ArticlePDF Available

Allocation and Forecasting of Global Ship Emissions

Allocation and Forecasting of Global Ship Emissions
James J. Corbett
Chengfeng Wang
James J. Winebrake
Erin Green
Prepared for the Clean Air Task Force
Boston, MA, USA
11 January 2007
University of Delaware; Newark, Delaware; USA.
Energy and Environmental Research Associates, LLP; Pittsford, NY; USA.
Rochester Institute of Technology; Rochester, NY; USA.
Green and McGrath, Associates; Rochester, NY; USA.
1 Introduction and Summary ..................................................................................................... 2
2 Inventory Development .......................................................................................................... 4
2.1 Overview......................................................................................................................... 4
2.2 Spatial Proxies ................................................................................................................ 5
2.3 Baseline Inventory, Comparison, and Validation........................................................... 9
2.4 Results................................................................................................................................. 11
3 Forecasting Trends................................................................................................................ 11
3.1 Overview....................................................................................................................... 11
3.2 Ship Installed Power as Emissions Trend Indicator ..................................................... 13
3.3 Growth Forecast Discussion......................................................................................... 14
3.4 Emissions Adjustments................................................................................................. 19
4 Conclusions........................................................................................................................... 19
5 References............................................................................................................................. 22
Table 1. Emissions Comparison with Different Regional Estimates (’000 metric tons).............. 10
Table 2. Global ship emissions (Unit: metric tons) ...................................................................... 11
Table 3. Sensitivity analysis of potential improvement in thermal efficiency (fuel economy)... 17
Table 4. Matrix of maximum annual percent fuel economy gain over from 1970 to 2005,
comparing thermal efficiencies in new versus replaced engines...................................... 18
Table 5. Projected SOx emissions under BAU and various global sulfur-control scenarios. ..... 20
Figure 1. Summary of ship emissions point estimates (20, 23, 35, 36). Box-plots represent the
and 95
percentile results from uncertainty analysis; whiskers extend to lower and
upper bounds (20, 25)......................................................................................................... 6
Figure 2. Illustration of oceangoing ship traffic, based on 2000-2002 ICOADS proxy................ 8
Figure 3. Comparison of ship emissions spatial proxies derived from power-weighted AMVER
2005 and improved ICOADS 2000-2002; value of each 0.1° by 0.1° grid cell represents
the difference between the two proxies (in millionth of global emissions); grid cells with
warm color have more emissions allocated by AMVER while grid cells with cool color
have more emissions allocated by ICOADS....................................................................... 8
Figure 4. Comparison between the results of STEEM, top-down approach using ICOADS,
AMVER, and the combination. Coastal zones resemble the 200 nautical miles Exclusive
Economic Zone (EEZ). Canadian West Coast includes northwest of Canada; and
Canadian East Coast includes the northeast of Canada.................................................... 10
Figure 5. Average installed power in the world fleet by year-of-build (65)................................ 15
Figure 6. Global indices for seaborne trade, ship energy/fuel demand, installed power............. 16
Figure 7. Comparison of BAU SOx trends with global sulfur controls at 1.5%, 1%, and 0.5%. 20
1 Introduction and Summary
1.1 This report presents updated global inventories of emissions from international shipping
using best practices, and projects future shipping emissions. More specifically, in this report we:
1. Summarize recent work with regard to global ship emissions inventories.
2. Describe how growth in trade activity affects fleet energy and emissions trends.
3. Present updated global ship emission inventories at high resolution (0.1 degree latitude
by 0.1 degree longitude).
4. Present a business-as-usual (BAU) forecast of fleet energy and emissions trends
consistent with the correlation between increased oceangoing trade and required power.
We believe this inventory represents the most complete and accurate global inventory of air
emissions from international shipping prepared to date. As such, the global ship emissions
presented in this report will facilitate analyses of shipping’s contribution to ambient
concentrations of these pollutants and their secondary atmospheric products. Combined, these
inventories and pending atmospheric modeling will support analyses and improved
understanding of the impacts to human health due to ship emissions.
1.2 This study was conducted to improve the understanding of the International Maritime
Organization (IMO) and other environmental regulatory authorities regarding air pollution
emitted from global shipping activity and to inform IMO deliberations on potential revisions to
standards for air pollution from ships set forth in Annex VI to MARPOL 73/78 (1-3). It should
also inform IMO deliberations regarding greenhouse gas (GHG) emissions from ships, as
reflected in the IMO Study on Greenhouse Gas Emissions from Ships (4, 5), related resolution (6)
and subsequent documents. IMO has conducted significant work to evaluate fleet environmental
impacts from propulsion emissions over more than a decade, as Annex VI was drafted, ratified,
and entered into force. Presently, IMO delegates are discussing revisions to Annex VI to better
achieve environmental stewardship goals for international shipping (7, 8), partly by continuing or
expanding its focus on reducing oxides of nitrogen (NOx), sulfur emissions (SOx and sulfate
aerosol formation), and related particulate matter (PM). These efforts include consideration of
the fate of particulate and aerosol emissions emitted by ships on major trade routes (9), industry
assessments of feasibility and benefits from expanded fuel-switching to lower-sulfur fuels
through additional SOx Emissions Control Areas (SECAs) and/or globally (10, 11), and
consideration of geographically non-uniform NOx standards that would reduce land-side
exposure (12). Work currently supporting IMO deliberations includes updated studies in Europe
on the air quality, health, and environmental impacts attributed to shipping by peer-reviewed
experts published in leading journals (13, 14). This work is part of European efforts to identify
cost-effective abatement measures to reduce emissions of air pollution from ships (15), and to
implement a new Thematic Strategy on Air Pollution that was adopted by the European
Commission in September 2005 (16).
1.3 IMO submittals by Friends of the Earth International (FOEI) have contributed
information related to air pollution from ships and assessments that show a number of
operational and technical measures for reducing such emissions from ships, feasible in both
short- and long-term contexts (17, 18). FOEI’s submittal to MEPC 53 (MEPC 53/4/1) discussed
potential health impacts and growth in fleet emissions, citing data showing that increased
international shipping emissions will overwhelm sulfur reductions in IMO-compliant SECAs and
that emissions in European sea areas will exceed total land-based emissions in the EU25
countries by as early as 2020.
1.4 In terms of assessments, IMO has received analyses of local and regional inventories and
impacts, and more qualitative global evaluations of potential costs and mitigation strategies.
Global inventories and forecasts are needed to enable larger scale assessment of impacts and
mitigation measures. The IMO-commissioned Study on Greenhouse Gas Emissions from Ships
(4, 5) used an activity-based inventory approach to better understand total energy use by ships,
fleetwide trends, and potential GHG reductions from technical and operational measures (19).
1.5 This report describes global ship activity data and best-practice methodologies for
producing emissions inventories. These best-practices include identification and use of installed
power characteristics, current power-based emissions factors, engine load service corrections,
and engine operating time (19-21). Improved spatial resolution is presented as well, updating
global representation of shipping in the IMO Study on Greenhouse Gas Emissions from Ships as
well as updating emissions estimates. The report uses ICOADS data, the source with the greatest
spatial detail and longest publicly available time series, to produce 2002 (baseline) inventories
for ship SOx, PM, and black carbon emissions. However, ICOADS data are improved by
trimming over-reporting vessels, by using multiple-year data, and by weighting ship observations
with ship installed power; these steps mitigate sampling bias, augment sample data set, and
account for ship heterogeneity. The report adopts updated emissions inventory values reported
and confirmed in peer-reviewed articles, reflecting total ship fuel use.
1.6 Using a power-based approach consistent with baseline inventory practices, we construct
a business-as-usual (BAU) forecast scenario that reveals faster growth rates for total energy
requirements in recent decades than previously reported (5, 22). We discuss and consider a
range of economic and energy trends that support this BAU forecast, demonstrating that the
updated forecast trends may be more conservative than direct extrapolation from the past three
1.7 This forecast confirms that current IMO policy does not reduce emissions even in a
short-term context. Furthermore, emission reductions in the 60% range will do little more than
offset through 2030 emissions growth accompanying projected increases in future shipping
activity. New-engine NOx reductions resulting from IMO-compliant fleet turnover since 2000
are less than increased NOx emissions over this period. In other words, controls reducing
fleetwide shipping emissions by at least 60% would need to be fully implemented for both new
and existing engines within the next two decades to maintain 2002 shipping pollution levels.
1.8 This study’s results provide a global context within which other IMO submittals focused
on regional fate and transport and human health can be considered. The inventories can be used
to model BAU impacts with atmospheric models, and can be adjusted to represent certain policy
controls (e.g., future SECAs). This context enables insightful consideration by IMO and
member nations of areas where air quality, human health effects, and environmental impacts may
support revision of MARPOL Annex VI, the NOx Technical Code and related guidelines. Given
that more aggressive technology and operational measures are necessary to offset increasing
emissions of greenhouse gases and air pollution, this report may also help guide the choice of
those technologies and strategies that have the potential to achieve necessary emissions
2 Inventory Development
2.1 Overview
2.1.1 Approaches applied in previous studies to produce spatially resolved ship emissions
inventories can be categorized as either top-down or bottom-up. Top-down methods apply non-
spatial estimates to a spatial domain using some proportionality assumptions; bottom-up methods
develop estimates directly from spatially resolved detail and aggregate this activity to represent
the inventory domain. Both of these types of approaches have strengths and weaknesses. A top-
down approach is capable of producing multi-scale inventories perhaps in a quicker and less
costly way and is less resource-demanding. However, the accuracy of a top-down approach is
limited by the “representativeness” of the spatial proxy of ship traffic and the accuracy of the
global inventory (20, 23-26). On the other hand, a bottom-up approach faces problems associated
with large numbers of ship movements and potentially dynamic shipping routes that are time-
dependent (26-28). Both methods have been applied in service of IMO policy development. In
our work, a global shipping network has been developed and applied to data for North America;
this hybrid approach combines the best of the bottom-up and top-down approaches (26, 29-32).
However, additional time is needed to build worldwide data sets to produce this type of
hybrid approach for global ship emissions inventories. Nevertheless, atmospheric modelers and
policy makers need a global inventory in the interim. After comparing statistically and
geographically two spatial proxies to produce top-down global ship emissions inventories, we
use the International Comprehensive Ocean-Atmosphere Data Set (ICOADS) to improve
representativeness of the top-down approach by addressing potential statistical and geographic
sampling bias caused by over-reporting and non-responding ships.
