Journal of the Air & Waste Management Association

ISSN: 1096-2247 (Print) 2162-2906 (Online) Journal homepage: https://www.tandfonline.com/loi/uawm20

NOx emission calculations for bulk carriers by

using engine power probabilities as weighting
factors

Chih-Wen Cheng, Jian Hua & Daw-Shang Hwang

To cite this article: Chih-Wen Cheng, Jian Hua & Daw-Shang Hwang (2017) NOx emission
calculations for bulk carriers by using engine power probabilities as weighting factors, Journal of the
Air & Waste Management Association, 67:10, 1146-1157, DOI: 10.1080/10962247.2017.1356763

To link to this article: https://doi.org/10.1080/10962247.2017.1356763

Published online: 05 Sep 2017.

Submit your article to this journal

Article views: 7348

View related articles

View Crossmark data

Citing articles: 4 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uawm20
JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION
2017, VOL. 67, NO. 10, 1146–1157
https://doi.org/10.1080/10962247.2017.1356763

NOx emission calculations for bulk carriers by using engine power probabilities
as weighting factors
Chih-Wen Cheng, Jian Hua, and Daw-Shang Hwang
Department of Marine Engineering, National Taiwan Ocean University, Keelung, Taiwan, ROC

ABSTRACT ARTICLE HISTORY

An important marine pollution issue identified by the International Maritime Organization (IMO) is Received April 28, 2017
NOx emissions; however, the stipulated method for determining the NOx certification value does not Revised July 6, 2017
reflect the actual high emission factors of slow-speed two-stroke diesel engines over long-term slow Accepted July 13, 2017
steaming. In this study, an accurate method is presented for calculating the NOx emission factors and
total amount of NOx emissions by using the actual power probabilities of the diesel engines in four
types of bulk carriers. The proposed method is suitable for all types and purposes of diesel engines, is
not restricted to any operating modes, and is highly accurate. Moreover, it is recommended that the
IMO-stipulated certification value calculation method be modified accordingly to genuinely reduce
the amount of NOx emissions. The successful achievement of this level of reduction will help improve
the air quality, especially in coastal and port areas, and the health of local residents.
Implications: As per the IMO, the NOx emission certification value of marine diesel engines having a
rated power over 130 kW must be obtained using specified weighting factor (WF)-based calculation.
However, this calculation fails to represent the current actual situation. Effective emission reductions of
6.91% (at sea) and 31.9% (in ports) were achieved using a mathematical model of power probability
functions. Thus, we strongly recommend amending the certification value of NOx Technical Code 2008
(NTC 2008) by removing the WF constraints, such that the NOx emissions of diesel engines is lower than
the Tier-limits at any load level to obtain genuine NOx emission reductions.

Introduction

Following the adoption of Annex VI of the International
Convention for the Prevention of Pollution from Ships
(MARPOL) by the International Maritime Organization
(IMO) in 1997, the regulations governing the emission of
NOx, SOx, and CO2 have gradually become more stringent.
After discussions over the course of several sessions at the
57th Session in 2008, the Marine Environment Protection
Committee (MEPC, the Committee) of the IMO approved
the proposed amendment that all new marine diesel
engines must comply with the three-tier NOx emission
regulation. The amendment was finally adopted by the
Committee at the 58th Session in the same year, as
shown in Figure 1. The goal of this regulation is to reduce
the level of NOx emissions caused by the international
shipping industry by 20% after 2011, and to reduce NOx
emissions in emission control areas (ECAs) by 80% after
2016 (IMO 2008a; 2008b; 2008c; 2008d). MARPOL Annex
VI also specifies that from 2012 onward, the sulfur content
in fuel oil for international shipping purposes must not
exceed 3.5%, and will be further limited to no more than
0.5% after 2020. Moreover, beginning in 2015, the sulfur
content in fuel oil within the sulfur emission control areas
(SECAs) must not exceed 0.1%. Synthesizing all kinds of
reasons, many large ships have practiced slow steaming in
response, in order to reduce emissions, and due to sub-
stantially increased fuel costs (Doudnikoff and Lacoste
2014; Notteboom and Vernimmen 2009).
Many scholars believe that shipping is a highly polluting
mode of transport because the ships emit large quantities of
NOx and SOx (Corbett and Fischbeck 1997; Corbett and
Koehler 2003; Endresen et al. 2003). The Third IMO
Greenhouse Gas (GHG) Study estimated that over the
years from 2007 to 2012, an average of 20.9 million and
11.3 million tonnes of NOx and SOx, respectively, were
emitted annually from all shipping activities, while inter-
national shipping was estimated to produce approximately
18.6 million and 10.6 million tonnes of NOx and SOx,
respectively, on an annual basis. The NOx and SOx emis-
sions from all shipping activities represented approxi-
mately 15% and 13% of global NOx and SOx emissions
from anthropogenic sources, respectively, as reported in
the latest Intergovernmental Panel on Climate Change

