Environmental Pollution 231 (2017) 1370e1379

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Assessment of shipping emissions on four ports of Portugal*

R.A.O. Nunes, M.C.M. Alvim-Ferraz, F.G. Martins, S.I.V. Sousa*
LEPABE e Laboratory for Process Engineering, Environment, Biotechnology and Energy, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias,
4200-465, Porto, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: In the last few years, ship emissions have attracted growing attention in the scientific community. The
Received 15 February 2017 main reason is the constant increase of marine emissions over the last twenty years due to the inten-
Received in revised form sification of port traffic. Thus, this study aimed to evaluate ship emissions (PM10, PM2.5, NOx, SO2, CO, CO2,
29 August 2017
N2O CH4, NMVOC, and HC) through the activity-based methodology in four of the main ports of Portugal
Accepted 30 August 2017
~es, Setúbal, Sines and Viana do Castelo) during 2013 and 2014. The analysis was performed ac-
(Leixo
Available online 13 September 2017
cording to ship types (bulk carrier, container, general cargo, passenger, Ro-Ro cargo, tanker and others)
and operational modes (manoeuvring, hotelling and during cruising). Results indicated that tankers were
Keywords:
Portuguese ports
the largest emitters in two of the four analysed ports. Regarding cruising emissions, container ships were
Activity-based method the largest emitters. . CO2, NOx and SO2 estimated emissions represented more than 95% of the cruising
Air pollution and in-port emissions. Results were also compared with the total national emissions reported by the
Shipping emissions Portuguese Environment Agency, and if the in-port emissions estimated in the present study would have
Climate change been taken into account to these totals, emissions of NOx and SO2 would increase 15% and 24% in 2013
and 16% and 28% in 2014. Summing up ships seem to be an important source of air pollution, mainly
regarding NOx and SO2.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction in 2016, 21 million tonnes were exported by sea, corresponding to

Over the last twenty years the increase of in-port traffic has
made shipping one of the most polluting industries in the world. In
the past, the environmental and health risks associated with
shipping were underestimated but nowadays these areas are
attracting the attention of the scientific community. Maritime
transport is considered a highly efficient mode of cargo trans-
portation from an energetic point of view (capacity of tons trans-
ported per kilometre travelled). According to the United Nations
Conference on Trade and Development (UNCTAD), shipping is
responsible for around 80% of global trade by volume and over 70%
of global trade by value worldwide (IMO, 2016a). Although mari-
time transport is an effective transport option, emissions from ships
contribute significantly to the total emissions in global transport
(Eyring et al., 2009; Saraçoǧlu et al., 2013; Yau et al., 2012).
In Portugal, according to the Report of the State of the Envi-
ronment 2016 made by the Portuguese Environment Agency (APA)
*
This paper has been recommended for acceptance by Charles Wong.
* Corresponding author.
E-mail address: sofia.sousa@fe.up.pt (S.I.V. Sousa).
53.6% of the total that was exported by the country. On the other
hand, maritime transport has also been responsible for 36.3 million
tonnes of imported goods, corresponding to 61.7% of all imported
goods at a national level (APA, 2016a). Portugal is a country with
942 km of coastline with a high activity in the maritime sector and
densely populated in coastal areas, which makes the accurate
quantification of emissions from ships that navigate and berth in
Portuguese ports very important.
Recent studies show that at least 70% of emissions from ships in
international routes occur within 400 km of the coast (Eyring et al.,
2009). Ships’ emissions can be easily transferred hundreds of kil-
ometres towards the mainland and their impacts can be evident
both on local and global scales, which may represent a significant
risk to human health (Corbett et al., 2007; Eyring et al., 2009). In
fact, the emissions from ships may cause adverse effects on air
quality during hotelling or manoeuvring and in transiting along the
coast (Kilic and Deniz, 2010).
Emissions during hotelling and manoeuvring represent only a
small part of the total emissions from shipping, however port areas
are recognizable receptors and inevitably point sources of
concentrated shipping emissions (Tzannatos, 2010).
The main pollutants emitted from ships are particulate matter

