D

e
s
i
g
n
,

l
a
y
o
u
t

a
n
d

p
r
i
n
t

p
r
o
d
u
c
t
i
o
n
:

E
r
i
k

T
a
n
c
h
e

N
i
l
s
s
e
n

A
S
,

0
6
/
2
0
1
2
P
r
i
n
t
e
d

o
n

e
n
v
i
r
o
n
m
e
n
t
a
l
l
y

f
r
i
e
n
d
l
y

p
a
p
e
r
.
Fuel cells for ships
Research and Innovation, Position Paper 13 - 2012
Contact details:
Kristine Bruun Ludvigsen - Kristine.Bruun.Ludvigsen@dnv.com
Eirik Ovrum - Eirik.Ovrum@dnv.com
Research and
Innovation in
DNV
This is
DNV
The objective of strategic research
is through new knowledge and
services to enable long term
innovation and business growth in
support of the overall strategy of
DNV. Such research is carried out
in selected areas that are believed
to be of particular signicance
for DNV in the future. A Position
Paper from DNV Research and
Innovation is intended to highlight
ndings from our research
programmes.
DNV is a global provider of services
for managing risk. Established
in 1864, DNV is an independent
foundation with the purpose of
safeguarding life, property, and the
environment. DNV comprises 300
ofces in 100 countries with 9,000
employees. Our vision is to create
a global impact towards ensuring a
safe and sustainable future.
Fuel cell technology has recently been proven successful in large maritime
demonstration projects. Although fuel cell technology is not new, this success
means that it has become relevant to discuss the potential for fuel cell
technology in on-board applications and the current status of the technology, as
done in the present paper. This paper also discusses certain safety aspects, and
highlights the development of mathematical models for assessing performance
and operational aspects of shipboard fuel cell systems via simulation.
The main drivers for developing maritime fuel cell technology are reduction
in fuel consumption and less local and global impacts of emissions to air from
ships. Additional benets include insignicant noise and vibration levels,
and lower maintenance requirements compared with traditional combustion
engines. Key challenges include the demand for clean, low carbon fuel and the
need to decrease investment costs, improve service lifetime, and reduce the
current size and weight of fuel cell installations.
DNV Research and Innovation has taken a leading role in facilitating
the demonstration of safe and reliable fuel cell applications for ships. In
the FellowSHIP project, a 330 kW fuel cell was successfully installed, and
demonstrated smooth operation for more than 7000 hours on board the
offshore supply vessel Viking Lady. This is the rst fuel cell unit to operate on
a merchant ship, and proves that fuel cells can be adapted for stable, high-
efciency, low-emission on-board operation. When internal consumption was
taken into account, the electric efciency was estimated to be 44.5 %, and no
NO
X
, SO
X
and PM emissions were detectable. When heat recovery was enabled,
the overall fuel efciency was increased to 55 %. Nevertheless, there remains
potential for further increasing these performance levels.
Although operational experiences have shown that fuel cell technology
performs well in a maritime environment, further R&D is necessary before fuel
cells can be used to complement existing powering technologies for ships.
Summary
4
RISING FUEL PRICES and impending environmental
regulations have created a pressure for ships to operate
more efciently and in an environmentally friendly
manner. Fuel cell power production is a technology that
can eliminate NO
X
, SO
X
and particle (PM) emissions,
and reduce CO
2
emissions compared with emissions from
diesel engines. Fuel cells powered by low carbon fuels
(e.g. natural gas) will have local and regional benets as
both emissions and noise are reduced. In the longer term,
hydrogen fuel generated from renewables could lead to
ships with zero carbon emissions.
The use of the fuel cell as an electricity generator was
invented by William Grove in 1842 (Vielstich et al., 2001).
Due to the success and efciency of combustion engines,
fuel cells have not been widely considered for general
use, and, until recently, fuel cells have been applied
only for special purposes, such as space exploration and
submarines. However, rising fuel prices and a strong focus
on reduction of global and local emissions have led to
an increasing focus on the development of fuel cells for
application in other areas as well. Recent market studies
(Fuel Cell Today, 2011) have revealed that fuel cells should
no longer be considered as a technology for the future;
they are already commercially available today for a diverse
range of applications (e.g. portable electronics, power
plants for residential use, and uninterruptible power
supply).
The FellowSHIP
1
project designed, developed, built, and
tested a 330 kW marine fuel cell power pack installed on
board the Norwegian supply vessel Viking Lady (owner:
Eidesvik Offshore ASA). With this rst large-scale fuel cell
1 www.vikinglady.no
pilot in operation, Viking Lady docked in Copenhagen
during the COP-15 international climate conference in
December 2009, (Biello, 2009), (Figure 1). This not only
demonstrated that fuel cells can operate successfully
in a marine environment, but also conrmed the long-
term efforts of DNV and our project partners towards
developing and facilitating the introduction of new
greener technologies (Eide and Endresen, 2010).
DNV recognized the potential of fuel cells in ships at
an early stage, and has taken a leading role in research,
development, and demonstration in order to facilitate
safe and reliable introduction of this technology on board
(Tronstad and Endresen, 2005; Mangset et al., 2008; DNV,
2008; Ovrum and Dimopoulos, 2011; Ludvigsen, 2012).
Early efforts to evaluate fuel cells for marine applications
include feasibility studies in the projects FCSHIP (FCSHIP,
2004) and FellowSHIP (Sandaker et al., 2005).
Large-scale marine concepts have been tested onshore in
