Global Trade and Fuels Assessment—
Additional EGA Modeling Scenarios
United States
Environmental Protection
Agency
-------
Global Trade and Fuels Assessment—
Additional EGA Modeling Scenarios
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
Prepared for EPA by
RTI International
and
EnSys Energy & Systems Inc.
EPA Contract No. EP-C-08-008
RTI Project Number 0211577.002.002
v>EPA
United States EPA-420-R-09-009
Environmental Protection .. „„_
Agency MaY2009
-------
CONTENTS
Section Page
1 Introduction 1-1
2 Review of the Bunker Fuel Demand Modeling 2-1
2.1 Summary of the Bunker Fuel Demand Modeling Approach 2-1
2.2 Results of Bunker Fuel Forecasts 2-2
3 U.S. and Canadian EGA Fuel Demand Estimates 3-1
3.1 Summary of Original EGA Fuel Demand Modeling Approach 3-1
3.2 Original EGA Fuel Demand Estimates for 2020 3-2
3.3 Inclusion of Environment Canada's 2020 Fuel Demand Estimates and
Innocent Passage of U.S.-Related Voyages in Canadian Waters 3-2
3.4 2020 Fuel Demand Estimates for the United States and Canada 3-4
4 Revised Business-as-Usual WORLD Model Cases 4-1
4.1 WORLD Model Enhancements 4-1
4.2 WORLD Model Revised Assumptions 4-2
4.2.1 IEO2008 Outlook—Supply/Demand/Price Basis 4-2
4.2.2 Product Quality 4-11
4.2.3 Regional Bunker Demands 4-12
4.2.4 Regulatory Outlook for Bunker Fuels 4-12
4.2.5 U.S. and Canadian EC A Affected Bunker Fuel Volume 4-14
4.2.6 Refinery Capacity and Projects 4-14
4.2.7 Refinery Technology and Costs 4-17
4.2.8 Transportation 4-18
4.3 Prices Input to the WORLD Model 4-20
4.3.1 Marker Crude Price 4-20
4.3.2 Natural Gas and Miscellaneous Input Prices 4-20
in
-------
4.4 Reporting 4-21
5 Revised WORLD Model for EGA 5-1
5.1 Supply-Demand Balance 5-1
5.2 Refining Capacity Additions 5-3
5.3 Refining Economics and Prices 5-6
5.4 EGA Costs 5-10
5.5 Refining and CC>2 Emissions Impacts 5-11
5.6 Marine Fuels Composition 5-13
References R-l
IV
-------
FIGURES
Number Page
2-1. Method for Estimating Bunker Fuel Demand 2-3
2-2. Worldwide Bunker Fuel Use 2-4
2-3. Annual Growth Rate in Worldwide Bunker Fuel Use 2-5
2-4. Worldwide Trade Flows 2-6
2-5. Annual Growth Rate in Worldwide Trade Flows 2-6
2-6. Worldwide IFO380 Use 2-8
2-7. Worldwide IFO180 Use 2-8
2-8. Worldwide MDO-MGO Use 2-9
2-9. Bunker Fuel Used by the International Cargo Fleet Importing to and Exporting
From the United States (by Region) 2-9
2-10. Annual Growth Rate in Bunker Fuel Used by the International Cargo Fleet
Importing to and Exporting from the United States (by Region) 2-10
2-11. Bunker Fuel Used by the International Cargo Fleet Importing to and Exporting
from the United States (by Vessel/Cargo Type) 2-10
2-12. Annual Growth Rate in Bunker Fuel Used by the International Cargo Fleet
Importing to and Exporting from the United States (by Vessel/Cargo Type) 2-11
2-13. U.S. Trade Flows—Imports Plus Exports 2-11
2-14. Annual Growth in U.S. Trade Flows—Imports Plus Exports 2-12
4-1. IEO 2008 Oil Price Projections 4-8
-------
TABLES
Number Page
3-1. 2020 EGA Fuel Demand Estimates, International Trading Ships 3-3
3-2. 2020 EGA Fuel Demand Estimates, All Ships 3-3
3-3. 2020 Canadian EGA Fuel Demand Estimates, All Ships (Including Innocent
Passage of U.S.-Related Voyages) 3-4
3-4. 2020 EGA Fuel Demand Estimates, All Ships, United States and Canada 3-5
4-1. IEO 2008 World Total Liquids Production, by Region and Country, Reference
Case, 1990-2030 (Million Barrels Oil Equivalent per Day) 4-3
4-2. IEO 2008 World Unconventional Liquids Production, by Region and Country,
Reference Case, 1990-2030 (Million Barrels Oil Equivalent per Day) 4-4
4-3. IEO 2008 World Total Liquids Production, by Region and Country, High Price
Case, 1990-2030 (Million Barrels Oil Equivalent per Day) 4-5
4-4. IEO 2008 World Unconventional Liquids Production, by Region and Country,
