1 by means of a pressure regulator (LPR), the filling level dependent tank
pressure is controlled to a variable or fixed predefined set point. In stage 2 by means
of a gas metering valve (LGV), the rail pressure is controlled to a set point that is
dependent on the operating conditions. The value of the rail pressure which feeds the
injectors is the results of the mass flow balance between the mass flow entering the
system and the mass flow which leaves the system (flow injected by each injector
according to the engine power demand).
This process is described by the following differential equation



Ninjectors
dm  �
=
−ṁout i  + ṁin
dt

(3)

i=0

here m is the mass of the gas within the control volume
.

m = ρcontrol volume Vcontrol volume
The injection pressure is regulated by operating the GMV via the electronic unit
control (ECU). The ECU uses the information provided by the pressure and
temperature sensors installed on the rail in combination with the required injection
quantities as per engine map to calculate the amount of gas to be delivered via the
GMV and controls thus the rail pressure.
A key component of this system layout is the GMV that serves two functions.
First the control of gas mass flow in proportional operation and second the shut-off
function between pressure control valve and rail.
The Liebherr H2 system is conceived for operating without any electronic pressure
relief valve. The idea is to keep the system layout as simple as possible as well as not
releasing any hydrogen gas in atmosphere during operation (which may be critical due
to safety restrictions or special application requirements).

Liebherr’s Approach to Hydrogen Fuel Injection Systems    103

In order to be used in many different applications, regardless of whether DI or PFI
engine are considered, a common system layout is proposed. Therefore, the design of
the system entails the same pressure regulation concept and the use of a single rail for
PFI and DI as can be seen in Fig. 5 (the shown layout of rail and injectors refers to the
DI system, for PFI the interface between rail and injectors can be different).

Fig. 5.  Liebherr H2 fuel injection system: in evidence the use of a common rail configured as
typical diesel like injection system

The development work to ensure that a single layout can be adapted to the
needs of different applications and engine configurations must be a tradeoff over the
following requirements, which sometime are in contradiction:
• System pressure gradient up & down from IDLE – full load (bar/s) and vice versa
• Pressure stability @ different engine operating points (IDLE, full load, torque
point)
• Hydrogen volume discharge at engine stop (mainly driven by safety requirements)
2.5 Experimental Results on the Hydrogen System
Liebherr has developed a system test bench with the aim of having a tool for research
and development activities at system level. The bench (shown in the Fig. 6) can be
used to test the complete system or the performances of the individual components
(through independent control of the injectors and GMV).
The bench is also equipped with a high speed data acquisition (DAQ) and is
designed in such a way that the pressure and temperature are continuously monitored
and feed backed to the ECU to enable the pressure and injection control functions.
The bench is also equipped with a back pressure system connected either to the tip of
the injectors or to the outlets of the other components to evaluate their performances
under variable and controlled pressure drop.

104    R. Pirkl et al.

Fig. 6.  On the left an overview of the H2 system bench, on the right the whole Liebherr H2
system installed including the MPRV up to the injectors

Additionally the system test bench permits also the full simulation of a machine
profile. In the Fig. 7 a typical case of a 4 cylinder engine running at a constant speed
of 2000 rpm at variable load is shown. This case is chosen to demonstrate the typical
performances and limitations, our system layout can achieve.
In light blue the rail pressure is plotted whereas in red the corresponding injected
quantity is plotted. Three relevant cases can be differentiated
1. Engine load increases. In this case the injected quantity is suddenly increased to
simulate the transient from IDLE condition to full load. The pressure set point also
is suddenly increased and the system pressure follows the set point with a response
time < 1ms. An overshoot of the system pressure is also visible, which is not linked
to the system layout, but determined mainly by the actual status of the calibration
and which is under further optimization.
2. Foot-off operation. in this phase (in the below graph, this event happens two times
@ 2a,2b) the engine load is decreased and the injected quantity is also suddenly
decreased to simulate the transient from a partial load or full load (2a and 2b) to
idle. In the plot it can be seen that the gradients of pressure decreasing are different.
In case 2a the idle pressure is reached in 2 seconds while in case 2b in about 5
seconds. This is essentially linked to two factors: the intrinsic characteristics of the
injection system and the absolute value of the injected quantity (mg/shot), in case
2b the pressure drops more slowly to the idle value because the injection quantity is
one third compared to case 2a and therefore the system depressurizes naturally at a
slower speed.
3. The last case illustrated is the case of cut-off or coasting. At this time of the machine cycle the injected quantity required is zero. This typically happens when the
vehicle is descending along the slope and the engine is used only as an inertial
brake. It can be observed how the rail pressure remains constant at the set point
value the system is gas tight as the GMV is completely closed to stop the flow of
gas. This behavior is linked to the system layout which, as previously explained,
does not use a pressure relief valve.
Focusing further on the foot-off operation (case 2), in the lower picture of Fig. 7 the
performances of the Liebherr system is compared to the performances of a typical
hydrogen PFI system.

Liebherr’s Approach to Hydrogen Fuel Injection Systems    105

Fig. 7.  Simulated machine operation, variable load & constant speed (see above); foot-off
benchmark results (see below)

In order to compare the two systems, under the same conditions and in a fair
manner a benchmark test has been performed. The benchmark is made: at constant
engine speed, requesting a reduction of system pressure of 12 bar, and imposing the
same amount of quantity injected. Therefore the pressure gradient is purely linked to
the physical characteristics of the system and independent of the pressure controller.
From the image, it can be observed that the proposed layout is in the order of
magnitude, seven times faster than a typical PFI system.
Design a system capable of fast transient operation appears contradictory to the
requirements of having a system which maximizes the stability of the rail pressure at
the different engine operating points. Due to this trade-off between dynamic behavior
and pressure stability these two aspects have been carefully evaluated during the
system concept phase. The outcome of the benchmark is showed in the Fig. 8 where,
for the sake of brevity, only the constant full load operation is analyzed.
In a gas injection system, at sonic or choked condition the injected quantity is
directly proportional to the injection pressure (see also equation number (1). Thus any
deviation of the controlled pressure from the desired set point will generate a fueling
deviation.
The gas injection system is (similar to the Diesel system) a dynamic system in
which the pressure waves into the rail are a result of the external perturbations
(injection and feeding events) as well as the natural response of the system
(determined by its Eigen-values). It is physically impossible to reach the ideal
conditions in which the system pressure is perfectly constant. Therefore the target
becomes the minimization of the pressure oscillations.

106    R. Pirkl et al.

Looking at the pressure traces shown in Fig. 8 it is possible to observe that the
performance of Liebherr solution is very close to the performance of a typical PFI
system but at the first glance inferior in term of absolute delta pressure (peak – valley)
value observed in the considered measuring window. In reality, unlike Diesel systems
in which a significant part of the rail pressure perturbations are generated by the high
pressure pump; in hydrogen injection systems the gas is fed by the pressure regulator
which has the primary purpose of generating a constant outlet pressure, by acting in
response to the pressure variation generated in the rail itself. The system turns out to
be self-phased. Therefore the variation of absolute pressure observable over time is
not as important as the pressure drop happening during the injection events. Therefore
the two systems are equivalent since pressure variation is limited within a band of
0.2bar for both.

Fig. 8.  Evaluation of the pressure traces measured

Summarizing because of the benchmark results, it is possible to state that the
system layout proposed by Liebherr guarantees an excellent dynamic response,
superior to typical PFI systems, and an equivalent performance in terms of pressure
stability. Furthermore, since the system is designed in such a way that the hydrogen volume discharged at engine stop is minimized; it is possible to affirm without a
doubt that the Liebherr system has all the options to become attractive solution for the
market.
2.6 Hydrogen Injector Platform
Like most hydrogen (and other gas) injector concepts that can be found in literature,
the Liebherr hydrogen injectors for PFI and DI, unlike a hydraulically actuated diesel
injector, a direct actuated injector. Hydraulically actuated concepts for H2 injectors

Liebherr’s Approach to Hydrogen Fuel Injection Systems    107

also have been shown, but they will require a working fluid in order to open and close
the valve and thus lead to a much more complex system structure [6].
Liebherr’s directly actuated H2 low-pressure DI (LDI injectors) and PFI injectors
(LPI injectors) are designed to meet the necessary flow rate requirements of
heavy-duty engines. The outer dimensions of the LDI injector are close to those of
corresponding diesel injectors, particularly with regard to the critical maximum outer
diameter. Both injector platforms (PFI and DI) are based on the same technical design
platform regarding, actuation, sealing and valve guidance.
Injector overview. Main features and design characteristics of the Liebherr H2
injectors are shown Fig. 9.

Fig. 9.  Liebherr injector overview

Actuator. The injector is directly actuated to open and close the needle. The
magnet is thoroughly dimensioned to enable precise actuation (opening and closing)
and being at the same time small enough to stay within the requirement limits for
the injector outer dimension in the magnet area. From the beginning on, this process
has been accompanied by magnet simulations for different magnet concepts and
materials. In addition, different magnet materials and installation situations have
been simulated to adjust the magnet force to enable on the one hand a proper opening
of the injector and on the other hand to reduce the closing delay to a minimum.
The magnet is hermetically insulated from the hydrogen to prevent any hydrogen
embrittlement risk for the magnet materials
Injector nozzle. The injector nozzle of the LDI injector is equipped with a
diffusor cap to enable spray formation and bending. In the current sample stage this
diffusor cap is exchangeable. This enables cost effective testing cold chamber or
single cylinder engine testing of different variants in order to define the best fitting
solution. A screw on solution as shown in Fig. 10 realizes the exchangeability of the
diffusor cap.

108    R. Pirkl et al.

Fig. 10.  Exchangeable diffusor cap

Valve design. Liebherr developed a two-valve concept for the DI injector in which
an active cold valve controls the hydrogen flow and a passive hot valve decouples the
cold valve from the combustion heat and combustion pressure. For the PFI injector
only the cold valve is required thanks to the lower temperatures and pressures in the
intake manifold.
The cold valve is defining the injector flow rate by its flow cross section. The
flow cross section is opened and closed by means of the inward opening needle. The
guidance concept and material combination is optimized for dry running conditions.
The good wear behavior of this concept has already been shown in multiple
endurance tests on the hydraulic test bench and also in currently running engine
tests. To achieve an effective sealing against hydrogen the valve plate is additionally
equipped with an elastomer sealing element.
The hot valve is integrated in the nozzle of the injector. The hot valve is an
outward opening passive safety valve that is sealing against the engine cylinder. The
combustion chamber pressure hereby supports the sealing function.
Injector test results. One major development target is the achievement of very
low leakage rates. Testing with forming gas consisting of 95% nitrogen and 5%
hydrogen on a vacuum leak test bench shows very good results for the Liebherr
injector concept. The overall external leakage rate is 1*10–6 mbar*l/s. This means
that the injector is considered as gas tight.
Reason for the very good leakage test results is the new developed Liebherr twovalve injector concept.
The injection rates were measured on a function test bench using an injection
analyzer. The test medium was nitrogen and the measured values were subsequently
converted into hydrogen values. The backpressure on the injection analyzer test bench
is always set to 5 bar. Below described test results refer to the LDI injector.
The measured injection rate of the current sample stage show a good stability
as well as a good opening and closure behavior (see Fig. 11). A multiple step
optimization of the current profile also helped to achieve these results. The linear
injection quantity curve allows a precise quantity control at different pressure levels
(see Fig. 12).

Liebherr’s Approach to Hydrogen Fuel Injection Systems    109

The required maximum amount of 110 mg is reached at a rail pressure of 30 bar.
Injection rates are up to 15 g/s @ 40 bar rail pressure. The minimum quantity targets
of ~2.5 mg per stroke are achieved at a rail pressure of 10 bar.
All tests were performed with a dry running injector, without the addition of
lubricating oil. Several variants of the needle guide were tested in advance in order

Fig. 11.  Injection rate curve

Fig. 12.  Injection quantity curve (30 bar configuration)

to find a solution with good dry-running properties that did not show any significant
wear in endurance tests.

3 Summary and Outlook
From today’s perspective the hydrogen combustion engine will definitely be one
of the powertrain solutions to enable the transition to zero CO2. There are different

110    R. Pirkl et al.

hydrogen injection solutions in evaluation. PFI is definitely the fastest and less
complex way, but transient performance and power density will show a gap to more
demanding requirements. This gap is especially evident for the applications where
a hydrogen combustion engine will have the most benefits compared to a fuel cell
like non-road mobile machinery and special purpose vehicles. In order to provide
solutions for different requirements Liebherr develops components and system for
direct and port fuel hydrogen injection.
Beside the hydrogen system solutions for heavy duty engines, Liebherr also
develops a system with higher flow rates for large engines. Below Fig. 13 shows the
future Liebherr hydrogen product portfolio.
In addition to H2 applications, the products currently in the development pipeline
also have the potenzial to be used with other alternative fuels such as ammonia and
methanol, which are already being evaluated in preliminary tests.

Fig. 13.  Liebherr injector portfolio

Liebherr’s Approach to Hydrogen Fuel Injection Systems    111

References
1. Weiss, U.: The off-highway sector in the field of tension of future power train concepts –
Which chances has the internal combustion engine (ICE) in this industry?, International
engine congress Baden-Baden 2022 (2022)
2. Dreisbach, R., Arnberger, A., et al.: The heavy-duty hydrogen engine and its realization until
2025, 42nd International Vienna Motor Symposium (2021)
3. Korn, T.: The new highly efficient hydrogen internal combustion engine as ideal powertrain
for the heavy-duty sector, International engine congress Baden-Baden (2019)
4. Lejsek, D., et.al.: Thermodynamic analysis of hydrogen engines with port fuel- and direct
injection, 18th Symposi-um “Sustainable Mobility, Transport and Power generation” Graz
(2021)
5. Eichlseder, H., Klell, M.: Wasserstofff in der Fahrzeugtechnik, 4. Aufl. Springer Verlag
GmbH, ISBN 978-3-658-20446-4 (2018)
6. Nogami, M., et. al.: Development of a Common-rail Type High Pressure Hydrogen Injector
with a Large Injection Rate and an Ability of Multiple Stage Injection, 18th World Hydrogen
Energy Conference (2010)

Hydrogen Dosing Systems for Large Engines:
Challenges and Potentials of Three Different
Approaches
Enrico Bärow(*), Michael Willmann, Andreas Kühner, and Rick Boom
Woodward L’Orange GmbH, Porschestraße 8, 70435 Stuttgart, Germany
{enrico.baerow,michael.willmann,
andreas.kuehner,rick.boom}@woodward.com

Abstract.  Hydrogen dosing systems for large engines are available as low
pressure gas admission valves in the intake manifold, as mid pressure port fuel
or direct injection systems and high pressure dual fuel systems. Here the first
three options are used to operate the engine in an Otto-cycle mode, where the
last injection system allows a Diesel-like combustion process.
All four engine concepts have their validity in their individual application.
Key aspects for choosing one of the combustion technologies are system
and operating costs – strongly related to the tank and periphery technologies
needed to provide a certain system pressure for the dosing system.
The pressure and power range as well as the functionality of all systems,
operating conditions and limitations will be discussed. Main challenges in the
development and the practical application on an engine are shown as well as
the corresponding technical solutions.
Where available, combustion results will be shared to support the working
hypothesis for the selection of individual concepts / systems. A final conclusion
will indicate individual benefits and will give an outlook, which system has to
be expected on which engine application in the field.
Keywords:  Hydrogen injection · High pressure Dual Fuel · Direct injection ·
Port fuel injection

1 Introduction
The costs for electric power generation from renewable energy sources have been
plummeting for years. Today, they are cost competitive even without subsidies in
most regions of the world. Political instabilities in the world seem to accelerate the
trend towards non-fossil fuels – the price gap between fossil and non-fossil fuels is
narrowing [1].

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 112–122, 2023.
https://doi.org/10.1007/978-3-658-41477-1_8

Hydrogen Dosing Systems for Large Engines …    113

It can be expected, that over the next years Power-to-X (P2X) technologies will
follow a similar cost decline as they scale up rapidly [2]. Legislative initiatives mainly
in Europe and the USA are also targeting a massive acceleration of market maturation
for renewables and especially green Hydrogen. On the basis of green hydrogen, it will
be possible to create cost competitive synthetic fuels with a zero carbon footprint.
These synthetic fuels easily can be used to achieve a dramatic reduction of greenhouse
gas emissions, especially in applications, where a direct electrification and battery
electric drives are not feasible. From today’s perspective each of the P2X fuels will
have favorable applications where its fuel characteristics offer advantages over the
competing P2X fuels.

2 Applications for Hydrogen Use
The fuels shown in Fig. 1 are the most widely discussed P2X fuels. From the
perspective of being GHG neutral they all have one commonality: they need to
be produced from green hydrogen. The conversion of hydrogen to more complex
chemicals improves the ease of storage but also increases the total costs and energy
demand of production, especially if carbon capture is needed to generate hydrocarbon
fuels. It is shown in Fig. 1, how much fuel volume is necessary to store a certain
amount of energy. It can be seen, that for LNG, Methanol and Ammonia, the volume
is almost two to three times as large as for Diesel or heavy fuel oil (HFO). For hydrogen, both liquid and compressed, the required volume is much larger. The increased
efforts for the insulation of cryogenic or a pressure vessel for the compressed hydrogen worsens this ratio. The storage of hydrogen in a liquid organic hydrogen carrier
(LOHC) seems to require less volume than for the storage of pure hydrogen. But there
are still some challenges left: e.g. the storage of the discharged oil, the reconversion
yield, the necessary additional infrastructure (e.g. a hydrogen buffer tank) and the
high energetic efforts to release the hydrogen.