2.1.2 Representing spatial (and temporal) activity of commercial shipping is fundamentally
similar to modeling any mobile source: the location and intensity of the fleet activity must be
depicted. This can be done using sampled data representing fleetwide spatial activity and its
temporal dynamics. Ideally, the reporting ship fleet would be representative of the world fleet
pattern in terms of ship type, size, installed power, operating profiles, etc. The number of ships
reporting should be statistically large or at least approximate a randomly sampled subset of the
world fleet, and locations should be reported regularly at equal intervals. For emission estimating
purposes, ship load or speed data are useful; in addition, a ship identifier can be used to adjust
ship reporting frequencies with related ship attributes. Two global ship reporting data sets,
ICOADS and the Automated Mutual-assistance Vessel Rescue system (AMVER) data set (33),
have been previously used as proxies of ship traffic to geographically resolve the global
emissions inventories (23, 34, 35).
2.1.3 Emissions inventories for oceangoing ship are calculated using activity-based
methodologies referenced above. There are, however, significant differences among various
global ship emission inventories; inventories estimated by one approach may be 50% higher than
inventories estimated by another (20, 24, 25). A reliable and up-to-date global ship emission
inventory is also critical to the accuracy of the spatially resolved inventory.
In this work, we
adopted the updated inventories produced by Corbett and Koehler (20), generally confirmed by
Eyring et al (36). The methodology is as follows (20):
Step 1: Identify the vessel(s) to be modeled, and engines in service
Step 2: Estimate the engine service hours for the voyage or voyage segment
Step 3: Determine the engine load profiles, including power and duty cycle
2.2.2 ICOADS is the world's largest data set for global marine surface observations voluntarily
reported by the Voluntary Observing Ships (VOS) fleet and taken from moored and drifting
buoys (only data reported by marine vessels were used for ship traffic and emissions analysis
purposes) (37, 38). Ships are recruited by Members of the World Meteorological Organization
(WMO) on the basis of the willingness of the ships' officers to perform the observations and the
regular route followed by the ships. Ships are asked to make observations at the standard
synoptic hours four times a day and send to a meteorological service as soon as possible.
However, analysis shows that many ships reported only once a year and the majority of ships
reported less than 280 times per year while a few ships reported more 6,000 times a year (39).
Due to changes in the global ocean observational systems, the number of ships on the Voluntary
Observing Ships (VOS) fleet list, which is the main source of ship observations for ICOADS,
has declined from a peak of 7,700 in 1984-1985 to about 4,000 ships worldwide in 2003, which,
by number, is about 13.8% of the world fleet of ships of 1,000 gross tons and greater, or about
4.4% of the world fleet (37, 40-42). ICOADS data is publicly available from This work is based on the analysis of 20-year of ICOADS data
Figure 2 illustrates ship traffic intensity derived from ICOADS data.
Where e
represents emissions of one mass unit of pollution (in tons, kilograms, or grams)
from cell i in one period; e
as the global ship emissions inventory (in tons, kilograms, or grams)
for that pollutant in that period; w
is the value of cell i represented by a global spatial proxy; and
is the sum of the value of all the grid cells of the proxy in that period.
2.2.1 Mapping emissions requires additional steps. Global ship emissions inventories are
allocated to each geographic grid cell proportional to the activity level in that cell represented by
spatial proxies. Emissions in each grid cell can be calculated with equation (1):
2.2 Spatial Proxies
2.1.4 These efforts yield a total value for the fleet emissions included in the scale of the
estimate (e.g., port-based, national, regional, global).
Figure 1 presents a summary of recent
activity-based estimates for NOx (as elemental nitrogen), SOx (as elemental sulfur), and
particulate matter (PM
). The inventory by Corbett and Fischbeck (35) used international
marine fuels and fuel-based emissions factors; Endresen et al (23) also used fuel-based emissions
factors, but included activity-based data such as operating hours, engine load, and specific fuel
Ranges depicted in Figure 1 suggest that NOx and SOx pollution from oceangoing ships
represent some 15-30% of global NOx emissions and 5-7% of global SOx emissions, while fuel
usage ranges 2-4% of world fossil fuels. Spatial representation is required to fully understand
environmental impacts at regional and local scales, especially for aerosols and particulate matter
Step 4: Apply emissions or fuel consumption rates for specific engine/fuel combinations
Step 5: Estimate emissions or fuel consumption for the voyage or voyage segment
Steps 6+: Assign emissions spatially and temporally both in and out of port regions
Cargo Fleet NOx (as N) Cargo Fleet SOx (as S) Cargo Fleet PM
Tg per Year
Eyring et al, 2005 Corbett and Koehler, 2003 Endresen et al, 2003 Corbett and Fischbeck, 1999
Registered Fleet NOx (as N) Registered Fleet SOx (as S) Registered Fleet PM
Tg per Year
Eyring et al, 2005 Corbett and Koehler, 2003 Endresen et al, 2003 Corbett and Fischbeck, 1999
World Fleet NOx (as N) World Fleet SOx (as S) World Fleet PM
Tg per Year
Figure 1. Summary of ship emissions point estimates (20, 23, 35, 36). Box-plots represent the 5
and 95
percentile results
from uncertainty analysis; whiskers extend to lower and upper bounds (20, 25).
2.2.3 AMVER, sponsored by the United States Coast Guard (USCG), is a global ship reporting
system used worldwide by search and rescue authorities to arrange for assistance to ships and
persons in distress at sea. Participation is free, voluntary, and open to merchant ships of all flags,
but, until recently, had been limited to ships over 1000 gross tons, on a voyage of 24 hours or
longer. Enrollment now has been expanded to accommodate vessels outside the traditional
criteria, such as cruise ships, research vessels and fish processors. AMVER vessels are requested
to report their position every 48 hours at sea and as soon as voyage information changes (voyage
deviation). At the discretion of the master, reports may be sent more frequently than requested
(43). Apparently, many ships report much less than requested and around 4,000 vessels with at
least 128 days on the AMVER plot in a calendar year are eligible for an award (44).Although
there are about 12,000 vessels from more than 100 nations in the AMVER database (23), around
4,000 vessels have actively participated in the AMVER system in recent years (44). The number
of unique vessels reporting to AMVER in a given year appears to vary from year to year, but
appears to demonstrate similar consistency to ICOADS data. AMVER is strictly confidential and
used only in a bona fide maritime emergency (45). Only a few researchers have been given
access to limited AMVER data under special agreement, and this data does not include ship
2.2.4 The number of observations made by one type of ship is a function of number of vessels,
operating profile (e.g., time at sea), and reporting frequency of that type of ship. Both ICOADS
and AMVER are statistically and spatially biased. Neither of the two data sets perfectly
represents the world fleet and its activity. An examination of ICOADS reveals that it over-
samples container ship traffic, and, to a lesser extent, refrigerated cargo ship (i.e., reefer) traffic,
and under-samples general cargo ship and tanker traffic, particularly general cargo ship traffic.
These vessels more typically require weather routing information than bulk vessels due to their
liner schedules, which may explain their increased participation in ICOADS. In contrast,
AMVER over-samples bulk carrier, tanker, and container ship traffic, especially bulk carrier and
tanker traffic and significantly under-samples general cargo ship, RO-RO, and reefer traffic.
AMVER’s sampling bias is consistent with its participating criteria, and apparently attracts
greater participation of vessels that operate outside of set liner schedules. Consequently, these
two proxies allocate the intensity of ship activity differently (as expected by routing differences
among ship types) and produce emissions inventories with significant differences in some
regions. These differences could significantly affect regional accuracy of atmospheric modeling
of ship emissions.
Figure 3 shows how the two proxies will allocate global ship emissions
Efforts have been taken to improve the two proxies (23, 31, 32). When ICOADS
observations are weighted by installed power, more emissions will be assigned to 0°-40° north
and less will be assigned to 40°-90° north, 0°-20° south and 40°-90° south than unweighted
ICOADS. In contrast, weighting AMVER data by gross tonnage matters less. This also implies
that AMVER covers a less broad range of vessels by size. ICOADS, whether weighted by ship
installed power or not, will assign significantly more emissions than AMVER to the region
between 40°-60° north and assign significantly fewer emissions to the region between 0°-20°
Figure 2. Illustration of oceangoing ship traffic, based on 2000-2002 ICOADS proxy.
Figure 3. Comparison of ship emissions spatial proxies derived from power-weighted AMVER 2005 and
improved ICOADS 2000-2002; value of each 0.1° by 0.1° grid cell represents the difference between the two
proxies (in millionth of global emissions); grid cells with warm color have more emissions allocated by
AMVER while grid cells with cool color have more emissions allocated by ICOADS.
2.2.5 We chose ICOADS as the spatial proxies for allocating ship emissions for this work for
three reasons:
ICOADS is free to researchers and has maintained a historical data set for
temporal analysis, while AMVER can be obtained only under special agreement
and does not maintain a historical data set;
ICOADS covers a wider range of ships in terms of ship type, size, and engine
power (132 – 74,640 kW), and thus it may better represent the world ship traffic
(46), while AMVER focuses on larger vessels due to its participation criteria; and
Individual ships reported to ICOADS can be identified to improve the proxy.
2.2.6 The three most recent years, 2000-2002, ICOADS data were used to augment the sample
size and to mitigate non-uniform response bias. Lloyd’s ship registry data set was used to
identify ships made ICOADS reports (46) based on unique ship identifiers, and only observations
made by ships with valid installed power were selected. The number of observations made by
each ship was examined and outliers were trimmed. Each observation was weighted by ship
installed power which is a more direct of indicator of ship emissions than with ship size (20, 21,
26). Power-weighted method increases emissions assigned to ocean-shipping lanes while
decreasing the emissions to coastal routes.