CONTACT Chih-Wen Cheng d98660001@ntou.edu.tw Department of Marine Engineering, National Taiwan Ocean University, No. 2, Beining Road,
Jhongjheng District, Keelung 202, Taiwan, ROC.
© 2017 A&WMA
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION
Figure 1. Three tiers of NOx emission limits.
(IPCC) Assessment Report. According to the IMO, NOx
and SOx emissions from international shipping repre-
sented approximately 13% and 12% of global NOx and
SOx totals, respectively (IMO 2014).
In the IMO Study, which is intended to serve as a key
reference on global shipping emissions, the statistics of
emissions in various categories were collected using
bottom-up and top-down methods, and the information
regarding ship speeds was gathered using the automatic
identification system (AIS). In addition, the fuel oil
consumption (FOC) and main engine (ME) power at
approximately 75% of full power were estimated, the
CO2 and SOx emissions were calculated, and the NOx
emissions were estimated using the fixed NOx emission
values as the emission factors (EFs).
The levels of SOx and CO2 emissions are dependent on
the FOC. The ratio of the CO2 EF to specific fuel oil
consumption (SFOC) can be constant. Heavy fuel oil
(HFO) has a relatively low carbon content, for which the
constant 3.114 (g CO2/g HFO) is widely adopted. In
contrast, the carbon content of marine diesel oil (MDO)
and marine gas oil (MGO) is higher than that of HFO, for
which the constant 3.206 (g CO2/g MDO or MGO) is
often used (IMO 2012). The FOC of the relative
horsepower can be accurately measured by mounting
magnetic-type fuel meters with 1% error margin at the
fuel inlet and return line of diesel engines; however, the
level of NOx emissions is associated with the engine load
and combustion temperature. A higher combustion tem-
perature and the time that combustion gases remain at
high temperature result in greater NOx emissions. In
addition, the NOx EF and combustion temperature vary
with engine load, fuel injection timing, and quantity,
1147
especially for the load factor (Lf) under 0.25. Therefore,
the factors that effectively influence the amount of NOx
emissions from ships are the engine load and output, and
not the speed. Although the relationship between the
speed of the ship and ME output follows the propeller
law, due to variations in the resistances of the ship for
different ship types and propeller slippages, which can be
attributed to variations in the sea state conditions, the
actual ME load and ship speed may vary at any point in
time (European Commission and ENTEC UK Limited
2002; Khan et al. 2013; IMO 2016). The ME power is not
proportional to the exact cubic power of the speed of the
ship. Furthermore, the propeller design and ME charac-
teristics may vary with the type of ship. For example, the
exponent values of container ships and bulk carriers may
be between 2.8 and 3.2 because of different thrust power,
hull, and propeller profile design. Thus, the actual ME
horsepower cannot be precisely determined from the
speed of the ship via the propeller law (Borkowski,
Kasyk, and Kowalak 2011). However, due to recent
advances in electronics, science, and technology, the
power meters that are built in to modern ME control
systems can accurately measure the real-time horsepower
variations during operation of the ship, thereby enabling
convenient and precise horsepower-associated calcula-
tions to be performed. Hence, the most simplified and
accurate method for calculating the exhaust emissions is
to compute the SOx and CO2 emissions using the mea-
sured FOC, and to calculate the NOx emissions based on
the actual engine load or output, instead of the speed.
Following the mandatory enforcement of MARPOL
Annex VI, each diesel engine with a rated power that
exceeds 130 kW is required to be tested in compliance
with the NOx Technical Code 2008 (NTC 2008) regulation,
and to acquire the NOx emission certification value via the
specified weighting factor (WF)-based calculation.
Different types of diesel engines possess various character-
istics and potentially different resulting NOx EFs. Before
the energy crisis, almost all commercial ships navigated at
75–85% high engine load levels at sea because of high
thermal efficiency. At lower loads, the two-stroke main
diesel engines tend to experience longer combustion
times, lower efficiencies, and lower power outputs, and
therefore generate relatively large amounts of NOx per
unit power, which leads to higher EFs when compared
with high load operations (Andereadis et al. 2011;
Corbett and Koehler 2003). This phenomenon is more
pronounced for the electrohydraulically controlled large
diesel engines and common rail engines that conform to
the Tier II regulations.
In order to manufacture engines that comply with the
Tier II emission regulations and to lower the SFOC to
satisfy market demand, engine manufacturers relaxed the
1148 C.-W. CHENG ET AL.
NOx emission control at low loads. A lower load can
contribute to a higher NOx EF, although the NOx emission
certification value will still be lower than the Tier II limits
according to the low-load, low-WF calculation specified in
the NTC 2008 regulations. However, this WF allocation
fails to properly represent the actual situation that has
developed over the last decade wherein large-scale vessels
experience long periods of slow steaming. Excessive calcu-
lation errors of NOx emissions will occur if NOx EFs are
regarded as fixed values while the actual load state and
operation time are not considered. Moreover, greater var-
iations in the ship speed and ME load will lead to higher
errors. The NOx EFs of Tier II MEs running at 25% load
are approximately 55% higher compared to the 100% load.
The relationships of NOx EFs and engine load are
explained later with detailed equations.
In this paper, the actual NOx EFs and total emissions for
four types of large bulk carriers (BCs) with over 200,000
deadweight (DWT) are calculated using the mathematical
model of power probability functions for the MEs in actual
navigation. The computed results are then compared to the
certification values and data published by the IMO.
Moreover, it is suggested that the certification value calcu-
lation method of NTC 2008 be amended by removing the
WF constraints, such that the NOx emissions of diesel
engines must be lower than the Tier-limits at any load
level for the purpose of genuine NOx emission reductions.
Based on these results, an effective emission reduction of
6.91% at sea and 31.9% in port areas, respectively, is shown.
The correct calculation of the actual EFs and total emis-
sions may facilitate the accurate inventory of global ship-
ping emissions, which may serve as key reference indexes
for effective emission reduction, life cycle assessment, and
policy formulation.
Methodology
Basic information and operation modes of four
types of BCs
The four types of BCs with fixed pitch propeller examined
in this work are mainly engaged in the transport of coal
Table 1. Particulars of four types of BCs.
Type BC1
LOA × breadth × depth (M) 299.9 × 50.0 × 24.4
DWT (MT) 203,510
ME type 6S70MC-C
ME rated power (kW@RPM) 18,700@91
Rated speed (knots) 16.5
ME NOx Tier limit (g/kWh) I (17.0)
ME certified value (g/kWh) 14.60
GE NOx Tier limit (g/kWh) I (12.1)
GE certified value (g/kWh) 11.8
Keel laid date June 2006
from Australia to Taiwan, with two to four sister ships for
each type. Once the estimated time of arrival (ETA) to a
port is determined, the ship speed, ME output, and revo-
lutions per minute (RPM) remain almost unaltered dur-
ing navigation. In addition, slow steaming is practiced in
most cases to reduce the FOC. The research data used in
this paper were extracted from data representing at least
12 months of operations of each representative vessel for
each BC type between 2013 and 2015, wherein the ME
load, power output, RPM, and generator engine (GE)
power output were extracted from the electronic control
system of the ME and the engine room (E/R) log book.
These engines were tested/witnessed to obtain important
data comprising the engine output, NOx EF, SFOC, and
so on during certification. The information on machine
testing, equipment, calibration records, and test results of
engines are recorded in detail in well-maintained NOx
Technical Files (Classification Societies 2014). For exam-
ple, the manufacturer of the dynamometer for engine
power measurement is Fuchino (serial number 91032)
with a maximum measurement range of 100,000 BHP.
The manufacturer of the NOx analyzer is Horiba (model
CLA-155) with a maximum measurement range of 2,000
ppm, while that of the weighting machine for FOC is CAS
(model 1W78).
In order to improve the propulsion efficiency and
reduce the fuel costs, the hull and machinery designs are
being continuously enhanced and optimized, thereby
enhancing the overall efficiency in order to respond to
market demands. The basic information of the four types
of BCs is shown in Table 1. The fuel and exhaust valves of
the BC1 ME are controlled using a chain and camshaft
transmission with a relatively high SFOC. For the BC2
ME, electrohydraulic control is applied to accurately reg-
ulate the opening and closing of the fuel and exhaust
valves, and the ME rated power and RPM are reduced,
leading to an improvement of the overall propulsion
efficiency and reduction of the SFOC. BC3 and BC4
featured a superior hull shape design and propulsion
efficiency, which provides a reduction in the submerged
resistance and the wind resistance above the surface of the
water. The rated power and RPM of the BC4 ME are
BC2 BC3 BC4
299.7 × 50.0 × 25.0 299.7 × 50.0 × 25.0 299.70 × 50.0 × 25.0
206,630 206,380 209,750
7S65ME-C(Mk8) 7S65ME-C8.2 7S65ME-C8.2
16,800@86 16,800@86 16,000@84
16.5 16.7 16.7
I (17.0) II (14.4) II (14.4)
15.40 13.68 13.10
I (11.3) II (9.2) II (9.2)
10.4 9.1 8.7
July 2010 January 2013 May 2014
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 1149