the US Ship Service Fuel Cell (SSFC) project (Hoffman,
2011) and in two EU projects, FELICITAS (2006) and
MC-WAP
2
. While the FellowSHIP installation on Viking
Lady adapted land-based technology for on-board
testing, the METHAPU
3
project developed solid oxide
fuel cell technology that was tailored for use in marine
applications. METHAPU resulted in a 20 kW installation
on a car-carrier. The currently on-going PaXell
4
project
aims towards developing and building fuel cell units for
power supply on board cruise vessels. A summary of fuel
cells installed on ships and boats until 2009 is provided
in McConnell (2010) where also a number of small-scale
2 www.mc-wap.cetena.it
3 www.methapu.eu
4 www.e4ships.de/e4ships-home.html
Introduction
5
applications such as pleasure boats, sailboats, ferries, and
water taxis are listed.
DNV has supported a number of the R&D projects
mentioned above (e.g., FCSHIP, FellowSHIP, METHAPU
and PaXell), conducting feasibility studies, safety and
risk assessments, and rule development. Through these
projects, signicant competence has been built, enabling
us to facilitate future introduction of fuel cells on DNV
classed vessels. DNV also initiated the development of a
modelling platform for analysing and optimising the new
increasingly complex energy systems launched for ships
(Kakalis and Dimopoulos 2012).
The rst part of this paper introduces fuel cell technology
and highlights its advantages and challenges for marine
applications. Operational experience gained through the
FellowSHIP project is presented in the following section,
and the FellowSHIP installation is used as a case study
illustrating how advanced modelling and simulations can
facilitate safe and optimal installation of future fuel cell
power packs in a ship environment. Finally, the paper
focuses on the safety of fuel cell installations.
Figure 1: Viking Lady in Copenhagen during COP-15, 15-17 December 2009 (Photo: Sten Donsby).
6
A FUEL CELL POWER PACK consists of a fuel and gas
processing system (the balance of plant), and a stack of
fuel cells that convert the chemical energy of the fuel to
electric power through electrochemical reactions. The
process can be described similar to that of a battery,
with electrochemical reactions occuring at the interface
between the anode or cathode and the electrolyte
membrane, but with continuous fuel and air supplies, see
Figure 2. Different fuel cell types are available, and can be
characterized by the materials used in the membrane. The
most relevant types of fuel cells for ship applications are
introduced in the next section. For further information on
fuel cell technology see e.g. Larimine (2003).
The main advantages and challenges related to introducing
fuel cell technology onto ships are presented below.
Figure 2: Basic principles of fuel cells (courtesy of FuelCellToday)
ADVANTAGES:
Improved efciency
Figure 3 shows how the direct electrochemical conversion
of fuel energy to electricity in fuel cells provides fewer
sources of loss than in combustion engines. At optimal
load, the best fuel cell stacks have an electric efciency
of 50-55 %, giving a fuel to electric efciency of 45-50 %
when internal consumption is included. These values are
slightly higher than the typical values of fuel to electric
efciency for state-of-the-art marine diesel generators,
which are just above 40 %. New gas engines claim to achieve
efciencies greater than 45 %. For part load operation,
where combustion engines have lower efciencies and
emissions of local pollutants are higher, fuel cell power
packs generally maintain or even increase their efciency.
Figure 3: Electricity from electrochemical vs. combustion process
(illustration from www.vikinglady.no)
Losses in the electrochemical conversion process generate
heat that is recoverable. Depending on the type of fuel cell
technology, the amount and quality of exhaust from fuel
cell stacks are high compared with combustion engines.
Thermal integration with steam turbines or some form of
Fuel cells
- advantages and challenges
7
a Rankine cycle (i.e., converting heat into work) can thus
increase the electric efciency signicantly, as discussed in
the next section.
Reduced emissions to air
CO
2
emissions lead to global warming. By using fuels such as
liquid natural gas (LNG) or methanol that have less carbon
content than conventional ship fuels, these emissions can
be reduced. CO and CH
4
emissions can occur from fuel
cells depending on the choice of fuel, but are signicantly
lower than for combustion engines running on LNG.
When hydrogen is used as fuel, no carbon compounds are
emitted.
PM, NO
X
, and SO
X
emissions from ships can result in severe
consequences to human health and the environment (e.g.
Corbett et al. 2007; Eide et al. 2012). In the long-term, the
potential uptake of fuel cells on board could contribute
to reducing these consequences. NO
X
is formed by
combustion at high temperatures, a process that does not
occur in fuel cells, and thus NO
X
emissions from fuel cells
are negligible. As sulphur must be removed from the fuel
before it is supplied to the fuel cell, SO
X
emissions are
eliminated. PM is not emitted from fuel cells, as the fuel
cannot contain heavy hydrocarbons.
Other advantages
Due to fewer moving parts, use of a fuel cell power
plant instead of a combustion engine will reduce
noise and vibrations, improving comfort for crew and
passengers. Fewer moving parts also lead to a reduction
in maintenance requirements during operation compared
with combustion engines. Fuel cell technologies that
have a small balance of plant, i.e. when limited space is
required for fuel and gas processing, can easily be installed
in independent modules. This makes the total installation
less vulnerable to single failures and, in principle, the
modules could be placed in several different locations
around the ship. Together with lower vibrations and noise,