Fig. 1.  Required volume in liter of P2X fuel for storage of 10kWh (without storage tank)

114    E. Bärow et al.

Due to the high efforts for the storage of pure hydrogen, this fuel can preferably
be used in short to intermediate range applications (see Fig. 2). The maps show a
systematic approach to identify the best P2X technologies for different applications,
divided in mobile (top) and stationary applications (without grid connection; bottom).
Grid based power generation does not apply to this categorization methodology
because they offer “unlimited” range. For short range and especially highly utilized
applications direct electrification or fuel cells might be the best solution because the
high system efficiency outweighs the higher investment costs for fuel cells or battery
energy storage. Internal combustion engines, on the other hand, offer a higher power
density and higher efficiency at high loads – and are available at lower investment
costs.

Fig. 2.  Schematic segmentation for choice of P2X technology, top: for mobile applications,
bottom: for (off-grid) stationary applications

Hydrogen Dosing Systems for Large Engines …    115

3 Injection Technologies

Fig. 3.  Product portfolio of WLO injection technologies for different injection pressures and
fuels

Woodward offers a broad range of injection technologies that are designed or adapted
to inject synthetic fuels, either gaseous or liquid. Figure 3 gives an overview of
injectors and valves for different pressures and different fuels. Basically, these are
differentiated into:
– Low pressure gas admission systems including SOGAV gas valves and EPRS
pressure regulation valves,
– low pressure atomizers for liquid fuels,
– medium pressure injectors for gaseous fuels (plate valve) or liquid fuels,
– high pressure dual fuel injectors
The whole injector portfolio not only covers different fuels but also different injection
arrangements either as port fuel injection (PFI) or direct injection (DI) into the
cylinder head. All injectors either were developed in the past years or are currently
under development in different projects.
This publication focuses on the injection systems for hydrogen for internal
combustion engines (see Fig. 3, bottom line). As mentioned before, each application
brings its own requirements regarding the fuel, the fuel injection system and hence the
combustion process.
The main aspects to take into account are threefold:

116    E. Bärow et al.

– the source of hydrogen (defining e.g. pressure, temperature, phase state)
– requirements regarding injection system (retro fit vs. new-built, simplicity with
respect to actuation)
– engine requirements (power density, efficiency, transient operation requirements)
Figure 4 illustrates schematically the recommended correlation between available
infrastructure, engine requirements and the most feasible injection systems. Other
combinations would also be possible and mainly depend on specific boundary
conditions.
On the next chapters the different hydrogen injection systems will be explained.
3.1 High Pressure – DI

Fig. 4.  Correlation between hydrogen infrastructure, expected engine efficiencies/break mean
effective pressures (BMEP) and recommended injection system

For high pressure dual fuel applications Woodward L’Orange introduced the new
high-pressure dual fuel (HPDF) injector family [3, 4]. It allows the injection of both,
diesel fuel at up to 2200 bar and a second (carbon neutral) fuel at 600 bar directly
into the cylinder. Although the injector was designed for the injection of natural
gas as well as liquid fuels, with modifications it can be adapted for the injection of
hydrogen. The diagram in Fig. 5 (right) shows the differences in the energy injection
rate. For the same injection pressure, operation conditions and spray configuration
the diagram compares the amount of energy that is injected for a certain injection
duration – basically taking into account the different velocities inside the spray holes,
the different state of the fuel (gaseous vs. liquid) and the different energy densities
of the fuel. Although of these differences it can be seen, that for 500 bar the injected
amount of energy with hydrogen is of the same order as for Methanol for the same
spray configuration.

Hydrogen Dosing Systems for Large Engines …    117

Fig. 5.  HPDF injector with 3-1-needle arrangement (left). Comparison of achievable engine
powers only with respect to chemical and physical properties of the fuels (right)

The high-pressure direct injection of hydrogen allows a diesel like combustion
process. On one hand this would allow highest efficiencies, best transient behavior
and high power densities. But this might also lead to one of two major drawbacks:
1. the inhomogeneous mixture would lead to a combustion close to stoichiometric
conditions. As a result, this would lead to high flame temperatures and hence to
a strong formation of NOx. With those emissions to be limited, this might have a
negative effect on cylinder power and efficiency.

Fig. 6.  Comparison of required power for compressing different fuels to high pressure

2. The compression of hydrogen to very high pressures is very energy demanding.
Figure 6 illustrates the theoretical power for different fuels to be compressed to a
certain pressure. The compression power is normalized by the mechanical engine
power output. It can be seen, that due to the lower energy content of the P2X fuels
the compression power is higher than for diesel fuel, even at lower end pressures.
It can also be seen, that for hydrogen there is a big difference, if hydrogen is either
cryogenic/liquid or gaseous at high pressure. Assuming a tank filling of 20% of
a 700 bar storage tank, which equals roughly 100 bar remaining tank pressure, a
compression to 600 bar injection pressure would be about three times as power
consuming as for liquid hydrogen. Higher tank fillings (tank pressures) would
reduce the required power/energy, whereas lower tank pressures would increase it
even further. In [8] it is shown, that the actual compression energy for hydrogen is
mostly twice as much as the theoretical.

118    E. Bärow et al.

Due to these two reasons a critical view on the field of high power density application
should be made when choosing a hydrogen HPDF concept. The right infrastructure
(availability of high pressure Hydrogen or liquid Hydrogen and Cryogenic pumps)
and a well-tuned combustion process (i.e. including exhaust aftertreatment) are
inherent for the use of HPDF with hydrogen.

Fig. 7.  Medium pressure injector for hydrogen injection (plate valve design)

3.2 Medium Pressure – DI & PFI
Woodward L’Orange has designed a new injector, especially made for the injection
of hydrogen (see Fig. 7). It features a direct electric actuation and can be operated
with system pressures of 15–60 bar. Due to the low density of hydrogen, a large cross
section in the needle seat needs to be opened. Therefore, the injector is designed
with a plate valve nozzle. It can be applied in two different ways: either for port fuel
injection (PFI) or direct injection (DI) in the cylinder head (see Fig. 8). Both arrangements have their benefits and disadvantages.
Arranged for port fuel injection, it allows a good mixing of the injected hydrogen with the air to achieve a homogeneous mixture. This can be achieved by placing
the injector close to the engine’s inlet valves and by optimizing the injection timing
towards mixing. This results in a low knock-tendency and can help to minimize
the risk of back fire. Also, with a homogeneous and lean mixture, the engine can
be operated at very high efficiencies and with almost no formation of nitric oxides.
Besides, as it does not require any space in the cylinder head, the mixture can be
ignited by a central ignition source (e.g. spark plug). This concept would be perfect
for the retrofit of an engine because the cylinder head does not need to be modified.
The downside of the PFI arrangement is a reduced cylinder filling with air – and so a
lower power density.

Hydrogen Dosing Systems for Large Engines …    119

On the other hand, the injector can be used for direct injection. In this
configuration, the injector is integrated into the cylinder head and the injection timing
is limited to a smaller time frame. There is less time for mixture formation until top
dead center, which will result in a more inhomogeneous mixture, higher NOx and
slightly higher risk of knocking. But this more extensive arrangement leads to some
other benefits. The injection timing and cylinder pressure conditions during injection
allow a higher filling rate (air+fuel) in the cylinder – and so a higher power density.
Due to the shorter delay between injection and ignition, the danger of pre-combustion
due to oil leakage in the cylinder is reduced. Last, but not least, this arrangement is
not sensitive to back firing and can perform faster load transitions (Fig. 9).

Fig. 8.  Arrangement of medium pressure injector for PFI or DI

3.3 Low Pressure – PFI

Fig. 9.  
Solenoid operated gas admission valve (SOGAV), left, and electronic pressure
regulation system (EPRS), right

120    E. Bärow et al.

For many years the SOGAVTM (Solenoid Operated Gas Admission Valve) port fuel
gas admission valves have been used for applications where gaseous fuels need to be
injected at low pressure upstream of the cylinder intake valves. The pressure range
typically is from 5 to 10 bar. It offers a direct actuation and so it is an easy to use
and well proven robust design gas admission valve. In combination with the ECU
(Woodward LECM) it is possible to detect open and closed position with a so called
virtual sensing capability. This virtual sensing makes it possible to monitor in real
time the position of the valve position. This makes it a good solution for pure gas or
dual fuel engines. Such low pressure dual fuel engines rely on Woodward L’Orange’s
diesel injectors as ignition source. These are able to inject smallest quantities of fuel
with extreme precision while also being able to run at full power in diesel-mode.
SOGAV has been designed to be used in a natural gas and LNG environment.
With hydrogen and also ammonia port fuel gas admission systems, the SOGAV fuel
gas specification has been expanded into the hydrogen and ammonia applications. The
so- called hydrogen and ammonia hardened SOGAV are available for alternative fuel
applications.
The SOGAV operating principle is based on a differential pressure across the
valve. With load changes of the engine the charge air pressure of the inlet manifold
is varying. In order to control the differential pressure this requires active pressure
control of the fuel gas pressure. The differential pressure across the SOGAV is also
used as a control parameter to control the opening duration and optimize the opening
duration depending on actual load conditions. I.e. at idle conditions the SOGAV
opening would be very short when the nominal differential pressure is maintained
and could cause combustion stability issues. Since the introduction of SOGAV these
pressure regulators have been based on the mechanical design diaphragm zeropressure regulators with a manipulated back pressure.
Modern engines require improved dynamic behavior, which drives more accurate
and faster pressure regulation. Alternative fuels, like hydrogen, require even more
accurate and fast pressure control. Woodward introduced an Electronic Pressure
Regulation System (EPRS) that meets the increased pressure control accuracies.
It has a model based control that includes the volume of the gas manifold, in order
to optimize the gas pressure response without over- or undershooting the pressure.
The EPRS is called a system as it interacts with the SOGAV command signal as a
feedforward parameter. A mechanical zero pressure regulator can only act upon
a pressure change, while the EPRS already acts on the SOGAV command signal
change. This is especially important for the hydrogen low pressure port fuel (i.e.
SOGAV) application.

4 Summary
The introduction of green hydrogen as an energy carrier for mobile and stationary
large engine applications is one of the major steps in the next years. Depending on
the specific application a broad range of combustion concepts are feasible due to the

Hydrogen Dosing Systems for Large Engines …    121

favorable fuel properties of hydrogen. Woodward L’Orange provides a broad range
of injection systems for those combustion concepts from highspeed engines with
≈100 kW/cylinder up to >1000 kW/cylinder.
A platform family concept with broad customization options allows efficient
development for each injection systems while also fully meeting the specific needs of
each individual engine concept.
Acknowledgements. The authors would like to thank all the Woodward members who closely
collaborate in the development of the future P2X injector families. We especially appreciate
the support we receive from Woodward’s Technology group (know-how and guidance in
combustion technology, simulation and design) and Woodward’s rapid prototype network,
especially: Greg Hampson, Domenico Chiera, Jessica Deblois, James Wood, John Karspeck
and Chuck Brennecke.

Abbreviations
DI	Direct injection
EPRS	Electronic pressure regulation system
HPDF	High pressure dual fuel
HPDI	High pressure direct injection
LNG	Liquefied natural gas
LH2	Liquefied hydrogen
LOHC	Liquid organic hydrogen carrier
MPI	Medium pressure injection
PC	Pre-combustion chamber
PFI	Port fuel injection
SOGAV	Solenoid operated gas admission valve

References
1. https://www.handelsblatt.com/unternehmen/erneuerbare-energien-gruener-wasserstoff-istzum-ersten-mal-guenstiger-als-wasserstoff-aus-erdgas/28251636.html. Accessed 27 April
2022
2. Zhao, Y., Setzler, B., Wang, J., Nash, J., Wang, T., Xu, B., Yan, Y.: An Efficient Direct
Ammonia Fuel Cell for Affordable Carbon-Neutral Transportation. Joule (2019). https://doi.
org/10.1016/j.joule.2019.07.005
3. Senghaas, C., Willmann, M., Berger, I.: New injector family for high pressure gas and low
caloric liquid fuels. 29th CIMAC Congress Vancouver (2019)
4. Bärow, E., Willmann, M., Aßmus, K., Redtenbacher, C., Wimmer, A.: Operating Experience
with a Combined High-Pressure Gas-Diesel Platform Injector. In Eichlseder, H. (ed.), 17.
Tagung Der Arbeitsprozess des Verbrennungsmotors: 17th Symposium The Working Process
of the Internal Combusition Engine, vol. 103, pp. 141–153. Verlag der Technischen Universität Graz (2019) (IVT-Mitteilungen)

122    E. Bärow et al.
5. Senghaas, C., Bärow, E.: Woodward L’Orange’s New Injector Generation – An Adaptable
Injector Family for Future Fuels. 9th AVL Large Engines TechDays, April 21–22, 2021
(2021)
6. Gleis, S., Frankl, S., Prager, M., Wachtmeister, G.: Optical analysis of the combustion of
potential future E-Fuels with a high pressure dual fuel injection system. June 23rd/24th,
2020, Kurhaus Baden-Baden (2020)
7. Frankl, S., Gelner, A., Gleis, S., Härtl, M., Wachtmeister, G.: Numerical Study on Renewable
and Sustainable Fuels for HPDF Engines, Proceedings of the 2020 28th Conference on
Nuclear Engineering, ICONE28-POWER2020–16438 (2020)
8. Gardiner, M.: Energy requirements for hydrogen gas compression and liquefication as related
to vehicle storage needs. DOE Hydrogen and Fuel Cells Program Record #9013. Department
of Energy, USA (2009)

Hydrogen Storage Technologies
Mathias Keck(*), Dirk Bessey, Frank Buehler, and Manuel Eugen Faiß
BIN Boysen Innovationszentrum Nagold GmbH & Co. KG, Nagold, Germany
{mathias.keck,dirk.bessey,
frank.buehler,manuel.faiss}@bin.boysen-online.de

Abstract.  In order to achieve the globally agreed carbon dioxide reductions
for a heavy-duty vehicle (HDV) several technologies are currently pursued,
battery electrical vehicles (BEV), fuel cell electrical vehicles (FCEV),
hydrogen-internal combustion engines (H2-ICE) and e-fuels. The total cost of
ownership (TCO) is a very important aspect in the transportation sector. The
fuel cell electrical vehicle (FCEV) is a very interesting alternative powertrain
technology especially for long haul applications. The hydrogen storage
technology is a key success factor to achieve the overall TCO requirements.
The paper describes the specific requirements and challenges for a hydrogen
storage system in a HDV application and compares the three different hydrogen storage technologies, such as compressed gaseous hydrogen (CGH2),
liquid hydrogen (LH2) and cryo-compressed hydrogen (CcH2).
Keywords:  Hydrogen · Heavy duty hydrogen storage tank system · Fuel cell
electrical vehicle

1 Introduction
The global reduction of carbon dioxide emissions is a great challenge in the next
decade. In [1] the breakdown of the carbon dioxide emissions by countries is shown.
Carbon dioxide emissions from Russia, Japan, EU27, USA, China and India have
remained at very high level over the last 20 years and have to be reduced.
The European Union has committed itself to reduce carbon dioxide emissions
until 2050 to be climate-neutral [2]. The decarbonization of the transportation sector
is an important part of the solution. Currently most heavy-duty vehicles (HDV) are
powered by a diesel combustion engine and thus emit exhaust gas. To reduce the
carbon dioxide emissions the diesel combustion engine has to be replaced by an alternative powertrain. There are different technical approaches for alternative powertrains
pursued. First of all, internal combustion engines (ICE) could be powered in future
with e-fuels or hydrogen. Specially e-fuels could be an interesting solution for the
existing fleet using the existing infrastructure. However, also the use of hydrogen for
ICE’s is an interesting solution, while only some adaptions on the current combustion
technology is necessary. Battery electrical vehicles (BEV) and fuel cell electrical

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 123–130, 2023.
https://doi.org/10.1007/978-3-658-41477-1_9

124    M. Keck et al.

vehicles (FCEV) are currently more or less equally discussed for future alternative
powertrains. Both solutions have their advantages and disadvantages. For electrical
powered HDV’s also catenary systems are under investigation for some niche
applications. Besides the range, usage and the infrastructure, the total cost of ownership (TCO) plays a very import factor for the owner. Furthermore, it is imperative that
whatever powertrain solution is considered that the energy carrier (battery, hydrogen, e-fuel) is produced by a renewable source, such as photovoltaic and/or wind
power. Green hydrogen provides an enormous potential to become the energy carrier
for the future. Therefore, the appropriate storage technology for hydrogen is keen
to achieve the overall requirements. The three most commonly discussed hydrogen
storage technologies will be presented and their advantages and disadvantages will be
discussed

2 Hydrogen Storage Technologies
2.1 Hydrogen
Hydrogen is the smallest atom in the periodic system and one of the smallest
molecules. At ambient pressure of 1 bara and ambient temperature of 293 K hydrogen
has a gaseous state with a very low density of about 0.082 kg/m3. Accordingly, to this
property hydrogen has to be stored at high pressure and / or at low temperature. For
low temperatures in cryogenic area, it is also important to differentiate between orthoand parahydrogen because of the different material properties. The critical point of
normal hydrogen is at a temperature of about 33 K and a pressure of about 13 bara.
Figure 1 shows the phase diagram of normal hydrogen [3] and the application area in
the different storage technologies of subcooled/liquid hydrogen (sLH2/LH2) [4], cryocompressed hydrogen (CcH2) [5, 6] and compressed gaseous hydrogen (CGH2).

