Proposal to Designate an Emission

            Control Area for Nitrogen Oxides,

            Sulfur Oxides and Particulate Matter


            Technical Support Document
&EPA
United States
Environmental Protection
Agency

-------
                Proposal to Designate an Emission
                Control Area for Nitrogen Oxides,
              Sulfur Oxides and Particulate Matter

                   Technical Support Document
                         Assessment and Standards Division
                         Office of Transportation and Air Quality
                         U.S. Environmental Protection Agency
&EPA
United States                               EPA-420-R-10-013
Environmental Protection                        .   4.omn
Agency                                  August 2010

-------
                               Table of Contents
EXECUTIVE SUMMARY

CHAPTER 1: DESCRIPTION OF POPULATION AND ENVIRONMENTAL AREAS AT
RISK	1-2
1.1   THE COMMONWEALTH OF PUERTO RICO	1-2
1.2   U.S. VIRGIN ISLANDS	1-7
1.3   CONCLUSION	1-9

CHAPTER 2:  EMISSION INVENTORY	2-2
2.1   Introduction	2-2
2.2   Description of Ships Included in the Analysis	2-2
2.3   Inventory Methodology	2-3
2.4   Development of 2002 Inventories	2-4
2.5   Development of 2020 Inventories	2-20
2.6   Inventories for Proposed ECA	2-30
2.7   Other Inventories	2-32
2.8   Conclusion	2-33
APPENDICES  	2-37
APPENDIX 2A: PORT COORDINATES	2-37
APPENDIX 2B: PORT METHODOLOGY AND EQUATIONS	2-38
APPENDIX 2C: EMISSION INPUTS TO STEEM	2-46
APPENDIX 2D: GROWTH FACTOR DEVELOPMENT	2-52

CHAPTER 3:  IMPACTS OF SHIPPING EMISSIONS ON AIR QUALITY, HEALTH
AND THE ENVIRONMENT	3-2
3.1   Pollutants Reduced by the ECA	3-2
3.2   Health Effects Associated with Exposure to Pollutants Reduced by the ECA	3-5
3.3   Ecosystem Impacts Associated with Exposure to Pollutants Reduced by the ECA. 3-13

-------
CHAPTER 4:  COSTS	4-2
4.1   Fuel Production Costs	4-2
4.2   Operational Costs	4-7
4.3   Vessel Costs	4-10
4.4   Total Estimated ECA Costs in 2020	4-10
4.5   Cost Effectiveness	4-11

CHAPTER 5:  ECONOMIC IMPACTS	5-2
5.1   The Purpose of an Economic Impact Analysis	5-3
5.2   Economic Impact Analysis Methodology	5-3
5.3   Expected Economic Impacts of the Proposed ECA	5-5
APPENDICES 	5-14
                                                                             in

-------
Executive Summary

Introduction

       On June 25, 2010, the United States submitted a proposal (MEPC 61/7/3) to the
International Maritime Organization (IMO) to designate an Emission Control Area (EGA) for
specific portions of the coastal waters around Puerto Rico and the U.S. Virgin Islands. This
action would control emissions of nitrogen oxides (NOx), sulfur oxides (SOx), and particulate
matter (PM) from ships.  Designation of the proposed EGA would significantly reduce emissions
from ships and deliver substantial benefits to the local population, as well as to marine and
terrestrial ecosystems

       Also submitted to the IMO is an Information Document (MEPC 61-INF.9), which
provides a complete analysis of how the proposal addresses the IMO's approval criteria. This
Technical Support Document provides additional detail on the technical analyses supporting
those submissions.

Description of Population and Areas

       Chapter 1 provides a description of The Commonwealth of Puerto Rico and the U.S.
Virgin Islands.  This includes information about geography, population and population densities,
special ecological areas, and the economies of these islands, and supplements information
contained in the Information Document prepared for the proposal  package.  The combination of
people  and sensitive ecosystems being located in close proximity to ports and areas of ship
activity with the high levels of ship activity in this area mean that emissions from ships are
contributing to ambient concentrations of air pollution and to adverse environmental impacts in
Puerto Rico and the U.S. Virgin Islands.

Emission Inventory

       Chapter 2 describes how U.S. emission inventories were developed to describe air
emissions from ships  operating in waters within the proposed EGA.  These inventories provide
the foundation upon which all the subsequent analyses were built, and address Criterion 6 of
Section 3, Appendix III to MARPOL Annex VI.  Beyond the level of detail provided in MEPC
61/7/3, Chapter 2 explains how the inputs were developed and what assumptions were made in
assessing what the emissions are from ships currently (2002 base year), what the emissions
would look like in 2020 without the proposed EGA, and what reductions can be  expected from
the proposed EGA.

       Chapter 2 describes the "bottom-up" methodology that was used, based on the latest state
of the art models and  inputs. This chapter describes which port-related emissions were included
and why, and how emissions were obtained for ships  while underway in U.S. waters.  This
chapter explains in great detail each parameter that went into the modeling and analyses,
including which ships are included, which fuels are used by those ships, which other (non-ECA)
emission controls are  in place for each scenario, and what growth rates are expected,
incorporating forecasts of the demand for marine transportation services in 2020.
                                                                                     IV

-------
Impacts of Emissions on Human Health and the Environment

       Chapter 3 describes the impact of ships' emissions on human health and ecosystems and
supports Section 5.4 of the Information Document.  Chapter 3 includes a description of the
pollutants proposed for control in the U.S. Caribbean EGA. The proposed EGA would not only
reduce direct emissions of NOx,  SOx and PM, but also secondarily formed ambient PM and
ground-level ozone.  Section 3.1  describes the nature of these pollutants, formation processes,
and relationship to ship emissions. Section 3.2 presents the health effects associated with
exposure to NOX,  SOX, PM and ground-level ozone, summarizing the key scientific literature.
Section 3.3 describes the impacts of emissions from ships on terrestrial and aquatic ecosystems
such as acidification, nutrient enrichment, ozone uptake and visibility degradation.

Cost Analyses

       Chapter 4 describes our estimates of the costs associated with the reduction of NOX, SOX,
and PM emissions from ships, not only to the shipping industry but also to marine fuel suppliers
and companies who rely on the shipping industry. This chapter provides additional detail
regarding the analyses conducted in support of Criteria 7 and 8 of Section 3, Appendix III to
MARPOL Annex  VI.  This chapter describes the analyses used to evaluate the cost impact of
Tier III NOx requirements combined with low sulfur fuel use on vessels operating within the
proposed EGA, including estimates of low sulfur fuel production costs and operating costs. This
chapter also presents cost per ton estimates for EGA-based NOx and fuel sulfur standards and
compares these with the costs of established land-based control programs.

Economic Impact Analysis

       Chapter 5 examines the economic impacts of the projected EGA costs on shipping
engaged in international trade. This chapter provides additional detail in support of Criterion 8
of Section 3, Appendix III to MARPOL Annex VI.  This chapter describes the econometric
methodology that  was used in estimating two aspects of the economic impacts: social costs and
how they are shared across stakeholders, and market impacts for the new engine and new vessel
markets.

-------
                                                                     Chapter 1

CHAPTER 1:   DESCRIPTION OF POPULATION AND ENVIRONMENTAL AREAS
AT RISK      	1-2
1.1   The Commonwealth of Puerto Rico	1-2
1.2   U.S. Virgin Islands	1-7
1.3   Conclusion	1-9
                                     1-1

-------
Emission Control Area Designation
CHAPTER 1: Description of Population and Environmental Areas
                   at Risk

       The proposed Emission Control Area (EGA) consists of the Commonwealth of Puerto
Rico and the U.S. Virgin Islands.  These islands are unincorporated territories of the United
States. They are situated where the Western Atlantic Ocean meets the Caribbean Sea, among the
chain of islands called the Antilles of the West Indies. This chapter describes the geography,
population, and economy of each of these U.S. territories.

  1.1 The Commonwealth of Puerto Rico

       The Commonwealth of Puerto Rico is an archipelago of the easternmost islands of the
Greater Antilles. Puerto Rico consists of the main island of Puerto Rico and several smaller
islands including Vieques, Culebra, Mona and Monito, Desecheo, Caja de Muertos, and La Isleta
de San Juan.

       The main island of Puerto Rico extends a maximum of about 180 km  east-west and 65
km north-south, with a total land area of 9,000 square km.  Puerto Rico's mountainous interior
rises to  a peak altitude of 1,339 meters at Cerro de Punta, part of La Cordillera Central mountain
range. Together with the Sierra de Luquillo and Sierra de Cayey ranges, mountainous terrain
covers most of the interior and roughly two-thirds of the entire island. An area of rugged
limestone or karst topography extends to the north of La Cordillera Central, occupying a large
portion  of north-central and northwest of the island. Over much of the remainder of the island,
flat coastal plains separate the mountains from the sea.1 Figure 1-1 identifies  the location of
these major landforms on Puerto Rico and the Virgin Islands.
 Mod tied Irom:
 Mon-of, W.H., 1970, Ths karat lindforma cf Rue-to Ri:o: U S. Geological Survey
  Prafesi onal Paper SE9, 99 p.                              £ a ,-  b b e ^ !
 Jordan, D.G.. 1372, Land-use effect on the water regimei cf the U.S. Virf in
  Islands:U.S.Geological Strvay Professional PaperSOO-D, D. D211-D216
 Base mod tied Irom U.S. G*-alojical Suivey digital data.
       EXPLANATION

     I  A lea of hirst topography

       Mountainous area
                                                              •
                                                               tk   SL ihifn^
                                                             Gildn
                                      il. trait ff^-
Discontinuous coaslal plain

Lcw-Eyng to gcnty EDling plain
A pptoxtmstt ax is of mountain chain
                   Figure 1-1: Topography of Puerto Rico and the Virgin Islands1

       The topography of Puerto Rico influences weather patterns, particularly in the mountain
ranges, in part by lifting the moist air masses and increasing rainfall. As a result, annual
                                            1-2

-------
                                                                                Chapter 1

precipitation varies directly with altitude.1 This phenomenon occurs especially on the eastern
side on the main island, as the prevailing winds from the east carry the moist air towards the
mountains. Puerto Rico's climate is marine tropical with seasonal variation in precipitation.2

       Puerto Rico's climate and geography contribute to a great deal of natural diversity.
Ecosystems range from bioluminescent bays and tropical mangrove swamps to coral reefs.
Puerto Rico has two areas classified by UNESCO as World Biosphere Reserves: Luquillo and
Guanica. The Luquillo Mountains in the northeast of Puerto Rico contain the only protected
tropical rainforest in the United States forest system, El Yunque.  The Guanica Reserve, located
in the southwest of the island, consists of several mangrove cays and a subtropical dry forest.3
Furthermore, Mona and Monito Island, 70 km off the west coast of the main island, has been
denoted as the Galapagos of the Caribbean.  It contains sensitive ecosystems and provides habitat
for several endangered species, for example the Mona Island ground iguana and the hawksbill
turtle.4

       The human population of Puerto Rico is densely clustered near the  coasts into highly
urbanized communities. Approximately 70 percent of the population lives within 10 kilometers
of the shore. Rural areas account for only a marginal percentage of the total population in Puerto
Rico.5 Figure 1-2 illustrates the high population density along the coast. This map also shows
the co-location of commercial ports with the most densely populated regions.  As a consequence
of their proximity to ports and the coastline, inhabitants of the islands Vieques, Culebra, and the
main island of Puerto Rico are clearly affected by ship  emissions.
                                                  San Juan Port
                                                 "~ Guayama Port

                  Figure 1-2 Port locations and population density of Puerto Rico6
       The U.S. Census Bureau estimates the total population of Puerto Rico at approximately
4.0 million people for 2009.7  If Puerto Rico were a U.S. state, it would rank 27th in population
size, after Kentucky and before Oregon.  Additionally, Puerto Rico ranks behind all 50 states
except Delaware and Rhode Island in land area, at approximately 9,000 square kilometers.  This
gives Puerto Rico an average population density of about 440 people per square kilometer,
second in the United States after New Jersey.8  Only around 20 countries in the world have a
higher population density.9 Not only is Puerto Rico densely populated, but the population is
heavily distributed among groups that are especially sensitive to air pollution, particularly
                                           1-3

-------
Emission Control Area Designation
children and the elderly. Over 1.5 million, or 38 percent of the population, is under 18 or over
the age of 65.
10
       There are approximately 1.2 million households in Puerto Rico. The average household
size is 3.2 people, with families making up 75 percent of the households. Just three percent of
those living in Puerto Rico are foreign born. Of the native-born population, more than 90 percent
was born in the municipioA where they currently reside.10

       Of the population above 16 years old, 47 percent is classified as in the labor force, which
is far lower than all US states. Major occupations include management/professional (29 percent
of the employed population), sales/office (28 percent) and service occupations (19 percent). The
leading industries by employment are education/health care/social assistance (22 percent of the
employed population), retail trade (13 percent) and manufacturing (11  percent). In 2008,
approximately 16 percent were unemployed.10

       About 45 percent of the population was classified at or below the poverty threshold in
2008. This is far above the national average in the United States of 13 percent. The poverty rate
for children in Puerto Rico is even higher: 56 percent.10  Moreover, much of this population lacks
adequate access to medical services.  Approximately 32 percent of the population, or 1.27
million Puerto Ricans, are considered medically underserved.11

       While the links between these socioeconomic conditions and risks from air pollution is
complex, when combined with the high concentration of these population groups in close
proximity to large sources of emission such as marine ports, there is little doubt the residents
face  an elevated risk.

       San Juan, Puerto Rico's largest city, has a population of about 420,000 within the
municipio boundaries and 2.6 million throughout the metropolitan area.12 The area has a
population density of about 950 people per square kilometer, making it 170th in population
density among urban areas worldwide.13 Located on the North shore of Puerto Rico, San Juan is
built along one of the biggest natural harbors in the Caribbean Sea. The Port of San Juan is the
center for goods movement on the island, moving approximately 11 million metric tons of
products on nearly 3,800 vessel trips in 2008. Based on tonnage, San Juan ranks 49th out of the
top 150 commercial ports in the United  States that year.14 However, the port is in the top 10 in
the country when ranked by container traffic. In 2009 the Port of San Juan moved 1.7 million
twenty-foot equivalent units15 of containerized cargo. San Juan is also a major destination for
cruise ship passengers. In 2008 an estimated 1.4 million passengers on over 500 port calls
visited the Port of San Juan, making it one of the top cruise destinations in the Caribbean.16  San
Juan is the fifth busiest cruise ship departure port in the United States.17

       The city of Ponce, located on the southwest side of the main island, is also the home to a
major port. Ponce is Puerto Rico's second largest municipio with 180,000 inhabitants.7 The port
complex, which will be renamed Port of the Americas, is undergoing large-scale redevelopment
in order to relieve the congestion in San Juan. The port ranked as the 83rd busiest port, by
 1A municipio is the primary legal subdivision of Puerto Rico.


                                           1-4

-------
                                                                               Chapter 1

tonnage, out of the top 150 commercial ports in the United States in 2008.14 When completed,
Port of the Americas will be capable of handling up to 1.5 million twenty-foot containers and
600,000 tons of general cargo each year.  As of 2008, 3.8 million metric tons of goods moved
through the port in Ponce.18

       Mayaguez, an industrial and port city on the west coast of the island, is home to over
90,000 people.7 The port in Mayaguez moved approximately 350,000 tons of cargo in 2008,
much of which was fuel  shipments.

       Similarly, the port city of Arecibo, just downwind (west) of San Juan, is home to a port
primarily used for importing fuel. In 2008 fuel shipments alone totaled 53,000 tons.  Arecibo
contains 100,000 residents.7 The city's terrain consists of hills surrounding the city, forming a
natural bowl.

       The inland city of Caguas in eastern Puerto Rico is another major commercial center.  It
is situated in a valley surrounded by mountains, where air pollutants tend to accumulate rather
than disperse.  Although not port city, Caguas is subject to air pollution carried downwind from
ship traffic along the east coast of Puerto Rico.

       Other major port cities include Guayama, Yabucoa and Fajardo. Population and density
figures for the  main  coastal and inland municipios are listed in Table 1-1.

  Table 1-1 Annual Estimates of the Resident Population and Population Density for Municipios of Puerto
                               Rico. Source data: Reference 7.
MUNICIPIO
Arecibo
San Juan
Fajardo
Yabucoa
Guayama
Ponce
Guayanilla
Mayaguez
Caguas
Culebras
Vieques
POPULATION
(2009)
102,770
420,326
42,365
48,615
45,372
178,346
23,752
92,156
143,274
2,156
9,311
POP. DENSITY
(PEOPLE/KM2)
315
3,394
548
339
270
600
217
458
944
72
71
       In the late 1940s and early 1950s, the government of Puerto Rico implemented a plan to
encourage economic development by transforming the island's economy from an agriculture-
based economy (primarily sugar production) to one based on industry. The plan involved
importing raw materials, utilizing local labor to manufacture products, and then distributing the
finished products throughout the U.S. market. Throughout the 1950s the plan proceeded, and by
the end of the decade, the gross domestic product of Puerto Rico had almost doubled.19
                                           1-5

-------
Emission Control Area Designation
       As a result, the economy of the territory has moved from agricultural (sugar production)
to industrial, with manufacturing currently accounting for about 45 per cent of GDP and
agriculture only about 1 per cent. The sector has evolved to be more capital intensive over the
past three decades, with pharmaceuticals production now comprising the largest share within the
sector. Employment within the sector has likewise shifted. The high dependency on skilled labor
has reduced the manufacturing sector workforce to just over 10% in 2007, down from 19% in
1991 20

       Puerto Rico has very strong economic links with the continental United States. Because
of its lack of natural resources, the territory obtains the raw materials as well as chemicals,
machinery and equipment, clothing, food, fish, and petroleum from outside the island, mainly
from the continental United States (55 per cent), Ireland (24 per cent) and Japan (5 per cent).20
Finished goods include chemicals, electronics, apparel, canned tuna, rum, beverage concentrates,
and medical equipment, and are mainly destined for the continental United States (90 per cent).2

       Chemicals, in turn, are the largest export product, also accounting for two-thirds of the
                             20
total value of export shipments.  In all, Puerto Rico reported approximately $80 billion in
manufacturers' shipments in 2007, 65 percent of which is related to chemical manufacture.
21
       Aside from a small fraction of renewable energy production, Puerto Rico relies entirely
on external shipments of petroleum, natural gas and coal to meet its energy needs. Petroleum
imports totaled approximately 190,000 barrels per day, 20 percent of which are in the form of
crude oil which is then refined on the island. In addition, nearly 30 billion cubic feet of natural
                                                               T?
gas and 1.7 million tons of coal are shipped to Puerto Rico each year.

       The customs district of Puerto Rico, including all of its major ports, rank in the top 25
ports in the United States in terms of foreign trade and value. In 2008 over 14 million metric tons
of goods traveled through Puerto Rican ports at an estimated value of over $13 billion.23

       In addition to ships entering Puerto Rican ports, there is a substantial amount of ship
activity around the island from vessels on their way to or from the Panama Canal and other
countries in the Americas. These ship operations are described in Section 7 of the Information
Document accompanying MEPC 61/7/3.

       Puerto Rico is not only frequented by ships transporting goods, but also people.
Approximately 1.4 million visitors arrived by cruise ship in 2008. Puerto Rico has historically
been a top destination for cruises. Cruises and other tourism constitute a vital component of
Puerto Rico's economy. Tourist expenditures in Puerto Rico approached $3.5 billion in 2007.
The industry also provides jobs for seven percent of the workforce.20 According to the UN World
Tourism Organization, Puerto Rico had about 4 million international tourist arrivals in 2007,
ranking it 50th in the world.24

       In sum, Puerto Rico's economy is highly dependent on marine transportation.  This
dependency along with the physical and human geography, place the population at an elevated
risk from ship-related air pollution.
                                           1-6

-------
                                                                                Chapter 1

  1.2 U.S. Virgin Islands

       The U.S. Virgin Islands are the westernmost islands of the Lesser Antilles, located
between Puerto Rico (about 90 miles west) and the British Virgin Islands, and near the Anegada
passage, a deep (2,300 meter) channel that connects the Atlantic Ocean with the Caribbean Sea.
The U.S. Virgin Islands are comprised of three main islands, Saint Thomas, Saint John, and Saint
Croix, as well as several dozen smaller islands. The entire island chain measures  about 45 km
east-west by about 11 km north-south.

       This area is geologically active, being near the boundary of the Caribbean and North
American plates.  The U.S. Virgin Islands are volcanic in origin and mostly hilly to rugged and
mountainous  with little level land, although jungle-like regions may be found on the elevated
plateaus. These islands are known for their white sand beaches and coral reefs. More than half
of the island of St. John has been managed by the U.S. National Park Service since the Virgin
Islands National Park was expanded in 1962 to include over 5,000 acres of submerged lands  to
protect and preserve coral gardens and seascapes. Several other natural areas have been
officially designated for preservation and conservation, including the Virgin Islands Coral Reef
Monument, whose reefs are sheltered by mangrove forests and seagrass beds.4 UNESCO has
designated over two-thirds of the island of St. Johns as a Biosphere Reserve.3

       Like Puerto Rico, geographic constraints result in the citizens of the U.S. Virgin Islands
being located in densely populated coastal areas.  Figure 1-3 illustrates the major cities of the
three main U.S. Virgin Islands and their population densities.  Also, like Puerto Rico, this map
shows that all inhabitants of the U.S. Virgin Islands live in close proximity to commercial ports
or the coasts and are clearly affected by ship emissions.
                                           1-7

-------
Emission Control Area Designation
         Charlotte Amalie Port
                           r. .->n if
                           &W
              n. I

              -, - _
                          St. Thomas
                             51,181
       St. John
        4,197
                                         Christiansted Port
                                          ™
                                                          r.-r-.'
                                                           **»
                                           «-, I*-,. ,!*•«-  • fpi- ^
                                            «*   _/%-
              Frederiksted Port
                                        Alucoix Port
                                        St. Croix
                                         53,234
 I ib'.'JISIW I
 POPULATION    c Capital
[~~l Omi 1C.OOO  ~ SuWisutet Boundary

 ~1 £000 4900  SOOOPopuldionof
P==,       ,. U.S. Virgin Islands:
 IlLass.MnZ.COO     ^^
               Figure 1-3 Port locations and population density of the U.S. Virgin Islands

       The total population of the U.S. Virgin Islands is about 109,000.  The population density
of the islands is about 360 people per square kilometer, ranking it 34th in the world among
nations25.  This population is spread between St. Thomas (51,000 people; 630 people per square
kilometer) and St. Croix (53,000 people; 280 people per square kilometer). An additional 4,000
people live in  St. John; the rest of the islands have small or no populations.5'26

       St. Thomas is the site of the U.S.  Virgin Islands' capital and largest city, Charlotte
Amalie. Approximately 19,000 people live in the capital. Charlotte Amalie, as is typical of St.
Thomas and St. John islands, is characterized by steep topography that tends to contain air
pollution.

       Charlotte Amalie is also the location of the largest port in the U.S. Virgin Islands, St.
Thomas Port.  In 2005, St. Thomas alone saw over two million cruise passengers and over 800
cruise ship calls.  In  addition to cruise ships, smaller ferries and other passenger vessels frequent
the port and the small islands across from Charlotte Amalie (Hassel Island and Water Island).  St.
Thomas is also a major transshipment port for cargo destined elsewhere in the Caribbean.  In
total, Virgin Island ports handled over one million tons of cargo in 2005.27

       St. Croix is the largest and most populous of the U.S. Virgin Islands and contains the
Ports of Frederiksted, Alucoix, and Christiansted.  The most heavily  populated areas of St. Croix
are located downwind of Christiansted Port.
                                            1-8

-------
                                                                                   Chapter 1

       The main industry in the U.S. Virgin Islands is tourism, which accounts for about 80
percent of the territory's income.25 Like Puerto Rico, the U.S. Virgin Islands are also engaged in
manufacturing, importing raw materials and exporting finished goods including textiles,
electronics, rum, and pharmaceuticals.  Consequently, the U.S. Virgin Islands are dependent
upon the shipping industry. The largest trade partners for the U.S. Virgin Islands are the United
States and Puerto Rico.

       Finally, St. Croix is the location of one of the world's largest petroleum refineries,
Hovensa. A joint venture between Hess Corporation and Petroleos de Venezuela, this refinery
supplies heating oil and gasoline to the U.S. Gulf and East coasts. In 2008, the U.S. Virgin
Islands sent 320,000 barrels per day of refined products to the United States. With a capacity of
about 500,000 barrels per day, Hovensa is one of the ten largest refineries in the world and the
largest private employer in the U.S. Virgin Islands.  This refinery is subject to U.S. domestic
environmental regulations.28

       In sum, the economy of the U.S. Virgin Islands is highly dependent on marine
transportation. This dependency, in combination with the physical and human geography of the
territory, place its population and environment at an elevated risk from ship-related pollution.

  1.3 Conclusion

       Both Puerto Rico and the U.S. Virgin Islands are densely populated islands that receive a
large amount of ship traffic, both from trade vessels and tourist vessels. Due to the topographic
and geographic makeup of these islands, most of the population is located around the coasts. The
two territories are heavily fueled by the manufacturing industry, exemplified by the import of
raw materials and export of finished goods.  As a result, there is a significant portion of the
population residing in and around the numerous port cities as workers in the manufacturing
industry. In addition, as described in  Section 7 of the Information Document, Puerto Rico and the
U.S. Virgin Islands are located in high transit areas. Ships voyaging from Europe, Africa, and
Asia operate in passages to the east and west of these islands. The emissions from the
considerable ship traffic in this region have  an impact  on air quality, human health, and the
environment, in Puerto Rico and the U.S. Virgin Islands.
1 U.S. Geological Survey (1999). Groundwater Atlas of the United States: Alaska Hawaii Puerto Rico and the Virgin
Islands - HA 730-N, 1999.

2 Central Intelligence Agency (May 27, 2010). The World Factbook: Puerto Rico. Available at
https://www.cia.gov/library/publications/the-world-factbook/geos/rq.html

3 United Nations Educational, Scientific and Cultural Organization (2008). Biosphere Reserves World Network,
Man and the Biosphere Programme. Available at: http://www.unesco.org/mab/doc/brs/BRList2010.pdf

4 Mac, M. I, P. A. Opler, C. E. Puckett Haecker, and P. D. Doran (1998). Status and trends of the nation's biological
resources, 2 vols, U.S. Department of the Interior, U.S. Geological Survey, Reston, Va. Available at
http://www.nwrc.usgs.gov/sandt/SNT.pdf
                                             1-9

-------
Emission Control Area Designation
5 U.S. Census Bureau (2002). TIGER/Line Files, UA Census 2000, Washington, DC. Available at
http://www.census.gov/geo/www/tiger/tigerua/ua tgr2k.html

6 Modified from Galarza, Javier Rodriguez (2008). Population Density of Puerto Rico, Census 2000

7U.S. Census Bureau (2010). Annual Estimates of the Resident Population for Municipios of Puerto Rico: April 1,
2000 to July 1, 2009, Available at: http://www.census.gov/popest/municipios/PRM-EST2009-01.html

8 Central Intelligence Agency (May 27, 2010). The World Factbook: Puerto Rico. Available at
https://www.cia.gov/librarv/publications/the-world-factbook/geos/rq.html

9 Population Reference Bureau (2009). 2009 World Population Data Sheet. Available at
http://www.prb.org/Publications/Datasheets/2009/2009wpds.aspx

10 U.S. Census Bureau (2010). 2006-2008 American Community  Survey 3-Year Estimates. Available at
http://factfinder.census.gov/home/saff/main.html? lang=en

11 U.S. Department of Health and Human Services, Health Resources and Services Administration (2010). HRSA
Geospatial Data Warehouse. Available at http://datawarehouse.hrsa.gov/hpsadetail.aspx

12 U.S. Census Bureau (2010). Metropolitan and Micropolitan Statistical Area Estimates - Annual Estimates of the
Population: April 1, 2000 to July  1, 2009, CBSA-EST2009-03. Available at:
http://www.census.gov/popest/metro/CBSA-est2009-annual.html

13 City Mayors Statistics (2010). The largest cities in the world by land area, population and density. Available at
http://www.citvmavors.com/statistics/largest-cities-densitv-250.html

14 U.S. Army Corps of Engineers Navigation Data Center (2010). Tonnage for Selected U.S. Ports in 2008.
Available at http://www.ndc.iwr.usace.armv.mil/wcsc/portname08.htm

15American Association of Port Authorities (2010). North American Port Container Traffic. Available at
http://aapa.files.cms-plus.com/Statistics/NORTHAMERICANPORTCONTAINERTRAFFIC2009.pdf

16 Puerto Rico Ports Authority (2009). Cruise  Ships Passengers Movement Port of San Juan. Available at:
http://www.prpa.gobierno.pr/uploads/Estadisticas/Estadisticas%202008/cruises%20hist%20fv.pdf

17 U.S. Department of Transportation Maritime Administration (2009). U.S. Water Transportation Statistical
Snapshot. Available at http://www.marad.dot.gov/documents/US_Water_Transportation_Statistical_snapshot.pdf

18 U.S Army Corps of Engineers (2008). Waterborne Commerce of the United States - Waterways and Harbors Gulf
Coast, Mississippi River System and Antilles. IWR-WCUS-08-2.
http://www.ndc.iwr.usace.armv.mil/wcsc/pdf/wcusmvgc08.pdf

19 Suarez, Nydia R (1998). The Rise and Decline of Puerto Rico's Sugar Economy, Economic Research Service,
U.S. Department of Agriculture, Sugar and Sweetener S&O/SSS-224/December 1998, 22-38. Available at
http://www.ers.usda.gov/Briefing/Sugar/sugarpdf/SSS224PuertoRico.pdf

20 Government Development Bank for Puerto  Rico (November 2008). Puerto Rico in Figures, 2007. Available at:
http://www.gdb-pur.com/publications-reports/prinfigures/Infigures2007.pdf
                                                 1-10

-------
                                                                                              Chapter 1
21 U.S. Census Bureau (2010). Economic Census of Island Areas. Available at
http://factfinder.census.gov/servlet/EconSectorServlet?caller=dataset&sv name=2007+Economic+Census+of+Islan
d+Areas&  Sectorld=*&ds name=EC0700Al

22 U.S. Energy Information Administration (2010). State Energy Profiles - Puerto Rico, Last Update: June 10, 2010.
Available at http://www.eia.doe.gov/state/territory_profile_pr.html#fnl

23 U.S. Census Bureau Foreign Trade Division (2010). U.S. Waterborne Foreign Trade by Custom District.
Available at http://www.marad.dot.gov/library  landing_page/data and statistics/Data and  Statistics.htm

24 UN World Tourism Organization (2009). UNWTO World Tourism Barometer, Volume 7 No. 3. Available at
http://unwto.org/facts/eng/pdf/barometer/UNWTO Barom09 3 en.pdf

25 Central Intelligence Agency (May 27, 2010). The World Factbook: Virgin Islands. Available at
https://www.cia.gov/library/publications/the-world-factbook/geos/vq.html

26 U.S. Census Bureau (2003). U.S. Virgin Islands: 2000, Social, Economic, and Housing Characteristics, 2000
Census of Population and Housing, PHC-4-VL  Available at http://www.census.gov/prod/cen2000/phc-4-vi.pdf

27 Virgin Islands Port Authority (2009). Aviation and Marine Traffic Statistics, VTPA Office of Public Relations.
Available at http://www.viport.com/statistics.html

28 U.S. Energy Information Administration, Independent Statistics and Analysis. Country Analysis Briefs -
Caribbean, 2009, http://www.eia.doe.gov/emeu/cabs/Caribbean/pdf.pdf
                                                  1-11

-------
CHAPTER 2:  EMISSION INVENTORY	2-2
2.1   Introduction	2-2
2.2   Description of Ships Included in the Analysis	2-2
2.3   Inventory Methodology	2-3
2.4   Development of 2002 Inventories	2-4
2.5   Development of 2020 Inventories	2-20
2.6   Inventories for Proposed ECA	2-30
2.7   Other Inventories	2-32
2.8   Conclusion	2-33
APPENDICES	2-37
APPENDIX 2A: PORT COORDINATES	2-37
APPENDIX 2B: PORT METHODOLOGY AND EQUATIONS	2-38
APPENDIX 2C: EMISSION INPUTS TO STEEM	2-46
APPENDIX 2D: GROWTH FACTOR DEVELOPMENT	2-52
                                    2-1

-------
CHAPTER 2: Emission Inventory

  2.1 Introduction

       This chapter presents our estimated air emission inventories for ships that operate in the
proposed EGA for the Commonwealth of Puerto Rico and the U.S. Virgin Islands (PR/USVI).
This chapter is organized in four parts.  First, we describe the domain of ships included in the
analysis. Second, we describe the modeling methodology. Third, we present the results of this
modeling, for the baseline inventory year of 2002 as well as the baseline and control scenarios
for 2020. Finally, we present inventories for other sources of emissions for comparison.

       Using the methodology described below, the estimated ship emissions inventories  for the
proposed U.S. Caribbean EGA for 2020 are as set out in Table 2-1.  Inventories for both the
reference (baseline) and the control scenarios are presented.  EGA designation is expected to
reduce emissions of NOx, SOx, and PM by 27 percent, 96 percent, and 86 percent, respectively,
in 2020.

                   Table 2-1  C3 Emission Inventories for Proposed ECA in 2020
EMISSION TYPE
Reference
Control
Delta Emissions
Delta Emissions (%)
ANNUAL EMISSIONS (METRIC TONNES)a'b
NOXP
36,950
27,032
-9,919
-27%
VI ioP
3,793
512
-3,342
-86%
M 2.5C
3,488
471
-3,017
-86%
HC
1,509
1,509
0
0%
CO
3,609
3,609
0
0%
SO2CO
29,568
1,075
-28,493
-96%
2
1,797,909
1,711,452
-86,457
-5%
  2.2 Description of Ships Included in the Analysis

       The ship inventories reported in this chapter are for vessels with Category 3 propulsion
engines (defined as engines with per cylinder displacement at or above 30 liters). These are the
ships that are most likely to be affected by the MARPOL Annex VI fuel sulfur limits since these
vessels are most likely to be designed to use residual fuel. While smaller vessels will also be
affected by the proposed ECA, it is also the case that many of these vessels (those flagged or
registered in the United States) are already subject to comparable U.S. marine diesel engine and
fuel requirements under the CAA. In either case, these smaller vessels are likely to be using
distillate fuel and therefore switching to a lower sulfur diesel fuel is not expected to impose a
significant burden on their owners.

       The ship inventories include emissions from both propulsion and auxiliary engines
installed  on board the Category 3 vessels included in the analysis.