http://dx.doi.org/10.1016/j.envpol.2017.08.112
0269-7491/© 2017 Elsevier Ltd. All rights reserved.
 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379
(PM), volatile organic compounds (VOC), carbon monoxide (CO),
nitrogen oxides (NOx), sulphur oxides (SOx) and carbon dioxide
(CO2). Particular concern is related to PM, VOC CO, NOx and SOx due
to their potential public health impacts, NOx and SOx due to their
contribution for rain acidification, and CO2 due to its role on
greenhouse effect (Matthias et al., 2010). Recent studies attribute to
ships around 2.2%, 15%, and 5e8% of the global anthropogenic CO2,
NOx and SOx levels, respectively (IMO, 2016b; Song, 2014;
Tzannatos, 2010). The international nature of the maritime sector
makes it difficult and complex to design and implement policies to
decrease emissions, however, the monitoring and control of air
pollution is a recognized part of the essential measures to minimize
the impacts of port activities (Sanabra et al., 2014; Tzannatos, 2010).
Shipping pollution is regulated by the International Maritime Or-
ganization (IMO) under the International Convention for the Pre-
vention of Pollution from Ships (MARPOL), which is the most
important international convention concerning the prevention of
pollution of the marine environment by ships both for operational
and accidental reasons. Currently, MARPOL is divided into 6 An-
nexes (IMO, 2016b), and according to Annex VI “Prevention of Air
Pollution from Ships” (in force since May 2005), IMO develops and
promulgates international regulations for atmospheric emissions
from ships. Since July 2010 and after the revision of Annex VI, IMO
established limits for SOx and NOx emissions, prohibited emissions
of stratospheric ozone depleting substances and created emission
control areas (ECAs). Currently, the global limit for sulphur content
of ships' fuel is 3.50% m/m, but IMO established a limit of 0.50% m/
m to be applied from January 2020 or January 2025, depending on
the availability of low sulphur fuel for ships’ use. Since January 2015
the sulphur limit for fuel used by ships in SOx Emission Control
Areas (SECAs) has been 0.10% m/m (IMO, 2016b). In Europe, pas-
senger ships that sail outside SECAs have to respect a limit of 1.5%
m/m sulphur in fuels (EEA, 2013). IMO regulates the NOx emissions
through a three tier scheme (Tier I, II and III). The Tier III, currently
in force, applies to new ships (built after January 2016) operating in
North American and U.S. Caribbean NOx emissions control areas
(NECAs) (IMO, 2016b). In 2021 the North Sea and the Baltic Sea will
become NECAs and Tier III will be mandatory for ships built after
January 2021 operating in those areas too (IMO, 2016b).
Significant progress has been made for estimating ship emis-
sions in different regions of the world. In the last few years several
port inventory studies have been conducted. Tzannatos (2010)
estimated the emission of the main pollutants from cruise ships
for the Port of Piraeus in Greece, between 2008 and 2009, using an
in-port ship activity-based methodology. Results indicated that
overall cruise ship emissions in Piraeus Port reached 2600 tonnes
annually. Deniz and Kilic (2009) calculated emissions from ships at
Ambarli Port in 2005, estimating annual emissions of NOx, SO2, CO,
CO2, VOC and PM of, respectively, 845 ty-1, 242 ty-1, 2127 ty-1, 78 590
ty-1, 504 ty-1 and 36 ty-1. They also reported that the maximum
values of SO2 and NOx concentrations modelled in a 2 km range
from the port exceeded 100 mg m
3 of NOx and 55 mg m
3 of SO2,
affecting around 60 000 people. Berechman and Tseng (2012)
conducted an emissions inventory at the Port of Kaohsiung in
Taiwan to estimate the associated emission costs of ships and
trucks that operate in that port during 2010. Using a bottom-up
methodology, authors found that tankers, containerships and
bulk ships were the major contributors to ship emissions. Liu et al.
(2014) also estimated emissions of SO2 at Port of Kaohsiung from
2006 to 2010, concluding that estimating methods based on cargo
capacity led to higher emissions than those based on vessel activity
time. Saraçoǧlu et al. (2013) studied ship emissions at Izmir Port in
Turkey for 2007 using a ship activity-based methodology. The
emissions of NOx, SO2, CO2, hydrocarbons (HC) and PM were,
respectively, 1923 ty-1, 1405 ty-1, 82 753 ty-1, 74 ty-1 and 165 ty-1
1371
and, according to the results, ships calling at the Izmir Port were an
important air pollution source towards Izmir city and its sur-
roundings. Song (2014) reported a ship emissions study of CO2,
methane (CH4), nitrous oxide (N2O), PM10, PM2.5, NOx, SOx, CO and
HC at Shanghai Yangshan port in China, investigating both in-port
ship emissions and the associated social costs during 2009. The
authors used an activity-based methodology, supported by data
from the automatic identification system (AIS). The emissions of
CO2, CH4, N2O, PM10, PM2.5, NOx, SOx, CO and HC were, respectively,
578 444 ty-1, 10 ty-1, 33 ty-1, 1078 ty-1, 859 ty-1, 10 758 ty-1, 5623 ty-1,
1136 ty-1 and 519 ty-1. Song and Shon (2014) conducted a study
during three different years (2006, 2008 and 2009) to estimate
emissions from ships at Busan Port in Korea. The largest emissions
were estimated during “in port” mode and for container ships.
Authors also estimated future ship emissions in 2020 and 2050 and
predicted a gradual increase of the emissions, except for SO2 and
PM. Tichavska and Tovar (2015) used a model called STEAM (Ship
Traffic Emission Assessment Model) developed by Jalkanen et al.
(2009, 2012) to estimate exhaust pollutants related with cruise
and ferry operations in an island context (Las Palmas Port, Spain)
during 2011. The authors highlighted that cruise ships were a sig-
nificant source of air pollution at Las Palmas Port. Saxe and Larsen
(2004), Maragkogianni and Papaefthimiou (2015) and Sanabra et al.
(2014) considered port clusters for their studies. Saxe and Larsen
(2004) developed a model to calculate the dispersion of air pol-
lutants (NOx, SO2 and PM) from ships in three Danish ports in 2001.
The authors concluded that in-port emissions of NOx possibly
induce health problems to the population near the Danish ports
and recommended monitoring air pollutants to suggest methods to
reduce air pollution from ships. Maragkogianni and Papaefthimiou
(2015) estimated emissions from cruise ships in the five busiest
Greek ports for 2013 and concluded that seasonality plays a major
role. In summer the emissions and associated impacts were
significantly amplified. Sanabra et al. (2014) assessed the local
impact of ship air pollution for the Spanish ports network in 2009.
Results indicated that the main emitted pollutant was NOx, repre-
senting 86% of total emissions and that PM2.5 and NOX were the
pollutants that presented higher externalities.
In Portugal the Commission for the Coordination and Develop-
ment of the Northern Region (CCDR-N) reported a study for the
North of Portugal in 2008, which included the emissions of CO2,
SO2, NOx, VOCs and PM from maritime transport produced at
Leixo~es Port. However, this study only considered the emissions
during manoeuvring, hotelling and loading/unloading (the latter
applied to tankers) (CCDR-N, 2014) and was performed with a top-
down approach. In 2016 the Portuguese Environment Agency (APA)
published the “Portuguese national inventory report on green-
house gases, 1990e2014” which included an inventory of atmo-
spheric emissions from domestic ship traffic between Portuguese
ports for the period between 1990 and 2014 (APA, 2016b). However,
interpolations of values for the activity of ships were made for the
periods of 1991e1994 and 1996 to 1999 and only CO2, CH4 and N2O
emissions for domestic ship transport between Portuguese ports in
mainland territory were considered. Thus, despite the work already
performed, it is important to conduct a more detailed inventory of
atmospheric emissions from ships in Portuguese ports to evaluate
the real impact of this transport mode on air quality and health. As
far as known, no previous paper concerning ship emissions has
been published with reference to ports in Portugal. Thus, the pre-
sent study aimed to estimate PM10, PM2.5, NOx, SO2, CO, CO2, N2O
CH4, NMVOC, and HC in-port (manoeuvring and hotelling) and
cruising emissions in four of the main ports of Portugal using a
Bottom-up methodology based on the movement and activity of
ships. To estimate emissions during cruising a buffer area was
drawn, which allowed to determine the distance between ports up
1372 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379
to 400 km from the Portuguese coast. Emissions during cruising
were estimated only for the ships that docked in the ports analysed
and for the part of the routes inside the buffer area. As far as known,
AIS data seem to be the most reliable and accurate approach to
describe ships activities (Nunes et al., 2017). Unfortunately, after
several attempts to get AIS data for the Portugal coastline, namely
through several Portuguese official organisations (without success)
and commercial databases (no funding available), it was decided to
perform the study even without AIS data, because in Portugal there