Fig. 1.  Phase diagram of normal hydrogen [3]

Hydrogen Storage Technologies    
125

In [7], a comparison of different technologies for energy storage is shown. An
important difference between the individual storage technologies is the specific
energy, the energy density of the system and the storage medium. In general, the
specific energy density of batteries for automotive applications is very low. In
comparison gasoline has a very high energy density and specific energy. The highest
specific energy can be achieved with hydrogen. Whereas gasoline storage systems
have a low system mass, hydrogen storage systems have a heavy composite pressure
vessel or a vacuum insulated pressure vessel made of stainless steel. But because of
the high specific energy of hydrogen the hydrogen storage tank system has a higher
specific energy. According to the state of the art technology battery storage systems
are not suitable for long haul trucks which have to cover long distances of up to
800 km per day without a stop for refueling. In the following the different hydrogen
storage technologies for commercial vehicles will be presented.
2.2 Liquid Hydrogen Storage System
State of the art for cryogenic hydrogen storage systems is the liquid storage technology. Liquid hydrogen at ambient pressure of 1 bara and a temperature of 20 K has a
very high density of over 70 kg/m3. But liquid hydrogen storage systems (LH2) have
the disadvantage of two receptacles at each vessel because the gaseous hydrogen must
be returned to the fueling station. The subcooled hydrogen storage tank (sLH2) has
only one receptacle at each vessel and it is possible to refuel both vessels through one
receptacle, because there is no necessity to return the gaseous hydrogen.
The particular challenge at liquid hydrogen or subcooled liquid hydrogen storage
systems is the low storage temperature through a passive cooling technology. The
storage systems consist of two vessels, an inner and an outer vessel which are made
of stainless steel. Between the inner and the outer vessel is a vacuum gap which is
filled with a multi-layer insulation (MLI) to minimize the heat input into the inner
tank. Also, the bearing and the piping in the vacuum gap needs to be optimized for
low heat input. The lower the heat input, the longer is the so-called hold time of the
tank, which is the time until the boil-off of the tank starts. When the temperature of
liquid hydrogen increases through the heat input as described above, it results also
in a pressure increase within the inner vessel. If the pressure reaches a critical level,
the hydrogen is released through a pressure relieve valve (PRV) out of the tank into
the boil-off management system (BOMS). The boil-off management system converts
the hydrogen into water. Additional challenges are the cryogenic components like
the receptacles, cryogenic valves, pressure relief valves or heat exchangers. Figure 2
shows a tank assembly of a sLH2 storage tank and the BOMS from BOYSEN. The
diameter of the outer vessel is about 710 mm, the length is about 2500 mm and the
inner vessel has a net volume of about 750 l.

126    M. Keck et al.

Fig. 2.  Subcooled liquid hydrogen storage system (front) and boil-off management system
(top)

This configuration enables a usable storage capacity of over 40 kg for each storage
tank depending of the operation strategy. The cryogenic connecting pipe, which is
also vacuum insulated, allows a refueling of both storage tanks via one receptacle.
The vacuum piping and the vessels have not only to be designed for the maximum
allowable working pressure, both of them have to be designed for the mechanical
stress through excitation caused by the road load. Figure 3 shows a test rig which is
simulating a dynamic road load and the test rig for the boil-off management system.
The dynamic road load test rig consists of seven hydraulic cylinders which are fixed
at a frame of a heavy-duty truck. By means of this test rig it is possible to simulate
a real dynamic road load which is measured on the road and adapted to the test rig.
The boil-off management test rig enables to simulate the boil-off event without a
storage tank system. Furthermore, it is possible to measure the tail pipe emissions,
the mixture of hydrogen and the air in front of the catalyst in every condition of
the boil-off management system. The pressure loss and temperature, including the
investigations of the light-off behavior of the catalyst, are important metrics for the
design of the boil-off management system.

Fig. 3.  LH2 storage system on test rig (left) and boil-off management test rig (right)

Hydrogen Storage Technologies    
127

2.3 Compressed Gaseous Hydrogen Storage Technology
State of the art for compressed hydrogen storage systems (CGH2) in heavy-duty
trucks is a pressure of 350 bar. Currently refueling procedures for this pressure level
are existing. At 350 bar and a temperature of 288 K the density of normal hydrogen
increases to 24 kg/m3.
Most pressure vessels for this application are type IV vessels. The basic design
concept of a type IV pressure vessel is shown in Fig. 4. The inner liner is made of
plastic to secure the gas tightness, while the outer skin is made of a carbon fiber to
build the load bearing structure.

Fig. 4.  Compressed gaseous hydrogen vessel type IV

Refueling procedures for 700 bar storage tank systems for heavy duty trucks above
10 kg hydrogen are in discussion. To increase the usable hydrogen mass BOYSEN
is developing a compressed hydrogen storage tank system including the vessel for an
application with 700 bar for heavy duty trucks. By increasing the pressure from 350
to 700 bar the density of hydrogen increases by 67%. Particular challenges include
the development of an on-tank valve (OTV) and a high-pressure regulator (HPR) for
a 700 bar application with high mass flow rates. Also, the sealing design between
the on-tank valve and the boss is an important part of the development. Furthermore,
the high control accuracy at low pressure and high mass flow rate is key to achieve.
Compared to the liquid hydrogen storage system the energy density is much lower.
Therefore, a frame for up to three or four vessels behind the driver’s cabin is needed
to guarantee an appropriate range of the vehicle. Figure 5 shows an example of a
concept with six vessels.

128    M. Keck et al.

Fig. 5.  Compressed gaseous tank system (CGH2)

2.4 Cryo-compressed Hydrogen Storage System
Cryo-compressed hydrogen storage systems (CcH2) for automotive applications
became famous with the investigations of BMW [6]. BMW has developed a storage
tank system with an operating pressure below 350 bar. The lowest temperature which
could be achieved was 33 K. Today the cryo-compressed hydrogen storage technology is still under development. New investigations [5] show that the pressure can
be increased to 450 bar. A particular challenge is the combination of high pressure
and low temperature. The system needs an inner high-pressure vessel and an outer
vessel to build a gap which is filled with a multi-layer insulation in vacuum ambient.
Because of the high-pressure level, the boil-off event is much lower prioritized as by
the liquid hydrogen storage system. BOYSEN is following the interesting development but there are only activities in preparation to investigate a proof of concept.

Hydrogen Storage Technologies    
129

3 Comparison Between the Different Storage Technologies
As described in Sect. 2.1 every storage technology has its advantages and disadvantages. Following in Fig. 6 the three hydrogen storage technologies will be
compared. Purely looking at the density, the CcH2 tank system seems to be the most
favorable one. Furthermore, a boil-off management system is currently not needed.
However, it is also the most complex and expensive system. Besides the density the
net volume carrying hydrogen is also very important.
The huge advantage of the CGH2 technology is that the infrastructure is already
existing. Meaning hydrogen refueling stations (HRS) for passenger cars (700 bar)
and commercial vehicles (350 bar) are available and even further developed. To be
an adequate solution for applications which require a medium range, definitely
700 bar tanks and the appropriate filling protocol for commercial vehicles needs to
be established. Furthermore, the design of OTV must be more robust against leakage
as a result of wear and tear. Nevertheless, to carry a higher amount of hydrogen per
vehicle, there are multiple numbers of tanks necessary in comparison to the sLH2 or
CcH2 technology.
The sLH2 technology offers from BOYSEN’s point of view the highest potential
for long-haul applications and/or applications with a high-power demand. The
TCO is superior against the CGH2 and CcH2 technology looking at the total vehicle
architecture and system cost. From a technical standpoint the heat loss needs to be
absolutely minimized by an extraordinary vacuum insulation in combination with
MLI. Furthermore, the bearing concept between the inner and outer tank needs to be
designed to carry the load on one hand and to minimize the heat transfer from the
outer into the inner tank. These measures sum up in a long hold time of the system.
However, a boil-off management system as a safety device is still required.

Fig. 6.  Comparison between the three common storage systems

130    M. Keck et al.

4 Conclusions + Outlook
Green hydrogen is one of the key elements for carbon dioxide-free road transport. The
development of suitable components for hydrogen application is one of the biggest
challenge which requires a high competence of the individual supplier. Components
such as cryogenic valves or high-pressure valves face automotive suppliers with new
challenges, but also offer them opportunities.
Another challenge is the certification of the hydrogen storage systems. In addition
to hydrogen compatible of the sensor, a safety system such as a BOMS are very
important developments in the future.
One of the most important advantages of hydrogen storage systems is the short refueling
process. The hydrogen storage system enables to cover long distances without long
refueling stops. When refueling is required, the sLH2 and CGH2 storage systems should be
refueled in the same timely matter like a diesel heavy-duty vehicle. For the use of hydrogen in the future, it is important that a corresponding infrastructure is quickly established. A
standardized refueling process and components such as the receptacle for refueling hydrogen storage systems are also very important for rapid acceptance of the technology. Both
the CGH2 700 bar technology and the sLH2 technology reduces the packaging problem in
vehicles for long ranges and offer a zero-emission strategy for long haul use.
There are several ways to achieve the zero-emission goal for commercial vehicles.
Hydrogen has a very high potential to become the future energy storage. But hydrogen needs to be produced from renewal energy such as photovoltaic and wind power.
Furthermore, the appropriate infrastructure needs to be established fast.
Given these premises and a technology open competition, BOYSEN strongly
believes that we will see CHG2 and sLH2 technologies in the market based upon
the individual application of the commercial vehicle. Therefore, BOYSEN develops
intensively the 700 bar CGH2 technology and the sLH2 technology further to achieve
market readiness within this decade.

References
1. Crippa, M. et al.: GHG emissions of all world countries, JRC SCIENCE FOR POLICY
REPORT, 2021 Report (2021)
2. The European Parliament and Council of the European Union.: Regulation (EU) 2021/1119
of the European Parliament and of the Council of 30 June 2021 establishing the framework
for achieving climate neutrality and amending Regulations (EC) No 401/2009 and (EU)
2018/1999 (‘European Climate Law’), Brussel (2021)
3. Lemmon, E.W., Huber, M.L., McLinden, M.O.: NIST REFPROP, Reference Fluid Thermodynamic and Transport Properties, NIST Standard Reference Database 23, Version 9.1 (2013)
4. Maus, S., Stanzel, N., Schäfer, S.: Clean Energy Partnership (CEP), Whitepaper process
sLH2 (2021)
5. Brunner, T., Forstner, C., Cardella, U.: Clean Energy Partnership (CEP), Whitepaper process
CcH2 (2021)
6. Kunze, K., Kircher, O.: Cryo-Compressed Hydrogen Storage. Cryogenic Cluster Day, Oxford
(2012)
7. Wunderlich, P.: Electrodes for Lithium-Oxygen Batteries, Dissertation, RWTH Aachen, (2019)

Well-to-Wheel CO2-Analysis of Different
Powertrain Systems on Representative
Heavy-Duty Mission Profiles
Nicolas Hummel(*), Tim Herold, and Christian Beidl
Institut für Verbrennungskraftmaschinen und Fahrzeugantriebe, Technische
Universität Darmstadt, Darmstadt, Deutschland
{hummel,beidl}@vkm.tu-darmstadt.de

Abstract.  Im Zusammenhang mit dem Ziel der Europäischen Kommission
bis 2030 die CO2-Emissionen um 55 % zu senken, ist besonders der Verkehrssektor gefordert alternative Antriebsysteme für unterschiedlichste
Anwendungen zu entwickeln.
Der Fokus dieser Arbeit liegt dabei auf der Untersuchung unterschiedlicher
Antriebssysteme in der Nutzfahrzeuganwendung. Neben dem Standard Dieselantrieb werden die Well-to-Wheel CO2-Emissionen eines Erdgasfahrzeugs,
eines batterieelektrischen Fahrzeugs und eines Wasserstoffverbrenners in verschiedenen Anwendungsfällen untersucht. Die Lastprofile erstrecken sich von
dem hochspezifischen Müllsammelbetrieb bis hin zur Warenlieferung auf der
Langstrecke. Dabei spielen die Zuladung und die Topographie ebenfalls eine
wichtige Rolle.
Methodologisch erfolgen die Untersuchungen aller Fahrszenarien nach
folgendem Schema. Eine im realen Fahrzeug aufgenommene Strecke wird in
die Virtuelle IPG-Truckmaker Umgebung überführt. Anschließend wird das
Modell des realen Referenzfahrzeugs erstellt und die erzeugte Simulationsumgebung wird anhand des Vergleichs von realen und simulierten Kraftstoffverbräuchen validiert. Das validierte Referenzfahrzeug dient von nun an als
Basis für den Aufbau der Nutzfahrzeuge mit alternativen Antrieben. In diesem
Zusammenhang können die Verbräuche der unterschiedlichen Antriebssysteme
in den ausgewählten Anwendungen verglichen werden. Zusammen mit den
ermittelten Well-to-Wheel Emissionen der betrachteten Energiespeicher erfolgt
schließlich die Bewertung des Anwendungsspezifischen CO2-Impakts der
einzelnen Antriebssysteme.
Das Ergebnis dieses Vergleichs ist neben der variierbaren Simulation durch
veränderte Randbedingungen oder Streckenprofile auch die Beantwortung
der für Logistikunternehmen aller Art relevanten Frage des ökologischen
Optimums. Sie dient im hierbei spezifischen Fall als Orientierungshilfe, um bei
einer möglichen Flottenanpassung den ökologischen Aspekt stärker zu berücksichtigen.
Keywords:  Alternative powertrains · Well-to-Wheel · CO2 · Simulation

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 131–142, 2023.
https://doi.org/10.1007/978-3-658-41477-1_10

132    N. Hummel et al.

1 Motivation
The mandatory goals set by the European Commission in the context of the “Green
deal” [1] requires a minimum reduction of the greenhouse gas emissions (GHG)
by 55% until 2030 in comparison to 1990 and climate-neutrality until 2050. These
goals apply to all sectors and especially in the mobility sector, new technologies and
methodologies to asses and rate energy paths are required to reach those goals.
In Europe, the transport sector causes about 25% of the GHG-emissions. Figure 1
displays the repartition in Germany: Heavy-duty emissions represented 8% of the
GHG-emissions in 2020. In this context a reduction of 42% has to be achieved in
the transportation section until 2030 while the prognosed heavy-duty traffic volume
is supposed to increase by 16.8% from 2020 to 2030 [2]. Those contradictory goals
underline the necessity to develop new energy concepts for the heavy-duty sector.

Fig. 1.  Greenhouse gas emissions in Germany 2020 [3]

In the current rating procedure of the CO2-impact of a powertrain system, only
the vehicle use is considered. In this context, 2019 a part of the heavy-duty trucks
fleet, representing about 65–70% of the sectors CO2-emissions, where monitored and
embedded in the VECTO simulation tool. This procedure is used as a benchmark for

Well-to-Wheel CO2-Analysis of Different Powertrain …    133

the future evaluation of Tank-to-Wheel CO2-emission reduction [4]. Nevertheless, in
order to evaluate and compare the climate-impact of different powertrain systems,
a comprehensive, system-embracing analysis has to be carried out. In this context
the production of the energy carrier plays a major role, as some propulsion systems
such as battery electric or hydrogen driven vehicles have zero Tank-to-Wheel CO2emission as the energy carrier does not contain any carbon atom. Nevertheless,
depending on the production path, the production of electricity and hydrogen can have
a significant CO2-emission factor. In the scope of this work, the production path of
every energy carrier is considered in order to compare the powertrains on a Well-toWheel basis and therefore allow a systematic rating of the CO2-impact of a powertrain
system. Furthermore, different assumptions for future energy production and CO2impact are made and applied to three different heavy-duty mission profiles.

2 Methodology
The reference vehicle, in this case a conventional EURO VI Diesel truck, is monitored
during real-life use on typical mission profiles using a data-logger with GPS-antenna
and CAN-communication to ECU-datasets. The GPS signal is then used to build a
digital copy of the route in IPG-Truckmaker. Depending on the GPS signal quality,
the altitude profile has to be smoothed or complemented with accurate datasets (i.e.
DGM) in order to obtain a continuous, drivable digital route. The recorded speed
profile of the real truck is then implemented on the digital track in form of speed
limitation. This v/s-approach secures the correlation of desired vehicle speed and altitude profile regardless of the powertrain and actual speed. In this context different
powertrains can evolve at different velocities over the same routes according to their
power reserve.
In the next step, a digital model of the truck is embedded in IPG-Truckmaker. In
this context the vehicle data and the engine maps are implemented. In the following
simulation validation step, real and simulated fuel consumptions are compared and
if necessary the simulation parameters corrected. At this point the validated vehicle
parameters allow to replace the standard Diesel engine with other powertrain
configurations and simulate them on the same route. The different driving resistances
and consumption values represent the output values of simulation environment and
are checked regarding their plausibility on the basis of the simulated vehicle weights
and expected powertrain efficiencies.
The multiplication with the Well-to-Wheel values of the different powertrains and
the assumed productions scenarios enable a comparison of powertrains systems on
different missions according to their CO2-impact (Fig. 2).