further reduced, whereas its DWT is increased, which
represents the optimal design of these vessels. In order
to obtain results more precisely, the engine output data
presented in Lf, as shown later in Eq. (6), on hourly basis
to minimize the gaps of transient power.
According to the IMO 3rd GHG Study, the average Lf
of the ME for 100,000+ DWT large BCs operating at sea
was reduced from 0.77 (2007) to 0.57 (2012), and daily
FOC was reduced from 55.5 t to 42.3 t. Likewise, the
average Lf was also reduced from 0.83 to 0.49 and from
0.69 to 0.32 for 120,000+ DWT tankers and 8,000+ TEU
(twenty-foot equivalent units) container carriers, and
daily FOC was reduced from 65.4 t to 39.4 t and from
200.0 t to 95.6 t, respectively (IMO 2014).
NOx emission certification value and weighting
factor
According to MARPOL Annex VI, the NOx emission
measurement shall apply to each marine diesel engine
with a power output of more than 130 kW installed on a
ship. The amount of NOx emissions obtained from the test
bed involves the emissions mass flow rate calculated using
the carbon balance method. Furthermore, the temperature,
pressure, humidity, and fuel oil composition are considered
when generating the final NOx emission certification value
according to NTC 2008 (IMO 2008d), as expressed in Eqs.
(1) and (2):
NOx emission certification value
¼ ðΣQi  Wf i Þ=ðΣPi  Wf i Þ (1)
Q ¼ Ugas  Cgas  Gmew  Khd (2)
where
Qi : NOx emission mass flow rate at mode i, g/h
Wf i : Weighting factor at mode i (according to the
test cycle in NTC 2008)
the certification values were calculated according to Eqs.
(1) and (2).
The NOx EFs of the MEs and the generator parent
engines of the four BCs at various loads are shown in
Figures 2 and 3, where the WFs assigned by NTC 2008 are
represented by the black dots. The WF exhibits a max-
imum value of 0.50 when the ME operates at a 75% load,
but the NOx EF is not that high. The SFOC at 75% load
level is generally fairly low, as shown in Figure 4. In the
early days, most ships operated at a high speed around
this load level because of high thermal efficiency, and
consumed a large amount of fuel oil at the same time.
However, under the present target of fuel cost reduction
and energy saving, a slow steaming strategy is widely
adopted, and there is an extremely low chance that the
ships will operate at high loads. Consequently, a lower
load results in a higher NOx EF. The EFs of Tier II MEs
running at 25% load are approximately 55% higher com-
pared to the 100% load. In addition, more pronounced
differences can be observed at power levels lower than
25%, where the WF declines to a minimum value of 0.15.
Therefore, this WF allocation and calculation approach
according to NTC 2008 cannot reflect the actual emis-
Figure 2. NOx EFs and WFs of two-stroke ME at different load levels.