High Price Case, 1990-2030 (Million Barrels per Day) 4-6
4-5. IEO 2008 World Liquids Consumption, by Region, Reference Case, 1990-
2030 (Million Barrels Oil Equivalent per Day) 4-7
4-6. IEO 2008 World Liquids Consumption, by Region, High Price Case, 1990-
2030 (Million Barrels Oil Equivalent per Day) 4-8
4-7. Annual Product Growth Rates Applied in the 2020 Reference and High Price
Cases 4-10
4-8. World Regional Bunker Sales 4-13
4-9. United States and Canada Affected Bunker Fuel Demand, 2020 4-15
4-10. Refinery Base Capacity and Assessed Additions 4-16
4-11. Key Price and Cost Premises for 2020 WORLD Cases 4-19
4-12. WORLD Tanker Rates 4-19
5-1. WORLD Reference and EGA Case Inputs and Summary Results—Supply,
2020 5-2
5-2. WORLD Reference and EGA Case Inputs and Summary Results—Demand,
2020 5-3
5-3. WORLD Input Refinery Base Capacity and Assessed Additions 5-4
5-4. WORLD Refinery Capacity Additions, Investments, and Utilizations 2020 5-5
5-5. 2020 Price Results From WORLD Case Analyses 5-7
VI
-------
5-6. 2020 Price Results From WORLD Case Analyses 5-8
5-7. 2020 WORLD Output Product Price Differentials 5-9
5-8. 2020 WORLD Output Product Supply Costs and CO2 Emissions 5-11
5-9. 2020 WORLD Output Refinery Investment, Throughput, and CO2 Emissions,
United States and Canada 5-12
5-10. 2020 WORLD Output Product Supply Costs, United States and Canada 5-13
5-11. 2020 Projected Bunker Fuel Demands 5-14
5-12. WORLD Projected Bunker Fuel Compositions 5-16
vn
-------
SECTION 1
INTRODUCTION
The U.S. Environmental Protection Agency (EPA), along with other regulatory bodies in
the United States and Canada, is considering whether to designate an Emission Control Area
(EGA) along the North American coastlines, as provided for by MARPOL Annex VI. This
addition to the international MARPOL treaty went into effect on May 19, 2005, and was
amended in October 2008. Annex VI places global limits on fuel sulfur levels and exhaust
emission rates for NOX In addition, countries participating in the treaty are also permitted to
request designation of EGAs, in which ships must comply with more stringent limits on fuel
sulfur levels and NOX emissions. The Baltic and North Sea areas have already been designated as
EGAs, and the effective dates of compliance in these bodies of water were 2006 and 2007,
respectively. The current fuel sulfur limit in EGAs is 1.5%. This limit is reduced to 1.0% Sulfur
in March 2010 and further reduced to 0.1% S in January 2015.
To evaluate possible recommendations regarding a North American EGA, EPA is
performing a thorough examination of potential responses by the petroleum-refining and ocean-
transport industries to such a designation, along with any resulting economic impacts. EPA
contracted with RTI International to provide a foundation for these recommendations through
developing the knowledge, data, and modeling capabilities needed for such an analysis; to assess
technology alternatives for reducing sulfur emissions from ships; and to estimate the impact a
EGA designation would have on the petroleum-refining and ocean transport industries.
Under the prior EPA contract number EP-C-05-040, the RTI, EnSys, and Navigistics
team undertook an extensive analysis of the potential impacts of a U.S. EGA. That study is
documented in the November 2008 report Global Trade and Fuels Assessment—Future Trends
and Effects of Requiring Clean Fuels in the Marine Sector (U.S. Environmental Protection
Agency [EPA], 2008). (Hereafter in this new report, the earlier document is referred to as the
"2007 Study.") The 2007 Study was initiated in 2006 and used premises and forecasts available
at that time to establish business-as-usual (BAU) outlooks, which in turn were used to establish
EGA cases and costs. Since then, global energy markets and economies have seen dramatic
changes.