are no publications concerning international shipping emissions,
thus, and although AIS was not used, this study is a very important
contribution to the existing literature fulfilling the existing lack in
this field of knowledge.
2. Material and methods
2.1. Study area and ports characteristics
Portugal is located in southwestern Europe and due to its Atlantic
centrality has a high geostrategic importance in the connection of
maritime traffic from the Americas and Africa to Europe. Fig. S1 (in
Supplementary Material e SM1) shows a shipping traffic density
map for 2015 that allows understanding the geostrategic importance
of Portugal in the international maritime traffic.
According to the Association of Portugal Ports, national ports
have essentially a commercial function (APP, 2016). Fig. 1 shows the
location of the studied ports (Leixo ~es, Setúbal, Sines and Viana do
Castelo) and the buffer area with 400 km from the Portuguese coast
line.
Leixo~es and Viana do Castelo ports are managed by the same
administration (APDL - Douro Ports Administration, Leixo ~es and
Viana do Castelo, S.A.). Leixo~es is the largest port infrastructure in
the Northern Region of Portugal and one of the most important in
the country. It represents 25% of Portuguese international trade and
moves about 14 million tonnes of goods per year. Viana do Castelo
Port is located near the mouth of Lima River, incorporating a
commercial port, an industrial port, a marina and a fishing port. The
commercial port can handle about 900 000 tonnes per year; the
industrial port consists of two plants: the shipyards of Viana do
Castelo (ENVC) and a component manufacturing industry for wind
turbines (ENERCON). The port is responsible for handling fractional
general cargo, dry bulk, liquid bulk and Ro-Ro cargo. APDL has
annual incomes of about 10 million V.
Setúbal Port is an Iberian solution in the Lisbon Region. Located
at the junction of major routes of intercontinental navigation
North-South and East-West, it has a direct connection to the main
road and rail routes in Portugal and Spain. This port is a leader in
fractional general cargo transportation, with about 43% of the na-
tional total and is also leader in the Ro-Ro traffic of the new light
vehicles, with around 90% of the national total (annual incomes of
about 3.5 million V).
Sines Port is the largest artificial port in Portugal, and a deep
water port. This port has specialized terminals that allow move-
ment of different types of goods. It is the main port on the Atlantic
coast of Portugal due to its geophysical characteristics and the main
power supply input port in Portugal (natural gas, coal, oil and oil
products). It is the Portuguese port with the highest annual in-
comes of around 12 million V (APP, 2016).
2.2. Calculation method
The methodology quoted as Tier 3 in the European Monitoring
and Evaluation Programme/European Environment Agency (EMEP/
EEA) air pollutant emission inventory guidebook 2016 was used for
the estimation of air polluting emissions from ships in this study.
This methodology also known as “activity-based” method or
“bottom-up” approach is based on the ship movement information
and involves the application of emission factors to a particular ship
activity. The emission factors are used to relate the emitted quantity
of a certain pollutant with the energy spent by the ship’ engines
(considering the specifications of operation) during a certain ac-
tivity. The activities here considered were: i) manoeuvring -
considered the average time when the ship was inside the port in
manoeuvre operations (given by port authorities); ii) hotelling -
considered all the time the ship was inside the port, with exception
of manoeuvring operations; and iii) cruising e considered when
the ship was at sea (up to 400 km from the Portuguese coast
accounted from the point of entrance/exit of the port). This last
operational mode only considers ships’ routes that entered/exited
the four ports here analysed.
For every ship call, the amount of air pollutants (PM10, PM2.5,
NOx, SO2, CO, CO2, N2O CH4, NMVOC, and HC) produced during
cruising up to 400 km from the Portuguese coast, manoeuvring and
hotelling were estimated using the following expressions,
respectively:
D
ECruising ðtonÞ ¼  ðME  LFME  ME EF þ AE  LFAE  EFÞ
V
(1)
EManoeuv ðtonÞ ¼ TManoeuv ðME  LFME  EF þ AE  LFAE  EF
þ AB  EFÞ
(2)
EHotell ðtonÞ ¼ THotell ðME LFME EF þ AE  LFAE EF þ AB  EFÞ
(3)
where ME is the maximum main engine power (kW), AE is the
auxiliary engine power (kW), AB is the auxiliary boiler energy
default, V is the ship average speed for each type of ship (km/h), D is
the distance between port up to 400 km from the Portuguese coast
line (km), LFME is the load factor of ME at cruising, manoeuvring
and hotelling (%), respectively for Equations (1)e(3), LFAE is the
load factor of AE at cruising, manoeuvring and hotelling (%),
respectively for Equations (1)e(3), EF is the emission factor for
different engine types/fuel combinations and operational modes
(cruising, manoeuvring and hotelling) (g/kWh), respectively for
Equations (1)e(3), TManoeuv is the average time spent during
manoeuvring (h) and THotell is the average time spent at berth (h).
2.3. Data requirements and data sources
The survey information and operational data, such as ships
names, arrival and departure information of each ship (used to
calculate the hoteling times) and manoeuvring times, as well as the
names/identification numbers, date of construction, gross weight
and gross tonnage (GT) of ships were obtained from port author-
ities. For Viana do Castelo port, it was not possible to obtain the
arrival and departure information of ships because port authorities
were not able to compile this information. Thus ships calling Viana
do Castelo port were not considered in the estimation of cruising
emissions. In this study seven primary categories of ships were
considered: i) bulk carrier; ii) container; iii) general cargo; iv)
passenger; v) Ro-Ro cargo; vi) tanker; and vii) others. The type of
each ship provided by the port authorities were assigned into the
above described seven primary categories according to Entec
(2010) study. In the “others” category, war vessels, tugs, fishing
boats, dredgers and unknown ships were considered.
 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379
Fig. 1. Location of the four studied ports and the buffer area with 400 km from the Portuguese coast line (Grey line).
2.3.1. Engine powers and load factors
Since it was not possible to obtain the engine powers by another
method, default tables were used to obtain the maximum engine
powers of main (ME) and auxiliary engines (AE) and auxiliary
boilers (AB), and the load factors for the engines for different
operating modes. According to the information present in the
EMEP/EEA Guidebook 2016, the ME maximum powers can be ob-
tained as a function of GT. Thus, ME maximum powers were ob-
tained according to Table S1 (Supplementary Material e SM2) and
used as the maximum continuous power of the propulsion engine
(MCR) with reference to the World fleet of 2010. For ships where GT
was not available in the data provided by Port Authorities, data
from Marine Traffic website was extracted (Marine Traffic, 2016).
The AE rated powers were obtained posteriorly from ME to AE ra-
tios by ship type developed by Trozzi et al. (2010) also with refer-
ence to 2010 world fleet (Table S2 e Supplementary Material SM2).
It should be noted that, the world fleet characterization was based
on an analysis of the available literature on the subject and on
statistical analysis of Lloyd's database for 1999 and 2010 in the
frame of MEET project and in the update of the EMEP/EEA Guide-
book. The AB energy defaults were obtained from the ICF (2009),
Starcrest Consulting Group (2014, 2015) reports and are resumed
in Table S3 (Supplementary Material SM2). It was assumed that
1373
boilers were not used during cruising. As far as known, AB are
usually used during cruising when ME loads are less than 20%,
however in this study a default load factor of 80% was assumed
because it was not possible to estimate ME load factors through
Propeller Law.
Consequently, since domestic literature regarding the activity of
ship engine loads in the Portuguese Coast and Ports is deficient, the
assumptions made by Entec (2002) were adopted according to
Table S4 (additional details of load factors are included as Supple-
mentary Information e SM3).
2.3.2. Cruising time and duration of in-port activities
Since it is not possible obtain the actual speeds for ships under
this study, average speeds by type of ship from Entec (2002) were
adopted according to Table S5 (Supplementary Material SM3). In
order to calculate cruising times from port to overseas-port dis-
tances up to 400 km from the Portuguese coast line, Dataloy Dis-
tance Table (DDT) online calculator (Dataloy Systems, 2016) was
used for each of the distinct international origin/destination com-
binations. According to the study performed by Fitzgerald et al.
(2011), DDT was considered one of the most accurate software to
calculate international port-to-port distances, because it considers
a great number of waypoints and logical relations between them.
1374 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379