modularity makes the engine room location less critical.
Given that safety issues are handled appropriately, this
provides a high degree of design exibility.
CHALLENGES:
New fuels
All fuel cell types require either pure hydrogen or fuels that
can be reformed to hydrogen and CO, either before entering
the fuel cell or inside the fuel cell. The gas entering the cells
must be sulphur-free, and low temperature fuel cells have
restrictions on the amount of CO that can be tolerated.
Some projects have aimed towards reforming marine
diesel oil (MDO) to hydrogen for use with fuel cells. These
projects have not yet been successful and it seems that fuel
cells have greater potential when alternative fuels to MDO
or heavy fuel oil (HFO) can be applied on board. Thus, the
relevant short-term options are fuels such as methanol or
LNG. However, the distribution network for such fuels is
currently limited. Should hydrogen become more readily
available in the future, this will also be a relevant option.
Investment costs
The current cost of fuel cells is high. This is due to limited
market penetration and because only a few large-scale
installations are in operation. For fuel cells to be relevant for
ships, fuel cell manufacturing costs must be reduced. The
investment costs are not expected to compete directly with
combustion engines at 3-400 $/kW, and thus the lifetime
costs of the installation must be compared (investments
and operation). Fuel cell prices vary signicantly between
different fuel cell technologies. For MCFC modules, prices
have been reported to be as low as 3000 $/kW, but also
signicantly higher. A target of 1500 $/kW has frequently
8
been used as a development goal for commercialisation of
fuel cells (Escombe, 2008). Fuel cell producers claim that
this target will be achieved between 2020 and 2025.
Lifetime
Daily maintenance requirements for fuel cells are low, but
stack replacement is necessary. Fuel cell stacks have not
yet reached the goal of 40000 operating hours without
suffering from signicant performance degradation. Due
to continuous R&D efforts, fuel cell lifetime is increasing.
System design must therefore allow for replacement of
the fuel cell stacks approximately every 5 years, while the
remaining balance of plant typically has a 20-year lifetime.
Operational costs
The costs of stack replacement can be partly offset by
reduced maintenance costs compared with a combustion
engine. Due to the higher investment outlay, fuel costs need
to be lower than that of a comparable combustion engine
over the lifetime of the installation. As indicated by Mangset
et al. (2008), reduced fuel costs due to increased efciency
and a shift to cheaper alternative fuels (e.g., LNG) may
favour fuel cells in terms of life cycle costs. A possible
introduction of carbon taxes may also mean that alternatives
to MDO become more protable. In general, economics of
installation are signicantly dependent on the assumptions
made about an individual ships operating prole and fuel
consumption, and each ship must be considered on a case-
by-case basis.
Life Cycle Assessments
Life Cycle Assessments for marine fuel cell applications
have been carried out, for example by Reenaas (2005) and
Alkaner and Zhou (2006). These assessments concluded
that the environmental footprint favours fuel cells over
conventional power generator sets (diesel engines), mainly
because higher fuel efciency is assumed for the fuel cells.
In comparison with diesel engines, the production phase
has signicantly higher impact on a fuel cell units life cycle
performance. If fuel consumption is not decreased when
replacing combustion engines with fuel cell technology,
then the environmental footprint will generally increase.
However, there is a considerable potential for lowering
energy consumption under the production phase as the
technology matures (Alkaner and Zhou, 2006).
Size
The size of fuel cell installations varies with the type of
technology chosen. However, in terms of volume and weight
per kW installed, it will be hard to compete with combustion
engines, especially for fuel cell types that require a complex
balance of plant. Estimated electric efciency based on
lower heating value of the relevant fuel, specic power, and
power density are compared for two types of fuel cells power
packs (cf. next section) and two types of internal combustion
engines (4-stroke diesel and lean burn gas) in Table 1.
Electric power
generator
Electric
efciency
(%)
Specic
power
(kW/m
3)
Power
density
(W/kg)
Fuel cell (MCFC) 45-50 3 15
Fuel cell (HTPEM) ~45 30 60
Marine diesel
(4-stroke)
40 80 90
Marine gas
(4-stroke)
45 80 90
Table 1: Characteristic properties of two fuel cell types and two types of
combustion engines. Numbers are roughly estimated based on available
product documentation for the fuel cells and DNV internal Report No
2010-0605 for the combustion engines.
9
SEVERAL FUEL CELL TYPES exist, and their names reect
the materials used in the electrolyte. The properties of
the electrolyte membrane affect the allowable operating
temperatures and the nature of electrochemical reactions
and fuel requirements (Larminie, 2003; U.S. DOE,
2011). During the last decades several different fuel cell
technologies have been proposed and developed, and
their levels of maturity, realistic efciency potential, and
future prospects vary signicantly. Table 2 shows operating
temperatures, and average reported fuel to electric
efciencies (when internal consumption is included), for
the fuel cell types considered to be of most relevance for
marine applications.
Fuel cell type
Temperature
(C)
Electric
efciency
(%)
Proton Exchange
Membrane (PEM)
30-100 35-40
High Temperature PEM
(HT-PEM)
160-200 ~45
Molten Carbonate
(MCFC)
~650 45-50
Solid Oxide (SOFC) 500-1100 45-50
Table 2: Fuel cell properties, (Escombe, 2008; McConnell, 2009).
Proton Exchange Membrane Fuel Cell (PEMFC) fuelled by