2.3 Baseline Inventory, Comparison, and Validation
2.3.1 Global emissions inventories were evaluated and validated by comparing with studies at
regional and port scales. Figure 4 presents the comparison of a top-down approach using
ICOADS, AMVER, and a combination of the two as spatial proxies with STEEM, a bottom-up
network model recently applied to estimate North America coastal emissions (30). STEEM
empirically routed more than 170,000 voyages and estimated emissions from activity-based
characteristics of ships using these routes; the spatial distribution from STEEM has no sampling
bias since it represents all available port call activity for oceangoing ships in international
service. As shown in Figure 4, ICOADS, AMVER, and their combination are each biased
around North America. None produces better inventories in all areas than the other two.
2.3.2 Accuracy of the ICAODS-based inventory was also examined by comparing it with
results of previous regional studies. Results of different reference years for different vessel
groups were extrapolated and/or projected to the same reference year, 1990, to enable the
comparison. Table 1 summarizes the comparison of our approach with the results from regional
studies by Lloyds Register and ENTEC, both using a bottom-up approach.
The Baltic Sea is defined in the Lloyds Register study as “the Gulf of Bothnia and the
Gulf of Finland” and the entrance to the Baltic Sea bounded by the parallel of the Skaw in the
Skagerrak at 57˚44.8’N” (27). The North Sea is defined as in MARPOL 73/78 ANNEX VI
(regulation 5(1)) (47). The Mediterranean and Black Sea area is defined in the Lloyds Register’s
study as “including the gulfs and seas there in, bounded to the west by the Straits of Gibraltar at
the meridian 5˚36’W” and the Black Sea (28, 48).
Our estimates for SO
for the Mediterranean/Black Sea agree very well with the Lloyds
study; our SO
results are about 10% higher than Lloyds’. Results for NO
and SO
for other
regions, except the Baltic Sea, agree fairly well with other studies. The discrepancies for the
Baltic Sea are significant, with our results about 50% lower than Lloyds Register study;
however, our results for all shipping in the Baltic/North Sea combination fall in the range of the
results derived from ENTEC, Lloyds, and IIASA studies (22, 27, 28, 47, 49).
US Alaska
US West Coast
US East Coast
US Great lakes
US Gulf Coast
US Hawaii
MX West Coast
MX Gulf Coast
CA West Coast
CA East Coast
CA Great Lakes
metric tons of SO2
Figure 4. Comparison between the results of STEEM, top-down approach using ICOADS, AMVER, and the
combination. Coastal zones resemble the 200 nautical miles Exclusive Economic Zone (EEZ). Canadian West
Coast includes northwest of Canada; and Canadian East Coast includes the northeast of Canada.
Table 1. Emissions Comparison with Different Regional Estimates (’000 metric tons)
Region Source
Year and fleet represented
Mediterranean/Black Sea This work 958 550 International cargo fleet, 2002
2,387 1,391 Estimated for all shipping, 1990
Lloyds (28) 1,725 1,246 All shipping movements, 1990
Baltic/North Sea
This work 502 287 International cargo fleet, 2002
1,601 929 Estimated for all shipping, 1990
ENTEC (22) 1,074 763 All shipping movements, 2000
1,074 763 Extrapolated for all shipping, 1990
Lloyds, IIASA (27, 47, 49)
N/A 1,400 Estimated for all shipping, 1990
North Sea
This work 445 255 International cargo fleet, 2002
1,421 826 Estimated for all shipping, 1990
IIASA (27, 49) N/A 439 International shipping, 1990
N/A 1,171 Extrapolated for all shipping, 1990
Baltic Sea This work 56 32 International cargo fleet, 2002
179 104 Estimated for all shipping, 1990
Lloyds (27) 353 229 All shipping movements, 1990
1. Extrapolations to common years for comparison are based on the cited sources.
2.3.3 Some discrepancies identified in Table 1 may be due to a number of factors, including
sampling bias, assignment of emissions by ship type, and inconsistent definition of terms.
Additionally, emissions from hotelling activity in and near ports are likely under-
represented in this work for at least two reasons. First, the top-down approach allocates a
calculated global inventory to the ICOADS (or other ship activity proxy) proportionally -
with adjustments for installed power, etc., as described above. Second, ship reporting
frequencies may decrease as operators prepare for arrival and/or departure, and ships may
not report locations to ICOADS or AMVER while at dock. Additionally, port-based
inventories would typically include emissions from related harbor craft activity (e.g.,
vessel-assist tugs) and may allocate ship activity along more resolved local navigation
routes. For these reasons, this global context may support but not replace detailed local
inventories using bottom-up methods.
2.3.4 Taking the foregoing into consideration, we do not consider this work to contradict the
inventory, fate and transport, and health effects estimates provided in other studies for
this region (13-15), and we conclude that our results either agree very well or fall in the
range of previous studies.
2.4 Results
2.4.1 Our final inventory results for various pollutants are shown in Table 2 with monthly
global ship emissions inventories produced using ICOADS 2000-2002 power-weighted spatial
Table 2. Global ship emissions (Unit: metric tons)
NOx (N) SOx (S) CO2 (C)
1 7.81% 390,743 368,861 13,754,144 44,857 92,997 84,400 5,580
2 7.33% 366,634 346,103 12,905,531 42,090 87,259 79,193 5,236
3 8.20% 410,136 387,168 14,436,788 47,084 97,612 88,589 5,857
4 8.10% 404,813 382,143 14,249,414 46,473 96,345 87,440 5,781
5 8.25% 412,597 389,492 14,523,415 47,366 98,198 89,121 5,892
6 7.87% 393,421 371,389 13,848,421 45,165 93,634 84,979 5,618
7 8.55% 427,683 403,733 15,054,458 49,098 101,789 92,380 6,107
8 8.68% 434,240 409,923 15,285,248 49,851 103,349 93,796 6,201
9 8.50% 424,797 401,008 14,952,847 48,767 101,102 91,756 6,066
10 8.90% 445,173 420,244 15,670,102 51,106 105,951 96,157 6,357
11 8.99% 449,502 424,330 15,822,462 51,603 106,981 97,092 6,419
12 8.81% 440,261 415,606 15,497,171 50,542 104,782 95,096 6,287
Total 100.00% 5,000,000 4,720,000 176,000,000 574,000 1,190,000 1,080,000 71,400
Note: We use Corbett and Koehler emissions inventory (base case) with 2000-2002 ICOADS monthly variation.
3 Forecasting Trends
3.1 Overview
3.1.1 Forecasts can differ depending on their purposes and scales. Some forecasts look to
reveal where timely investment and action at a local scale or by a single firm can produce the
most benefit (e.g., profit). Other forests are intended to be conservative or aggressive; that is,
they intend to be biased to serve the decision makers’ value and tolerance for risk and surprise.
This may describe large scale forecasts such as emissions or trade trends. One challenging class
of forecasts may be considered “difference” forecasts, where alternative scenarios illustrate how
a path taken” may differ from “a path not taken” rather than to determine which is most
probable. These kinds of forecasts are common in policy domains, such as energy, environment,
and economics (e.g., IPPC scenarios). Certainly, ship emissions forecasting presents one
challenging example, especially at the international or multinational scales, and especially when
considering policy actions like a SOx Emissions Control Area (SECA) under IMO MARPOL
Annex VI (2).
3.1.2 Previous studies available to IMO have described global growth rates for maritime
shipping, based on fleet size, trade growth, and/or cargo ton-km, mostly calibrated to linear or
conservative extrapolations of historic data. The IMO Study on Greenhouse Gas Emissions from
Ships (5) used fleet growth rates based on two market forecast principles, validated by historical
seaborne trade patterns: 1) World economic growth will continue; and 2) Demand for shipping
services will follow the general economic growth. The IMO study correctly described that
growth in demand for shipping services was driven by both increased cargo (tonnage) and
increased cargo movements (ton-miles), and considered that these combined factors make
extrapolation from historic data difficult. Nonetheless, their forecast for future seaborne trade
(combined cargoes in terms of tonnage) was between 1.5% and 3% annually. The IMO study
applied these rates of growth in trade to represent growth in energy requirements.
The ENTEC study (22) adopted growth rates from the IMO study. Eyring et al. (50)
estimated “future world seaborne trade in terms of volume in million tons for a specific ship
traffic scenario in a future year” using a linear fit to historical GDP data. Interestingly, this
represents one of the only studies to forecast growth in seaborne trade for energy and emissions
purposes at rates faster than GDP. The TREMOVE maritime model (51, 52) estimates fuel
consumption and emissions trends derived from forecast changes in ship voyage distances
(maritime movements in km) and the number of port calls. According to the TREMOVE report,
maritime “fleet and vehicle kilometres grow annually by 2.5% for freight and 3.9% for
passengers,” while “port callings grew by 8% compared to the previously used input figures.”
3.1.3 Except for the Eyring et al. work, these linear extrapolations appear to present growth
rates slower than the economy. Linear extrapolations are likely biased on the low side, because
shipping growth rates have actually grown faster than the economy. Freight transportation,
particularly international cargo movement, is an important and increasing contributor to global
and national economic growth, as well as state and regional economic growth in and around
major cargo ports. If growth in GDP and trade volumes is compounded as forecast by economic
and transportation demand studies, then growth in energy requirements should be non-linear
also. The U.S. Bureau of Transportation Statistics (BTS) recently released a report that describes
North American freight activity and trends (53). This document reports growth rates for North
America above 7.4% for international trade and above 7.2% across all measures of value, and
states that:
Since 1994, the value of freight moved among the three countries has averaged
almost 8 percent annual growth in both current and inflation-adjusted terms,
compared with about 7-percent growth for U.S. goods trade with all countries
(table 1). In 2005, both goods trade and gross domestic product (GDP) grew in
inflation-adjusted terms. Except in 2001 and 2002, during the past decade, U.S.
trade with Canada and Mexico has increased at a faster rate than U.S. GDP.
Growth in goods movement by dollar value may be expected to differ from growth in the
volume of goods moved, and in the change in activity by the multimodal fleets (ships, trucks,
trains, and aircraft) moving cargo. We confirmed that the contribution of international trade is
increasing as a proportion of U.S. gross domestic product (GDP) – i.e., freight transportation is
growing faster than U.S. GDP (53, 54). Economic activity related to imports and exports
together contribute about 22% of recent U.S. GDP in recent years; whereas, goods movement
contributed only about 10% of GDP in the 1970s. Moreover, the dominance of containerized
cargoes in seaborne trade suggests that truck and containerized shipments may double by 2025 or
sooner (55). GDP in the U.S. is growing at ~3.7% CAGR since 1980, and the freight sector is
growing at ~6.4% CAGR over the same period (54). This freight-sector growth rate in terms of
dollar value is reflected in the observed ~6.3% to 7.2% annual growth rates of “high-value”
containerized trade volumes, particularly from Asia (56).
