See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/237445845

Marine Technologies for Reduced Emissions

Conference Paper · April 2005

CITATIONS READS

20 2,878

2 authors, including:

German Weisser
Winterthur Gas & Diesel
60 PUBLICATIONS 459 CITATIONS

SEE PROFILE

All content following this page was uploaded by German Weisser on 02 June 2019.

The user has requested enhancement of the downloaded file.

 Marine Technologies for Reduced Emissions1
Heinrich Schmid2 and German Weisser3
Wärtsilä Switzerland Ltd, Winterthur

Summary
The exhaust emissions from ships are best controlled through optimisation of ship design and operation, in other words
by consideration of the ship as part of a whole transportation system. This paper reviews several aspects of ship design and
operation which influence the level of exhaust emissions from ships.
The paper also outlines the various technologies available, being developed or already in use in Sulzer low-speed two-
stroke marine diesel engines which have beneficial influences on further lowering of the exhaust emissions. These include
Low-NOX Tuning, electronically-controlled common-rail systems (Sulzer RT-flex engines), direct water injection with
exhaust gas recirculation, and aftertreatment by selective catalytic reduction (SCR). A new Total Heat Recovery concept
offers a route to a significant drop in overall emissions, including carbon dioxide (CO2).
With respect to the emissions of sulphur oxides (SOX), the paper also covers some aspects concerning the use of low-
sulphur fuels in low-speed two-stroke marine diesel engines.

types will be regulated, to what extent and when.

Introduction
Atmospheric emissions have been a major topic for many
years between political and shipping circles. More recently,
however, the political climate has hardened with the
subject being raised from a matter of local pollution to
one of ‘global warming’.
Exhaust emissions originating from land transport
and electricity generation are already heavily regulated
with very low limits, resulting in considerably reduced
emissions in absolute terms. By comparison, shipping has
not yet been greatly affected and, in fact, the emissions
in absolute terms are rising with the growth in sea-borne
trade. Shipping consumes some five per cent of global oil
consumption [1, 2] which leads to global NOX emissions
of about 12.57 million tonnes/year, and about 10.54
million tonnes/year global SOX emissions [1].
In the USA, ports have been identified as key polluters.
In Los Angeles alone, the ports daily emit some 31.5
tonnes of NOX which is equivalent to the NOX emissions
from about 650,000 cars [3].
Clearly, stricter air pollution control regulations will
come for shipping. Yet it is not known which emissions
1
This paper was presented at the 2nd Annual
Conference on Green Ship Technology, Amsterdam,
the Netherlands, 13/14 April 2005.
2
Heinrich Schmid is General Manager, Application
Development & Support, Ship Power, Wärtsilä
Switzerland Ltd.
3
German Weisser is Manager, Engine Performance,
Technology/R&D, Wärtsilä Switzerland Ltd.
The Kyoto Protocol concerning ‘greenhouse gases’
(mainly CO2) came into force on 16 February 2005.
IMO, as the responsible body for shipping under the
protocol, is already addressing the subject. In this context,
world shipping has been reported as generating some 438
million tonnes/year of CO2 which is equivalent to about
1.8 per cent of global CO2 emissions [4].
The first major regulation of air pollution from
ships, embodied in Annex VI of the MARPOL 73/78
convention which was established through IMO, will
come into force on 19 May 2005. In addition to global
regulations for NOX emissions and a cap on bunker fuel
sulphur contents, the MARPOL Annex VI regulations also
have provisions for tighter restrictions on SOX emissions in
designated SOX Emission Control Areas (SECA).
Marine engines have, since January 2000, been
manufactured in compliance with the NOX regulations of
this MARPOL Annex. As a result, ships delivered since
that date have NOX emissions of some 25 to 30 per cent
less than previous generations of ships.
Officials and politicians in the European Union and
other countries are already discussing the extension of the
first SECA and the introduction of other SECAs.
Yet we must always remember that shipping is the most
efficient form of freight transport and we are dependent
upon it for international freight transport. Despite the
great contribution to global emissions in absolute terms,
shipping also generates the least emissions by tonne-km of
freight transport, less than aircraft, road and rail.
Over past decades, the shipping industry has achieved
significant improvements in engine fuel efficiency and

