You are on page 1of 14

ATMOSPHERIC ENVIRONMENT: X 12 (2021) 100142

Contents lists available at ScienceDirect

Atmospheric Environment: X
journal homepage: www.journals.elsevier.com/atmospheric-environment-x

Development of exhaust emission factors for vessels: A review and


meta-analysis of available data
Achilleas Grigoriadis a, Sokratis Mamarikas a, Ioannis Ioannidis a, Elisa Majamäki b,
Jukka-Pekka Jalkanen b, Leonidas Ntziachristos a, *
a
Department of Mechanical Engineering, Aristotle University, P.O. Box 458, 54124, Thessaloniki, Greece
b
Finish Meteorological Institute, P.O. Box 503, 00101, Helsinki, Finland

A R T I C L E I N F O A B S T R A C T

Keywords: The maritime sector is characterized by high contribution in the emission of harmful pollutants such as NOX, PM,
Maritime sector and SOX. The sector is making decisive steps to drastically reduce its environmental footprint by applying new
Shipping emissions technology and fuels as a reaction to the increasingly stringent emissions regulatory framework. In order to
Emission inventories
calculate the contribution of shipping to current emission inventories and to project future developments under
Emission factors
different policy scenarios reliable and diversified emission factors have to be developed. In this context the
Emission models
Ship on-board measurements present paper develops a new set of emission factors for ships, based on a review and statistical processing of
Ship engine test-bed measurements shipping emission measurements available in literature. To offer significant resolution and versatility for use
under different operating conditions, the developed emission factors are expressed as a function of the ship
engine load. These emission factors are also distinguished according to the engine and fuel types that are met in
practice. The absolute values of the developed emission factors highlight the emission performance of ships per
pollutant, while the load dependent functions produced reveal how this emission performance of ships changes
under various operating conditions. Finally, a comparison of the developed load dependent emission factors with
modal emission factors retrieved from literature is performed specifically for NOX emissions. This comparison
indicates the differences and similarities between the two methods as a measure of uncertainty in the estimation
of emissions using the proposed emission factor dataset.

1. Introduction vicinity to coastal areas to high PM concentrations. Notably (Kuittinen


et al., 2020), calculated that Particle Number (PN) emissions from ves­
The maritime sector remains a challenge in terms of pollutants sels are at the same level with all continental anthropogenic activities
emission reductions, compared to other transportation sectors where together, quantified to 1,2 × 1028 particles in 2016.
substantial reductions have been achieved (EEA, 2020). Shipping was Efforts to regulate the sector have primarily focused on the reduction
responsible for 2,89% of total manmade Greenhouse Gases (GHG) in of sulphur oxide emissions, with the IMO’s most recent regulation of
2018 and this share is expected to increase towards 2050 (IMO, 2020). 0,5% fuel sulphur content limit being globally effective since January
Shipping is also characterized by high emissions of harmful pollutants, 2020 (IMO, 2016). NOX are regulated in the context of the certification
such as Sulphur Oxides (SOX), Nitrogen Oxides (NOX) and Particulate of marine diesel engines and are subject to restrictions with geospatial
Matter (PM), causing a significant deterioration of air quality, especially characteristics within low emission zones, which are currently appli­
in coastal areas (Corbett et al., 2007), (Matthias et al., 2010), (Viana cable to only few parts of the world where relevant Emission Control
et al., 2014). Lehtoranta et al. (2019b) estimated that vessels produce Area (ECAs) have been established (IMO, 1997). Existing NECAs are
15% of NOX and 5–8% of SOX emissions of all anthropogenic activities. located in the United States and in the Baltic and North Sea. CO2 is
Viana et al. (2014) found that maritime PM contribution in European currently being systematically monitored and reported by ship operators
coastal areas can reach 7% of total ambient PM10, and 14% of total PM2, at EU and international levels (EC, 2013), (IMO, 2011) and energy
5. Moldanova et al. (2009) reported that global PM from maritime efficient design obligations for specific ship categories have started to be
amounts 0,9-1,7 Mton annually, exposing population living in close applied, in the context of a detailed carbon emission reduction strategy

* Corresponding author.
E-mail address: leon@auth.gr (L. Ntziachristos).

https://doi.org/10.1016/j.aeaoa.2021.100142
Received 21 July 2021; Received in revised form 10 November 2021; Accepted 13 November 2021
Available online 16 November 2021
2590-1621/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

(IMO, 2018a,b). 2016). Two main EF databases and models exist and widely used
As a reaction to the increasingly stringent emissions regulatory currently in Europe: the EMEP/EEA inventory Guidebook (Trozzi and
framework, the maritime sector has been strongly investing in new De Lauretis, 2016) and the STEAM model (Jalkanen et al., 2012).
technology and fuels. Technological measures include vessel design and The EFs that are included in the EMEP/EEA inventory Guidebook are
operational changes (Balcombe, 2019), engine optimization techniques expressed in kg/ton of fuel, or in g/kWh (Trozzi and Vaccro, 2006). The
including exhaust gas recirculation (EGR) (Raptotasios et al., 2015), and Guidebook provides EFs for various pollutants, as well as energy con­
the installation and operation of exhaust aftertreatment devices such as sumption factors, which vary by the fuel type and engine type. When
scrubbers and selective catalytic reduction (SCR) systems (Nova and detailed activity data are available by the user of the Guidebook, then
Tronconi, 2014). Fuel changes involve the use of low sulphur fuel (Sarvi the estimation of emissions can be further specified by the operating
et al., 2008), (Sarvi et al., 2011) and, most recently, the use of alter­ mode of ship, which is distinguished to cruising and
native fuels (i.e. LNG, methanol, NH3) followed by all engine modifi­ maneuvering/hoteling.
cations these may require (Anderson et al., 2015), (Fridell et al., 2020), The STEAM model (Jalkanen et al., 2009) expresses EFs in g/kWh, as
(Hansson et al., 2020). a function of the engine load. The NOX load-dependent EFs also vary
This range of available options requires reliable and diversified with the nominal speed of the engine, as well as the engine build year.
Emission Factors (EFs) to calculate the contribution of shipping to cur­ Apart from NOX the following pollutants are also covered: CO, SOX,
rent emission inventories and to project future developments under VOC, PM, BC and CO2 and energy consumption factors.
different policy scenarios (Johnson et al., 2020). Existing EF datasets for In the United States (US), the USEPA’s inventory guidebook also
ships, such as the IMO ship emissions toolkit (IMO, 2018a,b), the expresses EFs in g/kWh (US EPA, 2009). The engine load is set to 83%
EMEP/EEA Air Pollutant Emission Inventory Guidebook (AEIG) (Trozzi for cruising mode. For the other operating phases, the load is estimated
and De Lauretis, 2016), the USEPA’s inventory guidebook (US EPA, by the propeller’s law principle, as the cube ratio of actual to maximum
2009), and the model STEAM (Jalkanen et al., 2012), and other com­ speed. EFs are provided for NOХ, PM, HC, CO, SOХ, CO2 emissions, along
mercial datasets require constant maintenance and updates to be in-line with energy consumption factors. EFs vary on the engine type (SSD,
with latest developments and allow users to make correct assessments MSD, HSD, steam turbines), fuel type (RO, MDO, MGO) and sulphur
related to vessel emissions. content for MGO (0,5% or 0,1%).
A number of research and development activities can be used to
produce new emission information from vessels, using a multitude of 2.2. Ship emissions’ performance
techniques and available tools. Measurements are conducted on marine
engines in the laboratory (Ntziachristos et al., 2016) or in field cam­ NOX emissions are formed in the combustion chamber during the
paigns on operational vessels (Winnes and Fridell, 2010). Techniques reaction of nitrogen content of air and any nitrogen included in fuel,
utilized range from in-funnel measurements using suitable devices with the oxygen from air, at high temperature (Agrawal et al., 2008b).
(Winnes and Fridell, 2010), to airborne (Murphy et al., 2009) or NOX emissions are generally considered to increase with engine load due
coast-side (Pirjola et al., 2014) measurement with single vessels’ emis­ to higher combustion temperatures (Winnes and Fridell, 2009), (Sanaz
sions detection by satellite monitoring (Georgoulias et al., 2020), being et al., 2018). However, some evidence (Woodyard, 2004), (Zetterdahl
one of the latest state-of-art techniques. Our literature review came up et al., 2016) indicates that NOX emissions may also decrease with engine
with more than 53 published research papers focusing on vessels emis­ load due to the decreasing oxygen content. Exhaust NOX emissions of
sions characterization since 2015. marine engines are dominated by NO, the primary reaction product,
In this context, the target of the present paper is to collect, analyze, rather than NO2 which forms at lower temperature (Kristensen, 2012).
process and develop EFs for various pollutants, ship engine and fuel Large marine engines in terms of rated power, accompanied with a low
types, based on existing datasets and latest findings available in the nominal rotational speed, produce higher specific NOx emissions than
public domain. To offer significant resolution and versatility for use lower rated power and higher speed engines (Sinha et al., 2003). This is
under different operation conditions, the developed EFs are expressed as due to the extended period that combustion gases remain in the cylinder
a function of the ship engine load. The outputs intend to provide evi­ in slow speed engines which provides time for additional NOX formation
dence on the existing environmental performance of ships, enhancing (Celo, 2015), (Lamas and Rodringuez, 2012). The effect of changing
the overall effort to develop accurate and up-to date EFs for the sector from heavy to lighter fuel oils on NOX emissions does not seem to be
and supporting existing models and emission inventory databases. consistent (Petzold et al., 2011). Specific fuel-related parameters, such
as cetane number and fuel aromaticity seem to be more important in
2. Status of emission factors and emission measurements of terms of NOX formation than the fuel density (Nabi and Hustrad, 2012).
ships CO emissions are a product of the incomplete combustion of carbon
in the fuel. CO emissions are mainly caused by fuel injected close to the
2.1. Overview cylinder walls where combustion takes place at lower temperature,
leading to lower reaction rate, so formed CO is not being further con­
In the maritime sector, emission factors are most commonly verted to CO2 (Woodyard, 2004). CO formation inversely depends on the
expressed in relation to energy output, i.e. mass of pollutant per unit of residence time of fuel to the oxidative combustion environment and, in
energy produced by the engine (g/kWh) or per unit of fuel consumed principle, engines with higher nominal engine speed should exhibit
(engine input, i.e. kg/ton). In a limited number of studies, distance- higher emissions. EFs for CO, in absolute values, are higher in high speed
based EFs in g/nm (nm: nautical mile) are also presented. EFs in the four-stroke engines and lower in slow speed two-stroke engines for an
literature are distinguished according to engine type (slow speed diesel - additional reason, i.e. the higher air to fuel ratio of the latter (Zhang
SSD, medium speed diesel - MSD, high speed diesel - HSD, steam and gas et al., 2016), (Celo, 2015). CO is in generally inversely linked to engine
turbine) and emission standard, as well as utility type (main or auxiliary load, as higher loads in diesel combustion continue to have a surplus of
engines) and fuel used (MDO - Marine Diesel Oil, MGO - Marine Gas Oil, oxygen and deliver a higher combustion temperature that both help to
HFO - Heavy Fuel Oil, Gasoline) (ENTEC, 2002), (CARB, 2008), (US oxidize CO (Lehtoranta et al., 2017), (Agraval et al., 2010), (Zetterdhal
EPA, 2009), (Trozzi and De Lauretis, 2016). By incorporating various et al., 2016). On the other hand, CO EFs seem not to be significantly
levels of activity, EFs can then be expressed as a function of the engine influenced by the fuel type (Petzold et al., 2011).
operating points or, at a more aggregated level, they become specific to Emissions of unburnt Hydrocarbons (HCs) are generally low from
activity modes, such as berthing/hoteling, maneuvering and cruising diesel-combustion engines. These are mainly influenced by the com­
that are met under the normal operation of ships (Trozzi and De Lauretis, bustion efficiency which is further affected by engine load. Inefficient