       The ship emission inventories are based on the U.S. Army Corps of Engineers (USAGE)
foreign traffic entrances and clearances data set. This is derived from U.S. Customs Vessels
Entrances and Clearances data.  The following vessels are required to file a Vessel Entrance or
Clearance Statement:
                                        2-2

-------
              Any vessel from a foreign port or place;
       •       Any foreign vessel from a domestic port;
       •       Any vessel of the United States arriving from another U.S. port and having
              merchandise on board being transported in bond (this does not include bonded
              ship's stores or supplies), or transporting unentered foreign merchandise; or
       •       Any vessel, either U.S. or foreign, which has visited a hovering vessel (19 USC
              1401(k)), or has delivered or received merchandise or passengers outside of U.S.
              waters.

       The Entrances and Clearances data sets cover only foreign cargo movements. However,
many ships tend to travel in circular routes (e.g., from Miami to Puerto Rico to Mexico to Brazil
and then back to Miami). Cargo moved from Miami to Brazil would be foreign cargo, but the
trip between Miami (origin) and the first port in Puerto Rico would be captured in the clearances
data since it shows the first port of call.

       Not included in this  data set are US/domestic ships operating solely between the
continental United States and Puerto Rico or the U.S. Virgin Islands. Most of that traffic is
thought to be on ships with Category 2 propulsion engines or tug/barge combination vessels, and
these smaller ships are already subject to U.S. marine diesel engine requirements, and the sulfur
content of fuel available in the U.S. ports in which they operate is also subject to federal controls.

  2.3 Inventory Methodology

       The methodology used to estimate the inventories for the proposed EGA is consistent
with the methodology used for the North American EGA. This methodology is summarized
below; more details about the methodology as well as the additional calculations and minor
changes required for the current application are  presented in later sections of this chapter.

       The inventory methodology consists of several parts.

       First, we developed an inventory for 2002 for a broader modeling domain (called the
inventory domain) consisting of the entire area around Puerto Rico and the U.S. Virgin Islands
that is subject to the sovereign authority of the United States and consists of the exclusive
economic zone surrounding these islands.  The year 2002 was chosen to be consistent with the
analysis performed for the North  American EGA and allows us to take advantage of much of the
work performed for that analysis, including inter-port emissions and estimated growth rates.

       The 2002 inventory consists of two parts that were estimated using two different
methods: port emissions and interport emissions.

              Port inventories consist of emissions that occur in a port, defined as up to 25 miles
              from the entrance of the port. Port inventories were developed for seven ports in
              Puerto Rico and five ports in the  U.S. Virgin Islands. Port-specific emissions
              were estimated using a "bottom-up" approach based on port-specific vessel calls,
              emission factors, and activity for each port.  For all other ports, estimates from the
              STEEM model are used (see below).
                                         2-3

-------
              Interport emissions consist of emissions that occur outside of ports but within the
              inventory domain. These inventories were obtained using the Waterway Network
              Ship Traffic, Energy and Environment Model (STEEM). STEEM also uses a
              "bottom-up" approach, estimating emissions from C3 vessels using historical
              shipping activity, ship characteristics, and activity-based emission factors.
              STEEM was used to quantify and geographically (i.e., spatially) represent
              interport vessel traffic and emissions for vessels traveling within the proposed
              EGA.

       The regional emission inventories produced by the current STEEM interport model are
most accurate for vessels while cruising in ocean shipping lanes; the near port inventories use
more detailed local port information and are significantly  more accurate near the ports.
Therefore, the inventories in this analysis are derived by merging together:  1) the near port
inventories, which extend 25 nautical miles, and 2) the remaining interport portion of the
STEEM inventory, which extends from the endpoint of the near port inventories to the outer
boundary of the Caribbean inventory domain.

       Merging these inventories requires spatially allocating the in-port emissions, removing
the data for the 12 ports from the STEEM inventory, and replacing it with the detailed port
inventories.  The STEEM port data was retained for all other Puerto Rican and Virgin Island
ports. The result of this process was a complete, spatially allocated inventory for 2002 covering
the entire inventory domain. Near some ports, a portion of the underlying STEEM emissions
were retained if it was determined that the STEEM emissions included ships traversing the area
near a port, but not actually entering or exiting the port.

       Next, baseline and control inventories were developed for the entire inventory domain for
2020. The baseline inventories for 2020 were estimated by applying a growth rate and emission
adjustment factors to the 2002 inventories.  The emission  adjustment factors account for
emission controls that will be in effect in 2020, including the MARPOL Annex VI Tier I and
Tier II NOx standards for new engines and the Regulation 13  NOx retrofit program.  The control
inventories for 2020 were estimated by applying the same growth rate as the 2020 baseline case
but a different set of emission adjustment factors that also account for the EGA engine and fuel
sulfur controls. The result of this process was a complete, spatially allocated inventory for 2020
covering the entire inventory domain, for both  the baseline and control scenarios.

       Finally, the inventories for the proposed EGA for the 2020 baseline and control scenarios
were developed by totaling the emissions within the proposed EGA boundaries.

     Inventories are presented for the following pollutants: oxides of nitrogen (NOx),
particulate matter (PM2.5 and PMio), sulfur dioxide (802), hydrocarbons (HC), carbon monoxide
(CO), and carbon dioxide (CO2). The PM inventories include directly emitted PM only.

  2.4 Development of 2002 Inventories

     The inventories for the proposed EGA are derived from inventories estimated for the
inventory domain consisting of the U.S. Exclusive Economic Zone around the islands of Puerto
Rico and the U.S. Virgin Islands. The year 2002 was chosen to be consistent with the analysis
                                         2-4

-------
performed for the North American EGA and allows us to take advantage of much of the work
performed for that analysis, including inter-port emissions and estimated growth rates. The total
inventories for 2002 are the total of port and interport inventories described in this section. The
result is a spatially allocated inventory for the inventory domain.

2.4.1 Near Port Emissions

       The outer boundaries of the ports are defined as 25 nautical miles (nm) from the terminus
of the reduced speed zone. Port emissions are estimated for different modes of operation and
then summed. Emissions for each mode are estimated using port-specific information for vessel
calls, vessel characteristics, and activity, as well as other inputs that vary instead by vessel or
engine type (e.g., emission factors). The methodology and port inventory development was
conducted under contract; details of the methodology as applied to the U.S. ports is described in
the  contractor report.l

2.4.1.1  Ports Modeled

       The 12 near port inventories are an improvement upon STEEM's near port results in
several ways. First, the precision associated with STEEM's use of ship positioning data may be
less accurate in some locations, especially as the lanes approach shorelines where ships would
need to follow more prescribed paths. Second, the STEEM model includes a maneuvering
operational mode (i.e., reduced speed) that is generally assumed to occur within a 20 kilometer
radius of each port.  In reality, the distance when a ship is traveling at reduced speeds varies by
port. Also, the distance a ship traverses at reduced speeds often consists of two  operational
modes:  a reduced speed zone (RSZ) as a ship enters or leaves the port area and  actual
maneuvering at a very low speed near the dock.  Third, the STEEM model assumes that the
maneuvering distance occurs at an engine load of 20 percent, which represents a vessel speed of
approximately 60 percent of cruise speed.  This is considerably faster than ships would maneuver
near the docks.  The single maneuvering speed assumed by STEEM also does not reflect the fact
that the reduced speed zone, and therefore emissions, may vary by port.  Fourth, and finally, the
STEEM model does not include the emissions from auxiliary engines during hotelling operations
at the port.  The near-port inventories correct these issues.

       Near port emissions were estimated for the ports listed in Table 2-2.  The 12 ports were
chosen because of the availability of call data from the USAGE entrance and clearance data.2
The port coordinates are provided in Appendix 2A.

                                   Table 2-2 Modeled Ports
                PUERTO RICO PORTS
                San Juan
                Fajardo
                Ponce
                Jobos
                Guayanilla
                Mayaguez
                Yabucao
U.S. VIRGIN ISLANDS PORTS
St. Thomas
Christiansted
Frederiksted
Port Harvey (alumina bauxite refinery)
Port Hess (oil terminal)
                                         2-5

-------
       These ports were chosen because they are the largest ports in Puerto Rico and the U.S.
Virgin Islands.  For all other ports in these territories, emissions inventories estimated by the
STEEM model are used.

2.4.1.2  Near Port Inventory Methodology

       Near port emissions for each port are estimated using a bottom-up approach based on the
number of vessel calls and vessel characteristics. Emissions are estimated for four modes of
operation:

             Hotelling:  Hotelling, or dwelling, occurs while the vessel is docked or anchored
             near a dock, and only the auxiliary engine(s) are being used to provide power to
             meet the ship's energy needs.
             Maneuvering: Maneuvering occurs within a very short distance of the docks.
             Reduced speed zone (RSZ): The RSZ varies from port to port, though generally
             the RSZ would begin and end when the pilots board or disembark, and typically
             occurs when the near port shipping lanes reach unconstrained ocean shipping
             lanes.
             Cruise:  The cruise mode emissions in the near ports  analysis extend 25 nautical
             miles beyond the end of the RSZ lanes.

       Emissions are calculated separately for propulsion and auxiliary engines.

       The basic equation used to estimate emissions for each engine at each mode is as follows:

                                       Equation 2-1

       Emission&od4eng] = (call$ x (P[eng]) x (hrsl callmoj x (LFmode[eng]) x (EF[eng]) x (Adj) x (10^ tonned g)
       Where:
             EmissionSmode [eng] = Metric tonnes emitted by mode and engine type
          -  Calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
          -  P[eng] = Total engine power by engine type, in kilowatts
             hrs/callmode = Hours per call by mode
             LFmode [eng] = Load factor by mode and engine type (unitless)
          -  EF[eng] = Emission factor by engine type for the pollutant of interest, in g/kW-hr
             (these vary as a function of engine type and fuel used, rather than activity mode)
             Adj = Low load adjustment factor, unitless (used when the load factor is below
             0.20)
             10"6 = Conversion factor from grams to metric tonnes

2.4.1.3  Data Inputs for Near Port Emission Inventories

       The following inputs  are required to estimate emissions inventories for each vessel at
four modes of operation (cruise, RSZ, maneuvering, and hotelling); these inputs are described in
more detail below.
                                         2-6

-------
       Number of calls and ship characteristics (main engine power, cruise speed, and load
       factors)
       Cruise distance
       RSZ distances and speeds for each port
       Auxiliary engine power and load factors
       Main emission factors
       Auxiliary emission factors
       Low load adjustment factors for main engines
       Maneuvering time-in-mode (hours/call)
       Hotelling time-in-mode (hours/call)

       Number of Calls and Ship Characteristics (main engine power, cruise speed, and load
factors)

       For this analysis, USAGE entrance and clearance data for 2002,3 together with Lloyd's
data for ship  characteristics,4 were used to identify average ship characteristics and calls by ship
type for each port. Information for number of calls, propulsion engine power, and cruise speed
were obtained from these data.

       The records from the USAGE entrances and clearances data base were matched with
Lloyd's data  on ship characteristics for each port. Calls by vessels that have either Category 1 or
2 propulsion  engines were eliminated from the data set.  This was accomplished by matching all
ship calls with information from Lloyd's Data, which is produced by Lloyd's Register-Fairplay
Ltd. Over 99.9 percent of the calls in the entrances and clearances data were directly matched
with Lloyd's data. The remaining 0.1 percent was estimated based upon ships of similar type
and size. Engine category was determined from engine make and model.  Engine bore and
stroke were found in the Marine Engine 2005 Guide5 and displacement per cylinder was
calculated. Ships with main propulsion engines with per cylinder displacement less than 30 liters
eliminated from the data set. Passenger ships and tankers have either diesel-electric or gas
turbine-electric engines that are used for both propulsion and auxiliary purposes and were
retained in the data set as they are subject to the EGA requirements.

       The dataset for vessels with Category 3 propulsion engines was then binned by ship type,
engine type and dead weight tonnage (DWT) range.  The number of entrances and clearances in
each bin are counted, summed together and divided by two to determine the number of calls (i.e.,
one entrance  and one clearance was considered a call). Propulsion power and vessel cruise speed
are also averaged for each bin.

       Main engine load factors are calculated directly from the propeller curve based upon the
cube of actual speed divided by maximum speed (at 100% maximum continuous rating [MCR]).
In addition, cruise mode activity is based on cruise distance and speed inputs. Appendix 2B
provides the specific equations used to calculate propulsion and auxiliary emissions for each
activity mode.

       Note that load factors for main engines are not listed  explicitly, since they are calculated
as a function of mode and/or cruise speed.
                                         2-7

-------
       Cruise Distance

       Cruise mode emissions are calculated assuming a 25 nautical mile distance into and out
of the port outside of the reduced speed and maneuvering zones.

       RSZ Distances and Speeds by Port

       The reduced speed zone (RSZ) distance and speed were 10 nautical miles and 10 knots,
respectively, for all PR/USVI ports.

       Auxiliary Engine Power and Load Factors

       Hotelling emissions are a significant part of port inventories, and it is important to
distinguish propulsion engine emissions from auxiliary engine emissions when estimating these
emissions. This is because hotelling emissions are generally generated by auxiliary engines.

       In the methodology used in this analysis, auxiliary engine maximum continuous rating
power and load factors were calculated separately from propulsion engines and different
emission factors (EFs) applied. All auxiliary engines were treated as Category 2 medium-speed
diesel (MSD) engines for purposes of this analysis.

       Auxiliary engine power is not contained in the USAGE database and is only sparsely
populated in the Lloyd's database; as a result, it must be estimated. The approach taken was to
derive ratios of average auxiliary engine power to propulsion power based on survey data.  The
California Air Resources Board (ARB) conducted an Oceangoing Ship Survey of 327 ships in
January 2005 that was principally used for this analysis.6 Average auxiliary engine power to
propulsion power ratios were estimated by ship type and are presented in Table 2-3. These ratios
by ship type were applied to the propulsion power data to derive auxiliary power for the ship
types at each port.
                                         2-8

-------
                 Table 2-3 Auxiliary Engine Power Ratios (ARE Survey, except as noted)
SHIP TYPE
Auto Carrier
Bulk Carrier
Container Ship
Passenger Ship3
General Cargo
Miscellaneous13
RORO
Reefer
Tanker
AVERAGE
PROPULSION
ENGINE (kW)
10,700
8,000
30,900
39,600
9,300
6,250
11,000
9,600
9,400
Average Auxiliary Engines
NUMBER
2.9
2.9
3.6
4.7
2.9
2.9
2.9
4.0
2.7
POWER
EACH
(kW)
983
612
1,889
2,340
612
580
983
975
735
TOTAL
POWER
(kW)
2,850
1,776
6,800
11,000
1,776
1,680
2,850
3,900
1,985
ENGINE
SPEED
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
AUXILIARY TO
PROPULSION
RATIO
0.266
0.222
0.220
0.278
0.191
0.269
0.259
0.406
0.211
        a Many passenger ships typically use a different engine configuration known as diesel-electric. These vessels
         use large generator sets for both propulsion and ship-board electricity. The figures for passenger ships
         above are estimates taken from the Starcrest Vessel Boarding Program.
        b Miscellaneous ship types were not provided in the ARB methodology, so values from the Starcrest Vessel
         Boarding Program were used.

        Auxiliary engine to propulsion engine power ratios vary by ship type and operating mode
roughly from 0.19 to 0.40. Auxiliary  load, shown in Table 2-4, is  used together with the total
auxiliary engine power to calculate auxiliary engine emissions. Starcrest's Vessel Boarding
Program7 showed that auxiliary engines are on all of the time, except when using shoreside
power during hotelling.

                         Table 2-4 Auxiliary Engine  Load Factor Assumptions
SHIP TYPE
Auto Carrier
Bulk Carrier
Container Ship
Passenger Ship
General Cargo
Miscellaneous
RORO
Reefer
Tanker
CRUISE
0.13
0.17
0.13
0.80
0.17
0.17
0.15
0.20
0.13
RSZ
0.30
0.27
0.25
0.80
0.27
0.27
0.30
0.34
0.27
MANEUVER
0.67
0.45
0.50
0.80
0.45
0.45
0.45
0.67
0.45
HOTEL
0.24
0.22
0.17
0.64
0.22
0.22
0.30
0.34
0.67
                                              2-9

-------
       Main Engine Emission Factors

       An analysis of emission data was prepared and published in 2002 by Entec.8 The
resulting Entec emission factors include individual factors for three speeds of diesel engines
(slow-speed diesel (SSD), medium-speed diesel (MSD), and high-speed diesel (HSD)), steam
turbines (ST), gas turbines (GT), and two types of fuel used here, residual marine (RM) and
marine distillate oil (MDO).  Table 2-5 lists the propulsion engine emission factors for NOx and
HC that were used for the 2002 port inventory development. The CO, PM, SO2 and CO2
emission factors shown in the table come from other data sources as explained below.   Since PM
and SC>2 emission factors are dependent on the fuel sulfur level, the fuel types and fuel sulfur
levels used in this analysis are described at the end of this section.

                Table 2-5 Emission Factors for OGV Main Engines using RM, g/kWh
ENGINE
SSD
MSD
ST
GT
ALL PORTS
NOX(
18.1
14.0
2.1
6.1
0
1.40
1.10
0.20
0.20
HC
0.60
0.50
0.10
0.10
CO2PB
620.62
668.36
970.71
970.71
1 10
1.4
1.4
1.5
1.5
PM25
1.3
1.3
1.4
1.4
S02
10.29
11.09
16.10
16.10
       CO emission factors were developed from information provided in the Entec appendices
because they are not explicitly stated in the text.  HC and CO emission factors were confirmed
with a recent U.S. Government review.9

       PM10A values were determined based on existing engine test data in consultation with
ARB.10 GT PMio emission factors were not part of the U.S. Government analysis but assumed
here to be equivalent to ST PMi0 emission factors. Test data shows PMi0 emission rates as
dependent upon fuel sulfur levels,  with base PMio emission rates of 0.23 g/kw-hr with distillate
fuel (0.24% sulfur) and 1.35 g/kw-hr with residual fuel (2.46% sulfur).11 The equation used to
generate emission factors based on sulfur content is shown below.  PM2.5 is assumed to be 92
percent of PMi0. While the US Government NONROAD model uses 0.97 for such conversion
based upon low sulfur fuels, a reasonable value seems to be closer to 0.92 because higher sulfur
fuels in medium and slow speed engines would tend to produce larger particulates than high
speed engines on low sulfur fuels.

            Equation 2-2 Calculation of PM10 Emission Factors Based on Fuel Sulfur Levels

                 PMEF = PMNom + [(SAct - SNom) x BSFC x FSC x MWR x 0.0001]

             where:
                    PMEF  = PM emission factor adjusted for fuel sulfur
 1PM10 is paniculate matter of aerodynamic diameter 10 micrometers or less.
                                        2-10

-------
                     PMNom  = PM emission rate at nominal fuel sulfur level
                            = 0.23 g/kW-hr for distillate fuel, 1.35 g/kW-hr for residual fuel
                     SACI    = Actual fuel sulfur level (weight percent)
                     SNOHI    = nominal fuel sulfur level (weight percent)
                            = 0.24 for distillate fuel, 2.46 for residual fuel
                     BSFC   = fuel consumption in g/kW-hr
                            = 200 g/kW-hr used for this analysis
                     FSC    = percentage of sulfur in fuel that is converted to direct sulfate PM
                            = 2.247% used for this analysis
                     MWR   = molecular weight ratio of sulfate PM to sulfur
                            = 224/32 = 7 used for this analysis

       SO2 emission factors were based upon a fuel sulfur to SC>2 conversion formula which was
supplied by ENVIRON.12 Emission factors for SO2 emissions were calculated using the formula
assuming that 97.753 percent of the fuel sulfur was converted to 862.13 The brake specific fuel
consumption (BSFC)B that was used for SSDs was 195  g/kWh, while the BSFC that was used
for MSDs was 210 g/kWh based upon Lloyds 1995. The BSFC that was used for STs and GTs
was 305 g/kWh based upon Entec.14

                     Equation 2-3 Calculation of SO2 Emission Factors, g/kWh

                      SO2 EF = BSFC x 64/32 x 0.97753 x Fuel Sulfur Fraction

       CO 2 emission factors were calculated from the BSFC assuming a fuel carbon  content of
86.7 percent by weight14 and a ratio of molecular weights of CO2 and C at 3.667.

                     Equation 2-4 Calculation of CO2 Emission Factors, g/kWh

                                 CO2 EF = BSFC x 3.667 x 0.867

       Fuel consumption was calculated from CO2 emissions based on a 1:3.183 ratio.
Approximately 3.183 tons of CC>2 emissions are assumed produced from one metric ton of fuel.

       SO2 emission factors were calculated using Equation 2-3 while PM emissions were
determined using Equation 2-2.

       Note on Fuel Types and Fuel Sulfur Levels: There are primarily three types of fuel used
by marine engines: residual marine (RM), marine diesel oil (MDO), and marine gas oil (MGO),
with varying  levels of fuel sulfur.15 MDO and MGO are generally described as distillate fuels.
For this analysis, RM and MDO fuels are assumed to be used. Since PM and SO2 emission
factors are dependent on the fuel sulfur level, calculation of port inventories requires  information
about the fuel sulfur levels associated with each fuel type, as well as which  fuel types are used by
propulsion and auxiliary engines.

       Based on an ARE survey,16 average fuel sulfur level for residual marine was set to 2.7
percent, which is what was assumed in the North American EGA application for the eastern and
 ' Brake specific fuel consumption is sometimes called specific fuel oil consumption (SFOC).
                                         2-11

-------
gulf coast portions of the U.S.  A sulfur content of 1.5 percent was used for MDO.17 While a
more realistic value for MDO used in the U.S. appears to be 0.4 percent, given the small
proportion of distillate fuel used by ships relative to RM, the difference should not be significant.
Sulfur levels in other areas of the world can be significantly higher for RM.  Table 2-6, based on
the ARE survey, provides the assumed mix of fuel types used for propulsion and auxiliary
engines by ship type.

                       Table 2-6 Estimated Mix of Fuel Types Used by Ships
SHIP
TYPE
Passenger
Other
FUEL USED
PROPULSION ^
100% RM
100% RM
U XILIARY
92% RM/8% MDO
71%RM/29%MDO
       Auxiliary Engine Emission Factors

       The most current set of auxiliary engine emission factors also comes from Entec except
as noted below for PM and SC>2.  Table 2-7 provides these auxiliary engine emission factors.

                  Table 2-7 Auxiliary Engine Emission Factors by Fuel Type, g/kWh
ENGINE F
MSD
UEL
RM
MDO
ALL PORTS
NOXC
14.70
13.90
0
1.10
1.10
HC
0.40
0.40
C02
668.36
668.36
PM10
1.4
0.6
PM25
1.3
0.55
S02
11.09
6.16
       Auxiliary engine power was estimated from average propulsion power using the ratio of
auxiliary power to propulsion power ratios described below.

       Using the ratios of RM versus MDO use as given in Table 2-6 together with the emission
factors shown in Table 2-7, the auxiliary engine emission factor averages by ship type are listed
in Table 2-8. As discussed above, this fuel sulfur level may be too high for the PR/USVI.
However, we do not believe this emission factor has a significant effect on the total emission
inventory estimates.

                 Table 2-8 Auxiliary Engine Emission Factors by Ship Type, g/kWh
SHIP TYPE
Passenger
Others
ALL PORTS
NOX
14.64
14.47
COH(
1.10
1.10
CO
0.40
0.40
2
668.36
668.36
PM10
1.4
1.2
PM25
1.3
1.1
SO2
10.70
9.66
                                         2-12

-------
       Low Load Adjustment Factors for Propulsion Engines

       Emission factors are considered to be constant down to about 20 percent load. Below
that threshold, emission factors tend to increase as the load decreases. This trend results because
diesel engines are less efficient at low loads and the brake specific fuel consumption (BSFC)
tends to increase. Thus, while mass emissions (grams per hour) decrease with low loads, the
engine power tends to decrease more quickly, thereby increasing the emission factor (grams per
engine power) as load decreases. Energy and Environmental Analysis Inc. (EEA) demonstrated
this effect in a study prepared for the U.S. Government in 2000.18  In the EEA report, equations
have been developed for the various emissions. The low-load adjustment factors were developed
based upon the concept that the BSFC increases as load decreases below about 20 percent load.

       Using these algorithms, fuel consumption and emission factors versus load were
calculated.  By normalizing emission factors to 20% load, low-load multiplicative adjustment
factors were calculated for propulsion engines and presented in Table 2-9. Due to how they are
operated, there is no need for a low load adjustment factor for auxiliary engines.

                 Table 2-9 Calculated Low Load Multiplicative Adjustment Factors
LOAD
(%) N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
OXHC
11.47
4.63
2.92
2.21
1.83
1.60
1.45
1.35
1.27
1.22
1.17
1.14
1.11
1.08
1.06
1.05
1.03
1.02
1.01
1.00

59.28
21.18
11.68
7.71
5.61
4.35
3.52
2.95
2.52
2.20
1.96
1.76
1.60
1.47
1.36
1.26
1.18
1.11
1.05
1.00
CO
19.32
9.68
6.46
4.86
3.89
3.25
2.79
2.45
2.18
1.96
1.79
1.64
1.52
1.41
1.32
1.24
1.17
1.11
1.05
1.00
PM
19.17
7.29
4.33
3.09
2.44
2.04
1.79
1.61
1.48
1.38
1.30
1.24
1.19
1.15
1.11
1.08
1.06
1.04
1.02
1.00
SO2CO
5.99
3.36
2.49
2.05
1.79
1.61
1.49
1.39
1.32
1.26
1.21
1.18
1.14
1.11
1.09
1.07
1.05
1.03
1.01
1.00
2
5.82
3.28
2.44
2.01
1.76
1.59
1.47
1.38
1.31
1.25
1.21
1.17
1.14
1.11
1.08
1.06
1.04
1.03
1.01
1.00
       Maneuvering andHotelling Time-in-Mode

       Specific information about the amount of time spent in maneuvering and hotelling modes
was not available for the 12 ports included in the ports inventory. Instead, we used the approach
                                        2-13

-------
that was used for the U.S. mainland ports, in which all commercial ports were mapped to one of
a smaller set of "typical ports" and the operating characteristics of the relevant typical port was
applied to the specific matched ports. For this analysis, Tampa was selected as the typical port
thought to be most representative of the PR/US VI ports, due to its location and mix of ship types
that call on the port. Time-in-mode data by ship type for the Tampa port were used directly.

2.4.1.4  2002 Near Port Inventories

       The resulting 2002 emission inventory for each of the 12 ports is provided in Table 2-10.

        Table 2-10  2002 Emissions Summary for Twelve Puerto Rico and U.S. Virgin Island Ports
PORT NAME
San Juan, PR
Fajardo, PR
St. Thomas, VI
Ponce, PR
Christiansted, VI
Jobos, PR
Guayanilla, PR
Mayaguez, PR
Yabucao, PR
Frederiksted, VI
Port Harvey, VI
Port Hess, VI
Total Port Emissions
ANNUAL EMISSIONS (METRIC TONNES)
NOXP
3,909
29
2,305
420
31
42
95
1,271
53
73
230
797
9,255
M 10 P
350
o
6
231
35
o
J
3
9
107
4
8
21
65
839
M 25
324
2
215
32
o
3
3
8
99
4
7
19
60
775
HC
122
1
74
14
1
1
o
3
38
2
2
8
26
292
CO
303
2
178
34
2
3
7
98
4
5
18
63
719
S02C(
2,980
20
2,030
263
24
26
71
940
33
69
169
501
7,727
> 2
177,477
1,297
112,970
17,272
1,392
1,694
4,595
58,241
2,170
3,967
10,678
32,515
424,271
2.4.2 Interport Emissions

       The second part of the emissions inventory is emissions from ships traveling outside of
the 25-mile port areas and for ports other than the 12 ports described above.  These emissions are
estimated using the Waterway Network Ship Traffic, Energy, and Environmental Model
(STEEM).19'20 This model geographically characterizes emissions from ships traveling along
shipping lanes to and from individual ports, in addition to the emissions from vessels transiting
near the ports. The shipping lanes were identified from actual ship positioning reports. The
model then uses detailed information about ship destinations, ship attributes (e.g., vessel speed
and engine horsepower), and emission factors to produce spatially allocated (i.e., gridded)
emission estimates for ships engaged in foreign commerce.

       This modeling was performed to estimate interport emissions from main propulsion and
auxiliary engines used by Category 3 ocean-going vessels operating in the modeling domain.
The modeling domain consists of the entire area around Puerto Rico and the U.S. Virgin Islands
that is subject to the sovereign authority of the United States.
                                        2-14

-------
2.4.2.1  Interport Inventory Methodology

       The interport emissions were estimated using the Waterway Network Ship Traffic,
Energy, and Environmental Model (STEEM).21'22  STEEM was developed by the University of
Delaware as a comprehensive approach to quantify and geographically represent interport ship
traffic, emissions, and energy consumption from large ocean-going vessels. The model estimates
emissions from main propulsion and auxiliary marine engines used on Category 3  vessels that
engage in foreign commerce using historical shipping activity, ship attributes (i.e.,
characteristics), and activity-based emission factor information.  These inputs are assembled
using a GIS platform that also contains an empirically derived network of shipping lanes.  It
includes the emissions for all ship operational modes from cruise in unconstrained shipping lanes
to maneuvering in a port.  The model, however, excludes hotelling operations while the vessel is
docked or anchored, and very low speed maneuvering close to a dock.  For that reason, STEEM
is referred to as  an "interport" model, to easily distinguish it from the near ports analysis.

       STEEM uses advanced ArcGIS tools and develops emission inventories in the following
way.  The model begins by building a spatially-defined waterway network based on empirical
shipping location information from two global ship reporting databases. The first is the
International Comprehensive Ocean-Atmosphere Data Set (ICOADS), which contains reports on
marine surface and atmospheric conditions from the Voluntary Observing Ships (VOS) fleet.23
There are approximately 4,000 vessels worldwide in the VOS system. The ICOADS project is
sponsored by the National Oceanic and Atmospheric Administration and National Science
Foundation's National Center for Atmospheric Research (NCAR).  The second database is the
Automated Mutual-As si stance Vessel Rescue (AMVER) system.24 The AMVER data set is
based on a ship search and rescue reporting network sponsored by the U.S. Coast Guard.  The
AMVER system is also voluntary, but is generally limited to ships over 1,000 gross tons on
voyages of 24 hours or longer. About 8,600 vessels reported to AMVER in 2004.

       The latitude and longitude coordinates for the ship reports in the above databases are used
to statistically create and spatially define the direction and width of each shipping lane in the
waterway network. Each statistical lane (route and segment) is given a unique identification
number for computational purposes. For the current analysis, STEEM used 20 years of ICOADS
data (1983-2002) and about one year of AMVER data (part of 2004 and part of 2005). This is
illustrated in Figure 1-1.
                                        2-15

-------
                            Figure 2-1 AMVER and ICOADS data

       Every port is also spatially located in the waterway network using ArcGIS software.

       As illustrated in Figure 2-2, the waterway network represented by STEEM resembles a
highway network on land.  It is composed of ports, which are origins and destinations of
shipping routes: junctions where shipping routes intersect, and segments that are shipping lanes
between two connected junctions. Each segment can have only two junctions or ports, and ship
traffic flow can enter and leave a segment only through a junction or at a port.  The figure
represents only a sample of the many routes contained in the model.
      Figure 2-2 Illustration of STEEM Modeling Domain and Spatial Distribution of Shipping Lanes

       The STEEM interport model also employs a number of databases to identify the
movements for each vessel (e.g., trips), individual ship attributes (e.g., vessel size and
horsepower), and related emission factor information (e.g., emission rates) that are subsequently
used in the inventory calculations.
                                        2-16

-------
       To allocate ships to the statistical lanes, STEEM uses ArcGIS Network Analyst tools
along with specific information on each individual ship movement to solve the most probable
path on the network between each pair of ports (i.e., a trip) for a certain ship size.  This is
assumed to represent the least-energy path, which in most cases is the shortest distance unless
prevented by weather or sea conditions, water depth, channel width, navigational regulations, or
other constraints that are beyond the model's capability to forecast.

       After identifying the shipping route and resulting distance associated with each unique
trip, the emissions are simply calculated for each operational mode using the following
generalized equation along with information from the ship attributes and emission factor
databases:

                                       Equation 2-5

       Emissions per trip = distance (nautical miles) / speed (nautical miles/hour) x horsepower (kW) x
                         fractional load factor x emission factor (g/kW-hour)

       In STEEM, emissions are calculated separately for distances representing cruise and
maneuvering operational modes.  Maneuvering occurs at slower speeds and load factors than
during cruise conditions.  In STEEM, maneuvering is assumed to occur within a 20 kilometers
radius of each port when a ship is entering or leaving a port. A ship is assumed to move at
maneuvering speed for an entire trip if the distance is less than 20 kilometers.

       Finally, the emissions along each shipping route  (i.e., segment) for all trips are
proportioned among the respective cells that are represented by the gridded modeling domain.
For this work, emissions estimates were produced at a cell resolution of 4 kilometers by 4
kilometers, which is appropriate for most atmospheric air quality models. The results  for each
cell are then summed, as appropriate, to produce emission inventories for the various  geographic
regions of interest in this analysis.

2.4.2.2  Data Inputs for Interport Emission Inventories

       The STEEM model includes detailed information about ship routes and destinations in
order to provide spatially allocated emissions of ships in transit.  The shipping lanes and
directions were empirically derived from ship positioning data in several datasets. The
International Comprehensive Ocean-Atmosphere Data Set (ICOADS) contains reports on marine
surface and atmospheric conditions from the Voluntary Observing Ships (VOS) fleet.25 STEEM
also uses a dataset derived from the Automated Mutual-Assistance Vessel Rescue (AMVER)
system,26 which is based on a ship search and rescue reporting network sponsored by  the U.S.
Coast Guard. Traffic along each of these lanes is derived from USAGE entrance and  clearance
            T7
data for 2002,   together with Lloyd's Register-Fairplay Ltd's data for ship characteristics.
Information for number of calls, ship characteristics, propulsion engine power, and cruise speed
were obtained from these data.

       The emission factors and load factors used  as inputs to  STEEM are very similar to those
used for the ports analysis. Additional adjustments were made to interport emission results for
      and SO2 in order to reflect recent U.S. Government review of available engine test data and
                                         2-17

-------
fuel sulfur levels.  Details of the STEEM emission inputs and adjustments are located in
Appendix 2C.

2.4.3 Total Ship Inventory for 2002

       The national and regional inventories in this study are a combination of the results from
the near ports analysis and the STEEM interport modeling. These two inventories are quite
different in form.  The STEEM characterizes emissions from vessels while traveling between
ports. That includes when a vessel is maneuvering to enter or exit a port, cruising near a port as
it traverses the area, or moving in a shipping lane across the open sea.  The results are spatially
reported in a gridded format that is resolved to a cell dimension of 4 kilometers by 4 kilometers.
The near port results, on the other hand, reflect emissions that occur inside of or within 25 miles
of twelve specific ports and are not reported in a gridded format.