Pi : Measured power at each mode, kW

Ugas : Ratio between density of exhaust component
and density of exhaust gas
Cgas : Concentration of the respective component in
the raw exhaust gas, ppm, wet
Gmew : exhaust mass flow, kg/hr, wet
Khd : NOx humidity correction factor
The WFs in NTC 2008 are based on the ISO 8178–4
standards, and were determined using the typical
operating mode and running time of conventional marine
diesel engines (IMO 2008d; ISO 2007). The emission tests
of the diesel engines of the four BCs were performed, and Figure 3. NOx EFs and WFs of four-stroke GE at different load levels.
1150 C.-W. CHENG ET AL.

Lðf2

Probability ðLf1 %Lf %Lf2 Þ ¼ f ðLf ÞdLf (3)

Lf1

ð1
Probability ð0%Lf %1Þ ¼ f ðLf ÞdLf ¼ 1 (4)
0

where
f ðLf Þ): power probability density function,
0 % Lf % 1; Lf : load factor, operating power/rated
power
In Figures 6–9, the area between the two arbitrary
load factors Lf denotes the accumulated probability of
Figure 4. SFOC of the two-stroke ME at different load levels.
occurrence within this load interval. The total area of
summation is 1.
Due to varying ship speeds, draft, sea conditions, and
cargo loading conditions, there is no specific function that
can represent all operating modes (Banks et al. 2013). The
IMO 3rd GHG Study defines an equation showing all of

Figure 5. SFOC of the four-stroke GE at different load levels.

sions of ships operating with a slow steaming strategy. In

contrast, the GEs operate within the load range of 25–75% Figure 6. Accumulated power probability of BC1 ME.
in most cases, and show a relatively uniform WF alloca-
tion in Figure 3. Although the WFs are 0.05 and 0.10 at
100% and 10% loads, respectively, the GEs do not run at
these load levels during actual operations. Figure 5 shows
the SFOC of four-stroke GEs at different load levels with a
high SFOC at low load and a low SFOC at high load
because of the thermal efficiency.

Power probability density function and emission

formulas
The power probability density function (PPDF) of
the MEs and GEs are obtained from actual opera-
tion. The definitions and constraints of the actual
operating PPDF are as follows: Figure 7. Accumulated power probability of BC2 ME.
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION
Figure 8. Accumulated power probability of BC3 ME.
Figure 9. Accumulated power probability of BC4 ME.
these parameters (IMO 2014). By applying this modifica-
tion, the results could be changed. However, the mathe-
matical model introduced in the following provides a
reasonable fit for the calculation of the actual NOx emis-
sions. Thus, the products of the NOx emission function
values and corresponding PPDF values can be summed,
thereby deriving the most practical power probability
emission factor (PPEF), as expressed in Eq. (5). Further,
the summation of the product of the PPEF with the
corresponding power and total operation time will yield
the accurate total NOx emissions, as denoted by Eq. (6):
X
L f ¼0
PPEF ¼ f ðLf Þ  EmðLf Þ (5)
Lf ¼1
Total NOx emissions ¼ P  T
 Σ fLf  f ðLf Þ  EmðLf Þg
(6)
where
1151
EmðLf Þ: function of NOx emission obtained from the
test bed
P: rated power, kW
T: total operation time, hours
Results
Engine load factors
The engine power is expressed as a load factor, and the
accumulated ME power probability distributions of the
four BCs in a year are illustrated in Figures 6–9. The
average Lf ranges from 0.42 to 0.50, and the vessels
operate at even lower loads for long periods. The accu-
mulated probability of an Lf greater than 0.7 is not high,
and there is no probability that the Lf would exceed 0.6
in BC4 during the entire operation. This illustrates the
fact that in order to lower the fuel costs, slow steaming is
practiced in most cases by large BCs operating at sea,
while long-term fast-steaming at over 75% ME load is
rarely seen.
Certification value and emission factor
The WF allocation of the E3 test cycle specified by NTC
2008 and the NOx EF measured under different loads in the
MEs of the four BCs are shown in Figure 2. Using Eqs. (1)
and (2), the BC1 ME has a certification value, 14.6 g/kWh,
far lower than the Tier I limit (17.0 g/kWh), measured NOx
EFs within 13.44–14.93 g/kWh, and a relatively high SFOC,
as shown in Figure 4. Although electrohydraulic control
has been adopted in the BC2 ME, which is only required to
comply with the Tier I limit because it was installed prior to
2011, the relaxed control on the NOx EF to reduce the
SFOC means that BC2 has the highest certification value
and the lowest SFOC among the four, as shown in Figures
2 and 4. The BC3 and BC4 MEs were installed after 2011,
and the respective certification values must be maintained
under the Tier II limit and must meet the market demands
in terms of the SFOC. Therefore, by taking advantage of the
low WF at lower loads, the NOx emission control at lower
loads can be relaxed to achieve a low SFOC. The certifica-
tion values of the MEs of the four BCs were calculated to be
14.60, 15.40, 13.68, and 13.10 g/kWh, respectively. It can be
seen from Figure 2 that, with the exception of BC1, the
other three MEs exhibit a sharp increase in the NOx EF at
25% load.
The WFs allocation of the GEs in the four types of BCs
in the D2 test cycle specified by NTC 2008 and the NOx EFs
measured at different load levels are illustrated in Figure 3.
The BC1 GE exhibits the lowest SFOC, as shown in
Figure 5, under the relaxed control on NOx emissions
because it is only required to comply with the Tier I
1152 C.-W. CHENG ET AL.