2
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

combustion conditions promote fuel and lubricating oil to escape un­ instrumentation, whereas particle collection in filters is also
burned leading to higher fraction of HC in the exhaust. HCs generally accomplished.
follow the trends and justification of CO in terms of load and engine Emission factors can be composed in relation to specific vessel ac­
speed dependance (Kristenssen, 2012), (Kalender and Ergin, 2017). tivity. Activity can be differentiated according to engine operating
PM emitted from ships consists of several components such as points or ship operating mode (i.e. maneuvering/hoteling, cruising). In
organic matter (OM), elemental carbon (EC), sulphates, metals and this process, emission factors become specific to conditions that can be
inorganic ions (Huang et al., 2018). For the purposes of the current quantified during the emission inventorying process. For example, one
study, OM is estimated as Organic Carbon (OC) increased by a factor of can discern the different phases of operation of a vessel in open-sea or in-
20% (OM = OC x 1,2) (Petzold et al., 2008). Particle emissions can also port operation and assign a proper emission factor to each of the indi­
be characterized by the number of particles and their size distribution. vidual phases. This overall process is schematically represented in Fig. 1.
Particles are further divided into solid (non-volatile) and (semi-)volatile In order to calculate emissions of single vessels, the actual activity
ones (Winnes et al., 2016). Both solid and volatile are dominated by profile (route, speed, location) of the ship is required and is distin­
ultrafine particles (under 100 nm) (Chu-Van et al., 2018). PM formation guished in separate trip phases (cruising, maneuvering, hoteling)
is highly related to diesel engine operation and fuel system character­ (Chu-Van et al., 2018). Cruising corresponds to the phase of voyage
istics, as the fuel to air mixture is not premixed in these compression between two destinations at open sea, maintaining an approximately
ignition engines, where the overall mixing is performed within the constant high speed which corresponds to steady engine load. Maneu­
cylinder. Therefore, during combustion, local insufficient mixing of air vering is related to the phase where a ship arrives to the port before
with fuel leads to the non-effective oxidation of fuel and therefore docking or departs from docking before this exits the port. This phase is
leading to the formation of soot particles. The elemental part (soot, black characterized by lower loads and some transient operation. During
carbon) is related to the insufficient oxidation of fuel. The organic part hoteling/at berth phase ships are moored at port (zero speed) and
mostly comprises of heavy (low volatility) fuel components as well as perform cargo or passenger operations. At this phase mainly auxiliary
lube oil species that escape combustion. Some conversion of lighter fuel engines are working to cover ship energy needs (Jalkanen et al., 2009).
components to heavier species, due to partial oxidation, may also take Cruise ships are a well-known exception to this rule where much higher
place during combustion. OM constitutes a major part of PM composi­ energy consumption is required to keep all amenities on the vessel
tion depending on the fuel type (Moldanova et al., 2009). OM emissions operational for their clients while anchored.
are influenced by the fuel type and the transition from residual to Exhaust emission measurements can also be performed on engine
distillate can reduce the mass, but not the number of PM (Fu et al., test-beds in the laboratory, thus linking emission rates with specific and
2013), (Zetterdahl et al., 2016), (Winnes and Fridell, 2009). The relation well-determined engine load points. In such tests, the engine is placed on
of PM emissions with load appears weak but an increase of load may the dynamometer and is operated at various constant loads at will. En­
show a slight reduction of PM (Winnes and Fridell, 2010), while the EFs gine loads are often selected from the ISO 8178 loading points. The
of PM present a higher level at low loads because of the fuel injection sampling system is similar to the one used in real-world on-board
pressure drop at these loads (Chu-Van et al., 2018). emission measurements (Ntziachristos et al., 2016b). Engine dyna­
CO2 emissions, as the main product in the reaction of HC oxidation, mometer measurements are convenient for comparing emissions with
are directly related to the carbon content of the fuel and overall fuel various types of fuels (distillate, LNG, biofuels, etc.) (Ntziachristos et al.,
consumption. Therefore, as far as the carbon content of various types of 2016b), (Lehtoranta et al., 2019b) and aftertreatment options (SCR)
fuels changes, then difference in CO2 emissions can be observed with the (Lehtoranta et al., 2015) over identical operation conditions, that may
fuel use. In line with the specific fuel oil consumption (SFOC) of marine be more difficult to achieve on-board.
diesel engines, CO2 emissions are high at low loads, are decreased in In the past few years, in-plume measurements have been conducted
medium loads presenting a minimum at approximately 75% of the via remote sniffer technologies, which are installed either in fixed po­
maximum load and are again increased as load tends to 100% (Wood­ sitions where increased ship activity exists (ports/harbor/bridges)
yard, 2004). With the introduction of non-conventional alternative (Ekmektizoglou et al., 2020) or on mobile means that chase ship plumes
fuels, such as LNG and methanol, CO2 emissions are reduced because of (vans, boats, aerial means) (Pirjola et al., 2014), (Murphy et al., 2009).
the lower carbon content of such new fuels. In addition, NH3 and H2 are New techniques are continuously being applied, such as drones carrying
considered as future solutions for the total decarbonization of the sector small sniffer systems that chase plumes and measure emissions. These
(DNV, 2020), due to the complete lack of carbon in the molecular sniffer-based methods measure in-plume content for various pollutants
structure of the fuel. and refer it to fuel consumption by parallel measurement of in-plume
SOX emissions primarily depend on the fuel type burned, in terms of CO2 content measurement. The conversion from content measurement
its content in sulphur, (Fuel Sulphur Content - FSC) which is oxidized into emission rate is accomplished by multiplying the content ratio of
during the combustion reaction (Miller et al., 2012). Measurements on the pollutant over the CO2, with the ratio of the molecular masses of
the same ship before and after the transition to low sulphur content fuel both gases and by estimating the type of fuel used (Pirjola et al., 2014).
showed that SO2 emissions have been reduced by more than 82% Specific attention should be given once developing emission rates
(Zetterdahl et al., 2016). SOX emissions in relation with the engine load through this method as the content of pollutants in the background has
present a similar trend with specific fuel consumption, increased at to be carefully removed.
lower loads and reduced as the engine load increases. SO2 accounts more
than 95% of total SOX emissions with the remaining part being 2.4. Current study objective
composed by SO3 and sulphates (SO4 and water) (Miller et al., 2012).
Due to the recent developments in availability of emissions mea­
2.3. Emission measurement methods surements from different vessels and the introduction of stringent
environmental regulations for ships at a global level, the target of the
Emission rates, in the context of real-world measurement campaigns, current study is twofold. First, to develop new EFs based on literature
can be extracted by measuring exhaust emissions in relation to fuel findings through statistical processing and, second, to quantify the effect
consumption or energy output of the vessel engines. The measurement of emission control regulations in emission rates. In particular, the
process mainly includes the involvement of high-end pollutant analyzers following questions become relevant in such an analysis:
or more simplified sensors, which are placed in the ship exhaust line
(before or after any emission control system). For PM and PN, a dilution
system is also required prior to the PM/PN measurement

3
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

Fig. 1. Schematic representation of Emission Factor development process through on-board measurements.