hydrogen is the most widespread and developed fuel cell
technology. It operates on hydrogen, and this needs to be
of high quality as impurities will damage the membranes.
High temperature PEM (HTPEM) is a modied version
of PEMFC with a novel membrane that can withstand
temperatures up to 200C.
Higher temperature of operation enable a simpler
balance of plant, because the needs for cooling, water
management, and purication are reduced compared
with PEMFC. HTPEM fuels cells also have a higher
tolerance for CO, and are therefore more suitable for
use with reformed fuels (methanol, natural gas, and
ethanol). The PEM technologies have excellent dynamic
capabilities, and electric efciencies of around 40 % have
been demonstrated. Higher efciencies (45-50 %) have
been claimed for HTPEM due to less internal energy
consumption, but data derived from operating experience
are limited (McConnell, 2009).
PEMFC are produced in smaller units, up to 100 kW,
and are thus suited for distributed power supply. Typical
applications are cars, stationary power generation, small-
scale power sources, and combined heat and power systems.
HTPEM fuel cell units consist of independent modules,
typically 5-15 kW, with small built-in reformer units. The
modules can easily be assembled into larger power packs,
and up to 1 MW has been suggested. These technologies
are signicantly more compact than the high temperature
technologies presented in the following section.
Submarines, yachts, ferries and recreational boats have
been tted with PEM fuel cells running on hydrogen.
Examples are the 2 x 50 kW units on the ferry FCS
Alsterwasser in Hamburg
5
and the 60-70 kW installation on
the ferry Nemo H
2
in Amsterdam
6
. A 12 kW HTPEM has also
been installed on the harbour ferry MF Vgen in Bergen,
Norway
7
. A larger installation of HTPEM on a cruise vessel
will be demonstrated through the PaXell project.
5 www.zemships.eu/en/index.php
6 www.lovers.nl/co2zero/factsheet/
7 www.tu.no/industri/article263210.ece (in Norwegian)
Fuel cell types
10
Molten Carbonate Fuel Cell (MCFC) and Solid Oxide
Fuel Cell (SOFC) technologies are high-temperature fuel
cells that are exible regarding choice of fuel: methanol,
ethanol, natural gas, biogas, and hydrogen are most
commonly used. MCFC is the more mature of these two
technologies, while SOFC is considered to have the greatest
potential in terms of efciency and power density. As an
example, Figure 4 shows the electrochemical reactions
inside a MCFC, in which both carbonate and hydrogen
ions are involved in electricity production.
An electric stack efciency of 50-55 % has been obtained
from both MCFC and SOFC installations, and when internal
consumption is included this is lowered to 45-50 %. High
operating temperatures lead to high exhaust temperatures
(400-800 C) that, together with a large volume ux of
exhaust, yield a signicant potential for heat recovery. The
fuel to electric efciency can be increased to 55-60 % for
MCFC plants and to above 60 % for SOFC plants when
heat recovery is included (Escombe, 2008).
A complex balance of plant to handle fuel and air
treatment is required for both technologies and larger
units are therefore preferred. MCFC units generally have
one fuel cell stack of about 200500 kW, while an SOFC
unit is built from several smaller stacks of 1-20 kW each.
The SOFC units can be built to be signicantly more
compact than MCFC units, but the complete power packs
remain large in volume compared with diesel generators.
High-temperature fuel cells must operate at stable
temperatures, and therefore have low tolerance to rapid
load changes. In general, these fuel cell types can only be
justied in applications where power and heat demands
are high and stable.
High-temperature fuel cells are currently used for
uninterruptable power supplies in hospitals and server
parks, as well as for power generation from landll or
industrial biogas. The lack of dynamic capabilities means
that these fuel cells are best suited for providing base-load
electric power on ships. A methanol-fuelled marine SOFC
plant of 20 kW was tested on board the car carrier Undine
in 2010
8
. The largest marine fuel cell installation to date is
the 330 kW MCFC installed on board Viking Lady
9
.
The uptake of these technologies is hard to project due
to high market uncertainty and current investment costs.
The most promising avenues, to date, are the following:
Harbour-mode solutions, with the possibility of
decreasing the detrimental health effects of chemical
emissions and noise from ship trafc in urban areas,
are currently of considerable interest. Fuel cells can be
used as an alternative to cold ironing for all ship types
that have space available for fuel cells as an auxiliary
unit. The preferred fuel cell alternatives would then
be HTPEM, or a hybrid combination of batteries and
MCFC or SOFC, all running on low carbon fuels. If
hydrogen is available, PEM or HTPEM would be the
preferred choices.
Ferries operating on short routes are suitable
candidates for the rst ships powered only by fuel
cells due to their relatively low power requirements
and frequent refuelling possibilities. The same is
true for ships operating on inland waterways. Hybrid
installations, with PEM fuel cells and batteries, already
exist as pilot installations, and would provide a zero
8 www.methapu.eu
9 www.vikinglady.no
11
emission alternative if hydrogen can be produced from
renewable sources.
Cruise ships will benet from the reduction of noise
and vibrations, as well as from reduced local emissions
while in port and cruising in environmentally sensitive
areas. Most cruise ships today are diesel-electric, and a
fuel cell installation could easily be integrated into the
designs. The public perspective of a soot-free cruise will
be a huge advantage for the rst fuel cell powered cruise
ships. HTPEM units are the most realistic alternatives
due to their high specic power, see Table 1.