—1— © Wärtsilä Corporation, April 2005

hull designs, and is using ships with larger cargo-carrying
capacities. The result has been reductions in both fuel
consumption and exhaust emissions. For example,
consider the developments in container shipping.
In 1972, the state-of-the-art was the ‘Liverpool Bay’
class container liners of OCL. These carried 2300 TEU
at a service speed of 26 knots. They each had twin-screw
propulsion by a steam turbine plant with a total output
of 59,664 kW. A typical fuel consumption for such
steam turbine plant was probably 270 g/kWh, giving a
consumption for the ‘Liverpool Bay’ class of 242 g/TEU-
mile.
Today’s state-of-the-art is a typical Post-Panamax
container liner carrying 8000 TEU and powered by a
Sulzer 12RT-flex96C common-rail diesel engine with
an maximum continuous power of 68,680 kW. It has a
service speed of 25 knots at 90 per cent service power,
with a BSFC (brake specific fuel consumption) of 166.8
g/kWh. This results in a fuel consumption of 51.5 g/TEU-
mile.
Thus with 15 per cent more power than the ‘Liverpool
Bay’ 30 years earlier, the modern ship moves about
248 per cent more containers. Combining this gain in
transportation efficiency with the change from steam
turbine plant to a low-speed diesel engine, and the gains
in engine efficiency over the years (a 15 per cent reduction
in BSFC from the Sulzer RND90M engine type of 1974
to the Sulzer RT-flex96C engine type today), the fuel
consumption in terms of TEU-mile is reduced to 21 per
cent of the state-of-the-art 30 years ago.
The emissions profiles of the two ships are not really
comparable as the only form of emissions control for the
steam plant was a ‘clear exhaust’ whereas the new ship
complies with the IMO regulations for NOX emissions.
Nevertheless the decrease in exhaust emissions would be
broadly equivalent to the reduction in fuel consumption as
many types of emissions are proportional to the quantity
of fuel consumed.
Exhaust emissions from shipping
The exhaust emissions from marine engines comprise the
following groups:
• NOX (nitrogen oxides)
These are fully regulated under Annex VI of the
MARPOL 73/78 convention.
They mainly relate to the combustion process in the
engine, but some originate from nitrogen in the fuel.
• SOX (sulphur oxides)
Although partially regulated under Annex VI of the
MARPOL 73/78 convention, they are seen as a local
problem and are principally regulated in SECA.
They are dependent completely on the fuel sulphur
content.
• Smoke
This is regulated locally. It is related to engine processes.
• CO2 (carbon dioxide)
This is regarded as a global problem and is a
‘greenhouse gas’ under the Kyoto Protocol but is not
regulated at present. It is the subject of discussions in
—2—
IMO. CO2 emissions are directly proportional to the
quantity of fuel burned.
• HC, CO, particulates
These are not regulated. They are related to engine
processes.
Although the individual types of emissions can, to
varying degrees, be controlled individually, there are trade-
offs between emission types which must be taken into
consideration. For example, a control solution for CO2
typically increases NOX emissions, and vice versa.
Ship parameters influencing emissions
Although technologies are being applied to marine engines
to reduce their exhaust emissions, and others are being
developed, much can be gained in terms of emissions by
optimising the ship design and its operation, even the
overall transport system.
For example, it can be shown that large ships need
less fuel to transport a unit of cargo (tdw-mile or TEU-
mile) and therefore discharge less emissions (Figs. 1 to
5). Furthermore large ships can be operated faster than
smaller ships for the same fuel requirement in terms of
tdw-mile or TEU-mile. This may appear counter to logic
but it can be demonstrated.
In the following considerations, each case refers to
individually optimised designs based on equivalent
technological levels. The charts used here are based on
containership examples but the approach can be applied
equally to other ship types.
The generated emissions are a function of the
propulsion power. The design propulsion power required
for ships increases with the design speed so that a power
factor can be calculated for a given class of container ships
based on a factor of 1.0 for the 26-knot ship (Fig. 1).
Similarly, a time factor can also be derived to express
the speed of cargo transport. For a given trade volume,
the faster ship can transport more cargo in a given time,
or more slower ships are needed to transport the same
quantity of cargo in the same given time. Thus the time
factor is inversely proportional to the design ship speed,
again with time factor 1.0 for the 26-knot ship (Fig. 2).
Multiplying the two factors, for propulsion power and
time, gives an energy factor indicating the energy needed
to move a given volume of cargo at different design ship
speeds (Fig. 3). By assuming that exhaust emissions are
proportional to the energy factor, it becomes apparent that
the faster ships generate greater quantities of emissions in
absolute terms.
When the propulsion power is put into relationship
with the cargo carried, namely the power requirement
per tonne deadweight (tdw) at a given speed, it becomes
apparent that larger ships require less power to transport
one cargo unit than smaller ships (Fig. 4). In turn, this
means that the atmospheric emissions from larger ships are
less, per cargo unit, than for smaller ships.
Superimposing the time factor on the power
requirement per tdw curves gives a series of curves of
energy requirement in kWh/tdw-mile at constant ship
© Wärtsilä Corporation, April 2005
 Propulsion power factor Time factor
1.2 1.6

1.0
1.4
0.8
0.6 1.2

0.4
1.0
0.2
0.0 0.8
17 18 19 20 21 22 23 24 25 26 27 17 18 19 20 21 22 23 24 25 26 27
Ship speed, knots Ship speed, knots

Fig. 1: Variation in propulsion power factor with design ship Fig. 2: Variation in time factor with design ship speed for
speed for container vessels. container vessels, indicating that more slower vessels are
[05#015] needed to transport the same cargo as fewer faster vessels.
[05#016]

Energy factor Ship service power

1.2 per ship capacity
1.8
1.0 1.6
0.8 1.4
1.2
0.6 1.0
0.8
0.4
0.6
0.2 0.4
0.2
0.0 0.0
17 18 19 20 21 22 23 24 25 26 27 0 20,000 40,000 60,000 80,000
Ship speed, knots Ship capacity, tdw