134    N. Hummel et al.

Fig. 2.  Methodological approach

The described methodology allows a comparison of energy systems without
having the real hardware onsite for real-life testing. In this context it is possible to
choose the most ecological powertrain according to specific mission requirements.

3 Simulation Environment
The simulation of the reference vehicle requires basic information’s about weight,
powertrain topology, gear ratio and engine data. Once the parametrization of the
hardware is complete, the shifting strategy is also embedded in the IPG-Truckmaker
environment and the comparison of real and virtual datasets allow to trim the vehicle
simulation parameters in several iteration steps. In the case of special applications
such as city buses or refuse collection vehicles, the weight variations and auxiliary
power demand during stops must be implemented via maneuver parametrization.
The comparison of the fuel consumptions in reality and in the simulation is used
to validate the reference vehicle model. Figure 3 displays the real and the simulated
Diesel consumption on a refuse collection route. The cumulated real and simulated
fuel consumptions match in the driving parts of the route as well as in the refuse
collection parts. In this context it can be assumed, that the assumptions made for the
vehicle and powertrain parametrization, the auxiliaries and the driving maneuvers
reflect real driving behavior.

Well-to-Wheel CO2-Analysis of Different Powertrain …    135

Fig. 3.  Real and simulated fuel consumptions on a refuse collection route

Based on the validated reference Diesel truck, further powertrain topologies are
implemented and parametrized in IPG-Truckmaker. According to the adapted weight
a plausibility check based on the cumulated motion energy is executed: the motion
energies of all powertrains should be in the same range but differ slightly according to
the vehicle weights.
The powertrain efficiencies define the consumption on each vehicle type on the
considered route. Figure 4 displays the consumptions of alternative powertrains on
three different heavy-duty mission profiles.

Fig. 4.  Consumption of different powertrain type on typical heavy-duty mission profiles

• In the waste collection route all three internal combustion engines have similar
consumptions (Diesel ~ 510 kWhDiesel) about twice as high as the battery electric
truck (BEV ~ 250 kWhElectric). On this mission type, with a mean velocity of
9.8 km/h and a standstill timeshare of 60% over the 60 km route, the combustion

136    N. Hummel et al.

engine vehicles are mainly operated in engine map areas with poor efficiencies.
On the other hand, the BEV is benefiting from a clutch free starting, a higher
powertrain efficiency at low power/torque output and regenerative braking.
• On the delivery route with a mean velocity of 37 km/h, a positive elevation of
18 m/km and v*apos about 1 m/s2 the advantage of the BEV is still significant
in terms of efficiency due to the challenging altitude profile of this route.
Nevertheless, the flatter the elevation profile gets, the lower the difference with be.
• The simulated long-haul route has very little elevation and a mean speed of
61 km/h over 90 km. In this type of use-case the Diesel truck consumes about 1/3
more than the BEV (~275 kWhDiesel vs. ~185 kWhElectric). The BEV still benefits
from lower powertrain losses but due to the constant speed travelling, only little
energy can be recuperated during braking.

4 Emission Factor of Different Energy Carrier
The equivalent CO2-emissions of an energy carrier can be divided into two main
phases:
• Well-to-Tank for the production path (WtT)
• Tank-to-Wheel for the usage of the energy carrier in the vehicle (TtW)
The sum of both terms results in the Well-to-Wheel (WtW) emission factor. In order
to compare energy systems in a systemic way, the WtW approach is the one to use as
the allocation of CO2-emissions is not equally distributed on WtT and TtW and differs
depending on the energy carrier.
4.1 Well-To-Tank
The Well-to-Tank part includes the extraction at the source, the transport, the
manufacturing and the distribution of the energy carrier. The values for the fossil
sources diesel and gas are taken from the JEC Well-to-Tank report v5 and are
respectively 62 g CO2,eq/kWhDiesel and 59 g CO2,eq/kWhCNG [5]. They are no major
technological improvement expected in the Well-to-Tank path for fossil sources. Only
the share of the transport to the marker could be reduces in the future if regenerative
fuels are used in the tanker ships which carry crude-oil to Europe.
Electricity is required for the production of all other energy carrier. In this context,
the emission factor of the energy sector is a major lever for CO2-neutral mobility.
In the scope of the European Green-deal and the “Fit for 55” announcement from
July 2021 [6], a CO2-emission reduction of 55% on the basis of the 1990 value
(764 g CO2,eq/kWhElectric) should lead to a maximum emission factor for the German
electricity mix of 343 g CO2,eq/kWhElectric in 2030 [7]. The scenarios for 2050 are
based on the goals of the European commission with 95% emission reduction
compared to 1990 and would lead to an emission factor of 23 g CO2,eq/kWhElectric.
This value is also used for the emission factor of local renewable source such as wind
or solar energy even before 2050 [8, 9].

Well-to-Wheel CO2-Analysis of Different Powertrain …    137

For the production of hydrogen, the energy required for the electrolysis and its
origin is the main contributor to the emission factor. The low temperature electrolyzer
efficiency is expected to reach 71% by 2030, which would lead to an emission factor
of 483 g CO2,eq/kWhH2 when using German 2030 electricity-mix or 32 g CO2,eq/
kWhH2 is only renewable energy source are used. In order to be used in a vehicle the
produced Hydrogen has to be transported and compressed. Those steps lead to total
Well-to-Tank emission factors of 579 g CO2,eq/kWhH2 and 47 g CO2,eq/kWhH2.
In order to obtain carbonated synthetic fuels, carbon atoms have to be separated
from a CO2 source and synthetized with hydrogen in a second process step. In this
work, only Direct Air Capture is treated as CO2 neutrality has to be reached in all
process steps. With future expected process efficiencies, renewable methane (eCH4)
can be produced with an emission factor of 611 g CO2,eq/kWhCH4 with the 2030
German electricity mix or 40 g CO2,eq/kWhCH4 with renewable sources [10, 11]. With
similar assumptions for the Fischer-Tropsch synthesis process, emission factors of
respectively 615 g CO2,eq/kWhFT-Di and 49 g CO2,eq/kWhFT-Di are reachable [11].
Hydrated vegetable oils (HVO) are biogenic fuels from the second generation
which only use waste oils (i.e. from the food industry) and therefore do not compete
with food production. The Well-to-Tank emission factor for HVO is taken from the
JEC Well-to-Tank reprt V5 and reaches 43 g CO2,eq/kWhHVO [5, S. 201].
Figure 5 summarizes the Well-to-Tank CO2,eq-emission factors for the considered
energy carrier.

g CO2,eq/kWh

Well-to-Tank
1400
1200
1000
800
600
400
200
0

1130
1146
994
611
579 615
343

485
65 59

43
2017

Diesel

CNG

Strom

65 59

43

65 59 23 47 40 49 43

2030-55% reducon
vs.1990

2050-Ambioniert-95

H2-VKM 700bar

eCH4

FT-Diesel

HVO100

Fig. 5.  Well-to-Tank CO2,eq-emission factors for the considered energy carrier

From Fig. 5 it becomes clear, that as soon as a sufficient amount or renewable energy
is used for the production, the type of energy carrier used will mainly depend on the
requirements during usage and not on the climate impact.

138    N. Hummel et al.

4.2 Tank-To-Wheel
For the calculation of the Tank-to-Wheel (TtW) CO2,eq emissions different aspects
have to be considered. The combustion of an energy carrier with carbon chain releases
CO2. Nevertheless, the carbon mass share is decisive regarding the amount of CO2
emitted during the combustion: therefore Diesel releases 264 g CO2,eq/kWhDiesel [12]
whereas CNG emits 202 g CO2,eq/kWhCNG [12].
In the case of electric or hydrogen powered vehicles the Tank-to-Wheel CO2emissions are Zero as no carbon-based combustion takes place in the vehicle.
Synthetic and biogenic fuels have a similar molecular structure as their fossil
counterparts therefore the combustion itself has a similar CO2 footprint than the fossil
fuel.
In addition to the CO2-emission due to the combustion of the energy carrier, the
usage of urea in the exhaust-aftertreatment system also has a CO2-impact. In this
context the SCR of a Diesel engine has further 2 g CO2,eq/kWhDiesel emission while
the SCR of a Hydrogen combustion engines emits about 4 g CO2,eq/kWhH2 (Table 1).
Table 1.  Tank-to-Wheel emissions
Tank-to-Wheel
g CO2,eq/kWh
Energy carrier
Urea
Total

Diesel

CNG

Strom

H2-VKM

eCH4

FT-Diesel

HVO100

264
2
266

202
0
202

0
0
0

0
4
4

202
0
202

255
1
256

255
1
256

4.3 Well-to-Wheel
The calculation of the Tank-to-Wheel CO2-emissions of synthetic and biogenic
fuels considers the combustion of the fuel, the so-called CO2-credits reflecting the
CO2 absorbed by the plant during its growth and the CO2 efficiency of the whole
production process. In this context HVO100 reaches a CO2-efficiency of 100% and
has a calculated Tank-to-Wheel emission factor of 0 g CO2,ep/kWhHVO [13] whereas
the CO2-footprint of Fischer-Tropsch-Diesel and e-Methane depends on the efficiency
of the CO2-capture process.
The Well-to-Wheel CO2-emission of the considered energy carrier depends on
the energy demand and the energy source used for the production and on the energy
conversion in the vehicle. The evolution of the emission factors according to the
origin of the electric energy used for production shows that e-fuels such as e-Methane
of Fischer-Tropsch-Diesel have no positive environmental impact if not produced
with at least 75% renewable electricity [14]. On the other hand, biogenic fuels such as
HVO100 have a high potential already today even though they probably will be used
as blends instead of pure fuels because of their limited capacity.
From Fig. 6 it can be taken, that from shortly after 2030 regenerative fuels are
going to break even with fossil diesel and gas if energy mixes evolve as expected
and planed from the European governments. Nevertheless sweet spot productions
allow the production of Fischer-Tropsch-Diesel and e-Methane with Well-to-Wheel

Well-to-Wheel CO2-Analysis of Different Powertrain …    139

emission factors in the range of 50 g CO2,eq/kWh in short future. The low emission
factor of HVO underlines the necessity to use all available stocks as blends in fossil
Diesel to reduce its emission factor.

g CO2,eq/kWh

Well-to-Wheel emission factors
1400
1200
1000
800
600
400
200
0

1146
1130
994
836

764

485
65
59

65
59
1990

Diesel

CNG

Strom

625
593 629
497
343

65
59
2017

H2-Elektrolyse

59 32 40
65 23 47 49

2030-55%
2050-95%
reducon vs.1990 reducon vs. 1990
H2-VKM 700bar

eCH4

FT-Diesel

Fig. 6.  Well-to-Wheel CO2,eq-emission factors for the considered energy carrier

5 Comparison of the CO2,eq-Emissions of Different
Powertrains
The multiplication of vehicle consumption from the IPG-Truckmaker simulation and
the emission factor of the energy carrier displays the CO2,eq-emission of a powertrain
system according to a production path and a mission profile.
Figure 7 displays the CO2,eq-emissions of long-haul trucks with different
powertrain systems in three energy system scenarios:
• A status quo situation in 2020 with an emission factor of 485 g CO2,eq/kWh for the
German electricity mix. The efficiency of the electrolysis process is assumed to be
at 64% and the Direct-Air-Capture has a CO2-extraction efficiency of 81%.
• In 2030, the assumed reduction of the GHG-emissions by 55% compared to 1990
would lead to an emission factor for the German electricity mix of 343 g CO2,eq/
kWh. Furthermore the efficiency of the hydrogen production and Direct-AirCapture are supposed to reach respectively 69% and 91%.
• 2050, all energy carrier are expected to have a low CO2-footprint due to the high
share of renewable electricity in the German and European mix. In this context
a reduction of 81% to 96% compared to a conventional Diesel powertrain is
achieved. Nevertheless, the values for Hydrogen, e-Methane and Fischer-Tropsch
Diesel can be reached already today if renewable energy from sweet spots is used
for the energy carrier production.

140    N. Hummel et al.
Diesel

400%
372%
361%

CNG

350%

HVO100

314%

German E-mix 2020

300%

H2-VKM 700bar
eCH4

WtW CO2,eq in %

250%
FT-Diesel

200%
193%
183%

200%

German E-mix 2030
H2-VKM 700bar German E-mix 2030

150%

100%

eCH4 German E-mix 2030
FT-Diesel German E-mix 2030

100%
83%

German E-mix 2050

59%
50%
13%
0%

66%
19%
16%
15%
4%

H2-VKM 700bar from renewable
sources
eCH4 from renewable sources
FT-Diesel from renewable sources

Fig. 7.  CO2,eq-emissions of different powertrain systems according to the energy carrier used

The CO2-saving potential of battery electric trucks over conventional Diesel trucks
varies according to the share of regenerative braking and stating operation of the
mission profile and directly translates into the powertrain efficiency. In a high
transient operation such as waste collection or city-bus applications the electric
powertrain reaches about 67% powertrain efficiency while a Diesel combustion
engine only reaches about 22%. With the German electricity mix of 2020, a CO2emission reduction of about 35% can be achieved with battery electric trucks in a city
environment and even 63% in a hilly delivery mission profile. The electric powertrain
matches all requirements of those kind of mission profiles while the drawbacks of
combustion engines are maximized: driving in poor powertrain efficiency areas and
local emissions due to cold exhaust aftertreatment systems. On the energetical side,
the electric powertrain is therefore first choice if the range requirements can be
reached.

Well-to-Wheel CO2-Analysis of Different Powertrain …    141

On the other hand, in a nearly stationary driving context of a long-haul truck a
Diesel engine has efficiency values of about 40% and with the lower regenerative
braking share the electric powertrain efficiency lays in the lower 60% which leads to
a reduction of about 30% when using the German electricity mix 2020. Nevertheless,
the importance of the requirements regarding the range of the vehicle could easily be
matched when using Fischer-Tropsch Diesel or HVO in a pure form or as blending
in Diesel combustion engines in long-haul applications. The CO2-impact can be
minimized to 13 to 19% of a conventional Diesel truck when using pure HVO or
Fischer-Tropsch Diesel. Furthermore, the local emissions of the truck are not an
issue due to operation temperatures of the exhaust aftertreatment system in which
conversion efficiencies are close to 100%.

6 Summary and Outlook
This study displays the importance of a diversification of powertrains to meet the
requirements of different heavy-duty applications. Furthermore, it becomes clear
that the energy form used in the vehicle is not a criterion for the climate impact of
a powertrain. Instead, the production path for the energy carrier and even more the
usage of every sustainable energy source is the key for a climate neutral mobility. In
this context the advantages of each energy carrier and the related powertrain must be
used efficiently: in urban applications electromobility with zero local emissions and
a high powertrain efficiency in all driving situations is without alternative. In longhaul missions, molecular renewable energy carrier have great potential due to their
high energy density, the already existing infrastructure und the scalable ramp up with
blending.
The future mobility must be prepared and solutions addressed today. Even more,
in order to reach climate neutrality by 2050, the energy system must be renewed and
build on all climate neutral energy types available.