       Therefore, to obtain the total inventory for 2002 it is necessary to spatially allocate the
emissions in a format that is compatible with the STEEM 4 kilometers by 4 kilometers gridded
output. Once that has been accomplished, the two inventories can be blended together. Both of
these processes are described below.  This work was conducted by ENVIRON International as a
subcontractor under the U.S. Government contract with ICF.

2.4.3.1  Spatial Location of the Near Port Inventories

       The hotelling, maneuvering, RSZ, and cruise emissions from the near port inventories
were spatially located by their respective latitude and longitude coordinates using ArcGIS
software. For this study, shapefiles were created that depicted the emission locations as
described above. These shapefiles and the STEEM output can be layered upon each other,
viewed in ArcMap, and analyzed together. The following sections provide a more detailed
description of how the shapefiles representing the ports, RSZ lanes,  and cruise lanes were
developed.

      Hotelling and Maneuvering emissions

      Each port, and thus the designated location for hotelling and maneuvering emissions, is
modeled as a single latitude/longitude coordinate point using the estimated port center. The
hotelling and maneuvering emissions represented by the latitude/longitude coordinate for each
port were subsequently assigned to a single cell in the gridded inventory where that point was
located. It should be noted that modeling a port as a point will over specify the location of the
emissions associated with that port if it occupies an area greater than one grid cell, or 4
kilometers by 4 kilometers.  The coordinates of the 12 ports used in  this work are shown in
Appendix 2A.

      RSZ emissions

       The RSZ routes associated with each of the 12 ports were modeled as lines.  Each RSZ
was assumed to be 10 nautical miles in length.

       The RSZ emissions were distributed evenly along the length of the line. The
latitude/longitude coordinates for each point along the line were  subsequently used to assign the
                                        2-18

-------
emissions to a grid cell based on the proportion of the line segment that occurred in the
respective cell.

       Cruise emissions

       The cruise mode links that extend 25 nautical miles from the end of the RSZ end point
were also modeled with line shapefiles.  These links were spatially described for each port
following the direction of the shipping lane evident in the STEEM data. Again, as with RSZ
emissions, the latitude/longitude coordinates for each point along the line were subsequently
used to assign the emissions to a grid cell based on the proportion of the line segment that
occurred in the respective cell.

2.4.3.2  2002 Inventory - Port and Nearport

       After spatially defining the geographic location of the near port emissions, but before
actually inserting them into the gridded STEEM inventory, it was necessary to determine if all of
the STEEM emissions within an affected cell should be replaced, or if some  of the emissions
should be retained. In this latter case, ships would be traversing the area near a port, but not
actually entering or exiting the port.

       The percentage of STEEM emissions that are attributable  to a port, and should be
removed and replaced, was approximated by dividing the STEEM emissions in the isolated
portion of the route that lead only to the port, with the STEEM emissions in the major shipping
lane.

       The actual merging of the two inventories was performed  by creating a number of
databases that identified the fraction of the near port inventory for each pollutant species and
operating mode that should be added to the grid cells for each port.  A similar database was also
created that identified how much of the original STEEM emissions should be reduced to account
for ship movements associated directly with a port, while preserving those that represented
transient vessel traffic. These databases were subsequently used to calculate the new emission
results for each affected cell in the original STEEM gridded inventory, resulting in the combined
inventory results for this study.

       For the San Juan port, the outer edges of the port inventories fell outside the Caribbean
inventory domain; that portion outside the domain was removed.  As a result, the port totals
presented in the next section are slightly less than  those reported in Section 2.4.1. The removed
portion represents less than 4 percent of the total port emissions.

       The total inventory was created by summing emission estimates for ships while at port
and while underway (interport). The total 2002 inventory for the  Caribbean inventory domain,
along with the relative contributions of the port and interport emissions are presented in Table
2-11.
                                        2-19

-------
                Table 2-11 2002 Total C3 Inventory for Caribbean Inventory Domain
EMISSION TYPE
Port
Interport
Total Emissions
ANNUAL EMISSIONS (METRIC TONNES)3
NOX PI
8,955
19,358
28,313
« 10 P
807
1,512
2,319
VI 2.5
742
1,391
2,134
HC
282
642
923
CO
697
1,511
2,208
SO2C(
6,869
11,219
18,088
1 2
408,456
691,419
1,099,875
        a The port emission totals in this table are slightly less than those in Table 2-10 due to the
        gridding process and trimming to include only port emissions that fall within the inventory
        boundaries.

       The interport and port inventories are about 70 percent and 30 percent of the total,
respectively.

  2.5 Development of 2020 Inventories

       To obtain the 2020 baseline and control inventories for the inventory domain, it is
necessary to adjust the 2002 inventories to account for activity level growth and the emission
reductions that would  occur in 2020 absent the EGA controls (baseline case) and with the EGA
controls (control case). This section describes how the adjustment factors were obtained and
presents the inventories for the inventory domain for 2020. The inventories for the proposed
U.S. Caribbean EGA are described in Section 2.6, below.

2.5.1 Adjustment Methodology

       We used  a multi-step approach to adjust the 2002 inventories to estimate the 2020
baseline and control scenarios for the inventory domain.  Specifically, we apply a growth factor
adjustment and an emission factor adjustment.

       The growth factor adjustment is derived from the growth factors that were estimated for
the North American EGA.

       The emission factor adjustments are derived by developing a new set of emission factors
based on the emission  programs that will be in place in the baseline and control scenarios; the
adjustment factor is the ratio of the 2002 emission factors to the 2020 emission factors.

2.5.1.1  Growth Factors for 2020

       The starting point for developing the 2020 inventories is to determine the average annual
growth rates from 2002 through 2020.  The average annual growth rate for the inventory domain
is derived from the average annual growth rates estimated for the North American EGA.  These
were estimated for seven regions within the U.S. EEZ. The seven regions are Alaska, East
Coast, Gulf Coast, Hawaii, North Pacific, South Pacific, and Great Lakes.  The definition of
these regions and the methodology used to derive these growth rates are described in Appendix
2D.
                                        2-20

-------
       From an examination of the shipping routes within the Caribbean EEZ, it appears that
ships enter from all of the regions except the Great Lakes. As a result, the growth rate for the
Caribbean EEZ was derived as a power-weighted average of the six regional growth rates.  The
growth rate is then compounded over the inventory projection time period for 2020 (i.e., 18
years). The growth rates and resulting multiplicative growth factors for each of the regions and
the Caribbean EEZ are provided in Table 2-12.

                     Table 2-12 Emission Inventory Growth Factors for 2020
REGION
Alaska
East Coast
Gulf Coast
Hawaii
North Pacific
South Pacific
Caribbean domain (wgt avg)
TOTAL
PROPULSION
POWER (MW)
14,931
865,085
319,976
38,353
94,796
571,433

2002-2020 AVERAGE
ANNUALIZED
GROWTH RATE (%)
3.3%
4.5%
2.9%
5.0%
3.3%
5.0%
4.3%
MULTIPLICATIVE
GROWTH FACTOR
RELATIVE TO 2002
1.79
2.21
1.67
2.41
1.79
2.41
2.16
       The multiplicative growth factor for the Caribbean inventory domain is applied to each of
the pollutant totals for 2002 to project emissions to 2020.  Additional adjustments are required to
account for emission controls, which are described in the following sections.

2.5.1.2  Emission Requirements Included in the Adjustment:  Baseline and Control

       The emission adjustment factor is developed to reflect the control programs that will be
in place in 2020 in both the baseline and control scenarios, compared to the 2002 scenario.

       By 2020, ships will be required to comply with the MARPOL Annex VI Tier INOX
standard for marine diesel engines that became effective in 2000, as well as the Tier II standard
that will become effective in 2011. Also included in the 2020 baseline inventories is the NOx
retrofit program for pre-controlled engines in regulation 13 of MARPOL Annex VI.

       The EGA requirements will add two other requirements.  First, ships  will be required to
use fuel with a sulfur content not to exceed 0.10%.  Although the 0.10% fuel sulfur requirement
goes into place for  all  vessels operating in EGAs  beginning in 2015, the use  of 2020 as the
analytic year will still provide a representative scenario for the impact of the 0.10% fuel sulfur
requirement on human health  and the environment.  This is because the fuel requirements of the
EGA go into effect all at once; there is no phase-in.  So the impacts of the fuel requirement in
2020 are expected to be the same as in 2015, with  a small increase due to growth.

       The EGA program also requires ships constructed on or after January 1, 2016 to be
equipped with engines that meet the Tier III NOx limits. While 2020 will include five years of
turnover to the Tier III standards, the long service lives of engines on ocean-going vessels mean
that these impacts will be small and affect less than 25% of the total fleet, assuming an average
20-year service life.  These NOx reductions would not increase the benefits  of the program by
very much, if any.
                                        2-21

-------
       The modelling presented here estimates the expected effect of shipping emissions in
2020. The year 2020 was chosen because it allows the use of detailed emission inventories that
were created for other emission sources (e.g., land-based stationary and mobile sources) as part
of wider scale air pollution modelling efforts.  This allows us to compare the ship emission
inventories to total anthropogenic emission inventories for Puerto Rico and the Virgin Islands.
The choice of 2020 is also consistent with the fuel cost analysis.

       The use of 2020 has two implications for the inventory analysis. First, with regard to the
impacts of the EGA fuel sulphur requirements, the choice of 2020 slightly over-estimates the
immediate benefits of the program in 2015.  However, since the fuel controls apply to all vessels
beginning in 2015 (there is no phase-in), the estimated impacts of the fuel requirement in 2020
are expected to be similar to the impacts in 2015, with the difference due to growth in the marine
transportation sector. Therefore, the use of 2020 as the analytic year will still provide a
representative scenario for the impact of the 0.1 percent fuel sulphur requirement on human
health and the environment.  Second, with regard to the NOx impacts, the use of 2020 includes
only five years of turnover to the Tier III standards.  Because of the long service lives of engines
on ocean-going vessels, this mean that the fleet will not be fully turned over for some time and
therefore the full benefits of the EGA NOx controls are not reflected in the analysis. In
conclusion, the choice of 2020 as the analytic year provides a balance between modelling too
early of a year where the Tier III NOx standards may not yet apply and modelling too late of a
year where there may be more uncertainty associated with projecting emissions into the future.
It should be noted that, although the 0.5% global fuel sulphur standard goes into effect in 2020,
we  did not include the global standard in the 2020 analysis. This approach provides an estimate
of benefits in the early (pre-2020) years of the program.

       The effects of these  controls are reflected in the 2020 emission inventories by applying
appropriate adjustment factors that reflect the percentage of the vessel fleet in those years that
are estimated to comply with the controls.  Adjustment factors are ratios of 2020 to 2002
calendar year (CY) emission factors (EFs). Adjustment factors are derived separately by engine
type for propulsion and auxiliary engines. The adjustment factors for propulsion engines are
applied to the propulsion portion of the port inventory and the interport portion of the inventory.
The adjustment factors for auxiliary engines are applied to the auxiliary portion of the port
inventory.

2.5.1.3  Emission Factors for 2020 Inventory Adjustments

       The emission factors for the 2020 inventory adjustments reflect the application of the
controls described  above. Note that gas  and steam turbine engines are not subject to any of the
NOx standards; however, these engines are not a large part of the inventory.

       For the NOx limits, the current Tier I controls, which are modeled as achieving an 11
percent reduction from Tier 0, apply to the 2000 through 2010 model year (MY) engines.  In
2011 thru 2015, Tier II controls are applied. Tier II controls are modeled as a 2.5 g/kW-hr
reduction from Tier I. In the EGA area only, for 2016 MY engines and beyond, Tier III controls
are applied.  Tier III controls are modeled as achieving an 80 percent reduction from Tier I
levels.  The NOx retrofit program for Tier 0 (pre-control) engines was modeled as 11 percent
control from Tier 0 for 80 percent of 1990 thru 1999 MY engines greater than 90 liters per
                                         2-22

-------
cylinder (L/cyl) starting in 2011.  The retrofit program was also modeled with a five year phase-
in.  Finally, control of fuel sulfur content within the EGA area to 0.10% affects both SC>2 and PM
emissions.

       The NOx emission factors (EFs) by engine/ship type and tier are provided in Table 2-13.
Tier 0 refers to pre-control.  There are separate entries for Tier 0/1/2 base and Tier 0/1/2 control,
since the control engines would be using distillate fuel, and there are small NOx emission
reductions assumed when switching from residual to distillate fuel.28  The NOx control EFs by
tier were derived using the assumptions described above.

                         Table 2-13 Modeled NOX Emission Factors by Tier
ENGINE/
SHIP
TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
NOX EF (g/kW-hr)
BASELINE CONTROL
TIERO

18.1
14
2.1
6.1

14.6
14.5
TIERO
RETROFIT

16.1
12.5
n/a
n/a

n/aa
n/aa
TIER
I

16.1
12.5
n/a
n/a

13.0
12.9
TIER
II

13.6
10.0
n/a
n/a

10.5
10.4
AREAS
TIERO

17
13.2
2
5.7

14.6
14.5
TIERO
RETROFIT

15.1
11.7
n/a
n/a

n/aa
n/aa
TIER
I

15.1
11.7
n/a
n/a

13.0
12.9
TIER
II

12.6
9.2
n/a
n/a

10.5
10.4
TIER
III

3
2.3
n/a
n/a

2.6
2.6
              a The retrofit program applies to engines over 90 L/cyl; auxiliary engines are smaller than
              this outpoint and would therefore not be subject to the program.

       Because this program phases in over time, it is necessary to estimate the adjustment
factor for each year to obtain the appropriate adjustment factor for 2020. This is done by using
vessel age distributions (Table 2-14) to generate calendar year NOx EFs by engine/ship type for
the base and control areas included in the scenarios. The adjustment factors for 2020 for the
baseline and control scenarios are presented in Table 2-15.
                                          2-23

-------
   Table 2-14 Vessel Age Distribution for Deep Sea Ports by Engine Type
AGE
GROUP
(years old)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35+
PROPULSION ENGINE TYPE a (Fraction of Total)
MSD SSD
0.00570
0.07693
0.10202
0.08456
0.08590
0.06427
0.06024
0.07867
0.06730
0.04181
0.04106
0.03100
0.04527
0.03583
0.03519
0.02921
0.00089
0.01326
0.00847
0.00805
0.00566
0.00495
0.00503
0.00676
0.00539
0.01175
0.00803
0.00522
0.00294
0.00285
0.00254
0.00084
0.00023
0.00117
0.00132
0.01967

0.02667
0.07741
0.07512
0.07195
0.05504
0.05563
0.04042
0.07266
0.05763
0.04871
0.04777
0.03828
0.03888
0.02787
0.02824
0.01466
0.01660
0.01582
0.02414
0.01982
0.02258
0.02945
0.01883
0.01080
0.01091
0.01099
0.01045
0.00835
0.00788
0.00370
0.00106
0.00113
0.00367
0.00582
0.00092
0.00013
GT
0.00000
0.07189
0.14045
0.05608
0.67963
0.04165
0.00000
0.00626
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00034
0.00370
0.00000
0.00000
0.00000
0.00000
0.00000
ST
0.00447
0.12194
0.16464
0.05321
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.04873
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00875
0.00883
0.00883
0.18029
0.11065
0.01395
0.08657
0.02907
0.05126
0.00605
0.07105
0.00000
0.00000
0.03172
ALL
AUXILIARY
ENGINES
0.01958
0.07670
0.08426
0.07489
0.07831
0.05685
0.04455
0.07150
0.05764
0.04475
0.04364
0.03538
0.04160
0.02909
0.02935
0.01869
0.01189
0.01462
0.01966
0.01550
0.01756
0.02260
0.01467
0.00943
0.00900
0.01224
0.01130
0.00738
0.00659
0.00349
0.00193
0.00096
0.00322
0.00419
0.00098
0.00598
a MSD is medium speed diesel, SSD is slow speed diesel, GT is gas turbine, ST is steam
turbine.
                              2-24

-------
            Table 2-15 Modeled NOX Emission Factors by Calendar Year and Control Type
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
CY NOX EF (g/kW-hr)
20022

18.1
14
2.1
6.1

14.6
14.5
020 BASE

14.7
10.9
2.1
6.1

11.7
11.5
2020 ECA
CONTROL

10.8
7.7
2.0
5.7

8.6
8.6
       The PMand SO2 EFs are a function of fuel sulfur level.  For the baseline portions of the
inventory, the residual fuel sulfur level modeled is 27,000 ppm.  The baseline distillate fuel
sulfur level assumed for all areas is 15,000 ppm.  As discussed previously, for the baseline, main
engines use residual fuel and auxiliary engines use a mix of residual and distillate fuel. For the
control areas, there is one level of distillate fuel sulfur assumed to be used by all engines: 1,000
ppm for the ECA control areas.

       Table 2-16 provides the PMio EFs by engine/ship type and fuel sulfur level. For
modeling purposes, PM2.5  is assumed to be 92 percent of PMio.  The PM EFs are adjusted to
reflect the appropriate fuel  sulfur levels using Equation 2-2.

                           Table 2-16 Modeled PM10 Emission Factors
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
PM10 EF (g/kW-hr)
BASELINE CO
27,000 ppm S

.40
.40
.50
.50

.40
.20
NTROL AREAS
ECA
1,000 ppm S

0.19
0.19
0.17
0.17

0.19
0.19
       Table 2-17 provides the modeled SC>2 EFs.  SC>2 emission reductions are directly
proportional to reductions in fuel sulfur content.
                                         2-25

-------
                           Table 2-17 Modeled SO2 Emission Factors*
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
SO2 EF (g/kW-hr)
BASELINE CO
27,000 ppm S

10.29
11.09
16.10
16.10

10.70
9.66
VTROL AREAS
ECA
1,000 ppm S

0.36
0.39
0.57
0.57

0.39
0.39
       For the CO2 emission factors, CO2 is directly proportional to fuel consumed.  Table 2-18
provides the modeled CC>2 and brake specific fuel consumption (BSFC) EFs. Due to the higher
energy content of distillate fuel on a mass basis, the switch to distillate fuel for the control areas
results in a small reduction to BSFC and, correspondingly, CC>2 emissions.29

                  Table 2-18 Modeled Fuel Consumption and CO2 Emission Factors
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
EF (g/kW-hr)
BASELINE
BSFCC

195
210
305
305

210
210
) 2BS

620
668
970
970

668
668
CONTROL
AREAS
FCCO

185
200
290
290

200
200
2

589
637
923
923

636
636
       The HC and CO emission factors are assumed to remain unchanged from the 2002
scenario, since there are no emission standards or requirements for those pollutants. The ECA
NOx and fuel sulfur requirements are anticipated to reduce the NOx, SC>2 and PM emission
factors. The switch to lower sulfur distillate fuel use is also expected to lower CC>2 emissions
slightly.

2.5.1.4  Port Emission Adjustment Factors

       The EF adjustment factors are a ratio of the control EF to the 2002 EF. Table 2-19
through Table 2-23 provide the EF adjustment factors for each pollutant for the 2020 baseline
and control scenarios.
                                         2-26

-------
Table 2-19 NOX EF Adjustment Factors by Engine/Ship Type and Control Type"
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
2020
BASE

0.8130
0.7804
1.0000
1.0000

0.7985
0.7972
2020 ECA
CONTROL

0.5967
0.5515
0.9524
0.9344

0.5869
0.5940
             1NOX adjustment factors are a ratio of future base or control EFs to 2002 EFs
Table 2-20 PM10 EF Adjustment Factors by Engine/Ship Type and Control Type"
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
2020
BASE

1.0000
1.0000
1.0000
1.0000

1.0000
1.0000
2020 ECA
CONTROL

0.1352
0.1328
0.1108
0.1108

0.1328
0.1550
            PM10 adjustment factors are a ratio of the control EFs to the 2002
          EFs. PM is not adjusted for the future baseline because fuel sulfur
          levels are only assumed to change within the ECA.
Table 2-21 PM2.5 EF Adjustment Factors by Engine/Ship Type and Control Type"
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
2020
BASE

1.0000
1.0000
1.0000
1.0000

1.0000
1.0000
2020 ECA
CONTROL

0.1339
0.1316
0.1092
0.1092

0.1316
0.1555
          a PM2 5 adjustment factors are a ratio of the control EFs to the 2002
          EFs. PM is not adjusted for the future baseline because fuel sulfur
          levels are only assumed to change within the ECA. The PM2 5
          adjustment factors are slightly different from those for PM10 due to
          rounding.
                                 2-27

-------
            Table 2-22 SO2 EF Adjustment Factors by Engine/Ship Type and Control Type3
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
2020
BASE

.0000
.0000
.0000
.0000

.0000
.0000
2020 ECA
CONTROL

0.0351
0.0353
0.0352
0.0352

0.0365
0.0405
                     a SO2 adjustment factors are a ratio of the control EFs to the
                     2002 EFs. SO2 is not adjusted for the future baseline because
                     fuel sulfur levels are only assumed to change within the ECA.
            Table 2-23 CO2 EF Adjustment Factors by Engine/Ship Type and Control Type3
ENGINE/
SHIP TYPE
Main
SSD
MSD
ST
GT
Aux
Pass
Other
2020
BASE

1.0000
1.0000
1.0000
1.0000

1.0000
1.0000
2020 ECA
CONTROL

0.9488
0.9531
0.9509
0.9509

0.9525
0.9525
                     a CO2 adjustment factors are a ratio of the control EFs to the
                     2002 EFs. CO2 is not adjusted for the future baseline because
                     fuel consumption (BSFC) is only assumed to change within the
                     ECA.

2.5.1.5  Interport Emission Adjustment Factors

       Since the interport portion of the inventory is not segregated by engine or ship type, it
was necessary to develop a different set of emission adjustment factors for those emissions.  This
was done using the port-specific cruise emissions for the propulsion engines as a surrogate for
interport emissions.  This is appropriate because the majority of emissions while underway are
from propulsion, not auxiliary, engines. Also, the cruise mode best represents ship operation
while underway at sea.

       The port-specific cruise emissions for the 2020 baseline and control scenarios were
summed and ratios of these scenario totals to the 2002 totals were developed for each pollutant.
This analysis was performed separately for each of the Puerto Rico and Virgin Island ports.
These ratios were then adjusted to remove growth by dividing each by the growth factor (2.16).
Composite EF adjustment factors were then developed for all PR/USVI ports combined by
                                          2-28

-------
weighting each port's adjustment factors by the fraction of total propulsion installed power for
that port.

       The resulting EF adjustment factors applied to the 2002 interport portion of the inventory
are provided in Table 2-24 below.

                      Table 2-24 EF Adjustment Factors for 2020 Scenarios"
POLLUTANT :
NOX
PM10
PM25
SO2
C02
002
1.0000
1.0000
1.0000
1.0000
1.0000
2020
BASE
0.7986
1.0000
1.0000
1.0000
1.0000
ECA
CONTROL
0.5820
0.1320
0.1308
0.0355
0.9517
                       a Adjustment factors are ratios of future base or control EFs
                       to 2002 EFs. These adjustment factors are used to adjust
                       the interport portion of the 2002 inventory.

2.5.1.6  2020 Near Port and Interport Inventories

       The 2020 near port and interport inventories were developed by applying the growth
factors and emission factor adjustments to the 2002 inventories. These inventories were then
combined to obtain the 2020 total inventories, for the baseline and control cases.

       The interport inventories were scaled by a growth factor to 2020, as previously described,
and the emission adjustment factors were applied.

       The near port inventories were created by applying the growth and emission adjustment
factors to the 2002 near port inventories. The near port inventories were then converted into a
gridded format using the same approach as for the 2002 inventory.  Using this grid, STEEM
values were removed from near port cells and near port emissions were used as replacement
values. In cases where the emissions near ports were only partially attributable to port traffic, the
STEEM inventory was reduced rather than removed.

       Interport and near port emissions were then aggregated to form regional totals. The
resulting baseline and control inventories for 2020 are presented in Table 2-25.  The inventories
include all emissions within the Caribbean inventory domain.

          Table 2-25 Category 3 Vessel Inventories in the Inventory Domain for 2020 Scenarios"
SCENARIO
Baseline
ECA Control
ANNUAL EMISSIONS (METRIC TONNES)
NOX Pfl
48,782
35,685
I ioP
5,006
676
M 25
4,605
618
HC
1,993
1,992
CO
4,764
4,765
S02CO
39,036
1,419
2
2,373,593
2,259,323
     1 These inventories include all emissions within the Caribbean inventory domain.
                                         2-29

-------
2-26.
       The fuel consumption by fuel type in the baseline and EGA cases is presented in Table
               Table 2-26 Fuel Consumption by Category 3 Vessels for 2020 Scenarios"
SCENARIO
Baseline
ECA Control
METRIC TONNES FUEL
DISTILLATE ]
40,446
709,809
IESIDUAL
705,263
0
TOTAL
745,709
709,809
                    a These inventories include all emissions within the Caribbean inventory domain.

  2.6 Inventories for Proposed ECA

       The size and shape of the proposed ECA differs from that of the Caribbean inventory
domain.  The inventory domain used in the above consists of the entire area around Puerto Rico
and the U.S. Virgin Islands that is subject to the sovereign authority of the United States, which
is the exclusive economic zone surrounding these islands. The proposed ECA is a subset of this
area, and includes waters adjacent to coasts of the Commonwealth of Puerto Rico and the U.S.
Virgin Islands. The northern and southern boundaries of the proposed area would extend
roughly 50 nm and 40 nm, respectively, from the territorial sea baseline of the main island of
Puerto Rico.  The western edge of the proposed area would generally run north-south, about half
way between the Puerto Rican island of Mona, and the west coast of the main island.  The
eastern edge of the proposed area would generally run north-south, but extend eastward through
the area between the U.S. Virgin Islands and the British Virgin Islands and also eastward through
the area between Saint Croix and Anguilla and Saint Kitts.

       Because the port and interport inventories described above are spatially allocated, with
every location assigned an appropriate quantity of emission and with the total inventory
equivalent to the sum of all locations, it is a straightforward matter to estimate the inventories for
the proposed ECA.

       Specifically, to estimate the inventories for the proposed ECA, the boundaries of the
proposed ECA, as described in Section 5 of the Information Paper, were overlaid upon the
spatially explicit inventory (Figure 2-3).  All gridded emissions cells within this ECA boundary
were summed, with the totalled grid cells being equivalent to the inventory of the Caribbean
ECA.
                                        2-30

-------
                 Inventory Domain

                 Proposed EGA
                                                                          O
0 25  50    100
          INM
                  Figure 2-3: Boundary of the Inventory Domain and Proposed ECA

       The 2020 inventories for the proposed ECA, for the baseline and control scenarios, are
presented in Table 2-27. Also presented are the tones reduced and the percent reductions for
each pollutant.  This information shows that the proposed ECA includes about 75 percent of the
total emissions in the inventory domain.  More importantly, as shown in Chapter 3, these
emissions are most likely to reach shore.
                   Table 2-27 C3 Emission Inventories for Proposed ECA in 2020
EMISSION TYPE
Reference
Control
Delta Emissions
Delta Emissions (%)
ANNUAL EMISSIONS (METRIC TONNES)a'b
NOXP
36,950
27,032
-9,919
-27%
VI ioP
3,793
512
-3,342
-86%
M 2.5C
3,488
471
-3,017
-86%
HC
1,509
1,509
0
0%
CO
3,609
3,609
0
0%
S02CO
29,568
1,075
-28,493
-96%
2
1,797,909
1,711,452
-86,457
-5%
                                         2-31

-------
        " The ship inventories include emissions within the proposed EGA.
        * For this analysis, the commercial marine vessel emissions inventory does not include ships
        powered by "Category 1" or "Category 2" (i.e., <30 L/cyl) engines. These smaller engines
        installed on U.S.-flag vessels are already subject to strict national standards affecting NOX, PM,
        and fuel sulphur content. Engines above 130 kW but less than 30 L/cyl on foreign-flag vessels
        are covered by Annex VI; however, the Annex VI reductions for these vessels have not been
        included in the analysis.
        c The PM25 inventories include directly-emitted PM25 only.
  2.7 Other Inventories

       Inventories were developed for other types of air emissions sources, to calculate the
percent that C3 marine vessels would contribute to the sum of emissions affecting populations
and the environment in Puerto Rico and the U.S. Virgin Islands in 2020. The categories of
sources considered in this analysis include land-based mobile and stationary sources, including
aircraft.

       The U.S. EPA periodically updates its national emission inventory (NEI) forecasts, and
                                                                   30
                                                                     However, for some
much of the data were taken from these nationally prepared inventories.
sources, additional calculations were made, as described below.

2.7.1 Overview of 2020 Non-C3 Emission Inventories
       The emissions from mobile non-road and on-road sources were taken directly from the
NEI projections for 2020. These account for expected growth as well  as current domestic
regulations that will apply in 2020.  The emissions from non-point and small stationary sources
were also taken directly from NEI projections for 2020.

       The total emissions projected to be emitted from non-C3 sources in 2020 are presented in
Table 2-28. The methods for estimating the major stationary source and aircraft emissions are
described in the next section.

   Table 2-28: Projected 2020 Emissions from Non-C3 Sources in Puerto Rico and the U.S. Virgin Islands
SOURCE CATEGORY
Stationary Sources
Highway and Nonroad Gasoline
and Diesel Vehicles
Aircraft
Total Non-C3 Metric Tonnes
2020 ANNUAL EMISSIONS
NOX
44,000
15,000
2,200
62,000
PM2 5 SO
8,700
1,000
80
9,800
2
51,000
200
60
52,000
2.7.1.1  Major stationary sources

       Emissions from major stationary sources (industries, utilities), were gathered from annual
reports submitted by the facilities to local authorities for the reporting year 2008.31  It is general
practice in air quality planning to assume no net growth in stationary source air emission
inventories. Overall, growth from these sources is assumed to be balanced by improved
                                         2-32

-------
emission controls that must be applied when facilities are expanded.  Thus, the 2020 emissions
projections presented above are equal to the 2008 emissions calculated based on reported data
from these sources.

       For these sources, PM emissions are typically expressed in terms of primary filterable
PMio plus an estimate of the mass of particles that are formed by condensation after the exhaust
gases exit the stack. This expression of PM differs from the standard expression of PM2 5.
Furthermore, the ratio of PM2.5 to PMio varies by source and fuel type. For this inventory
analysis, the PM emissions estimates collected from the major stationary sources were assumed
equal to PM2.5.  With most major stationary sources well controlled for PM emissions with either
some form of aftertreatment or the use of clean fuels, this is a fair assumption. With the
uncertainty in these PMio to PM2.5 conversions, it is possible the resulting inventory may slightly
underestimate or overestimate the contribution of C3 marine vessels to total man-made PM2.s
emissions. For the 2020 case without the EGA, the confidence interval for the marine percent of
all man-made sources is plus or minus ten percent.  For the 2020 case with the EGA, the
confidence interval is plus or minus two percent.

2.7.1.2  Aircraft

       Estimates of PM, 862, and some NOx emissions in 2020 from aircraft in Puerto Rico and
the U.S. Virgin Islands  were taken directly from the NEI. While it is expected that these values
may be underestimated, no better data are available. It is noted that the resulting inventory may
slightly overestimate the contribution of C3 marine vessels to total man-made SO2 and PM
emissions.

       Estimates of NOx emissions from itinerant commercial air  carrier operations were
calculated by applying an emission factor (in grams of NOx per landing and take off, g/LTO) to
reported aircraft operation statistics.  For the U.S. Virgin Islands, the NEI estimate was taken to
represent local aircraft operations, as no significant long range commercial aircraft  operate on
these small islands. For Puerto Rico,  non-military  aircraft operations data for 2002 were taken
from the U.S. Federal Aviation Association.32  The itinerant air carrier operations data were
multiplied by an emission factor of 10,968 g/LTO,  a 2009 fleet average,33 while the NEI estimate
was taken to represent local operations.  To project 2020 emissions for Puerto Rico, the 2002
aircraft operations data  was multiplied by  the aircraft NOx growth rate from the NEI, 1.3.

  2.8 Conclusion

       An emission inventory for ships in PR/USVI was developed based on the latest state of
the art models and inputs, using a "bottom-up" methodology. The inventory includes emissions
for 12 ports, as well as emissions for ships while underway within  the area of the proposed EGA.
In addition, an emission inventory for other man-made pollution sources in Puerto Rico and the
U.S. Virgin Islands was developed for purposes of comparison. The analysis shows that a
comprehensive  review was made of air emissions sources, and that ships are contributing
significantly to  air pollution in Puerto Rico and the U.S.  Virgin Islands.
                                        2-33

-------
1ICF International (October 2007). Commercial Marine Port Inventory Development, prepared for the U.S.
Environmental Protection Agency, EPA Report Number EPA-420-R-07-012c, Docket ID EPA-HQ-OAR-2007-
0121-0063.1.

2 U.S. Army Corps of Engineers Navigation Data Center, Vessel Entrances and Clearances, 2002, available at
http://www.iwr.usace.army.mil/ndc/db/entclrn/data/entrclrn02/

3  U.S. Army Corps of Engineers Navigation Data Center, Vessel Entrances and Clearances, 2002, available at
http://www.iwr.usace.army.mil/ndc/db/entclrn/data/entrclrn02/

4  ICF International (October 2007). Commercial Marine Port Inventory Development, prepared for the U.S.
Environmental Protection Agency, EPA Report Number EPA-420-R-07-012c, Docket ID EPA-HQ-OAR-2007-
0121-0063.1.

5  Nexus Media Communications,  The Motor Ship's Guide to Marine Diesel Engines 2005, available at
http://www.motorship.com/

6 California Air Resources Board (September 2005). 2005 Oceangoing Ship Survey, Summary of Results.

7 Starcrest Consulting Group (June 2004). Port-Wide Baseline Air Emissions Inventory, prepared for the Port of Los
Angeles

8 Entec UK Limited (2002). Quantification of Emissions from Ships Associated with Ship Movements between
Ports in the European Community, prepared for the European Commission, Docket ID EPA-HQ-OAR-2007-0121-
0059.

9 U.S. Environmental Protection Agency (January 2009). Main Engine CO and HC Emission Factors in C3 Model
and Current Literature, Memorandum from Ari Kahan to Docket EPA-HQ-OAR-2007-0121.

10 U.S. Environmental Protection Agency (September 2007). Estimation of Paniculate Matter Emission Factors for
Diesel Engines on Ocean-Going Vessels, Memorandum from Mike Samulski to Docket EPA-HQ-OAR-2007-0121,
Docket ID EPA-HQ-OAR-2007-0121-0060.