• How do parameters such as the fuel type, nominal engine power and average emission performance and the dependence of emission rate with
nominal rotational speed affect power-based EFs of gaseous pollut­ engine load. As a first step, a single engine load point had to be selected
ants and PM/PN? in order to use this as a reference. This reference load point per pollutant
• What is the effect of emission standard regulations in EF trends over was selected to be the one most frequently observed in the available
time? literature. Depending on engine type and pollutant, the most frequent
• How power-based EFs for gaseous pollutants vary in relation with the load point encountered was found to be at 50% of engine load for the
engine load? vast majority of cases or 75% for only a handful of pollutants and engine
types. Average values for each pollutant (named Base EFs - BEFs) have
3. Materials and methods been calculated for the three categories of marine engines (slow, me­
dium, high speed) and for different fuel types (residual and distillate) for
3.1. Data collection and organization this reference load point.
The emission rates found in other load points than the reference one,
The suitability of engine load as a variable on which to base emission have been normalized over the emissions rate at the reference load
factors is justified by the fact that the number of engine load points of point. This normalization process has been separately applied for each
ships during real world operation is limited, as ships spend most of their single literature source. This produces a family of points which are
operating time at specific steady state loads. It is indicative that, on centered around 1 for every single literature source. By using this
major ship classes, 94–99% (depending on vessel type) of the total method over all literature sources and combining, one obtains the
operating time is on cruising or hoteling/anchoring, in which conditions relative effect of load on emissions regardless of the absolute value of
the engine operates in steady state. In contrast, maneuvering is domi­ emission reported in each study. A single function of load correction is
nated by engine load changes (transient operation), but represents only then produced by means of regression over the combined set of values.
1–6% of the voyage time (Comer et al., 2017). Thus, the engine load is The overall process for estimating the absolute emission levels and
not a complicate explanatory variable for shipping emissions and can be the emission dependance with the engine load is schematically
easily estimated according to phase of operation. explained in Fig. 2. All literature sources that have been used are listed
The methodology for developing EFs is primarily based on a Litera­ in Annex I. The same Annex indicates which exact sources have been
ture Review (LR), which resulted in the collection of 157 papers, studies used for the creation of tables and figures results of the current study.
and reports focusing on vessel emissions, from which finally 59 were
utilized, with over 150 individual emission measurements’ results 4. Results
deploying various methods. These mainly comprised on-board mea­
surements (37 measurements) and marine engine placed on test-beds 4.1. Baseline emission factors
(19 measurements). Also 2 identified in-plume measurements and 1
on-board and in-plume measurement were considered. For each test Averaging of emission rates from various sources over the reference
configuration, various emission rates, including gaseous pollutants engine load point led to the estimation of the baseline emission factor for
(NOX, CO, SO2, THC, etc.), greenhouse gases (CO2) and PM/PN, as well various gaseous pollutants (NOX, CO, HC) and the baseline specific fuel
as energy consumption rates have been collected. All these data have oil consumption factor (SFOC). Table 1 provides the absolute level of
been organized in a unique database and divided in category groups emissions and consumption depending on engine type, together with the
(fuel type, engine type, ship category etc.). number of measurements and the standard deviation of each value. It is
reminded that these reference values correspond to 50% engine load.
The same table also provides the grand average value of all measure­
3.2. Development of load-dependent emission factors ments collected over the reference load point under the “any type”
column, which is suggested to be used in case the engine type is not
The individual emission values collected were used to deliver an

4
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

Fig. 2. Schematic representation of the Emission Factor development process.

Table 1 Table 2
Base Emission or Consumption Factor together with the standard deviation and Base NOX emission factors, standard deviation (and number of individual values
the number of individual values in parenthesis per engine type at 50% engine in parenthesis), per engine type and TIER standards for 50% engine load.
load. NOX BEF (g/ TIER 0 TIER I TIER II TIER III
Pollutant/ Base Emission (BEF) or Consumption (BCF) factor kWh)
Consumption
Any type SSD MSD HSD SSD 17,7 ± 2,14 14,4 ± 2,75 11,3 ± 2,53 2,00 (1)
(20) (11) (14)
NOX (g/kWh) 13,2 ± 3,98 14,4 ± 3,60 12,4 ± 4,08 11,7 ± 3,68 MSD 10,8 ± 1,60 10,5 ± 2,01 8,28 ± 0,148 1,02 ± 0,815
(35) (15) (15) (5) (6) (27) (4) (9)
CO (g/kWh) 0,898 ± 0,714 ± 0,974 ± 1,10 ± HSD 8,53 ± 2,71 7,41 ± 1,72 5,95 ± 1,15 1,25 ± 0,150
0,571 (38) 0,631 (14) 0,530 (18) 0,361 (6) (4) (6) (3) (2)
HC (g/kWh) 0,440 ± 0,358 ± 0,405 ± 0,662 ±
0,350 (20) 0,234 (7) 0,204 (9) 0,599 (4)
SFOC (MJ/kWh) 8,83 ± 8,48 ± 8,42 ± 9,74 ±
cannot be estimated.
0,841 (10) 0,266 (3) 0,493 (4) 0,891 (3)
CO2 and SO2 can be calculated from SFOC values if the fuel energy
content and speciation in carbon and sulphur are known. If such values
known. are not known, the emission rates in Table 3 can be used instead. These
The calculated mean values for CO and HC show that emissions are have been derived on the basis of Table 1. SFOC values in combination
higher for high-speed engines. This is most probably the result of the fact with averaged fuel characteristics found in the literature. Specifically,
that slow speed engines operate at higher air-to-fuel ratios and provide the SO2 calculation also takes into account the fact that part of the SO2 is
more time for oxidation of species after combustion than faster engines converted to SO4 and corrects for this (more on this correction follows
do. In terms of fuel consumption, it seems that SSD and MSD engines with the discussion around Fig. 5).
little differ at 50% load, while fuel efficiency drops significantly for HSD CO2 is directly proportional to the SFOC and therefore high-speed
engines. engines produce higher emissions, due to lower efficiency, compared
The impact of engine type on NOX is rather opposite than CO, HC and
SFOC. In this case, SSD exhibit higher NOX emission rates, again as a
result of the higher oxygen and time availability. In particular NOX from Table 3
vessels are controlled over different so-called Tier levels, depending on CO2 and SO2 Base Emission Factor, per engine and fuel type for 50% engine load.
their date of manufacture. Hence, Table 1 values are only to be used Pollutant Fuel type Base Emission Factor (BEF)
when the vessel age or Tier level are not known. Base NOX emission
All types SSD MSD HSD
factors split according to Tier and engine type are presented in Table 2.
In all engine types, baseline emission factors drop as the Tier level ad­ CO2a (g/kWh) Residual 678 651 646 747
CO2a (g/kWh) Distillate 645 619 615 711
vances, a finding that shows the positive impact of regulatory control on
SO2b (g/kWh) Residual 4,43 4,31 4,29 4,70
actual emission levels. In particular Tier III appears to be some 90% SO2b (g/kWh) Distillate 0,372 0,358 0,355 0,410
lower than pre-Tier levels, at least for SSD and MSD engines. In partic­ a
CO2 has been calculated by the SFOC, considering a fuel carbon content of
ular, all NOX Tier III values correspond to SCR-equipped vessels.
86,8% and LHV of 41,5 MJ/kg for residual fuel and 86,5% and 43,4 MJ/kg for
Although, in principle, Tier III can be achieved with EGR alone all the
distillate.
measurements found in literature were obtained on SCR-equipped en­ b
SO2 has been calculated by the SFOC, considering FSC of 1,42% and LHV of
gines so the actual performance of EGR on real world emission rates 41,5 MJ/kg for residual fuel and 0,0931% and 43,4 MJ/kg for distillate.

5
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

Table 4 4.2. Load dependent functions


PM Base Emission Factors, per engine and fuel type for 50% engine load.
Pollutant Fuel type Base Emission Factor (BEF) The dimensionless dependence of emissions rates on engine load for
various pollutants (NOX, CO, HC) and SFOC is depicted in Fig. 3. As
All types SSD MSD HSD
earlier explained, the most frequent engine load point for which emis­
PMa (g/kWh) Residual 0,960 0,934 0,930 1,03 sion rates were available in the literature was the 50% load, therefore all
PMa (g/kWh) Distillate 0,217 0,215 0,215 0,221
Dry PMb (g/kWh) Residual 0,637 0,624 0,622 0,669
other emission rates have been normalized over the emission rates at
Dry PMb (g/kWh) Distillate 0,197 0,196 0,196 0,199 this engine load. Fig. 3 presents all individual normalized points
a together with the best-fit regression line. The exact equation for each
PM has been calculated by the sum of organic mass, elemental carbon, ash,
regression line is provided as Annex II of the paper. Emission rates from
hydrated sulphates (SO4 +6,5H2O).
b
Dry PM has been calculated by the sum of organic mass, elemental carbon,
all engine types (SSD, MSD, HSD) have been included in the same pool,
ash and dry sulphates (SO4). as our analysis could not identify specific engine type effects that would
give substantially differentiated results.
It is observed that the general emission trends described in section
to medium and slow speed. The higher heating value of distillate fuels in
2.2 of the current study are reflected to the shape of the regression lines
combination with the slightly lower carbon content than residual fuels,
derived. SFOC, which depends on the efficiency of marine engines,
leads to lower CO2 emissions. For estimating SO2 emissions, the distil­
shows to drop as load increases towards a minimum at 75% and then this
late fuel that on average contains 0,09% sulphur produces significantly
increases again. This is typical behavior for diesel engines and in
lower SO2 emissions than residual, where average FSC is estimated at
particular for marine ones that are optimized for optimum fuel efficiency
1,42%. It is clarified that these values are relevant for residual fuels
at cruise conditions that occur around 75% load (Yang et al., 2019),
when no exhaust gas cleaning systems (scrubbers) are used.
(Comer et al., 2017).
For PM, the engine type was not found to have a significant impact
CO emissions strongly increase as load drops below 30%. This
on specific emission rate. On the contrary, PM was largely affected by
behavior can be explained by the lower combustion temperature at low
the use of residual or distillate fuels that were shown to have an impact
loads – the CO to CO2 conversion is known to be very strongly linked to
on all individual PM components (elemental carbon, organic mass, ash
temperature (Sippula et al., 2014). The behavior of HC is similar to that
and sulphates). In particular for the contribution of sulphates, total PM
of CO but of lesser value as the HC oxidation is not as strongly dependent
has been calculated on the basis of hydrated sulphates (SO4 + 6,5H2O)
on temperature as that of CO. One may also observe some slight increase
or as dry PM, which includes sulphates but no addition of water. Total
in HC for 100% load. Although there is large scatter of data at this load
PM values would be relevant if one would wish to compare the calcu­
point, slight increase of HC may be expected due to the long duration of
lated values in the current study with measured values collected after
injection at 100% load and the increased probability of hydrocarbon
typical dilution ratios (e.g. 100:1). However, in plume simulation
interaction with the piston and cylinder walls (Yang et al., 2019).
studies, the dry PM would be the starting point and then the model
For NOX the trend shows that emissions decrease with engine load, as
would account for the partition of water between the particulate and the
this approaches 50%. For higher loads than 50%, a constant level against
gaseous phases as particles age in the atmosphere. Calculated BEFs for
engine load is seen within uncertainty. At lower loads, NOX specific
PM (total and dry) are presented in Table 4.
emissions may increase because of leaner air fuel mixtures (Kalender
Table 5 provides further speciation of PM to individual components
and Ergin, 2017) and the lower fuel efficiency at this engine operating
and some average physical characteristics of particle distributions. As
area (Gysel et al., 2017) which is the denominator. Nevertheless, at
PM is constituted by EC and OM, a further analysis of its speciation was
higher loads (above 50%) the statistical processing verifies the con­
performed on the basis of the statistical analysis of emission rates of
trasting findings regarding the NOX increase or decrease with the engine
these two groups. For the purposes of this analysis, EC and BC were
load, already recognized in the literature review section. There is a
considered equivalent and were used interchangeably. Although it is
respective number of emission rates either above or below the formed
well recognized that EC and BC are measured with different techniques,
steady line above 50%. This behavior indicates that NOX emissions may
the range of values encountered in the literature were largely over­
be influenced by parameters that counteract each other, as previously
lapping and this allowed pooling them together. Particle number (PN)
stated in the review section. The NOX increase effect due to the tem­
distinguished to total and non-volatile (solid) particles is included in
perature rise at high loads may in some cases overtake the influence of
Table 5, without distinction to fuel type due to lack of data. Base EFs for
NOX decrease, because of the enrichment of the fuel mix at high loads,
OC and EC are indicated at 50% engine load, while SO4 conversion rate,
while in other cases the reduction of the air to fuel ratio seems to
total PN and solid PN are provided at 75% load.
overcome the effect of temperature. This finding leads to the consider­
From Table 5, it can be observed that residual fuel leads to more OM,
ation that parameters such as the engine age, technology maintenance
EC and SO4 compared to distillate fuel use.