References
1. European Commission: European Green Deal. Europäischer Grüner Deal, (2022)
2. Bundesministerium für Digitales und Verkehr: Verkehrsverflechtungsprognose 2030, https://
www.bmvi.de/SharedDocs/DE/Artikel/G/verkehrsverflechtungsprognose-2030.html. (2022)
Zugegriffen: 29. Juni 2022
3. Bundesministerium für Umwelt, Naturschutz und Nukleare Sicherheit und WWW.BMU.
DE. Klimaschutz in Zahlen – Fakten, Trends und Impulse deutscher Klimapolitik, Ausgabe
(2021)
4. European Commission: Verordnung (EU) 2019/1242 Des Europäischen Parlaments und des
Rates vom 20. Juni 2019 zur Festlegung von CO2-Emissionsnormen für neue schwere Nutzfahrzeuge und zur Änderung der Verordnungen (EG) Nr. 595/2009 und (EU) 2018/956 des
Europäischen Parlaments und des Rates sowie der Richtlinie 96/53/EG des Rates, 25. Juli
2019

142    N. Hummel et al.
5. European Commission. Joint Research Centre: JEC well-to-tank report V5: JEC well to
wheels analysis : well to wheels analysis of future automotive fuels and powertrains in the
European context: Publications Office, (2020)
6. 'Fit for 55': delivering the EU's 2030 Climate Target on the way to climate neutrality. In:
COM(2021) 550 final, 14. Juli 2021
7. Umweltbundesamt: Klimaschutz im Stromsektor 2030 – Vergleich von Instrumenten zur
Emissionsminderung. Endbericht, (2017)
8. Öko-Institut E. V. und Fraunhofer ISI. Klimaschutzszenario 2050 – Zusammenfassung des
2. Endberichts, (2. Endbericht), (2015)
9. Bundesregierung Deutschland: Klimaschutzgesetz 2021. Generationenvertrag für das
Klima, (2021)
10. Umweltbundesamt: System comparison of storable energy carriers from renewable
energies. Final report, (2019)
11. Forschungsvereinigung Verbrennungskraftmaschinen E. V. Zukünftige Kraftstoffe: FVVKraftstoffstudie IV [online]. Transformation der Mobilität im klimaneutralen und postfossilen Zeitalter. FINAL REPORT 1269 | 2021 – Frankfurt am Main, www.fvv-net.de.
(2021)
12. Umweltbundesamt: CO2-Emissionsfaktoren für fossile Brennstoffe, (2016)
13. European Commission: Richtlinie (EU) 2018/ 2001 des Europäischen Parlaments und des
Rates – vom 11. Dezember 2018 – zur Förderung der Nutzung von Energie aus erneuerbaren Quellen, 21. Dezember 2018
14. Öko-Institut E. V.: Strombasierte Kraftstoffe: die Zukunft von PtX

Hydrogen in the Gas Network – Challenges
and Solutions for High Performance Engines
for Power Generation
Clément Leroux(*), Robert Böwing, Bernadet Hochfilzer,
Alexander Zuschnig, and Manuel Behr
INNIO Jenbacher GmbH & Co. OG, Jenbach, Austria
{clement.leroux,Robert.Boewing,
Bernadet.Hochfilzer,Alexander.Zuschnig,
Manuel.Behr}@innio.com

Abstract.  The global drive to increase the use of renewable energy requires
new approaches for energy storage and transportation to be developed.
Blending hydrogen into the existing natural gas pipeline network is one
strategy to store surplus electric energy in the form of green hydrogen. This
strategy is currently being seriously considered in both the US and Europe.
The INNIO approach is to enable INNIO’s entire Jenbacher product line
for power generation to run on pipeline gas – hydrogen mixtures, with the
permitted hydrogen content allowed to fluctuate between 0 and 25% vol. The
challenges for engine development are as follows: The solution must enable
operation with varying hydrogen contents and be applicable to the entire pipeline gas product program. Engine parameters like NOx, peak firing pressure,
power control reserve, combustion knock margin, and turbocharger surge
margin must stay within defined limits. Critical events like mega knock at high
BMEP (engine damage) and autoignition during fault ride though events (grid
code non-compliance) must be avoided.
The selected strategy involves the measurement of the hydrogen content in
natural gas and the pre-definition of various engine operating parameters. The
final concept consists of a hardware/software package that includes additional
sensors as well as hydrogen compensation software. Gas train size/version and
fuel metering size/version must be checked and adapted to each individual
engine if required.
The development project ran over two years and included hydrogen sensor
selection and validation, component certification, controls strategy development, and single and multi-cylinder engine testing. The testing was carried
out with defined mixtures of natural gas, propane, and hydrogen on various
Jenbacher engines from INNIO platforms and included steady-state and
transient operation. The final solution is retrofittable.

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings S. 143–157, 2023.
https://doi.org/10.1007/978-3-658-41477-1_11

144    C. Leroux et al.
INNIO’s Jenbacher engines are now ready for hydrogen ad-mixed in the
gas pipeline. The applied technical concept offers robust power plant operation
at best performance. Any potenzial customer demands with hydrogen in natural
gas contents above 25% (up to 100%) also can be handled but will require
further engine hardware and software packages affecting pure natural gas
operation.
Keywords:  Hydrogen · Natural Gas · Renewable Energy · Power Generation

1 Introduction
Increasing the use of renewable energy requires new approaches to energy storage and
transport. Blending hydrogen into the existing gas pipe-line network is a strategy for
storing and supplying renewable energy to customers in the short-term. This strategy
is seriously being considered in both the US and Europe.
The resulting engineering task is thus to enable INNIO’s entire Jenbacher product
line to operate on natural gas – hydrogen mixtures, with the hydrogen content
fluctuating between 0 and 25%. The challenges for engine development are as
follows:
• The solution needs to enable operation with varying hydrogen contents and be
applicable to the entire Jenbacher product portfolio,
• Disadvantages for standard gas operation must be avoided
• Engine parameters like NOx, peak firing pressure, control reserve, knock margin
and turbocharger surge margin must stay within defined limits
• Critical events like mega knock at high BMEP (engine damage) and autoignition
during fault ride though (grid code non-compliance) must be avoided
This challenge differs somewhat from past developments, where engines for
special gas applications have been designed specifically for the unique, specific gas
characteristics without the need to consider unchanged engine performance with
natural gas. Some challenges like varying gas quality or potenzial uncontrolled
combustion events are similar, so the INNIO’s broad Jenbacher experience and
knowledge in special gases are key to development of a hydrogen energy solution for
the entire product portfolio in a fast and efficient manner.
The development activities described in this paper are thus not only based on
INNIO’s Jenbacher experiences in engine operation with pure hydrogen or add-mixed
hydrogen as covered in [1–4] but also on expertise in special gas engine design as
illustrated in [5–12]. The last-mentioned papers covering engine development
activities and operating experiences with various special gases containing hydrogen,
such as gases from steel production like coke gas, blast furnace gas and converter gas
(with hydrogen contents up to 65% vol.); from gasification processes such as wood
gas (with hydrogen contents up to 45% vol.); and from chemical industry processes
(with hydrogen contents up to 25%). The knowledge gained by developing customer

Hydrogen in the Gas Network …    145

solutions for these special gases includes multi-gas operation (switching between
different gases) and mixed-gas operation (e.g., add-mixing a high-calorific gas to a
low-calorific gas).
These experiences have been key concerning the new challenge of hydrogen in the gas pipeline, especially considering the long-term operating behavior in
the field. Nevertheless, it shall be pointed out again, that the boundary condition to
not compromise the given engine performance in pure natural gas operation mode
(which is assumed to remain the dominant customer operation mode) excludes major
hardware adaptation measures (compression ratio, Miller cam timing, combustion
system, port injection etc.) and makes other adaptation measures rather unattractive
(turbocharger modification, mixture temperature reduction, special oil formulations,
etc.).
Figure 1 shows some examples of already existing hydrogen in natural gas
customer applications. Please note that more than 200 MW of INNIO’s installed fleet
are running on syn gases with up to 70% hydrogen.

Fig. 1.  Examples of INNIO’s Hydrogen Admixing Customer Projects

2 Challenges of Hydrogen Admixing on Engine Operation
2.1 Introduction
Admixing hydrogen to the natural gas in the pipeline affects the engine operation if no
counter measures are taken. Table 1 shows the qualitative impact of 25% hydrogen in
natural gas with no engine-setting adaptations are done.

146    C. Leroux et al.
Table 1.  Impact of H2 in NG on Engine Operation (Qualitative Change vs. pure NG Operation)
Parameter
NOx emissions
HC emissions
Peak firing pressure
Control reserve
Knock margin
Surge margin

25% H2 Delta vs. NG
No Adaptation
↑↑
–
↑
↓↓
↓↓
↓↓

2.2 Engine Operating Window
INNIO’s Jenbacher engines are characterized by a lean-burn combustion system
developed for high efficiency and low NOx emissions. A typical operating window
of a lean-burn engine is shown in Fig. 2. This operating window is restricted by four
limits:
•
•
•
•

Knocking
Misfiring
Minimum compressor bypass (CBP) reserve
Maximum exhaust gas temperature

Fig. 2.  Operating Window of a Lean-Burn Engine

Hydrogen in the Gas Network …    147

The admixing of hydrogen to the natural gas affects the limits of the operating
window. This will be discussed further in the following section.
2.3 Effects of Hydrogen Admixing on Combustion
Hydrogen admixing results in significant changes of the air-fuel mixture properties
affecting the engine operation. Table 2 shows a comparison of selected methane and
hydrogen characteristics. Hydrogen admixing thus leads to a reduction in the methane
number, hence in the knock resistance. The lower minimum ignition energy of hydrogen makes the mixture more prone, for instance, to pre-ignition. The higher laminar
flame speed accelerates the combustion process and the wider ignition limits improve
the lean limit.

Table 2.  Comparison of the Properties of Methane and Hydrogen
Parameter

Unit
–
kg/Nm3
MJ/kg
ms

Methane
CH4
100
0.66
50
1.33

Hydrogen
H2
0
0.09
120
0.11

Methane number
Density
Heating value
Ignition delay
(λ = 1, T = 1200 K, p = 30 bar)
Laminar flame speed
(λ = 1, T = 300 K, p = 1 bar)
Minimum ignition energy
(At stoichiometric conditions)
Self-ignition temperature
Ignition limits (λ)

m/s

0.37

2.1

mJ

0.28

0.016

K
–

859
0.7–2.1

780
0.5–10.5

Hydrogen admixing to natural gas reduces the methane number of the fuel
mixture as shown in Fig. 3. This must be considered for any compensation actions.

148    C. Leroux et al.
95

NG Base - MN92

Methane Number [-]

90

NG Base - MN82

85

NG Base - MN72

80

NG Base - MN65

75
70
65
60
55
50
0

5

10

15

20

25

30

H2 Content [%Vol.]

Fig. 3.  Methane Number of Natural Gas – Hydrogen Mixtures

Hydrogen admixing also leads to an increase in the laminar flame speed of the
fuel mixture as shown in Fig. 4. At constant boundary conditions, the higher laminar
flame speed results in a faster and earlier positioned combustion process characterized
by higher peak firing pressure, higher NOx emissions, and lower exhaust gas enthalpy
for the turbine of the turbocharger.

18

Laminar flame speed
Relave change

17

16
14

16.5

12

16

10
8

15.5

6

15

4

14.5

Relave Change [%]

Laminar Flame Speed [cm/s]

17.5

2

14

0
0

5

10

15

20

25

30

H2 Content [%Vol.]

Fig. 4.  Laminar Flame Speed of Natural Gas – Hydrogen Mixtures

The low minimum ignition energy of hydrogen improves the lean air-fuel ratio
limit of the engine, but also increases the risk of auto-ignition (pre-ignition, backfiring, mega-knock, etc.). This must be considered when selecting spark timing, airfuel ratio, and maximum power output.

Hydrogen in the Gas Network …    149

Considering the main effects of hydrogen admixing on combustion described
above, a strategy to compensate the impact of hydrogen in natural gas thus could be
to operate the engine with a higher air-fuel ratio and a later spark timing. This will
have a positive effect on NOx-emissions, peak-firing pressure, power control reserve,
knock margin and turbocharger surge margin. Figure 5 shows the impact of 25%
hydrogen in natural gas on combustion with constant boundary conditions and with
adapted air-fuel ratio and spark timing.

Burn Rate [kJ/m³.°CA]

NG
NG + 25% H2 no adaptaon
NG + 25% H2 with adaptaon

Crank Angle [°CA]

Fig. 5.  Combustion Process – Impact of Hydrogen and Engine Settings

3 Development of a Hardware & Software Solution
The focus of the engineering activities was to develop and validate a hardware and
software solution for reliable engine operation with best performance at fluctuating
hydrogen contents in the natural gas grid. Therefore, a hydrogen and multi-gas
mixing infrastructure was added to the testing facilities enabling an extensive testing
campaign, including both single cylinder engine (SCE) and multi-cylinder engine
(MCE) tests, to cover the INNIO’s entire Jenbacher product portfolio. The engine
tests were performed with up to 30% vol. hydrogen in the gas on various engine types.
To cover the full range of gas qualities including the most critical cases with low base
gas methane number, the base gas used during the testing campaign was a mixture of
natural gas and propane, to which hydrogen was admixed (Fig. 6).

150    C. Leroux et al.

Fig. 6.  Hydrogen Admixing Test at INNIO in Jenbach (H2 Trailer, J612 Engine, Gas Mixing
Unit)

For thorough validation, the testing campaign included multiple types of
investigations, some of which are listed below:
Steady-state investigations:
•
•
•
•

knock and misfire border
energy balance and emissions
cylinder pressures and component temperatures
power control reserve and turbocharger layout

Transient investigations:
•
•
•
•

engine start capability and emergency stop
island mode (load acceptance and load rejection)
static and dynamic grid code
hydrogen content change rates

System tests:
• provoke intake backfire to measure flame and pressure waves
• provoke exhaust deflagration to understand risks for aftertreatment systems
All components of the engine and the balance of plant were checked, validated, and
released for hydrogen readiness in close cooperation with involved suppliers. The
parts included gas train, gas dosing valve, port injection valve (when applicable), prechamber gas compressor, compressor bypass valve, throttle valve, and turbocharger.
Although a universal, standardized hardware and software solution for hydrogen
in natural gas has been developed, some individual engine checks are still required
depending on the customer application. For example, the gas train layout must be

Hydrogen in the Gas Network …    151

checked for its sizing considering up to 20% higher gas flow rate required when
operating the engine with hydrogen in natural gas.
The extensive validation also included artificially provoked intake backfire
and exhaust deflagration tests to ensure system robustness in the field in case of
unexpected events such as engine operation outside of specification. INNIO’s
experience with special gases and field engines was very beneficial in developing and
validating a robust technical solution.
As explained in Sect. 2.3, the low minimum ignition energy of hydrogen makes
the air-fuel mixture more prone to auto-ignition. Depending on the hydrogen content,
the air-fuel ratio and the engine load, early combustion resulting in heavy knocking
can suddenly occur in principle. Figure 7 shows the cylinder pressure trace during an
uncontrolled combustion event when running without the developed compensation
software.
The final engine operating conditions were defined to stay away from any
uncontrolled combustion events and ensure reliable operation by adapting air-fuel
ratio, spark timing, or engine load. Several hardware configurations were tested,
including variation of the compression ratio, adaptation of the turbocharger, and usage
of aged components to cover all engine versions and engine aging effects in the field.
Aged components included, for example, pistons with 30,000 operating hours and
heavy deposit build up.

Cylinder Pressure [bar]

Normal Combuson
Uncontrolled Combuson

Carnk Angle [°CA]

Fig. 7.  Uncontrolled Combustion Event when Running without Compensation Software

Operating the engine with hydrogen in natural gas mixture and an increased airfuel ratio improves the combustion stability, both in transient and steady state mode.
Depending on the engine type and version, the turbocharger can be adapted to provide
sufficient power control reserve for the increased air-fuel ratio, if desired.

152    C. Leroux et al.

4 INNIO’s Jenbacher Solution for Hydrogen Admixing
Operation
4.1 Hydrogen Sensor
A core feature of the solution is to measure the fluctuating hydrogen content in
the natural gas and to predefine spark timing, air-fuel ratio and power (if required)
as function of the hydrogen content. To determine the hydrogen content in the gas
supply, a sensor is implemented upstream of the engine. Important selection criteria
for the hydrogen sensor are its response time, measurement range, accuracy, and
potential cross dependencies. The sensor cost and its lifetime and maintenance
interval are the main drivers to the lifecycle cost of this component. After performing
market research, only a few manufacturers could be identified that already have an
available product. A lot of supplier development is being done in the field of sensing
hydrogen, especially optimizing costs and reducing cross dependence, by applying
different physical principles.
The measurement principle of the chosen sensors is based on a determination
of the thermal conductivity. Since hydrogen has a much larger thermal conductivity
than natural gas as shown in Fig. 8, any positive change of this value corresponds to a
change of the hydrogen content. For this measurement principle, the knowledge of the
thermal conductivity of the base gas is key information. This, however, cannot always
be ensured since the gas quality and, hence, the thermal conductivity of the base gas,
is changing. Therefore, the hydrogen sensor was optimized by having a calibration on
a base gas that has an average value of the field gas quality. An improvement of this
approach could be having an additional sensor for the natural gas quality on site that
can handle hydrogen content.

Thermal Conducvity [mW/m.K]

180
160
140
120
100
80
60
40
20
0
H2

N2

O2

Air

CO2

CH4

C2H6

C3H8 C4H10 C5H12 C6H14

H2 Content [%Vol.]

Fig. 8.  Thermal Conductivity of Different Gases at 0 °C [14]

Hydrogen in the Gas Network …    153

During the engine test, the measurement values of the sensors were always
averaged over a certain amount of time during steady-state engine operation. Via a
gas flow meter in the hydrogen admixing pipe, the reference hydrogen content was
determined. The methane number was calculated using the gas composition with a gas
chromatograph and varied by admixing propane. The test plan included a variation
of the hydrogen content, the methane number, as well as dynamic responses of the
admixed hydrogen content. The results and the already explained cross dependency
of the base gas quality can be seen in Fig. 9, where the hydrogen content was kept
constant and a methane number variation of the base gas was made.

Supplier A

H2 Sensor Deviaon [%]

Lower / upper specififcaon limit

70

75

80

85

90

95

Base Gas Methane Number [-]

Fig. 9.  H2 Sensor Deviation to Test Bench Gas Meter vs. Gas Base Methane Number at 16%
Vol. H2 Content

4.2 Hydrogen Compensation Software
To compensate for the impact of hydrogen in the natural gas on engine operation,
different approaches were defined and implemented. The defined controls strategy
resulted in new software to meet the requirements. The three major points of the
hydrogen compensation software are as follows:
• Ignition timing point change
• Boost pressure change (i.e., air-fuel ratio change)
• Power reduction (if required)

154    C. Leroux et al.

For the hydrogen compensation software, the H2-signal, either from the H2-sensor or
provided as a customer signal in vol% is used. If the signal is not valid, an internal
default value of the H2 content is used to run the engine in stable and safe operation
conditions and prevent unstable combustion events until valid H2 content again can be
provided.
Ignition Timing Change
The nominal ignition timing in degrees crank angle depends on the fuel gas,
compression ratio and other factors. To counter the exhaust gas temperature decrease
with rising hydrogen content, the ignition timing is retarded. This also increases the
distance to the knocking border and, therefore, counteracts the higher possibility of
knock events due to the decrease of the methane number. An example of ignition
timing retardation depending on hydrogen content is shown in Fig. 10.