Fig. 3: Variation in energy factor, given by multiplying the Fig. 4: Trend of ship service power per unit cargo according to
propulsion power and time factors, with design ship speed for ship size based on a design ship speed of 22 knots.
container vessels. [05#017] [05#018]

Fig. 5: Combination of multiple ship power curves, with each curve being for a different constant speed (as in figure 4), reveals that
for a given energy expenditure, larger ships can sail faster than smaller ships, reducing the air emissions in terms of tdw-mile or
TEU-mile. In the examples shown, the same energy per tdw-mile is required for:
• 80,000 tdw vessel (about 7000 TEU) at 26 knots,
• 47,000 tdw vessel (about 5000 TEU) at 24 knots,
• 30,000 tdw vessel (about 3000 TEU) at 22 knots,
• 20,000 tdw vessel (about 1800 TEU) at 20 knots. [05#019]

Energy per cargo unit, kWh/tdw-mile

0.12

0.10
26 knots
0.08 24 knots

0.06 22 knots

20 knots
0.04

0.02

0.0
0 20,000 40,000 60,000 80,000
Ship capacity, tdw

—3— © Wärtsilä Corporation, April 2005

speeds (Fig. 5). These show that, for a given energy
expenditure and thus a given emissions quantity, larger
ships can sail faster than smaller ships.
Not only do those ship designs have the same energy
usage, they also have the same environmental impact in
terms of emissions per cargo unit.
Operating large, fast ships is thus of clear economic
benefit to ship operators. It is also advantageous to operate
large, fast ships from an environmental standpoint. The
larger, faster ships reduce the air emissions in terms of
mass per tdw-mile or TEU-mile.
Engine technologies influencing emissions
Once a ship in terms of its overall design and operating
pattern is optimised for economy and low emissions, it
still needs a low-emissions propulsion plant.
A range of technologies are now available or are
under development to control exhaust emissions. These
predominantly concern NOX emissions, but some affect
the overall level of all emissions. Some technologies, such
Pre-injection Triple-injection Sequential injection
–20 0 20 40 60 –20 0 20 40 60 –20
as Low-NOX Tuning, concern the adaptation of engine
processes. Other forms of emissions control such as SCR
aftertreatment and waste heat recovery involve add-on
systems. The technologies incorporated in RT-flex engines
call for new directions in diesel development, while low-
sulphur fuels present logistic problems.
The benchmarks for comparison of these emissions
control technologies are the standard Sulzer RTA and
RT-flex engines. As standard, all Sulzer RTA and RT-flex
low-speed marine engines manufactured today comply
with the IMO regulation on NOX emissions as set out
in Annex VI to the MARPOL 73/78 convention. This is
achieved by IMO Tuning which uses internal measures to
adapt the engine tuning and fuel injection.
The measures for complying with the then pending
IMO regulation were introduced for Sulzer engines
towards the end of 1996, and the first engines certified to
comply with this regulation were built in 1999 [5].
No additional equipment is applied to the engines for
IMO compliance. It involves applying a series of standard
measures, namely increased compression ratio, delayed
injection timing and adapted exhaust valve timing, as
well as different fuel nozzles. These are applied in various
combinations according to the degree of NOX reduction
necessary for the specific engine type and its rating point.
This solution is simple and effective yet has no detrimental
effect on engine reliability and has only minimal effect on
fuel consumption.
The precision and flexibility in engine setting achieved
in the Sulzer RT-flex common-rail engines (see below) also
facilitate their compliance with the IMO NOX regulation.
Low-NOX Tuning and Low-NOX Injection
The measures employed in IMO Tuning necessary to
comply with the IMO regulation can be applied in
both RTA and RT-flex engine types to achieve further
reductions in NOX emissions to perhaps five per cent
below the IMO limit. This technique is designated Low-
NOX Tuning and can incur a fuel penalty of some
2 g/kWh greater BSFC.
One must always remember that there is a trade-off
between fuel consumption and NOX emission levels.
Decreasing one causes the other to increase, and vice versa.
This is caused by the underlying fundamentals of the laws
Fig. 6: Injection characteristics
for three different fuel injection
Needle lift patterns which can be obtained
using the Sulzer RT-flex common-
rail fuel injection system. In
pre-injection and triple injection,
Fuel rail pressure
the three injectors in each cylinder
operate in unison, while for
Cylinder pressure
sequential injection the three
Injection pressure injectors operate in turn separately.
0 20 40 60
[05#020]
Crank angle, degrees
of physics and combustion chemistry.
Sulzer RT-flex common-rail engines, however, have
greater capabilities than the mechanically-controlled
RTA engines to reach lower NOX emission levels through
tuning measures alone. The fully-integrated electronic
control systems of RT-flex engines give them greater
flexibility and accuracy in engine setting than is possible
with the fixed timing of the camshaft systems in RTA
engines. At the same time, the freedom in RT-flex engines
for selection of the fuel injection pressure enables them to
have better combustion.
Selective injection patterns in RT-flex engines (Fig. 6)