Table 5
OC, EC and PN Base Emission Factors, standard deviation (and number of individual values in parenthesis). OC and EC are provided for 50% engine load, while SO4
conversion rate, PN (total and non-volatile) at 75%.
Pollutant Base Emission Factor (BEF)

All types Residual Distillate

OC (g/kWh) 0,232 ± 0,158 (14) 0,248 ± 0,163 (12) 0,138 ± 0,0617 (2)
ECa (g/kWh) 0,0182 ± 0,0181 (16) 0,0206 ± 0,0193 (13) 0,00827 ± 0,00319 (3)
SO4 conversion rate (%) 3,94 ± 1,50 (6) 4,56 ± 0,627 (5) 0,827 (1)
PN (#/kWh) 3,58E+15 ± 1,87E+15 (3) N/A N/A
PN non-volatile (#/kWh) 6,33E+14 ± 2,67E+14 (2) N/A N/A
a
EC equivalent to BC.

6
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

Fig. 3. Engine load impact for normalized emission rates (NOX, CO, HC) and Specific Fuel Oil Consumption at 50% load.

Fig. 4. Engine load formed trend of pollutants for NOX normalized emission
rates for three engine technology Tier standards. Fig. 5. Normalized SO4 conversion ratio in relation with the engine load.

and specific engine tuning further play a vital role in NOX performance understand the actual mechanism that leads to the different perfor­
(Fridell et al., 2008), (Sanaz et al., 2018). Therefore, an additional mance, a fact which is out of the scope of the present paper.
analysis has been conducted to investigate the influence of Tier standard The dependance of SO2 on engine load is related to the respective
regulations on NOX load dependence. Emissions are separated to three dependencies of fuel consumption and sulphur conversion rate for the
categories, according to the Tier classes I, II and to engines built before formation of SO4. More specifically, the conversion rate to SO4 in rela­
the application of Tier regulation (referred as Tier 0 - constructed before tion with the engine load is provided in Fig. 5. The referred SOХ
2000). This classification is considered convenient to reflect the effect of represent the total sulphur emissions. The conversion rate shows the
age and technology evolution in engine manufacturing, parameters that amount of SO4 emissions in total SOX quantities. As evident the
can work as drivers of marine engines’ NOX performance differentiation. remaining part of sulphur found in fuel is converted to SO2.
The formed trends for all three types of engines are presented in Fig. 4. The expression of the conversion rate in relation to engine load
The comparison of the formed trends over the emission rates for the shows that less SO4 is produced at low loads, while the percentage of SO4
three engine technology levels (Tiers 0, I, II) verifies the hypothesis that in total emitted SOX emissions increases with load. The overall contri­
newer engines exhibit a different NOX behavior at higher loads bution of SO4 to total SOX is estimated to be between 2 and 5%,
compared to the older ones. In particular, for Tiers 0 and I energy- depending on the engine operating condition. The SO2 dependence on
specific emissions increase at higher engine loads, while in newer the engine load is actually a combination of the dependencies of the SO2 to
load rise is accompanied with a reduction of emissions. The contradic­ SO4 conversion rate and the SFOC. This relationship varies with fuel
tory findings of literature therefore seem to be dependent on the engine type (residual or distillate) as a result of different fuel properties. The
type technology, while further investigation should be made in order to trend is depicted in Fig. 6, where the SO2 and SO4 load dependent

7
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

Fig. 6. SO2 & SO4 load dependent functions for residual and distillate fuels.
Fig. 8. Ash content (%) in relation with the FSC.

functions have been calculated for these two main fuel categories, based
on the SFOC, their % FSC and the SO4 conversion rate.
In particular, for the distillate fuel where the FSC is low, the SO2
trend follows approximately the one of the SFOC. With residual fuel use,
the SO2 dependance on engine load is different than that of SFOC due to
the substantial formation of SO4 at high loads. SO4 load dependency,
also displayed in Fig. 6, as expected follows the convention rate trend in
combination with the SFOC. The establishment of a conversion rate
dependance on engine load is not only important for the estimation of
SO2 in the emitted gaseous pollutant mix, but also for the quantification
of sulphates which contribute to total PM.
PM, apart from sulphates, is also constituted by OM, EC and ash.
Already the SO4 dependency has presented in Fig. 6, while the respective
correlations of EC and OC are provided in Fig. 7.
The emission rates of OC and EC with the engine load are quite
similar, presenting a gradual improvement of these emissions with the
load increase. Thus, OC and EC are minimized at full load because of the Fig. 9. Hydrated PM load dependance for residual and distillate fuel types.
more efficient oxidation due to high combustion temperatures at this higher engine loads.
load range. For the remaining component of PM, ash is based on the fuel Particle emissions are also characterized by their number. The
ash content and SFOC. Specifically, the fuel ash content, which is a fuel identified measurements were conducted by equipment capable of
property, appears to be linearly correlated with the FSC, based on measuring particles with a diameter higher than 5 nm. The dependency
literature findings’ processing (Fig. 8). of the number with engine load as regards the total and the non-volatile
The total particle mass, dry or hydrated, as already stated is calcu­ part of particles is displayed in Fig. 10.
lated as the sum of species; hydrated sulphates (water and SO4) or dry PN is reduced at full load due to the complete combustion conditions
sulphates (SO4), OM, EC and ash, at a specific engine load. Therefore, that are being formed, while the increase of PN at low loads may be
the dependency of the pollutant with the engine load is a combination of attributed to lower temperature conditions at this load area. However,
the partial dependencies of each substance (Fig. 9 for residual and the PN correlations are less strong than the respective of PM because of
distillate). less available studies that examined the particle number compared to
The overall calculation for two fuel types reveals that for distillate those of considering the particle mass.
fuel, PM is reduced with the increase of engine load due to the major PN load dependance is influenced by the OC and sulphates (SO4). In
contribution of the OM in the PM, while the dominant presence of hy­ the low load area, the existence of lower quantities of SO4 leads to a
drated sulphates at the residual fuel use leads to higher emissions at lower PN increase, compared to the increase of OC. Also, in higher loads

Fig. 7. OC & EC engine load dependencies.

8
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

Fig. 10. PN (total & solid) performance in relation with the engine load.

the decrease of PN with the load increase is lower than the respective Various literature sources present NOX emissions rates in relation to
decrease of OC as the SO4 drastically increases at this load area. operation mode (Peng et al., 2016), (Fu et al., 2013), (Wang et al.,
2019), (Liu et al., 2018), (Winnes and Fridell, 2010). They indicate that
4.3. Engine load dependence in comparison with operating mode based power-specific NOX levels during cruising are higher compared to the
EFs phases of maneuvering and hoteling. Fig. 11 summarizes the average
emission levels calculated from all the literature sources that reported
The overall dependance of pollutants on engine load is the basis of emission rates in relation with the operating mode. For comparison
the calculation of power-based EFs, which can then be specified for purposes, we have also included in Fig. 11 the corresponding emission
different vessel operation phases. However, such power-based emission levels using the power-based load dependent emission factors developed
factors are developed by statistical analysis of measurements over in the current work. To distinguish load dependent factors between
steady-state engine points. Although marine engines are designed to cruising and maneuvering/at berth, we have assumed that cruising
primarily operate at steady states, they also experience transient oper­ corresponds to 80% engine load and that maneuvering/hoteling corre­
ation during load variations which are more frequent during maneu­ sponds to the average 30% load engine load. This assumption is based on
vering than berthing/unberthing. These modes include a sequence of the findings of Ntziachristos et al. (2016) and Trozzi and De Lauretis
operating points, as well as the transition from one point to another. The (2016) shipping inventory guidebook. Error bars in engine load data
suitability of power-based emission factors to express emission levels display the variance of emissions in loads closed to the selected one
under specific operating modes requires some further consideration. (70% and 90% for cruising, 20% and 40% for maneuvering/at berth). In

Fig. 11. Engine load in comparison with operating mode NOX EFs.