As above referred, a buffer area with 400 km from the Portuguese 3. Results and discussion
coast line was drawn and for each ship call only the track inside this
area (using the Google Earth software) was considered and the 3.1. In-port emissions
ship-travelled distance corresponded to the part of the routes in-
side this area (Fig. 1). To avoid duplication of the routes, half of the Emissions of PM10, PM2.5, NOx, SO2, CO, CO2, N2O CH4, NMVOC,
travelled distances was considered for the ships that travelled be- and HC during manoeuvring and hotelling were estimated for the
tween the studied ports. The cruising time of each trip was calcu- selected ship categories and ports during 2013 and 2014. Table 1
lated by the quotient between the distance travelled inside of the summarizes the total amounts of emitted air pollutants as well
buffer area and the speed of each ship. Hotelling times were as, the number of ship calls at each port during 2013 and 2014. In
calculated by the difference between arrival and departure times terms of in-port emissions, the ships calling at Sines port during
(considering the first and the last berthing locations). Average 2014 have produced the higher emission amounts (3.4Eþ05 ton
manoeuvring times for each port were obtained from Port Au- y
1) followed by ships at Setúbal port (3.4Eþ05 ton y
1), at Leixo ~es
thorities. It should be noted that all values mentioned above were port (3.4Eþ05 ton y
1), and finally at Viana do Castelo port
verified through interviews with experienced members of the (8.9Eþ03 ton y
1). It should be noted that emissions were higher
shipping sector and by port authorities’ responsible. during 2014 than during 2013, except at Leixo ~es where they were
similar for both years. Nevertheless, when considering total GT of
the ships calling each port, Sines was the one with the lowest
emissions per total GT, followed by Leixo ~es, Viana do Castelo and
Setúbal. As seen from Table 1, number of ships at Leixo ~es and Sines
2.3.3. Engine/fuel type profiles, fuel sulphur content and emission
were significantly higher than at the other ports. The total number
factors
of ships at Leixo~es port was the highest (2641 ship calls) in 2014.
Emission factors (EF) are largely dependent on engine/fuel type
The port of Viana do Castelo was the only one where the total
profiles and fuel sulphur content. For this study engine/fuel type
number of ships was higher in 2013 than in 2014.
profiles have been obtained from the LMIS dataset from Entec
Regarding emissions by type of pollutant, those of CO2 were
(2010) study which considered 14255 ships (Entec, 2010). The
found to be dominant (more than 93% of total), followed by NOx
classification of ME was made based on GT of the respective ship
emissions (3.0e4.2% of total), SO2 emissions (1.5e1.6%) and PM10
types according to Table S6 (Supplementary Material SM4). For AE,
and PM2.5 emissions (0.2% of total) for all analysed ports during the
it was assumed that all ships' categories had MSD or HSD engines
studied period. Emissions of NMVOCs, CH4 and N2O were the
without distinction (M/H SD) (See Supplementary Material e SM4).
lowest corresponding to <0.1% of total estimated emissions.
Ships were assumed to use three fuel types: i) Residual Oil (RO)
In general, the categories that registered more calls in 2013 and
with a sulphur content of 2.7% by mass; ii) Marine Diesel Oil (MDO)
2014 were general cargo, container ships and tankers. Leixo ~es port
with a sulphur content of 1.0% by mass; and iii) Marine Gas Oil
was the only one that registered a considerable number of calls of
(MGO) with a sulphur content of 0.5% by mass. Fuel sulphur con-
passenger ships. At Setúbal, the number of calls of general cargo
tents have been based on data provided to the International
was higher than 50% of the total. Furthermore, both in 2013 and
Maritime Organisation's Marine Environment Protection Commit-
2014 Setúbal port was the one that registered more calls of Ro-Ro
tee (IMO MEPC) (2001) according to Entec (2002). The fuel types
cargo and bulk carriers. This was expected because Setúbal port is
used by ships (for ME and AE) in the study area of this work were
the main port in Portugal in the Ro-Ro traffic of new light vehicles.
assigned according to ship type based on information from the
At Sines, container ships and tankers were responsible for the
Entec (2010) (Table S7, Supplementary Material SM4). In this study
majority of calls, which seems to be related with this port being the
it was assumed that all auxiliary boilers (AB) used RO.
main receptor of goods from the energy sector (natural gas, coal, oil
Since that at the moment there are no locally derived emission
and oil products). Viana do Castelo port registered the lowest
factors for the Portuguese coast, EF used in this study were taken
number of ship calls. In this port general cargo ships were
from other overseas studies (EMEP/EEA, 2016a; Entec, 2002; ICF,
responsible for more than 80% of the calls in both years. These
2009; IVL, 2004). The EF of SO2, CO2, HC and the specific fuel con-
differences in terms of calls by ship type suggest that pollutant
sumption (SFC) for main and auxiliary engines were taken from the
emissions may be different in each port. In fact, at Leixo ~es and
Entec 2002 study (Entec, 2002), while those of CO, CH4 and N2O
Setúbal, ship calls increased 1.9% and 12.1% from 2013 to 2014
were derived from the study performed by IVL Swedish Environ-
respectively, which lead to an increase of 3.0% and 13.1% in total
mental Research Institute (IVL, 2004). EF for PM10 were taken from
pollutant emissions. At Sines although the number of ship calls was
the ICF (2009) study based on a formulae that is function of fuel
similar in 2013 and 2014, total emissions increased 8.8%. This in-
type and sulphur content. PM2.5 EF were obtained from a PM10 to
crease seems to be related to the large number of containers ships
PM2.5 conversion factor of 0.92 (ICF, 2009). From EMEP/EEA 2016a,b
that berthed in this port during 2014 (containers ship’ calls
study EF of NMOC and NOX were assumed. Three different EF for
increased 4.8%). At Viana do Castelo, although port ship calls
NOX emissions were assumed according to the construction date of
decreased 9% total emissions increase 2.8%. This increase seems to
the ships. According to EMEP/EEA 2016a,b report, NOX EF for 2000
be related with the higher ME and AE rated powers that were
(Entec, 2002) are representative of the fleet before application of
higher for the ships that berth in Viana do Castelo during 2014
IMO NOx Technical Code, while 2005 and 2010 values (according to
which led to higher emissions.
Entec, 2007) were obtained from the year 2000 NOx EF with a
Fig. 2 shows in-port ship emissions by ship types for 2014 at: a)
reduction of 3.4% and 6.8% to account for the new engines intro-
Leixo~es Port, b) Setúbal Port, c) Sines Port and d) Viana do Castelo
duced by 2005 and 2010, respectively. EF for the different pollut-
Port. The same information for 2013 can be found in Fig. S2 (Sup-
ants were assigned according to engine type (ME, AE and AB),
plementary Material SM6). Regarding emissions from different ship
operational modes (cruising, manoeuvring and hotelling) and fuel
types, at Leixo~es emissions from container ships and tankers were
type (RO, MDO and MGO) (additional details of EF is included as
similar and corresponded to around 30% of estimated emissions.
Supplementary Material SM5 Table S8).
 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379 1375