together with Low-NOX Tuning techniques are expected
to give an option of reducing NOX emissions to perhaps
20 per cent below the MARPOL limit. This Low-NOX
Injection option could be appropriate for times when
lower NOX emissions are required by local regulation.
Again, owing to the inevitable trade-off between NOX
emissions and BSFC, there is a consequent fuel penalty of
about 4 g/kWh compared with the standard, MARPOL-
compliant engines. Though the normal fuel economy
would be possible when operating elsewhere on the world’s
seas.
—4— © Wärtsilä Corporation, April 2005
Sulzer RT-flex engines
In recent years a series of six Sulzer RT-flex engine types
with fully-electronically controlled common-rail systems
has been introduced to suit all types of ocean-going ships,
with maximum continuous power outputs from 5800 to
80,080 kW.
Their common-rail systems provide for the principal
functions of fuel injection and exhaust valve operation.
Fuel oil (at nominally 1000 bar) and servo oil (at
pressures up to 200 bar) are delivered by a gear-driven
supply unit mounted on the engine side, at the driving
end for engines with up to seven cylinders and at a mid
gear drive for greater cylinder numbers. The fuel supply
pumps are at one side of the drive gear with the hydraulic
servo-oil pumps on the other side. The numbers, size and
arrangement of pumps are adapted to the engine type and
cylinder number.
Control oil at a constant pressure of 200 bar is
delivered by duplicated motor-driven pumps. Both servo
and control oil are drawn from the engine lubricating oil
Fig. 7: Usual operation:
Sulzer RT-flex engines are unique in that all nozzles
in unison
individual fuel injectors can be cut out as
required. At most engine speeds the engines
operate on all fuel injectors acting in unison in
each cylinder. As speed is reduced, at first one
injection valve is shut off and then at very load
speeds another injection valve is shut off leaving
the engine running on all cylinders but with just
one injector in operation in each cylinder.
[03#118]
system through an automatic self-cleaning fine filter.
The supply unit delivers to the common rails arranged
in the rail unit mounted along the side of the cylinder
covers. The electronic control system is arranged in a
number of control units mounted on the outside of the
rail unit.
Fuel rail pressure is regulated according to engine
needs by suction control of the fuel supply pumps with
helix-controlled filling volume regulation. Heated fuel is
delivered from the fuel rail through a separate injection
control unit for each engine cylinder to the standard
fuel injection valves which are hydraulically operated by
the high-pressure fuel oil. The injection control units,
using quick-acting rail valves, regulate the timing of fuel
injection, control the volume of fuel injected, and set the
shape of the injection pattern. The RT-flex common rail
system provides free selection of these engine settings,
together with precision in the volume of fuel injected to
individual cylinders.
The three fuel injection valves in each cylinder cover
are separately controlled so that they may be programmed
to operate separately or in unison as necessary (Fig. 7).
—5—
The possibility of separate operation of the multiple fuel
injection valves in an engine cylinder is a unique feature of
Sulzer RT-flex engines.
Exhaust valves are operated by a hydraulic ‘push rod’ as
in RTA engines but the actuating energy now comes from
the servo oil. There is thus complete flexibility also in valve
timing and operation.
The principal benefits of Sulzer RT-flex electronically-
controlled common-rail engines are:
• Smokeless operation at all running speeds
• Very low, stable running speeds at 10 per cent nominal
speed
• Reduced part-load fuel consumption
• Easy engine setting for less maintenance
• Longer times between overhauls (TBO) expected,
primarily through better load balance between
cylinders and cleaner combustion at all loads.
Most of the benefits of the RT-flex system stem from
the complete flexibility in engine setting possible with
electronically-controlled common-rail systems.
Selective cut-out of injection valves
Smokeless operation at low speed
Two nozzles
Single nozzle
Time
It has become clear from the development of Sulzer
RT-flex engines that common-rail technology has opened
up a completely new era in the development of low-speed
marine engines. The benefits of the flexibility in engine
setting given by common-rail technology are certainly
applicable to any ship type. They offer interesting ship
power solutions for both today’s and future ship needs.
Smokeless operation
Smokeless operation is a visible benefit of the flexible
engine setting in RT-flex engines. Smokeless operation
at all running speeds is achieved by maintaining the fuel
injection pressure at the optimum level right across the
engine speed range. In addition, selective shut-off of single
injectors and an optimised exhaust valve timing help to
keep smoke emissions below the visible limit at very low
speeds (Fig. 8). The selective shut off of injectors gives
more balanced engine operation than cutting out whole
cylinders as required by systems with individual fuel
injection pumps.
The superior combustion performance indicated by
smokeless operation also results in a cleaner engine with
© Wärtsilä Corporation, April 2005
 Filter Smoke Number, FSN
0.50