Figure 4: Chemical processes inside an MCFC (courtesy of MTU Onsite Energy)
12
THE SUPPLY VESSEL Viking Lady is the rst merchant
ship to have a large-scale fuel cell installation operating
on board. It is also the rst vessel to use high-temperature
fuel cell technology. The choice of fuel cell technology
was based on fuel availability and the maturity of the
available technologies. The MCFC power unit, developed
by MTU in Germany, had been successfully demonstrated
for several land-based installations, and was modied for
operation in a marine environment. The FellowSHIP
project (phase II)
10
was responsible for the modication,
installation, testing, and operation of the power pack.
When the project closed in July 2010, the system had
operated for 7000 hours, demonstrating unequivocally
that existing fuel cell technology can be integrated into a
ship environment.
LNG is the main fuel in the gas-electric propulsion system
of Viking Lady, and the vessel therefore provided a good
test-bed for MCFCs, since no additional fuel system was
needed. In the current installation, the MCFC delivers
power to a direct current (DC) link that is connected to
the ships alternating current (AC) bus through power
converters. The ships electric propulsion system therefore
consume fuel cell power equivalently to power provided by
the main generators.
The fuel cell stack, together with the required balance
of plant, is located in a large, purpose-built container
(13 x 5 x 4.4 m). Project-specic electrical components
(transformers, converters and DC bus) designed to protect
the fuel cell from potentially harmful disturbances on the
10 Partners of the FellowSHIP project, phase II, were the ship-owner Eidesvik
Offshore ASA, Wrtsil Ship Design AS, Wrtsil Norway AS, MTU Onsite
Energy, and DNV Research & Innovation, nancially supported through the
Research Council of Norway, Innovation Norway, and the German Federal
Ministry of Economics and Technology.
power grid, are situated in a standard 20-ft container; see
Figure 5. The total weight of the containers is 110 tons, but
both weight and volume could be signicantly reduced
in future fully integrated systems. Since the FellowSHIP
installation was retrotted, the current design allows for
temporary installation on deck, as well as an onshore test
period for the whole power pack, and therefore the size
and weight were not optimised.
An onshore test period ensured proper functionality of all
power and control interfaces between the two containers
and the overall safety systems, and minimised the time
needed for hook-up and modications on board. Viking
Lady was delivered for operation on the North Sea in
April 2009, and, in September of the same year, the 330
kW MCFC power pack was installed. The fuel cell was
connected to the main switchboard for the rst time in
December 2009. After initial testing, Viking Lady became
the rst vessel to obtain the class notation FC-Safety, as
described in the DNV Rules (DNV, 2008).
During its rst year in operation, the fuel cell stack showed
no signs of degradation, indicating that the measures
taken to protect the fuel cells were appropriate. The stack
was protected against electric disturbances. In addition,
ship movements, hull vibrations, and air salinity were also
taken into consideration when designing the fuel cell
stack and its container and support systems. In January
2012, the fuel cell was cooled down and conserved for
future demonstration projects. A total of 18 500 operating
hours were logged without signs of severe performance
degradation. Approximately half of the time logged was
in idling mode.
Fully loaded, the fuel cells produced electricity at a
measured electric efciency of 52.1 % based on the lower
Fuel cell testing
on-board Viking Lady
13
Figure 5: Stord, Norway, September 2009: Installing containers on Viking Lady, the white for fuel cells and the blue for power electronics.
heating value of the LNG. Although exact measurements
of gas to grid efciency were not possible for the
current system setup, this was estimated to be 48.5 %
including internal consumption, and 44.5 % when DC/
AC conversion was also accounted for. A heat exchanger
that produced warm water from the fuel cell exhaust was
tested, with about 80 kW heat recovered. This increased
the overall fuel efciency to slightly above 55 %. With
optimal system integration, there is the potential for
increasing the electrical efciency to close to 50 %, and
the fuel efciency up to 60 %.
The FellowSHIP installation is not classed as main or
auxiliary power, but is considered as supplementary power.
14
Viking Ladys main engines must always be on-line. In
some operating conditions, e.g., in calm weather and while
docked in harbour, a load reduction is required for the
fuel cell in order to avoid the main engines operating at
unfavourable load conditions. Load changes on an MCFC
must be slow, but when load changes are performed in
a controlled manner, operating at low-load conditions is
not dangerous for these fuel cells and may even prolong
the lifetime of the stack. Nevertheless, the aim is for
continuous operation at close to full load in order to allow
full exploration of the environmental benets and fuel-
saving potential of this technology.
Although the cost, weight, and volume of the test installation
were high, the feasibility of installing and operating a fuel
cell power pack in a marine environment was successfully
demonstrated on board Viking Lady. In future marine
MCFC designs, more focus should be directed towards
thermal integration, utilizing the high quality exhaust
heat, and on including some form of energy storage to
allow for stable load conditions for the fuel cell.
In order to explore the potential benets of combining
fuel cells and engines with energy storage, Viking Lady will
host another large-scale technology demonstrator. The
FellowSHIP project is to be continued, and a battery pack