9
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

the operating mode data, error bars correspond to the standard devia­ absolute emission values and of the engine load-dependent trends.
tion of the sample. Emission levels are further distinguished by vessel A significant number of measurements has been taken into account
Tier level. For consistency between the two methods, the same fleet for most of the pollutants, the number of which is already depicted into
synthesis in terms of engine types was used. Tables 1, 2 and 5. The same tables also include the calculated standard
The comparison of the two methods exhibits rather similar absolute deviation. The significant number of measurements in most cases in­
values but with differences when comparing emission factors in cruising creases the confidence of the BEF values for most pollutants. To better
and maneuvering. When load-dependent factors are considered quantify uncertainty for practical applications Table 6 shows the
maneuvering/at berth emission rates appear higher than cruising for all calculated 95% confidence intervals, we calculated the 95% confidence
Tier levels. Operation-mode emission factors exhibit an opposite interval for each pollutant.
behavior, at least for Tier 0 and Tier I, with maneuvering/at berth The second element of the proposed emission factors is their load
appearing lower than cruising. dependence which was developed by least squares regression of
At this stage, it is not clear why this difference is observed. Part of the normalized load-dependent value. The goodness of fit of the regression
difference may perhaps be explained by the impact of load transition for each pollutant is also provided in Table 6. Moreover, an ANOVA was
during maneuvering, which is not included in load-dependent emission conducted to test the statistical significance of the dependance to the
factors. However, such transient conditions for marine engines are not as engine load. The null hypothesis made was that the mean below and
dramatic as for smaller engines used for road vehicles. For operation- above the 50% of engine load are the same. This would mean that
mode emission factors, a source of uncertainty may also come from emissions do not depend on the engine load. A p-value of less than 0,05
the calculation of emission rates in g/kWh. This comes from indepen­ leads us to reject the null hypothesis of equal means between the two
dent measurement of engine power and emission rates and it may be groups, while a p-value higher than 0,05 to accept the hypothesis.
more difficult to calculate an average value as load changes. It is also SFOC and SO4 conversion rate present the best metrics both for the
important to observe that such differences are not met for Tier II en­ BEF and the load dependency, thus a high confidence level. Good con­
gines. This may have to do with the different shape of the load depen­ fidence also exists for NOX both for the overall values of the pollutant
dence of NOx shown in Fig. 4 for Tier II than earlier engine types which and the specific breakdown per engine Tier technology. On the other
suggests a different engine calibration. So, engine tuning between steady hand, a relatively poor confidence exists for pollutants that their abso­
state and load changes may have a role to play in the difference. Finally, lute emission levels are low (CO and HC) and therefore they are more
the overall uncertainty is probably also enhanced by the diversity met in sensitive to a range of parameters that can introduce variance, namely
literature, regarding the definition of operating modes, as there is not a experimental errors, fuel properties, effect of engine wear and mainte­
common standard to be followed to distinguish maneuvering and at nance, engine design differentiations per manufacturer etc. The ANOVA
berth conditions while studying ship activity profiles. results in terms of the provided p-values highlighted a statistically sig­
With regard to PM, CO and HC, in contrast to NOX, maneuvering/at nificant difference of the mean above and below 50% engine load for all
berth results to higher emission levels than cruising (Peng, 2016), (Fu pollutants except for PN (Total and non-volatile) and NOx Tier 0. This
2013). This trend is in line with the engine-load analysis outputs pro­ confirms that energy-specific emission rates depend on engine load.
vided in Fig. 3 in which these pollutants are higher at low loads and are The emission factors developed aim at characterizing the perfor­
decreased with the load increase. mance of a fleet of vehicles with the same characteristics hence they
Although this analysis could not locate the origin of the difference in should be suitable for inventorying purposes. Obviously, deviations may
the different approaches, it is although suggestive that load-dependent be expected when these emission factors are used to evaluate the per­
and mode-dependent emission factors may differ and that this needs formance of a specific vessel. In that case, parameters such as the engine
to be further studied to understand implications in future emission type, age and maintenance level as well as fuel specificities are not
model works. explicitly taken into account and may result to performance beyond the
ranges estimated.
5. Discussion
6. Conclusions
The previous meta-analysis developed a new set of EFs for ships
based on a statistical processing of emission measurements found in The study provided mean emission rates and their engine load
literature. This new set can be used in various applications named dependence for key pollutants produced by marine diesel engines. These
emission inventories, studies as well as any other application that in­ values have been derived by combining information from a variety of
volves the estimation of shipping emissions at an aggregated level. literature sources in a consistent way. The final values and functions
However, a critical question is related to the uncertainty of the provided produced can be used in the framework of inventorying activities for

Table 6
Base Emission Factor 95% Confidence Interval, load dependent function regression R2 and ANOVA calculated p-value per each engine and fuel type.
Parameter BEF 95% Confidence interval Engine load dependency

Any type SSD MSD HSD Residual Distillate R2 ANOVA p-value

NOX (g/kWh) 11,8–14,6 12,4–16,5 10,1–14,8 6,57-16,8 – – 53,2 2,70E-04


CO (g/kWh) 0,708-1,09 0,337-1,09 0,703-1,25 0,641-1,56 – – 61,6 3,72E-07
HC (g/kWh) 0,272–0,608 0,124–0,592 0,238–0,571 0,00–1,76 – – 61,8 6,20E-08
SFOC (MJ/kWh) 8,20-9,47 7,67-9,29 7,52-9,33 7,03–12,4 – – 77,6 1,58E-04
NOX TIER 0 (g/kWh) – 16,7–18,7 8,94-12,6 3,55-12,5 – – 49,2 0,800
NOX TIER I (g/kWh) – 12,5–16,4 9,71-11,3 5,43-9,39 – – 60,7 0,0268
NOX TIER II (g/kWh) – 9,78-12,8 8,00–8,55 2,45-9,46 – – 64,8 0,00481
NOX TIER III (g/kWh) – – 0,357-1,68 0,00–3,16 – – – –
OC (mg/kWh) 138–327 – – – 139–356 0–922 67,7 2,39E-04
EC (mg/kWh) 8,30-28,2 – – – 8,43-32,7 0–18 87,9 4,50E-03
SO4 conversion rate (%) 2,21-5,67 – – – 3,69-5,43 – 75,2 1,34E-05
PN × 1015 (#/kWh) 0,00–9,27 – – – – – 51,2 0,0569
PN non-volatile × 1015 (#/kWh) 0,00–4,03 – – – – – 67,8 0,0974
Ash content (%) – – – – – – 79,1 3,58E-05

10
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

shipping emissions if knowledge of engine type, fuel type and Tier CRediT authorship contribution statement
standard are available. In particular for PM, emission factors were
expressed as composite values of individual key PM categories Achilleas Grigoriadis: Conceptualization, Methodology, Writing –
(elemental carbon, organic mass, ash and sulphates) in order to retain original draft. Sokratis Mamarikas: Conceptualization, Methodology,
consistency and provide values that can be proven useful to emission Writing – original draft. Ioannis Ioannidis: Investigation, Data cura­
and air quality modelers. tion. Elisa Majamäki: Writing – review & editing. Jukka-Pekka Jal­
Emission rates for all pollutants were found to strongly depend on kanen: Methodology, Writing – review & editing. Leonidas
engine load with g/kWh values dropping as load increased. Tier stan­ Ntziachristos: Conceptualization, Methodology, Writing – review &
dard was found as the most significant variable for the estimation of NOX editing, Writing – original draft, Supervision, Funding acquisition.
emissions with Tier III values appearing at a fraction of 10% of pre-Tier
level. Slow speed engines were also found to be higher emitters of NOX Declaration of competing interest
than faster ones. PM, on the other hand, mostly depended on fuel use
rather than on engine type or Tier standard. The authors declare that they have no known competing financial
Despite efforts to use all available engine sources and to combine interests or personal relationships that could have appeared to influence
those in the same pool of data, marine engines emission performance is the work reported in this paper.
still bound to uncertainty and more research is required to understand
fuel, engine and operation mode impacts on emission profiles. In Acknowledgement
particular, this study identified some inconsistency when expressing
emission factors according to their load dependence or operation mode This work was conducted in the framework of SCIPPER project. The
dependence. This should be further clarified in future studies as this is SCIPPER project has received funding from the European Union’s Ho­
potentially important for correctly estimating in-port emission contri­ rizon 2020 research and innovation programme under grant agreement
bution of vessels. Nr.814893.

Appendix C. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.aeaoa.2021.100142.