The objective of this report is to supplement the prior work with additional analysis and
refinery modeling to support EPA's EGA implementation analyses. First, RTI revised intra-ECA
fuel consumption estimates to include estimates generated by Environment Canada (EC) and
innocent passage of ships bound for U.S. ports in Canadian waters. The original analysis
1-1
-------
included only innocent passage of U.S.-flagged ships off the coast of British Columbia. EC
provided RTI with fuel consumption estimates based on Canadian port calls. RTI included this
data in the EGA fuel consumption estimates and new innocent passage estimates for U.S.-related
voyages passing in Canadian waters. This revised fuel consumption demand for both the United
States and Canada was supplied to EnSys for additional WORLD modeling cases.
Second, this new study and report were to reassess the potential cost and other impacts of
a U.S. and Canadian EGA under updated assumptions and recognizing current uncertainty in the
outlook for the world oil market and refining. The original 2007 Study analyzed scenarios
against horizons of 2012 and 2020. This new 2008 study focused solely on 2020 but incorporated
EGA analyses against two global projections for 2020, one based on the Energy Information
Administration (EIA) International Energy Outlook 2008 (U.S. Department of Energy [DOE],
2008) reference case and the second on the corresponding high price case. The results from
additional modeling scenarios provide projections of the cost, refining, and CO2 emissions
impacts of a U.S. and Canadian EGA in the year 2020 and against two different world oil price,
supply, and demand outlooks.
1-2
-------
SECTION 2
REVIEW OF THE BUNKER FUEL DEMAND MODELING
This section discusses the demand side of the marine fuels market. The consumption
forecasts in this section provide a baseline for the WORLD model, against which the shipping
industry's possible response to the adoption of a U.S. or North American EGA regulation could
be evaluated.
2.1 Summary of the Bunker Fuel Demand Modeling Approach
The final report, entitled Global Trade and Fuels Assessment—Future Trends and Effects
of Requiring Clean Fuels in the Marine Sector (EPA, 2008), presents a detailed description and
discussion of the bunker fuel demand modeling approach and its results. The following
discussion is a summary that is presented for readers' convenience. In general, the approach used
to estimate marine bunker fuel use can be described as an "activity-based" approach with a focus
on the international cargo vessels that represent the majority of fuel consumption. Components
of the estimation included the following:
• identifying major trade routes,
• estimating volumes of cargo of various types on each route,
• identifying types of ships serving those routes and carrying those cargoes,
• characterizing types of engines used by those ships, and
• identifying the types and estimated quantities of fuels used by those engines.
Implementing this approach involves combining information from a variety of sources:
data on the existing fleet of shipping vessels from Clarksons (2005), information from Corbett
and Wang (2005) and various industry sources on engine characteristics, and projections of
future global trade flows from Global Insights (2005). The data on vessels and engines provide a
characterization of fuel use associated with delivering a particular load of cargo, and the data on
trade flows control how many times, and over what distances, these loads have to be delivered.
Estimating fuel consumption through an activity-based methodology that combines data
on specific vessels with data on engine characteristics is similar to the approaches used in
Corbett and Koehler (2003, 2004), Koehler (2003), Corbett and Wang (2005), and Gregory
(2006). The approach in this report extends previous analyses by linking these ship data to
projections of worldwide trade flows to determine the total number of trips undertaken in each
year—and hence, fuel use—rather than using estimates of the number of hours a ship/engine
typically runs in a year.
2-1
-------
Accordingly, the model estimates fuel consumption based on an underlying economic
model's projections of international trade by commodity category (Global Insights, 2005).
Demand for marine fuels is derived from the demand for transportation of various types of
cargoes by ship, which, in turn, is derived from the demand for commodities that are produced in
one region of the world and consumed in another. The flow of commodities is matched with
typical vessels for that trade (characterized according to size, engine horsepower, age, specific
fuel oil consumption, and engine load factors). Next, typical voyage parameters are assigned,
including average ship speed, round-trip mileage, tonnes of cargo shipped, and days in port. Fuel
consumption for each trade route and commodity type thus depends on commodity projections,
ship characteristics, and voyage characteristics.
Figure 2-1 illustrates the broad steps involved in developing baseline projections of
marine fuel consumption. It is a multistep process that relies on data and forecasts from
numerous sources, some of which are listed above, to inform the projections. The flow chart in
the figure illustrates the relationships to be profiled in characterizing baseline marine fuel
consumption by cargo vessels.
Also, although the focus of this analysis of bunker fuel forecasts is on projecting use by
vessels carrying cargo among international ports, it includes other vessel types when estimating
total demand for bunker fuels, as discussed below. These vessel types include passenger vessels,
such as ferries and cruise ships; service vessels, such as tugs and offshore supply vessels (OSVs);
and military vessels.