11 U.S. Environmental Protection Agency (September 2007). Estimation of Paniculate Matter Emission Factors for
Diesel Engines on Ocean-Going Vessels, Memorandum from Mike Samulski to Docket EPA-HQ-OAR-2007-0121,
Docket ID EPA-HQ-OAR-2007-0121-0060.
 2 Memo from Chris Lindhjem of ENVIRON, PM Emission Factors, December 5, 2005.
13 U.S. Environmental Protection Agency, Exhaust and Crankcase Emission Factors for Nonroad Engine Modeling -
Compression Ignition (April 2004). Appendix C, EPA- 420-P-04-009, available online at
http://www.epa.gov/otaq/models/nonrdmdl/nonrdmdl2004/420p04009.pdf. Docket ID EPA-HQ-OAR-2003-0190-
0411.

14 Entec UK Limited (2002). Quantification of Emissions from Ships Associated with Ship Movements between
Ports in the European Community, prepared for the European Commission, Docket ID EPA-HQ-OAR-2007-0121-
0059.
                                              2-34

-------
15 RTI International (December 2006). Global Trade and Fuels Assessment - Future Trends and Effects of
Designation Requiring Clean Fuels in the Marine Sector: Task Order No. 1, Draft Report, prepared for the U.S.
Environmental Protection Agency, EPA Report Number EPA420-D-07-006, Docket ID EPA-HQ-OAR-2007-0121-
0063.3.
16
  California Air Resources Board (September 2005). 2005 Oceangoing Ship Survey, Summary of Results.
  Entec UK Limited (2002). Quantification of Emissions from Ships Associated with Ship Movements between
Ports in the European Community, prepared for the European Commission, Docket ID EPA-HQ-OAR-2007-0121-
0059.

18 Energy and Environmental Analysis Inc. (February 2000). Analysis of Commercial Marine Vessels Emissions and
Fuel Consumption Data, EPA420-R-00-002, available online at http://www.epa.gov/otaq/models/nonrdmdl/c-
marine/r00002.pdf.

19 Corbett, J.  et al. (April 2007). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Final Report, prepared by University of Delaware for the California Air Resource Board, Contract
Number 04-346, and the Commission for Environmental Cooperation in North America, Contract Number 113.111,
Docket ID EPA-HQ-OAR-2007-0121-0063.2.

20 Corbett, J.  et al. (May 2006). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Tasks 1 and 2:  Baseline Inventory and Ports Comparison, Final Report, prepared by University of
Delaware for the California Air Resource Board, Contract Number 04-346, and the Commission for Environmental
Cooperation in North America, Contract Number 113.111, May 2006, Docket ID EPA-HQ-OAR-2007-0121-0013.

21 Corbett, J.  et al. (April 2007). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Final Report, prepared by University of Delaware for the California Air Resource Board, Contract
Number 04-346, and the Commission for Environmental Cooperation in North America, Contract Number 113.111,
Docket ID EPA-HQ-OAR-2007-0121-0063.2.

22 Corbett, J.  et al. (May 2006). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Tasks 1 and 2:  Baseline Inventory and Ports Comparison, Final Report, prepared by University of
Delaware for the California Air Resource Board, Contract Number 04-346, and the Commission for Environmental
Cooperation in North America, Contract Number 113.111, May 2006, Docket ID EPA-HQ-OAR-2007-0121-0013.

23 Corbett, J.  et al. (May 2006). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Tasks 1 and 2:  Baseline Inventory and Ports Comparison, Final Report, prepared by University of
Delaware for the California Air Resource Board, Contract Number 04-346, and the Commission for Environmental
Cooperation in North America, Contract Number 113.111, May 2006, Docket ID EPA-HQ-OAR-2007-0121-0013.

24 Corbett, J.  et al. (May 2006). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Tasks 1 and 2:  Baseline Inventory and Ports Comparison, Final Report, prepared by University of
Delaware for the California Air Resource Board, Contract Number 04-346, and the Commission for Environmental
Cooperation in North America, Contract Number 113.111, May 2006, Docket ID EPA-HQ-OAR-2007-0121-0013.

25 Corbett, J.  et al. (May 2006). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Tasks 1 and 2:  Baseline Inventory and Ports Comparison, Final Report, prepared by University of
Delaware for the California Air Resource Board, Contract Number 04-346, and the Commission for Environmental
Cooperation in North America, Contract Number 113.111, May 2006, Docket ID EPA-HQ-OAR-2007-0121-0013.
                                              2-35

-------
26 Corbett, J. et al. (May 2006). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Tasks 1 and 2:  Baseline Inventory and Ports Comparison, Final Report, prepared by University of
Delaware for the California Air Resource Board, Contract Number 04-346, and the Commission for Environmental
Cooperation in North America, Contract Number 113.111, May 2006, Docket ID EPA-HQ-OAR-2007-0121-0013.

27 U.S. Army Corps of Engineers Navigation Data Center (2002), Vessel Entrances and Clearances available at
http://www.iwr.usace.army.mil/ndc/db/entclrn/data/entrclrn02/

28 Entec UK Limited (2002). Quantification of Emissions from Ships Associated with Ship Movements between
Ports in the European Community, prepared for the European Commission, Docket ID EPA-HQ-OAR-2007-0121-
0059.

29 Entec UK Limited (2002). Quantification of Emissions from Ships Associated with Ship Movements between
Ports in the European Community, prepared for the European Commission, Docket ID EPA-HQ-OAR-2007-0121-
0059.

30 U.S. Environmental Protection Agency (2007) National Emission Inventory Data, Version 3 of 2002 NEI, State
Tier 2 Sector Summary. Available at http://www.epa.gov/ttn/chief/net/2002inventory.html#inventorydata. Filename
"pf02v3\projections\2020cc-2002cc_20070925.xls."

31 Email dated June 10, 2009 from Luis R. Sierra Torres, Chief Inspection and Compliance Division, Air Quality
Area, Puerto Rico Environmental Quality Board, with attached file "E.xls;" and email dated September 3, 2009 from
Verline Marcellin of the US VI Department of Planning and Natural Resources, submitting the stationary source
emission inventory for 2004 to 2008 via attached file "emissions-revised.xls."

32 Federal Aviation Administration, APO Terminal Area Forecast Detail 2002 Report,
http://aspm.faa.gov/main/taf.asp, accessed September 2009.

33 "Aircraft NOx Emissions Limitation", submitted by the United States to the ICAO Committee on Aviation
Environmental Protection, Working Group 3: Emissions Technical (CAEP/8-WG3-WP/6-14, October 2009).
                                               2-36

-------
Appendices
Appendix 2A:  Port Coordinates
                                   Table 2A-1 Port Coordinates"
Port Name
Ponce, PR
San Juan, PR
Fajardo, PR
Jobos, PR
Guayanilla, PR
Mayaguez, PR
Yabucao, PR
Christiansted, St. Croix, VI
St. Thomas, VI
Frederiksted, St. Croix, VI
Port Harvey, St. Croix, VI
Port Hess, St. Croix, VI
USAGE
CODE
C2151
C2136
C2139
n/ac
n/a
n/a
n/a
C2157
C2143
n/a
n/a
n/a
PORT COORDINATES
Longitude
-66.716670
-66.166670
-65.648401
-66.1876488
-66.7530155
-67.1564198
-65.834481
-64.732420
-64.899990
-64.8857045
-64.7709
-64.7473556
Latitude
18.001260
18.666670
18.324173
17.95325968
17.98942757
18.20260974
18.0534
17.752710
18.350010
17.7142481
17.70753889
17.69643056
              a US Army Corps of Engineers (USAGE) data from http://www.iwr.usace.army.mil/ndc/db/pport/dbf/
              Other locations from http://maps.google.com and online information searches.
              Harvey is an alumina bauxite refinery and Hess is an oil terminal.
              b n/a = not applicable
                                           2-37

-------
Appendix 2B:  Port Methodology and Equations

       Near port emissions for each port are calculated for four modes of operation: 1) hotelling, 2)
maneuvering, 3) reduced speed zone (RSZ), and 4) cruise. Hotelling, or dwelling, occurs while the
vessel is docked or anchored near a dock, and only the auxiliary engine(s) are being used to provide
power to meet the ship's energy needs. Maneuvering occurs within a very short distance of the
docks.  The RSZ varies from port to port, though generally the RSZ would begin and end when the
pilots board or disembark, and typically occurs when the near port shipping lanes reach
unconstrained ocean shipping lanes. The cruise mode emissions in the near ports analysis extend 25
nautical miles beyond the end of the RSZ lanes for the PR/US VI deep water ports.

       Emissions are calculated separately for propulsion and auxiliary engines. The basic
equation used is as follows:
Emissionsmode[eng]

 Where:
                                        Equation 2B-1
                  = (calls)x(P[eng] )x(hrs I call mode)x(LFmode[eng] )x(EF[eng] )x(^')x(10^ tonnes I g)
                  ng] = Metric tonnes emitted by mode and engine type
   Calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
   P[eng] = Total engine power by engine type, in kilowatts
   hrs/callmode = Hours per call by mode
   LFmode [eng] = Load factor by mode and engine type (unitless)
   EF[eng] = Emission factor by engine type for the pollutant of interest, in g/kW-hr
       (these vary as a function of engine type and fuel used, rather than activity mode)
   Adj = Low load adjustment factor, unitless (used when the load factor is below 0.20)
   10"6 = Conversion factor from grams to metric tonnes

       Main engine load factors are calculated directly from the propeller curve based upon the
cube of actual speed divided by maximum speed (at 100% maximum continuous rating [MCR]).  In
addition, cruise mode activity is based on cruise distance and speed inputs. The following sections
provide the specific equations used to calculate propulsion and auxiliary emissions for each activity
mode.

Cruise

       Cruise emissions are calculated for both propulsion (main) and auxiliary engines.  The basic
equation used to calculate cruise mode emissions for the main engines is:
                                        Equation 2B-2
 Emissionscmise[mam] = (calls) x (P[mam]) x (hrs I callcmise) x (LFcrmse[mam]) x (EF[mam]) x (1 (T6 tonnes I g)

   Where:
   Emissionscruise [main] = Metric tonnes emitted from main engines in cruise mode
   Calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
   P[main] = Total main engine power, in kilowatts
                                          2-38

-------
          hrs/callcraise = Hours per call for cruise mode
          LFcruise [main] = Load factor for main engines in cruise mode (unitless)
          EF[ma;n] = Emission factor for main engines for the pollutant of interest, in g/kW-hr (these vary
              as a function of engine type and fuel used, rather than activity mode)
          10"6 = Conversion factor from grams to metric tonnes

          In addition, the time in cruise is calculated as follows:

                                               Equation 2B-3
               Hrs I callcruise =   Cruise Distance [nmiles ] I Cruise Speed [knots ] x 2 trips I call

          Where:
          Cruise distance = one way distance (25 nautical miles)
          Cruise speed = vessel service speed, in knots
          2 trips/call = Used to calculate round trip cruise distance

              Main engine load factors are calculated directly from the propeller curve based upon the
       cube of actual speed divided by maximum speed (at 100% maximum continuous rating [MCR]):

                                               Equation 2B-4
                  LoadFactor cmise,. , = (Cruise  Speed [knots ] I Maximum Speed [knots ])

              Since cruise speed is estimated at 94 percent of maximum speed34, the load factor for main
       engines at cruise is 0.83.

              Substituting Equation  2B-3 for time in cruise into Equation 2B-2, and using the load factor
       of 0.83, the equation used to calculate cruise mode emissions for the main engines becomes the
       following:

                             Equation 2B-5 Cruise Mode Emissions for Main Engines
Emissionscrmse[mam] = (calls) x (P[mam]) x (CruiseDisiance/CruiseSpeed) x (2 trips/call) x 0.83 x (EF[mam]) x (10~6 tonnes/g)

          Where:
          EmissionScruise [main]= Metric tonnes emitted from main engines in cruise mode
          calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
          P[main] = Total main engine power, in kilowatts
          Cruise distance = one way distance (25 nautical miles)
          Cruise speed = vessel service speed, in knots
          2 trips/call = Used to calculate round trip cruise distance
          0.83 = Load factor for main engines in cruise mode, unitless
          EF [main] = Emission factor for main engines for the pollutant of interest, in g/kW-hr (these vary
              as a function of engine type and fuel used, rather than activity mode)
          10"6 = Conversion factor from grams to metric tonnes

              The equation used to calculate cruise mode emissions for the auxiliary engines is:
                                                 2-39

-------
                            Equation 2B-6 Cruise Mode Emissions for Auxiliary Engines
Emissions cmise[ai
-------
                                        Equation 2B-8
             Hrs I call ^  =  RSZ Distance [nmiles ] / RSZ Speed [knots ] X 2 trips I call

   Load factor during the RSZ mode is calculated as follows:

                                        Equation 2B-9
                     LoadFactorRSZ,ain, = (RSZ Speed I Maximum Speed')
In addition:
                                       Equation 2B-10
                            Maximum Speed = Cruise Speed 10.94

   Where:
   0.94 = Fraction of cruise speed to maximum speed

Substituting Equation 2B-10 into Equation 2B-9, the equation to calculate load factor becomes:

                                       Equation 2B-11
                    LoadFactorRSZ[mam] = (RSZ Speed X 0.94 / Cruise Speed)3

   Where:
   0.94 = Fraction of cruise speed to maximum speed

       Load factors below 2 percent were set to 2 percent as a minimum.

       Substituting Equation 2B-8 for time in mode and Equation 2B-11  for load factor into
Equation 2B-7 , the expression used to calculate RSZ mode emissions for the main engines
becomes:

                       Equation 2B-12 RSZ Mode Emissions for Main Engines
         Emissions RSZ[aux] =  (calls )x(P[aux^)x(RSZ Distance/ RSZ Speed )x (2 trips /call)
         x (RSZ Speed x 0.94 /Cruise Speed )3 x (EF[aux])x (Adj)x(10"6 tonnes Ig)

   Where:
   EmissionsRsz[main] = Metric tonnes emitted from main  engines in RSZ mode
   calls = Round-trip visits (i.e., one entrance and one clearance is  considered a call)
   P[main] = Total main engine power, in kilowatts
   RSZ distance = one way distance, in nautical miles (10 nm for all PR/USVI ports)
   RSZ speed = speed, in knots (10 knts for all PR/USVI ports)
   2 trips/call = Used to calculate round trip RSZ distance
   Cruise speed = vessel service speed, in knots
   EF[ma;n] = Emission factor for main engines for the pollutant of interest,  in g/kW-hr (these vary
       as a function of engine type and fuel used, rather than activity mode)
   Adj = Low load adjustment factor, unitless (used when the load  factor is below 0.20)
   10"6 = Conversion factor from grams to tons
   0.94 = Fraction of cruise speed to maximum speed
                                          2-41

-------
         Emission factors are considered to be relatively constant down to about 20 percent load.
  Below that threshold, emission factors tend to increase significantly as the load decreases. During
  the RSZ mode, load factors can fall below 20 percent.  Low load multiplicative adjustment factors
  were developed and applied when the load falls below 20 percent (0.20). If the load factor is 0.20
  or greater, the low load adjustment factor is set to 1.0.

         The equation used to calculate RSZ  mode emissions for the auxiliary engines is:

                        Equation 2B-13 RSZ Mode Emissions for Auxiliary Engines
EmissionsRSZ[awc] = (calls)x(P[awc])X(RSZ Distance/RSZ Speed)X(2 trips/call)X(LFBSZ[aa])X(EF[aa])X(10~6 tonnes Ig)

      Where:
      EmissionsRsz[aux] = Metric tonnes emitted from auxiliary engines in RSZ mode
      calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
      P[aux] = Total auxiliary engine power, in kilowatts
      RSZ distance = one way distance, in nautical miles (10 nm for all PR/USVI ports)
      RSZ speed = speed, in knots (10 knts for all PR/USVI ports)
      2 trips/call = Used to calculate round trip cruise distance
      LFRSZ [aux] = Load factor for auxiliary engines in RSZ mode, unitless (these vary by ship type
         and activity mode)
      EF[aux] = Emission factor for auxiliary engines for the pollutant of interest, in g/kW-hr (these
         vary as a function of engine type and fuel used, rather than activity mode)
      10"6 = Conversion factor from grams to  metric tonnes

         Unlike main engines, there is no need for a low load adjustment factor for auxiliary engines,
  because of the way they are generally operated.  When only low loads are needed, one or more
  engines are shut off, allowing the remaining engines to maintain operation at a more efficient level.

         The inputs of calls, RSZ distance, and RSZ speed are the same for main and auxiliary
  engines. Relative to the main engines, auxiliary engines have separate inputs for engine power,
  load factor, and emission factors.  The RSZ distances are assumed to be 10 nm for all PR/USVI
  ports. RSZ speed is assumed constant at 10 knots for all ships entering the harbor area.

  Maneuvering

         Maneuvering emissions are calculated for both propulsion (main) and auxiliary engines.
  The basic equation used to calculate maneuvering mode emissions for the main engines is:

                                          Equation 2B-14
       EmiSsionsmat{mam] = (ca/fa)x(^H])x(/^/ca/^^                               tonnes/ g)

      Where:
      Emissionsman[main] = Metric tonnes emitted from main engines in maneuvering mode
      calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
      P[main] = Total main engine power, in kilowatts
      hrs/callman = Hours per call for maneuvering mode
      LFman [mam] = Load factor for main engines in maneuvering mode, unitless
                                             2-42

-------
   EF[main] = Emission factor for main engines for the pollutant of interest, in g/kW-hr (these vary
       as a function of engine type and fuel used, rather than activity mode)
   Adj = Low load adjustment factor, unitless (used when the load factor is below 0.20)
   10"6 = Conversion factor from grams to metric tonnes

       Maneuvering time-in-mode is estimated based on the distance a ship travels from the
breakwater or port entrance to the pier/wharf/dock (PWD).  Maneuvering times also include shifts
from one PWD to another or from one port within a greater port area to another. Average
maneuvering speeds vary from 3 to 8 knots depending on direction and ship type. For consistency,
maneuvering speeds were assumed to be the dead slow setting of approximately 5.8 knots.

   Load factor during maneuvering is calculated as follows:

                                       Equation 2B-15
               LoadFactorman,ain, = (Man Speed[knots]/ Maximum Speed[knots])


In addition:
                                       Equation 2B-16
                          MaximumSpeed = Cruise Speed[knots] 70.94

   Where:
   0.94 = Fraction of cruise speed to maximum speed

Also, the maneuvering  speed is 5.8 knots.  Substituting Equation 2B-16 into Equation 2B-15, and
using a maneuvering speed of 5.8 knots, the equation to calculate load factor becomes:

                                       Equation 2B-17
                          LoadFactorman[mam} = (5.45 /Cruise Speed)3


       Load factors below 2 percent were set to 2 percent as a minimum.

       Substituting Equation 2B-17 for load factor into Equation 2B-14,  the expression used to
calculate maneuvering mode emissions for the main engines becomes:

                    Equation 2B-18 Maneuvering Mode Emissions for Main Engines
Emissions mm[main] = (calls) X (P[main]) X (hrs I callmm) X (5.45 / Cruise Speed)3 X (EF[main]) X (Adj) X (1(T6 tonnes I g)
   Where:
   Emissionsman[main] = Metric tonnes emitted from main engines in maneuvering mode
   calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
   P[main] = Total main engine power, in kilowatts
   hrs/callman = Hours per call for maneuvering mode
   Cruise speed = Vessel service speed, in knots
   EF[main] = Emission factor for main engines for the pollutant of interest, in g/kW-hr (these vary
       as a function of engine type and fuel used, rather than activity mode)
   Adj = Low load adjustment factor, unitless (used when the load factor is below 0.20)
                                          2-43

-------
    10"6 = Conversion factor from grams to metric tonnes

       Since the load factor during maneuvering usually falls below 20 percent, low load
adjustment factors are also applied accordingly. Maneuvering times are not readily available for all
12 ports. For this analysis, maneuvering times and load factors available for Tampa were used to
calculate maneuvering emissions for the PR/USVI ports.

       The equation used to calculate maneuvering mode emissions for the auxiliary engines is:

                  Equation 2B-19 Maneuvering Mode Emissions for Auxiliary Engines
   Emissions man[aux] = (calls) x (P[aux]) x (hrs I callman) x (LFman[aux]) x (EF[aa]) x (1 (T6 tonnes I g)

    Where:
    Emissionsman[aux] = Metric tonnes emitted from auxiliary engines in maneuvering mode
    calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
    P[aux] = Total auxiliary engine power, in kilowatts
    hrs/callman = Hours per call for maneuvering mode
    LFman [aux] = Load factor for auxiliary engines in maneuvering mode, unitless (these vary by ship
       type and activity mode)
    EF[aux] = Emission factor for auxiliary engines for the pollutant of interest, in g/kW-hr (these
       vary as a function of engine type and fuel used, rather than activity mode)
    10"6 = Conversion factor from grams to metric tonnes

       Low load adjustment factors are not applied for auxiliary  engines.

Hotelling

       Hotelling emissions are calculated for auxiliary engines only, as main engines are not
operational during this mode. The equation used to calculate hotelling mode emissions for the
auxiliary engines is:

                    Equation 2B-20 Hotelling Mode Emissions for Auxiliary Engines
   Emissionshotel[aux]  = (calls) x (P[aux]) x (hrs I callhotel) x (LFhotel[aux]) x (EF[aux]) x (1 (T6 tonnes I g)

    Where:
    Emissionshotei[aux] = Metric tonnes emitted from auxiliary engines in hotelling mode
    calls = Round-trip visits (i.e., one entrance and one clearance is considered a call)
    P[aux] = Total auxiliary engine power, in kilowatts
    hrs/callhotei = Hours per call for hotelling mode
    LFhotei [aux] = Load factor for auxiliary engines in hotelling mode, unitless (these vary by ship
       type and activity mode)
    EF[aux] = Emission factor for auxiliary engines for the pollutant of interest, in g/kW-hr (these
       vary as a function of engine type and fuel used, rather than activity mode)
    10"6 = Conversion factor from grams to metric tonnes

       Hotelling times are not readily available for the 12 ports.  For this analysis, hotelling times
available for Tampa were used to calculate hotelling emissions for the PR/USVI ports.
                                           2-44

-------
34  Starcrest Consulting Group (June 2004). Port-Wide Baseline Air Emissions Inventory, prepared for the Port of Los
Angeles
                                                 2-45

-------
Appendix 2C:  Emission Inputs to STEEM
       The STEEM waterway network model relies on a number of inputs to identify the
movements for each vessel, individual ship attributes, and related emission factor information.
Each of these databases is described separately below.

Shipping Movements

       The shipping activity and routes database provides information on vessel movements or
trips.  It is developed using port entrances and clearances information from the USAGE report for
the U.S. and the Lloyd's Maritime Intelligence Unit (LMIU) for Canada and Mexico.35 These
sources contain information for each vessel carrying foreign cargo at each major port or waterway
that, most importantly for this analysis, includes:

          Vessel name
          Last port of call (entrance record) or next port of call (clearance record)

       The database then establishes unique identification numbers for each ship, each port pair,
and each resulting trip.

Ship Attributes

       The ship attributes data set contains the important characteristics of each ship that are
necessary for the STEEM interport model to calculate the emissions associated with each trip. The
information in this data set is matched to each previously assigned ship identification number. The
following information comes from the USAGE entrances and clearances report for each ship
identification number:

          Ship type
          Gross registered tonnage (GRT)
          Net registered tonnage (NRT)

       The ship attributes data set contains the following information from Lloyd's Register-
Fairplay for each ship identification number.

          Main propulsion engine installed power (horsepower)
          Service speed (cruise speed)
          Ship size (length, wide, and draft)

       Sometimes data was lacking from the above references for ship speed.  In these instances,
the missing information was developed for each  of nine vessel types and the appropriate value was
applied to each individual ship of that type.  Specifically, the missing  ship speeds for each ship
category were obtained from the average speeds used in a Lloyd's Register study of the Baltic Sea
and from an Entec UK Limited study for the European Commission.36'37  The resulting vessel cruise
speeds for ships with missing data are shown in Table 2C-1.
                                         2-46

-------
                       Table 2C-1 Average Vessel Cruise Speed by Ship Type"
AVERAGE CRUISE
SHIP TYPE SPEED (knots)
Bulk Carrier
Container Ship
General Cargo
Passenger Ship
Refrigerated
Cargo
Roll On-Roll Off
Tanker
Fishing
Miscellaneous
14.1
19.9
12.3
22.4
16.4
16.9
13.2
11.7
12.7
                          a Used only when ship specific data were missing from the
                          commercial database references.

       The average speed during maneuvering is approximately 60 percent of a ship's cruise speed
based on using the propeller law described earlier and the engine load factor for maneuvering that is
presented later in this section.

       As with vessel cruise speed, main engine installed power was sometimes lacking in the
Lloyd's Register-Fairplay data set.  Here again, the missing information was developed for nine
different vessel types and the appropriate value was applied to each individual ship of that type
when the data were lacking. In this case, the missing main engine horsepower was estimated by
regressing the relationships between GRT and NRT, and between installed power and GRT for each
category. This operation is performed internally in the model and the result applied to each
individual ship, as appropriate.

       The ship  attributes database also contains information on the installed power of engines used
for auxiliary purposes.  However, this information is usually lacking in the Lloyds data set, so an
alternative technique was employed to estimate the required values.  In short, the STEEM model
uses a ratio of main engine horsepower to auxiliary engine horsepower that was determined for
eight different vessel types using information primarily from ICF International.38  (The ICF report
attributed these power values to a study for the Port of Los Angeles by Starcrest Consulting.34) The
auxiliary engine  power for each individual vessel of a given ship type is then estimated by
multiplying the appropriate main power to auxiliary power ratio and the main engine horsepower
rating for that individual ship.  The main  and auxiliary power values and the resulting auxiliary
engine to main engine ratios are shown in Table 2C-2.
                                          2-47

-------
                             Table 2C-2 Auxiliary Engine Power Ratios
VESSEL TYPE
Bulk Carrier
Container Ship
General Cargo
Passenger Ship
Refrigerated Cargo
Roll On-Roll Off
Tanker
Miscellaneous
AVERAGE
MAIN ENGINE
POWER (kW)
7,954
30,885
9,331
39,563
9,567
10,696 c
9,409
6,252
AVERAGE
AUXILIARY ENGINE
POWER (kW)
1,169
5,746
1,777
39,563 a
3,900 b
2,156C
1,985
1,680
AUXILIARY TO
MAIN ENGINE
POWER RATIO
0.147
0.186
0.190
1.000
0.136
0.202
0.211
0.269
              a The ICF reference reported a value of 11,000 for auxiliary engines used on passenger
               ,  38
          vessels.
              b The STEEM used auxiliary engine power as reported in the ARE methodology document.
              0 The STEEM purportedly used values for Roll On-Roll Off main and auxiliary engines that
              represent a trip weighted average of the Auto Carrier and Cruise Ship power values from the
              ICF reference.

       Finally, the ship attributes database provides information on the load factors for main
engines during cruise and maneuvering operation,  in addition to load factors for auxiliary marine
engines. Main engine load factors for cruise operation were taken from a study of international
shipping for all ship types, except passenger vessels.39 For this analysis, the STEEM model used a
propulsion engine load factor for passenger  ship engines at cruise speed of 55 percent of the total
installed power. This is based on engine manufacturer data contained in two global shipping
studies.39'40  During maneuvering, it was assumed that all main engines, including those for
passenger ships, operate at 20 percent of the installed power. This is consistent with a study done
by Entec UK for the European Commission. The main engine load factors at cruise speed by ship
type are shown in Table 2C-3.

       Auxiliary engine load factors, except for passenger ships, were obtained from the ICF
International study referenced above. These values are also shown in Table 2C-3. For cruise
mode, neither port nor interport portions of the inventory were adjusted for low load operation, as
the low load adjustments are only applied to propulsion engines with load factors below 20%.
            Table 2C-3 Main and Auxiliary Engine Load Factors at Cruise Speed by Ship Type
SHIP TYPE
Bulk Carrier
Container Ship
General Cargo
Passenger Ship
Refrigerated Cargo
Roll On-Roll Off
Tanker
Miscellaneous
AVERAGE MAIN
ENGINE LOAD
FACTOR (%)
75
80
80
55
80
80
75
70
AVERAGE
AUXILIARY ENGINE
LOAD FACTOR (%)
17
13
17
25
20
15
13
17
                                           2-48

-------
Emission Factor Information

       The emission factor data set contains emission rates for the various pollutants in terms of
grams of pollutant per kilowatt-hour (g/kW-hr).  The main engine emission factors are shown in
Table 2C-4. The speed specific factors for NOx, HC, and SC>2 were taken from several recent
analyses of ship emissions in the  U.S., Canada, and Europe.41'42'43'44  The PM factor was based on
discussions with the California Air Resources Board (ARB) staff. The fuel specific CO emission
factor was taken from a report by ENVIRON International.45 The STEEM study used the composite
emission factors shown in the table because the voyage data used in the model do not explicitly
identify main engine speed ratings, i.e., slow or medium, or the auxiliary engine fuel type, i.e.,
marine distillate or residual marine.  The composite factor for each pollutant is determined by
weighting individual emission factors by vessel engine population data from a 2005 survey of
ocean-going vessels that was performed by ARB.46
                   Table 2C-4 Main Engine Emission Factors by Ship and Fuel Type
ENGINE
TYPE FUE1
Slow Speed
Medium Speed
Composite EF
MAIN ENGINE EMISSION FACTORS (g/kW-hr)
TYPE
Residual
Marine
Residual
Marine
Residual
Marine
NOXP
18.1
14
17.9
M 10
1.5
1.5
1.5
PM25a
1.4
1.4
1.4
HCC
0.6
0.5
0.6
0
1.4
1.1
1.4
SO2
10.5
11.5
10.6
                      Estimated from PM10 using a multiplicative adjustment factor of 0.92.
       The emission factors for auxiliary engines are shown in Table 2C-5. The fuel specific main
emission factors for NOx and HC were taken from several recent analyses of ship emissions in the
U.S., Canada, and Europe, as  referenced above for the main engine load factors. The PM factor for
marine distillate was taken from a report by ENVIRON International, which was also referenced
above. The PM factor for residual marine was based on discussions with the California Air
Resources Board (ARB) staff. The CO factors are from the Starcrest Consulting study of the Port
of Los Angeles.34  For SO2, the fuel specific emission factors were obtained from Entec and
Corbett and Koehler:39 The composite emission factors displayed in the table are discussed below.
                  Table 2C-5 Auxiliary Engine Emission Factors by Ship and Fuel Type
ENGINE TYPE
Medium Speed
Medium Speed
Composite EF
AUXILIARY ENGINE EMISSION FACTORS (g/kW-hr)
FUEL TYPE
Marine
Distillate
Residual
Marine
Residual
Marine
NOXP
13.9
14.7
14.5
M 10
0.3
1.5
1.2
PM25a
0.3
1.4
1.1
HCC
0.4
0.4
0.4
0
1.1
1.1
1.1
SO2
4.3
12.3
**
                a Estimated from PM10 using a multiplicative adjustment factor of 0.92.
                b See Table 2C-6 for composite SO2 emission factors by vessel type.
                                          2-49

-------
       As for main engines, the STEEM study used the composite emission factors for auxiliary
engines.  For all pollutants other than SC>2, underlying data used in the model do not explicitly
identify auxiliary engine voyages by fuel type, i.e., marine distillate or residual marine.  Again, the
composite factor for those pollutants was determined by weighting individual emission factors by
vessel engine population data from a 2005 survey of ocean-going vessels that was performed by
ARE.
     47
       For SO2, composite emission factors for auxiliary engines were calculated for each vessel
type.  These composite factors were determined by taking the fuel specific emission factors from
Table 2C-5 and weighting them with an estimate of the amount of marine distillate and residual
marine that is used by these engines. The relative amount of each fuel type consumed was taken
from the 2005 ARB survey. The relative amounts of each fuel type for each vessel type and the
resulting 862 emission factors are shown in Table 2C-6.
              Table 2C-6 Auxiliary Engine SO2 Composite Emission Factors by Vessel Type
VESSEL TYPE
Bulk Carrier
Container Ship
General Cargo
Passenger Ship
Refrigerated Cargo
Roll On-Roll Off
Tanker
Miscellaneous
RESIDUAL
MARINE
(%)
71
71
71
92
71
71
71
0
MARINE
DISTILLATE
(%)
29
29
29
8
29
29
29
100
COMPOSITE
EMISSION FACTOR
(g/kW-hr)
9.98
9.98
9.98
11.66
9.98
9.98
9.98
4.3
Adjustments to STEEM PM and SO2 Emission Inventories

       The interport emission results contained in this study for PMio and 862 were taken from the
STEEM inventories and then adjusted to reflect the U.S. Government's recent review of available
engine test data and fuel sulfur levels for the near port analysis. In the near ports work, a PM
emission factor of 1.4 g/kW-hr was used for most main engines, e.g., slow speed diesel and medium
speed diesel engines, all of which are assumed to use residual marine. A slightly higher value was
used for steam turbine and gas turbine engines, and a slightly lower value was used for most
auxiliary engines. However, these engines represent only a small fraction of the total  emissions
inventory. The STEEM study used an emission factor of 1.5 g/kW-hr for all main engines and a
slightly lower value for auxiliary engines.  Here again, the auxiliary engines comprise only a small
fraction of the total emissions from these ships. Therefore, for simplicity, the interport PM
inventories were adjusted by multiplying the STEEM results by the ratio of the two primary
emission factors, i.e., 1.4/1.5 or 0.933, to approximate the difference in fuel effects.
                                         2-50

-------
35 Corbett, J. et al. (May 2006). Estimation, Validation and Forecasts of Regional Commercial Marine Vessel
Inventories, Tasks 1 and 2:  Baseline Inventory and Ports Comparison, Final Report, prepared by University of
Delaware for the California Air Resource Board, Contract Number 04-346, and the Commission for Environmental
Cooperation in North America, Contract Number 113.111, May 2006, Docket ID EPA-HQ-OAR-2007-0121-0013.

36 Lloyd's Register and International Maritime Organization, Marine Exhaust Emission Quantification Study - Baltic
Sea, inMEPC 45/INF.7. 1998.

37 Entec UK Limited (2002). Quantification of Emissions from Ships Associated with Ship Movements between Ports in
the European Community, prepared for the European Commission, Docket ID EPA-HQ-OAR-2007-0121-0059.

38ICF International (January 5, 2006). Current Methodologies and Best Practices in Preparing Port Emission
Inventories, Final Report, prepared for the U.S. Environmental Protection Agency, available online at
http://www.epa.gov/sectors/port^p_portemissionsfmal.pdf.

39 Corbett, J.J. and H.W. Koehler (2003). Updated Emissions from Ocean Shipping, Journal of Geophysical Research,
108(020); p. 4650.

40 Corbett, J.J. and H.W. Koehler (2004). Considering Alternative Input Parameters in an Activity-Based Ship Fuel
Consumption and Emissions Model: Reply to Comment by Oyvind Endresen et al. on "Updated Emissions from Ocean
Shipping," Journal of Geophysical Research. 109(D23303).