Studies for Southern California (San Pedro Bay) ports agree that growth in cargo
volumes equivalent to 6-7% compounding annual growth rates is expected (57-60). However,
increased cargo may not produce a corresponding increase in port calls, as some studies interpret
(58). Historic data on port calls to San Pedro Bay have shown the number of ship calls remained
between 5,000 and 7,000 calls per year since the 1950s (61).
3.1.4 Freight energy use is correlated to increased goods movement, unless substantial energy
efficiency improvements are being made within a freight mode (e.g., U.S. rail) or across the
logistics supply network. Even assuming that efficiency improvements from economies of scale
reduce energy intensity and emissions rather than being directed to larger and faster ships (e.g.,
containerships), compounding increases in trade volumes outstrip energy conservation efforts
unless technological or operational breakthroughs in goods movement emerge. Furthermore,
proportional relationships between environmental impacts and goods movement trends are
reflected in recent port and regional studies of economic activity and goods transportation,
particularly those focused on Southern California ports (57, 62-64).
3.2 Ship Installed Power as Emissions Trend Indicator
3.2.1 Given that energy used and emissions produced during goods movement increases at a
rate correlated to growth in
activity, a number of proxies may be used to estimate inventory
growth rates. These include: economic activity (GDP and imports/exports value), trade activity
(tons and ton-miles), and fuel usage (sales and estimates). All of these are indirect proxies
(second or higher order) of the activity that produces emissions. Except for fuel usage statistics,
none directly describe power requirements for shipboard power plants (propulsion and auxiliary
engine systems). Best practices for ship emissions inventories typically use power-based (or fuel-
based) emissions factors, because of the implicit proportionality between engine load and
pollutant emissions – especially for uncontrolled sources (20, 21). Therefore, we derive
emissions trends directly from installed power data for cargo ships in the world fleet.
3.2.2 Assumptions we must make to use trends in installed power are rather simple: 1)
international vessels in cargo service generally design power systems to satisfy trade route speed
and cargo payload requirements; in other words, there is no economic reason to design
propulsion systems for containerships, tankers, etc., with more power than their cargo transport
operation requires; 2) international vessels operate under duty cycles that are well understood,
especially at sea speeds, which for most vessel types utilize the majority of installed power as
reflected in best practice methodologies for activity based inventories of energy and emissions
from ships; and 3) ships in commercial cargo service on major trade routes reflect the best fit of
ship design to service requirements; in other words, the trends revealed in installed power of
ships reveals fleet trends in speed and size. With these assumptions, trends in installed power
reveal the correlated trend in energy use by ships.
3.2.3 We considered the trend in average installed power by year of build covering 1970 to
2003 for the world fleet of ships. Where data were missing in the installed power field for some
vessels, we used linear regression statistics within each vessel type associating gross registered
tonnage (GRT) and rated power to fill data gaps. That describes one homogeneous world-
average growth rate, which will result in underestimating future emissions where trade increases
strongly and overestimating emissions in declining trade routes.
Observed trends in average installed power by year-of-build are shown in Figure 5. The
compound annual growth rate (CAGR) for installed power since 1985 is ~10.7% per year, more
than twice the rate of world seaborne trade growth, driven by increases in containership power
which grew at more than 16% CAGR over these two decades. While the slope before 1980
appears similar to the slope after 1985, one can observe the significant fleet restructuring
(particularly for tankers) during the economic recession in the early 1980s. Choosing a period
since 1970 inclusive of the shipping recession, the rate of installed power growth for the world
fleet ~5.1% CAGR; even so, power growth rates for the liner fleet over this period were still
greater than 9% CAGR.
3.3 Growth Forecast Discussion
3.3.1 Admittedly, the quality of forecasts of maritime shipping and trade is somewhat limited
(66), and thus forecasting of environmental impact from shipping is constrained by the quality of
shipping and trade forecasts. This section compares results of alternative measures of growth at
multiple scales, demonstrating general similarity among power-based and other ocean shipping
trends. At the global scale, we evaluate available trends in energy use and/or emissions from
published literature with the seaborne cargo and trade data discussed earlier. Eyring et al
estimate fuel usage and emissions over a historical period from 1950 to 2000 and forecasts for
2020 and 2050 using an activity-based approach describing a BAU scenario and a number of
alternate scenarios combining different ship traffic and technology assumptions (36, 50). For
comparison purposes, we use their BAU scenario for a diesel-only fleet.
3.3.2 We compare world fleet trends in installed power (derived from average power by year of
build) with energy trends (Eyring work and fuel sales), with trade-based historical data (tons and
ton-miles). Activity-based energy results for similar base-years (2001 or 2002) are within close
agreement (20, 25, 36).
This allows us to index trends to nearly the same value and year, to
index trade-based trends similarly, and to compare these with trends in installed power, as
summarized in
Figure 6.
3.3.3 Three insights emerge from this global comparison.
1) Extrapolating past data (with adjustments) produces a range of BAU trends that is
bounded and reveals convergence around a set of similar trends; in other words, while
the range of growth may vary within bounds of a factor of two, one cannot get “any
An exception is work by Endresen et al, that tends to adjust parameters to agree with international marine fuel sales
statistics; their results are within uncertainty ranges described in other work (23, 24, 67).
forecast they want” out of the data. If we consider that global trade and technology
drivers mutually influence future trends, then we may interpret convergence within the
bounds as describing a likely forecast of global shipping activity.
2) World shipping activity and energy use are on track to double from 2002 by
about 2030 (~2015 if one considers seaborne trade since 1985, ~2050 if one
considers Eyring’s BAU trend). Growth rates are not likely to be reduced
without significant changes in freight transportation behavior and/or changes in
shipboard technology.
3) Confirming earlier discussion, trends in installed power are clearly coupled with trends
in trade and energy. This reinforces the analysis of installed power as a proxy for
forecasting growth, not only for use in baseline inventory estimates.
1965 1970 1975 1980 1985 1990 1995 2000 2005
Average Installed Power (kW)
Figure 5. Average installed power in the world fleet by year-of-build (65).
3.3.4 An important question is whether forecasts that directly apply seaborne trade growth rates
to energy and emissions trends should assume any change in the fleet-average energy intensity
over the coming decades. A common belief is technological change improves energy efficiency
in ocean freight transportation (i.e., reduces energy intensity) over time; rationale for this belief
may extend from two historical facts about shipping and energy use: 1) shipping has
traditionally been less energy intensive than other freight modes (especially trucking), and 2)
marine propulsion engineering developments over the past century produced what are arguably
the most fuel-efficient internal combustion (diesel) engines in the world (68).
Depending on change in energy intensity and/or emissions through investments in
economies of scale, fuel conservation measures, or emissions control measures, the rate of
change in energy and emissions could be a modified growth curve from the growth in cargo
activity. If so, one indication would be different rates of change for installed power on ships
providing goods movement compared to changes in cargo volume. In other words, if a fleet of
ships can carry more cargo without a proportional increase in installed power, then it must be
adopting improved technologies (e.g., hull forms, engine combustion systems, plant efficiency
designs) or innovating its cargo operations (e.g., payload utilization).
1950 1960 1970 1980 1990 2000 2010 2020 2030
Seaborne Trade (tons) Seaborne Trade (ton-miles)
OECD HFO Int'l Sales Seaborne Trade (trend since 1985)
World Marine Fuel (Eyring, 2005) Installed Power-This work
Extrapolating trends since ~1980-85
depending on data source
Figure 6. Global indices for seaborne trade, ship energy/fuel demand, installed power.
3.3.5 In fact, the opposite trend is observed over the past 20 to 30 years, where fleet installed
power has grown at rates faster than global trade growth. Rephrasing, ocean shipping may have
become more energy intensive, not more energy conserving. This seemingly counter-intuitive
observation is typical of other transportation modes, particularly onroad freight and passenger
vehicles, and readily explainable in terms of trade globalization and containerization of
international trade. Globalization produced longer shipping routes, and containerization serves
just-in-time (or at least on-time) liner schedules; both of these drivers motivated economic
justification for larger and faster ships which require greater power to perform their service.
Introduction of the fastest, largest ships first occurred on the most valuable trade routes (e.g.,
serving North America and Europe) where economics most justified the higher performing
freight services. Increasingly over the past two decades, ships serving all routes became faster
and larger through intentional expansion and aging fleet transition from prime routes to
secondary markets.
3.3.6 However, technological change can offset this trend, if fleets can achieve greater
efficiencies while increasing installed power. In other words, if fuel economy (energy input) is
not directly proportional over time to energy output (proportional to rated power), then improved
propulsion technologies can explain the decoupling of increased power and fuel use. Ship diesel
engines have achieved substantial improvements in thermal efficiency since the 1970s (69-71),
and fleet turnover has introduced more efficient ships into the world fleet. Using thermal
efficiency as an indicator of this, we construct a sensitivity table that illustrates the potential for
engine designers to offset trade-driven energy requirements through more efficient fleet
modernization. Table 3 shows that replacing an old ship with inefficient engine with a new ship
using the most efficient engine of equivalent power has potential to double fuel economy.
Table 3. Sensitivity analysis of potential improvement in thermal efficiency (fuel economy)
High Low
Efficiency 55% 50% 45% 40% 35% 30% 25%
Low-low-bound 25% 2.20 2.00 1.80 1.60 1.40 1.20 0.00
Low-bound 30% 1.83 1.67 1.50 1.33 1.17 0.00
Low 35% 1.57 1.43 1.29 1.14 0.00
High 40% 1.38 1.25 1.13 0.00
High-bound 45% 1.22 1.11 0.00
High-high-bound 50% 1.10 0.00
Max 55% 0.00
In practice fleet modernization is not as extreme as the illustration in Table 3; fleetwide
average gains in efficiency occur slowly over time because operators replace vessels according
to other economic factors in addition to fuel economy. For example, if one were to assume for
illustration that every new ship replaced the least efficient ship in the fleet since 1970, an upper
bound improvement in energy efficiency can be illustrated (Table 4). Emphasizing that this
over-estimates fleetwide improvements,
Table 4 would suggest that growth rates in installed
power may be adjusted downward by no more than 3% to 4% per year, and with fleet-average
adjustments for fuel economy improvements between 0.5% and 2% more likely given fleet
turnover rates.