2.2 Results of Bunker Fuel Forecasts
Figure 2-2 shows estimated worldwide bunker fuel consumption by vessel type. Fuel
consumption in year 2001 was equal to 278 million tonnes, which can be compared to the
estimate in Corbett and Koehler (2004) of 289 million tonnes. By 2020, bunker fuel demand
approaches 500 million tonnes per year. Note that the "historical" bunker fuel data shown going
back to 1995 are also model estimates based on historical Global Insights trade flows.
(Comparisons of these estimates with others in the literature are discussed in more detail in
Section 4.2 of the 2007 Study, given their importance to the modeling of the petroleum-refining
industry in the WORLD model.)
2-2
-------
Ship Analysis: by Vessel Type and Size Category
Inputs
Outputs
Deadweight for all Vessels of
Given Type & Size3
Average Cargo
Carried (Tons)
«•{ A
Horsepower, Year of Build
for all Vessels of Given
Type & Size3
Specific Fuel Consumption
(g/SHP-HR) by Year of Buildb
Average Daily Fuel
Consumption
(Tons/Day)
Engine Load Factors0
Average Daily Fuel
Consumption (Tons/Day)
- Main, Aux. Engine at Sea
-Aux. Engine in Port
4 B )
Trade Analysis: by Commodity and Trade Route
Inputs
Outputs
Average Ship Speed0
Round Trip Mileage
Tons of Cargo Shipped6
Days at Sea and in
Port, per Voyage
Average Cargo Carried/''
per Ship Voyage (. '
Number of Voyages
Total Estimated Bunker Fuel Demand
Average Daily Fuel Consumption
(Tons/Day)
- Main, Aux. Engine at Sea
-Aux. Engine in Port
B
Total Days at Sea
and in Port ^
Bunker Fuel
Demand
Driven by changes in engine efficiency.
Driven by growth in
commodity flows.
a - Clarksons Ship Register Database
b - Engine Manufacturers' Data, Technical Papers
c - Corbett and Wang (2005) "Emission Inventory Review: SECA Inventory Progress Discussion"
d - Combined trade routes and heavy leg analysis
e - Global Insight Inc. (Gil) Trade Flow Projections
Figure 2-1. Method for Estimating Bunker Fuel Demand
2-3
-------
600 !
500
400
300
200 -
100
d Military Vessels
• Natural Gas
DDryBulk
• Passenger Ships
DPetroleum
D General Car go
• Fishing Vessels
D Chemicals
H Container
D Other
BCrudeOil
Figure 2-2. Worldwide Bunker Fuel Use
Source: Global Insights, Inc. 2005. World Trade Service. Customized Data Export.
Figure 2-3 shows the annual growth rates by vessel type/cargo that underlie the
projections in Figure 2-2. The projected total annual growth is generally between 2.5% and 3.5%
over the time period between 2006 and 2020 and generally declines over time, resulting in an
average annual growth of around 2.6%. As shown in the "container" categories in Figures 2-2
and 2-3, fuel consumption by container ships is the fastest-growing component of worldwide
bunker fuel demand; in 2004, consumption by container ships was around 75 million tonnes,
growing to 87 million tonnes by 2006 and close to 180 million tonnes by 2020. (The historical
estimates can be compared to Gregory [2006], which places container-ship consumption in 2004
at 85 million tonnes, based on installed power.) While overall growth is less than 3% per year,
growth in container-ship demand remains above 5% per year on an average annual basis for the
next 15 years. Across all vessel types, growth in bunker fuel consumption is somewhat lower
than the worldwide GDP growth forecasts from EIA (International Energy Outlook 2005 [DOE,
2005]) of around 3.9% per year but higher than International Energy Agency (IEA) estimates of
overall fuel consumption growth (around 1.6% in the World Energy Outlook 2005). Demand
estimates reflect the long-term outlook based on forecasted needs and not short-term economic
scenarios. The estimated growth in marine bunker demand over the next 15 yearslis consistent
with the historical growth rate of 2.7% per year observed in LEA data between 1983 and 2003.
2-4
-------
10%
-2%
vo
00
-O- Total
-A- Crude Oil
— Other
o r-i
o o
-*~ Container
Chemicals
Fishing Vessels
^H ^H *-H
O O O
—*~ General Cargo -•" Dry Bulk
H(- Petroleum ^~ Natural Gas
Passenger Ships Military Vessels
c
r
c
r
Figure 2-3. Annual Growth Rate in Worldwide Bunker Fuel Use
Source: Global Insights, Inc. 2005. World Trade Service. Customized Data Export.