ANNEX I. LR Evidence

Table Pollutant Literature Sources


/Figure /SFOC

Table 1/Fig. 3 NOX Cooper (2001), Agrawal et al. (2008a), Agrawal et al. (2008b), Fridell et al. (2008), Sarvi et al. (2008), Winnes et al. (2009), Winnes et al. (2010),
Agrawal et al. (2010), Petzold et al. (2011), Miller et al. (2012), Gysel et al. (2014), Sippula et al. (2014), Johnson et al. (2016), Ryu et al. (2016),
Zetterdahl et al. (2016), Gysel et al. (2017), Kalender et al. (2017), Zhou et al. (2017), Huang et al. (2018), Johnson et al. (2018), Shen et al.
(2020), Winnes et al. (2020), Yang et al. (2019)
CO Cooper (2001), Agrawal et al. (2008a), Agrawal et al. (2008b), Fridell et al. (2008), Sarvi et al. (2008), Winnes et al. (2009), Winnes et al. (2010),
Agrawal et al. (2010), Petzold et al. (2011), Miller et al. (2012), Lehtoranta et al. (2013), Gysel et al. (2014), Sippula et al. (2014), Chu Van et al.
(2016), Johnson et al. (2016), Zetterdahl et al. (2016), Gysel et al. (2017), Kalender et al. (2017), Zhou et al. (2017), Huang et al. (2018), Johnson
et al. (2018), Shen et al. (2020), Winnes et al. (2020)
HC Cooper (2001), Fridell et al. (2008), Sarvi et al. (2008), Petzold et al. (2011), Kristensen (2012), Miller et al. (2012), Lehtoranta et al. (2013),
Sippula et al. (2014), Kalender et al. (2017), Huang et al. (2018), Sanaz et al. (2018), Yang et al. (2019), Winnes et al. (2020)
SFOC Cooper (2001), Winnes et al. (2009), Gysel et al. (2014), Ryu et al. (2016), Sanaz et al. (2018), Yang et al. (2019), Winnes et al. (2020)
Table 2 NOX Cooper (2001), Agrawal et al. (2008a), Fridell et al. (2008), Moldanová et al. (2009), Murphy et al. (2009), Winnes et al. (2009), Agrawal et al.
(2010), Winnes et al. (2010), Miller et al. (2012), Hallquist et al. (2013), Lehtoranta et al. (2013), Moldanová et al. (2013), Gysel et al. (2014),
Celo et al. (2015), Chu Van et al. (2016), Fridell et al. (2016), Johnson et al. (2016), Winnes et al. (2016), Zetterdahl et al. (2016), Gysel et al.
(2017), Kalender et al. (2017), Chu Van et al. (2018), Johnson et al. (2018), Huang et al. (2018), Sanaz et al. (2018), Lehtoranta et al. (2019a),
Lehtoranta et al. (2019b), Yang et al. (2019), Shen et al. (2020), Winnes et al. (2020), McCaffery et al. (2021)
Table 5/Fig. 7 OC Agrawal et al. (2008a), Agrawal et al. (2010), Petzold et al. (2011), Khan et al. (2012), Miller et al. (2012), Johnson et al. (2016), Gysel et al.
(2017), Huang et al. (2018), Johnson et al. (2018), Winnes et al. (2020)
EC Agrawal et al. (2008a), Agrawal et al. (2010), Petzold et al. (2011), Khan et al. (2012), Khan et al. (2013), Sippula et al. (2014), Johnson et al.
(2016), Gysel et al. (2017), Johnson et al. (2018), Winnes et al. (2020)
Table 5/ PN Petzold et al. (2011), Chu Van et al. (2016), Winnes et al. (2020)
Fig. 10 PN solid Petzold et al. (2011), Winnes et al. (2020)
Fig. 4 NOX Agrawal et al. (2008a), Agrawal et al. (2010), Winnes et al. (2010), Miller et al. (2012), Zetterdahl et al. (2016), Gysel et al. (2017), Kalender et al.
(2017), Johnson et al. (2018), Sanaz et al. (2018), Yang et al. (2019), Shen et al. (2020), Winnes et al. (2020)
Table 5/Fig. 5 SO4 Agrawal et al. (2008a), Agrawal et al. (2010), Petzold et al. (2011), Khan et al. (2012), Miller et al. (2012)
Fig. 8 Ash Cooper (2003), Agrawal et al. (2008a), Agrawal et al. (2008b), Moldanová et al. (2009), Murphy et al. (2009), Agrawal et al. (2010), Petzold et al.
(2010), Petzold et al. (2011), Miller et al. (2012), Moldanová et al. (2013), Chu Van et al. (2016), Ntziachristos et al. (2016), Winnes et al. (2016),
Zetterdahl et al. (2016), Zhang et al. (2016), Chu Van et al. (2018), Huang et al. (2018), Liu et al. (2018), Wu et al. (2018), Shen et al. (2020),
Winnes et al. (2020)
Fig. 11 NOX (Oper. Fridell et al. (2008), Moldanová et al. (2009), Murphy et al. (2009), Winnes et al. (2009), Winnes et al. (2010), Khan et al. (2012), Celo et al.
Mode) (2015), Fridell et al. (2016), Chu Van et al. (2018), Huang et al. (2018), Sanaz et al. (2018), Bai et al. (2020)

ANNEX II. Load dependent functions for Diesel fuels

11
A. Grigoriadis et al.
Pollutant Equation Engine load Load Correction Base EF/SFOC
type* (%)
a b c All types SSD MSD HSD Residual Distillate SSD MSD HSD

TIER TIER TIER TIER III TIER TIER TIER TIER III TIER TIER TIER TIER III
0 I II 0 I II 0 I II

NOX (g/kWh) 1 <50 4,14 − 4,14 2,03 13,2 14,4 12,4 11,7 – – 17,7 14,4 11,3 2,00 10,8 10,5 8,28 1,02 8,53 7,41 5,95 1,25
NOX (g/kWh) 4 ≥50 1
CO (g/kWh) 2 0–100 0,507 0,981 0,898 0,714 0,974 1,10 – – – – – – – – – – – – – –
HC (g/kWh) 1 0–100 3,21 − 5,24 2,82 0,440 0,358 0,405 0,662 – – – – – – – – – – – – – –
SFOC (MJ/kWh) 1 0–100 1,48 − 2,22 1,74 8,83 8,48 8,42 9,74 – – – – – – – – – – – – – –
CO2** (g/MJ) 5 0–100 EF CO2= (SFOC x 1000 x MCO2)/(MC x LHV)
Total S*** (g/kWh) 5 0–100 EF S = (FSC x SFOC x 1000)/LHV
SO4 conversion rate (%) 1 0–100 0,814 0 0,542 3,94% – – – 4,56 0,827 – – – – – – – – – – – –
SO2**** (g/kWh) 5 0–100 EF SO2 = (EF S - EF SO4) x MSO2/MS
SO4***** (g/kWh) 5 0–100 EF SO4 = (EF S x SO4 conversion rate) x MSO4/MS
OC (g/kWh) 2 0–100 0,661 0,597 – 0,232 – – – 0,248 0,138 – – – – – – – – – – – –
EC (BC) 2 0–100 0,420 1,25 – 0,0182 – – – 0,0206 0,00827 – – – – – – – – – – – –
(g/kWh)
Ash (g/kWh) 5 0–100 EF Ash = (Ash Cont. x SFOC x 1000)/LHV
Ash (conv. from FSC)****** 3 0–100 0,0179 –
12

– – – – – – – – – – – – – – – – – – –
(%)
Dry PM (g/kWh) 5 0–100 Dry PM = OM + EC + SO4 + Ash
Hydrated PM (g/kWh) 5 0–100 Hydrated PM = OM + EC + (SO4 +6,5H2O) + Ash
PN Total (g/kWh) 1 0–100 2,03 − 3,84 2,74 3,58E+15 – – – – – – – – – – – – – – – – –
PN Solid (Non-Volatile) (g/ 2 0–100 0,853 0,553 – 6,33E+14 – – – – – – – – – – – – – – – – –
kWh)

*Equation type: 1. polyonymic - y = ax^2 + bx + c, 2. power - y = ax^(-b), 3. linear - y = ax, 4. steady - y = c, 5. specific expression provided within table.
**CO2: FCC - Fuel Carbon Content, MCO2 - Molar weight of CO2, MC - Molar weight of Carbon, LHV - Lower Heating Value.
***Total S: FSC - Fuel Sulphur Content.
****SO2: MSO2 - Molar weight of SO2, MS - Molar weight of sulphur.
*****SO4: MSO4 - Molar weight of SO4, MS - Molar weight of sulphur.
******Ash (conv. from FSC): Ash content can be calculated as a function of FSC.