0.40 HFO
380 cSt
3% sulphur
0.30 0.1% ash
ON OFF Aux. Blower Fig. 8:
0.20
Smoke measurements from the Sulzer
Smoke visibility limit
6RT-flex58T-B of the ‘Gypsum
0.10
Conventional low-speed engine Centennial’ during sea trials, compared
with an equivalent Sulzer RTA
6RT-flex58T-B with Common Rail engine. According to a combinator
0.00
0 10 20 30 40 50 60 70 80 90 100 characteristic.
Engine Load, % [02#010]

both less deposits of combustion residues inside and less
fouling of exhaust-gas economisers.
Delta Tuning
Advantage has been taken of the flexibility of engine
setting with RT-flex engines to provide an alternative
fuel consumption curve as standard to give lower BSFC
in what is for many ships the main operating range (Fig.
9). Through Delta Tuning [6], the BSFC is lowered in
the mid- and low-load operating range below 90 per
cent engine power. The consequent increase in NOX in
that operating range is compensated by reducing NOX
emissions in the higher-load range. The result is that, for
both BSFC curves, the engines comply with the IMO
NOX regulation.
Delta Tuning was first applied to a series-built engine in
the first eight-cylinder RT-flex96C engine which was shop
tested in April 2004. The results were most satisfactory.
It was anticipated that the BSFC in the mid-load range
could be reduced by some 2 g/kWh though this reduction
varies between engine types. The BSFC measured during
the shop testing of the 8RT-flex96C engine was, however,
slightly better than expected.
Water-fuel emulsions
It has long been well known that introduction of water
into the combustion chamber reduces NOX formation.
Of the various possible techniques for doing this, water-
fuel emulsion is the longest studied. It was first tested on
Sulzer engines about 30 years ago and soon after that was
commercially applied on a Sulzer stationary low-speed
engine in a US power plant.
Running an engine on water-fuel emulsion makes it
theoretically possible to reduce NOX emissions by up to 50
per cent. The proportion of water added for the water-fuel
emulsion is primarily limited by the maximum delivery
capacity of the fuel injection pumps so that, in practice,
the engine has either to be derated or the maximum
achievable reduction of NOX limited to about 10 or 20
per cent. To obtain the maximum NOX reduction also at
full load, it would probably be necessary to redesign not
only the injection system but also the camshaft with all the
consequences in the camshaft drive, etc.
The proportion of water added is also limited by
the viscosity of the emulsion and the degree of heating
required to reduce the viscosity for injection. This is
a property of the water-fuel emulsion and cannot be
addressed by engine or system design.
Another aspect of water-fuel emulsion is that the
injection nozzle design (hole diameter, etc.) has to be
adapted to the increased quantity of liquid injected. With
this modified nozzle design, both fuel consumption and
component temperatures may be penalised when the
engine is running without water.

∆BSFC, g/kWh
4

0 171 g/kWh Fig. 9:

The new alternative BSFC
-2 curve for Sulzer RT-flex96C
RTA96C engines engines given by Delta Tuning
-4
RT-flex96C engines: compared with the original
-6 BSFC curves. All curves shown
Standard tuning
-8 Delta Tuning
are for engines complying with
the IMO NOX regulation.
-10 [04#102]
50% 75% 100%
Load

—6— © Wärtsilä Corporation, April 2005

 With DWI
Water injection
Fuel injection
Fig. 10: CFD simulation of the combustion process with direct water injection (DWI) indicates the levels of NOX. These pictures
show that the water is sprayed (left) directly into the area of highest NOX concentration. The NOX concentrations are calculated for
12 degrees crank angle after the start of fuel injection.
[05#021 & 05#022]
Whereas camshaft-controlled engines are restricted in
the timing of injection, electronically-controlled RT-flex
common-rail engines have a flexibility in engine setting
which is of great benefit when using water-fuel emulsions.
The electronic control system can readily be adapted to
given different, optimised injection control characteristics
for operation on water-fuel emulsions and fuel alone.
Furthermore, the engine settings can be optimised
according to engine load and speed. The electronic control
system would also be extended to provide load-dependent
mixing of water and fuel in the emulsifier.
RT-flex common-rail engines are similarly restricted
by the maximum capacity of the engines’ fuel systems
in the proportion of water that can be added in the
emulsion. However, they generally have a greater spare
capacity in their fuel pumps than do camshaft-controlled
RTA engines and it may be convenient to re-design new
engines to have a greater pumping capacity for water-fuel
emulsions.
NOX emissions, g/kWh
16
14
12
Fig. 11: 10
The NOX emissions measured in the Sulzer 8
RT-flex research engine when using Direct
Water Injection for different water-fuel 6
ratios in a common-rail system. The testbed 4
results show that 50% reduction in NOX is
possible with 70% water. 2
[05#023] 0
0 20
—7—
Without DWI NO mass fraction
Fuel injection
With the pumping capacity in RT-flex engines as
currently designed, it should be possible to reduce NOX
emissions to some 20 per cent below the MARPOL limit.
It is expected that the NOX reduction may be increased to
30 per cent if water-fuel emulsions are also combined with
the Low-NOX Injection technology of selective injection
patterns and Low-NOX Tuning techniques.
Direct water injection
Another technique for introducing water into the
combustion process for lowering NOX emissions is to
inject the water directly into the combustion chamber
separately from the fuel (Fig. 10). Under development
for Sulzer low-speed engines at Wärtsilä since 1993, this
direct water injection (DWI) technique directly reduces
cycle temperatures and thus NOX formation. Unlike other
techniques for introducing water into the combustion
process, direct injection enables the water to be injected at
the right time and place to obtain the greatest reduction in
50% NOX reduction
achieved
40 60 80 100
Water/fuel ratio, %
© Wärtsilä Corporation, April 2005