Growth in fuel use by container ships and the overall contribution by these vessels to
worldwide demand are driven by several factors. The first is overall growth in worldwide GDP
mentioned above. This growth leads to increases in international trade flows over time (shown in
Figures 2-4 and 2-5). These figures illustrate that, although container trade is smaller in total
volume than other categories, it is the fastest-growing component of the trade flows. Measuring
trade flows in tonnes of goods, as shown in Figure 2-4, also does not provide a good proxy for
the fuel consumption needed to transport the goods. Liquids and dry bulk are much denser than
container goods, for example. It is estimated that utilization rates for container ships (comparing
deadweight tonnes of capacity to actual cargo transported) are around 50%. Thus, it takes
approximately twice as many ships to transport the same amount of container tonnes compared
to liquid/dry bulk tonnes. This relationship tends to influence total bunker fuel use and weight it
toward container trade. In addition, growth rates in particular trade flows, such as Asia to the
United States, will also influence overall fuel consumption, especially as related to container
ships, as discussed below in relation to U.S. regional trade flows.
2-5
-------
9,000
8,000
@ Container ID General Cargo D Dry Bulk H Crude Oil D Chemicals D Petroleum D Natural Gas
Figure 2-4. Worldwide Trade Flows
Source: Global Insights, Inc. 2005. World Trade Service. Customized Data Export.
10%
0%
§
-O- Total
-*- Crude Oil
1 Container
Chemicals
1 General Cargo
1 Petroleum
• Dry Bulk
• Natural Gas
Figure 2-5. Annual Growth Rate in Worldwide Trade Flows
Source: Global Insights, Inc. 2005. World Trade Service. Customized Data Export.
2-6
-------
Figures 2-6 through 2-8 show estimated consumption of specific grades of bunker fuels
from Figure 2-2.
Figures 2-9 through 2-12 present estimates of fuel use by the international cargo fleet
engaged in delivering trade goods to and exporting trade goods from the United States. These
estimates comprise part of the total worldwide bunker fuel use shown in Figure 2-2 and do not
include fuel used for domestic navigation. The results in Figure 2-9 show estimated historical
bunker fuel use in year 2001 of around 47 million tonnes. (Note that, while this fuel is used to
carry trade goods to and from the United States, it is not necessarily all purchased in the United
States and is not all burned in U.S. waters.) This amount grows to over 90 million tonnes by
2020, with the most growth occurring on trade routes from the East Coast and the "South
Pacific" region of the West Coast.
Figure 2-10 shows the annual growth rate projections for the fuel consumption estimates
in Figure 2-9. The South Pacific and East Coast regions of the United States are growing the
fastest, largely as the result of container ship trade (see Figures 2-11 and 2-12). Overall, the
average annual growth rate in marine bunkers associated with future U.S. trade flows is 3.4%
between 2005 and 2020. This growth rate is somewhat higher than worldwide totals but is
similar to estimated GDP growth in the United States of 3.1% between 2005 and 2020 (DOE,
2006) and is influenced by particular components of U.S. trade flows.
The projected growth rate in bunker fuel consumption related to U.S. imports and exports
is driven by container ship trade (see Figures 2-13 and 2-14), which grows by more than 4% per
year. U.S. trade volumes are also influenced by high worldwide growth in GDP and resulting
demand for U.S. goods. Along with the fact that container ships use a disproportionately large
amount of fuel to move a given number of tonnes of cargo, fuel use by container ships is also
influenced by shifts in trading routes over time. In the future, trade is expected to shift to the
Pacific region (an increase in Asia-U.S. routes), which causes the average distance per voyage to
increase. Thus, while ship efficiency is increasing over time as older ships retire, this effect is
dominated by the increase in voyage distance, leading to higher bunker fuel growth.
2-7
-------
400
350
n Military Vessels •Passenger Ships •Fishing Vessels QOther
• NaturalGas DPetroleum BChemicals DCrudeOil
DDryBulk IE General Car go BContainer
o
r^
o
Figure 2-6. Worldwide IFO380 Use
Source: Global Insights, Inc. 2005. World Trade Service. Customized Data Export.
o
o
o
El Military Vessels
• NaturalGas
DDryBulk
•ft
o
o
• Passenger Ships
DPetroleum
H General Car go
o
o
• Fishin g Vessels
DChemicals
HContainer
•ft
o
Q Other
HCrudeOil
o
r^<
o
Figure 2-7. Worldwide IFO180 Use
Source: Global Insights, Inc. 2005. World Trade Service. Customized Data Export.
2-8
-------
120
100
g
B
o
o
§
(S
El Military Vessels
• Natural Gas
DDryBulk
|