Table 1
In-port Emissions of air pollutants and ship calls by port during 2013 and 2014 (in t y
1).

Port Year NOx SO2 PM10 PM2.5 HC NMVOC CO CO2 CH4 N2O Total Calls
~es
Leixo 2013 8.1Eþ03 3.3Eþ03 3.5Eþ02 3.2Eþ02 2.6Eþ02 1.0Eþ02 3.1Eþ02 1.9Eþ05 2.7 9.9 2.1Eþ05 2564
2014 8.2Eþ03 3.3Eþ03 3.6Eþ02 3.3Eþ02 2.7Eþ02 1.1Eþ02 3.1Eþ02 2.0Eþ05 2.8 1.0Eþ01 2.1Eþ05 2612
Setúbal 2013 9.8Eþ03 4.2Eþ03 4.7Eþ02 4.3Eþ02 3.8Eþ02 1.6Eþ02 3.7Eþ02 2.6Eþ05 3.8 1.3Eþ01 2.8Eþ05 1350
2014 1.1Eþ04 4.9Eþ03 5.5Eþ02 5.1Eþ02 4.4Eþ02 1.8Eþ02 4.2Eþ02 2.9Eþ05 4.4 1.5Eþ01 3.1Eþ05 1514
Sines 2013 9.4Eþ03 5.0Eþ03 5.6Eþ02 5.2Eþ02 4.4Eþ02 1.8Eþ02 3.9Eþ02 3.0Eþ05 4.5 1.5Eþ01 3.1Eþ05 1982
2014 1.0Eþ04 5.5Eþ03 6.1Eþ02 5.6Eþ02 4.9Eþ02 2.0Eþ02 4.3Eþ02 3.2Eþ05 4.9 1.6Eþ01 3.4Eþ05 1975
Viana do Castelo 2013 3.6Eþ02 1.4Eþ02 1.5Eþ01 1.4Eþ01 1.1Eþ01 4.1 1.4Eþ01 8.1Eþ03 1.1E-01 4.1E-01 8.6Eþ03 199
2014 3.5Eþ02 1.4Eþ02 1.6Eþ01 1.4Eþ01 1.2Eþ01 4.6 1.4Eþ01 8.3Eþ03 1.1E-01 4.1E-01 8.9Eþ03 181