Ignition Timning [°CAbTDC]

Standard setting
H2 compensation
Switching point

H2 Content [%Vol.]

Fig. 10.  Ignition Timing Adaptation vs. H2 Content

The starting point to adapt the ignition timing can be set to a value higher than 0%
vol. H2 to account for sensor accuracy. The ignition timing is retarded with increasing
hydrogen content. The slope can vary depending on engine type and version. A
switching point of the ignition timing slope can be set to adapt the characteristic if
needed. This functionality is used on some engine versions.
Air-Fuel Ratio Adaptation
As explained above, combustion stability generally is improved when admixing
hydrogen to the base gas, enabling engine operation at higher air-fuel ratio. Moreover,
operation at higher air-fuel ratio allows safe operation in regard to uncontrolled
combustion events. The air-fuel ratio is increased, depending on the measured hydrogen content. Figure 11 illustrates a possible adaptation of the air-fuel ratio based on
the hydrogen content and engine load.

Hydrogen in the Gas Network …    155

The software allows for the adjustment of the air-fuel ratio with a certain slope to
counteract the impact of hydrogen. The starting point to adapt the air-fuel ratio can
be set to a value higher than 0% vol. to account for sensor accuracy. After a defined
switching point, it is possible to overcompensate the hydrogen impact by further
leaning out the engine.
The necessary increase in air-fuel ratio also depends on the engine load. If the
engine load is lower, the needed air-fuel ratio increase is lower. Therefore, the air-fuel
ratio increase is interpolated, depending on the engine load.

Standard seng - 100% load
H2 compensa on - 100% load

Air Fuel Rao [-]

Switching point

H2 Content [%Vol.]

Fig. 11.   Air-Fuel Ratio Adaptation vs. H2 Content

Power Reduction (only in Case of Excessive H2 Content)
The target engine power is set with the engine control HMI (human machine interface) or is provided as an external signal from the customer system. The nominal
engine power that can be achieved depends on the engine type and engine version,
as well as the cylinder number of the engine. With a power reduction, the maximum
allowed engine power can be reduced to prevent a potential engine trip and keep the
engine operational until the boundary conditions change, and the power reduction is
not needed any more. The reason for an active power reduction can vary. Possible
reductions of engine power can include oil or water temperatures and high or low grid
frequency among other factors.
For hydrogen admixing, an additional power reduction is introduced that depends
on the actual hydrogen content in the fuel gas. Depending on the specific engine
type and version, it may be necessary to limit the maximum power output to ensure

156    C. Leroux et al.

stable engine operation and decrease the occurrence of unstable combustion events
by decreasing the cylinder peak pressure. Additionally, the software can protect the
engine when the hydrogen admixing exceeds the upper limit (H2 > 25% vol.) by first
applying a power reduction and then shutting down the engine.
4.3 Engine Operation Results with Hydrogen Compensation Software
The developed solution covers INNIO’ entire Jenbacher engine product portfolio with
Type 2, 3, 4, 6, and 9 engines, including 50 Hz and 60 Hz applications. Subject to
engine version, base gas quality (MN), and site conditions, the engine power output
will stay unchanged up to 25% hydrogen in natural gas respectively will be slightly
adjusted according to the hydrogen signal. A potenzial version-specific power
adjustment at very high hydrogen contents can be compensated with a turbocharger
adaptation if desired.
Table 3 shows an example of a successful application of INNIO’s Jenbacher
solution for hydrogen admixing. The developed solution ensures safe and reliable
operation conditions for all engine components. It brings a slight reduction in peak
firing pressure and a moderate gain in knock margin. The control reserve can be fully
recovered, and the compressor surge margin stays unchanged. The NOx emissions
are kept unchanged while the HC emissions are reduced. An adaptation of the engine
datasheet is not required when changing from natural gas to 25% hydrogen in natural
gas operation, as engine efficiency and heat balance do not change with the applied
hydrogen compensation software.

Table 3.  Engine Operation with and without Hydrogen Compensation Software
Parameter

Unit

NOx emissions
HC emissions
Peak firing pressure
Control reserve
Knock margin
Surge margin

%
%
%
%
%
%

25% H2 in NG 25% H2 in NG
No Adaptation λ Adaptation
only
+60
±0
−20
−10
+10
+10
−25
−50
−85
−50
−20
−15

25% H2 in NG
H2 Compensation Software
±0
−25
−5
±0
+25
±0

5 Summary
Targeted activities of controls, performance, and testing engineers resulted in a
robust technical solution for INNIO’s Jenbacher engines that enables operation with
fluctuating hydrogen content up to 25% in the pipeline gas. A new engine controls
logic was developed, including a closed loop control using fuel gas and emissions

Hydrogen in the Gas Network …    157

sensors. The new logic adapts the operating settings to enable engine operation at
maximum performance while avoiding abnormal combustion events and ensuring
exhaust emissions and grid code compliance. During the application and validation
process, steady-state and transient engine testing with gas mixtures of natural gas,
propane, and hydrogen were carried out, including back-fire investigations. The
realized increase of the allowed hydrogen content in natural gas up to 25% for the
INNIO’s entire Jenbacher engine portfolio supports our continuous efforts to build
a more secure, affordable, and clean energy value chain – harnessing the power of
engineering, technology, digitization, and green fuels.

References
1. Baumann, Z.: Machen Sie sich bereit für eine Zukunft mit Wasserstoff, Dessau Gas Engine
Conference, Mai (2022)
2. Kunz, A.: Gas Engines: Renewable Power with Operational Flexibility, Key-Note, 9th AVL
Large Engines TechDays, Fully Digital, April (2021)
3. Hochfilzer, B.: Decarbonization of Europe’s Power Generation – ready for Hydrogen Engine Power Plants for 25% Hydrogen in Natural Gas, CIMAC Cascades in Graz,
September (2021)
4. Laiminger, S., et al.: Hydrogen as Future Fuel for Gas Engines, CIMAC Vancouver (2019)
5. Böwing, R.: Use of Special Gases in Power Plant Engines” in chapter “Off-HighwayGas Engines in Natural Gas and Renewable Methane for Powertrains, ISBN: 978-3-31923225-6 (2016)
6. Prankl, S. et al.: Fortschritte bei Sondergasen – GE Gasmotoren mit hoher Leistung für
wasserstoff- und kohlenmonoxidreiche Gase, 4. Rostocker Großmotorentagung, September
(2016)
7. Zauner, S. et al.: Nutzung von Gichtgas im Großmotor mit Hilfe eines auf Zweigasbetrieb
angepassten Regelungskonzeptes; 3. Rostocker Großmotorentagung, September (2014)
8. Chmela, F. et al.: Simulation of the combustion rate trend during operation with special
gases; 8th Dessau Gas Engines Conference, Germany, March (2013)
9. Amplatz, E. et al.: Utilization of special gases in stationary gas engines – A manufacturer’s
experience , MTZ heavy duty conference, Hamburg, November (2011)
10. Amplatz, E. et al.: Sondergase aus Industrieprozessen – neue Ressourcen für Energieerzeugung mit Verbrennungsmotoren; Tagung Gasfahrzeuge Stuttgart (2009)
11. Schneßl, E. et al.: Großgasmotorenkonzepte für Gase mit extrem niedrigem Heizwert; 6.
Dessauer Gasmotoren-Konferenz, März (2009)
12. Schneßl, E. et al.: „Optimierung von Brennverfahren für Sondergasanwendungen auf Basis
Simulation und Versuch am Einzylinder-Forschungsmotor“; Heavy-Duty, On-and Off-Highway Engines, MTZ Konferenz in Friedrichshafen, Nov. (2009)
13. Lopez, F., Vogl, L., Spyra, N., Böwing, R., Krainz, G.: New requirements for dynamic
Grid Code regulations and the impact on turbocharging concepts, 21. Aufladetechnische
Konferenz (2016)
14. Ingenieur Buch Website: https://www.ingenieur-buch.de/media/blfa_files/9783658106867Leseprobe.pdf. Accessed: 24. Sept. 2022

Safe and Sustainable Testing of Hydrogen
Powertrains
Nicolas Weyland(*)
KST-Motorenversuch GmbH & Co. KG, Bad Dürkheim, Germany
nicolas.weyland@kst-motorenversuch.de

Abstract.  Hydrogen-based powertrains are hopeful prospects in the development of climate-friendly transport systems of the future. Testing in a safe way
is absolutely essential for the development and testing of hydrogen-based
powertrains.
Our attention is to ensure sustainable testing.
Green hydrogen and green electricity, as well as feeding them back into
the power grid, are important elements in the issue of sustainability. The
production of green hydrogen through electrolysis is a goal we will achieve in
2024.
Safety is a priority issue, when testing with hydrogen. This is in our
interest, but also in the interest of all our customers. A comprehensive safety
package for the operation of hydrogen powertrains serves as a basis for this.
This includes hazard assessment, fire and explosion protection documents, as
well as a safety strategy.
The results of the risk assessment are the cornerstone for the safety
concept.
The risk assessment reveals the hazards that can occur during hydrogen testing. With the help of various countermeasures, the probability of
occurrence or risks of these events can be reduced.
These countermeasures have been worked out for a wide range of
scenarios and include requirements for sensors, redundancies, as well as room
air exchange rates and line evacuations.
The entire spectrum was then summarized in the safety strategy and is
being implemented for all hydrogen test benches. This strategy relates to
the test bench side. Special walls, explosion protection documents and ram
protection are also installed for trailer stations in order to the safety of the
hydrogen suplly process.
Keywords:  Hydrogen · Testing · Sustainability

1 Preface

Current trends such as emission‐free vehicles or CO2‐neutral development and
production demonstrate the growing importance of sustainability in the automotive
© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 158–166, 2023.
https://doi.org/10.1007/978-3-658-41477-1_12

Safe and Sustainable Testing of Hydrogen Powertrains    159

industry. This industry segment is going through a significant phase of change.
Key elements of this transformation are electromobility, autonomous driving and
connected vehicles. The growing demand for electric cars and innovative mobility
concepts shows how important sustainability has become.
The transformation focuses on climate change and CO2 emissions as well as the
sustainability of the value chain. Climate change and its effects have long been part
of a broad social debate. As a result of this, the legislature has significantly tightened
the requirements for CO2 emissions in recent years. Automotive manufacturers need
to reduce CO2 emissions both in vehicle production and for the vehicle fleet. In this
context, the use of alternative drive trains such as fuel cells play an important role.
As natural resources are becoming increasingly scarce, sustainable material chains
are becoming increasingly important to the automotive industry. These are based on
the principle of reusing and recycling resources. Automotive manufacturers count on
biodegradable components and sustainable processes in research, development, and
production. A sustainable supply chain and environmentally friendly transport are
essential in order to fully live up to the responsibility of human rights and environmental protection.
We support the system change to e‐mobility, combustion engines with hydrogen
and fuel cell technology based on hydrogen to deliver innovative solutions to protect
the environment and safeguarding the climate.
With our motto “Testing Powertrains to Move the Future”, we see ourselves as a
partner in innovation together with our national and international customers – mainly
manufacturers and suppliers from the automotive industry – to shape the future. We
continuously analyse the market developments to identify trends at an early stage and
support our partners with innovations in the field of testing.
•
•
•
•
•
•

How will drive the car of the future?
How efficient and environmentally friendly are electric cars?
What are the alternatives?
How powerful are hydrogen engines?
What is the lifetime of a gearbox?
How quickly do the engines wear off?

These are the type of questions we work on to find answers for.
Sustainable actions are one of the greatest challenges of our time.

2 Our Strategy to Sustainability
At KST, sustainability is based on three structural pillars: economically successful
development & testing, environmental protection and social responsibility. As a
responsible and successful company, we are equally committed to each.

160    
N. Weyland

2.1 Certifications
In accordance with the related regulations, we are accredited with various
certifications, which are regularly checked and audited. This includes a quality
management system certified according to ISO 9001: 2015 by TÜV SÜD. KST
is TISAX certified. This is an information protection standard defined by the automotive industry. The Deutsche Akkreditierungsstelle GmbH confirms our competence
according to DINEN ISO/IEC 17025:2018 to perform motor CEC test procedures.
On April 29th, 2020, KST received the certification of the French rating
“accredited R&D service provider”, issued by the French Ministry of Education
and Research. With this accreditation for the years 2020, 2021 and 2022, KST is
recognized as an innovative research service provider for companies in France.
KST as a accredited service provider must annually demonstrate its active work on
innovative R&D projects in order to receive this government certification.
2.2 Service Portfolio
By creating an environmentally friendly identity for our company, we are
strengthening our customers’ trust in our performance and capabilities for innovation
and wish to further encourage customer and supplier groups to an environmentally
conscious behaviour.
KST operates a test facility in Bad Dürkheim with state-of-the-art test benches for
passenger cars, commercial vehicles and large engines as well as for the development
and testing of fuel cells and motor vehicle powertrains.
The range of services includes functional and development work as well as
endurance testing for the following areas:
•
•
•
•
•
•
•
•
•
•

internal combustion engines till 4 MW
electric and hybrid drives from 48 V to 1200 V
powertrains till 4-wheels configuration (BEV/Hybrid/Conventional)
inverters and electronic components
hydrogen drives, H2 engines, H2 fuel cell systems, stacks and H2 components
large engines – industry/railway/marine
turbochargers
exhaust systems
transmissions
driving tests and fleet aging

2.3 Sustainability
Against the background of volatile markets and the political requirement to promote
alternative drive concepts, KST is intensifying its research and development activities.
As a result, not only test capacities for power electronics were developed and made
available, but also test capacities for hydrogen capability increased and the development of system efficiency test benches are driven forward.

Safe and Sustainable Testing of Hydrogen Powertrains    161

Fuel, electricity and hydrogen are key to operating the range of test benches
successfully. Fuels for the internal combustion engines are stored in tanks. Operating
supplies such as AdBlue, Glysantin etc. as well as oils are obtained through
thoroughly monitored channels and stored in appropriate containers. Hazardous
substances are stored in type-approved tanks which comply with the corresponding
regulations and are subject to regular monitoring. We have been purchasing 100%
green electricity from a regional supplier since 2022. Hydrogen is supplied daily by
trailer deliveries. KST has a trailer station on the plant’s premises since May 2021.
Another trailer station was going into operation in October 2022. A hydrogen storage
tank farm is also being planned.
2.3.1
•
•
•
•

How Do We Understand Ecology?

protection and efficient use of natural resources
controlled CO2 emissions with the aim of reducing them
controlled fuel consumption
controlled energy consumption with the aim of purchasing 100% green electricity
and natural gas to achieve CO2 neutrality

2.3.2 Which are the Measured Environmental Impacts of Our Activities?
•
•
•
•
•
•

controlled emissions
controlled fuel consumption
controlled power consumption
pre-sorted waste
controlled discharge of odour and noise
use of soil, energy, water and other resources

2.3.3 What Measures Have We Taken?
• in 2015, an energy management system was introduced to record every major
source of power consumption at the test benches, refrigeration systems and cooling
water reservoirs in order to measure and control the energy consumption
• partial renovation of the company buildings to reduce energy consumption
• conversion of the heating systems from night storage heaters and fuel combustion
systems to natural gas central heating
• conversion of all light sources to energy-efficient technology, mostly LEDs
• conversion of the fuel farm’s fire extinguishing system to use environmentally
friendly extinguishing agents (as a measure for calamity control)
• refurbishment of the fuel farm to guarantee optimal and permanent sealing to
prevent leakage
• retrofitting of ventilation/exhaust systems and refrigeration units of test benches
with frequency converters to achieve significant energy savings
• conversion of all refrigeration systems to environmentally friendly coolants
• conversion of the forklift fleet from diesel to electric and retrofitting of the
remaining diesel forklifts with diesel particulate filters; in 2021, 70% of fork lifts
are already electric forklifts

162    
N. Weyland

•
•
•
•

2 photovoltaic systems with approx. 90 kWp output
renewal of the exhaust silencer for noise reduction
consolidation of process cooling recirculation to increase efficiency
in 2020, starting of modernizing and retrofitting five hydrogen-based test benches
which were completed in the first quarter of 2021; seven more hydrogen test
benches are completed since the quarter of 2022, hydrogen tests in conjunction
with fuel cells and combustion engines are already possible since the second
quarter of 2021
• for the operation of hydrogen systems, specific safety measures were implemented
in order to continue to guarantee the present high safety standard
• completion of a hydrogen trailer station in Q2 2021 -> conversion from H2 cylinder
bundles to H2 trailer station for increased efficiency
• conversion of the previous purchase plan from conventional natural gas to
natural gas with CO2 neutralization in 2021 -> natural gas purchase with CO2
neutralization from 2022
2.3.4 What are the Future Measures?
• energy-related optimization of systems; replacement of old machines with more
energyefficient machines; optimization of measurement- control- and regulation
technology
• maintaining our high standards of safety through training at all times
• to further increase efficiency and save energy the centralization of compressed
air production and retrofitting the circuit with frequency converters is planned for
2023
• the centralization of a modern heat recovery system and retrofitting frequency
converters in the circuit is planned for 2023 to further increase efficiency and save
energy
• an hydrogen storage tank solution is planned for 2023 to reach a more efficient use
of the H2 resources
• a common project with the local electric supplier Pfalz-Werke is planned for 2024
to supply KST with green hydrogen from an electrolyser which will be installed in
the near of the KST test facility and connected with a pipeline; this will increase
the efficienty of delivering of H2
• reducing the company’s CO2 footprint by the use of 50% electric mobility vehicles
by 2025; in 2020, 20% of our fleet already consisted of e-models
2.3.5