regulations. For the BC3 and BC4 GEs, which are required Table 2. NOx PPEF of ME and GE.
to follow the Tier II limit, the trade-off between the SFOC BC1 BC2 BC3 BC4
and EF was examined under different power levels, includ- ME 14.13 16.17 14.99 14.65
GE #1 11.92 10.37 8.92 8.62
ing 10%, 25%, 50%, 75%, and 100%. The NOx emissions GE #2 12.06 10.38 8.90 8.95
control of the BC2 GE has reached the Tier II standards GE #3 12.11 10.72 8.95 8.61
level; however, it only needs to conform to Tier I standards Note. Units: g/kWh.
due to the early GE installation date. As a result, the
emissions control can be slightly relaxed to reduce the Table 3. Total NOx emissions from ME of four BCs.
SFOC. The certification values of the GEs of the four BCs BC1 BC2 BC3 BC4
are 11.80, 10.40, 9.10, and 8.70 g/kWh, respectively. In At sea 643.72 700.03 804.30 579.38
Maneuvering 7.87 10.89 6.76 9.05
Figure 3, except for the BC1 GE, the other three GEs exhibit Total 651.59 710.92 811.06 588.43
a sharp increase in the NOx EF at a 10% load. Note. Units: metric tons.

NOx PPEF and total emissions

The operation hours of the four BCs are shown in
Figure 10. It can be seen that the average operation
hours at sea and at berth account for 64.4% and 32.6%
of the total time, respectively, and the average Lf of the
MEs in the 9,000-hr continuous operation ranges from
0.42 to 0.50, as shown in Figures 6–9. The shipping emis-
sions primarily originate from the ME, whereas the GE is
the major contributor to the emissions at berth. Although
the PPDF of a diesel engine is a random function that
varies from ship to ship, the NOx EFs measured from the
test bed are assumed as a linear function at different loads,
as shown in Figures 2 and 3, and linear interpolation is
used for the other engine loads. Therefore, the corre-
sponding NOx emissions can be calculated once the load
or power output data have been obtained from the diesel
engine control system.
The PPEF calculated using the PPDF of the ME and
GE operation is displayed in Table 2. The total NOx
emissions from the MEs of the four BCs that were derived
both at sea and during maneuvering operations in port
are shown in Table 3.
As shown in Figure 4 and Table 2, the BC1 ME with
conventional mechanical control has a relatively high
SFOC and the smallest PPEF of the four MEs. The BC2
Figure 10. Operation hours of four BCs.
ME has an excessively high PPEF due to the long-time
operation at around 40% load while the EF exhibits its
maximum value, as shown in Figure 2. The BC3 and BC4
MEs exhibit lower SFOCs than the BC1 ME when working
at low loads to satisfy market demands. An optimal balance
is also obtained between the NOx EF and the SFOC. The
NOx emission certification values of these two MEs
obtained from the WF calculation based on the NTC
2008 regulation comply with the Tier II regulations.
However, due to long-term low-power operation, the
results calculated based on the PPDF of the actual opera-
tion eventually exceed the Tier II limit, 14.40 g/kWh, and
reached 14.99 and 14.65 g/kWh, respectively.
Based on the hull dimensions, DWT, and ME rated
power of the four BCs, it is evident that the overall effi-
ciency of these ships has gradually increased from genera-
tion to generation. For example, as compared to BC1, BC4
has increased by 3% in terms of DWT, whereas there has
been a 14% drop in the ME rated power. In terms of NOx
emissions, the total NOx emissions are associated with the
ME power probability, NOx EF, and operation hours.
Generally speaking, electronically controlled two-stroke
diesel engines always possess a higher NOx EF at lower
loads. BC3 ME has the highest operation hours (6397.3 hr)
with an ME power probability mostly ranging from 0.40 to
0.50 load factors, thus leading to the highest total NOx
emissions amounts to 804.30 metric tons (MT) at sea, as
shown in Table 3. The BC4 ME, despite its operation at sea
for 5872.8 hr, has the lowest rated power (16,040 kW),
with the power probability largely falling within 0.38 to
0.46 load factors, and a relatively lower EF, as shown in
Figure 2, which are attributed to the overall propulsion
efficiency and ME enhancements, thereby generating the
least amount of NOx emissions (only 579.38 MT). The
BC1 ME possesses the lowest propulsion efficiency and
highest ME rated power (18,660 kW), the lowest PPEF
(14.13 g/kWh) and operation hours. Thus, BC1 shows a
higher amount of total NOx emissions than that of BC4.
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION
Figure 11. Total NOx emissions from GEs and running hours of SG.
NOx emissions of the GE
The total NOx emissions of the GEs and shaft generator
(SG) running hours are shown in Figure 11; the total
emission is the summation of the emission at berth,
which is represented by horizontally striped bars, and
the emission at sea, which is represented by diagonally
striped bars, and the triangular dots represent the SG
running hours, from which it can be seen that the total
NOx emissions are related to the SG running hours at
sea. BC3 has the highest SG running hours, and a rela-
tively low NOx PPEF. The total NOx emissions from its
three GEs are only 14.03 MT (accounting for 1.7% of the
total emissions of BC3), the lowest among the four BCs.
Despite the use of the SG at sea, due to the highest PPEF,
BC1 still has the highest total NOx emissions from its
three GEs among the four BCs at approximately 33.51
MT, although this only accounts for 4.9% of the total
emissions of BC1. BC2 and BC4 have total NOx emis-
sions of 32.09 MT and 30.27 MT, respectively, with only
short-term or no use of the SG, which accounts for 4.3%
and 4.9% of the total amounts.
Discussion
PPEF and NTC 2008 certification value
The certification value for each ME of the BCs obtained
from the fixed-WF calculation based on the test results
according to NTC 2008, and the PPEF calculated using