41 Levelton Consultants  Ltd. (2006). Marine Emission Inventory Study Eastern Canada and Great Lakes - Interim
Report 4: Gridding Results, prepared for Transportation Development Centre, Transport Canada.

42 California Air Resources  Board (October 2005). Emissions Estimation Methodology for Ocean-Going Vessels.

43 ICF International (January 5, 2006). Current Methodologies and Best Practices in Preparing Port Emission
Inventories, Final Report, prepared for the U.S. Environmental Protection Agency, available online at
http://www.epa.gov/sectors/port^p-portemissionsfinal.pdf.

44 Corbett, J.J. and H.W. Koehler (2003).  Updated Emissions from Ocean Shipping, Journal of Geophysical Research,
108 (D20); p. 4650.

45 ENVIRON International Corporation (2002). Commercial Marine Emission Inventory Development, prepared for the
U.S. Environmental Protection Agency, EPA Report Number: EPA420-R-02-019.

46 California Air Resources  Board (September 2005). 2005 Oceangoing Ship Survey, Summary of Results.

47 California Air Resources  Board (September 2005). 2005 Oceangoing Ship Survey, Summary of Results.
                                                2-51

-------
Appendix 2D:  Growth Factor Development

     This appendix describes the development of growth factors for various U.S. Regions that were
used as the basis for the PR/US VI growth rate.  The derivation of the PR/US VI growth rate is
described in section 2.5.1.

Geographic Regions

     The geographic area reflects ship operations that occur within 200 nautical miles (nm) from
the official U.S. baseline but excludes operations in Exclusive Economic Zones of other countries.
The official U.S. baseline is recognized as the low-water line along the coast as marked on the
official U.S. nautical charts in accordance with the articles of the Law  of the Sea.  The boundary
was mapped using geographic information system (GIS) shapefiles obtained from the National
Oceanic and Atmospheric Administration, Office of Coast Survey.48 The accuracy of the NOAA
shapefiles was verified with images obtained from the U.S. Geological Survey. The confirmed
NOAA shapefiles were then combined with a shapefile of the U.S. international border from the
National Atlas.49

       The resulting region was further subdivided for this analysis to create regions that were
compatible with the geographic scope of the regional growth rates, which are used to project
emission inventories for the year 2020.

       The Pacific Coast region was split into separate North Pacific and South Pacific regions
       along a horizontal line originating from the Washington/Oregon border (Latitude 46°  15'
       North).

       The East Coast and  Gulf of Mexico regions were divided along a vertical line roughly drawn
       through Key Largo (Longitude 80° 26' West).

       The Alaska region was divided into separate Alaska Southeast  and Alaska West regions
       along a straight line intersecting the cities of Naknek and Kodiak.  The Alaska Southeast
       region includes most of the State's population, and the Alaska West region includes the
       emissions from ships on a great circle route along the Aleutian Islands between Asia and the
       U.S. West Coast.

       For the Great Lakes domain, shapefiles were created containing all the ports and inland
       waterways in the near port inventory and extending out into the lakes to the international
       border with Canada. The modeling domain spanned from Lake Superior on the west to  the
       point eastward in the State of New York where the St. Lawrence River parts from U.S. soil.

       The Hawaiian domain was subdivided so that a distance of 200 nm beyond the southeastern
       islands of Hawai'i, Maui, O'ahu, Moloka'i, Ni'ihau, Kaua'i, Lanai, and Kahoolawe was
       contained in Hawaii East.  The remainder of the Hawaiian Region was  then designated
       Hawaii West.

       This methodology resulted in nine separate regional modeling domains that are identified
below and shown in Figure 2D-1. U.S. territories are not included in this analysis.

          South Pacific (SP)
          North Pacific (NP)


                                         2-52

-------
          East Coast (EC)
          Gulf Coast (GC)
          Alaska Southeast (AE)
          Alaska West (AW)
          Hawaii East (HE)
          Hawaii West (HW)
          Great Lakes (GL)
                             Figure 2D-1 Regional Modeling Domains

Growth Factors by Geographic Region

       The growth factors that are used to estimate future year emission inventories are based on
the expected demand for marine bunker fuels that is associated with shipping goods, i.e.,
commodities, into and out of the U.S.  This section describes the growth factors that are used to
project the emissions to 2020 for each of the nine geographic regions evaluated in this analysis.
The use of bunker fuel as a surrogate for estimating future emissions is appropriate because the
quantity of fuel consumed by C3 engines is highly correlated with the amount of combustion
products, i.e., pollutants that are emitted from those vessels. The term bunker fuel in this report also
includes marine distillate oil and marine gas oil that are used in some auxiliary power engines.

       The remainder of this section first summarizes the development of growth rates by RTI
International (RTI) for five geographic regions of the U.S., as performed under contract to the U.S.
government.50'51  This is followed by the derivation of the growth factors for the nine geographic
regions of interest.

Summary of Regional Growth Rate Development

       RTI developed fuel consumption growth rates for five geographic regions of the U.S. These
regions are the East Coast, Gulf Coast, North Pacific, South Pacific, and Great Lakes. The amount
of bunker fuel required in any region and year is based on the demand for transporting various types
of cargo by Category 3 vessels.  This transportation demand is in turn driven by the demand for
commodities that are produced in one location and consumed  in another, as predicted by an
econometric model. The flow of commodities is matched with typical vessels for trade routes
                                         2-53

-------
(characterized according to cargo capacity, engine horsepower, age, specific fuel consumption, and
engine load factors). Typical voyage parameters are then assigned to the trade routes that include
average ship speed, round trip mileage, tons 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 2D-2 illustrates the approach to developing
baseline projections of marine fuel consumption.

       As a means of comparison, the IMO Secretary General's Informal Cross
Government/Industry Scientific Group of Experts presented a growth rate that ranged from 3.3% to
3.7%.52 RTFs overall U.S. growth rate was projected at 3.4%, which is consistent with that range.
Ship Analysis: by Vessel Type and Size Category

Inputs Outputs
Deadweight for all Vessels of
Given Type & Size8

Horsepower, Year of Build
for all Vessels of Given
Type & Size3

Specific Fuel Consumption
(g/SHP-HR) by Year of Build"

Engine Load Factors0

1 Average Cargo /T~\
1 Carried (Tons) V_y

Average Daily Fuel
* Consumption
(Tons/Day)

Average Daily Fuel
Consumption (Tons/Day) J^rT}
- Main, Aux. Engine at Sea **V_y
-Aux. Engine in Port

Trade Analysis: by Commodity and Trade Route

Inputs
Average Ship Speed0

Round Trip Mileaged

Tons of Cargo Shipped6

Average Cargo Carried/^~^N
per Ship Voyage \/V/

Outputs
Days at Sea and in
, Port, per Voyage

Total Days at f^\
J 	 'V _ _ 	 H L»
-. Sea and in Port V_y

» Number of Voyages

Total Estimated Bunker Fuel Demand


/* N
Average Daily Fuel Consumption
(Tons/Day) Total Days at Sea Bunker Fuel
- Main, Aux. Engine at Sea f^\ and in Port f^r^\ Demand
- Aux. Engine in Port v 	 / ^_s
Driven by changes in engine efficiency. Dr!ven b™r™th in
commodity flows.
             a - Clarksons Ship Register Database
             b - Engine Manufacturers' Data, Technical Papers
             c - Corbett and Wang (2005) "Emission Inventory Revi
             d - Combined trade routes and heavy leg analysis
             e - Global Insight Inc. (Gil) Trade Flow Projections
i: SECA Inventory Progress Discussion"
                  Figure 2D-2 Illustration of Method for Estimating Bunker Fuel Demand
                                             2-54

-------
Trade Analysis
       The trade flows between geographic regions of the world, as illustrated by the middle
portion of Figure 2D-2 were defined for the following eight general types of commodities:

       -   liquid bulk - crude oil

       -   liquid bulk - refined petroleum products

       -   liquid bulk - residual petroleum products

       -   liquid bulk - chemicals (organic and inorganic)

       -   liquid bulk -gas (including LNG and LPG)

       -   dry bulk (e.g., grain, coal, steel, ores and scrap)

       -   general cargo (e.g., lumber/forest products)

       -   containerized cargo

       The analysis specifically evaluated trade flows between 21 regions of the world. Table 2D-1
shows the countries associated with each region.

                       Table 2D-1 Aggregate Regions and Associated Countries
AGGREGATE
REGIONS
U.S. Atlantic Coast
U.S. Great Lakes
U.S. Gulf Coast
E. Canada3
W. Canada3
U.S. Pacific North
U.S. Pacific South
Greater Caribbean
South America
Africa - West
Africa-North/East-
Mediterranean
Africa-East/South
Europe-North
Europe-South
Europe-East
Caspian Region
Russia/FSU
Middle East Gulf
Australia/NZ
BASE COUNTRIES / REGIONS
U.S. Atlantic Coast
U.S. Great Lakes
U.S. Gulf Coast
Canada3
Canada3
U.S. Pacific North
U.S. Pacific South
Colombia, Mexico, Venezuela, Caribbean Basin, Central America
Argentina, Brazil, Chile, Peru, Other East Coast of S. America, Other
West Coast of S. America
Western Africa
Mediterranean Northern Africa, Egypt, Israel,
Kenya, Other Eastern Africa, South Africa, Other Southern Africa
Austria, Belgium, Denmark, Finland, France, Germany, Ireland,
Netherlands, Norway, Sweden, Switzerland, United Kingdom
Greece, Italy, Portugal, Spain, Turkey, Other Europe
Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovak Republic
Southeast CIS
The Baltic States, Russia Federation, Other Western CIS
Jordan, Saudi Arabia, UAE, Other Persian Gulf
Australia, New Zealand
                                          2-55

-------
AGGREGATE
REGIONS
Japan
Pacific-High Growth
China
Rest of Asia
BASE COUNTRIES / REGIONS
Japan
Hong Kong S.A.R., Indonesia, Malaysia, Philippines, Sinj
Korea, Taiwan, Thailand
japore, South
China
Viet Nam, India, Pakistan, Other Indian Subcontinent
          a Canada is treated as a single destination in the GI model. Shares of Canadian imports from
          and exports to regions of the world in 2004 are used to divide Canada trade into shipments
          to/from Eastern Canada ports and shipments to/from Western Canada ports.53

       The overall forecast of demand for shipping services and bunker fuel was determined for
each of the areas using information on commodity flows from Global Insight's (GI) World Trade
Service.  Specifically, GI provided a specialized forecast that reports the flow of each commodity
type for the period 1995-2024, based on a proprietary econometric model.  The general structure of
the GI model for calculating trade flows assumes  a country's imports from another country are
driven by the importing country's demand forces  (given that the exporting  country possesses
enough supply capacity), and affected by exporting the country's export price and importing
country's import cost for the commodity. The model then estimates demand forces, country-specific
exporting capacities, export prices, and import costs.

       The GI model included detailed annual region-to-region trade flows for eight composite
commodities from 1995 to 2024, in addition to the total trade represented by the commodities.
Table 2D-2 illustrates the projections for 2012 and 2020, along with baseline data for 2005.  In
2005, dry bulk accounted for 41 percent of the total trade volume, crude oil accounted for 28
percent, and containers accounted for 12 percent.  Dry bulk and crude oil shipments are expected to
grow more slowly over the forecast period than container shipments. By 2020, dry bulk represents
39 percent of the total, crude oil is 26 percent, and containers rise to 17 percent.

     Table 2D-2 Illustration of World Trade Estimates for Composite Commodities, 2005,2012, and 2020
COMMODITY TYPE
Dry Bulk
Crude Oil
Container
Refined Petroleum
General Cargo
Residual Petroleum and Other Liquids
Chemicals
Natural Gas
Total International Cargo Demand
CARGO (millions of tons)
20052
2,473
1,703
714
416
281
190
122
79
5,979
0122
3,051
2,011
1,048
471
363
213
175
91
7,426
020
3,453
2,243
1,517
510
452
223
228
105
8,737
Ship Analysis by Vessel Type and Size

       Different types of vessels are required to transport the different commodities to the various
regions of the world. Profiles of these ships were developed to identify the various vessel types and
size categories that are assigned to transport commodities of each type along each route.  These
                                          2-56

-------
profiles include attributes such as ship size, engine horsepower, engine load factors, age, and engine
fuel efficiency. This information was subsequently used to estimate average daily fuel consumption
for each typical ship type and size category.

       The eight GI commodity  categories were mapped to the type of vessel that would be used to
transport that type of cargo using information from  Clarkson's Shipping Database.54 These
assignments are shown in Table 2D-3.

                       Table 2D-3 Assignment of Commodities to Vessel Types
COMMODITY SHIP CATEGORY VESSEL TYPE
Liquid bulk - crude oil
Liquid bulk - refined
petroleum products
Liquid bulk - residual
petroleum products
Liquid bulk - chemicals
(organic and inorganic)
Liquid bulk - natural gas
(including LNG and LPG)
Dry bulk (e.g. grain, coal,
steel, ores and scrap)
General cargo (including
neobulk, lumber/forest
products)
Containerizable cargo
Crude Oil Tankers
Product Tankers
Product Tankers
Chemical Tankers
Gas Carriers
Dry Bulk Carriers
General Cargo
Container Ships
Tanker
Product Carrier
Product Carrier
Chemical & Oil Carrier
LNG Carrier, LPG Carrier, Chemical & LPG Carrier,
Ethylene/LPG, Ethylene/LPG/Chemical,
LNG/Ethylene/LPG, LNG/Regasification, LPG/Chemical,
LPG/Oil, Oil & Liquid Gas Carrier
Bulk Carrier
General Cargo Liner, Reefer, General Cargo Tramp, Reefer
Fish Carrier, Ro-Ro, Reefer/Container, Ro-Ro
Freight/Passenger, Reefer/Fleet Replen., Ro-Ro/Container,
Reefer/General Cargo, Ro-Ro/Lo-Lo, Reefer/Pallets
Carrier, Reefer/Pass./Ro-Ro, Reefer/Ro-Ro Cargo
Fully Cellular Container
       Each of the vessel types were classified by their cargo carrying capacity or deadweight tons
(DWT). The size categories were identified based on both industry definitions and natural size
breaks within the data.  Table 2D-4 summarizes the size categories that were used in the analysis
and provides other information on the general attributes of the vessels from Clarkson's Shipping
Database.  The vessel size descriptions are also used to define shipping routes based on physical
limitations that are represented by canals or straits through which ships can pass.  Very large crude
oil tankers are the largest by DWT rating, and the biggest container ships (Suezmax) are also very
large.
                                          2-57

-------
                                Table 2D-4 Fleet Characteristics
SHIP
TYPE
Container
General
Cargo
Dry Bulk
Crude Oil
Tanker
Chemical
Tanker
Petroleum
Product
Tanker
Natural
Gas
Carrier
Other
Total
SIZE BY
DWT
Suezmax
PostPanamax
Panamax
Intermediate
Feeder
All
Capesize
Panamax
Handymax
Handy
VLCC
Suezmax
AFPvAmax
Panamax
Handymax
Coastal
All
AFPvAmax
Panamax
Handy
Coastal
VLGC
LGC
Midsize
All
-
MINIMUM
SIZE
(DWT)
83,000
56,500
42,100
14,000
0
MAXIMUM
SIZE
(DWT)
140,000
83,000
56,500
42,100
14,000
All
79,000
54,000
40,000
0
180,000
120,000
75,000
43,000
27,000
0
0
79,000
54,000
40,000
0
180,000
120,000
75,000
43,000
27,000
All
68,000
40,000
27,000
0
60,000
35,000
0
0
68,000
40,000
27,000
0
60,000
35,000
All
-
-
NUMBER
OF SHIPS
101
465
375
1,507
1,100
3,214
715
1,287
991
2,155
470
268
511
164
100
377
2,391
226
352
236
349
157
140
863
7,675
26,189
TOTAL
DWT
(millions)
9.83
30.96
18.04
39.8
8.84
26.65
114.22
90.17
46.5
58.09
136.75
40.63
51.83
10.32
3.45
3.85
38.8
19.94
16.92
7.9
3.15
11.57
6.88
4.79
88.51
888.4
TOTAL
HORSE-
POWER
(millions)
8.56
29.3
15.04
32.38
7.91
27.07
13.81
16.71
10.69
19.58
15.29
5.82
8.58
2.17
1.13
1.98
15.54
3.6
4.19
2.56
1.54
5.63
2.55
3.74
53.6
308.96
TOTAL
KILO-
WATTS
(millions)
6.38
21.85
11.21
24.14
5.90
20.18
10.30
12.46
7.97
14.60
11.40
4.34
6.40
1.62
0.84
1.48
11.59
2.68
3.12
1.91
1.15
4.20
1.90
2.79
39.96
230.36
       The average fuel consumption for each vessel type and size category was estimated in a
multi-step process using individual vessel data on engine characteristics.  Clarkson's Shipping
Database Register provides each ship's total installed horsepower (HP), type of propulsion (diesel
or steam), and year of build.  These characteristics are then matched to information on typical
specific fuel consumption (SFC), which is expressed in terms of grams of bunker fuel burned per
horsepower-hour (g/HP-hr, which is equivalent to 1.341 g/kW-hr).

       The SFC values are based on historical data from Wartsila Sulzer, a popular manufacturer of
diesel engines for marine vessels. RTI added an additional 10 percent to the reported "test bed" or
"catalogue" numbers to account for the guaranteed tolerance level and an in-service SFC
                                          2-58

-------
differential.  Overall, the 10 percent estimate is consistent with other analyses that show some
variation between the "test bed" SFC values reported in the manufacturer product catalogues and
those observed in actual service.  This difference is explained by the fact that old, used engines
consume more fuel than brand new engines and in-service fuels may be different than the test bed
fuels.55

       Figure 2D-3 shows SFC values that were used in the model regarding the evolution of
specific fuel oil consumption rates for diesel engines over time. Engine efficiency in terms of SFC
has improved over time, most noticeably in the early 1980s in response to rising fuel prices.
However, there is a tradeoff between improving fuel efficiency and reducing emissions.
Conversations with engine manufacturers indicate that it is reasonable to assume SFC will remain
constant for the projection period of this study, particularly as they focus on meeting NOx emission
standard as required by MARPOL Annex VI, or other potential pollution control requirements.
Post-2000 SFC values are constant at approximately 135 g/hp-hr (180 g/kW-hr).
          200 -,
          180
          120
          100
           80
           60
           40
           20
            1950  1955  1960  1965  1970  1975 1980 1985  1990  1995  2000  2005  2010 2015 2020


                           Figure 2D-3 Diesel Engine Specific Fuel Consumption
       RTI assumed a fixed SFC of 220 g/FtP-hr (295 g/kW-hr) for steam engines operating on
bunker fuel.

       Using the above information, the average daily fuel consumption (AFC), expressed in metric
tons of fuel at full engine load, for each vessel type and size category is found using the following
equation:

                                        Equation 2D-1


                          Fleet AFCV>, =— ^[SFCVS xHPvs xlO'6 tonnes I g]
                                          2-59

-------
       Where:
             Fleet AFC = Average daily fuel consumption in metric tonnes at full engine load
             v = Vessel type
          -  s = Vessel size category
             N = Number of vessels in the fleet
             SFC = Specific fuel consumption in grams of bunker fuel burned per horsepower-
             hour in use(g/FIP-hr)
             FTP = Total installed engine power, in horsepower (FTP)
             106 tonnes/g = Conversion from grams to metric tonnes

       As previously noted, AFC values calculated in the above equation are based on total
horsepower; therefore, they must be scaled down to reflect typical operation using less than 100
percent of the horsepower rating, i.e., actual engine load.  Table 2D-5 shows the engine load factors
that were used to estimate the typical average daily fuel consumption (tons/day) for the main
propulsion engine and the auxiliary engines when operated at sea and in port.56

                        Table 2D-5 Main and Auxiliary Engine Load Factors




VESSEL TYPE
Container Vessels
General Cargo Carriers
Dry Bulk Carriers
Crude Oil Tankers
Chemical Tankers
Petroleum Product Tankers
Natural Gas Carrier
Other
MAIN
ENGINE
LOAD
FACTOR
(%)
80
80
75
75
75
75
75
70


AUXILIARY
ENGINE AS % OF
MAIN ENGINE
22.0
19.1
22.2
21.1
21.1
21.1
21.1
20.0

AUXILIARY
ENGINE AS % OF
MAIN ENGINE
AT SEA
11.0
9.5
11.1
10.6
10.6
10.6
10.6
10.0
       The RTI analysis also assumed that the shipping fleet changes over time as older vessels are
scrapped and replaced with newer ships. Specifically, vessels over 25 years of age are retired and
replaced by new ships of the most up-to-date configuration.  This assumption leads to the following
change in fleet characteristics over the projection period:

          New ships have engines rated at the current SFC, so even though there are no further
          improvements in specific fuel consumption, the fuel efficiency of the fleet as a whole
          will improve over time through retirement and replacement.
          New ships will weigh as much as the average ship built in 2005, so the total cargo
          capacity of the fleet will increase over time as smaller ships retire and are replaced.
          Container ships will increase in size over time on the trade routes between Asia to either
          North America or Europe.

Trade Analysis by Commodity Type and Trade Route

       Determining the total number of days at sea and in port requires information on the relative
amount of each commodity that is carried by the different ship type size categories on each of the
                                         2-60

-------
trade routes. For example, to serve the large crude oil trade from the Middle East Gulf region to the
Gulf Coast of the U.S., 98 percent of the deadweight tonnage is carried on very large oil tankers,
while the remaining 2 percent is carried on smaller Suezmax vessels. After the vessel type size
distribution was found, voyage parameters were estimated. Specifically, these are days at sea and in
port for each voyage (based on ports called, distance between ports, and ship speed), and the
number of voyages (based on cargo volume projected by GI and the DTW from Clarkson's
Shipping Database).  The length of each voyage and number of voyages were used to estimate the
total number of days at sea and at port, which is a parameter used later to calculate total fuel
consumption for each vessel type and  size category over each route and for each commodity type.
(More information on determining the round trip distance for each voyage that is associated with
cargo demand for the U.S. is provided in the next section.)
speed:
       The days at sea were calculated by dividing the round trip distance by the average
vessel
                   Days at Sea Per Voyagev _
                                       Equation 2D-2

                                                round trip distance route
                                          Joute
                                                    speedv s x 24 hrs
       Where:
             v = Vessel type
             s = Vessel size category
             route = Unique trip itinerary
             round trip route distance = Trip length in nautical miles
             speed = Vessel speed in knots or nautical miles per hour
             24 hrs = Number of hours in one day
                                                                            56
       Table 2D-6 presents the speeds by vessel type that were used in the analysis.   These values
are the same for all size categories, and are assumed to remain constant over the forecast period.

                                Table 2D-6 Vessel Speed by Type
VESSEL TYPE
Grade Oil Tankers
Petroleum Product Tankers
Chemical Tankers
Natural Gas Carriers
Dry Bulk Carriers
General Cargo Vessels
Container Vessels
Other
SPEED (knots)
13.2
13.2
13.2
13.2
14.1
12.3
19.9
12.7
       The number of voyages along each route for each trade was estimated for each vessel type v
and size category s serving a given route by dividing the tons of cargo moved by the amount of
cargo (DTW) per voyage:
                                         2-61

-------
                                       Equation 2D-3
        ,     „ T.               total metric tonnes of cargo moved
    Number of Voyagesv s trade =	
                              fleet average DWTv s x utilization rate

       Where:
              v = Vessel type
              s = Vessel size category
              trade = Commodity type
              Fleet average DWT = Median dead weight tonnage carrying capacity in metric tons
              Utilization rate = Fraction of total ship DWT capacity used

       The cargo per voyage is based on the fleet average ship size from the vessel profile analysis.
For most cargo, a utilization rate of 0.9 is assumed to be constant throughout the forecast period.
Lowering this factor would increase the estimated number of voyages  required to move the
forecasted cargo volumes, which would lead to an increase in estimated fuel demand.

       In addition to calculating the average days at sea per voyage, the average days in port per
voyage was also estimated by assuming that most types of cargo vessels spend four days in port per
voyage. RTI notes, however, that this can vary somewhat by commodity  and port.

Worldwide Estimates of Fuel Demand

       This section describes how the information from the vessel and trade analyses were used to
calculate the total annual fuel demand associated with international cargo trade.  Specifically, for
each year^ of the analysis, the total bunker fuel demand is the sum of the fuel consumed on each
route of each trade (commodity). The fuel consumed on each route of each trade is in turn the sum
of the fuel consumed for each route and trade for that year by propulsion main engines and auxiliary
engines when operated at sea and in port. These steps are illustrated by the following equations:
                                       Equation 2D-4
           —  y  y
           —  ^j  ^j
             trade route


             trade route
Days at Seatrade)route)y + AFCtade)route!yatport x Days at Porttade route y ]
       Where:
             FC = Fuel consumed in metric tonnes
             y = calendar year
             trade = Commodity type
             route = Unique trip itinerary
             AFC = Average daily fuel consumption in metric tonnes
             yatsea = Calendar year main and auxiliary engines are operated at sea
             yatport = Calendar year main and auxiliary engines are operated in port
                                         2-62

-------
                                        Equations 2D-5

            e.route.yatsea = E (Percent of trade along route)v s [Fleet AFCV s x (MELF+AE at sea LF)]

            e.route.yatport = v £ r (Percent of trade along route)v s [Fleet AFCV s x AE import LFJ

       Days at Seatad   t   = £ (Percent of trade along route)  [Days at sea per voyage  x Number of voyages  1
                        v,s,t,r                        ' L                  '                   ' J
       Days at Porttmde route y = £ (Percentof trade along route)vs [Days at port per voyage xNumberof voyages]
       Where:
              AFC = Average daily fuel consumption in metric tones
              trade = Commodity type
              route = Unique trip itinerary
           -   yatsea = Calendar year main and auxiliary engines are operated at sea
           -   yatport =  Calendar year main and auxiliary engines are operated in port
           -   y = calendar year
           -   v = Vessel type
           -   s = Vessel size category
              t = Trade
              r = Route
           -   Fleet APC = Average daily fuel consumption in metric tonnes at full engine load
           -   MELF = main engine load factor, unitless
              AE at sea  LF = auxiliary engine at-sea load factor, unitless
              AE in port LF = auxiliary engine in-port load factor, unitless

       The inputs for these last four equations are all derived from the vessel analysis and the trade
analysis previously described.

Worldwide Bunker Fuel Consumption

       Based on the methodology outlined above, estimates of global fuel consumption over time
were computed, and growth rates determined from these projections.
                                           2-63

-------
                     H Container     ffi General Cargo  D Dry Bulk     H Crude Oil
                     D Chemicals     D Petroleum     • Natural Gas   D Other
                     • Fishing Vessels  • Passenger Ships D Military Vessels
                             Figure 2D-4 Worldwide Bunker Fuel Consumption
       Figure 2D-4 shows estimated world-wide bunker fuel consumption by vessel type. Figure
2D-5 shows the annual growth rates by vessel-type/cargo that are used in the projections shown in
Figure 2D-4. Total annual growth is generally between 2.5 percent and 3.5 percent over the time
period between 2006 and 2020 and generally declines over time, resulting in an average annual
growth of around 2.6 percent.
                                           2-64

-------
            10%
            8%
_e

-------
       Voyage distances for container vessels are based on information from Containerization
International Yearbook (CIY)58 and calculations by RTI. That reference provides voyage
information for all major container services. Based on the frequency of the service, number of
vessels assigned to that service, and the number of days in operation per year, RTI estimated the
average length of voyages for the particular bilateral trade routes in the Global Insights trade
forecasts.

       The distance information developed above was combined with the vessel speeds previously
shown in Table 2D-6 to find the length of a voyage in days. Table 2D-7 presents the day lengths for
non-containerized vessel types and Table 2D-8 shows the same information for container vessels.

         Table 2D-7 Day Length for Voyages for Non-Container Cargo Ship (approximate average)
GLOBAL INSIGHTS TRADE
REGIONS
Africa East-South
Africa North-Mediterranean
Africa West
Australia-New Zealand
Canada East
Canada West
Caspian Region
China
Europe Eastern
Europe Western-North
Europe Western-South
Greater Caribbean
Japan
Middle East Gulf
Pacific High Growth
Rest of Asia
Russia-FSU
Rest of South America
DAYS PER VOYAGE
US South
Pacific
68
49
56
48
37
11
95
41
61
53
54
26
35
77
52
68
64
51
US North
Pacific
75
56
63
47
46
5
89
36
68
60
61
33
31
72
48
64
71
30
US East
Coast
57
37
36
65
7
40
41
73
38
24
30
16
65
56
67
66
38
41
US Great
Lakes
62
43
46
81
18
58
46
87
45
32
37
29
81
65
76
64
46
46
US Gulf
54
47
43
63
19
39
48
69
46
34
37
17
62
83
88
73
48
44
                  Table 2D-8 Day Length for Voyages for Container-Ship Trade Routes
ORIGIN - DESTINATION REGIONS
Asia - North America (Pacific)
Europe - North America (Atlantic)
Mediterranean - North America
Australia/New Zealand - North America
South America - North America
Africa South - North America (Atlantic)
Africa West - North America (Atlantic)
DAYS PER
VOYAGE
37
37
41
61
48
54
43
                                          2-66

-------
ORIGIN - DESTINATION REGIONS
Asia - North America (Atlantic)
Europe - North America (Pacific)
Africa South - North America (Pacific)
Africa West - North America (Pacific)
Caspian Region - North America (Atlantic)
Caspian Region - North America (Pacific)
Middle East/Gulf Region - North America (Atlantic)
Middle East/Gulf Region - North America (Pacific)
DAYS PER
VOYAGE
68
64
68
38
42
38
63
80
Bunker Fuel Consumption for the United States

       Figure 2D-6 and Figure 2D-7 present the estimates of fuel use for delivering trade goods to
and from the U.S. The results in Figure 2D-6 show estimated historical bunker fuel use in year
2001 of around 47 million tonnes (note: while this fuel is used to carry trade goods to and from the
U.S., it is not necessarily all purchased in the U.S. 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.
                 5 US North Pacific G US Great Lakes D US Gulf S US East Coast 0 US South Pacific
            Figure 2D-6 Bunker Fuel Used to Import and Export Cargo by Region of the United States

       Figure 2D-7 shows the estimated annual growth rates for the fuel consumption that are used
in the projections shown in Figure 2D-6. Overall, the average annual growth rate in marine bunkers
associated with future U.S. trade flows is 3.4 percent between  2005 and 2020.
                                          2-67

-------
           10%
            8%
            6%
          e
          e
          a
                            • United States
                            • US Great Lakes
• US South Pacific

• US Gulf
• US North Pacific

• US East Coast
       Figure 2D-7 Annual Growth Rates for Bunker Fuel Used to Import and Export Cargo by Region of the
                                        United States
2020 Growth Factors for Nine Geographic Regions

       The results of the RTI analysis described above are used to develop the growth factors that
are necessary to project the 2002 base year emissions inventory to 2020. The next two sections
describe how the five RTI regions were associated with the nine regions analyzed in this report, and
how the specific growth rates for each of the nine regions were developed.

        Mapping the RTI Regional Results to the Nine Region Analysis

       The nine geographic regions analyzed in this study  were designed to be consistent with the
five RTI regional  modeling domains.  More specifically, four of the nine geographic areas in this
study, i.e., Alaska East, Alaska West, Hawaii East, and Hawaii West are actually subsets of two
broader regional areas that were analyzed by RTI, i.e., the North Pacific for both Alaska regions and
South Pacific for Hawaii. Therefore, the growth rate information from the related larger region was
assumed to be representative for that state.

       The nine geographic regions represented in the emission inventory study are presented in
Figure 2-1. The association of the RTI regions to the emission inventory regions is shown in Table
2D-9.
                                          2-68

-------
             Table 2D-9 Association of the RTI Regions to the Nine Emission Inventory Regions
CONSUMPTION REGION
North Pacific
North Pacific
North Pacific
South Pacific
South Pacific
South Pacific
Gulf
East Coast
Great Lakes
CORRESPONDING
EMISSION INVENTORY
REGION
North Pacific (NP)
Alaska East (AE)
Alaska West (AW)
South Pacific (SP)
Hawaii East (HE)
Hawaii West (HW)
Gulf Coast (GC)
East Coast (EC)
Great Lakes (GL)
         Growth Factors for the Emission Inventory Analysis

       Emission inventories for 2020 are estimated by multiplying the 2002 baseline inventory for
each region by a corresponding growth factor that was developed from the RTI regional results.
Specifically, the average annual growth rate from 2002-2020 was calculated for each of the five
regions.  Each regional growth rate was then compounded over the inventory projection time period
for 2020, i.e., 18 years.  The resulting multiplicative growth factors for each emission inventory
region and the associated RTI average annual growth rates are presented in Table 2D-10 for 2020.

                    Table 2D-10 Regional Emission Inventory Growth Factors for 2020
EMISSION
INVENTORY
REGION
Alaska East (AE)
Alaska West (AW)
East Coast (EC)
Gulf Coast (GC)
Hawaii East (HE)
Hawaii West (HW)
North Pacific (NP)
South Pacific (SP)
Great Lakes (GL)
2002-2020 AVERAGE
ANNUALIZED GROWTH
RATE (%)
3.3
3.3
4.5
2.9
5.0
5.0
o o
J.J
5.0
1.7
MULTIPLICATIVE
GROWTH FACTOR
RELATIVE TO 2002
1.79
1.79
2.21
1.67
2.41
2.41
1.79
2.41
1.35
48 National Oceanic and Atmospheric Administration, Exclusive Economic Zone, Available online at
http://nauticalcharts.noaa.gov/csdl/eez.htm.

49 U.S. Department of Interior, North American Atlas - Political Boundaries, Available online at
http://www.nationalatlas.gov/mld/boundOm.html.
                                            2-69

-------
50 RTI International (December 2006). Global Trade and Fuels Assessment - Future Trends and Effects of Designation
Requiring Clean Fuels in the Marine Sector:  Task Order No. 1, Draft Report, prepared for the U.S. Environmental
Protection Agency, EPA Report Number EPA420-D-07-006, Docket ID EPA-HQ-OAR-2007-0121-0063.3.

51 RTI International (April 24, 2006). RTI Estimates of Growth in Bunker Fuel Consumption, Memorandum with
spreadsheet from Michael Gallaher and Martin Ross, RTI, to Barry Garelick and Russ Smith, U.S. Environmental
Protection Agency.

52 DVIO.  Revision of MARPOL Annex VI and the NOX technical code. Input from the four subgroups and individual
experts to the final report of the Informal Cross Government/Industry Scientific Group of Experts. BLG/INF.10
12/28/2007

53 Transport Canada; Transportation in Canada Annual Report 2004. 2004. (Tables 3-26 and 8-27).
http://www.tc.gc.ca/pol/en/report/anre2004/8F_e.htm.