for hybrid operation with the combustion engines and the
fuel cell will be installed within 2013 (Eason, 2012). The
project will also make modications to allow the ship to be
powered only by the combined fuel cell and battery power
pack during certain test runs and in harbour mode.
15
IN THE FELLOWSHIP PROJECT, DNV developed a detailed
mathematical model describing the thermodynamic
behaviour and transport phenomena inside the marine
MCFC installed on Viking Lady. The main focus of the model
development was to analyse operation and performance of
the fuel cells under steady and dynamic conditions. The
model was calibrated by measurements on board Viking
Lady, and details of the implemented model and results
have been published in Ovrum and Dimopoulos (2011).
A similar modelling approach will be followed for the
HTPEM fuel cells in the on-going PaXell project.
Modelling and simulation have been used extensively
to analyse fuel cells and their potential in marine power
systems (Bruun, 2009; Bensaid et al., 2009; San et al.,
2010). Our modelling work follows the key concepts and
approach developed in the DNV COSSMOS (acronym
for Complex Ship Systems Modelling & Simulation)
framework (Dimopoulos and Kakalis, 2010). Within
this framework, DNV develops model-based methods
and a computer tool for the synthesis, design, and
optimisation of integrated marine machinery systems
(Kakalis and Dimopoulos, 2012). The fuel cell models
enable investigation of thermal and electrical integration
of the fuel cell for design and optimisation of ship power
production systems for different modes of operation.
Figure 6 shows that the model predictions agreed well with
the actual MCFC unit on board Viking Lady. The model
was calibrated against a range of data from on-board
measurements, and nally validated against a different set
of test data. The model shows an average error of about
4 %, as seen in the gure below, where all the values are
relative to a reference value and therefore dimensionless.
The following examples show the model in use:
Dynamic modelling
of fuel cells
Figure 6: Calibration and validation of the MCFC (Ovrum and Dimopoulos, 2011).
16
Reliability of power production is of paramount
importance in the marine setting, as a ship should always
have the power to return to port. Studies of temperature
distribution, shown in Figure 7, can provide valuable
information on how the use of the fuel cell affects its
lifetime, since hot spots and large temperature variations
in time degrade the cell performance. By simulating the
dynamics of the system, the loads and fuel utilization that
the fuel cell must endure on ships with different operating
proles can also be investigated. Thus, modelling can
manage, and potentially improve, the lifetime and
reliability of a fuel cell system.
Capability of representing transients is essential in order
to estimate how a novel system responds to critical events.
As an example, loss of fuel ow can be simulated as shown in
Figure 8. The fuel ow is reduced by 1 % of its original ow
every second, increasing the fuel utilization, until after 40
seconds the system regains control. Another critical event
is overheating of a fuel cell system due to loss of gas ow.
With such high values of fuel utilization or temperatures,
the MCFC might be degraded, reducing its performance
and lifetime. These are examples of using modelling to
supplement potentially destructive experiments for design
studies of novel marine machinery systems.
Figure 7: Temperature distribution in fuel cell membrane at 100 % load
(Ovrum and Dimopoulos, 2011).
Figure 8: Loss of fuel ow and the consequences for current density
(Ovrum and Dimopoulos, 2011).
17
Preliminary studies of marine fuel cells (FCSHIP, 2004)
focused on gaining a common understanding of basic
safety requirements. The studies concluded that safe fuel
cell systems are technically possible, but that there was
a lack of standard guidelines and rules to facilitate the
design and approval process. The rst rules for fuel cells
were published by DNV in 2008 (DNV, 2008), and class
guidelines were issued by GL in 2003 (GL, 2003). In the
DNV rules, there are two different class notations for fuel
cells; FC-SAFETY is mandatory for all fuel cell installations,
and, if the fuel cell unit is used for main or auxiliary
power, the class notation FC-POWER is also mandatory
(Table 3). An important part of the FellowSHIP project
was to develop and implement these rules to allow for safe
installation on Viking Lady.
The main safety hazard to be handled with on-board fuel
cells is the introduction of new fuels with low ammability
limits such as LNG, methanol, or hydrogen. This sets
requirements for sufcient ventilation, alarm systems, and
re protection, as well as introducing other measures to
limit the likelihood and consequences of a gas leakage.
LNG as a fuel is well-covered by rules, and there is
signicant experience with such installations.
However, there is far less experience with ship-borne
hydrogen or methanol installations. The use of methanol
on board was demonstrated in the METHAPU project. In
the FellowSHIP project, in which LNG was used as main
fuel and hydrogen as auxiliary gas, safety measures were
implemented that resulted in a FC-SAFETY notation for
Viking Lady.
Reliability and redundancy are very important issues if
fuel cells provide propulsion or auxiliary power, according
to the FC-POWER notation. This also sets requirements
to the control systems and the interface with the ships
overall power distribution.
In order to stay ahead of the technological development,
DNV is positioned at the forefront of development of
rules for new technologies. In this context, a continuation
of the FellowSHIP project was kicked-off in 2011. This
project, named HybridShip, is concerned with introducing
batteries for on-board energy storage, integrated with fuel
cells and gas engines.