the varying ME power probability as the coefficient, are
shown in Figure 12. The weight allocation of the former
has more emphasis at the 75% load level, while that of the
latter is dependent on the actual power probability and
operation hours of the diesel engine. An electronically
controlled Tier II ME would provide a false indication
of low certification values due to the allocation pattern of
1153
Figure 12. Comparison of PPEF and NOx certification values of MEs.
low EF for high load and high EF for low load. Therefore,
certification values cannot correctly represent the emis-
sions per unit work (kWh) when the MEs are under long-
term low-load operation. Instead, the PPEF and total NOx
emissions calculated using the actual power probabilities
of each ship as the variation coefficient can truly represent
the actual emissions of the diesel engine under any load
variations at any time. It can be seen in Figure 12 that the
PPEFs of the electrohydraulically controlled MEs of BC2,
BC3, and BC4 in actual slow steaming conditions are
higher than the certification values, while resutls for
BC1 are just the opposite. This indicates that the NOx
emissions from traditional mechanically controlled MEs
(Tier I engines) are lower than those from electrohydrau-
lically controlled MEs (Tier II engines) under long-time,
low-load operating conditions. Besides, the NOx
emissions of the Tier II engines (BC3 and BC4) were
originally expected to be lower but are actually higher
than 14.40 g/kWh under long-time low-load operating
conditions, and therefore fail to comply with the essence
of the NOx Tier II regulation of MARPOL.
The IMO NOx limits are fairly arbitrary and based
on the given weighting factors; if different weighting
factors were used, then different NOx limits could be
defined by IMO. However, the four BCs’ load profiles
cannot be taken to be representative of global fleet.
Obviously, if the ship engines are operating at low
load they could generate more NOx emissions than
the IMO rated value.
Comparison of total NOx emissions by different EFs
For NOx EFs, the Third IMO GHG Study applied fixed
values of 17.0 and 15.3 g/kWh for Tier I and II slow-speed
diesel engines, respectively (IMO 2014). The GREET
model of the Argonne National Laboratory in the
United States adopted 16.1 g/kWh, whereas the TEAM
model used 8.46 g/kWh (Winebrake, Corbett, and Meyer
2006; 2012). These values were calculated using the WFs
1154 C.-W. CHENG ET AL.
Figure 13. Comparison of total NOx emissions with different EFs.
with an emphasis on diesel engines operated at high
power levels. However, as demonstrated in this study,
this approach cannot provide the accurate emission esti-
mation of diesel engines in low-load operations.
The total NOx emissions from the MEs of the four
BCs were calculated using the published IMO values,
the PPEF mathematical model expressed in Eqs. (5) and
(6), and the NTC 2008 certification values illustrated in
Figure 13. Generally, the certification values of the large
slow-speed diesel engines that conform to the Tier I or
Tier II regulations are around 15.0 g/kWh or 13.0–
14.0 g/kWh, respectively. The total NOx emissions cal-
culated using the fixed values published by IMO,
namely, 17.0 and 15.3, are 3.0–20.3% higher than
those in the PPEF model, as illustrated by the horizon-
tally striped bars in the chart in Figure 13. The diag-
onally striped bars to the extreme right in Figure 13
plot the total NOx emissions calculated based on the
emission certification values of the four MEs under
slow steaming operation. As the EF is usually higher
at lower loads, greater differences can be observed with
lower certification values. For instance, the certification
value of BC4 is 13.10 g/kWh, which leads to an under-
estimation in the total NOx emissions by about 10.4%
as compared to the PPEF model. BC2 and BC3 have
certification values of 15.40 g/kWh and 13.68 g/kWh,
respectively, which result in an underestimation of the
total NOx emissions by 4.0% and 7.9%, respectively, in
comparison to the PPEF model. BC1 has a certification
value of 14.60 g/kWh, which is slightly higher than that
for the PPEF, which was 14.13 g/kWh. Thus, the total
NOx emissions calculated based on the certification
values exhibit a slight overestimation of approximately
3.3%. These discrepancies may be even greater with
longer slow steaming operations.
Emission reduction by limiting the NOx EF at any
load levels
The effect of operation at low loads on the NOx emis-
sions will become more obvious as the ME rated power
increases. Large oil tankers and BCs have similar navi-
gation modes. Based on the IMO statistical data in
2012, BCs and oil tankers with 10,000+ kW rated
power account for 6,181 ships and 62.5% of the total
installation power of these two types. The sum of the
exhaust emissions from these two types of vessels is
higher than that of container ships, and these are there-
fore considered to be the primary source of global
marine pollution (IMO 2014). Hence, the total amount
of emissions can be effectively reduced by controlling
the exhaust emissions from ships by limiting the EF of
the diesel engines at any load level, especially under
slow steaming operation. Taking the four BCs studied
in this work as an example, the EFs that exceed the Tier
I or Tier II limits are uniformly reduced to 17.0 or
14.4 g/kWh, thereby complying with the MARPOL
convention at any load level. As shown in Figure 14,
the BC3 and BC4 MEs are the most commonly used
Tier II electrohydraulically controlled engines at pre-
sent in comparison to those of BC1, which demonstrate
the most effective level of emissions reduction at about
6.9% on average, which is equivalent to a 2,950,000 MT
reduction in NOx emissions per year, according to IMO
statistical data. In addition, the NOx EF of the BC3 and