The average technology in the fleet may not change that much from its current path
over the next 35 years without strong policy incentives or substantial changes in fleet energy
pricing and supply. A linear growth rate does not match known or expected technology changes
relative to cargo growth; a linear trend in energy use would imply less power required to achieve
cargo throughput – where compounded growth is forecast for cargo volumes. In a BAU context,
fleet propulsion technologies will remain more similar than different to the current profile at least
through 2040. Moving more cargo will require more power and more energy, even with
anticipated thermal efficiency improvements for new engines.
If the economics of globalization and containerization sustained the high rate of growth
in installed power over the past decades, one could expect growth over the next decades more
similar to growth in trade as smaller, slower ships are scrapped thereby reducing variation among
the world fleet characteristics. In other words, the prime routes will continue to attract ship
designs best suited to the balance between freight performance and operating economy. Even so,
energy intensity is (at best) holding constant with increasingly globalization and containerization
of trade logistics. If fuel prices remain high or continue to increase, and if the pace of liner
freight continues to be satisfied by ship speeds ranging between 20-27 knots, then technical
improvements in propulsion and hull design may again be devoted to improving energy intensity.
Table 4. Matrix of maximum annual percent fuel economy gain over from 1970 to 2005,
comparing thermal efficiencies in new versus replaced engines.
High Low
Efficiency 55% 50% 45% 40% 35% 30% 25%
Low-low-bound 25% 6.29% 5.71% 5.14% 4.57% 4.00% 3.43% 0.00%
Low-bound 30% 5.24% 4.76% 4.29% 3.81% 3.33% 0.00%
Low 35% 4.49% 4.08% 3.67% 3.27% 0.00%
High 40% 3.93% 3.57% 3.21% 0.00%
High-bound 45% 3.49% 3.17% 0.00%
High-high-bound 50% 3.14% 0.00%
Max 55% 0.00%
Note: Obtained by dividing the Table 3 matrix by 35, representing 1970 to 2005. Actual fleetwide average
improvements in engine efficiency are smaller.
3.3.7 We believe that the unconstrained exponential trend and the linear trend define bounding
limits for expected change in ship installed power and energy use due to expected trade growth
and fleet technology improvements. Averaging these curves defines an arbitrary middle-growth
trend, which implicitly describes a mix of positive and negative drivers for ship energy
requirements without articulating a detailed scenario of conditional events. After adjustment, we
estimate a world growth trend ranging between 3.8% and 4.5% CAGR. This averaging
conservatively forecasts energy and emissions trends; it implicitly combines two assumptions: 1)
transition to containerization followed by larger, faster containerships will level-off as the world
market scraps older ships; and 2) improvements in propulsion and engine efficiency may
continue, more clearly decoupling energy demand from cargo service than has been observed
over recent decades.
3.3.8 Coincidentally, averaging bounding extrapolations yields between 3.8% and 4.5% CAGR
growth in installed power, nearly the same 4.1% CAGR as observed for past world seaborne
trade. In other words, this explains and confirms the use of seaborne trade growth to project ship
fuel use and emissions, as other studies have done. Therefore, we consider this BAU forecast to
be informed by observed past trends and consistent with adjustments intended to avoid overly
aggressive growth estimates. Consistent with the market-forecast principles reflected in the
IMO study, and given the strong relationship observed between cargo moved (work done)
and maritime emissions (fuel energy used), we adopt for our forecasts the world average
growth rate of 4.1%.
3.4 Emissions Adjustments
3.4.1 While we grow each pollutant inventory by this 4.1% annual rate, we also make several
adjustments for existing IMO regulatory requirements for NOx and sulfur. First, we adjust for
IMO NOx standards and fleet modernization rates introducing cleaner IMO-compliant engines.
Pursuant to MARPOL Annex VI, engines installed on ships constructed on or after 1 January
2000 or engines which undergo a major conversion on or after 1 January 2000 should meet the
requirements of the Technical Code on Control of Emission of Nitrogen Oxides from Marine
Diesel Engines (2). This means that after that date, the increase and replacement of the fleet
should be IMO-compliant for NOx. Using the U.S. EPA estimate that IMO compliant engines
will emit about 17% less NOx emissions than uncontrolled emissions (72), we adjust NOx
emissions for fleet scrapping and new ship orders. This results in a downward adjustment from
uncontrolled projections; for 2012 this adjustment is 3.6% of total NOx emissions. Other studies
suggest the downward adjustment may be greater, perhaps ranging from 5.5-5.6% in 2012 to 8.3-
8.4% in 2020. We accept that there can be a range of NOx reductions attributed to IMO
MARPOL Annex VI, and consider the difference in these estimates to be small.
3.4.2 We also adjust future inventories spatially to reduce forecast emissions in SECA regions
to comply with IMO SECA standards of 1.5% fuel. (The Baltic Sea SECA was implemented
and operational since 19 May 2006; the North Sea SOx Emission Control Area (SECA) comes
into effect on 22 November 2007.) To do this, we forecast emissions globally under the
assumption that world residual fuel sulfur levels remain constant (~2.7% world average), and
multiply emissions in SECA regions by 0.66, representing an average 44% reduction in fuel-
sulfur content. Adjustments are implemented for both the Baltic and North Seas for IMO-SECA
inventories for 2012.
4 Conclusions
4.1 These results help reveal insights important for future policy:
1. There are emissions reductions from an IMO-compliant (1.5% fuel-sulfur SECA) over BAU
trends; and
2. Shipping emissions and resultant health effects and/or other impacts that may be offset in a
base year by implementing a SECA will return to base-year levels within one or two decades.
3. An estimation of benefits from reducing ship emissions can be made using the global data set
we report here, or using more refined regional and local data sets.
4.2 These insights appear robust, regardless of the range in possible forecasts. Using the
forecast trend derived in this work, trade growth offsets baseline (2002) SOx emissions under a
global 1.5% marine fuel-sulfur cap before 2017. Using the range of growth rates reported by
Eyring et al. (2.6%, 3.1%, 3.4%, and 4.0%, annually), emissions within a SECA return to 2002
levels by 2025, 2022, 2020, and 2017, respectively; this range captures the 3% growth rate in the
IMO study on GHGs from ships (4, 5), and is consistent with findings for North America (29).
Figure 7 and Table 5 illustrate projected global sulfur emission trajectories under a
number of different scenarios (for illustration, we assume global reductions take effect in 2010):
business as usual (4.1% growth), with existing IMO ANNEX VI regulations in place;
BAU using 3% growth, consistent with IMO GHG study;
global marine-fuel reduction to 1.5% fuel-sulfur, for both 4.1% and 3% growth rates;
global marine-fuel reduction to 1.0% fuel-sulfur (for 4.1% growth); and
global marine-fuel reduction to 0.5% fuel-sulfur (for 4.1% growth).
2000 2005 2010 2015 2020 2025 2030 2035
Metric Tons SO2 (global)
4.1% Growth (This work) 1.5% Fuel-sulfur at 4.1% Growth
1.0% Fuel-sulfur at 4.1% Growth 0.5% Fuel-sulfur at 4.1% Growth
IMO GHG-study growth (3%) 1.5% Fuel-sulfur at 3% Growth
2002 Baseline
Figure 7. Comparison of BAU SOx trends with global sulfur controls at 1.5%, 1%, and 0.5%.
Table 5. Projected SOx emissions under BAU and various global sulfur-control scenarios.
2002 2010 2015 2020 2025 2030
BAU: 4.1% Growth (This work) 4.72 6.51 7.96 9.73 11.89 14.54
1.5% Fuel-sulfur at 4.1% Growth 4.72 3.62 4.42 5.40 6.61 8.08
1.0% Fuel-sulfur at 4.1% Growth 4.72 2.41 2.95 3.60 4.40 5.39
0.5% Fuel-sulfur at 4.1% Growth 4.72 1.21 1.47 1.80 2.20 2.69
BAU: IMO GHG-study growth (3%) 4.72 5.98 6.93 8.04 9.32 10.80
1.5% Fuel-sulfur at 3% Growth 4.72 3.32 3.85 4.46 5.18 6.00
Note: Implementation of global controls assumed to begin as early as 2010 for illustration. Shaded cells represent
uncontrolled BAU growth rates (this work or IMO GHG study)
4.4 This illustrates that more substantial emissions reductions will last longer into the future
under reasonable growth assumptions. Thus, policies requiring a global 0.5% fuel-sulfur limit or
control technologies achieving equivalent reductions would offset trade growth continuing to the
early 2040s under a 4.1% CAGR. However, a 2010 global sulfur limit of 1.5% would offset
trade growth only until approximately 2017 to 2022, depending on whether a 4.1% or a 3%
growth rate is applied.
4.5 With respect to NOx, growth in emissions has exceeded expected new-engine NOx
reductions resulting from IMO-compliant fleet turnover since application of the existing Annex
VI NOx standards to year 2000 and later ships. This is primarily due to the low scrappage rates
of the vessel fleet; in other words, new engine standards take along time to be fully incorporated
into the fleet due to the lengthy fleet turnover time.
To take a hypothetical example, a 20% reduction in emissions for a fleet that has a 2%
scrappage rate would imply only a 0.4% (20% x 2%) reduction of annual fleet emissions; this
per-year reduction is an order of magnitude smaller than annual emissions growth (~4%) due to
increased seaborne trade activity. Even a 50% reduction in emissions from new vessels leads to
only a 1% overall annual reduction under a 2% scrappage rate scenario. Controls reducing
fleetwide shipping emissions by at least 60% would need to be fully implemented for both new
and existing engines within the next two decades, in order to maintain 2002 global shipping
pollution levels. Achieving fleetwide reductions will involve more aggressive reductions “per
ship” if part of the fleet is left uncontrolled during transition or phase-in years, and reductions of
this magnitude cannot be achieved through new-engine standards alone.