Water injection
Residual gas
Fig. 12:
Principle of the WaCoReG process in which NOX formation
is restricted partly by degrading the purity of the cylinder
charge by increasing the residual gas content, and partly
by direct water injection. The water is injected early in the
compression stroke to cool the cylinder charge.
[01#165]
NOX emissions.
For DWI, the water is handled by a fully independent,
second common-rail injection system under electronic
control. It offers the possibilities firstly of injecting very
large amounts of water without the requirement to derate
the engine and secondly to inject the water and fuel
injection with different timings. The water can be injected
in parallel with the fuel and/or during the compression
stroke, so that optimising injection timing, with respect to
fuel and water consumptions, NOX emissions and other
emissions, such as hydrocarbons and carbon monoxide, is
possible without affecting engine reliability.
When DWI is applied to an RT-flex common-rail
engine, the fuel injection system can be provided with
different engine settings so that it can run with the water
injection turned on or off, without affecting fuel injection
behaviour.
Direct water injection does not impose any restriction
on the quantity of water injected. It could even be more
than 100 per cent, that is a 1:1 ratio of water:fuel.
With about 70 per cent water/fuel ratio, DWI has been
shown to be capable of reducing NOX emissions down to
around 8 g/kWh, or to some 50 per cent below the IMO
limit (Fig. 11).
Whichever system of injecting water is employed,
consideration must be particularly given to the logistics of
providing sufficient fresh water on board ship. When 20
—8—
to 50 per cent water addition is anticipated, the quantities
of water will be substantial. Fresh water generators can
be heated using the engine cooling water, or using steam
from an exhaust gas economiser. In addition there must be
sufficient tank capacity for the water, with the necessary
handling system.
Combining water injection and exhaust gas
recirculation
Although DWI can be applied alone, it might be more
interesting to apply it in combination with internal
exhaust gas recirculation (EGR), as in WaCoReG (water-
cooled residual gas) by which we expect to obtain up to
70 per cent reduction in NOX emissions below the IMO
limit. This would bring NOX emissions down to about
5g/kWh.
Internal recirculation normally increases the thermal
load of the engine, so the water injection is applied to
reduce temperature levels, thereby keeping thermal loads
much the same as when running without internal EGR.
With WaCoReG, the water is injected earlier in the
compression stroke than with DWI (Fig. 12).
Exhaust gas recirculation reduces NOX formation at
source by reducing the oxygen available in the engine
cylinder and increasing the heat capacity of the cylinder
charge. In contrast to four-stroke engines in which it is
common practice to recirculate exhaust gases through
external manifolds, in low-speed two-stroke engines
we prefer to adapt the engine scavenging process to
decrease the purity of gas in the cylinder at the start of
compression. This is achieved by reducing the height
of scavenge ports to reduce the scavenge air quantity
flowing through the cylinder. One benefit is that smaller
turbochargers are required for the reduced gas flows. The
lower scavenge ports also have the benefit of allowing
greater expansion in the cylinder and thus improving fuel
consumption.
The reduced gas flows have the further benefit that
exhaust gas temperatures are raised which is helpful for
heat recovery systems (see below).
Selective catalytic reduction (SCR)
Then, for those cases in which the minimum NOX
emissions are required, we can apply selective catalytic
reduction (SCR). This can provide up to 90 per cent
reduction in NOX emission levels.
SCR technology is already well-established, involving
the metered injection of urea solution into the exhaust
gas flow before the catalyst unit. With two-stroke engines,
the SCR unit is arranged preferrably between the engine’s
exhaust manifold and the turbine inlet of the turbocharger
(Fig. 13). This location is chosen to ensure sufficiently
high exhaust gas temperatures for the catalyst process.
At present, SCR is only being applied in special
cases. For example, three Ro-Ro paper products carriers
delivered in 1999/2000 are powered by single Sulzer
7RTA52U engines equipped with SCR equipment. Their
NOX emissions are just 2 g/kWh or less.
If a Sulzer RT-flex common-rail engine is equipped
with SCR then the engine could be optimised for the
© Wärtsilä Corporation, April 2005
 Static mixers
∆p
Urea injection
Air
Urea
Flow
dresser
TI
Engine exhaust gas receiver
TI TI TI
Fig. 13:
Schematic of SCR (selective catalytic reduction) plant integrated with the turbocharging system in a Sulzer low-speed two-stroke
engine. The SCR unit is arranged between the engine’s exhaust gas manifold and the turbocharger.
[05#024]
lowest possible fuel consumption using the full flexibility
of the RT-flex concept, leaving the SCR to ensure
minimum NOX emissions.
Low-sulphur fuels and SOX emissions
SOX emissions are completely dependent upon the
sulphur content of the fuel burned, and the overall fuel
consumption. There are technologies for removing SOX
gases from engine exhaust gases but they involve bulky
and expensive equipment and further raise the question
of effluent disposal. Thus, the most practical solution for
reducing SOX emissions from shipping is expected to be in
simply reducing the sulphur content of the fuel oil.
The first step has been taken in Annex VI of the
MARPOL 73/78 convention, with a global limit of 4.5
per cent maximum sulphur for marine bunkers. However,
ships in SOX Emission Control Areas (SECA) must not
use fuel with a sulphur content exceeding 1.5 per cent (by
mass).
Modern engines can burn low-sulphur fuels (less than
1.5 per cent sulphur) without difficulties providing that
attention is given to the cylinder lubricating oil grade and
feed rate, and the jacket cooling water temperatures [7].
Sulzer RTA and RT-flex engines can run for limited
periods on low-sulphur fuels while continuing to feed
cylinder lubricating oil of the usual 70BN grade. However,
TI
12–30 bar starting air
for dust blowing
SCR
To NO analyser
TI
Engine
TI TI
SCR system
Shipyard piping
Turbocharger
TI TI TI TI
in such circumstances, the cylinder oil feed rate should be
kept to the minimum advised in Wärtsilä guidelines.
For longer periods of operation on low-sulphur fuels, it
is advised to use cylinder lubricating oil with a lower base
number (BN).
Irrespective of the fuel sulphur content it is essential
that jacket water temperatures are kept at recommended
values. Sulzer RTA and RT-flex engines built to the latest
standards have liner temperatures that are optimised to be
greater than the dew point throughout their height. This
makes the engines fairly insensitive to variations in sulphur
content.
The introduction of low-sulphur fuels will, of course,
call for suitable adaptations in ships’ fuel storage, handling
and treatment systems to accommodate separate grades
of heavy fuel oil. Segregation is advisory as bunkers of
different grades can be incompatible. Arrangements may
also be necessary for storing and handling a second grade
of cylinder lubricating oil.
Total Heat Recovery Plant
The overall level of exhaust emissions can be significantly
reduced by employing a waste heat recovery system.
Wärtsilä has developed a Total Heat Recovery Plant
for Sulzer low-speed engines which can recover some 11
per cent of the engine shaft power as electrical power for
—9— © Wärtsilä Corporation, April 2005
 Exhaust gas
economiser Ship service steam