the BC4 ME may reach 22.23 and 20.14 g/kWh, respec-
tively, at 10% engine load by linear extrapolation. The
emission reduction effect during maneuvering opera-
tions in port is more significant, where the emissions
can be reduced by approximately 31.9%, which is
Figure 14. NOx emission reduction by limiting the NOx EF at
any load levels.
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION
equivalent to a 150,000 MT annual NOx emission
reduction. These reductions will help to significantly
improve the air quality in coastal and port areas, as well
as the health of local residents.
Effect of fuel oil on NOx emissions
The fuel used in the test bed was of ISO DMA grade,
whereas the cargo ships used HFO grade at sea. The use
of HFO by the slow-speed diesel engine increases the
NOx emissions by 6.4% (CARB 2008; IMO 2014). The
heating value and carbon content of the HFO are lower
than those of DMA grade, and the efficiency of diesel
engines in actual navigation is generally lower than that
in the test results obtained from the test bed. Actually,
the SFOC is always 10–15% higher than the test values
(IMO 2014). Therefore, the actual emissions from navi-
gation would be greater than the results (Tables 2 and
3) calculated using the test values. Using the measure-
ment and calculation methods specified in NTC 2008,
Agrawal et al. (2010) and Khan et al. (2013) conducted
on-board measurements on Tier 0 MEs, where the
calculated NOx EFs turned out to be as high as 19.60
and 19.77 g/kWh, respectively. Figure 15 plots the EFs
in comparison to those of BC1 and summarized from
the related studies where the research objects were
composed exclusively of Tier 0 large diesel engines,
and the calculations were performed based on NTC
2008. Generally, Tier 0 diesel engines manufactured
before 2000 were all focused on reducing the SFOC,
although the NOx emissions are relatively high. The
NOx emissions of Tier I diesel engines manufactured
after 2000 and operated at low load levels are lower
than those of Tier 0 engines. If the calculation is con-
ducted based on the related study results, as shown in
Figure 15, the total NOx emissions from ships are likely
to be overestimated.
Figure 15. EFs comparison of BC1 and related studies.
1155
Conclusion
The NOx certification values obtained from the fixed-
WF calculation specified in NTC 2008 do not correctly
represent the emissions per unit work (kWh) when
slow-speed two-stroke diesel engines are under long-
term low-load operation because the WF allocation
places more emphasis on the 75% power level. In this
study, the PPEF and total NOx emissions calculated by
using the proposed mathematical model of the actual
power probabilities of each ship as the variation coeffi-
cient accurately represent the actual emissions of the
diesel engines under any load variation at any point in
time.
The total NOx emissions calculated using the fixed
values published by the IMO, namely, 17.0 and 15.3 for
Tier I and Tier II two-stroke diesel engines, are 3.0 to
20.3% higher than those in the PPEF model, and calcu-
lations that are based on the certification values may
result in an underestimation of up to 10.4%.
The NOx emissions of Tier II engines with
expected lower emissions are actually higher than
14.4 g/kWh under long-time and low-load operating
conditions, as demonstrated in this study using the
actual PPDF and mathematic model, and therefore
fail to comply with the essence of the MARPOL Tier
II regulations.
The mathematical model is useful, but depends
on representative data on the load factors of a
representative sample of the global fleet. Assume
that the NOx EFs that exceed the Tier I or Tier II
limits at any load levels should be limited to 17.0 or
14.4 g/kWh, or even lower, and regardless of
weighting factors, which would allow them to com-
ply with the MARPOL convention under any engine
load conditions. Considering BC3 and BC4 ME load
factor profiles as examples and applying these to
IMO global fleet statistics, this will result in an
effective and significant emission reduction of 6.9%
and 31.9% at the sea and in port areas, respectively,
which would substantially improve the air quality in
coastal and port areas and the health of local resi-
dents. However, the direct use of clean fuels remains
the most effective means to reduce emissions.
The correct calculation of the actual EFs and total emis-
sions by using engine power probabilities as weighting
factors, as demonstrated in this study, is very useful and
applicable to the global shipping emissions inventory for
purposes of life cycle assessment, effective emission reduc-
tion, and policy formulation. Future work will investigate
the total emissions of pollutants and GHGs over the 25-
year life cycle for large-scale ships using the mathematic
model presented in this paper.
1156 C.-W. CHENG ET AL.
Acknowledgments
The authors thank the classification societies and shipping
companies in Taiwan for providing ship navigation and tech-
nical data.
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Conflicts of interest
The authors have no competing interest to declare.
About the authors
Chih-Wen Cheng is a Phd student at the National Taiwan
Ocean University.
Jian Hua and Daw-Shang Hwang are Associate Professors at
National Taiwan Ocean University.
References
Agrawal, H., W. A. Welch, S. Henningsen, J. W. Miller, and
D. R. Cocker III. 2010. Emissions from main propulsion
engine on container ship at sea. Journal of Geophysical
Research 115:1–7. doi:10.1029/2009JD013346.
Andereadis, P., A. Zompanakis, C. Chryssakis, and L.
Kaiktsis. 2011. Effects of the fuel injection parameters on
the performance and emissions formation in a large-bore
marine diesel engine. International Journal of Engine
Research 12:14–29.
Banks, C., O. Turan, A. Incecik, G. Theotokatos, S. Izkan, C.
Shewell, and X. Tian. 2013. Understanding ship operation