5 References
1. International Maritime Organization, Draft Text of the Annex VI to MARPOL 73/78 for
Prevention of Air Pollution From Ships, M. 39/6, Editor. 1996, International Maritime
Organization: London, UK.
2. International Maritime Organization, Annex VI of MARPOL 73/78 and NOx Technical
Code. IMO-664E ed. 1998, London, UK: International Maritime Organization. 150.
3. International maritime Organization, Air pollution rules to enter into force in 2005. 2004,
International Maritime Organization: London, UK. p. News release.
4. International Maritime Organization, Report on the outcome of the IMO Study on
Greenhouse Gas Emissions from Ships. MEPC 45/8, 2000.
5. Skjølsvik, K.O., et al., Study of Greenhouse Gas Emissions from Ships (MEPC 45/8
Report to International Maritime Organization on the outcome of the IMO Study on
Greenhouse Gas Emissions from Ships). 2000, MARINTEK Sintef Group, Carnegie
Mellon University, Center for Economic Analysis, and Det Norske Veritas: Trondheim,
6. International Maritime Organization and Marine Environment Protection Committee,
Assembly Resolution A.963(23): IMO Policies and Practices Related to the Reduction of
Greenhouse Gas Emissions From Ships. 2004, International Maritime Organization:
London, UK. p. 3.
7. IMO Working Group on Air Pollution from Ships, PREVENTION OF AIR POLLUTION
FROM SHIPS: Report of the Working Group, IMO Submittal, Editor. 2004, MARINE
ENVIRONMENT PROTECTION COMMITTEE, 52nd session, Agenda item 4.
8. IMO Working Group on Air Pollution from Ships, REVIEW OF MARPOL ANNEX VI
AND THE NOx TECHNICAL CODE: Report of the Working Group, IMO Submittal,
AIR POLLUTION, 1st session, Agenda item 2.
RELATED GUIDELINES: Basic information on particulate matter emitted from marine
engines - the fate in the atmosphere of diesel particulate and secondary particulate
matter from NOx and SOx, IMO Submittal, Editor. 2006, INTERSESSIONAL
Agenda item 2.
10. Intertanko, Revisions of MARPOL Annex VI, the NOx Technical Code and Relevant
Guidelines, in Intersessional Meeting of the BLG-WGAP, BLG Working Group on Air
Pollution, IMO submittal, Editor. 2006, INTERSESSIONAL MEETING OF THE BLG
WORKING GROUP ON AIR POLLUTION, 1st session, Agenda item 2: London, UK.
CODE AND RELATED GUIDELINES: Environmental Impact of SOx emissions from
international shipping and considerations on the supply of low sulphur bunker fuels,
WORKING GROUP ON AIR POLLUTION, 1st session, Agenda item 2.
RELATED GUIDELINES: Study on geographically based standard for NOx regulation,
WORKING GROUP ON AIR POLLUTION, 1st session, Agenda item 2.
13. Derwent, R.G., et al., The Contribution from Shipping Emissions to Air Quality and Acid
Deposition in Europe. AMBIO: A Journal of the Human Environment, 2005. 34(1): p.
14. Cofala, J., et al., Interim report: Analysis of Policy Measures to Reduce Ship Emissions in
the Context of the Revision of the National Emissions Ceilings Directive, International
Institute for Applied Systems Analysis, MET.NO, and Entec UK Limited, Editors. 2006,
International Institute for Applied Systems Analysis: Laxenburg, Austria.
POLLUTION FROM SHIPS: The role of ships in meeting the health & environmental
objectives of a new thematic strategy on air pollution, IMO Submittal, Editor. 2006,
POLLUTION, 1st session, Agenda item 2.
16. Commission of the European Communities, Communication from the Commission to the
Council and the European Parliament: Thematic Strategy on air pollution. 2005.
17. Friends of the Earth International, PREVENTION OF AIR POLLUTION FROM SHIPS:
Greenhouse Gas Emissions, IMO Submittal, Editor. 2001, MARINE ENVIRONMENT
PROTECTION COMMITTEE, 47th Session, Agenda item 4.
18. Friends of the Earth International, PREVENTION OF AIR POLLUTION FROM SHIPS:
Reducing Shipping Emissions of Air Pollution - Feasible and Cost-effective Options,
COMMITTEE, 53rd Session, Agenda item 4.
19. UNFCCC and Subsidiary Body for Scientific and Technological Advice, Methodological
issues relating to emissions from international aviation and maritime transport; Note by
the secretariat. 2004, United Nations Framework Convention on Climate Change,
Subsidiary Body for Scientific and Technological Advice: Bonn, Germany. p. 11.
20. Corbett, J.J. and H.W. Koehler, Updated Emissions from Ocean Shipping. Journal of
Geophysical Research - Atmospheres, 2003. 108(D20): p. 4650-4666.
21. ICF Consulting, Best Practices in Preparing Port Emission Inventories: Draft for
Review, I.C. Louis Browning and E.P.S.L. Kathleen Bailey, Editors. 2005, Prepared for
Office of Policy, Economics and Innovation, United States Environmental Protection
Agency: Fairfax, Virginia. p. 39.
22. European Commission and ENTEC UK Limited, Quantification of emissions from ships
associated with ship movements between ports in the European Community. 2002,
European Commission, DG ENV.C1, Rue de la Loi, 200, B-1049: Brussels, Belgium.
23. Endresen, Ø., et al., Emission from International Sea Transportation and Environmental
Impact. Journal of Geophysical Research, 2003. 108(D17): p. 4560.
24. Endresen, Ø., et al., Substantiation of a Lower Estimate for the Bunker Inventory:
Comment on ‘‘Updated emissions from ocean shipping’’ by James J. Corbett and Horst
W. Koehler. Journal of Geophysical Research - Atmospheres, 2004. 109(D23302).
25. Corbett, J.J. and H.W. Koehler, Considering Alternative Input Parameters In An Activity-
Based Ship Fuel Consumption And Emissions Model. Journal of Geophysical Research -
Atmospheres, 2004. 109(D23303): p. 8.
26. Wang, C., J.J. Corbett, and J. Firestone, Modeling Energy Use and Emissions from North
American Shipping: an Application of Ship Traffic, Energy and Environment Model
(revision in review). Environmental Science & Technology, 2006.
27. Lloyd's Register and International Maritime Organization, Marine Exhaust Emission
Quantification Study - Baltic Sea, in MEPC 45/INF.7. 1998.
28. Lloyd's Register, Marine Exhaust Emissions Quantification Study - Mediterranean Sea.
29. Corbett, J.J., C. Wang, and J. Firestone, Estimation, Validation, And Forecasts Of
Regional Commercial Marine Vessel Inventories, Tasks 3 And 4: Forecast Inventories for
2010 and 2020, Draft Report. 2006, State of California Air Resources Board:
Sacramento, CA.
30. Corbett, J.J., C. Wang, and J. Firestone, Estimation, Validation, And Forecasts Of
Regional Commercial Marine Vessel Inventories, Tasks 1 And 2: Baseline Inventory And
Ports Comparison, Final Report. 2006, State of California Air Resources Board and
Commission for Environmental Cooperation in North America: Sacramento, CA, and
Montreal, Quebec. p. 42.
31. Wang, C., A Study of Geographical Characterization of Ship Traffic and Emissions and a
Cost-Effectiveness Analysis of Reducing Sulfur Emissions from Foreign Waterborne
Commerce for the U.S. West Coast, in Marine Policy Program, College of Marine
Studies. 2006, University of Delaware: Newark, DE.
32. Wang, C. and J.J. Corbett, Geographical Characterization of Ship Traffic and Emissions.
Transportation Research Record, 2005. 1909.
33. AMVER, Welcome to AMVER, 2005.
34. Corbett, J.J. and P.S. Fischbeck, Emissions from Ships. Science, 1997. 278(5339): p. 823-
35. Corbett, J.J., P.S. Fischbeck, and S.N. Pandis, Global Nitrogen and Sulfur Emissions
Inventories for Oceangoing Ships. Journal of Geophysical Research, 1999. 104(D3): p.
36. Eyring, V., et al., Emissions from international shipping: 1. The last 50 years. Journal of
Geophysical Research - Atmospheres, 2005. 110(D17): p. D17305.
37. Voluntary Observing Ship Program, The WMO Voluntary Observing Ships (VOS)
Scheme. 2005.
38. ICOADS, The International Comprehensive Ocean-Atmosphere Data Set (ICOADS).
39. Wang, C., A Study of Geographical Characterization of Ship Traffic and Emissions and
Cost-effectiveness of Reducing Sulfur Emissions from Foreign Waterborne Commerce for
the U.S. West Coast, in College of Marine Studies. 2006, University of Delaware:
40. MARAD, Fleet Statistics. 2005.
41. Maritime Business Strategies LLC, Growth of the World Fleet Since WWII. 2005.
42. Worley, S.J., et al., ICOADS release 2.1 data and products. International Journal of
Climatology, 2005. 25(7): p. 823-842.
43. AMVER, Amver Ship Reporting System Manual. 2005.
44. AMVER, Amver Summary Statistics. 2005.
45. Wang, C. and J.J. Corbett, Geographical Characterization of Ship Traffic and Emissions.
Transportation Research Record: Journal of the Transportation Research Board, 2005.
46. Lloyd's Register - Fairplay, Lloyd’s Register of Ships on CD ROM. 2004.
47. International Maritime Organization, Designation of the North Sea area as a SOx
Emission Control Area, MEPC 44/11/4. 1999, International Maritime Organization:
London, UK.
48. Lloyds Register Engineering Services, E.E. Department, and K.A. Lavendar, Marine
Exhaust Emissions Quantification Study - Meditrranean Sea. 2001, Lloyds Register of
Shipping: London, UK. p. 23.
49. Amann, M., et al., Cost-effective Control of Acidifiaction and Ground-Level Ozone. 1996,
International Institute for Applied Systems Analysis: Laxenburg, Austria.
50. Eyring, V., et al., Emissions from international shipping: 2. Impact of future technologies
on scenarios until 2050. Journal of Geophysical Research - Atmospheres, 2005.
110(D17): p. D17306.