Atmospheric Environment: X 12 (2021) 100142


A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

References IMO, 2018a. GloMEEP Project Coordination Unit, Institute of Marine Engineering,
Science and Technology (IMarEST). Ship Emissions Toolkit. Guide No.1: Rapid
Assessment of Ship Emissions in the National Context.
Agrawal, H., Welch, W.A., Miller, J.W., Cocker, D.R., 2008a. Emission measurements
IMO, 2018b. Initial IMO Strategy on Reduction of GHG Emissions from Ships.
from a crude oil tanker at sea. Environ. Sci. Technol. 42 (19), 7098–7103.
https://wwwcdn.imo.
Agrawal, H., Malloy, Q.G., Welch, W.A., Miller, J.W., Cocker III, D.R., 2008b. In-use
org/localresources/en/OurWork/Environment/Documents/Resolution%
gaseous and particulate matter emissions from a modern ocean going container
20MEPC.304(72)_E.pdf.
vessel. Atmos. Environ. 42 (21), 5504–5510.
IMO, 2020. Fourth IMO GHG study. Full report.
Agrawal, H., Welch, W.A., Henningsen, S., Miller, J.W., Cocker III, D.R., 2010. Emissions
Jalkanen, J.P., Brink, A., Kalli, J., Pettersson, H., Kukkonen, J., Stipa, T., 2009.
from main propulsion engine on container ship at sea. J. Geophys. Res.: Atmosphere
A modelling system for the exhaust emissions of marine traffic and its application in
115 (D23).
the Baltic Sea area. Atmos. Chem. Phys. 9 (23), 9209–9223.
Anderson, M., Salo, K., Fridell, E., 2015. Particle-and gaseous emissions from an LNG
Jalkanen, J.P., Johansson, L., Kukkonen, J., Brink, A., Kalli, J., Stipa, T., 2012. Extension
powered ship. Environ. Sci. Technol. 49 (20), 12568–12575.
of an assessment model of ship traffic exhaust emissions for particulate matter and
Bai, C., Li, Y., Liu, B., Zhang, Z., Wu, P., 2020. Gaseous emissions from a seagoing ship
carbon monoxide. Atmos. Chem. Phys. 12 (5), 2641–2659.
under different operating conditions in the coastal region of China. Atmosphere 11
Johnson, K., Miller, W., Durbin, T., Jiang, Y., Yang, J., Karavalakis, G., Cocker, D., 2016.
(3), 305.
Black Carbon Measurement Methods and Emission Factors from Ships. International
Balcombe, P., Brierley, J., Lewis, C., Skatvedt, L., Speirs, J., Hawkes, A., Staffell, I., 2019.
Council on Clean Transportation (ICCT).
How to decarbonise international shipping: options for fuels, technologies and
Johnson, K., Miller, W., Yang, J., 2018. Evaluation of a Modern Tier 2 Oceangoing Vessel
policies. Energy Convers. Manag. 182, 72–88.
Equipped with a Scrubber. Final report prepared for CARB.
CARB, 2008. Emissions Estimation Methodology for Ocean-Going Vessels.
Jonson, J.E., Gauss, M., Schulz, M., Jalkanen, J.P., Fagerli, H., 2020. Effects of global
Celo, V., Dabek-Zlotorzynska, E., McCurdy, M., 2015. Chemical characterization of
ship emissions on European air pollution levels. Atmos. Chem. Phys. 20 (19),
exhaust emissions from selected Canadian marine vessels: the case of trace metals
11399–11422.
and lanthanoids. Environ. Sci. Technol. 49 (8), 5220–5226.
Kalender, S.S., Ergin, S., 2017. An experimental investigation into the particulate
Chu Van, T., Rainey, T., Ristovski, Z., Pourkhesalian, A.M., Garaniya, V., Abbassi, R.,
emissions of a ferry fuelled with ultra-low sulfur diesel. J. Mar. Sci. Technol. 25 (5),
Brown, R., 2016. Particle emissions from ships at berth using heavy fuel oil. In:
499–507.
Proceedings of the 17th International Association of Maritime Universities (IAMU)
Khan, M.Y., Giordano, M., Gutierrez, J., Welch, W.A., Asa-Awuku, A., Miller, J.W.,
Annual General Assembly (AGA), pp. 235–243.
Cocker III, D.R., 2012. Benefits of two mitigation strategies for container vessels:
Chu-Van, T., Ristovski, Z., Pourkhesalian, A.M., Rainey, T., Garaniya, V., Abbassi, R.,
cleaner engines and cleaner fuels. Environ. Sci. Technol. 46 (9), 5049–5056.
Brown, R.J., 2018. On-board measurements of particle and gaseous emissions from a
Khan, M.Y., Ranganathan, S., Agrawal, H., Welch, W.A., Laroo, C., Miller, J.W.,
large cargo vessel at different operating conditions. Environ. Pollut. 237, 832–841.
Cocker III, D.R., 2013. Measuring in-use ship emissions with international and US
Comer, B., Olmer, N., Mao, X., Roy, B., Rutherford, D., 2017. Black carbon emissions and
federal methods. J. Air Waste Manag. Assoc. 63 (3), 284–291.
fuel use in global shipping. In: 2015. The International Council on Clean
Kristensen, H.O., 2012. Energy demand and exhaust gas emissions of marine engines.
Transportation, p. 1225.
Clean Shipping Curr. 1 (6), 18–26.
Cooper, D.A., 2001. Exhaust emissions from high speed passenger ferries. Atmos.
Kuittinen, N., Jalkanen, J.P., Alanen, J., Ntziachristos, L., Hannuniemi, H., Johansson, L.,
Environ. 35 (24), 4189–4200.
Karjalainen, P., Saukko, E., Isotalo, M., Aakko-Saksa, P., Lehtoranta, K., Keskinen, J.,
Cooper, D.A., 2003. Exhaust emissions from ships at berth. Atmos. Environ. 37 (27),
Simonen, P., Saarikoski, S., Asmi, E., Laurila, T., Hillamo, R., Mylläri, F.,
3817–3830.
Lihavainen, H., Timonen, H., Rönkkö, T., 2020. Shipping Remains a Globally
Corbett, J.J., Winebrake, J.J., Green, E.H., Kasibhatla, P., Eyring, V., Lauer, A., 2007.
Significant Source of Anthropogenic PN Emissions Even after 2020 Sulfur
Mortality from ship emissions: a global assessment. Environ. Sci. Technol. 41 (24),
Regulation. Environmental Science & Technology.
8512–8518.
Lamas, M.I., Rodriguez, C.G., 2012. Emissions from marine engines and NOx reduction
DNV, 2020. Energy transition outlook. Maritime Forecast to 2050. Full report.
methods. J. Marit. Res. 9 (1), 77–81.
EEA, 2020. Greenhouse Gas Emissions from Transport in Europe. https://www.eea.
Lehtoranta, K., Turunen, R., Vesala, H., Nyyssönen, S., Soikkeli, N., Esselström, L., 2013.
europa.eu/data-and-maps/indicators/transport-emissions-of-greenhouse-gases-7/a
Testing SCR in high sulphur application. In: 27th CIMAC World Congress on
ssessment. (Accessed 13 October 2021).
Combustion Engine Technology.
Ekmekçioğlu, A., Kuzu, S.L., Ünlügençoğlu, K., Çelebi, U.B., 2020. Assessment of
Lehtoranta, K., Vesala, H., Koponen, P., Korhonen, S., 2015. Selective catalytic reduction
shipping emission factors through monitoring and modelling studies. Sci. Total
operation with heavy fuel oil: NO x, NH3, and particle emissions. Environ. Sci.
Environ. 743, 140742.
Technol. 49 (7), 4735–4741.
European Commission, 2013. Integrating Maritime Transport Emissions in the EU’s
Lehtoranta, K., Murtonen, T., Vesala, H., Koponen, P., Karjalainen, P., Ntziachristos, L.,
Greenhouse Gas Reduction Policies. https://ec.europa.eu/clima/sites/default/files/
Kuittinen, N., Alanen, J., Ronkko, T., Keskinen, J., Aurela, M., Timonen, H., 2017.
transport/shipping/docs/com_2013_479_en.pdf.
Particle and gaseous emissions from a dual-fuel marine engine. In: 22nd
Fridell, E., Salo, K., 2016. Measurements of abatement of particles and exhaust gases in a
International Transport and Air Pollution Conference. TAP, 2017.
marine gas scrubber. Proc. IME M J. Eng. Marit. Environ. 230 (1), 154–162.
Lehtoranta, K., Aakko-Saksa, P., Murtonen, T., Vesala, H., Ntziachristos, L., Rönkkö, T.,
Fridell, E., Steen, E., Peterson, K., 2008. Primary particles in ship emissions. Atmos.
Timonen, H., 2019a. Particulate mass and nonvolatile particle number emissions
Environ. 42 (6), 1160–1168.
from marine engines using low-sulfur fuels, natural gas, or scrubbers. Environ. Sci.
Fridell, E., Salberg, H., Salo, K., 2020. Measurements of emissions to air from a marine
Technol. 53 (6), 3315–3322.
engine fueled by methanol. J. Mar. Sci. Appl. 1–6.
Lehtoranta, K., Aakko-Saksa, P., Murtonen, T., Vesala, H., Kuittinen, N., Rönkkö, T.,
Fu, M., Ding, Y., Ge, Y., Yu, L., Yin, H., Ye, W., Liang, B., 2013. Real-world emissions of
Ntziachristos, L., Karjalainen, P., Timonen, H., Teinilä, K., 2019b. Particle and
inland ships on the Grand Canal, China. Atmos. Environ. 81, 222–229.
gaseous emissions from marine engines utilizing various fuels and aftertreatment
Georgoulias, A.K., Boersma, K.F., van Vliet, J., Zhang, X., Zanis, P., de Laat, J., 2020.
systems. In: 29th CIMAC World Congress on Combustion Engine, p. 399.
Detection of NO2 pollution plumes from individual ships with the TROPOMI/S5P
Liu, Y., Ge, Y., Tan, J., Fu, M., Shah, A.N., Li, L., Ding, Y., 2018. Emission characteristics
satellite sensor. Environ. Res. Lett. 15 (12), 124037.
of offshore fishing ships in the Yellow Bo Sea, China. J. Environ. Sci. 65, 83–91.
Gysel, N.R., Russell, R.L., Welch, W.A., Cocker III, D.R., Ghosh, S., 2014. Impact of
Matthias, V., Bewersdorff, I., Aulinger, A., Quante, M., 2010. The contribution of ship
sugarcane renewable fuel on in-use gaseous and particulate matter emissions from a
emissions to air pollution in the North Sea regions. Environ. Pollut. 158 (6),
marine vessel. Energy Fuels 28 (6), 4177–4182.
2241–2250.
Gysel, N.R., Welch, W.A., Johnson, K., Miller, W., Cocker III, D.R., 2017. Detailed
McCaffery, C., Zhu, H., Karavalakis, G., Durbin, T.D., Miller, J.W., Johnson, K.C., 2021.
analysis of criteria and particle emissions from a very large crude Carrier using a
Sources of air pollutants from a Tier 2 ocean-going container vessel: main engine,
novel ECA fuel. Environ. Sci. Technol. 51 (3), 1868–1875.
auxiliary engine, and auxiliary boiler. Atmos. Environ. 245, 118023.
Hallquist, Å.M., Fridell, E., Westerlund, J., Hallquist, M., 2013. Onboard measurements
Miller, W., Khan, Y., Agrawal, H., Ranganathan, S., Welch, W., Cocker, D., June 2012. In-
of nanoparticles from a SCR-equipped marine diesel engine. Environ. Sci. Technol.
use Emissions Test Program at VSR Speeds for Oceangoing Container Ship, Report
47 (2), 773–780.
for California Air Resources Board. Contract #08-412.
Hansson, J., Brynolf, S., Fridell, E., Lehtveer, M., 2020. The potential role of ammonia as
Moldanová, J., Fridell, E., Popovicheva, O., Demirdjian, B., Tishkova, V., Faccinetto, A.,
marine fuel—based on energy systems modeling and multi-criteria decision analysis.
Focsa, C., 2009. Characterisation of particulate matter and gaseous emissions from a
Sustainability 12 (8), 3265.
large ship diesel engine. Atmos. Environ. 43 (16), 2632–2641.
Huang, C., Hu, Q., Wang, H., Qiao, L., Wang, H., Zhou, M., Zhu, S., Lou, S., Li, L., Tao, S.,
Moldanová, J., Fridell, E., Winnes, H., Holmin-Fridell, S., Boman, J., Jedynska, A., et al.,
Li, Y., Lou, D., 2018. Emission factors of particulate and gaseous compounds from a
2013. Physical and chemical characterisation of PM emissions from two ships
large cargo vessel operated under real-world conditions. Environ. Pollut. 242,
operating in European Emission Control Areas. Atmos. Measur. Tech. 6 (12),
667–674.
3577–3596.
IMO, 1997. International Convention for the Prevention of Pollution from Ships
Murphy, S.M., Agrawal, H., Sorooshian, A., Padró, L.T., Gates, H., Hersey, S., Welch, W.,
(MARPOL) Annex VI. https://www.imo.org/en/OurWork/Environment/Pages/Air
Jung, H., Miller, W., Cocker, D., Nenes, A., Jonsson, H., Flagan, R., Seinfeld, J.H.,
-Pollution.aspx.
2009. Comprehensive simultaneous shipboard and airborne characterization of
IMO, 2011. Inclusion of Regulations on Energy Efficiency for Ships in MARPOL Annex VI.
exhaust from a modern container ship at sea. Environ. Sci. Technol. 43 (13),
https://www.imo.org/en/OurWork/Environment/Pages/Technical-and-Operat
4626–4640.
ional-Measures.aspx.
Nabi, M.N., Hustad, J.E., 2012. Investigation of engine performance and emissions of a
IMO, 2016. Marpol Annex VI, sulphur oxides (SOx) and particulate matter (PM) –
diesel engine with a blend of marine gas oil and synthetic diesel fuel. Environ.
regulation 14. Resolution MEPC 280 (70). https://www.classnk.or.jp/hp/pdf/activit
Technol. 33 (1), 9–15.
ies/statutory/soxpm/resmepc280-70.pdf.