A 200 Class notation
201 Ships where the fuel cell power is used for essential,
important or emergency services shall satisfy the
requirements in this rule chapter and will be given class
notation FC-POWER.
202 Ships where the fuel cell power is not used for essential,
important or emergency users shall satisfy the safety and
environmental requirements. Installations complying with the
requirements in this chapter, except Section 2 will be given
class notation FC-SAFETY.
Table 3: Extract from DNV Rules for classication of Ships, Pt.6 Ch.23:
Fuel cell installations
Safe operation
18
The FellowSHIP demonstration project, led by DNV,
developed and installed a 330 kW fuel cell power pack
on board the offshore supply vessel Viking Lady. This
was the rst large-scale fuel cell unit to be installed on a
commercial ship. The system delivered power to the ship
grid for over 7000 hours, demonstrating unequivocally
the applicability of fuel cells for ships. The fuel cell unit
on board Viking Lady had an overall efciency of above
55 % when heat recovery was included. DNV rules for
introducing fuel cells were developed, and the DNV
class notation FC-SAFETY was obtained by Viking Lady.
This ensured safe integration of new fuels (LNG and
hydrogen), as well as safe integration of fuel cells into the
ships power system.
High fuel efciency over a wide range of loads and
elimination of emissions of SO
X
, NO
X
, and PM, thereby
avoiding local consequences of air pollution from ships
on human health and the environment, are, together
with reductions in noise and vibrations, the main benets
from introducing fuel cells to ships. CO
2
emissions are
also reduced, or even completely eliminated if hydrogen
from renewables becomes available, thus lowering the
contribution from shipping to global warming.
The electrical efciency of fuel cell stacks depends upon
the fuel cell technology, with values ranging from 35-50
%. This efciency is only slightly higher than the values
claimed from generating electricity using state-of-the-
art combustion engines. Therefore, optimal system
integration, resulting in additional electric and thermal
power, is essential. Signicant reduction in costs is also
required if the fuel cell technologies discussed in this
paper are to become competitive for ships. With the
recent commercialisation of certain land-based fuel cell
applications, there is reason to believe that costs will fall.
For ship applications, reductions in size and weight are
also of immense importance.
DNV recognizes that fuel cells can become a part of
the future power production on ships. By leading and
participating in large R&D and pilot projects, we have
built competence and developed rules, thereby paving
the way for safe and smooth introduction of fuel cells
for ships. Methods to enable assessments of new fuel cell
designs and their system integration through modelling
and simulation have also been developed, in support of
DNVs class and advisory services. We recognize that it will
take time before fuel cells can become a realistic on-board
alternative; this is mostly because of price, but also because
of limited product development tailored to the maritime
market. National and regional incentive schemes for
environmentally friendly technologies could also play a
central role regarding when fuel cells can become cost-
competitive. Increased availability of alternative fuels, such
as LNG and hydrogen, may also accelerate introduction.
The FellowSHIP project has taken some important rst
steps towards a possible future for fuel cells on ships.
It is concluded that fuel cells for shipping require further
R&D before this technology can complement existing
powering technologies. However, in the near future we
might expect to see successful niche applications for some
specialised ships, particularly with hybrid systems.
Conclusions
19
References
Alkaner, S. and Zhou, P. (2006). A comparative study on life cycle analysis
of molten carbon fuel cells and diesel engines for marine application.
Journal of Power Sources, 158 (1). pp. 188-199. ISSN 0378-7753.
Bensaid, S., et al. (2009). MCFC-based marine APU: Comparison between
conventional ATR and cracking coupled with SR integrated inside the
stack pressurized vessel. International Journal of Hydrogen Energy, Vol.34,
Issue 4, pp.2026-2042.
Biello, D. (2009). Worlds First Fuel Cell Ship Docks in Copenhagen.