~es Port, b) Setúbal Port, c) Sines Port and d) Viana do Castelo Port.
Fig. 2. In-port ship emissions according to ship types in 2014 at: a) Leixo

Although the number of calls from tankers was lower (around 18%
of the total) than calls from container ships (more than 40% of the
total), tankers emitted higher amounts of pollutants. At Setúbal
port emissions from tankers represented more than 50% of total
emissions, nevertheless this type of ship only accounted with
around 11e12% of total ship calls. General cargo ships recorded the
largest number of calls (more than 50% of the total), but only
accounted for 18% of total emissions. Both in Leixo~es and Setúbal,
this seems related to the higher hotelling times and AE loads,
during this mode of operation for tankers. At Sines container ships
1376 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379

emitted the highest amount of pollutants in 2013 and 2014 with
more than 50% of total emissions. This result seems to be related
with the high engine powers of container ships. Yau et al. (2012), Ng
et al. (2013), Saraçoǧlu et al. (2013) and Song and Shon (2014)
concluded the same in their studies. At Viana do Castelo general
cargo ships were the highest emitters responsible for more than
75% of in-port emissions. For all the analysed ports and years,
emissions during manoeuvring were lower compared to emissions
during hotelling, nevertheless these were important to the total
amounts of in-port emissions. This information can be found in
Tables S9eS16 (Supplementary Material SM7).
3.2. Cruising emissions
Emissions of PM10, PM2.5, NOx, SO2, CO, CO2, N2O CH4, NMVOC,
and HC for cruising up to 400 km from the Portuguese coast for the
ships that berthed in the four ports analysed in this study during
2013 and 2014 were estimated. Table 2 summarizes the amounts of
emitted air pollutants by ship type as well as, the number of ships
for each category during 2013 and 2014. For cruising, container
ships were the largest emitters responsible for 8.2Eþ05 ton y
1 and
9.1Eþ05 ton y
1 which corresponded to 59% and 65% of total
emissions in 2013 and 2014, respectively. This results were related
with the higher distances travelled at sea up to 400 km from the
Portuguese coast for container ships, which led to higher times of
permanency in open sea. As seen from Table 2, and similar to in-
port emissions, CO2 emissions were found to be dominant (95% of
total in 2013 and 2014), followed by NOx emissions (3% of total) and
SO2 emissions (2%). Emissions of PM10, PM2.5, NMVOCs, CH4 and
N2O represent less than 1% of the total estimated cruising
emissions.
3.3. Contribution of the calculated ship emissions to the national
emissions
The in-port and cruising estimated emissions of NOx, SO2, CO2
and NMVOC for 2013 and 2014 were compared with the total na-
tional emissions reported on the Portuguese National Inventory
Report on Greenhouse Gases, 1990 e 2014 performed by APA
submitted under the United Nations Framework Convention on
Climate Change (UNFCCC) and the Kyoto Protocol (APA, 2016b). This
report is the official annual accounting of all anthropogenic (hu-
man-induced) emissions of greenhouse gases and other gases
species such as NOx, SO2 and NMVOC in Portugal. In this report,
emissions were estimated taking into account five large sectors
(Energy, Industrial Processes and Product Uses, Agriculture, Land
Use, Land-Use Change and Forestry (LULUCF) and Waste) and were
calculated using the Guidelines from Intergovernmental Panel on
Climate Change (IPCC) whenever possible. CO2 ship emissions were
accounted in this report in the Energy sector as domestic ship traffic
between Portuguese ports. Although this assessment is very
important, it underestimates the total emissions from shipping in
Portugal, because it does not include emissions from international
navigation. In order to assess the contribution of the ship emissions
calculated in this study to the total of national emissions above
referred, CO2 ship emissions calculated by APA for 2013 and 2014
were subtracted from the total of emissions. For NOx, SO2 and
NMVOC emissions, the totals of emissions reported by APA were
considered.
Table 3 shows the contribution of total and in-port ship air
pollutant emissions (NOx, SO2, CO2 and NMVOC) to the total na-
tional emissions during 2013 and 2014.
As can be seen in Table 3, if the total (in-port and cruising)
emissions calculated in the present study were accounted in the
National Inventory Report made by APA, emissions of NOx, SO2, CO2
and NMVOC would increase 29%, 40%, 4% and <1% in 2013 and 31%,
45%, 5% and <1% in 2014. If only in-port emissions were accounted,
the contribution of the estimated emissions remains significant and
the emissions of NOx, SO2, CO2 and NMVOC would increase 14%,
19%, 2% and <1% in 2013 and 15%, 23%, 2% and <1% in 2014 These
percentages could be even higher if emissions were calculated for
all Portuguese ports and including ships that just pass along the
Portuguese coast up to 400 km. Accordingly, ships seem to be an
important source of SO2 and NOx emissions when compared with
land-based sources. These results are consistent with those of other
studies which suggested that ship emissions are a significant source
of airborne pollution (mainly regarding NOx and SO2) (Fan et al.,
2016; Goldsworthy and Goldsworthy, 2015; Kilic and Deniz, 2010;
Saraçoǧlu et al., 2013). Moreover, according to the Air quality in
Europe - 2016 report, performed by the European Environment
Agency for EU-28, emissions from international shipping within
European seas, may increase in 50% the NOx emissions and in 75%
the SOx emissions (EMEP/EEA, 2016b). In fact, NOx emissions from
ships are relatively high due to the high temperatures and pres-
sures in which marine engines operate, in most cases without
reduction technologies. SO2 emissions are high because of high
average sulphur content in some marine fuels used by most ships
(Eyring et al., 2009). These results highlight the necessity of
improving the monitoring and control of ship emissions in
Portugal.

Table 2
Cruising Emissions of air pollutants and ship calls by ship type during 2013 and 2014 (in t y
1).