Recovery of Electrical Current on the Test Bench

In combustion engines, the energy is chemically bound in the fuel (petrol, diesel,
gas, hydrogen) and is converted into kinetic energy by combustion in the engine.
The kinetic energy is converted into a speed-dependent alternating current by the
test bench loading machine (generator). The frequency converter from the generator
adapts this current to the house network or public network and feeds it in. Depending

Safe and Sustainable Testing of Hydrogen Powertrains    163

on the test operation, stationary or dynamic, up to 85% of the energy can be
recovered.
The fuel cell system in the test stand is supplied with hydrogen. The reaction
of oxygen and hydrogen produces electricity directly. The direct current is then
converted into alternating current by the battery simulator and made it available again
to the home network. Depending on the test operation, stationary or dynamic, up to
90% of the energy can be recovered.
2.3.6 From the Bottle Bundle of Grey Hydrogen to the Green Hydrogen
Electrolyzer
Hydrogen is still mostly produced from steam reforming and this production causes a
large amount of CO2 (approx. 9 tons of CO2 per ton of hydrogen); this is considered
as grey hydrogen.
With electricity from renewable sources, however, hydrogen can also be produced
CO2-free through the electrolysis of water. Only this green hydrogen can become an
energy carrier of the future.
We started to operate our first hydrogen test benches at the beginning of 2021
and used the hydrogen that could be supplied to us at that time. First small quantities
in cylinder bundles (a cylinder bundle has a content of 12 kg of hydrogen which is
enough for a maximum of one day of tests on a small fuel cell). The first trailer station
was put into operation in May 2021; this was urgently needed due to the increasing
demand for hydrogen.
In the first 18 months, we were operating our hydrogen test facility from 100%
grey hydrogen. In order to meet the increasing demand for hydrogen and to obtain
more delivery reliability, we took a second hydrogen supplier on board in September
2022, this time with hydrogen coming from an electrolyzer produced partly with
electricity from renewable sources.
As we aim to use CO2-free hydrogen supply, it is obvious that a nearly hydrogen
production plant is an important option. Now a regional energy supplier is investing
in an electrolyser plant next to our test facility. KST is pursuing this forward-looking
project as a future customer that will be directly supplied with green hydrogen.
In favor of meeting the planned demand for hydrogen, an on-site electrolyzer with
a capacity of 6 to 10 MW is needed. To cope with fluctuations in the availability of
green electricity, a large storage facility with a capacity of several tons is required.
This is the only way to ensure a continuous hydrogen supply for our test field. In
order to optimize the use of the electrolyzer, the heat generated will also be used as
process heat.

164    
N. Weyland

3 Testing
3.1 H2 Engines
When developing hydrogen engines, existing diesel or gas engines are often used
as a basis, on which both the engine software (control unit or data statuses) and the
hardware are then adapted for operation with hydrogen.
As for the hardware, there is a variety of components to test.
Some components are validated directly on the engines:
• newly designed piston head shape for an optimized combustion chamber and
optimal air turbulence; an optimized hydrogen engine can achieve an efficiency of
around 45%
• new intake valves (tightness not yet guaranteed)
• special spark plugs for hydrogen combustion
• cylinder head with modified valve train, adapted to avoid dry running
• new blow-by separation system to reduce the oil content in the combustion air
• larger exhaust gas turbochargers (single-stage or double-stage charging) to achieve
a better filling of the engine
• exhaust after-treatment systems for NOx reduction and to catch the soot from the
oil
Other components are mostly tested separately on component test benches
before they are used in the engines and further validated:
• the injection systems: there are also two options available, Direct Injection (DI)
and Port Injection (PFI)
• hydrogen pressure regulator: optimising it enables better regulation of the rail
pressure and thus more accurate injection
• the compatibility of exhaust gas probes and engine sensors fuel-carrying parts with
hydrogen have to be tested; fuel filters for hydrogen have to be tested over the long
term, as well
3.2 Fuel Cells
In the development of fuel cell systems, both, the software (control unit) and the
hardware have to go into endurance testing for operation with hydrogen. As far as the
hardware is concerned, a variety of components have to be tested.
Some components are validated directly on the fuel cell system:
• material properties (hydrogen embrittlement) and/or coatings
• tests on admixtures of gases
• dynamics tests of different humidification units

Safe and Sustainable Testing of Hydrogen Powertrains    165

• emergency stop procedures
• humidifier
• climate driving tests (freeze-start properties):
– freezing and warm-up
– freezing and heating via cooling medium
– H2 admixture in cathode air
• condensate separator
• testing of various purging strategies
Other tests on the fuel cell:
• testing of different hydrogen qualities
• harmful gases in the air
• particles in air and H2
Other components are usually tested separately on component test benches
before being used and further validated in the fuel cell system:
•
•
•
•
•
•

stack or short stack
compressor
recirculation pumps
condensate valves
testing of various pressure maintenance valves
testing of cathode air filter systems

4 Safety
Safety for the protection of the staff, the company infrastructure and the safety of the
specimen are topics that we have developed responsibly.
4.1 Safety from a Test Bench Perspective
The preparation of safety documents is a very important process:

• explosion protection and fire protection documents have been prepared in
cooperation with TÜV with safety testing of the test bench and specimen during
commissioning (incl. leakage test of all H2-carrying components)
• risk assessment was developed independently
• safety concept was also developed independently
The reason for the high safety requirements becomes apparent when one takes a
closer look at the properties of hydrogen and their effects.
If hydrogen is dispersed in the ambient air, it already forms an explosive mixture
at a hydrogen content of 4% by volume. Since hydrogen is odourless and colourless,
the gas warning system is a central element within the safety concept.

166    
N. Weyland

The test bench ventilation increases the room air exchange rate, when gas is
detected. In combination with an explosion-proof ceiling fume hood, the hydrogen
can thus be diluted or evacuated.
In addition, hydrogen can also ignite directly at the point of leakage (e.g. pipe
or fitting), generating an extremely hot flame (2500 °C). For example, the spark of a
dropped tool is sufficient as ignition energy.
Compared to a fossil fuel fire, the hydrogen flame contains no carbon atoms.
Therefore, in addition to the high temperature, the flame is not visible to the human
eye. Accordingly, the infrared spectrum is not or only barely present. The IR flame
detectors from the test bench are therefore supported by UV flame detectors, which
detect hydrogen fires. These and many other topics were summarized in the selfdeveloped safety strategy and serve as an internal guideline for the retrofitting of
hydrogen test benches.
In addition to the test benches, however, safety must also be taken into account in
the delivery of hydrogen. For this purpose, several trailer stations have now been built
under strict conditions. This included structural measures such as protective walls,
explosion protection zones, special parking spaces, lightning arrester and much more.
4.2 Safety Measures to Protect the Test Parts
• pop-up valves or bursting discs in the intake air system or in the charge air duct
protect internal combustion engines in the event of backfiring
• installation of hydrogen sensors in the exhaust system for combustion engines and
fuel cells and in the blowby system for combustion engines; measurements show
up to 6% hydrogen content
• in the case of combustion engine applications, knock monitoring is realised via
installed indexing measurement systems and an immediate automatic shutdown of
the H2 supply to the test specimen is initiated in the event of knock events
• possibility of manual shutdown of the H2 supply to the test specimen by means
of a button on the test stand console in case of undesired events that are not
automatically monitored

Hybrid PEM Fuel Cell Systems
Sönke Gößling(*), Matthias Bahr, and Felix Smyrek
ZBT – Zentrum für Brennstoffzellen-Technik GmbH, Duisburg, Germany
{S.Goessling,M.Bahr,F.Smyrek}@zbt.de

Abstract.  PEM fuel cell drivetrains in heavy duty applications are mostly
implemented using multiple fuel cell systems and as battery hybrid systems.
The complex architecture of those hybrid system results in various degrees
of freedom in the operating strategy. Those have a major influence on aspects
such as dynamics, efficiency and lifetime of the entire system. In order to be
able to analyze and evaluate the capability of an efficiency optimized operating
strategies, the ZBT fuel cell model is integrated into a fuel cell system and
a full vehicle simulation. The dimensioning and also the load profiles are
inspired by the Hyundai XCIENT Fuel Cell with 20 t total permitted mass.
Starting from a rudimentary and purely functional operating strategy, an
efficiency optimized operating strategy is presented. A mere optimization of
the fuel cell system operation is not sufficient, since the losses in the battery
system must also be considered. The numerous effects of the operating strategy
are described.
Keywords:  PEM fuel cell model · Fuel cell system simulation · Fuel cell
vehicle simulation · Hybrid control strategy

1 Introduction
The development of electrified and CO2-free heavy-duty vehicles shows a rapid
change process. Even if optimal long-term solutions still require a considerable
amount of development, solutions must be made available in the short and medium
term. In order to put heavy-duty vehicles on the road nowadays, for example several
fuel cell systems are now used in current vehicles with dimensions essentially
comparable to those of passenger cars. The reason for this is the availability of the
system components as well as the fuel cells itself in this performance class. The
batteries in the vehicles are dimensioned relatively large, which allows the fuel cells
to be operated carefully and thus with a low degradation to enable the increased
lifetime requirements compared to passenger cars. As a result of this configuration,
several energy sources are available to cover the power requirements of the electric
drive. A significant part of the operating strategy is to choose the power distribution
between the battery and fuel cell that the most dynamic coverage possible is ensured
with the simultaneous highest possible efficiency and the lowest possible degradation

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 167–171, 2023.
https://doi.org/10.1007/978-3-658-41477-1_13

168    S. Gößling et al.

of the entire system. To demonstrate the potential offered by this flexibility in energy
distribution, the configuration of the Hyundai XCIENT Fuel Cell with a total mass of
20 t is replicated in a simulation model and examined in detail on the basis of publicly
available data [1].

2 Simulation Setup
The simulation is organized on three levels. On top level, the entire vehicle is
connected together with the two separate fuel cell systems with 95 kW each and the
73 kWh battery system. The systems are connected together via their respective DC/
DC converter to a 650 V high-voltage bus. This high-voltage bus supplies the 190 kW
electric motor, which in turn is connected to a 2-speed transmission. The vehicle is
operated via a distance and speed profile including idle times and altitude profile. A
general overview of the top level is shown in Fig. 1.

Fig. 1.  Screenshot of the AVL Cruise™ M top level vehicle simulation

On second level, the fuel cell media systems are simulated. Active recirculation
takes place at the anode by using a recirculation blower [2]. A Celeroton-2212000-GB compressor is used on the cathode side, the system is pressurized by means
of a throttle and a controllable bypass is also implemented for partial load operation.
There is no humidifier installed in the cathode system. The cooling system consists
of a small and large circuit and controls the temperature at the inlet as well as the
temperature rise across the fuel cell stack. The top level simulation of the vehicle as
well as the simulation of the media systems is done using the software AVL Cruise™.
The fuel cell system is generally not limited in operation at low power requirements.
However, if the power requirement approaches zero, the efficiency will also approach
zero (Fig. 2). The part-load range in the region of 10% of the possible maximum
system power achieves the highest efficiency of approx. 60%. As the system power
increases further, the efficiency decreases.

Hybrid PEM Fuel Cell Systems    169

On the third level, the simulation of the transport processes within the fuel cell
and the power of the fuel cell is carried out. At this level, a model developed by ZBT
in MATLAB/Simulink is used, which is integrated into the CruiseM environment via
functional mock-up interface (FMI) to enable a co-simulation. The model provides
a local resolution of the internal electrochemical and transport processes while still
being able to provide 1 ms real-time capability. It is based on the model published in
[3], but it has been extended to include various physical correlations and, in particular,
dynamic components [4, 5]. The presented fuel cell stack has an active area of
300 cm2 and consists of 400 cells.

Fig. 2.  Efficiency of the fuel cell system

The vehicle is operated with different VECTO load cycles [6] to be able to
analyze the distinction of different operating strategies.

3 Results
The reference operating strategy is mainly designed as a load-following strategy for
the fuel cell systems; the output of the fuel cell systems has the objective of providing
the required propulsion power. The battery is controlled in order to stabilize the highvoltage bus. If deviations occur in load-following operation due to limited maximum
power and dynamic boundary conditions of the fuel cell system, the difference is
automatically covered by the battery through the control concept. As a result, the state
of charge (SoC) of the battery changes. In this case, the reference operating strategy
adapts the load-following operation of the fuel cells to provide a surplus or less power
to control the SoC to the target of 65%. In Fig. 3, the resulting SoC of this strategy is
shown. This strategy enables to fully match the performance requirements and keep
the SoC within a relatively small operating range.
The reference operating strategy is not affected by the efficiency of the fuel
cell systems, which can vary considerably over the operating range. This influence
is additionally taken into account in the following efficiency-optimized operating
strategy. The new operating strategy now attempts to keep the SoC within the

170    S. Gößling et al.

specified operating range while operating the fuel cell systems at high efficiency.
The SoC as a result of the efficiency-optimized operating strategy is shown in Fig. 4.
The average power demand of the load profile is higher than the power supply of
the fuel cell systems when they are operated at optimum efficiency. As a result of
the operation of the fuel cell systems with priority on efficiency and consequently
lower power supply compared to the average demand, the SoC of the battery, which
compensates for the deviation, drops to a lower level. It can also be observed that the
SoC of the battery is subject to higher dynamics.

Fig. 3.  SoC for an operation with the reference operating strategy

Fig. 4.  SoC of the operating strategy, which focuses on prioritized operation of the fuel cells in
higher efficiency range

In order to compare and evaluate the results, the hydrogen consumption of the
vehicle is calculated. In addition, however, the SoC at the start and end of the cycle
must also be considered, which has a significant influence due to the large installed
capacity of the battery. In order to ensure that the SoC is properly taken into account
and that no simplifying assumptions have to be made here, the cycle is run through
several times in sequence. One cycle contains of a distance of 100 km and a total of
five of these cycles are carried out. In Figs. 3 and 4, it can be seen that the absolute

Hybrid PEM Fuel Cell Systems    171

level of SoC changes significantly in the first cycle, but a comparable pattern appears
for each cycle thereafter.
Comparing the consumption of the overall system for the two shown operating
strategies, no clear increase in efficiency can be seen here. Although the fuel cell
systems can be operated more efficiently, they are not solely responsible for the losses
in efficiency. All DC/DC and also the battery itself operate with specific power losses.
The new operation management has an impact on all these electrical systems. In
order to achieve an increase in efficiency, these must also be taken into account in the
operating strategy.
Acknowledgements. The research project was carried out in the framework of the industrial
collective research program (IGF no. 61 EWN). It was supported by the Federal Ministry
for Economic Affairs and Climate Action (BMWK) through the AiF (German Federation of
Industrial Research Associations eV) based on a decision taken by the German Bundestag.

References
1. Hyundai Motor Company (8. July. 2020): Erste Brennstoffzellen-Lkw Hyundai Xcient Fuel
Cell kommen nach Europa., [online] https://www.hyundai.news/de/articles/press-releases/
erste-brennstoffzellen-lkw-hyundai-xcient-fuel-cell-kommen-nach-europa.html. Accessed:
13. Oct. 2022
2. Klunker, C., Nachtigal, P.: REZEBT: Rezirkulationsgebläse-Entwicklung für die Brennstoffzellen-Technologie, FC3 Fuel Cell Conference Chemnitz (2022)
3. Gößling, S.: 2-D + 1-D ortsaufgelöste Modellierung von PEM-Brennstoffzellen. Dissertation, Universität Duisburg-Essen (2019)
4. Gößling, S., Nickig, N., Bahr, M.: 2-D + 1-D PEM fuel cell model for fuel cell system
simulations. Int. J. Hydrogen Energy 46(70), 34874–34882 (2021). https://doi.org/10.1016/j.
ijhydene.2021.08.044
5. Tinz, S.; Wick, M.; Gößling, S.; Bahr, M.: Modulare Simulationsumgebung für Brennstoffzellensysteme mit Fokus auf den Membranwasserhaushalt, MTZ83. 11, 52–65 (2022)
6. European Commission: Vehicle Energy Consumption calculation TOol – VECTO, https://
climate.ec.europa.eu/eu-action/transport-emissions/road-transport-reducing-co2-emissionsvehicles/vehicle-energy-consumption-calculation-tool-vecto_en. Accessed: 14. Oct. 2022

Fuel Cell System Development for Heavy
Duty Application
Stephan Schnorpfeil(*), Arne Kotowski, Hauke Sötje,
and Guido Hartmann
SEGULA Technologies GmbH, Rüsselsheim am Main, Germany
{stephanjohannes.schnorpfeil,arne.kotowski,
hauke.soetje,guido.hartmann}@segulagrp.de

Abstract.  Fuel Cell propulsion systems offer high energy density and short
refueling times. These are the key facts for their application to heavy-duty
vehicles. In the fuel cell propulsion system the layout of the fuel cell system
itself is crucial. Based on the requirements like lifetime – and efficiency
targets as well as package boundary conditions the components in the fuel cell
system like stack and the so called BoP (Balance of plant) components such as
compressor, hydrogen gas injector and humidifier are defined. Besides this the
high temperature – and low temperature circuits and the high voltage system
with DCDC’s and bleeding resistor are specified.
The paper describes the fundamental aspects and processes for defining
the layout of a fuel cell system with all its subcomponents. Starting with 1D
simulation the BoP components and their interactions are optimized for overall
high system efficiency and performances. The system packaging ensures the
integration in the heavy-duty vehicle and the connection to the vehicle air
intake – and hydrogen storage system, high voltage architecture and thermal
management. The software of the fuel cell system adjusts the system actuators
based on the targeted fuel cell system state like idle, nominal power, peak
power and transient operation. Finally, the system is tested on the fuel cell test
bench.
Keywords:  Fuel Cell · Balance of Plant · Simulation

1 Motivation
Fuel cell systems are more and more entering the market in different sectors. For
example developments are announced e.g. for marine – [1] and aviation applications
[2]. However, mainly in the heavy-duty sector the technology gets more and more
momentum. Major suppliers and OEM’s have started or even accomplished projects
for fuel cell vehicle and/or fuel cell system production [3, 4].
These heavy-duty applications need a fuel cell system that is dedicated development to their requirements especially concerning durability, total costs of ownership

© Der/die Autor(en), exklusiv lizenziert an Springer Fachmedien Wiesbaden GmbH,
ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022 , Proceedings, S. 172–183, 2023.
https://doi.org/10.1007/978-3-658-41477-1_14

Fuel Cell System Development for Heavy Duty Application    173

and efficiency. Besides this, certification and safety regulations need to be fulfilled.
The realization of such a heavy-duty fuel cell system is described in this paper.