13
A. Grigoriadis et al. Atmospheric Environment: X 12 (2021) 100142

Nova, I., Tronconi, E. (Eds.), 2014. Urea-SCR Technology for deNOx after Treatment of Sippula, O., Stengel, B., Sklorz, M., Streibel, T., Rabe, R., Orasche, J., Zimmermann, R.,
Diesel Exhausts, vol. 5. Springer, New York. 2014. Particle emissions from a marine engine: chemical composition and aromatic
Ntziachristos, L., Saukko, E., Lehtoranta, K., Rönkkö, T., Timonen, H., Simonen, P., emission profiles under various operating conditions. Environ. Sci. Technol. 48 (19),
Karjalainen, P., Keskinen, J., 2016. Particle emissions characterization from a 11721–11729.
medium-speed marine diesel engine with two fuels at different sampling conditions. Trozzi, C., De Lauretis, R., 2016. International Navigation, National Navigation, National
Fuel 186, 456–465. Fishing, EMEP/EEA Emission Inventory Guidebook.
Ntziachristos, L., Saukko, E., Rönkkö, T., Lehtoranta, K., Timonen, H., Hillamo, R., Trozzi, C., Vaccaro, R., 2006. June). Methodologies for estimating air pollutant emissions
Keskinen, J., 2016b. Impact of sampling conditions and procedure on particulate from ships: a 2006 update. In: Poster Presented at 2nd Environment & Transport
matter emissions from a marine diesel engine. In: In 28th CIMAC World Congress, Conference (Including 15th Transport and Air Pollution Conference), pp. 12–14.
p. 165. Uk Limited, Entec, 2002. Quantification of Emissions from Ships Associated with Ship
Peng, Z., Ge, Y., Tan, J., Fu, M., Wang, X., Chen, M., Ji, Z., 2016. Emissions from several Movements between Ports in the European Community. Final report July 2002.
in-use ships tested by portable emission measurement system. Ocean Eng. 116, U.S. Environmental Protection Agency, April 2009. Current Methodologies in Preparing
260–267. Mobile Source Port-Related Emission Inventories. Final Report Prepared for U.S.
Petzold, A., Hasselbach, J., Lauer, P., Baumann, R., Franke, K., Gurk, C., Weingartner, E., Environmental Protection Agency by ICF International.
2008. Experimental studies on particle emissions from cruising ship, their Viana, M., Hammingh, P., Colette, A., Querol, X., Degraeuwe, B., de Vlieger, I., van
characteristic properties, transformation and atmospheric lifetime in the marine Aardenne, J., 2014. Impact of maritime transport emissions on coastal air quality in
boundary layer. Atmos. Chem. Phys. 8 (9), 2387–2403, 52. Europe. Atmos. Environ. 90, 96–105.
Petzold, A., Weingartner, E., Hasselbach, J., Lauer, P., Kurok, C., Fleischer, F., 2010. Wang, C., Hao, L., Ma, D., Ding, Y., Lv, L., Zhang, M., Ge, Y., 2019. Analysis of ship
Physical properties, chemical composition, and cloud forming potential of emission characteristics under real-world conditions in China. Ocean Eng. 194,
particulate emissions from a marine diesel engine at various load conditions. 106615.
Environ. Sci. Technol. 44 (10), 3800–3805. Winnes, H., Fridell, E., 2009. Particle emissions from ships: dependence on fuel type.
Petzold, A., Lauer, P., Fritsche, U., Hasselbach, J., Lichtenstern, M., Schlager, H., J. Air Waste Manag. Assoc. 59 (12), 1391–1398.
Fleischer, F., 2011. Operation of marine diesel engines on biogenic fuels: Winnes, H., Fridell, E., 2010. Emissions of NOX and particles from manoeuvring ships.
modification of emissions and resulting climate effects. Environ. Sci. Technol. 45 Transport. Res. Transport Environ. 15 (4), 204–211.
(24), 10394–10400. Winnes, H., Moldanová, J., Anderson, M., Fridell, E., 2016. On-board measurements of
Pirjola, L., Pajunoja, A., Walden, J., Jalkanen, J.P., Rönkkö, T., Kousa, A., Koskentalo, T., particle emissions from marine engines using fuels with different sulphur content.
2014. Mobile measurements of ship emissions in two harbour areas in Finland. Proc. IME M J. Eng. Marit. Environ. 230 (1), 45–54.
Atmos. Measur. Tech. 7 (1), 149–161. Winnes, H., Fridell, E., Moldanová, J., 2020. Effects of marine exhaust gas scrubbers on
Raptotasios, S.I., Sakellaridis, N.F., Papagiannakis, R.G., Hountalas, D.T., 2015. gas and particle emissions. J. Mar. Sci. Eng. 8 (4), 299.
Application of a multi-zone combustion model to investigate the NOx reduction Woodyard, D., 2004. Pounder’s Marine Diesel Engines and Gas Turbines, Eight edition.
potential of two-stroke marine diesel engines using EGR. Appl. Energy 157, Butterworth-Heinemann.
814–823. Wu, D., Li, Q., Ding, X., Sun, J., Li, D., Fu, H., Chen, J., 2018. Primary particulate matter
Ryu, Y., Lee, Y., Nam, J., 2016. Performance and emission characteristics of additives- emitted from heavy fuel and diesel oil combustion in a typical container ship:
enhanced heavy fuel oil in large two-stroke marine diesel engine. Fuel 182, 850–856. characteristics and toxicity. Environ. Sci. Technol. 52 (21), 12943–12951.
Sanaz, J., Natalia, N., Kiril, T., 2018. Empirical testing of inventories applying on-board Yang, Z., Tan, Q., Geng, P., 2019. Combustion and Emissions Investigation on Low-Speed
measurements of exhaust emissions at port and at sea. J. Sustain. Develop. Trans. Two-Stroke Marine Diesel Engine with Low Sulfur Diesel Fuel. Polish Maritime
Logis. 3 (2) (5). Research.
Sarvi, A., Fogelholm, C.J., Zevenhoven, R., 2008. Emissions from large-scale medium- Zetterdahl, M., Moldanova, J., Pei, X., Pathak, R.K., Demirdjian, B., 2016. Impact of the
speed diesel engines: 2. Influence of fuel type and operating mode. Fuel Process. 0.1% fuel sulfur content limit in SECA on particle and gaseous emissions from
Technol. 89 (5), 520–527. marine vessels. Atmos. Environ. 145, 338–345.
Sarvi, A., Lyyränen, J., Jokiniemi, J., Zevenhoven, R., 2011. Particulate emissions from Zhang, F., Chen, Y., Tian, C., Lou, D., Li, J., Zhang, G., Matthias, V., 2016. Emission
large-scale medium-speed diesel engines: 1. Particle size distribution. Fuel Process. factors for gaseous and particulate pollutants from offshore diesel engine vessels in
Technol. 92 (10), 1855–1861. China. Atmos. Chem. Phys. 16 (10), 6319–6334.
Shen, F., Li, X., 2020. Effects of fuel types and fuel sulfur content on the characteristics of Zhou, J., Zhou, S., Zhu, Y., 2017. Characterization of particle and gaseous emissions from
particulate emissions in marine low-speed diesel engine. Environ. Sci. Pollut. Control marine diesel engines with different fuels and impact of after-treatment technology.
Ser. 27 (30), 37229–37236. Energies 10 (8), 1110.
Sinha, P., Hobbs, P.V., Yokelson, R.J., Christian, T.J., Kirchstetter, T.W., Bruintjes, R.,
2003. Emissions of trace gases and particles from two ships in the southern Atlantic
Ocean. Atmos. Environ. 37 (15), 2139–2148.

14

You might also like