Ship type Year NOx SO2 PM10 PM2.5 HC NMVOC CO CO2 CH4 N2O Total Calls

Bulk Carrier 2013 1.3Eþ03 6.5Eþ02 8.6Eþ01 8.0Eþ01 3.5Eþ01 1.8Eþ01 3.6Eþ01 3.8Eþ04 3.6E-01 1.9 4.1Eþ04 199
2014 1.6Eþ03 8.1Eþ02 1.1Eþ02 9.9Eþ01 4.3Eþ01 2.2Eþ01 4.5Eþ01 4.8Eþ04 4.5E-01 2.3 5.0Eþ04 234
Container Ship 2013 2.4Eþ04 1.3Eþ04 1.7Eþ03 1.6Eþ03 6.9Eþ02 3.5Eþ02 8.2Eþ02 7.8Eþ05 6.9 3.8Eþ01 8.2Eþ05 2094
2014 2.6Eþ04 1.5Eþ04 1.9Eþ03 1.8Eþ03 7.6Eþ02 3.8Eþ02 9.1Eþ02 8.7Eþ05 7.7 4.2Eþ01 9.1Eþ05 2235
General Cargo 2013 4.8Eþ03 2.8Eþ03 3.5Eþ02 3.4Eþ02 1.4Eþ02 6.6Eþ01 2.0Eþ02 1.6Eþ05 1.3 7.8 1.7Eþ05 1661
2014 4.9Eþ03 2.8Eþ03 3.6Eþ02 3.5Eþ02 1.4Eþ02 6.8Eþ01 2.1Eþ02 1.7Eþ05 1.4 8.0 1.8Eþ05 1695
Tanker 2013 6.4Eþ03 3.4Eþ03 4.4Eþ02 4.1Eþ02 1.8Eþ02 9.1Eþ01 2.0Eþ02 2.0Eþ05 1.8 9.6 2.1Eþ05 1552
2014 6.6Eþ03 3.4Eþ03 4.5Eþ02 4.2Eþ02 1.8Eþ02 9.3Eþ01 2.0Eþ02 2.0Eþ05 1.9 9.8 2.1Eþ05 1495
Passenger Ship 2013 3.9Eþ02 4.3Eþ01 8.1 8.6 1.2Eþ01 5.2 2.7Eþ01 1.6Eþ04 1.0E-01 7.7E-01 1.7Eþ04 68
2014 5.2Eþ02 5.4Eþ01 1.1Eþ01 1.1Eþ01 1.6Eþ01 7.0 3.6Eþ01 2.2Eþ04 1.4E-01 1.0 2.2Eþ04 78
Ro-Ro Cargo 2013 4.2Eþ03 2.2Eþ03 2.9Eþ02 2.7Eþ02 1.2Eþ02 5.8Eþ01 1.4Eþ02 1.3Eþ05 1.2 6.3 1.4Eþ05 239
2014 5.0Eþ03 2.6Eþ03 3.4Eþ02 3.2Eþ02 1.4Eþ02 7.0Eþ01 1.5Eþ02 1.5Eþ05 1.4 7.5 1.6Eþ05 285
Others 2013 1.5Eþ02 1.4Eþ01 3.1 3.1 3.9 1.8 8.1 5.1Eþ03 3.6E-02 2.5E-01 5.3Eþ03 83
2014 1.3Eþ02 2.1Eþ01 3.7 3.6 3.6 1.7 6.8 4.6Eþ03 3.3E-02 2.2E-01 4.7Eþ03 79
Total 2013 4.1Eþ04 2.2Eþ04 2.9Eþ03 2.7Eþ03 1.2Eþ03 5.9Eþ02 1.4Eþ03 1.3Eþ06 1.2Eþ01 6.5Eþ01 1.4Eþ06 5896
2014 4.5Eþ04 2.4Eþ04 3.2Eþ03 3.0Eþ03 1.3Eþ03 6.4Eþ02 1.6Eþ03 1.5Eþ06 1.3Eþ01 7.1Eþ01 1.5Eþ06 6101
 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379 1377

Table 3
Shipping emissions contribution of the four studied ports to the national emissions in 2013 and 2014 (in t y
1).

Pollutant Portugal emissions Portugal emissions Total ship In-port ship Total shipping emissions In-port shipping emissions
from APA report, in from APA report, in emissions (in emissions (in contribution to the national contribution to the national emissions
2013 2014 2013; 2014) 2013; 2014) emissions in 2013; 2014 in 2013; 2014

NOx 164 000 167 000 2.4Eþ05; 1.9Eþ05; 1.9Eþ05 29%; 31% 14%; 15%
2.4Eþ05
SO2 53 000 47 000 8.8Eþ04; 6.6Eþ04; 6.1Eþ04 40%; 45% 19%; 23%
8.5Eþ04
CO2 47 235 000 47 047 100 4.9Eþ07; 4.8Eþ07; 4.8Eþ07 4%; 5% 2%; 2%
4.9Eþ07
NMVOC 174 000 170 000 1.8Eþ05; 1.7Eþ05; 1.7Eþ05 <1%; <1% <1%; <1%
1.7Eþ05

3.4. Comparison between calculated ship emissions and other
foreign ports
Shipping emissions assessed in this study were compared with
other estimated for ports around the world (Table 4). All the studies
used for comparison applied the activity-based method to estimate
ship emissions, although the focus of the estimations differ with
respect to geographical boundaries (only in-port waters, or also
territorial waters), ship types, port activities/operations and pol-
lutants that were considered. Despite these constraints, some
comparisons can be useful. Table 4 shows the total amount of ship
emissions on ports, number of ship calls, operational modes and
ship routes considered in the different studies.
The in-port ship emissions at Leixo ~es, Setúbal and Sines ports
were higher than those estimated at Copenhagen Port, Elsinore
Port, Køge Port, Ambarlı Port and Izmir Port, despite the lower
number of ship calls at Portuguese ports (except for port of Køge).
The emissions at Las Palmas Port were similar to Leixo ~es. It should
be noted that the emissions reported by Saxe and Larsen (2004) for
the Denmark ports were lower than emissions calculated for all
ports in the present study. This seems related to the fact that for
Danish ports, CO2 and CO emissions were not included in the in-
ventory. Emissions at Busan Port were higher than those calculated
for the studied ports. Moreover, for the ship emissions inventory of
Busan Port the emissions of CO, CH4, N2O and HC were not
estimated.
Ship emissions of all pollutants calculated in this study were
considerably lower than those at Shanghai Port, which is one of the
largest port in the world (Yang et al., 2007). Although in the present
study cruising emissions were calculated only for ships that
anchored in four Portuguese ports up to 400 km from the coast, the
emissions of this mode of operation were higher than all foreign
ports emissions, except for Shanghai, which shows the importance
of the calculation of emissions from ships at open sea.
4. Uncertainties
The accuracy of this emission inventory is dependent on the
input data used and the assumptions made. In this study, the
number of calls, arrival and departure information of each ship

Table 4
Total amounts of ship emissions ports (in ton yr-1), number of ship calls, operational modes and ship routes considered in different studies.

Port Inventory Ship calls Operation Ship navigation routes Pollutants analysed Emissions Study
period mode (ton y
1)
~es Port
Leixo 2013, 2014 2 614, 2641 Ca, Mb, Hc In-port waters NOx, SO2, PM10, PM2.5, NMVOC, CO, CO2, 2.1Eþ05, This study
CH4, N2O and HC 2.1Eþ05
Setúbal Port 1 350, 1514 2.8Eþ05,
3.1Eþ05
Sines Port 1 982, 1975 3.1Eþ05,
3.4Eþ05
Viana do Castelo 199, 182 M, H 8.6Eþ03,
Port 8.9Eþ03
Cruising e e Up to 400 km from the coast 1.4Eþ06;
1.5Eþ06
Copenhagen Port, 2001 5729 M, H In-port waters NOx, SOx and TSP 698.3 Saxe and Larsen
Denmark (2004)
Elsinore Port, 45 226 1096.6
Denmark
Køge Port, 543 49.3
Denmark
Shanghai Port, 2003 1 280 128 C, M, H 129 km  102 km Shanghai Port NOx, SO2, PM, CO2 and HC 3.1Eþ06 Yang et al. (2007)
China study domain
Ambarlı Port, 2005 5432 C, M, H In-port waters 8.2Eþ04 Deniz and Kilic
Turkey (2009)
Izmir Port, Turkey 2007 2803 C, M, H A cruising distance of 128.8 km NOx, SO2, PM, CO2 and HC 8.6Eþ04 Saraçoǧlu et al.
(2013)
Busan Port, Korea 2008 50 402, 58 858, C, M, H In-port waters NOx, SO2, VOC, PM and CO2 5.7Eþ05 Song and Shon
49 736 (2014)
Las Palmas Port, 2011 3183 C, M, H In-port waters NOx, SOx, PM2.5 and CO2 2.2Eþ05 Tichavska and
Spain Tovar (2015)