Scientic American 21. December, 2009. http://www.scienticamerican.
com/article.cfm?id=worlds-rst-fuel-cell-ship
Bruun, K. (2009). Bond Graph Modelling of Fuel Cells for Marine Power
Plants. Norwegian University of Science and Technology, Department of
Marine Technology, Doktoravhandlinger ved NTNU 2009:161.
Corbett, J.J., Winebrake, J.J., Green, E.H., Kasibhatla, P., Eyring V. and
Lauer A. (2007). Mortality from Ship Emissions: A Global Assessment.
Environ. Sci. Technol., 2007, 41 (24), pp 85128518
Dimopoulos,G. G. and Kakalis, N. M. P. (2010). An integrated modelling
framework for the design, operation and control of marine energy
systems. Proceedings of 26th CIMAC World Congress, Bergen, Norway.
DNV (2008). DNV Rules for classication of Ships, Pt.6 Ch.23: Fuel cell
installations July 2008.
Eason, C. (2012). Rechargeable shipping. Lloyds List Tuesday March 20, 2012.
Eide, M.S. and Endresen, . (2010). DNV Position Paper No.5
Assessment of measures to reduce future CO2 emissions from
shipping. http://www.dnv.com/resources/position_papers/assessment_
future_emissions_co2_shipping.asp
Eide et al. (2012). Reducing CO
2
from shipping Do non-CO
2
effects
matter? Submitted, Atmospheric Chemistry and Physics.
Escombe, F. (2008). Fuel Cells: The Sourcebook, EscoVale Reports: REF
3408.
FCSHIP (2004). Fuel Cells in Ships Synthesis of open problems and
Roadmap for future RTD, European Commission. Final Report.
FELICITAS (2006). FELICITAS - fuel cells for heavy duty applications.
Warship Technology. May: 12-13.
Fuel Cell Today (2011). The Fuel Cell Today Industry Review 2011.
Available at: http://www.fuelcelltoday.com/analysis/industry-review/2011/
the-industry-review-2011
GL (2003): GL Rules and Guidelines VI: Additional Rules and Guidelines,
Part 3, Chapter 11 Guidelines for the Use of Fuel Cell Systems on Board
of Ships and Boats.
Hoffman, D. (2011). SYSTEM DESIGN: Lessons Learned, Generic Concepts,
Characteristics & Impacts. Presentation available at: http://www1.eere.
energy.gov/hydrogenandfuelcells/pdfs/apu2011_11_hoffman.pdf
Kakalis, N. and Dimopoulos, G. (2012). DNV Position Paper No.11
Managing the complexity of marine energy systems.
Larminie, J. and Dicks, A. (2003). Fuel Cell Systems Explained, Second
Edition. ISBN-10: 0768012597
Ludvigsen, K. Bruun (2012). Fuel cells as an alternative power source: -
results and learning from the FellowSHIP project, Green Ship Technology
2012, 27 -29 March, Copenhagen
Mangset, L.E., Longva, T. and Tronstad, T. (2008). Analysis of the
Economy of Fuel Cells Used in Merchant Ships. 30th Propulsion &
Emissions Conference, Gothenburg 2008-05-21.
McConnell, V.P. (2010). Now, voyager? The increasing marine use of fuel
cells. Fuel Cells Bulletin, Volume 2010, Issue 5, May 2010, Pages 12-17.
McConnell, V.P. (2009). High-temperature PEM fuel cells: Hotter, simpler,
cheaper. Fuel Cells Bulletin, Volume 2009, Issue 12, December 2009,
Pages 12-16.
Ovrum E. and G. Dimopoulos (2011). A validated dynamic model of the
rst marine Molten Carbonate Fuel Cell. Applied Thermal Engineering,
Available online 24 September 2011, ISSN 1359-4311
Reenaas, M. (2005). Solid Oxide Fuel Cell Combined with Gas Turbine
versus Diesel Engine as Auxiliary Power Producing Unit onboard a
Passenger Ferry. NTNU Master Thesis 2005.
San, B., Zhou, P. and Clealand, D. (2010). Dynamic modeling of
tubular SOFC for marine power system. Journal of Marine Science and
Application, Volume 9, Number 3, 231-240.
Sandaker, K., Svardal, B., Vestbstad, L., Tronstad T. (2005). FellowSHIP
Phase 1 Summary Report, FellowSHIP Project Report no: 1-0-D-2005-1
Tronstad, T. and Endresen . (2005). FellowSHIP - Fuel Cells for Low
Emission Ships. Published at: Technologies for Sustainable Energy
Development in the Long Term, Ris National Laboratory, Denmark 23-25
May 2005.
U.S. Department of energy (2011). Type of fuel cells. Available at http://
www1.eere.energy.gov/hydrogenandfuelcells/fuelcells/fc_types.html
Vielstich, W., H.A. Gasteiger, A. Lamn (Eds.), (2003). Handbook of Fuel
CellsFundamentals, Technology and Applications. VOL. 1, John Wiley.
ISBN: 978-0-471-49926-8.
Det Norske Veritas
NO-1322 Hvik, Norway
Tel: +47 67 57 99 00
www.dnv.com
D
e
s
i
g
n
,

l
a
y
o
u
t

a
n
d

p
r
i
n
t

p
r
o
d
u
c
t
i
o
n
:

E
r
i
k

T
a
n
c
h
e

N
i
l
s
s
e
n

A
S
,

0
6
/
2
0
1
2
P
r
i
n
t
e
d

o
n

e
n
v
i
r
o
n
m
e
n
t
a
l
l
y

f
r
i
e
n
d
l
y

p
a
p
e
r
.