54 RTI International (December 2006). Global Trade and Fuels Assessment - Future Trends and Effects of Designation
Requiring Clean Fuels in the Marine Sector:  Task Order No. 1, Draft Report, prepared for the U.S. Environmental
Protection Agency, EPA Report Number EPA420-D-07-006, Docket ID EPA-HQ-OAR-2007-0121-0063.3.

55 RTI International (December 2006). Global Trade and Fuels Assessment - Future Trends and Effects of Designation
Requiring Clean Fuels in the Marine Sector:  Task Order No. 1, Draft Report, prepared for the U.S. Environmental
Protection Agency, EPA Report Number EPA420-D-07-006, Docket ID EPA-HQ-OAR-2007-0121-0063.3.

56 Corbett, James and Chengfeng Wang (October 26, 2005). Emission Inventory Review SECA Inventory Progress
Discussion, p 11, memorandum to California Air Resources Board.

57 RTI International (December 2006). Global Trade and Fuels Assessment - Future Trends and Effects of Designation
Requiring Clean Fuels in the Marine Sector:  Task Order No. 1, Draft Report, prepared for the U.S. Environmental
Protection Agency, EPA Report Number EPA420-D-07-006, Docket ID EPA-HQ-OAR-2007-0121-0063.3.

58 RTI International (December 2006). Global Trade and Fuels Assessment - Future Trends and Effects of Designation
Requiring Clean Fuels in the Marine Sector:  Task Order No. 1, Draft Report, prepared for the U.S. Environmental
Protection Agency, EPA Report Number EPA420-D-07-006, Docket ID EPA-HQ-OAR-2007-0121-0063.3.
                                               2-70

-------
CHAPTER 3:  IMPACTS OF SHIPPING EMISSIONS ON AIR QUALITY, HEALTH
AND THE ENVIRONMENT	3-2
3.1   Pollutants Reduced by the ECA	3-2
3.2   Health Effects Associated with Exposure to Pollutants Reduced by the ECA	3-5
3.3   Ecosystem Impacts Associated with Exposure to Pollutants Reduced by the ECA. 3-13
                                      3-1

-------
CHAPTER 3: Impacts of Shipping Emissions on Air Quality, Health
                  and the Environment

       Designation of this Emission Control Area (EGA) would significantly reduce emissions
of NOx, SOx and PM2.5 and thereby reduce ambient levels of particulate matter and ground-level
ozone in Puerto Rico and the U.S. Virgin Islands.  The improvement in ambient air quality
would result in benefits to human health and the environment.  This chapter describes the
pollutants that would be reduced due to the EGA designation and their impacts on human health
and the environment.

3.1 Pollutants Reduced by the ECA

3.1.1 Particulate Matter

       Ships that operate in the proposed ECA generate emissions that increase on-land
concentrations  of harmful air pollutants such as particulate matter (PM) .  PM is a generic term
for a broad class of chemically and physically diverse substances. It can be principally
characterized as discrete particles that exist in the condensed (liquid or solid) phase spanning
several orders of magnitude in size. Since 1987, EPA has delineated that subset of inhalable
particles small  enough to penetrate to the thoracic region (including the tracheobronchial and
alveolar regions) of the respiratory tract (referred to as thoracic particles) . Current national
ambient air quality standards (NAAQS) use PM2.5 as the indicator for fine particles (with PM2.5
referring to particles with a nominal mean aerodynamic diameter less than or equal to 2.5 pm),
and use PMio as the indicator for purposes of regulating the coarse fraction of PMio (referred to
as thoracic coarse particles  or coarse-fraction particles; generally including particles with a
nominal mean aerodynamic diameter greater than  2.5 pm and less than or equal to 10 pm, or
PMio-z.s). Ultrafine particles (UFPs) are a subset of fine particles, generally less than 100
nanometers (0.1 ^im) in aerodynamic diameter.
       Particles span many sizes and shapes and consist of numerous different chemicals.
Particles originate from sources and are also formed through atmospheric chemical reactions; the
former are often referred to as "primary" particles, and the latter as "secondary" particles. In
addition, there are also physical, non-chemical reaction mechanisms that contribute to secondary
particles. Particle pollution also varies by time of year and location and is affected by several
weather-related factors, such as temperature, clouds, humidity, and wind. A further layer of
complexity comes from a particle's ability to shift between solid/liquid and gaseous phases,
which is influenced by concentration, meteorology, and temperature.

       Fine particles are produced primarily by combustion processes and by transformations of
gaseous emissions (e.g., NOx, SOx and volatile organic compounds (VOC)) in the atmosphere.
The chemical and physical properties of PM2.s may vary greatly with time, region, meteorology
and source category. Thus, PM2.5 may include a complex mixture of different chemicals
including sulfates, nitrates, organic compounds, elemental carbon and metal compounds. These
particles can remain in the atmosphere for days to weeks and travel through the atmosphere
hundreds to thousands  of kilometers. l
                                          3-2

-------
3.1.2 Ozone

       Ground-level ozone pollution is typically formed by the reaction of VOC and NOx in the
lower atmosphere in the presence of sunlight. These pollutants, often referred to as ozone
precursors, are emitted by many types of pollution sources such as highway and nonroad motor
vehicles and engines, including ships, power plants, chemical plants, refineries, makers of
consumer and commercial products, industrial facilities, and smaller area sources.

       The science of ozone formation, transport, and accumulation is complex.  Ground-level
ozone is produced and destroyed in a cyclical set of chemical reactions, many of which are
sensitive to temperature and sunlight. When ambient temperatures and sunlight levels remain
high for several days and the air is relatively stagnant, ozone and its precursors can build up and
result in more ozone than typically occurs on a single high-temperature day.  Ozone can be
transported hundreds of miles downwind of precursor emissions, resulting in elevated ozone
levels even in areas with low VOC or NOx emissions.

       The highest levels of ozone are produced when both VOC and NOx emissions are present
in significant quantities on clear summer days. Relatively small amounts of NOx enable ozone
to form rapidly when VOC levels are relatively high,  but ozone production is quickly limited by
removal of the NOx. Under these conditions NOx reductions are highly effective in reducing
ozone while VOC reductions have little effect. Such  conditions are called "NOx-limited."
Because the contribution of VOC emissions from biogenic (natural) sources to local ambient
ozone concentrations can be significant, even some areas where man-made VOC emissions  are
relatively low can be NOx-limited.

       Ozone concentrations in an area also can be lowered by the reaction of nitric oxide (NO)
with ozone, forming nitrogen dioxide (N02);  as the air moves downwind and the cycle continues,
the N02 forms additional ozone.  The importance of this reaction depends,  in part, on the relative
concentrations of NOx, VOC, and ozone, all of which change with time and location. When
NOx levels are relatively high and VOC levels relatively low, NOx forms inorganic nitrates (i.e.,
particles) but relatively little ozone.  Such conditions  are called "VOC-limited." Under these
conditions, VOC reductions are effective in reducing  ozone, but NOx reductions can actually
increase local ozone under certain circumstances. Even in VOC-limited urban areas, NOx
reductions are not expected to increase ozone levels if the NOx reductions are sufficiently large.
Rural areas are usually NOx-limited, due to the relatively large amounts of biogenic VOC
emissions in such areas. Urban areas can be either VOC- or NOx-limited, or a mixture of both,
in which ozone levels exhibit moderate sensitivity to changes in either pollutant.

3.1.3 NO2 and  SO2

       Sulfur dioxide (S02), a member of the sulfur oxide (SOx) family of gases, is formed from
burning fuels containing sulfur (e.g., coal or oil), extracting gasoline from oil, or extracting
metals from ore.  Nitrogen dioxide (NO2) is a member of the nitrogen oxide (NOx) family of
gases.  Most N02 is formed in the air through the oxidation of nitric oxide (NO) emitted when
fuel is burned at a high temperature. Ships emit both  N02 and S02. S02 andN02 can dissolve in
water vapor and further oxidize to form sulfuric and nitric acid which reacts with ammonia to
form sulfates and nitrates, both of which are important components of ambient PM. The health
                                          3-3

-------
effects of ambient PM are discussed in Section 3.2.1.  NOx along with non-methane
hydrocarbons (NMHC) are the two major precursors of ozone. The health effects of ozone are
covered in Section 3.2.2.

3.1.4 Diesel Exhaust PM

       Ship emissions contribute to ambient levels of air toxics known or suspected as human or
animal carcinogens, or that have noncancer health effects. The population experiences an
elevated risk of cancer and other noncancer health effects from exposure to air toxics.2 These
compounds include diesel PM.

       Marine diesel engines emit diesel exhaust (DE), a complex mixture comprised of carbon
dioxide, oxygen, nitrogen, water vapor, carbon monoxide, nitrogen compounds, sulfur
compounds and numerous low molecular-weight hydrocarbons.  A number of these gaseous
hydrocarbon components are individually known to be toxic including aldehydes, benzene and
1,3-butadiene. The diesel particulate matter  (DPM) present in diesel exhaust consists  of fine
particles (< 2.5pm), including a subgroup with a large number of ultrafine particles (< 0.1 pm).
These particles have a large surface area, which makes them an excellent medium for  adsorbing
organics, and their small size makes them highly respirable. Many of the organic compounds
present in the gases and on the particles, such as polycyclic organic matter (POM), are
individually known to have mutagenic and carcinogenic properties. Marine diesel engine
emissions consist of a higher fraction of hydrated sulfate (approximately 60-90%) due to the
higher sulfur levels of the fuel, organic carbon  (approximately 15-30%),  and metallic ash
(approximately 7-11%) than are typically found in land-based engines.3  In addition, while toxic
trace metals emitted by marine diesel engines represent a very small portion of the national
emissions of metals (less than one percent) and are a small portion of DPM (generally much less
than one percent of DPM), we note that several trace metals of potential toxicological
significance and persistence in the environment are emitted by diesel engines.4  These trace
metals include chromium, manganese, mercury, and nickel.  In addition,  small amounts of
dioxins have been measured in highway engine diesel exhaust, some of which may partition into
the particulate phase. Dioxins are a major health concern but diesel engines are a minor
contributor to overall dioxin emissions.

       Diesel exhaust varies significantly in chemical composition and particle sizes between
different engine types (heavy-duty, light-duty), engine operating conditions  (idle, accelerate,
decelerate), and fuel formulations (high/low sulfur fuel).   Also, there  are emissions differences
between on-road and nonroad engines because  the nonroad engines are generally of older
technology.  This is especially true for marine diesel engines.5 After being emitted in the engine
exhaust, diesel exhaust undergoes dilution as well as chemical and physical changes in the
atmosphere. The lifetime for some of the compounds present in diesel exhaust ranges from
hours to days.
                                          3-4

-------
3.2 Health Effects Associated with Exposure to Pollutants Reduced by the
    ECA

3.2.1 PM Health Effects

       This section provides a summary of the health effects associated with exposure to
ambient concentrations of PM.A The information in this section is based on the information and
conclusions in the Integrated Science Assessment (ISA) for Particulate Matter (December 2009)
prepared by EPA's Office of Research and Development (ORD).B

       The ISA concludes that ambient concentrations of PM are associated with a number of
adverse health effects.0 The ISA characterizes the weight of evidence for different health effects
associated with three PM size ranges: PM2.5, PMio-z.s, and UFPs. The discussion below
highlights the ISA's conclusions pertaining to these three size fractions of PM, considering
variations in both short-term and long-term exposure periods.

        Information specifically related to health effects associated with exposure to diesel
exhaust PM is included in Section 3.2.5 of this document.

3.2.1.1  Effects Associated with Short-term Exposure to PM2.s

       The ISA concludes that cardiovascular effects and all-cause cardiovascular- and
respiratory-related mortality are causally associated with short-term exposure to PM2.5.6  It also
concludes that respiratory effects are likely to be causally associated with short-term exposure to
PM2.5, including respiratory emergency department (ED) visits and hospital admissions for
chronic obstructive pulmonary disease (COPD), respiratory infections, and asthma; and
exacerbation of respiratory symptoms in asthmatic children.

3.2.1.2  Effects Associated with Long-term Exposure to PM2.s

       The ISA concludes that there are causal associations between long-term exposure to
PM2.s and cardiovascular effects, such as the development/progression of cardiovascular disease
(CVD), and premature mortality, particularly from cardiopulmonary causes.7  It also concludes
that long-term exposure to PM2.5 is likely to be causally associated with respiratory effects, such
as reduced lung function growth, increased respiratory symptoms, and asthma development.  The
ISA characterizes the evidence as suggestive of a causal relationship for associations between
A Personal exposure includes contributions from many different types of particles, from many sources, and in many
different environments. Total personal exposure to PM includes both ambient and nonambient components; and
both components may contribute to adverse health effects.
B The ISA is available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=216546
c The ISA evaluates the health evidence associated with different health effects, assigning one of five "weight of
evidence"  determination:  causal relationship, likely to be a causal relationship, suggestive of a causal relationship,
inadequate to infer a causal relationship, and not likely to be a causal relationship. For definitions of these levels of
evidence, please refer to Section 1.5 of the ISA. The following text summarizes only those health effects with at
least a "suggestive" weight of evidence determination.
                                            3-5

-------
long-term PM2.s exposure and reproductive and developmental outcomes, such as low birth
weight and infant mortality. It also characterizes the evidence as suggestive of a causal
relationship between PM2.5 and cancer incidence, mutagenicity, and genotoxicity.

3.2.1.3  Effects Associated with PMi0-2.s

       The ISA summarizes evidence related to short-term exposure to PMio-z.s- PMio-2.5 is the
fraction of PMio particles that is larger than PM2.5.8 The ISA concludes that available evidence
is suggestive of a causal relationship between short-term exposures to PMio-2.5 and
cardiovascular effects, such as hospitalizations for ischemic heart disease. It also concludes that
the available evidence is suggestive of a causal relationship between short-term exposures to
PMio-2.5 and respiratory effects, including respiratory-related ED visits and hospitalizations and
pulmonary inflammation. The ISA also concludes that the available literature suggests a causal
relationship between short-term exposures to PMi0-2.5 and mortality. Data are inadequate to
draw conclusions regarding health effects associated with long-term exposure to PMio-2.5.9

3.2.1.4  Effects Associated with Ultrafine Particles

       The ISA concludes that the evidence is suggestive of a causal relationship between short-
term exposures to ultrafine particles (UFP) and cardiovascular effects, including changes in heart
rhythm and vasomotor function (the ability of blood vessels to expand and contract).

       The ISA also concludes that there is suggestive evidence of a causal relationship between
short-term UFP exposure and respiratory effects. The types of respiratory effects examined in
epidemiologic studies include respiratory symptoms and asthma hospital admissions, the results
of which are not entirely  consistent. There is evidence from toxicological and controlled human
exposure studies that exposure to UFPs may increase lung inflammation and produce small
asymptomatic changes in lung function. Data are inadequate to draw conclusions regarding
health effects associated with long-term exposure to UFPs.11

3.2.2 Ozone Health Effects

       Exposure to ambient ozone contributes to a wide range of adverse health effects.0 These
health effects are well documented and are critically assessed in the EPA ozone air quality
criteria document (ozone AQCD)  and EPA staff paper.12'13  We are relying on the data and
conclusions in the ozone  AQCD and staff paper, regarding the health effects associated with
ozone exposure.

       Ozone-related health effects include lung function decrements, respiratory symptoms,
aggravation of asthma, increased hospital and emergency room visits, increased asthma
medication usage, and a variety of other respiratory effects. Cellular-level effects, such as
inflammation of lungs, have been documented as well.  In addition, there is suggestive evidence
D Human exposure to ozone varies over time due to changes in ambient ozone concentration and because people
move between locations which have notable different ozone concentrations. Also, the amount of ozone delivered to
the lung is not only influenced by the ambient concentrations but also by the individuals breathing route and rate.
                                            3-6

-------
of a contribution of ozone to cardiovascular-related morbidity and highly suggestive evidence
that short-term ozone exposure directly or indirectly contributes to non-accidental and
cardiopulmonary-related mortality, but additional research is needed to clarify the underlying
mechanisms causing these effects.  In a recent report on the estimation of ozone-related
premature mortality published by the National Research Council (NRC), a panel of experts and
reviewers concluded that short-term exposure to ambient ozone is likely to contribute to
premature deaths and that ozone-related mortality should be included in estimates of the health
benefits of reducing ozone exposure.14  People who appear to be more susceptible to effects
associated with exposure to ozone include children, asthmatics and the elderly. Those with
greater exposures to ozone, for instance due to time spent outdoors (e.g., children and outdoor
workers),  are also of concern.

       Based on a large number of scientific studies, EPA has identified several key health
effects associated with exposure to levels of ozone found today in many areas of the country.
Short-term (1 to 3 hours) and prolonged exposures (6 to 8 hours) to ambient ozone
concentrations have been linked to lung function decrements, respiratory symptoms, increased
hospital admissions and emergency room visits for respiratory problems.    16'17'18'19'20
Repeated exposure to ozone can increase susceptibility to respiratory infection and lung
inflammation and can aggravate preexisting respiratory diseases, such as asthma.21'22' 3'24'25
Repeated exposure to sufficient concentrations of ozone can also cause inflammation of the lung,
impairment of lung defense mechanisms, and possibly irreversible changes in lung structure,
which over time could affect premature aging of the lungs and/or the development of chronic
respiratory illnesses, such as emphysema and chronic bronchitis.26'27'28'29

       Children and adults who are outdoors and active during the summer months, such as
construction workers, are among those most at risk of elevated ozone exposures.30  Children and
outdoor workers tend to have higher ozone exposure because they typically are active outside,
working, playing and exercising, during times of day and seasons (e.g., the summer) when ozone
levels are  highest.31 For example, summer camp studies in the Eastern United States and
Southeastern Canada have reported statistically significant reductions in lung function in
children who are active outdoors.32'33'34'35'36'37'   '39  Further, children are more at risk of
experiencing health effects from ozone exposure than adults because their respiratory systems
are still developing. These individuals (as well as people with respiratory illnesses, such as
asthma, especially asthmatic children)  can experience reduced lung function and increased
respiratory symptoms, such as chest pain and cough, when exposed to relatively low ozone levels
during prolonged periods of moderate exertion.40'41'42'43

3.2.3 SO2 Health Effects

       This section provides an overview of the health effects associated with SOz. Additional
information on the health effects of S02 can be found in the EPA Integrated Science Assessment
for Sulfur Oxides.44 Following an extensive evaluation of health evidence from epidemiologic
and laboratory studies, the U.S. EPA has concluded that there is a causal relationship between
respiratory health effects and short-term exposure to SOz. The immediate effect of SOz on the
respiratory system in humans is bronchoconstriction. Asthmatics are more sensitive to the effects
of SOz likely resulting from preexisting inflammation associated with this disease.  In laboratory
studies involving controlled human exposures to S02, respiratory effects have consistently been
                                           3-7

-------
observed following 5-10 min exposures at S02 concentrations > 0.4 ppm in asthmatics engaged
in moderate to heavy levels of exercise, with more limited evidence of respiratory effects among
exercising asthmatics exposed to concentrations as low as 0.2-0.3 ppm.  A clear concentration-
response relationship has been demonstrated in these studies following exposures to S02 at
concentrations between 0.2 and  1.0 ppm, both in terms of increasing severity of respiratory
symptoms and decrements in lung function, as well as the percentage of asthmatics adversely
affected.

       In epidemiologic studies, respiratory effects have been observed in areas where the mean
24-hour S02 levels range from 1 to 30 ppb, with maximum 1 to 24-hour average S02 values
ranging from 12 to 75 ppb.  Important new multicity studies and several other studies have found
an association between 24-hour average ambient S02 concentrations and respiratory symptoms
in children, particularly those with asthma. Generally consistent associations also have been
observed between ambient S02 concentrations and emergency department visits and
hospitalizations for all respiratory causes, particularly  among children and older adults (> 65
years), and for asthma.  A limited subset of epidemiologic studies have examined potential
confounding by copollutants using multipollutant regression models. These analyses indicate
that although copollutant adjustment has varying degrees of influence on the S02 effect
estimates, the effect of S02 on respiratory health outcomes appears to be generally robust and
independent of the effects of gaseous and particulate copollutants,  suggesting that the observed
effects of S02 on respiratory endpoints occur independent of the effects of other ambient air
pollutants.

       Consistent associations between short-term exposure to S02 and mortality have been
observed in epidemiologic studies, with larger effect estimates reported for respiratory mortality
than for cardiovascular mortality.  While this finding is consistent with the demonstrated effects
of S02 on respiratory morbidity, uncertainty remains with respect to the interpretation of these
associations due to potential confounding by various copollutants.  The U.S. EPA has therefore
concluded that the overall evidence is suggestive of a causal relationship between short-term
exposure to S02 and mortality.   Significant associations between short-term exposure to S02
and emergency department visits and hospital admissions for cardiovascular diseases have also
been reported. However, these findings have been inconsistent across studies and do not provide
adequate evidence to infer a causal relationship between S02 exposure and cardiovascular
morbidity.

3.2.4 NO2 Health Effects

       Information on the health effects of N02 can be found in the EPA Integrated Science
Assessment (ISA) for Nitrogen Oxides.45 The EPA has concluded that the findings of
epidemiologic, controlled human exposure, and animal toxicological studies provide evidence
that is sufficient to infer a likely causal relationship between respiratory effects and short-term
N02 exposure. The ISA concludes that the strongest evidence for such a relationship comes from
epidemiologic studies of respiratory effects including symptoms, emergency department visits,
and hospital admissions. The ISA also draws two broad conclusions regarding airway
responsiveness following N02 exposure. First, the ISA concludes that N02 exposure may
enhance the sensitivity to allergen-induced decrements in lung function and increase the
allergen-induced airway inflammatory response following 30-minute exposures of asthmatics to
                                           3-8

-------
     concentrations as low as 0.26 ppm. In addition, small but significant increases in non-
specific airway hyperresponsiveness were reported following 1-hour exposures of asthmatics to
0.1 ppm NOz.  Second, exposure to NOz has been found to enhance the inherent responsiveness
of the airway to subsequent nonspecific challenges in controlled human exposure studies of
asthmatic subjects.  Enhanced airway responsiveness could have important clinical implications
for asthmatics since transient increases in airway responsiveness following N02 exposure have
the potential to increase symptoms and worsen asthma control.  Together, the epidemiologic and
experimental data sets form a plausible, consistent, and coherent description of a relationship
between NOz exposures and an array of adverse health effects that range from the onset of
respiratory symptoms to hospital admission.

       Although the weight of evidence supporting a causal relationship is somewhat less certain
than that associated with respiratory morbidity, N02 has also been linked to other health
endpoints. These include all-cause (nonaccidental) mortality, hospital admissions or emergency
department visits for cardiovascular disease, and decrements in lung function growth associated
with chronic exposure.

3.2.5 Diesel Exhaust PM Health Effects

       A large number of health studies have been conducted regarding diesel exhaust.  These
include epidemiologic studies of lung cancer in groups of workers and animal studies focusing
on non-cancer effects. Diesel exhaust PM (including the associated organic compounds which
are generally high molecular weight hydrocarbons but not the more volatile gaseous hydrocarbon
compounds)  is generally used as a surrogate exposure measure for whole diesel exhaust.

       Diesel exhaust has been found to be of concern by several groups worldwide including
the U.S. government. The IPCS (International Programme on Chemical Safety) has established
environmental health criteria for diesel fuel and exhaust emissions.  In these criteria, the IPCS
recommends that for the protection of human health diesel exhaust emissions should be
controlled. The IPCS explicitly states that urgent efforts should be made to reduce emissions,
specifically of particulates, by changing exhaust train techniques,  engine design and fuel
composition.46

3.2.5.1  Potential Cancer Effects of Exposure to Diesel Exhaust

       The U.S. EPA's 2002 final "Health Assessment Document for Diesel Engine Exhaust"
(the EPA Diesel HAD) classified exposure to  diesel exhaust as likely to be carcinogenic to
humans by inhalation at environmental exposures, in accordance with the revised draft
1996/1999 U.S. EPA cancer guidelines.47'48 In accordance with earlier U.S. EPA guidelines,
exposure to diesel exhaust would similarly be classified as probably carcinogenic to humans
(Group Bl) .49'50  A number of other agencies (National Institute for Occupational Safety and
Health, the International Agency for Research on Cancer, the World Health Organization,
California EPA, and the U.S. Department of Health and Human Services) have made similar
classifications.51'52'53'54'55 The Health Effects Institute has prepared numerous studies and
reports on the potential carcinogenicity of exposure to diesel exhaust.56'57'58
                                          3-9

-------
       More specifically, the U.S. EPA Diesel HAD states that the conclusions of the document
apply to diesel exhaust in use today including both onroad and nonroad engines including marine
diesel engines present on ships.  The U.S. EPA Diesel HAD acknowledges that the studies were
done on engines with generally older technologies and that "there have been changes in the
physical and chemical composition of some DE [diesel exhaust] emissions (onroad vehicle
emissions) over time, though there is no definitive information to show that the emission changes
portend significant toxicological changes."  In any case, the diesel technology used for marine
diesel engines typically lags that used for onroad engines, which have been subject to PM
standards since 1998.  Thus, it is reasonable to assume that the hazards identified from older
technologies may be largely applicable to marine engines.

       For the Diesel  HAD, the U.S. EPA reviewed 22 epidemiologic studies on the subject of
the carcinogenicity of exposure to diesel exhaust in various occupations, finding increased lung
cancer risk, although not always statistically significant, in 8 out of 10 cohort studies and 10 out
of 12 case-control studies which covered several industries. Relative risk for lung cancer,
associated with exposure, ranged from 1.2 to 1.5, although a few studies show relative risks as
high as 2.6.  Additionally, the Diesel HAD  also relied on two independent meta-analyses, which
examined 23 and 30 occupational studies respectively, and found statistically significant
increases of 1.33 to 1.47 in smoking-adjusted relative lung cancer risk associated with diesel
exhaust. These meta-analyses demonstrate the effect of pooling many studies and in this case
show the positive relationship between diesel exhaust exposure and lung cancer across a variety
of diesel exhaust-exposed occupations.59'60'61

       The U.S. EPA recently assessed air  toxic emissions and their associated risk (the
National-Scale Air Toxics Assessment or NATA for 1996 and 1999),  and  concluded that diesel
exhaust ranks with other emissions that the national-scale assessment  suggests pose the greatest
relative risk.62'63 This national assessment estimates average population inhalation exposures to
DPM for nonroad and on-highway sources. These exposures are the sum of ambient levels
weighted by the amount of time people spend in each of the locations.

       In summary, the likely hazard to humans together with the potential for significant
environmental risks leads us to conclude that diesel exhaust emissions from marine engines
present public health issues of concern.

3.2.5.2  Other Health Effects of Exposure to Diesel Exhaust

       Noncancer health effects of acute and chronic exposure to diesel exhaust emissions are
also of concern. The Diesel HAD established an inhalation Reference Concentration (RfC)
specifically based on animal studies of diesel exhaust exposure.  An RfC is defined by the U.S.
EPA as "an estimate of a continuous inhalation exposure to the human population, including
sensitive subgroups, with uncertainty spanning perhaps an order of magnitude, which is likely to
be without appreciable risks of deleterious noncancer effects during a lifetime." The U.S. EPA
derived the RfC from consideration of four well-conducted chronic rat inhalation studies
showing adverse pulmonary effects.64'65'66'67  The diesel RfC is based on a "no observable
adverse effect" level of  144 pg/m3 that is further reduced by applying uncertainty factors of 3
for interspecies extrapolation and 10 for human variations in sensitivity. The resulting RfC
derived in the Diesel HAD is 5 pg/m3 for diesel exhaust, as measured by DPM. This RfC does
                                          3-10

-------
not consider allergenic effects such as those associated with asthma or immunologic effects.
There is growing evidence that exposure to diesel exhaust can exacerbate these effects, but the
exposure-response data is presently lacking to derive an RfC. The Diesel HAD states, "With
DPM [diesel particulate matter] being a ubiquitous component of ambient PM, there is an
uncertainty about the adequacy of the existing DE [diesel exhaust] noncancer database to
identify all of the pertinent DE-caused noncancer health hazards" (p. 9-19).

       While there have been relatively few human studies associated specifically with the
noncancer impact of exposure to DPM alone, DPM is a component of the ambient particles
studied in numerous epidemiologic studies. The conclusion that health effects associated with
ambient PM in general are relevant to DPM is supported by studies that specifically associate
observable human noncancer health effects with exposure to DPM. As described in the Diesel
HAD, these studies identified some of the same health effects reported for ambient PM, such as
respiratory symptoms (cough, labored breathing, chest tightness, wheezing), and chronic
respiratory disease (cough, phlegm, chronic bronchitis and suggestive evidence for decreases in
pulmonary function). Symptoms of immunological effects such as wheezing and increased
allergenicity are also seen. Studies in rodents, especially rats, show the potential for human
inflammatory effects in the lung and  consequential lung tissue damage from chronic diesel
exhaust inhalation exposure.  The Diesel HAD concludes "that acute exposure to DE [diesel
exhaust] has been associated with irritation of the eye, nose, and throat, respiratory symptoms
(cough and phlegm), and neurophysiological symptoms such as headache, lightheadedness,
nausea, vomiting, and numbness or tingling of the extremities."68 There is also evidence for an
immunologic effect such as the exacerbation of allergenic responses to known allergens and
asthma-like symptoms.69'70'71

       The Diesel HAD briefly summarizes health effects associated with ambient PM and
discusses the PM2.5 NAAQS. There is a much more extensive body of human data, which is also
mentioned earlier in the health effects discussion for PM2.s (Section 3.2.1.1 of this document),
showing a wide spectrum of adverse  health effects associated with exposure to ambient PM, of
which diesel exhaust is an important  component.  The PM2.s  NAAQS is designed to provide
protection from the non-cancer and premature mortality effects of PM2.s as a whole.

3.2.5.3  Exposure to Diesel  Exhaust PM

       Exposure of people to diesel exhaust depends on their various activities, the time spent in
those activities, the locations  where these activities occur, and the levels of diesel exhaust
pollutants in those locations.  The major difference between ambient levels of diesel particulate
and exposure levels for diesel particulate is that exposure levels account for a person moving
from location to location, the proximity to the emission source, and whether the exposure occurs
in an enclosed environment.

       Occupational exposures to diesel exhaust from mobile sources, including marine diesel
engines, can be several orders of magnitude greater than typical exposures in the non-
occupationally exposed population. Over the years, diesel particulate exposures have been
measured for a number of occupational groups resulting in a wide range of exposures from 2 to
1280 pg/m3 for a variety of occupations. As discussed in the Diesel HAD, the National Institute
of Occupational Safety and Health (NIOSH) has estimated a total of 1,400,000 workers are
                                          3-11

-------
occupationally exposed to diesel exhaust from on-road and nonroad vehicles including marine
diesel engines.

3.2.5.3.1  Elevated Concentrations and Ambient Exposures in Mobile Source-Impacted Areas

       While occupational studies indicate that those working in closest proximity to diesel
exhaust experience the greatest health effects, recent studies are showing that human populations
living near large diesel emission sources such as major roadways,72 rail yards,73 and marine
ports 4 are also likely to experience greater exposure to PM and other components of diesel
exhaust than the overall population, putting them at a greater health risk. The percentage of total
port  emissions that come from ships varies by port. However, ships contribute to the  DPM
concentrations at ports, and elsewhere, which influence exposures.

       Regions immediately downwind of marine ports may experience elevated ambient
concentrations of directly-emitted PM2.s from diesel engines. Due to the nature of marine ports,
emissions from a large number of diesel engines are concentrated in a small area. Recent studies
conducted in the continental United States have looked at air quality impacts of diesel engine
emissions from ports. Although this proposed EGA is for Puerto Rico and the U.S. Virgin
Islands, the contribution from ports to elevated ambient concentrations  of diesel exhaust in
populated areas on the U.S. mainland is relevant since there are also ports near populated areas
of Puerto Rico and the U.S. Virgin Islands.

       A study from the California Air Resources Board (CARB) evaluated air quality impacts
of diesel engine emissions within the  Port of Long Beach and Los Angeles in California, one of
the largest ports in the U.S.75  The port study employed the ISCST3 dispersion model. With
local meteorological  data used in the modeling, annual average concentrations of DPM were
substantially elevated over an area exceeding 200,000 acres. Because the Ports are located near
heavily-populated areas, the modeling indicated that over 700,000 people lived in areas with at
least 0.3 pg/m3 of port-related DPM in ambient air, about 360,000 people lived in areas with at
least 0.6 pg/m3 of DPM, and about 50,000 people lived in areas with at least 1.5 pg/m3 of
ambient DPM emitted directly from the port. This port study highlights the substantial
contribution these facilities make to ambient concentrations of DPM in large, densely populated
areas.

       The U.S. EPA updated an initial screening-level analysis76'77 of selected marine port
areas to better understand the  populations that are exposed to diesel particulate matter (DPM)
emissions from these facilities.E The results of this study are summarized here and are also
available in the public docket.78'79 In summary, the screening-level analysis found that for the
45 U.S. marine ports studied,  al least  18 million people live in the vicinity of these facilities and
are exposed to ambient DPM  levels from all port emission sources that are at least 0.2 pg/m3
above those found in areas further from these facilities.  If only Category 3 engine DPM
emissions are considered, then the number of people exposed is 6.5 million.
E This type of screening-level analysis is an inexact tool and not appropriate for regulatory decision-making; it is
useful in beginning to understand potential impacts and for illustrative purposes.
                                           3-12

-------
3.2.6 Puerto Rico Asthma Rates

       Emissions of PM, SOx and NOx from ships contribute to ambient concentrations of PM,
ozone, SOx and NOx. As explained above, there are well established links between ambient
concentrations of PM, ozone, SOx and NOx and asthma. Two studies by the Puerto Rico
Department of Health in collaboration with the Puerto Rico Asthma Coalition found a higher
asthma mortality rate in Puerto Rico than for the continental United States for the period since
1980. For the period 1980 to 1998, these researchers find that the asthma mortality rate for
Puerto Rico was 2.5 times higher than in the continental U.S. While the Puerto Rican asthma
mortality rate experienced a decreasing and then stable pattern from 2000-2004, it remains about
two times higher than that in the continental U.S. for that same time period. The more recent of
the two studies also looked  at the lifetime asthma prevalence in Puerto Rico, defined as  those
individuals who at some time in their life have been diagnosed with asthma. This study found
the lifetime asthma prevalence rate over the period 2000-2007 to be 1.5 times higher in Puerto
Rico than in the continental U.S. The reductions in PM, NOx, and SOx emissions as  a result of
this proposed EGA would aid in reducing the prevalence of and mortality from asthma in Puerto
Rico. In addition to helping reduce asthma rates, lowering ships emissions of NOx, SOx, and
PM would also have a positive impact on the many other serious health problems detailed in this
section.