profiles with an aim to improve energy efficient operations.
London, UK: Low Carbon Shipping Conference.
Borkowski, T., L. Kasyk, and P. Kowalak. 2011. Assessment of
ship’s engine effective power, fuel consumption and emis-
sion using the vessel speed. Journal of KONES 18:31–9.
CARB. 2008. Emissions estimation methodology for ocean-
going vessels. http://www.arb.ca.gov/regact/2008/fue
logv08/appdfuel.pdf
Corbett, J. J., and P. Fischbeck. 1997. Emissions from ships.
Science 278:823–4. doi:10.1126/science.278.5339.823.
Corbett, J. J., and H. W. Koehler. 2003. Updated emissions
from ocean shipping. Journal of Geophysical Research 108
(D20):1–15. doi:10.1029/2003JD003751.
Classification Societies. 2014. NOx technical files of diesel
engines. Taipei, Tawain, ROC: Classification Societies.
Doudnikoff, M., and R. Lacoste. 2014. Effect of a speed
reduction of containerships in response to higher energy
costs in sulfur emission control areas. Transportation
Research Particle D: Transport and Environment 28:51–
61. doi:10.1016/j.trd.2014.03.002.
Endresen, Ø., E. Sørgård, J. K. Sundet, S. B. Dalsøren, I. S.
Isaksen, T. F. Berglen, and G. Gravir. 2003. Emission from
international sea transportation and environmental
impact. Journal of Geophysical Research-Atmospheres 108
(D17). doi:10.1029/2002JD002898.
European Commission and ENTEC UK Limited. 2002.
Quantification of emissions from ships associated with ship
movements between ports in European community.
Brussels, Belgium: DG ENV.C1, European Commission.
International Organization for Standardization, ISO 8178,
Reciprocating internal combustion engines—Exhaust
emission measurement—Part 4: Steady-state test cycles
for different engine applications.
International Maritime Organization. 2008a. International
Maritime Organization, Conference of Parties to the
International Convention for the Prevention of Pollution
from Ships, 1973, as Modified by the Protocol of 1978
Relating Thereto, MP/CONF. 3/34, 28 October 1997,
Consideration and adoption of the protocol of 1997 to
amend the International Convention for the Prevention
of Pollution from Ships, 1973, as modified by the protocol
of 1978 relating thereto. Retrieved from https://docs.imo.
org/Search.aspx?keywords=%22MP%2FCONF%203%
2F34%22 (accessed August 2017).
International Maritime Organization. 2008b. International
Maritime Organization, Marine Environment Protection
Committee, MEPC 57/21, 7 April 2008, Report of the
Marine Environment Protection Committee on its fifty-
seventh session. Retrieved from https://www.uscg.mil/imo/
mepc/docs/mepc57-report.pdf (accessed August 2017).
International Maritime Organization. 2008c. International
Maritime Organization, Marine Environment Protection
Committee, MEPC 58/23/Add.1, Annex 13,
RESOLUTION MEPC.176(58). Adopted on 10 October
2008, Amendments to the annex of the protocol of 1997
to amend the International Convention for the Prevention
of Pollution from Ships, 1973, as modified by the protocol
of 1978 relating thereto (Revised MARPOL Annex VI).
Retrieved from http://www.gc.noaa.gov/documents/
Annex_VI_Air_Pollution.pdf (accessed August 2017).
International Maritime Organization. 2008d. International
Maritime Organization, Marine Environment Protection
Committee, MEPC 58/23/Add.1, Annex 14, Resolution
MEPC.177(58). Adopted on 10 October 2008, amend-
ments to the technical code on control of emissions of
nitrogen oxides from marine diesel engines. (NOx techni-
cal code 2008). Retrieved from http://www.imo.org/en/
OurWork/Environment/PollutionPrevention/
AirPollution/Documents/Air%20pollution/Resolution%
20MEPC.177(58)%20NOx%20Technical%20Code%
202008.pdf (accessed August 2017).
International Maritime Organization. 2012. International
Maritime Organization, Marine Environment Protection
Committee, MEPC 63/23 Annex 8, 2012 Guidelines on
the method of calculation of the attained energy efficiency
design index (EEDI) for new ships. Retrieved from http://
www.imo.org/en/OurWork/Environment/
PollutionPrevention/AirPollution/Documents/212%2863%
29.pdf (accessed August 2017).
International Maritime Organization. 2014. International
Maritime Organization, Marine Environment Protection
Committee, MEPC 67/INF.3, Reduction of GHG emissions
from ships, Third IMO GHG Study 2014-Final Report.
Retrieved from http://www.imo.org/en/OurWork/
Environment/PollutionPrevention/A irPollution/
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION
Documents/MEPC%2067-INF.3%20-%20Third%20IMO%
20GHG%20Study%202014%20-%20Final%20Report%20
(Secretariat).pdf (accessed August 2017).
International Maritime Organization. 2016. International
Maritime Organization, Marine Environment Protection
Committee, MEPC 69/19/1, Development of guidelines
for the use of more than one engine operational profile,
submitted by Norway. Retrieved from https://docs.imo.
org/Search.aspx?keywords=%22MEPC%2069-19-1%22
(accessed August 2017).
Khan, M. Y., S. Ranganathan, H. Agrawal, W. A. Welch,
C. Laroo, J. W. Miller, and D. R. Cocker III. 2013.
Measuring in-use ship emissions with international
and U.S. federal methods. Journal of the Air & Waste
1157
Management Association 63:284–91. doi:10.1080/
10962247.2012.744370.
Notteboom, T. E., and B. Vernimmen. 2009. The effect of
high fuel costs on liner service configuration in container
shipping. Journal of Transport Geography 17 (5):325–37.
doi:10.1016/j.jtrangeo.2008.05.003.
Winebrake, J. J., J. J. Corbett, and P. E. Meyer. 2006. Total
fuel-cycle emissions for marine vessels: A well-to-hull ana-
lysis with case study. 13th CIRP International Conference
on Life Cycle Engineering, Belgium, 1 June, LCE2006.
Winebrake, J. J., J. J. Corbett, and P. E. Meyer. 2012.
Energy use and emissions from marine vessels: A total
fuel life cycle approach. Journal of the Air & Waste
Management Association 57:102–10. doi:10.1080/
10473289.2007.10465301.