51. Ceuster, G.D., B.v. Herbruggen, and S. Logghe, TREMOVE: Description of model and
baseline version 2.41, Draft Report, TRANSPORT & MOBILITY LEUVEN, Editor.
2006, European Commission: Brussels, Belgium.
52. Zeebroeck, B.V., G.D. Ceuster, and B.V. Herbruggen, TREMOVE 2: Maritime model
and runs, TRANSPORT & MOBILITY LEUVEN, Editor. 2006, European Commission:
Brussels, Belgium.
53. U.S. Department of Transportation, Research and Innovative Technology Administration,
and Bureau of Transportation Statistics, North American Freight Transportation. 2006,
Bureau of Transportation Statistics: Washington, DC: 2006.
54. Bureau of Economic Analysis, Table 1.1.6. Real Gross Domestic Product, Chained
Dollars, in National Income and Product Accounts Tables. 2006, U.S. Department of
Commerce: Washington, DC. p. Today is: 11/3/2006 Last Revised on October 27, 2006
Next Release Date November 29, 2006.
55. TRB Executive Committee, G. Giuliano, and S. Godwin, Critical Issues in
Transportation. 2006, Transportation Research Board: Washington, DC. p. 12.
56. Vickerman, J., North American Marine Trade and Transportation Trends, in Transystems
Industry Presentations. 2006: Kansas City, MO.
57. NNI Task Force, Report to Mayor Hahn and Councilwoman Hahn by the No Net
Increase Task Force. 2005, Port of Los Angeles: Los Angeles, CA. p. 603.
58. Mercator Transport Group, Herbert Engineering Corporation, and MDS Transmodal Ltd,
Forecast of Container Vessel Specifications and Port Calls Within San Pedro Bay: Final
Report. 2005, Cunningham Report: Bellevue, WA. p. 139.
59. Meyer Mohaddes Associates Inc., Port of Los Angeles Baseline Transport Study. 2004,
Port of Los Angeles: Los Angeles. p. 70.
60. Parsons Transportation Group, San Pedro Bay Ports Rail Market Study, TM-1b, Draft
Report. 2004, Port of Los Angeles: Los Angeles. p. 30.
61. Port of Los Angeles, et al., Control of Ship Emission in the South Coast Air Basin:
Assessment of the Proposed Federal Implementation Plan Ship Fee Emission Fee
Program. 1994, Port of Los Angeles: Los Angeles, CA.
62. California Air Resources Board, Goods Movement and Ports. 2006, California Air
Resources Board: Sacramento, CA.
63. Southern California Association of Governments, Goods Movement. 2006, Southern
California Association of Governments: Los Angeles, CA.
64. Port of Los Angeles, Goods Movement Projects. 2006.
65. Lloyds Maritime Information System (LMIS), The Lloyds Maritime Database. 2002,
Lloyd's Register - Fairplay Ltd.
66. Stopford, M., Maritime Economics. Second ed. 1997, London, UK: Routledge, Taylor
and Francis Group. 562.
67. Corbett, J.J., Environmental Impacts of Ocean Shipping in a Freight Context: A Summary
of Current Assessments, Future Trends, and Approaches to Mitigation, Draft, Working
Group on Transport, Editor. 2006, Organisation for Economic Cooperation and
Development (OECD): Paris, France.
68. Corbett, J.J., Marine Transportation and Energy Use, in Encyclopedia of Energy, C.J.
Cleveland, Editor. 2004, Elesvier Science: San Diego, CA. p. 745-748.
69. Wartsila NSD, Wartsila 32: Technology Review. 1998, Wartsila NSD Corporation:
Annapolis, MD.
70. Harrington, R.L., ed. Marine Engineering. 1992, Society of Naval Architects and Marine
Engineers: Jersey City, NJ. 953.
71. Heywood, J.B., Internal Combustion Engine Fundamentals. 1988, New York, NY:
McGraw-Hill, Inc. 930.
72. Environmental Protection Agency, 40 CFR Parts 89, 92, and 94: Control of Emissions of
Air Pollution from New CI Marine Engines at or above 37 kW: Final Rule. 1999, Federal
Register: Washington, DC.
... In particular, most of the SSS operates near the coastal and populated areas, and therefore, the negative health effects caused by emission to air are more severe than for deep sea shipping (Hjelle, 2010). According to Corbett et al. (2007) shipping related emissions cause 60,000 annual premature deaths globally, and more than 40% (27,000) of these deaths occur in Europe. ...
... Based on an estimation [60], vessels have contributed to 2.8 million tons of NO x , 1.7 million tons of SO 2 , and 195,000 tons of PM per annum. Each vessel emits between 5 and 6.9 tons of NO x , between 4.7 and 6.5 tons of SO 2 , as well as between 1.2 and 1.6 tons of PM per annum [61], [62]. In short, the maritime transportation system is accounted for 1 billion tons of GHG emission, or 3% of the GHG emission worldwide, from 2007 to 2012 [11]. ...
Full-text available
In this digital age, ports face stiff competition in global supply chain. Smart ports, as high performing ports, utilize information and communications technology (ICT) to provide a wide range of smart applications, resulting in vastly improved vessels and container management among others, which subsequently improve the competitiveness and sustainability of the national economy. While various novel solutions, such as information system and locating system, have been proposed to improve smart port activities, there are several key issues pertaining to ports and port operations that warrant specific attention, particularly greenhouse gases emission, which has accelerated to an alarming level. The urgent need to address such issues is lacking. This article aims to offer a review of the research literature on smart ports, including Internet of Things platform, greenhouse gases emission reduction, energy efficiency enhancement, and so on. The objective is to establish a foundation of existing research conducted on smart ports in order to motivate new research interests in this area. Open issues are also presented to foster new research initiatives on smart ports.
... Despite all the aforementioned advances in global shipping network analysis from multiple perspectives, what remains lacking is a precise cartography of its geographic distribution and spatial evolution. This may appear surprising, given the surge of interest for recent or even instant visualizations of global shipping flow data in various academic works (Corbett et al., 2007) and on online portals such as or to tackle, among other issues, ship emissions, or climate change back in time (Garcia-Herrera et al., 2017). Geographers, whose interest was initially more to understand globalization, trade, and port dynamics were unsurprisingly pioneers in providing the first-ever cartographies of global maritime flows back in the 1940s (Siegfried, 1940;Ullman, 1949). ...
... Even though the atmospheric concentrations of BC in remote areas, such as the Arctic Region, are generally low, their effects on the regional climate may be substantial (e.g. Flanner 2013;Winiger et al. 2016;Sand et al. 2016;AMAP 2011;Corbett et al. 2007). ...
States sought sovereignty over the Arctic Region by discovering the vastness of this uncharted territory. Coastal states developed measurement techniques to take the biggest share of the region. One of these measurement techniques is the Sector Principle that the Canadian senator Pascal Poirier introduced in 1907. Other Arctic states, such as the United States of America and Norway, objected to this technique. However, Russia also adopted and started to use this principle in order to draw Arctic borders in 1926. Before the Sector Principle was introduced, the Median Line Principle had been used and is still in use. Therefore, this new technique created political disputes on the controversial areas in the Arctic. Thus, another problem occurred apart from the unsolved disputed regions; the states also argued their way of measuring and calculating while preparing their Arctic claims to the UNCLOS. The Law of the Sea (1982) brought rules for gaining sovereignty for the 5 coastal states in the Arctic. On the other hand, currently, climate change threat increases the immense geopolitical importance of the region regarding petroleum, oil & gas and especially new shipping routes opportunities. Therefore, sovereignty rights in the region became much more significant for littoral states. Accordingly, this paper will try to see how technical systems have impacted on political claims – especially on shipping routes – and will analyse the history of acquisition of the sovereignty in the Arctic by two measuring techniques. The focus will be on Sector Principle within the sovereignty concept and geopolitical framework.
... Even though the atmospheric concentrations of BC in remote areas, such as the Arctic Region, are generally low, their effects on the regional climate may be substantial (e.g. Flanner 2013; Winiger et al. 2016;Sand et al. 2016;AMAP 2011;Corbett et al. 2007). ...
Maritime transportation covers approximately 90% of the global traffic volumes. The global fleet consists of approximately 100,000 diesel ships, around 250 LNG ships, and a smaller number of methanol or even electric ferries. When it comes to maritime transportation, the Arctic sea route is becoming more and more interesting for the shipping industry as it has been estimated that the Northeast Passage can shorten the travelling distance significantly compared to Suez Canal.
... Even though the atmospheric concentrations of BC in remote areas, such as the Arctic Region, are generally low, their effects on the regional climate may be substantial (e.g. Flanner 2013;Winiger et al. 2016;Sand et al. 2016;AMAP 2011;Corbett et al. 2007). ...
The Sami “minority” of Finland is the smallest indigenous community of this specific Arctic group in Nordic countries. Finnish Sami constitute a cultural, linguistic and territorialized minority. Finland recognized Sami as a “people” in 1995, nevertheless without ratifying the ILO Convention 169 Concerning Indigenous and Tribal Peoples. Besides the fact that Finnish Sami Parliament (Saamelaiskäräjät) has been recognized since 1973, and the Sami linguistic rights have been established since 1982, Sami do not possess territorial rights, especially at economic level. One of the main economic sectors where Sami are active is tourism in Lapland. The debate among the defendants of indigenous rights but also among some Sami prominent leaders are today about the effectiveness of the tourism in the survival of Sami way of life and culture. While some observers denounce the folklorization process of “saminess” through exasperated touristic exploitation, others see in tourism the only way to prevent complete assimilation and fade-out. This chapter will explore the role of tourism in the preservation of Sami culture in Finnish context, by using a field research conducted in July–August 2018 in Inari, Ivalo and Rovaniemi.
... The following GHGs are emitted (per year) to the atmosphere from international shipping [5]: 1.7 million tons of SO 2 , 2.8 million tons of NO x , and 195 000 tons of PM 2.5. In the case of ocean going vessels, they usually emit to the atmosphere between 1.2 and 1.6 metric tons of PM with diameter less than 10 μm, between 4.7 and 6.5 tons of SO 2 , and between 5 and 6.9 tons of NO x [6], [7]. Furthermore, studies reveal that PM 2.5 emitted by ships have been categorized to cause a major effect of cardiopulmonary and lung cancer mortalities in populations exposed in coastal areas [3], [8]. ...