Steam turbine
Ship service power
Fig. 14:
G
Schematic of the Total Heat
Power turbine
Recovery Plant. The recovered
G Aux. engine
energy is used to generate
electricity which can be employed
G Aux. engine
either for ship’s services, including
M Main engine
users such as reefer containers, or
G Aux. engine
for adding to the propulsion power
through a shaft motor.
G
[04#040]
Aux. engine

additional ship propulsion and shipboard services [8]. The
plant can thus give equivalent benefits in terms of better
fuel economy, reduced exhaust emissions or increased
propulsion power.
In this concept, exhaust gas energy across the load
range is increased by using a different turbocharger
matching when engine air is drawn from the ambient air
instead of from the engine-room. Usually marine engines
are designed for intake temperatures of up to 45°C for
tropical conditions with turbochargers drawing intake air
from the engine room. If instead the intake air is drawn
from outside the engine room thorough an air intake duct,
the maximum intake temperature can be assumed to be no
more than 35°C.
The lower air intake temperature allows the
turbochargers to be rematched to return the thermal load
of the engine back down to what prevails for the intake
temperature at 45°C. The thermal load of the adapted
engine will then be no greater than that of the usual
engine so as not to jeopardise engine reliability.
The rematched turbochargers allow more exhaust gas to
be branched off compared with the conventional tuning.
Therefore the rematched system gives both an increased
exhaust gas temperature for an exhaust gas economiser and

Fig. 15: Comparison of heat balances for Sulzer 12RT-flex96C engines without heat recovery (left) and with the Total Heat
Recovery Plant (right) showing an 11.4% gain in overall efficiency for the Total Heat Recovery Plant.
[04#048]

Shaft power 49.3% Overall efficiency Shaft power 49.0% Electrical power 5.9% Overall efficiency
49.3% 54.9%
Gain = 11.4%
Condenser
8.6%

Exhaust gas Exhaust gas

25.4% 12.6%

Scavenge air Scavenge air

cooling water cooling water
14.1% 12.9%

Jacket water Jacket water

6.3% 6.2%
Lubricating oil Lubricating oil
4.3% 4.2%

Radiation Radiation
0.6% 0.6%

Fuel input 100% Fuel input 100%

(171 g/kWh) (172 g/kWh)