3.3 Ecosystem Impacts Associated with Exposure to Pollutants Reduced by
    the ECA

       In addition to their health impacts, emissions of NOx, SOx, and PM from ships are also
of concern in Puerto Rico and the U.S. Virgin Islands because they cause harm to ecosystems.
As mentioned above, Puerto Rico and the U.S. Virgin Islands are rich in biodiversity  and contain
many sensitive ecosystems ranging from bioluminescent bays and tropical mangrove  swamps to
coral reefs. This section looks at ecosystem impacts of NOx, SOx and PM emissions  including
deposition, acidification and nutrient enrichment, ozone impacts on plants and trees and visibility
degradation.

3.3.1 Deposition

       Ship engines emit large amounts of NOX, SOX and direct PM over a wide area.
Depending on prevailing winds and other meteorological conditions, these emissions  may be
transported hundreds and even thousands of kilometers across the  entire width of Puerto Rico
and the U.S. Virgin Islands  and impact not only ambient air concentrations but also contribute to
deposition of nitrogen and sulfur in many sensitive ecological areas.

       Ships operating on high sulfur fuel emit both SOz and sulfate PM.  The sulfur  in marine
fuel is primarily emitted as sulfur dioxide (SO2), with a small fraction (about two percent) being
converted to sulfur trioxide  (SOs).80  SOs almost immediately forms sulfate, which is emitted as
primary PM by the engine, and consists of carbonaceous material, sulfuric acid, and ash (trace
metals). These particles also react in the atmosphere to form secondary PM such as sulfuric acid
aerosols, or sulfate particles.
                                         3-13

-------
       Ships also emit large amounts of nitric oxide (NO) and nitrogen dioxide (NO2) which are
carried into the atmosphere where they may be chemically altered and transformed into new
compounds.  For example, N02 can be further oxidized to nitric acid (HN03) and can contribute
in that form to the acidity of clouds, fog, and rain water and can also form ambient particulate
nitrate (pNOs) which may be deposited either directly onto terrestrial and aquatic ecosystems or
deposited onto land surfaces where it subsequently runs off and is transferred into downstream
waters.

       Deposition can occur either in a wet or dry form.  Wet deposition includes rain, snow,
sleet, hail, clouds, or fog. Dry deposition includes gases and dust.  Wet and dry atmospheric
deposition of PM2.5 delivers a complex mixture of metals (such as mercury, zinc, lead, nickel,
arsenic, aluminum, and cadmium), organic compounds (such as polycyclic organic matter,
dioxins, and furans) and inorganic compounds (such as nitrate, sulfate). Together these
emissions from ships are deposited onto terrestrial and aquatic ecosystems across Puerto Rico
and the U.S. Virgin Islands contributing to ecosystem problems.

       The chemical form of deposition is determined by ambient conditions (e.g., temperature,
humidity, oxidant levels) and the pollutant source. Chemical and physical transformations of
ambient particles occur in the atmosphere and in the media (terrestrial or aquatic) on which they
deposit.  These transformations influence the fate, bioavailability and potential toxicity of these
compounds.  The atmospheric deposition of metals and toxic compounds is implicated in severe
ecosystem effects.81

       The lifetimes of particles vary with particle size. Accumulation-mode particles such as
sulfates and nitrates are kept in suspension by normal air  motions and have a lower deposition
velocity than coarse-mode particles; they can be transported thousands of kilometers and remain
in the atmosphere for a number of days.

       Particulate matter is a factor in acid deposition. Particles serve as cloud condensation
nuclei and contribute directly to the acidification of rain.  In addition, the gas-phase species that
lead to the dry deposition of acidity are also precursors of particles.  Therefore, reductions in
NOx and SOx emissions will decrease both acid deposition and PM concentrations, but not
necessarily in a linear fashion. Sulfuric acid, ammonium nitrate, and organic particles also are
deposited on surfaces by dry deposition and can contribute to environmental effects.82

3.3.1.1   Nitrogen and Sulfur Deposition

       Nitrogen and sulfur interactions in the environment are highly complex. Both are
essential, and sometimes limiting, nutrients needed for growth and productivity. Excess of
nitrogen or sulfur can lead to acidification and nutrient enrichment.

       Deposition of nitrogen and sulfur species causes acidification, which alters
biogeochemistry and affects animal and plant life in terrestrial  and aquatic ecosystems across
Puerto Rico and the U.S. Virgin Islands.  Major effects include a decline in sensitive tree species
and a loss of biodiversity of fishes, zooplankton, and macro invertebrates.  The sensitivity of
terrestrial and aquatic  ecosystems to acidification from nitrogen and sulfur deposition is
predominantly governed by the earth's geology.
                                           3-14

-------
       Biological effects of acidification in terrestrial ecosystems are generally linked to
aluminum toxicity and decreased ability of plant roots to take up base cations. Decreases in acid
neutralizing capacity and increases in inorganic aluminum concentration also contribute to
declines in coral reefs, zooplankton, macro invertebrates, and fish species richness in aquatic
ecosystems. Across Puerto Rico and the U.S. Virgin Islands, ecosystems continue to be acidified
by current NOx and SOx emissions from stationary sources, area sources, and mobile sources.

       In addition to the role nitrogen deposition plays in acidification, it also causes ecosystem
nutrient enrichment and eutrophication that alters biogeochemical cycles and harms animal and
plant life such as native lichens and alters biodiversity of terrestrial ecosystems, such as forests
and grasslands.  Nitrogen deposition contributes to eutrophication of estuaries and coastal waters
which result in toxic algal blooms and fish kills.

       The addition of nitrogen to most ecosystems causes changes in primary productivity and
growth of plants and algae, which can alter competitive interactions among species.  Some
species grow more than others, leading to shifts in population dynamics, species composition,
and community structure.  The most extreme effects of nitrogen deposition include a shift of
ecosystem types in  terrestrial ecosystems, and hypoxic zones that are devoid of life in aquatic
ecosystems.8

3.3.1.1.1 Ecological Effects of A cidification
       As described in the INF paper, ambient air quality monitoring data collected from Puerto
Rico and the U.S. Virgin Islands between 2002 and 2007 indicate that wet deposition levels of
both sulfate and nitrate compounds are significant and elevated, especially for sulfate.F  The
principal factor governing the sensitivity of terrestrial and aquatic ecosystems to acidification
from nitrogen and sulfur deposition is geology  (particularly surficial geology).84 Geologic
formations having low base cation supply generally underlie the watersheds of acid-sensitive
lakes and streams.  Bedrock geology has been used in numerous acidification studies.85'86'87'88'89
Other factors contributing to the sensitivity of soils  and surface waters to acidifying deposition,
include: topography, soil chemistry, land use, and hydrologic flow path.

       Terrestrial

       Acidifying deposition can alter biogeochemical processes by increasing the nitrogen and
sulfur content of soils, accelerating nitrate and sulfate leaching from soil to drainage waters,
depleting base cations (especially calcium and magnesium) from soils, and increasing the
mobility of aluminum. Inorganic aluminum is toxic to some tree roots. Plants affected by high
levels of aluminum from the soil often have reduced root growth, which restricts the ability of
the plant to take up water and nutrients, especially calcium.90 These direct effects can, in turn,
influence the response of these plants to climatic stresses such as droughts and cold
temperatures.  They can  also influence the sensitivity of plants to other stresses, including insect
F The National Atmospheric Deposition Program (NADP)/National Trends Network operated by the University of
Illinois (http://nadp.sws.uiuc.edu) serves as the repository for annual data for wet deposition for the entire U.S.
including Puerto Rico and the U.S. Virgin Islands.


                                            3-15

-------
pests and disease91 leading to increased mortality of canopy trees. Emissions from ships
contribute to nitrogen and sulfur deposition levels and can therefore impact trees and forests.

       Lichens and bryophytes are among the first components of the terrestrial ecosystem to be
affected by acidifying deposition.92  There are over 1000 species of lichens known to occur in
Puerto Rico and related offshore islands.93  Vulnerability of lichens to increased nitrogen input is
generally greater than that of vascular plants.94 Effects of sulfur dioxide exposure to lichens
includes: reduced photosynthesis and respiration, damage to the algal component  of the lichen,
leakage of electrolytes, inhibition of nitrogen fixation, reduced K absorption, and  structural
changes.95 Additional research has concluded that the sulfur:nitrogen exposure ratio is as
important as pH in causing toxic effects on lichens.  Thus, it is not clear to what extent acidity
may be the principal stressor under high levels of air pollution exposure. The toxicity of sulfur
dioxide to several lichen species is greater under acidic conditions than under neutral
conditions.96 Emissions from ships contribute to nitrogen and sulfur deposition levels and can
therefore impact lichens.

       Aquatic Ecosystems

       Aquatic effects of acidification have been well studied at various trophic levels.  These
studies indicate that aquatic biota have been affected by acidification at virtually all levels of the
food web in acid sensitive aquatic ecosystems. Effects  have been most clearly documented for
fish, aquatic insects,  other invertebrates, and algae.

       Biological effects are primarily attributable to a combination of low pH and high
inorganic aluminum  concentrations. Such conditions occur more frequently during rainfall and
snowmelt that cause  high flows of water and less commonly during low-flow conditions, except
where  chronic acidity conditions are severe. Biological effects of episodes include reduced fish
condition factor, changes in species composition and declines  in aquatic species richness across
multiple taxa, ecosystems and regions. These conditions may also result in direct mortality.97
Biological effects in  aquatic ecosystems can be divided into two major categories: effects on
health, vigor, and reproductive success; and effects on biodiversity.
       Coral reef ecosystems in Puerto Rico and the U.S. Virgin Islands comprise diverse
habitats, including coral reefs, sea grass beds, and mangroves, that host abundant and diverse
marine organisms (Rohmann, 2005). These biologically rich ecosystems play an important role
in the socio-economic activities of coastal areas.  For example, the reef habitats support the
valuable fishing and tourism industries. However, the reef habitats are negatively impacted by
these industries, including emissions from ships including cruise ships.

        Complex reef ecosystems in Puerto Rico and the U.S. Virgin Islands with significant
amount of live coral have experienced steep declines in overall population and in coral species
                                           3-16

-------
(Waddell and Clarke, 2008).  As a result, the percentage of mean live hard coral cover0 today is
no greater than 10%.H Increases in C02, NOx and SOx emissions likely contribute to ocean
acidification which results in less available calcium carbonate for shell deposition and growth of
marine organisms. If this trend continues it may prevent future coral reef growth altogether and
result in the permanent alteration  of these important ecosystems.

       Coral Reefs in Puerto Rico

       Along with the main island of Puerto Rico, there are two uninhabited small islands off the
east coast (Culebra and Vieques), and three uninhabited islands (Mona, Monito, Desecheo) off
the west coast.  Most coral reefs occur on the east, south and west coasts of the main island, with
fringing reefs being the most common type. The western two-thirds of the north coast consists of
mainly hard ground and reef rock with low to very low coral cover and some small, sparse, low
coral colonies. Coral reefs cover approximately 3,370 km2 within three nautical miles of the
coasts. The main islands of Puerto Rico, including Culebra and Vieques, are almost completely
encircled by reefs, although coral reef abundance is highly variable, depending on the local
conditions.  Figure 3-1 shows the distribution of coral reefs in Puerto Rico as developed by the
U.S. National Oceanic and Atmospheric Administration (NOAA).
G Coral cover is a measure of the proportion of reef surface covered by live stony coral instead of sponges, algae, or
other organisms.
H NOAA's Healthy Reefs for Healthy People program defines coral cover levels of 10% or lower as 'red flags' and
recognizes a target level of 30% and above for reefs in the Mesoamerican Reef Region
(http://healthyreefs.org/healthy-reef-indicators/coral-cover.html)
                                           3-17

-------
                                                              Benthic Habitat
                                                              ^|  Coral Reef and Colonized Hardbottoin
                                                              ^|  Uncolonized Hardbottom
                                                              |  |  Submerged Vegetation
                                                              |  |  Unconsolidated Sediments
                                                              I  I  Other Delineations
                                                                I  Land
       Figure 3-1 Distribution and Extent of Coral Reef Ecosystems in Puerto Rico a
 Notes:
 a Map developed by NOAA's Center for Coastal Monitoring and Assessment, Biogeography Team (CCMA-
 BT) based on visual interpretation of aerial photography and hyperspectral imagers. For more information, see:
 http://biogeo.nos.noaa.gov.

       The Puerto Rico Coral Reef Ecosystem Monitoring Program monitors 12 reefs from six
Marine Preserve Areas (MPAs), and is sponsored by NOAA and has been administered by
Puerto Rico's Department of Natural and Environmental Resources (DNER) since 1999 (Garcia,
2008). The MPAs include reef sites at Isla Desecheo, Rincon, Mayaguez, Guanica, Isla Caja de
Muerto, and Ponce. Data from the program show that the benthic community at some of the reef
systems are experiencing decline - including decline in live coral cover as well as a general trend
of decline in the abundance of fish populations (statistically significant in seven of the 12 reef
stations surveyed).

       The declines in the health of key reef-building corals have become a concern to the U.S.
Government. In 2004, NOAA received a petition to protect Elkhorn (Acropora palmata),
Staghorn (A. cervicornis) and fused staghorn (A. prolifera) corals under the Endangered Species
Act (ESA) of 1973, as amended. NOAA found that petition had merit and convened a Biological
Review Team  (BRT) to review the status of these species.  Based on the results of the status
review, in 2006 NOAA's National Marine  Fisheries Service issued a final rule listing Elkhorn
and Staghorn corals as threatened throughout their known range.
                                           3-18

-------
        Coral Reefs in the U.S.  Virgin Islands

        In the U.S. Virgin Islands, coral reefs occur around all the major islands of St. Croix, St.
John, and St. Thomas, as well as the offshore bays, as depicted in Figure 3-2 below.  Fringing
reefs, deep reefs (wall and shelf-edge), patch reefs, and spur and groove formations are present,
although only St. Croix has barrier reefs. Bank reefs and scattered patch reefs with high coral
diversity occur deeper offshore. The U.S. Departments of Interior, and Commerce, and the
Virgin Islands Government have jurisdiction over submerged lands with coral reefs within the
U.S. Virgin Islands. In 2001, NOAA completed maps of U.S. Virgin Islands coral reefs and
associated ecosystems to a depth of 20 m. Of the 485 km2, 61% consisted of coral reefs and hard-
bottom habitats ,  33% were seagrass beds (labeled as submerged vegetation), and the rest was
sand or rock.
                                                                      Bent hi c H:ib tai
                                                                         Coral Reef and
                                                                         Colonized Hardbottom
                                                                         Uncolonized Kardbottom

                                                                         Submerged Vegetation
                                                                         UnconsoSidated Sediments

                                                                         Other Delineations

                                                                         Land
        Figure 3-2 Benthic Habitat Maps" - Distribution and Extent of Coral Reef Ecosystems in U.S. Virgin
                                   Islands (Rothenberger, 2008)
1 Sonar technology was used to generate the benthic habitat maps in Figures 2-1 and 2-2 and does not distinguish
whether or not coral reefs exist on hard-bottom substrate; nor does it distinguish live coral from denuded skeleton.
Hard-bottom substrate does not necessarily have corals on it nor does a reef necessarily exist where hard-bottom
substrate exists. Hard-bottom is the only substrate where coral reefs might exist (but don't necessarily exist).
                                              3-19

-------
    Notes:
    " Near shore benthic habitat maps were developed by NOAA's Center for Coastal Monitoring and
    Assessment, Biogeography Team (CCMA-BT) based on visual interpretation of aerial photography and
    hyperspectral imagers. For more information, see: http://biogeo.nos.noaa.gov.

    Coral reefs in the U.S. Virgin Islands have changed dramatically in the last three decades.
Insights into these changes come from long-term monitoring of sites ranging in depth from sea
level  to 40 m. Live coral cover has declined; coral diseases have become more numerous and
prevalent; macroalgal cover has increased; fish of some species are smaller, less numerous or
only rarely seen; and the long spined black sea urchins Diadema antillarum are less abundant.
Coral cover has declined on most if not all reefs in the U.S. Virgin Islands for which there are
quantitative data.  In the 1970s and 1980s coral cover on some reefs was over 40% and even
higher in some shallow Elkhorn coral zones (Gladfelter et al. 1977, Gladfelter 1982, Rogers et al.
1983, Edmunds 2002).  By the 1990s, many long-term monitoring sites had coral cover of about
25% or less, and macroalgal cover, although variable, often reached much higher values than in
the past. Coral cover continued to decline or remain stable until the major 2005 bleaching
/disease event.J Now coral cover is less than 12% on many reefs, including five long term study
sites in St. John and St. Croix covering over 10 ha of reefs that formerly had high coral cover and
diversity.

3.3.1.1.2  Ecological Effects of Nutrient Enrichment
       In general, ecosystems that are most responsive to nutrient enrichment from atmospheric
nitrogen deposition are those that receive high levels of nitrogen loading, are nitrogen-limited, or
contain species that have evolved in nutrient-poor environments.  Species that are adapted to low
nitrogen supply will often be more readily outcompeted by species that have higher nitrogen
demands when the availability of nitrogen is increased.98    too>101 AS a consequence, some
native species can be eliminated by nitrogen deposition.102'103'104'105 Note the terms "low" and
"high" are relative to the amount of bioavailable nitrogen in the ecosystem and the level of
deposition.

Terrestrial

       Ecological effects of nitrogen deposition occur in a variety of taxa and ecosystem types
including: forests, grasslands, arid and semi-arid areas, deserts, lichens, alpine, and mycorrhizae.
Atmospheric inputs of nitrogen can alleviate deficiencies and increase  growth of some plants at
the expense of others.  Nitrogen deposition alters the competitive relationships among terrestrial
plant species and therefore alters species composition and diversity.106'107'10  Wholesale shifts in
species composition are easier to detect in short-lived terrestrial ecosystems such as annual
grasslands, in the forest understory, or mycorrhizal associations, than for long-lived forest trees
J Coral bleaching is associated with a variety of stresses including increased sea surface temperatures. This causes
the coral to expel symbiotic micro-algae living in their tissues - algae that provide corals with food. Losing their
algae leaves coral tissues devoid of color, and thus appearing to be bleached. Prolonged coral bleaching (over a
week) can lead to coral death and the subsequent loss of coral reef habitats for a range of marine life. August 2005
saw the beginning of a record-breaking coral bleaching event throughout the Caribbean region. The U.S. Virgin
Islands were hit particularly hard: up to 95 percent of the corals bleached, and  some areas saw 40 percent of the
coral area killed.
                                            3-20

-------
where changes are evident on a decade or longer time scale. Note species shifts and ecosystem
changes can occur even if the ecosystem does not exhibit signs of nitrogen saturation.

       Most terrestrial ecosystems are nitrogen-limited, therefore they are sensitive to
perturbation caused by nitrogen additions.    The factors that govern the vulnerability of
terrestrial ecosystems to nutrient enrichment from nitrogen deposition include the degree of
nitrogen limitation, rates and form of nitrogen deposition, elevation, species composition, length
of growing season, and soil nitrogen retention capacity. Figure 3-3 below indicates some of the
terrestrial ecosystems located on Puerto Rico.
                                        PUERTO RICO
                                  Figure 3-3 Puerto Rico Ecozones
Freshwater Aquatic
       Nitrogen deposition alters species richness, species composition and biodiversity in
freshwater aquatic ecosystems.110  Evidence from multiple lines of research and experimental
approaches support this observation, including paleolimnological reconstructions, bioassays,
mesocosm and laboratory experiments.  Increased nitrogen deposition can cause a shift in
community composition and reduce algal biodiversity.

Wetland

       Given the relatively small size of Puerto Rico and the U.S. Virgin Islands, the acreage of
wetlands cannot be compared to those in the mainland United States. For instance wetlands in
the U.S. Virgin Islands are confined to narrow coastal fringes. Although small, these wetlands
are vital to migratory birds and native wildlife.111  Nitrogen deposition alters species richness,
species composition and biodiversity in  wetland ecosystems. The effect of nitrogen deposition
on these ecosystems depends on the fraction of rainfall in its total water budget. Excess nitrogen
                                          3-21

-------
deposition can cause shifts in wetland community composition by altering competitive
relationships among species, which potentially leads to effects such as decreasing biodiversity,
increasing non-native species establishment and increasing the risk of extinction for sensitive and
rare species.

Estuarine Aquatic

       Nitrogen deposition also alters species richness, species composition and biodiversity in
estuarine ecosystems.112  Nitrogen is an essential nutrient for estuarine and marine fertility.
However, excessive nitrogen contributes to habitat degradation, algal blooms, toxicity, hypoxia
(reduced dissolved oxygen), anoxia  (absence of dissolved oxygen), reduction of sea grass
habitats, fish kills, and decrease in biodiversity.113'114'115'116'  ^118  Each of these potential
impacts carries ecological and economic consequences.  Ecosystem services provided by
estuaries include fish and shellfish harvest, waste assimilation, and recreational activities.119
Increased nitrogen deposition can cause shifts in community composition, reduced hypolimnetic
DO, reduced biodiversity, and mortality of submerged aquatic vegetation. The form of deposited
nitrogen can significantly affect phytoplankton community composition in estuarine and marine
environments.

       Estuaries and coastal waters tend to be nitrogen-limited and are therefore inherently
                                                i 9fl 191
sensitive to increased atmospheric nitrogen loading.   '    The U.S. EPA issued the National
Estuary Program Coastal Condition Report (NEPCCR) in June 2007.122  The NEPCCR
concludes that 37% of estuaries in the National Estuary Program are in poor condition, including
Puerto Rico's San Juan Bay Estuary. This rating is based on four indicators of estuarine
condition - a water quality index, a sediment quality index, a benthic index  and a fish tissue
contaminants index.  The report notes that water quality is rated fair for San Juan Bay but that
one of the most common and widespread impairments to the estuary's waters are nutrient
enrichment/eutrophication.  The significant contribution by ships to emission inventories in
Puerto Rico and the U.S. Virgin Islands means that these ships also have a significant
contribution to nitrogen deposition levels which can contribute to nutrient enrichment and
eutrophication.

          Historically Puerto Rico had 60,000 acres of estuaries, 30,000 of which were
mangroves.123 Marshes and mangroves support a great variety of juvenile fish and  invertebrates
and provide food and nesting habitat for many different bird species. The preservation of marsh
and mangrove habitats is an objective in  the San Juan Bay Management Plan.124 The NEPCCR
also includes information on Puerto Rico's ecosystems of seagrass and submerged aquatic
vegetation, some of the most diverse ecosystems in the North Atlantic Ocean.

3.3.1.2 Deposition of Particulate Matter

       Current international shipping emissions of PM2.s contain small amounts of metals:
nickel, vanadium, cadmium, iron, lead, copper, zinc, aluminum.125'126'127  Investigations of trace
metals near roadways and industrial facilities indicate that a substantial burden of heavy metals
can accumulate on vegetative surfaces.  Copper, zinc, and nickel are shown  to be directly toxic to
vegetation under field conditions.128  While metals typically exhibit low solubility,  limiting their
bioavailability and direct toxicity, chemical transformations of metal compounds occur in the
                                           3-22

-------
environment, particularly in the presence of acidic or other oxidizing species.  These chemical
changes influence the mobility and toxicity of metals in the environment.  Once taken up into
plant tissue, a metal compound can undergo chemical changes, accumulate and be passed along
to herbivores or can re-enter the soil and further cycle in the environment.

       Ships also emit air toxics, including polycyclic aromatic hydrocarbons (PAHs) -- a class
of polycyclic organic matter (POM)  that contain compounds which are known or suspected
carcinogens. Since the majority of PAHs are adsorbed onto particles less than  1.0 ^im in
diameter, long range transport is possible. Particles of this size can remain airborne for days or
even months and travel distances up to 10,000 km before being deposited on terrestrial or aquatic
surfaces.129'130'131'132'133 PAHs tend to accumulate in sediments and reach high enough
concentrations in some coastal environments to pose an environmental health threat that includes
cancer in fish populations, toxicity to organisms living in the sediment and risks to those (e.g.,
migratory birds) that consume these  organisms.134'135  PAHs tend to accumulate in sediments
and bioaccumulate in freshwater, flora and fauna.

       The effects of the deposition of atmospheric pollution, including ambient PM, on
materials are related to both physical damage and impaired  aesthetic qualities. The deposition of
PM (especially sulfates and nitrates) can physically affect materials, adding to the effects of
natural weathering processes, by potentially promoting or accelerating the corrosion of metals,
by degrading paints, and by deteriorating building materials such as concrete and limestone.
Only chemically active fine particles or hygroscopic coarse particles contribute to these physical
effects. In addition, the deposition of ambient PM can reduce the aesthetic appeal of buildings
and culturally important articles through soiling. Particles consisting primarily of carbonaceous
compounds cause soiling of commonly used building materials and culturally important items
such as statues and works of art.

3.3.2 Impacts of Ozone on Plants and Ecosystems

       There are a number of environmental or public welfare effects associated with the
presence of ozone in the ambient air.136 Ship emissions of NOx contribute  to ambient ozone
concentrations in Puerto Rico and the U.S. Virgin Islands.  In this section we discuss the impact
of ozone on plants, including trees, agronomic crops and urban ornamentals.

       The Air Quality Criteria Document for Ozone and related Photochemical Oxidants notes
that "ozone affects vegetation throughout the United States, impairing crops, native vegetation,
and ecosystems more than any other air pollutant".137 Like carbon dioxide (COz) and other
gaseous substances, ozone enters plant tissues primarily through apertures (stomata) in leaves in
a process called "uptake".138 Once sufficient levels of ozone, a highly reactive substance, (or its
reaction products) reaches the interior of plant cells, it can inhibit or damage essential cellular
components and functions, including enzyme activities, lipids, and cellular membranes,
disrupting the plant's osmotic  (i.e., water) balance and energy utilization patterns.139'140  If
enough tissue becomes damaged from these effects, a plant's capacity to fix carbon to form
carbohydrates, which are the primary form of energy used by plants is reduced,141 while plant
respiration increases. With fewer resources available, the plant reallocates  existing resources
away from root growth and storage,  above ground growth or yield, and reproductive processes,
toward leaf repair and maintenance,  leading to reduced growth and/or reproduction. Studies
                                          3-23

-------
have shown that plants stressed in these ways may exhibit a general loss of vigor, which can lead
to secondary impacts that modify plants' responses to other environmental factors. Specifically,
plants may become more sensitive to other air pollutants, more susceptible to disease, insect
attack, harsh weather (e.g., drought, frost) and other environmental stresses.  Furthermore, there
is evidence that ozone can interfere with the formation of mycorrhiza, essential symbiotic fungi
associated with the roots of most terrestrial plants, by reducing the amount of carbon available
for transfer from the host to the symbiont.14 '143

       This ozone damage may or may not be accompanied by visible injury on leaves, and
likewise, visible foliar injury may or may not be a symptom of the other types of plant damage
described above.  When visible injury is present, it is commonly manifested as chlorotic or
necrotic spots, and/or increased leaf senescence (accelerated leaf aging).  Because ozone damage
can consist of visible injury to leaves, it can also reduce the aesthetic value of ornamental
vegetation and trees in urban landscapes, and negatively affects scenic vistas in protected natural
areas.

       Ozone can produce both acute and chronic injury in sensitive species depending on the
concentration level and the duration of the exposure. Ozone effects also tend to accumulate over
the growing season of the plant, so that even lower concentrations experienced for a longer
duration have the potential to create chronic stress on sensitive vegetation. Not all plants,
however, are equally sensitive to ozone.  Much of the variation in sensitivity between individual
plants or whole species is related to the plant's ability to regulate the extent of gas exchange via
leaf stomata (e.g., avoidance of ozone uptake through closure of stomata)144'14^146  Other
resistance mechanisms may involve the intercellular production of detoxifying substances.
Several biochemical substances capable  of detoxifying ozone have been reported to occur in
plants, including the antioxidants ascorbate and glutathione. After injuries have occurred, plants
may be capable of repairing the damage  to a limited extent.147

       Because of the differing sensitivities among plants to ozone, ozone pollution can also
exert a selective pressure that leads to changes in plant community composition.  Given the range
of plant sensitivities and the fact that numerous other environmental factors modify plant uptake
and response to ozone, it is not possible to identify threshold values above which ozone is
consistently toxic for all plants. The next few paragraphs present additional information on
ozone damage to trees, ecosystems, agronomic crops and urban ornamentals.

       Ozone also has been conclusively shown to cause discernible injury to forest trees.148'149
In terms of forest productivity and ecosystem diversity, ozone may be the pollutant with the
greatest potential  for regional-scale forest impacts. Studies have demonstrated repeatedly that
ozone concentrations commonly observed in polluted areas can have substantial impacts on plant
function.150'151

       Because plants are at the base of the food web in many ecosystems, changes to the plant
community can affect associated organisms and ecosystems (including the suitability of habitats
that support threatened or endangered species and below ground organisms living in the root
zone).  Ozone impacts at the community and ecosystem level  vary widely depending upon
numerous factors, including concentration and temporal variation of tropospheric ozone, species
composition, soil  properties and climatic factors.15  In most instances, responses to chronic or
                                           3-24

-------
recurrent exposure in forested ecosystems are subtle and not observable for many years. These
injuries can cause stand-level forest decline in sensitive ecosystems.153'154'155 It is not yet
possible to predict ecosystem responses to ozone with much certainty; however, considerable
knowledge of potential ecosystem responses has been acquired through long-term observations
in highly damaged forests in the United States.

       Laboratory and field experiments have also shown reductions in yields for agronomic
crops exposed to ozone, including vegetables (e.g., lettuce)  and field crops (e.g., cotton and
wheat). The most extensive field experiments, conducted under the National Crop Loss
Assessment Network (NCLAN) examined 15 species and numerous cultivars. The NCLAN
results show that "several economically important crop species are sensitive to ozone levels
typical of those found in the United States."156  In addition, economic studies have shown
reduced economic benefits as a result of predicted reductions in crop yields associated with
observed ozone levels.157'158'159

       Urban ornamentals represent an additional vegetation category likely to experience some
degree of negative effects associated with exposure to ambient ozone levels.  It is estimated that
more than $20 billion (1990 dollars) are spent annually on landscaping using ornamentals, both
by private property owners/tenants and by governmental units responsible for public areas.160
This is therefore a potentially costly environmental effect. However, in the absence of adequate
exposure-response functions and economic damage functions for the potential range of effects
relevant to these types of vegetation, no direct quantitative analysis has been conducted.

       Air pollution can have noteworthy cumulative impacts on forested ecosystems by
affecting regeneration, productivity, and species composition.161 In the U.S., ozone in the lower
atmosphere is one of the pollutants of primary concern. Ozone injury to forest plants can be
diagnosed by examination of plant leaves. Foliar injury is usually the first visible sign of injury
to plants from ozone exposure and indicates impaired physiological processes in the leaves.162

       This indicator is based on data from the  U.S. Department of Agriculture (USDA) Forest
Service Forest Inventory and Analysis (FIA) program.  As part of its Phase 3 program, formerly
known as Forest Health Monitoring, FIA examines ozone injury to ozone-sensitive plant species
at ground monitoring sites in forest land across the country. For this indicator, forest land does
not include woodlots and urban trees. Sites are  selected using a systematic sampling grid, based
on a global sampling design.163'164 At each site that has at least 30 individual plants of at least
three ozone-sensitive species and enough open space to ensure that sensitive plants are not
protected from ozone exposure by the forest canopy,  FIA looks for damage on the foliage of
ozone-sensitive forest plant species. Because ozone injury is cumulative over the course of the
growing season, examinations are conducted in July and August, when ozone injury is typically
highest.

       Monitoring of ozone injury to plants by the USDA Forest Service has expanded over the
last 10 years from monitoring sites in ten states  in 1994 to nearly 1,000 monitoring sites in 41
states in 2002. The data underlying this indicator are based on averages of all observations
collected in 2002, the latest year for which data are publicly available at the time the study was
conducted, and are broken down by EPA Region.  Ozone damage  to forest plants  is classified
using a subjective five-category biosite index based on expert opinion, but designed to be
                                          3-25

-------
equivalent from site to site. Ranges of biosite values translate to no injury, low or moderate
foliar injury (visible foliar injury to highly sensitive or moderately sensitive plants, respectively),
and high or severe foliar injury, which would be expected to result in tree-level or ecosystem-
level responses, respectively.    '166

3.3.3 Visibility

       Good visibility increases the quality of life where individuals live and work, and where
they engage in recreational activities. Airborne particles degrade visibility by scattering and
absorbing light.  Ship emissions of primary PM2.5 and SOx and NOx (which contribute to the
formation of secondary PlV^.s) contribute to poor visibility in Puerto Rico and the U.S. Virgin
Islands.

       The U.S. Government places special emphasis on protecting visibility in national parks
and wilderness areas. Section 169 of the Clean Air Act requires  the U.S. Government to address
existing visibility impairment and future visibility impairment in the 156 national parks and
wilderness areas which are categorized as Mandatory Class I Federal areas. Virgin Islands
National Park is a Mandatory Class I Federal area. The national  park covers over 5,900 hectares,
approximately 60% of the island of Saint John in the U.S. Virgin Islands, plus a few isolated
sites on the neighboring island of St. Thomas.

       Studies done  for the continental U.S. have shown that ship emissions contribute to sulfate
particles, which degrade visibility in Mandatory Class I Federal areas. For instance, one study
concluded that shipping and port emissions from the Pacific Coast showed significant
contributions to atmospheric sulfate concentrations over large areas of the western U.S. and that
reducing those emissions is important in controlling haze at Mandatory Class I Federal areas.167

       The emissions reductions associated with this proposed EGA would improve visibility in
Puerto Rico and the U.S.  Virgin Islands as a whole, as well as in sensitive areas  such as the
Virgin Islands National Park.

3.3.3.1  Visibility Monitoring

       In conjunction with the U.S. National Park Service, the U.S. Forest Service, other federal
land managers, and State organizations in the U.S., the U.S.  EPA has supported  visibility
monitoring in national parks and wilderness areas since 1988. The monitoring network was
originally established at 20 sites, but it has now been expanded to 110 sites that represent all but
one of the 156 Mandatory Federal Class I areas across the country. This long-term visibility
monitoring network is known as IMPROVE (Interagency Monitoring of PROtected Visual
Environments).

       Annual mean deciview levels and natural haze (or background levels of visibility that
would occur without manmade air pollution) levels for the Virgin Islands National Park are
available on the IMPROVE website. Figure 2-4 below presents annual mean deciview data from
2001-2008 alongside natural haze levels.  The EGA emission reductions being proposed here
will help Virgin  Islands National Park to reach natural haze levels.
                                           3-26

-------
                   Virgin Islands National Park IMPROVE Data
                                                                      • annual mean
                                                                      O background
           2001
2002
2003
2004
2005
2006
2008
Note:
                  Figure 3-4 Virgin Islands National Park IMPROVE Data
Data from http://vista.cira.colostate.edu/improve/Data/IMPROVE/summary_data.htm
                                    3-27

-------
1 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009.