N. a e Not applicable.
a
C e Cruising.
b
M e Manoeuvring.
c
H e Hotelling.
1378 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379
(used to calculate the hoteling times) and manoeuvring times, as
well as the names/identification numbers, date of construction,
gross weight and gross tonnage (GT) of ships were obtained from
port authorities. All ship characteristics were confirmed according
to Marine Traffic website which is a complete and reliable source
for basic ship information. Empirical values for ship speeds, engine
powers and load factors were used in this study, which may induce
a certain degree of uncertainty. The uncertainties in the emissions
estimations arise primarily from the following: i) the use of non-
linear regressions as a function of gross tonnage (GT) to obtain
ME powers, the determination of AE powers as a fraction of ME
power and the use of default AB powers; ii) the assignment of the
engine/fuel type profiles based on the ship fleet used in Entec UK
Limited study 2010, and iii) the use of default and constant ship
speeds and engine loads, which is known to change during ship
activities and for different weather conditions. In order to minimize
as much as possible the uncertainties associated with the use of
these default values, interviews with experienced members of the
shipping sector and with port authorities were performed. The
used EF also produce uncertainties, because they were based on
other overseas studies. Despite this, the selection was performed to
be the most representative for the Portuguese coast.
Further work is needed to improve the outcomes of this study
and minimize the scale of uncertainties. Improves could be ach-
ieved with the use of higher-resolution input data, especially on the
actual power of ME and EF.
5. Conclusions
This study allowed estimating the ship emissions of PM10, PM2.5,
NOx, SO2, CO, CO2, N2O CH4, NMVOC, and HC based on the activity-
based methodology in four of the main ports of Portugal (Leixo ~es,
Setúbal and Sines and Viana do Castelo) during 2013 and 2014.
According to the results, Leixo~es was the port that received more
ship calls in 2013 and 2014, followed by Sines, Setúbal, and Viana do
Castelo. The ship types that registered more ship calls were
container ships, general cargo and tankers.
At Leixo ~es emissions from container ships and tankers were
similar, corresponding to around 30% of estimated emissions. At
Setúbal Port emissions from tankers represents more than 50% of
total emissions which seems related to the higher hotelling times
and AE loads for these ships. At Sines container ships emitted the
highest amount of pollutants in 2013 and 2014 with more than 50%
of total emissions which seems related with the high engine
powers of these ships. At Viana do Castelo general cargo ships were
the highest emitters responsible for more than 75% of in-port
emissions.
Regarding cruising emissions, container ships were the largest
emitters responsible for around 59% and 65% of total emissions in
2013 and 2014, respectively. This results were related with the
higher distances travelled at sea up to 400 km from the Portuguese
coast for container ships, which led to higher times of permanency
in open sea.
CO2, NOx and SO2 were the pollutants for which the highest
amounts of cruising and in-port emissions were estimated. These
three pollutants represents more than 95% of the cruising and in-
port emissions. If the total emissions from ships calculated in the
present study, would have been taken into account in the National
Inventory Report on Greenhouse Gases, 1990e2014 performed by
APA, emissions of NOx, SO2, CO2 and NMVOC would increase 29%,
40%, 4% and <1% in 2013 and 31%, 45%, 5% and <1% in 2014. Thus, it
was possible to conclude that ships can be an important air
pollution source in Portugal, especially for NOx and SO2 emissions,
even not considering all Portuguese ports in the study. Neverthe-
less, maritime transport is highly efficient in an energetic point of
view, thus considering an increase on this type of transportation in
the following years, and consequently the increase in the associated
emissions, mitigation strategies should be considered regarding the
control of emissions namely for example on the fuel sulphur con-
tent, the use of different less pollutant combustibles, the creation of
NOx control areas, the increase of engines efficiency regarding fuel
consumption, the use of shore power during hotelling and others.
In the future, it is important to study emissions from all Portu-
guese ports as well as those from all ships sailing up to 400 km from
the Portuguese coast, which will allow making a total assessment of
atmospheric emissions from ships in the Portuguese territory and
in its coastal waters.
Conflict of interests
The authors declare no conflict of interests.
Acknowledgements
The authors are grateful to the port authorities of Viana do
~es, Setúbal and Sines for kindly providing data on
Castelo, Leixo
ship's movements and characteristics. This work was financially
supported by: Project POCI-01-0145-FEDER-006939 (Laboratory
for Process Engineering, Environment, Biotechnology and Energy e
LEPABE funded by FEDER funds through COMPETE2020 - Programa
Operacional Competitividade e Internacionalizaça ~o (POCI) and
grant SFRD/BPD/91918/2012 for SIV Sousa, funded by FCT, POPH/
QREN and European Social Fund (ESF).
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.envpol.2017.08.112.
References
APA, 2016a. Relato rio do Estado do Ambiente 2016. Age ^ncia Portuguesa do Ambi-
ente. Available from: http://sniamb.apambiente.pt/infos/geoportaldocs/REA/
REA2016/REA2016.pdf.
APA, 2016b. Portuguese National Inventory Report on Greenhouse Gases, 1990-
2014. Amadora Age ^ncia Portuguesa do Ambiente. Available from: http://unfccc.
int/national_reports/annex_i_ghg_inventories/national_inventories_
submissions/items/8108.php.
APP, 2016. Portos de Protugal. Available from: http://www.portosdeportugal.pt/
web/app/home.
Berechman, J., Tseng, P.H., 2012. Estimating the environmental costs of port related
emissions: the case of Kaohsiung. Transp. Res. Part D. Transp. Environ. 17,
35e38. http://dx.doi.org/10.1016/j.trd.2011.09.009.
CCDR-N, 2014. Inventa rio de emisso~es de poluentes atmosfe ricos na regi~
ao Norte -
rio Final elaborado no a
Relato ^mbito do Protocolo de colaboraç~ ao estabelecido
entre a CCDR Norte e a FCT/UNL. Comissa ~o de Coordenaça ~o e Desenvolvimento
Regional do Norte.
Corbett, J.J., Winebrake, J.J., Green, E.H., Kasibhatla, P., Eyring, V., Lauer, A., 2007.
Mortality from ship emissions: a global assessment. Environ. Sci. Technol. 41,
8512e8518. http://dx.doi.org/10.1021/es071686z.
Dataloy Systems, 2016. Dataloy Distance Table (DTT).
Deniz, C., Kilic, A., 2009. Estimation and assessment of shipping emissions in the
region of Ambarlı Port, Turkey. Environ. Prog. Sustain. Energy 107e115. http://
dx.doi.org/10.1002/ep.10373.
EEA, 2013. EMEP/EEA Air Pollutant Emission Inventory Guidebook 2013: Technical
Guidance to Prepare National Emission Inventories. European Environment
Agency, Luxembourg. http://dx.doi.org/10.2800/92722.
Entec, 2002. Quantification of Emissions from Ships Associated with Ship Move-
ments between Ports in the European Community. European Commission, Final
Report, Northwich, UK.
Entec, 2007. Ship Emissions Inventory Mediterranean Sea, Final Report for
Concawe.
Entec, 2010. Defra UK Ship Emissions Inventory Final Report, London, UK.
EMEP/EEA, 2016a. International Maritime Navigation, International Inland Navi-
gation, National Navigation (Shipping), National Fishing, Military (Shipping),
and Recreational Boats.
EMEP/EEA, 2016b. EMEP/EEA Air Pollutant Emission Inventory Guidebook 2016.
Eyring, V., Isaksen, I., Berntsen, T., Collins, W., Corbett, J., Endresen, O., Grainger, R.,
Moldanova, J., Schlager, H., Stevenson, D., 2009. Transport impacts on
 R.A.O. Nunes et al. / Environmental Pollution 231 (2017) 1370e1379
atmosphere and climate: Shipping. Atmos. Environ. 44, 4735e4771. http://
dx.doi.org/10.1016/j.atmosenv.2009.04.059.
Fan, Q., Zhang, Y., Ma, W., Ma, H., Feng, J., Yu, Q., Yang, X., Ng, S.K.W., Fu, Q., Chen, L.,
2016. Spatial and seasonal dynamics of ship emissions over the Yangtze river
delta and East China sea and their potential environmental influence. Environ.
Sci. Technol. http://dx.doi.org/10.1021/acs.est.5b03965 acs.est.5b03965.
Fitzgerald, W.B., Howitt, O.J.A., Smith, I.J., 2011. Greenhouse gas emissions from the
international maritime transport of New Zealand's imports and exports. Energy
Policy 39, 1521e1531. http://dx.doi.org/10.1016/j.enpol.2010.12.026.
Goldsworthy, L., Goldsworthy, B., 2015. Modelling of ship engine exhaust emissions
in ports and extensive coastal waters based on terrestrial AIS data - an
Australian case study. Environ. Model. Softw. 63, 45e60. http://dx.doi.org/
10.1016/j.envsoft.2014.09.009.
ICF, 2009. Current Methodologies in Preparing Mobile Source Port Related Emission
Inventories (Final Report).
IMO, 2016a. Marine Environment. International Maritime Organization. Available
from: http://www.imo.org/en/OurWork/Environment/Pages/Default.aspx.
IMO, 2016b. Prevention of Air Pollution from Ships. International Maritime Orga-
nization. Available from: http://www.imo.org/en/OurWork/Environment/
PollutionPrevention/AirPollution/Pages/Air-Pollution.aspx.
IVL, 2004. Methodology for Calculating Emissions from Ships: 1. Update of emission
factors Assignment for Swedish Environmental Protection Agency.
Jalkanen, J.-P., Brink, a., Kalli, J., Pettersson, H., Kukkonen, J., Stipa, T., 2009.
A modelling system for the exhaust emissions of marine traffic and its appli-
cation in the Baltic Sea area. Atmos. Chem. Phys. Discuss. 9, 15339e15373.
http://dx.doi.org/10.5194/acpd-9-15339-2009.
Jalkanen, J.P., Johansson, L., Kukkonen, J., Brink, A., Kalli, J., Stipa, T., 2012. Extension
of an assessment model of ship traffic exhaust emissions for particulate matter
and carbon monoxide. Atmos. Chem. Phys. 12, 2641e2659. http://dx.doi.org/
10.5194/acp-12-2641-2012.
Kilic, A., Deniz, C., 2010. Inventory of shipping emissions in izmit gulf. Turk. Environ.
Prog. Sustain. Energy 29, 221e232. http://dx.doi.org/10.1002/ep.
Liu, T.K., Sheu, H.Y., Tsai, J.Y., 2014. Sulfur dioxide emission estimates from merchant
vessels in a Port area and related control strat.egies. Aerosol Air Qual. Res. 14,
413e421. http://dx.doi.org/10.4209/aaqr.2013.02.0061.
Maragkogianni, A., Papaefthimiou, S., 2015. Evaluating the social cost of cruise ships
air emissions in major ports of Greece. Transp. Res. Part D. Transp. Environ. 36,
10e17. http://dx.doi.org/10.1016/j.trd.2015.02.014.
Marine Traffic, 2016. Available from: http://www.marinetraffic.com/en/ais/home/
centerx:-8/centery:41/zoom:5.
Matthias, V., Bewersdorff, I., Aulinger, A., Quante, M., 2010. The contribution of ship
1379
emissions to air pollution in the North Sea regions. Environ. Pollut 158,
2241e2250. http://dx.doi.org/10.1016/j.envpol.2010.02.013.
Ng, S.K.W., Loh, C., Lin, C., Booth, V., Chan, J.W.M., Yip, A.C.K., Li, Y., Lau, A.K.H., 2013.
Policy change driven by an AIS-assisted marine emission inventory in Hong
Kong and the Pearl River Delta. Atmos. Environ. 76, 102e112. http://dx.doi.org/
10.1016/j.atmosenv.2012.07.070.
Nunes, R.A.O., Alvim-Ferraz, M.C.M., Martins, F.G., Sousa, S.I.V., 2017. The activity-
based methodology to assess ship emissions e a review. Environ. Pollut. 231,
87e103. http://dx.doi.org/10.1016/j.envpol.2017.07.099.
Sanabra, M.C., Usabiaga Santamaría, J.J., Martínez De Ose s, F.X., 2014. Manoeuvring
and hotelling external costs: enough for alternative energy sources? Marit.
Policy Manag. 41, 42e60. http://dx.doi.org/10.1080/03088839.2013.782441.
Saraçoǧlu, H., Deniz, C., Kiliç, A., 2013. An investigation on the effects of ship
sourced emissions in Izmir port, Turkey. Sci. World J. 2013 http://dx.doi.org/
10.1155/2013/218324.
Saxe, H., Larsen, T., 2004. Air pollution from ships in three Danish ports. Atmos.
Environ. 38, 4057e4067. http://dx.doi.org/10.1016/j.atmosenv.2004.03.055.
Song, S., 2014. Ship emissions inventory, social cost and eco-efficiency in Shanghai
Yangshan port. Atmos. Environ. 82, 288e297. http://dx.doi.org/10.1016/
j.atmosenv.2013.10.006.
Song, S.K., Shon, Z.H., 2014. Current and future emission estimates of exhaust gases
and particles from shipping at the largest port in Korea. Environ. Sci. Pollut. Res.
21, 6612e6622. http://dx.doi.org/10.1007/s11356-014-2569-5.
Starcrest Consulting Group, 2014. Port of Los Angeles Inventory of Air Emissions e
2013 Report.
Starcrest Consulting Group, 2015. Port of Los Angeles Inventory of Air Emissions e
2014 Report.
Tichavska, M., Tovar, B., 2015. Port-city exhaust emission model: an application to
cruise and ferry operations in Las Palmas Port. Transp. Res. Part A Policy Pract.
78, 347e360. http://dx.doi.org/10.1016/j.tra.2015.05.021.
Trozzi, C., Consulting, T., Gregorio, V., Curbastro, R., Roma, I., 2010. Emission Esti-
mate Methodology for Maritime Navigation 6.
Tzannatos, E., 2010. Ship emissions and their externalities for the port of Piraeus e
Greece. Atmos. Environ 44, 400e407. http://dx.doi.org/10.1016/
j.atmosenv.2009.10.024.
Yang, D.Q., Kwan, S.H., Lu, T., Fu, Q.Y., Cheng, J.M., Streets, D.G., Wu, Y.M., Li, J.J., 2007.
An emission inventory of marine vessels in Shanghai in 2003. Environ. Sci.
Technol. 41, 5183e5190. http://dx.doi.org/10.1021/es061979c.
Yau, P.S., Lee, S.C., Corbett, J.J., Wang, C., Cheng, Y., Ho, K.F., 2012. Estimation of
exhaust emission from ocean-going vessels in Hong Kong. Sci. Total Environ.
431, 299e306. http://dx.doi.org/10.1016/j.scitotenv.2012.03.092.