— 10 — © Wärtsilä Corporation, April 2005

 NOX control technologies
0 20
Basis, IMO Tuning
Low-NOX Tuning
Low-NOX Injection
Water-fuel emulsion
RT-flex + emulsion
Direct Water Injection
WaCoReG
SCR
Total Heat
Recovery
–18 –2 0 +2
gain BSFC, g/kWh
Fig. 16: Summary of the changes in NOX emissions and specific fuel consumption (BSFC) for various emissions control technologies.
[05#025]
an increased branched-off exhaust flow for an exhaust gas
power turbine at ISO reference conditions.
Exhaust energy can thus be recovered and applied
in both a steam turbine and exhaust-gas power turbine
to generate electrical power, equivalent to about 11%
of engine power (Figs. 14 and 15). The electrical power
can be employed either in a shaft motor/generator or in
supplying shipboard services. The generated power can
thus contribute significant savings in both fuel costs and
exhaust emissions (CO2, NOX, SOX, etc.). The payback
time would depend upon the installation design and its
operating costs, but would be expected to be less than five
years.
The Total Heat Recovery Plant also offers an attractive
possibility for powering the larger container ships which
need more propulsion power than is available from the 14-
cylinder Sulzer RT-flex96C engine. The standard engine
gives an MCR power of 80,080 kW, thereby having a
continuous service output (85% load) of 68,068 kW. At
this engine rating, the Total Heat Recovery Plant would
contribute a shaft power of 7390 kW. The combined
service power for propulsion would therefore be 75,458
kW, equivalent to an engine MCR power of 88,770 kW
for a plant without heat recovery.
Thus a single Sulzer 14RT-flex96C engine with a
Total Heat Recovery Plant would be sufficient for the
propulsion of single-screw container ships up to 12,000
TEU capacity. It thereby allows these larger ships to be
powered by engines of well-established, proven design.
— 11 —
NOX emissions, % NOX
40 60 80 100
—
– 5%
– 20%
– 20%
– 30%
– 50%
– 70%
– 90%
– 11%
4 6 8
penalty
Conclusion, with summary of emissions
control measures
Today Wärtsilä already has a range of technology solutions
to cater for future emissions control needs of Sulzer
low-speed engines. Continuing development will ensure
that these technologies shall not adversely affect engine
reliability and component lifetimes. At the same time, care
will be taken to ensure the most desirable balance between
further improved fuel economy and emissions reduction.
Technologies such as Direct Water Injection or
WaCoReG will be available to achieve major reductions
in NOX emissions while minimum NOX emissions can
still be achieved by employing SCR. All these solutions
are applicable to conventional engines with mechanically-
operated fuel injection. However, when they are applied
to Sulzer RT-flex common rail engines, the engines
could then also be optimised to give the minimum
fuel consumption for the least possible CO2 emissions.
Indeed, Sulzer RT-flex technology is an important key. In
its basic form, as now entering service, smoke emissions
are already reduced so that they are no longer visible
at any ship speed. With further development it will
become the platform for propulsion plants to power the
environmentally-friendly ships of the future.
The selection of which technology to employ depends
very much upon the degree of NOX emissions reduction
required. The various technologies available and being
developed for Sulzer RTA and RT-flex engines are
© Wärtsilä Corporation, April 2005
 summarised in figure 16, showing both the degree of References
NOX emissions reduction achieved and the associated
1. Dr M.E. Davies et al, ‘Study on the economic, legal,
fuel consumption penalties. The current standard engines
environmental and practical implications of a European
built in compliance with the IMO regulation for NOX
Union system to reduce ship emissions of SO2 and
emissions are taken as the basis.
NOX’, BMT Murray Fenton Edon Liddiard Vince Ltd,
The emphasis in shipping on the NOX emissions leads
UK, Report No.3623, August 2000.
to a strategic point. The trade-off between NOX level and
2. ‘BP Statistical Review of World Energy’, June 2004.
BSFC means that with engine internal measures there is
3. Diane Bailey, ‘Harboring pollution: Strategies to clean
an increase in other air emissions (CO2, SOX, etc.) when
up US ports’, National Resources Defense Council,
NOX emissions are reduced. There is thus clear potential
USA, August 2004.
for conflicts in emissions targets and dilemmas in setting
4. ‘GHG emissions from international shipping and
the priorities for the technical development of engines.
aviation’, ECON Centre for Economic Analysis, Oslo,
For there to be an overall reduction in all air emissions
for Swedish Environmental Protection Agency, January
to tackle both ‘greenhouse’ gases and air pollution,
2003.
consideration will need to be given to external control
5. Rudolf Holtbecker, ‘Sulzer RTA engines comply with
measures such as SCR and waste heat recovery. Yet more
IMO emissions regulations’, Marine News 1–1999,
effect on emissions will most probably be gained by
pp12–14, Wärtsilä Corporation.
focusing instead upon ship operation and optimisation of
6. German Weisser, ‘Wider choice in fuel consumption
the whole transportation chain.
for Sulzer engines’, Marine News 3–2004, pp24–26,
Wärtsilä Corporation.
7. Michael Welsh, ‘Considerations for using low-sulphur
fuels’, Wärtsilä Corporation, 2002.
8. Heinrich Schmid, ‘Less emissions through waste heat
recovery’, Green Ship Technology Conference, London,
2004.

Published April 2005 by:

Wärtsilä Switzerland Ltd
PO Box 414
CH-8401 Winterthur
Tel: +41 52 262 49 22
Fax: +41 52 262 07 18
www.wartsila.com

— 12 — © Wärtsilä Corporation, April 2005

View publication stats