2 Basic Fuel Cell System Layout
The fuel cell system is the new heart of future powertrain for heavy-duty fuel cell
trucks (compare Fig. 1).

(a)

(c)
(e)

(b)

(d)
Fig. 1.  HyBatt truck component arrangement: a Hydrogen storage system (HSS) b Fuel cell
systems c Traction battery d E-Axle e HV Electronics [5]

The main task of the fuel cell system is to convert the hydrogen coming from the
HSS (hydrogen storage system) into electrical energy for the electrical powertrain.
Therefore, a well-specified setup of the fuel cell system is required (illustrated in
Fig. 2).

Fig. 2.  Simplified fuel cell system layout [6]

174    S. Schnorpfeil et al.

The fuel cell system consists of hydrogen supply (anode), air supply (cathode),
exhaust system, low temperature and high temperature coolant circuit as well as the
high voltage and low voltage system. Here, each component in the periphery of the
stack that is necessary to run the reaction in the stack is called a Balance of Plant
(BoP) component.
In the anode path the main components are hydrogen injector, jet pump and
recirculation pump. The hydrogen injector introduces the hydrogen into the hydrogen
recirculation. This injection provokes the motion of the hydrogen towards the stack
where it is converted into current based on the inner stack reaction. However, not the
complete hydrogen can be transformed. The rest is coming towards the stack anode
outlet and is back fed to the hydrogen circulation mainly by the jet pump which is
basically working after the ejector principle. Only in conditions where the hydrogen
flow is stalling the recirculation pump will support the back flow to the jet pump.
The cathode path consists mainly of the components air filter, compressor, heat
exchanger and humidifier. The air filter is extremely important in order to sort out
NOX, NH3 and SO2 since the stack will be contaminated with these pollutants. The
compressor supplies necessary air to the stack. This is on the one hand essential to get
in the start-up phase the reaction in the stack running. Here the compressor needs to
be supplied from a battery until the stack produces its own current. On the other hand
the compressor is the enabler to increase the power density of the fuel cell system. A
high boosting power in combination with a high conversion density rate in the stack
can lead to a potential for reducing stack size. Consequently, the compressed air need
to be cooled in the intercooler. Beyond the mentioned tasks of the compressor in the
shut-down phase the afterun of the compressor is necessary to dry the stack. However,
in normal operation it is necessary to humidify the stack. This is carried out by the
humidifier in case the stack has no internal humidification. Therefore, the exhaust gas
of the stack can be fed back to the humidifier.
For the operation of the fuel cell system a dedicated thermal management is
necessary. This normally includes a low temperature – and high temperature coolant
cycle. Whereas the high temperature coolant cycle is necessary to cool the stack with
a supply temperature of around 70 °C, the low temperature coolant cycle provides the
coolant for the intercooler and the high voltage electronics with a supply temperature
of around 40 °C.
The high voltage circuit layout is also dedicated for the application. One example
is shown in Fig. 3.

Fuel Cell System Development for Heavy Duty Application    175

Fig. 3.  Exemplary principle of high voltage system for a fuel cell system

In the start up phase the pre charge resistor relay is closed in order to limit current.
In normal system operation the pre charge resistor relay is open and main relay is
closed so that fuel cell system can feed the high voltage battery and/or the electric
engine. In system shut down the main relay and pre charge relay is opened and the
relay at the depolarization resistor is closed. This enables the depolarization resistor to
discharge the stack which is reacting the residuals of hydrogen and air.

3 Fuel Cell System Specification
3.1 Fuel Cell System Simulation
The fuel cell system development and specification is essentially following a V-cycle
approach, as it is known from other developments (Fig. 4).

176    S. Schnorpfeil et al.

Fig. 4.  Fuel Cell System development V-cycle

For the system simulation and especially for the BoP (Balance of plant)
component definition a system model is set up (Fig. 5). The model reflects the main
components of the BoP but besides this in the sub models also all throttles and valves
that are relevant for the system operation.

Fig. 5.  BoP (Balance of Plant) simulation model overview [6]

Fuel Cell System Development for Heavy Duty Application    177

The driving parameter of the fuel cell system simulation model is the electrical
load. The electrical load is determined in the superordinated fuel cell vehicle model
based on the driving cycle and vehicle operation strategy (Fig. 6).

Fig. 6.  Superordinated fuel cell vehicle model [7]

In terms of heavy duty application this leads to a configuration with
approximately 210 kW continuous power from the fuel cell system and a 50 kWh
– 70 kWh battery capacity, as presented in [5]. This enables the constant highway
driving to be covered by the fuel cell system only (Fig. 7).

Fig. 7.  Simulated drive cycle for Heavy duty truck tractor [5]

178    S. Schnorpfeil et al.

In order to archive the electrical load target the necessary air – and hydrogen flow
needs to be established in the stack. This drives the layout of anode and cathode path
with all subcomponents.
For example the layout of the passive humidifier is one of the most complicated
parts in the fuel cell system development process and is a challenge for multiphysical
simulation. Passive means, that water from the cathode return line will be used to
humidify the air from the cathode feedline. In general, a high humidification level is
required for a high stack efficiency. In the simulation of the passive humidifier points
like water transportation from wet to dry side as well as the thermal coupling between
both sides need to be taken into account. Target is to size the passive bypass throttle
in order to maintain a specific mass flow through the humidifier in all operating
conditions. Figure 8 shows the humidifier flow chart.

Fig. 8.  Passive humidifier flow chart [6]

Besides the dimension, also special requirements for the components need to be
taken into account. NOX, NH3 and SO2 as well as oil droplets would contaminate the
stack. Therefore, air filter needs to be installed which can reduces NOX, NH3 and SO2
to an acceptable minimum. Besides that the compressor needs to run oil free. Thus,
often compressors with air bearings are installed. Figure 9 shows the assembly of a
fuel cell compressor [8].

Fuel Cell System Development for Heavy Duty Application    179

Air bearing
Aerodynamics

Motor

Motor winding

Fig. 9.  Fuel cell compressor assembly with air bearings [8]

Important is to ensure the system function and do a validation of the simulation
model as shown in the development V-cycle (Fig. 4). Therefore the subcomponents
are tested by the supplier and the characteristic curves and/or maps are implemented
in the model. Often it is recommended to test the stack directly and/or do a stack
drive-in before it is implemented into the system. Furthermore, the interaction of the
components needs to be taken care of. This point can be determined on the fuel cell
system test bench where the complete fuel cell system is tested. Figure 10 shows a
stack testing – and fuel cell system testing setup.
Stack testing

Fuel Cell System testing

Fig. 10.  Stack and fuel cell system testing setup

180    S. Schnorpfeil et al.

Besides the already mentioned points, the fuel cell system simulation is an
important engineering tool for setting up the software & controls environment.
Coupling the fuel cell system model with the operation software modules the software
modules are tested and calibrated in the simulation environment (model in the loop).
3.2 Total Costs of Ownership
Besides the purely dimensioning of the components other factors in the layout of
a fuel cell system need to be taken into account. One critical point is the total cost
of ownership (TCO) of the fuel cell vehicle and therewith of the fuel cell system.
Typically, a holistic approach for the complete fuel cell system is applied. This means
that the same stack platform is used for different applications – the stack is only
multiplied. Thus, the production number of stack units increases and this drives lower
production costs per unit. Besides, in some applications the whole fuel cell system
is used multiple times – e.g. for reaching a power output of 200 kW two fuel cell
systems with 100 kW power output each is applied. Both approaches are exemplarily
shown in Fig. 11.

Fig. 11.  Fuel Cell Stack and – System multiplication

Both approaches have their advantages and disadvantages, e.g. multiplying
the stack reduces the amount of BoP components significantly. On the other side
the development of application-dedicated BoP components like humidifier and
particularly compressor may be necessary. In addition, hybrid forms between both
approaches are in development. Here a standardization of the fuel cell system can help
amongst others to lower TCO for end user. One example is StasHH – Standard Sized
FC module for Heavy Duty applications [9] (Fig. 12).

Fuel Cell System Development for Heavy Duty Application    181

Fig. 12.  StasHH Key Terminology and system definition [10]

There amongst others, the size of the package box for a dedicated fuel cell system
power output range as well as interfaces to application and the content of the package
box is defined (Fig. 13).

Fig. 13.  Dimension table based on StasHH [10]

Also for the BoP components a modularity approach can be followed. Thereby,
it is important that the components can be applied for more than one fuel cell
application with a high same-part ratio. For example, this can be demonstrated on the
fuel cell compressor. While the compressor wheel and volute is changed in order to
achieve a higher mass flow, the rest of the compressor parts such as air bearings and
E-motor are equal. Thus, a same part of ratio of up to 80% can be achieved [8]. This is
shown in Fig. 14 for three different compressor types.

182    S. Schnorpfeil et al.

Fig. 14.  3 different fuel cell compressor types with different air mass flow [8]

3.3 Safety and Certification
During the development it is important to take safety and certification requirements
into account from the beginning. Standard requirements come from the ISO 26262.
This particularly means that the functional safety process needs to be realized.
Starting with the DIA (Development Interface Agreement) between the projects
partners, steps like HARA (Hazard Analysis and Risk Assessment), Functional- and
Technical Safety Concept as well as Technical Hardware – and Technical Software
Safety need to be passed.
Up to now, the consolidation for an applicable standard regarding fuel cell
systems in vehicles is ongoing. Guidance can be given by standards that are existing
from other areas like ECE R134, ECE R100, ECE R10, regulation (EC) 79/2009 and
regulation (EU) 2021/535. However, safety and certification requirements can have
significant influence especially on the design of the system and therefore need to be
carefully reviewed before starting the conception phase. Some guidance is given e.g.
by white paper from TÜV Süd [11].
3.4 Additional Requirements
Furthermore, development also needs to take serviceability -, durability – and crash
requirements into consideration. Regarding serviceability, it is important upfront
to define components that need to be exchanged in service. One example is the ion
exchanger. Accessibility needs to be ensured for exchanging in service.
The durability target for heavy duty applications is a lifetime of more than 30,000
hours. The requirements for every BoP component can be estimated with the help of
the simulation model. Based on driving cycle it can be estimated how often and in
which modulation a component is actuated. This can help the supplier to represent the
lifetime driving on a dedicated test bench for the component.

Fuel Cell System Development for Heavy Duty Application    183

4 Summary
The paper demonstrates how in fuel cell system development the upfront simulation
of the system is a crucial step. Here, the basic components are defined. From
the vehicle data and drive cycle the requirements for each component can be
derived. Also, all operating modes of the fuel cell system can be simulated and the
component’s functionality can be checked. Besides, also the modulation of each
component can be estimated based on the driving cycle over lifetime. MiL simulation
helps to identify gaps in fuel cell system controls and software.
Next to operational aspects there are other boundary conditions like package
space-, certification-, safety- and total costs of ownership requirements. They also
need to be validated in the beginning of the development phase. Moreover, special
requirements like oil-free bearings in the compressor need to be recognized. Overall,
the above mentioned items need a careful review from the beginning of the development phase onwards.
Finally, the fuel cell system performance is tested on the fuel cell system test
bench. Here the interaction of the components can be validated. In order to be able
to check all operating modes a fuel cell system test bench with battery simulator is
recommended.

References
1. About Travel: https://abouttravel.ch/reisebranche/kreuzfahrtschiffe-und-die-klimafrageerste-brennstoffzellen-tests-kommen/. Accessed: 5. Oct. 2022
2. Airbus: https://www.airbus.com/en/newsroom/news/2020-10-hydrogen-fuel-cells-explained.
Accessed: 5. Oct. 2022
3. Daimler Truck: https://media.daimlertruck.com/marsMediaSite/en/instance/ko/Developmentmilestone-Daimler-Truck-tests-fuel-cell-truck-with-liquid-hydrogen.xhtml?oid=51975637.
Accessed: 5. Oct. 2022
4. Freudenberg: https://www.fst.com/de/fuel-cell/. Accessed: 5. Oct. 2022
5. Weiss, E., Schnorpfeil, S., Nickel, D.: Fuel Cell System Integration for Heavy-Duty
Applications. In: ATZ Heavy Duty Conference, Rostock (2021)
6. Hartmann, E., Rohde, N., Muthusamy, V., Schnorpfeil, S.: Fuel Cell Systems – A Challenge
of Multiphysical Simulations. Matlab Expo (2022), 2022/05/17
7. Schnorpfeil, S., Hartmann, E., Kotowski, A., Kapadia, B., Sötje, H.: Fuel Cell Propulsion
System Layout. In: ATZ Conference – Der Antrieb von morgen (2021)
8. Schnorpfeil, S., Glahn, C., Zwyssig, C., Fröhlich, P.: Standardized compressors for fuel cell
applications. In: 26. Aufladetechnische Konferenz, Dresden (2022)
9. Stashh: https://www.stashh.eu/. Accessed: 15. Sept. 2022
10. Stashh:
https://www.stashh.eu/sites/default/files/StasHH%20standard%20%28part%20
1%29%20-%20Size%20definition.pdf. Accessed: 15. Sept. 2022
11. Tüv Süd: https://www.tuvsud.com/en-us/resource-centre/white-papers/eu-type-approval-ofhydrogen-powered-vehicles. Accessed: 15. Sept. 2022

Autorenverzeichnis

A
Allcorn, Roger, 53
Atkins, Penny, 53
B
Bahr, Matthias, 167
Bärow, Enrico, 112
Behr, Manuel, 143
Beidl, Christian, 131
Bennet, David, 53
Bessey, Dirk, 123
Boom, Rick, 112
Böwing, Robert, 143
Buchholz, Bert, 24
Buehler, Frank, 123
Bülte, Heiner, 1
C
Calero, Alvaro Concepcion, 53
Coles, Nicholas, 53
D
D’Onofrio, Mario, 95
Davis, Jonathan, 53
Downes, Trevor, 53
F
Faiß, Manuel Eugen, 123
Fox, Nigel, 53
Friedrich, Thomas, 40
Funke, Carsten, 1
G
Gößling, Sönke, 167
H
Hartmann, Guido, 172
Harvey-Bush, Jackson, 53

Herold, Tim, 131
Herrmann, Dennis, 95
Hochfilzer, Bernadet, 143
Holz, Oswald, 76
Hughes, John, 53
Hummel, Nicolas, 131
K
Kapusta, Lydia, 95
Keck, Mathias, 76, 123
Kotowski, Arne, 172
Kühner, Andreas, 112
L
Leroux, Clément, 143
Loiudice, Angela, 53
M
Mante, Till, 24
Mauss, Fabian, 24
Mayer, Florian, 40
Mestre, Laura, 24
N
Nork, Benedikt, 1
O
Osborne, Richard, 53
P
Penning, Richard, 53
Pirkl, Richard, 95
Prehn, Sascha, 24
Presti, Manuel, 76
R
Rabanser, Peter, 53

© Der/die Herausgeber bzw. der/die Autor(en), exklusiv lizenziert an Springer
Fachmedien Wiesbaden GmbH, ein Teil von Springer Nature 2023
A. Heintzel (Hrsg.): HDENGI 2022, Proceedings, S. 185–186, 2023.
https://doi.org/10.1007/978-3-658-41477-1

Autorenverzeichnis

186
S
Sailer, Dennis, 76
Saroop, Agam, 53
Schneider, Simon, 40
Schnorpfeil, Stephan, 172
Seba, Bouzid, 13
Seidel, Lars, 24
Smyrek, Felix, 167
Sötje, Hauke, 172
Stiehl, Roman, 40
T
Theile, Martin, 24
Töpfer, Georg, 1
Trabold, Christian, 40

V
Valenta, Lukáš, 53
W
Weiss, Ulrich, 13
Weller, Fabian, 40
Weyland, Nicolas, 158
Willmann, Michael, 112
Z
Zuschnig, Alexander, 143