... Shipping, although being an energy efficient mode of transport, is a significant contributor to global emissions such as green house gases (GHG), NO x , SO x and PM (Buhaug et al., 2009;Corbett et al., 2007a;Eyring et al., 2005;Klimont et al., 2017), which are affecting global climate, human health and the environment (Eyring et al., 2010;Corbett et al., 2007b;Kampa and Castanas, 2008). The International Maritime Organization (IMO) is addressing the negative impact of shipping on global climate, human health and the environment by issuing regulations, requiring emission reduction such as the NO x emission limits of MARPOL Annex IV and the Energy Efficiency Design Index (EEDI). ...
Full-text available
In this paper, we present and investigate important factors that influence a vessels fuel consumption during operation and what model fidelity that is required to adequately capture these factors in fuel consumption estimations. Our study focuses on evaluating effect of model fidelity and the methodology used to assess the ship behaviour in operational conditions. Operational performance is affecting cost of operation and estimations of operational performance is used in design development, operational research and as basis for emission analysis, affecting decision making at an operational and technical level. A comprehensive case study is presented where we compare fuel consumption and engine operational profile for time domain and discrete-event simulations, and a static statistical model. Of the factors that have been included in the study, variation in propeller loading and consequently propulsion efficiency is the most prominent physical factor for estimation of required power and fuel consumption. Further, the ability to replicate realistic scenarios using simulators has a significant effect on our understanding of how operational and environmental factors affect operational performance.
Rapid economic growth has contributed to India's rising share in international trade. An increase in the number of ships visiting the port areas would enhance marine emissions, which would be harmful for densely populated port areas in India. The study developed a comprehensive emissions inventory for years 2011–2019 of air pollutants and greenhouse gases (CO2+CH4+N2O) from shipping vessels for Kolkata Dock System. The results revealed significant emissions from ocean-going vessels (OGVs), which currently do not follow the regulatory sulphur limit (0.5% m/m) as stated in Annex-VI and its amendments mandated by the International Maritime Organization's Prevention of Air Pollution from ships. The study revealed a notable increase of total annual marine NOx emissions from 1488 tons to 2074 tons during 2011–2019, a significant growth of 28%. The forecasted values of emissions using time series model ARIMA for year 2019–20 exhibit a rise of 11% from observed value of previous year. Different emission control strategies such as fuel substitution with higher grade fuel or low sulphur diesel fuel and alternative fuel such as liquefied natural gas (LNG) led to notable reduction of emissions for both air pollutants and greenhouse gases. This signified that if emissions were unregulated there would be noticeable growth every year. The study also estimated the environmental impact of air pollutants and greenhouse gases using dispersion model AERMOD. The modeled NOx and SOx concentrations from marine vessels constitute maximum impacts of (15–20)% and (32–55)% of total concentrations at two monitoring sites located within 4 km of Kolkata docks. Emissions controlled modeled runs show marked reductions of NOx, SOx and greenhouse gases concentrations. The directionality of emission controls would enable ports to introduce policies and programs to address these emissions and facilitate sustainable growth.
In this article, a performance evaluation of a novel system solution combining a hybrid turbocharger and a pre-turbine selective catalytic NO x reduction system is carried out. Pre-turbine selective catalytic system are used with marine two-stroke diesel engines to comply with International Maritime Organization Tier III. The system solution focuses on expanding the selective catalytic reduction operation range which is limited by fuel sulphur content by increasing exhaust temperature at low engine loads. The extended operation range is to be achieved while minimizing any fuel consumption penalties. Increasing the operation range brings improvements to emission levels during manoeuvring operations which are often carried out close to populated areas. It also provides flexibility by enabling emission reduction during slow steaming operations in which mitigating fuel consumption penalties is paramount. In addition to system evaluation in still water conditions, furthermore evaluations have been carried out taking into consideration the effect of waves on the system performance. Investigating the effect of operating in waves bring additional insight that is relevant for predicting performance in operational conditions. Analysis of the system solution found that improvements in selective catalytic reduction operation range can be achieved while also improving fuel consumption. Fuel consumption is significantly improved in the high load range. Effect of realistic operation conditions where found to affect performance; however, significant effects are only found for harsh sea states in the load range below the design point.
Full-text available
We present geographically resolved global inventories of nitrogen and sulfur emissions from international maritime transport for use in global atmospheric models. Current inventories of globally resolved sources of natural and anthropogenic emissions do not include the significant contribution of SO2 or NOx from oceangoing ships [Benkovitz et al., 1996]. We estimate the global inventory of ship emissions, using current emission test data for ships [Carlton et al., 1995] and a fuel-based approach similar to that used for automobile inventories [Singer and Harley, 1996]. This study estimates the 1993 global annual NOx and SO2 emissions from ships to be 3.08 teragrams (Tg, or 1012g) as N and 4.24 Tg S, respectively. Nitrogen emissions from ships are shown to account for more than 14% of all nitrogen emissions from fossil fuel combustion, and sulfur emissions exceed 5% of sulfur emitted by all fuel combustion sources including coal. Ship sulfur emissions correspond to about 20% of biogenic dimethylsulfide (DMS) emissions. In regions of the Northern Hemisphere, annual sulfur emissions from ships can be of the same order of magnitude as estimates of the annual flux of DMS [Chin et al., 1996]. Monthly inventories of ship sulfur and nitrogen emissions presented in this paper are geographically characterized on a 2°×2° resolution. Temporal and spatial characteristics of the inventory are presented. Uncertainty in inventory estimates is assessed: the fifth and ninety-fifth percentile values for global nitrogen emissions are 2.66 Tg N and 4.00 Tg N, respectively; the fifth and ninety-fifth percentile values for sulfur emissions are 3.29 Tg S and 5.61 Tg S, respectively. We suggest that these inventories, available via the Ship Emissions Assessment (SEA) web site, be used in models along with the Global Emissions Inventory Activity (GEIA) inventories for land-based anthropogenic emissions and modeled with ocean-biogenic inventories for DMS.
Full-text available
Emission generated by the international merchant fleet has been suggested to represent a significant contribution to the global anthropogenic emissions. To analyze the impacts of these emissions, we present detailed model studies of the changes in atmospheric composition of pollutants and greenhouse compounds due to emissions from cargo and passenger ships in international trade. Global emission inventories of NOx, SO2, CO, CO2, and volatile organic compounds (VOC) are developed by a bottom-up approach combining ship-type specific engine emission modeling, oil cargo VOC vapor modeling, alternative global distribution methods, and ship operation data. Calculated bunker fuel consumption is found in agreement with international sales statistics. The Automated Mutual-assistance Vessel Rescue system (AMVER) data set is found to best reflect the distributions of cargo ships in international trade. A method based on the relative reporting frequency weighted by the ship size for each vessel type is recommended. We have exploited this modeled ship emissions inventory to estimate perturbations of the global distribution of ozone, methane, sulfate, and nitrogen compounds using a global 3-D chemical transport model with interactive ozone and sulfate chemistry. Ozone perturbations are highly nonlinear, being most efficient in regions of low background pollution. Different data sets (e.g., AMVER, The Comprehensive Ocean-Atmosphere Data Set (COADS)) lead to highly different regional perturbations. A maximum ozone perturbation of approximately 12 ppbv is obtained in the North Atlantic and in the North Pacific during summer months. Global average sulfate loading increases with 2.9%, while the increase is significantly larger over parts of western Europe (up to 8%). In contrast to the AMVER data, the COADS data give particularly large enhancements over the North Atlantic. Ship emissions reduce methane lifetime by approximately 5%. CO2 and O3 give positive radiative forcing (RF), and CH4 and sulfate give negative forcing. The total RF is small (0.01–0.02 W/m2) and connected with large uncertainties. Increase in acidification is 3–10% in certain coastal areas. The approach presented here is clearly useful for characterizing the present impact of ship emission and will be valuable for assessing the potential effect of various emission-control options.
The Commercial Marine Vessel Traffic and Air Emissions Model (CMV-TAEM) estimates and geographically represents offshore vessel traffic and emissions based on actual shipping activities. The CMV-TAEM has three modules: ship traffic, ship emissions, and policy analysis. The model establishes empirical ship traffic network on the basis of ship observations derived from the International Comprehensive Ocean-Atmosphere Data Set and shipping activity records. Geographical representations of ship traffic intensities and emissions can be produced through the math-ematic manipulation of matrices of ship traffic network, shipping activity, and ship characteristic data. Overall, although seasonal changes are apparent, the global ship traffic pattern does not change much annually. The ship traffic pattern changes regionally, with a net increase in some areas and net decrease in others. Multiple-year observations are combined to make traffic patterns for major shipping lanes smoother and clearer. Results indicate that 84.5% of global ship traffic occurs north of the equator and two-thirds of global ship traffic within 200 nautical miles of the shore. About 10% of global ship traffic occurs in U.S. coastal waters; shipping along the East Coast accounts for more than one-fifth of the U.S. coastal traffic. Adequate data are available to determine ship activities and ship attributes and to implement the model.
The comments of Endresen et al [Endresen et al., 2004a], offer differing assumptions and clarifying information that may reduce the estimate of ship emissions by Corbett and Koehler [Corbett and Koehler, 2003]. The authors claim to substantiate a lower estimate for marine fuel consumption that more closely agrees with published fuel statistics and therefore agrees with all previous studies that have assumed reported marine fuel statistics are essentially accurate and complete [Benkovitz et al., 1996; Corbett and Fischbeck, 1997; Corbett et al., 1999; Endresen et al., 2003; Olivier and Peters, 1999; Skjølsvik et al., 2000]. Importantly, Endresen et al do not argue fundamentally against the accuracy of an activity-based methodology; their comments entirely focus on whether the input data obtained or developed for our model are reasonable. In this regard, they focus on exactly the issue identified in our uncertainty analysis - namely, the need for accurate engine activity parameters such as operating hours, load profile, and specific fuel consumption. Their revised parameters indeed reduce the estimate of maritime fleet fuel consumption, but not as much as claimed and certainly within our uncertainty bounds. Here we incorporate their assumptions into our activity-based model to confirm our earlier conclusions that fuel used by ships listed in international registries use more fuel than reported in international marine fuel statistics.