2 U.S. EPA. (2006). National-Scale Air Toxics Assessment for 1999.  This material is available electronically at
http://www.epa.gov/ttn/atw/natal999/.

3 Agrawal, H., Malloy, Q.G.J., Welch, W.A., Miller, J.W., Cocker, D.R. (2008). In-use gaseous and paniculate
matter emissions from a modern ocean going container vessel. Atmospheric Environment, 42,  5504-5510.

4 Hu, S., Polidori, A., Arhami, M., Shafer, M.M., Schauer, J.J., Cho, A., Sioutas, C. (2008). Redox activity and
chemical speciation of size fractionated PM in the communities of the Los Angeles-Long Beach Harbor.
Atmospheric Chemistry and Physics Discussions, 8,  11683-11672.

5 U.S. EPA (2002). Health Assessment Document for Diesel Engine Exhaust. EPA/600/8-90/057F Office of
Research and Development, Washington DC. Retrieved on March 17, 2009 from
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060.  pp. 1-1 1-2.

6 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.1.1.

7 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.1.2.

8 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.4.

9 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Table 2-6.

10 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.  Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Section 2.3.5.1.

11 U.S. EPA (2009). Integrated Science Assessment for Particulate Matter (Final Report). U.S.  Environmental
Protection Agency, Washington, DC, EPA/600/R-08/139F, 2009. Table 2-6.

12 U.S. EPA. (2006;. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC:  U.S. EPA.

13 U.S. EPA. (2007;. Review of the National Ambient Air Quality Standards for Ozone: Policy Assessment of
Scientific and Technical Information, OAQPS Staff Paper. EPA-452/R-07-003. Washington, DC, U.S. EPA.

"National Research Council (NRC), 2008. Estimating Mortality Risk Reduction and Economic Benefits from
Controlling Ozone Air Pollution. The National Academies Press:  Washington, D.C.
                                                 3-28

-------
15Bates, D.V., Baker-Anderson, M, Sizto, R. (1990). Asthma attack periodicity: a study of hospital emergency
visits in Vancouver. Environ. Res., 57,51-70.

16Thurston, G.D., Ito, K., Kinney, P.L., Lippmann, M. (1992). A multi-year study of air pollution and respiratory
hospital admissions in three New York State metropolitan areas: results for 1988 and 1989 summers. J. Exposure
Anal. Environ. Epidemiol, 2,429-450.

17Thurston, G.D., Ito, K., Hayes, C.G., Bates, D.V., Lippmann, M. (1994) Respiratory hospital admissions and
summertime haze air pollution in Toronto, Ontario: consideration of the role of acid aerosols. Environ. Res., 65,
271-290.

18Lipfert, F.W., Hammerstrom, T. (1992). Temporal patterns in air pollution and hospital admissions. Environ. Res.,
59,374-399.

19Burnett, R.T., Dales, R.E., Raizenne, M.E., Krewski, D., Summers, P.W., Roberts, G.R., Raad-Young, M.,
Dann,T., Brook, J.  (1994). Effects of low ambient levels of ozone and sulfates on the frequency of respiratory
admissions to Ontario hospitals. Environ. Res., 65, 172-194.

20 U.S. EPA. (2006;. Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA.

21 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA.

22Devlin, R. B., McDonnell, W. F., Mann, R., Becker, S., House, D. E., Schreinemachers, D., Koren, H. S. (1991).
Exposure of humans to ambient levels of ozone for 6.6 hours causes cellullar and biochemical changes in the lung.
Am. J. Respir. CellMol. Biol, 4,  72-81.

23 Koren, H.  S., Devlin, R. B., Becker, S., Perez, R., McDonnell, W. F. (1991). Time-dependent changes of markers
associated with inflammation in the lungs of humans exposed to ambient levels of ozone. Toxicol. Pathol,  19, 406-
411.

24Koren, H.  S., Devlin, R. B., Graham, D. E., Mann, R., McGee, M. P., Horstman, D. H., Kozumbo, W. J., Becker,
S., House, D. E., McDonnell, W. F., Bromberg, P. A. (1989). Ozone-induced inflammation in the lower airways of
human subjects. Am. Rev. Respir. Dis., 39, 407-415.

25Schelegle, E.S., Siefkin, A.D., McDonald, R.J. (1991). Time course of ozone-induced neutrophilia in normal
humans. Am. Rev. Respir. Dis., 143,1353-1358.

26 U.S. EPA. (1996). Air Quality Criteria for Ozone and Related Photochemical Oxidants. EPA600-P-93-004aF.
Washington. D.C.:  U.S. EPA.

27Hodgkin, J.E., Abbey, D.E., Euler, G.L., Magie, A.R. (1984). COPD  prevalence in nonsmokers in high and  low
photochemical air pollution areas. Chest, 86, 830-838.

28Euler, G.L., Abbey, D.E., Hodgkin, J.E., Magie, A.R. (1988). Chronic obstructive pulmonary disease symptom
effects of long-term cumulative exposure to ambient levels of total oxidants and nitrogen dioxide in California
Seventh-day Adventist residents. Arch. Environ. Health, 43, 279-285.
                                                 3-29

-------
29 Abbey, D.E., Petersen, F., Mills, P.K., Beeson, W.L. (1993). Long-term ambient concentrations of total
suspended particulates, ozone, and sulfur dioxide and respiratory symptoms in a nonsmoking population.  Arch.
Environ. Health, 48, 33-46.

30 U.S. EPA. (2007). Review of the National Ambient Air Quality Standards for Ozone: Policy Assessment of
Scientific and Technical Information, OAQPS Staff Paper. EPA-452/R-07-003. Washington, DC, U.S. EPA.

31 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA.

32 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA.

33 Avol, E.L., Trim, S. C., Little, D.E., Spier, C.E.,  Smith, M. N., Peng, R.-C., Linn, W.S., Hackney, J.D.,  Gross,
K.B., D'Arcy, J.B., Gibbons, D., Higgins, I.T.T. (1990 June). Ozone exposure and lung function in children
attending a southern California summer camp.  Paper no. 90-150.3. Paper presented at the 83rd annual meeting and
exhibition of the Air & Waste Management Association, Pittsburgh, PA.

34 Higgins, I. T.T., D'Arcy, J. B., Gibbons, D. I., Avol, E. L., Gross, K.B. (1990). Effect of exposures to ambient
ozone on ventilatory lung function in children. Am. Rev. Respir. Dis., 141, 1136-1146.

35Raizenne, M.E., Burnett, R.T., Stern, B., Franklin, C.A., Spengler, J.D. (1989) Acute lung function responses to
ambient acid aerosol exposures in children. Environ. Health Perspect.,  79,179-185.

36 Raizenne, M.: Stern, B.: Burnett, R.; Spengler, J. (1987 June) Acute respiratory function and transported air
pollutants: observational studies. Paper no. 87-32.6. Paper presented at the 80th annual meeting of the Air Pollution
Control Association, New York, NY.

37 Spektor, D. M., Lippmann, M. (1991). Health effects of ambient ozone on healthy children at a summer camp. In:
Berglund, R. L.:  Lawson, D. R.; McKee, D. J.,  eds. Tropospheric ozone and the environment: papers from an
international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management Association:
pp. 83-89.  (A&WMA transaction series no. TR-19).

38Spektor, D. M., Thurston, G.D., Mao, J., He, D., Hayes, C., Lippmann, M.  (1991).  Effects of single- and multiday
ozone exposures on respiratory function in active normal children. Environ. Res, 55,107-122.

39Spektor, D. M., Lippman, M., Lioy, P. J., Thurston, G. D., Citak, K., James, D. J., Bock, N., Speizer, F. E., Hayes,
C. (1988). Effects of ambient ozone on respiratory function in active, normal children. Am. Rev. Respir. Dis., 137,
313-320.

40 U.S. EPA. (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). EPA/600/R-
05/004aF-cF. Washington, DC: U.S. EPA.

41 Hazucha, M. J., Folinsbee, L. J., Seal, E., Jr.  (1992). Effects of steady-state  and variable ozone concentration
profiles on pulmonary function. Am. Rev. Respir. Dis., 146, 1487-1493.

42Horstman, D.H., Ball, B.A., Folinsbee, L.J., Brown, J., Gerrity, T. (1995) Comparison of pulmonary responses of
asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone.  Toxicol. Ind.
Health., 11(4), 369-85.
                                                  3-30

-------
43Horstman, D.H.,; Folinsbee, L.J., Ives, P.J., Abdul-Salaam, S., McDonnell, W.F. (1990). Ozone concentration and
pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10, and
0.12 ppm.Am. Rev. Respir. Dis., 142, 1158-1163.

44 U.S. EPA (2008). Integrated Science Assessment (ISA) for Sulfur Oxides - Health Criteria (Final Report).
EPA/600/R-08/047F. Washington, DC,: U.S.EPA.

4 U.S. EPA (2008). Integrated Science Assessment for Oxides of Nitrogen - Health Criteria (Final Report).
EPA/600/R-08/071. Washington, DC,: U.S.EPA.

46IPCS (1996) Environmental Health Criteria 171: Diesel Fuel and Exhaust Emissions, World Health Organization,
Geneva, 1996.

47 U.S. EPA (2002). Health Assessment Document for Diesel Engine Exhaust. EPA/600/8-90/057F Office of
Research and Development, Washington DC.

48 U.S. EPA. (1999). Guidelines for Carcinogen Risk Assessment.  Review Draft. NCEA-F-0644, July. Washington,
DC: U.S. EPA.
49
  U.S. EPA. (1986). Guide lines for carcinogen risk assessment. EPA/630/R-00/004. Washington, DC: U.S. EPA.
50 National Institute for Occupational Safety and Health (NIOSH). (1988).  Carcinogenic effects of exposure to
diesel exhaust. NIOSH Current Intelligence Bulletin 50. DHHS (NIOSH) Publication No. 88-116. Atlanta, GA:
Centers for Disease Control.

51 International Agency for Research on Cancer - IARC. (1997). Silica, some silicates, coal dust and para-aramid
fibrils. In Monographs on the evaluation of carcinogenic risks to humans. Vol. 68. Lyon, France: IARC, pp. 362-
375.

52 National Institute for Occupational Safety and Health (NIOSH). (1988).  Carcinogenic effects of exposure to
diesel exhaust. NIOSH Current Intelligence Bulletin 50. DHHS (NIOSH) Publication No. 88-116. Atlanta, GA:
Centers for Disease Control.

53 World Health Organization International Program on Chemical Safety (1996). Diesel fuel and exhaust emissions.
In Environmental Health Criteria Vol.  171. Geneva: World Health Organization.

54 California Environmental Protection Agency (Cal EPA, OEHHA). (1998). Health risk assessment for diesel
exhaust.  Public and Scientific Review Draft. Sacramento, CA: Cal EPA.

55 National Toxicology Program (NTP). (2000). 9th Report on Carcinogens. Research Triangle Park,  NC: Public
Health Service, U.S. Department of Health and Human Services. Available from: http://ntp-server.niehs.nih.gov.

 6 Health Effects Institute (HEI).  (1995). Diesel exhaust: a critical analysis of emissions, exposure, and health
effects. A Special Report of the Institute's Diesel Working Group. Cambridge, MA: Health Effects Institute.

 7 Health Effects Institute (HEI).  (1999). Diesel emissions  and lung cancer: epidemiology and quantitative risk
assessment.  A special report of  the Institute's Diesel Epidemiology Expert Panel. Cambridge, MA: Health
Effects Institute (HEI).
                                                 3-31

-------
58 Health Effects Institute (HEI).  (2002). Research directions to improve estimates of human exposure and risk
assessment.  A special report of the Institute's Diesel Epidemiology Working Group. Cambridge, MA: Health
Effects Institute.

59 U.S. EPA (2002). Health Assessment Document for Diesel Engine Exhaust. EPA/600/8-90/057F Office of
Research and Development, Washington DC.
60
  Bhatia, R., Lopipero, P., Smith, A. (1998). Diesel exposure and lung cancer. Epidemiology, 9(1), 84-91.
61 Lipsett, M. Campleman, S. (1999). Occupational exposure to diesel exhaust and lung cancer: a meta-analysis. Am
JPublic Health, 80(7), 1009-1017.

62 U.S. EPA (2002), National-Scale Air Toxics Assessment for 1996.  This material is available electronically at
http://www.epa.gov/ttn/atw/nata/.

63 U.S. EPA. (2006). National-Scale Air Toxics Assessment for 1999. This material is available electronically at
http://www.epa.gov/ttn/atw/natal999/.

64 Ishinishi, N. Kuwabara, N. Takaki, Y., et al. (1988). Long-term inhalation experiments on diesel exhaust. In:
Diesel exhaust and health risks. Results of the HERP studies. Ibaraki, Japan: Research Committee for HERP
Studies: pp.11-84.

65 Heinrich, U., Fuhst, R., Rittinghausen, S., et al. (1995). Chronic inhalation exposure of Wistar rats and two
different strains of mice to diesel engine exhaust, carbon black, and titanium dioxide. Inhal Toxicol, 7, 553-556.

66 Mauderly, J.L., Jones,  R.K., Griffith, W.C., et al. (1987). Diesel exhaust is a pulmonary carcinogen in rats
exposed chronically by inhalation. Fundam. Appl. Toxicol., 9, 208-221.

67 Nikula, K.J., Snipes, M.B., Barr, E.B., et al. (1995). Comparative pulmonary toxicities and carcinogenicities of
chronically inhaled diesel exhaust and carbon black in F344 rats. Fundam. Appl.  Toxicol, 25,80-94.

68 U.S. EPA (2002). Health Assessment Document for Diesel Engine Exhaust. EPA/600/8-90/057F Office of
Research and Development, Washington DC. Retrieved on March 17, 2009 from
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060.  p. 9-9.

69 Reger, R., Hancock, J., Hankinson, J., et al. (1982). Coal miners exposed to diesel exhaust emissions. Ann Occup
Hyg, 26,  799-815.

70 Attfield, MD. (1978). The effect of exposure to silica and diesel exhaust in underground metal and nonmetal
miners. In Kelley, W.D., (ed.), Industrial hygiene for mining and tunneling: proceedings of a topical symposium;
Cincinnati, OH: The American Conference of Governmental Industrial Hygienists, Inc.

71 Wade, J.F., III, Newman, L.S. (1993) Diesel asthma: reactive airways disease following overexposure to
locomotive exhaust. J OccupMed, 35, 149-154.

72 U.S. EPA. (2007). Chapter 3: Air Quality and  Resulting Health and Welfare Effects of Air Pollution from Mobile
Sources.  In Mobile Source Air Toxics Rule (Control of Hazardous Air Pollutants from Mobile Sources: 72 FR 8428,
February 26, 2007) Regulatory Impact Analysis. Retrieved March 19, 2009 from
http://www.epa.gov/otaq/regs/toxics/420r07002.pdf.
                                                  3-32

-------
73 State of California Air Resources Board. (2009 March). Rail Yard Health Risk Assessments and Mitigation
Measures. Retrieved March 19, 2009 from http://www.arb.ca.gov/railyard/hra/hra.htm.

74 Di, P., Servin, A., Rosenkranz, K., Schwehr, B., Tran, H., (2006). Diesel Particulate Matter Exposure Assessment
Study for the Ports of Los Angeles and Long Beach. Sacramento, CA: California EPA, California Air Resources
Board (CARB). Retrieved March 19, 2009 from http://www.arb.ca.gov/regact/marine2005/portstudy0406.pdf.

75 Di, P., Servin, A., Rosenkranz, K., Schwehr, B., Tran, H., (2006). Diesel Particulate Matter Exposure Assessment
Study for the Ports of Los Angeles and Long Beach. Sacramento, CA: California EPA, California Air Resources
Board (CARB). Retrieved March 19, 2009 from http://www.arb.ca.gov/regact/marine2005/portstudy0406.pdf.

76ICF International. September 28,  2007. Estimation of diesel particulate matter concentration isopleths for marine
harbor areas and rail yards.  Memorandum to EPA under Work Assignment Number 0-3, Contract Number EP-C-
06-094.  This memo is available in Docket EPA-HQ-OAR-2007-0121.

77 ICF International. September 28,  2007. Estimation of diesel particulate matter population exposure near selected
harbor areas and rail yards.  Memorandum to EPA under Work Assignment Number 0-3, Contract Number EP-C-
06-094.  This memo is available in Docket EPA-HQ-OAR-2007-0121.

78 ICF International. December 1, 2008. Estimation of diesel particulate matter concentration isopleths near selected
harbor areas with revised emissions (revised). Memorandum to EPA under Work Assignment Number 1-9, Contract
Number EP-C-06-094. This memo  is available in Docket EPA-HQ-OAR-2007-0121.

79 ICF International. December 10, 2008.  Estimation of diesel particulate matter population exposure near selected
harbor areas with revised harbor emissions (revised). Memorandum to EPA under Work Assignment Number 2-9,
Contract Number EP-C-06-094. This memo  is available in Docket EPA-HQ-OAR-2007-0121.

80U.S. EPA (2004, October). Air Quality Criteria for Particulate Matter (Final Report) U.S. Environmental
Protection Agency, Washington, DC, EPA 600/P-99/002aF-bF.

81 Gao, Y., E.D. Nelson, M.P. Field, etal. (2002). Characterization of atmospheric trace elements on PM2.5
particulate matter over the New York-New Jersey harbor estuary. Atmos. Environ. 36: 1077-1086.

82 U.S. EPA (2004, October).  Air Quality Criteria Document for Particulate Matter. (Final Report), U.S.
Environmental Protection Agency, Washington, DC, EPA 600/P-99/002aF-bF.

83U.S. EPA (2008). Nitrogen  Dioxide/Sulfur Dioxide Secondary NAAQS Review: Integrated Science Assessment
(ISA).(Final). U.S. EPA, Washington D.C., EPA/600/R-08/082F.

84 U.S. EPA (2008). Nitrogen Dioxide/Sulfur Dioxide Secondary NAAQS Review: Integrated Science Assessment
(Z&i)(Final). U.S. EPA, Washington D.C., EPA/600/R-08/082F.

85 Bricker OP: Rice KC (1989). Acidic deposition to streams: a geology-based method predicts their sensitivity.
Environ Sci Technol, 23, 379-385.

86 Stauffer RE (1990). Granite weathering and the sensitivity of alpine lakes to acid deposition. Limnol Oceanogr,
35(5), 1112-1134.
                                                 3-33

-------
87 Stauffer RE; Wittchen BD (1991). Effects of silicate weathering on water chemistry in forested, upland, felsic
terrain of the USA. Geochim Cosmochim Acta, 55, 3253-3271.

88 Vertucci FA; Eilers JM (1993). Issues in monitoring wilderness lake chemistry: a case study in the Sawtooth
Mountains, Idaho. Environ MonitAssess, 28, 277-294.

89 Sullivan TJ; Webb JR; Snyder KU; Herlihy AT; Cosby BJ (2007). Spatial distribution of acid-sensitive and acid-
impacted streams in relation to watershed features in the southern Appalachian mountains. Water Air Soil Pollut,
182,57-71.

90 U.S. EPA (2008). Nitrogen Dioxide/Sulfur Dioxide Secondary NAAQSReview: Integrated Science Assessment
(ISA)  (Final).  U.S. EPA, Washington D.C., EPA/600/R-08/082F.

91 Joslin JD; Kelly JM; van Miegroet H (1992). Soil chemistry and nutrition of North American spruce-fir stands:
evidence for recent change. JEnviron Quality, 21, 12-30.

92 U.S. EPA (2008). Nitrogen Dioxide/Sulfur Dioxide Secondary NAAQS Review: Integrated Science Assessment
(ISA)  (Final).  U.S. EPA, Washington D.C., EPA/600/R-08/082F.

93 Mercado-Diaz J (2009), Preliminary checklist of lichens and lichenicolous fungi of Puerto Rico.  Available online
at: http://lichensofpuertorico.herbario.upr.edu/lichensofpuertorico.pdf

94 Fremstad E; Paal J; Mols T (2005). Impacts of increased nitrogen supply on Norwegian lichen-rich alpine
communities: A 10 year experiment. JEcol, 93, 471-481.

95 Farmer AM; Bates JW; Bell JNB (1992). Ecophysiological effects of acid rain on bryophytes and lichens. In:
Bates JW; Farmer AM (Eds.), Bryophytes and lichens in a changing environment. Oxford, UK: Claredon Press; and
Fields RF. (1988). Physiological responses of lichens to air pollutant fumigations. In: Nash TH III; Wirth V (Eds.),
Lichens, bryophytes and air quality. Volume 30; Bibl. Lichenol. (pp 175-200). Berlin/ Stuttgart, Germany: Cramer
Publisher.

96 Scott MG; Hutchinson TC; Feth MJ (1989). A comparison of the effects on Canadian boreal forest lichens of
nitric  and sulfuric acids as sources of rain acidity. New Phytol, 111, 663-671.; Scott MG; Hutchinson TC; Feth MJ.
(1989b). Contrasting responses of lichens and Vaccinium angustifolium to long-term acidification of a boreal forest
ecosystem. CanJBot, 67, 579-588.

97 Van Sickle J; Baker JP; Simonin HA; Baldigo BP; Kretser WA; Sharpe WE (1996). Episodic acidification of
small streams in the northeastern United States: Fish mortality in field bioassays. EcolAppl, 6, 408-421.

98 Aerts  R. (1990). Nutrient use efficiency in evergreen and deciduous species from heathland. Oecologia, 84, 391-
397.

99 Aerts  R; Berendse F;  De Caluwe H; Schmits M (1990). Competition in heathland along an experimental gradient
of nutrient availability. Oikos, 57, 310-318.

100 Krupa SV  (2003). Effects of atmospheric ammonia (NH3).on terrestrial vegetation: A review. Environ Pollut,
124,179-221.
                                                  3-34

-------
101 Tilman D; Wedin D. (1991). Dynamics of nitrogen competition between successional grasses. Ecology, 72, 1038-
1049.

102 Ellenberg H. (1985). Veranderungen der floa mitteleuropas unter dem einfluss von diingung und immissionen.
Schweiz Z Forstwesten, 136, 19-39.

103 Falkengren-Grerup U (1986). Soil acidification and vegetation changes in deciduous forest in southern Sweden.
Oecologia, 70, 339-347. Falkengren-Grerup U. (1989). Soil acidification and its impact on ground vegetation.
Ambio, 18, 179-183.

104 Roelofs JGM. (1986). The effect of airborne sulfur and nitrogen deposition on aquatic and terrestrial heathland
vegetation. Experientia, 42, 372-377.

105 Stevens CJ; Dise NB; Mountford OJ; Cowing DJ (2004). Impact of nitrogen deposition on the species richness of
grasslands. Science, 303, 1876-1878.

106 Ellenberg H (1987). Floristic changes due to eutrophication. In Asman WAH; Diederen HSMA (Eds.).
Proceedings of the ammonia and acidification symposium of the European association for the science of air
pollution (EURASAP), held in Bilthoven, The Netherlands. April 13 - 15,  1987; (pp 301-308). European
Association for the Science of Air Pollution (EURASAP).

107 Kenk G; Fischer H (1988). Evidence from nitrogen fertilisation in the forests of Germany. Environ Pollut, 54,
199-218.

108 U.S. EPA (1993). Air Quality Criteria Document for Oxides of Nitrogen (Final). U.S. Environmental Protection
Agency. EPA/600/8-91/049aF-cF; 3 Volumes.

109 LeBauer DS; Treseder KK (2008). Nitrogen limitation of net primary productivity in terrestrial ecosystems is
globally distributed. Ecology, 89, 371-379.

110 U.S. EPA (2008).  Nitrogen Dioxide/Sulfur Dioxide Secondary NAAQS Review: Integrated Science Assessment
(ISA) (Final). U.S. EPA, Washington D.C., EPA/600/R-08/082F.
in
   U.S. DOI (1996). Impact of Federal Programs on Wetlands. Available at www.doi.oepc/wetlands2/v2chlO.html
112 U.S. EPA (2008).  Nitrogen Dioxide/Sulfur Dioxide Secondary NAAQS Review: Integrated Science Assessment
(ISA) (Final). U.S. EPA, Washington D.C., EPA/600/R-08/082F.

113 Boynton WR; Garber JH; Summers R; Kemp WM (1995). Inputs, transformations, and transport of nitrogen and
phosphorus in Chesapeake Bay and selected tributaries. Estuaries, 18, 285-314.

114 Howarth RW; Billen G; Swaney D; Townsend A; Jaworski N; Lajtha K; Downing JA; Elmgren R; Caraco N;
Jordan T; Berendse F; Freney J; Kudeyarov V; Murdoch PS; Zhao-Liang Z (1996). Regional nitrogen budgets and
riverine N & P fluxes for the drainages to the North Atlantic Ocean: natural and human influences. Biogeochemistry,
35,75-139.

115Paerl H (1995). Coastal eutrophication in relation to atmospheric nitrogen deposition: current perspectives.
Ophelia, 41,237-259.
                                                  3-35

-------
116 Paerl H  (1997). Coastal eutrophication and harmful algal blooms: importance of atmospheric deposition and
groundwater as "new" nitrogen and other nutrient sources. Limnol Oceanogr, 42, 1154-1162.

117 Valiela I; Costa JE (1988). Eutrophication of Buttermilk Bay, a Cape Cod coastal embayment: concentrations of
nutrients and watershed nutrient budgets. Environ Manage, 12, 539-553.

118 Valiela I; Costa JE; Foreman K (1990). Transport of groundwater-borne nutrients from watersheds and their
effects on coastal waters. Biogeochemistry, 10, 177-197.

119 Costanza R; d'Arge R; de Groot R; Farber S; Grasso M; Hannon B; Limberg K; Naeem S; O'Neill RV; Paruelo J;
Raskin RG; Sutton P; Van Den Belt M (1997). The value of the world's ecosystem services and natural capital.
Nature, 387, 253-259.

120 D'Elia CJ; Sanders JG; Boynton WR (1986). Nutrient enrichment studies in a coastal plain estuary:
phytoplankton growth in large-scale, continuous cultures. Can J Fish Aquat Sci,  43, 397-406.

121 Howarth RW; Marino R (2006). Nitrogen as the limiting nutrient for eutrophication in coastal marine
ecosystems:  evolving views over three decades. Limnol Oceanogr, 51, 364-376.

122 U.S. EPA (2007). National Estuary Program Coastal Condition Report.  Chapter 7: Puerto Rico National
Estuary Program Coastal Condition, pp 385-400.  The entire report can be downloaded from
http://www.epa.gov/owow/oceans/nepccr/index.html

123 U.S. DOI (1996). Impact of Federal Programs on Wetlands. Available at www.doi.oepc/wetlands2/v2chlO.html

124 U.S. EPA (2000). Comprehensive Conservation and Management Plan for the San Juan Bay Estuary.  Report is
available at:  http://www.estuario.org/

125AgrawalH.,Malloy Q.G.J., Welch W.A., Wayne Miller J., Cocker III D.R. (2008) In-use gaseous and paniculate
matter emissions from a modern ocean going container vessel. Atmospheric Environment, 42(21), 5504-5510.

126 Isakson J., Persson T.A., E. Selin Lindgren E. (2001) Identification and assessment of ship emissions and their
effects in the harbour of Gteborg, Sweden. Atmospheric Environment, 35(21),  3659-3666.

127 Miller, W., et al. (2008 June 10). Measuring Emissions from Ocean Going Vessels. Presentation presented at the
Fuel, Engines, and Control Devices Workshop, San Pedro, California.

128 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.

129 U.S. EPA (2004). Air Quality Criteria for P articulate Matter. Volume I EPA600/P-99/002aF and Volume II
EPA600/P-99/002bF. Retrieved on March 19, 2009 from Docket EPA-HQ-OAR-2003-0190 at
http://www.regulations.gov/.

130 Dickhut R.M., Canuel E.A., Gustafson K.E., Liu K., Arzayus K.M., Walker S.E., Edgecombe G., Gaylor M.O.,
MacDonald E.H. (2000). Automotive Sources of Carcinogenic Polycyclic Aromatic Hydrocarbons Associated with
Particulate Matter in the Chesapeake Bay Region. Environmental Science & Technology,  34(21), 4635-4640.
                                                 3-36

-------
131 Simcik M.F., Eisenreich, S.J., Golden K.A., et al. (1996) Atmospheric Loading of Polycyclic Aromatic
Hydrocarbons to Lake Michigan as Recorded in the Sediments. Environmental Science and Technology, 30, 3039-
3046.

132 Simcik M.F., Eisenreich S.J., Lioy P.J. (1999)  Source apportionment and source/sink relationship of PAHs in the
coastal atmosphere of Chicago and Lake Michigan. Atmospheric Environment, 33, 5071-5079.

133 Poor N., Tremblay R., Kay H., et al. (2002) Atmospheric concentrations and dry deposition rates of polycyclic
aromatic hydrocarbons (PAHs) for Tampa Bay, Florida, USA. Atmospheric Environment, 38, 6005-6015.

134 Simcik M.F., Eisenreich, S.J., Golden K.A., et al. (1996) Atmospheric Loading of Polycyclic Aromatic
Hydrocarbons to Lake Michigan as Recorded in the Sediments. Environmental Science and Technology, 30, 3039-
3046.

135 Simcik M.F., Eisenreich S.J., Lioy P.J. (1999)  Source apportionment and source/sink relationship of PAHs in the
coastal atmosphere of Chicago and Lake Michigan. Atmospheric Environment, 33, 5071-5079.

136 U.S. EPA (1999). The Benefits and Costs of the Clean Air Act, 1990-2010. Prepared for U.S. Congress by U.S.
EPA, Office of Air and Radiation, Office of Policy Analysis and Review, Washington, DC, November; EPA report
no. EPA410-R-99-001.

137 U.S. EPA (2006). Air Quality Criteria Document for Ozone and Related Photochemical Oxidants (Final). U.S.
EPA, Washington, DC, EPA/600/R-05/004aF-cF, 2006.

138 Winner WE; Atkinson CJ (1986). Absorption of air pollution by plants, and consequences for growth. Trends in
Ecology and Evolution 1:15-18.

139 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

140 Tingey DT; Taylor GE (1982). Variation in plant response to ozone: a conceptual model of physiological events.
In: Effects of Gaseous Air Pollution in Agriculture and Horticulture (Unsworth, M.H., Omrod, D.P., eds.) London,
UK:  Butterworth Scientific, pp. 113-138.

141 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

142 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

143 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

144 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

145 Ollinger SV; Aber JD; Reich, PB (1997). Simulating ozone effects on forest productivity: interactions between
leaf canopy and stand level processes. Ecological Applications 7:1237-1251.
                                                 3-37

-------
146 Winner WE (1994). Mechanistic analysis of plant responses to air pollution. Ecological Applications, 4(4):651-
661.

147 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

148 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

149 Fox S; Mickler R A (1996). Impact of Air Pollutants on Southern Pine Forests. Springer-Verlag, NY, Ecol.
Studies,Vol. 118,513pp.

150 De Steiguer; Pye J ; Love C (1990). Air Pollution Damage to U.S. Forests. Journal of Forestry, Vol 88 (8) pp.
17-22.

151 Pye JM (1988). Impact of ozone on the growth and yield of trees: A review. Journal of Environmental Quality 17
pp.347-360.

152 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

153 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

154 McBride JR; Miller PR; Laven RD (1985). Effects of oxidant air pollutants on forest succession in the mixed
conifer forest type of southern California. In:  Air Pollutants Effects On Forest Ecosystems, Symposium
Proceedings, St. P, 157-167.

155 Miller PR; Taylor OC; Wilhour RG (1982). Oxidant air pollution effects on a western coniferous forest
ecosystem. Corvallis, OR: U.S. Environmental Protection Agency, Environmental Research Laboratory. EPA600-D-
82-276.

156 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

157 Kopp RJ; Vaughn WJ; HazillaM; Carson R  (1985). Implications of environmental policy for U.S. agriculture:
the case of ambient ozone standards. J. Environ. Manage. 20:321-331.

158 Adams R M; Hamilton S A; McCarl B A (1986).  The benefits of pollution control: the case of ozone and U.S.
agriculture. Am. J.  Agric. Econ. 34: 3-19.

159 Adams R M; Glyer J D; Johnson S L; McCarl BA (1989). A reassessment of the economic effects of ozone on
U.S. agriculture. JAPCA  39:960-968.

160 Abt Associates, Inc (1995).  Urban ornamental plants: sensitivity to ozone and potential economic losses. U.S.
EPA, Office of Air Quality Planning and Standards, Research Triangle Park. Under contract to RADIAN
Corporation, contract no. 68-D3-0033, WA no. 6. pp. 9-10.
                                                 3-38

-------
161 U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

162 Grulke NE (2003). The physiological basis of ozone injury assessment attributes in Sierran conifers. In:
Bytnerowicz, A., M.J. Arbaugh, and R. Alonso, eds. Ozone air pollution in the Sierra Nevada: Distribution and
effects on forests. New York, NY: Elsevier Science, Ltd. pp. 55-81.

163 White D; Kimerling AJ;  Overton WS (1992). Cartographic and geometric component of a global sampling
design for environmental monitoring. Cartogr. Geograph. Info. Sys.  19:5-22.

164 Smith G; Coulston J;  Jepsen E; Prichard T (2003). A national ozone biomonitoring program—results from field
surveys of ozone sensitive plants in Northeastern forests (1994-2000). Environ. Monit. Assess. 87:271-291.

165 Coulston JW; Riitters KH; Smith GC (2004). A preliminary assessment of the Montreal process indicators of air
pollution for the United States. Environ. Monit. Assess. 95:57-74.

166U.S. EPA (2006). Air Quality Criteria for Ozone and Related Photochemical Oxidants (Final). U.S. EPA,
Washington, DC, EPA/600/R-05/004aF-cF.

167 Xu et al (2006).  Attribution of sulfate aerosols in Federal Class I areas of the western United States based on
trajectory regression analysis. Atmospheric Environment 40, pg 3433-3447.
                                                  3-39

-------
CHAPTER 4:   COSTS	4-2
4.1   Fuel Production Costs	4-2
4.2   Operational Costs	4-7
4.3   Vessel Costs	4-10
4.4   Total Estimated ECA Costs in 2020	4-10
4.5   Cost Effectiveness	4-11
                                        4-1

-------
CHAPTER 4: Costs

       The reduction of NOx, SOx, and PM emissions from ships has an associated cost that can
reach beyond the shipping industry to marine fuel suppliers and companies who rely on the
shipping industry.  Though these cost impacts do exist, analyses presented in this document
indicate that the costs associated with the proposed EGA are expected to have a minimal
economic impact and to be relatively small compared to the resulting improvements in air
quality. This chapter describes the analyses used to evaluate the cost impacts of Tier III NOx
requirements and the use of lower sulfur fuel on vessels operating within the proposed EGA;
including estimates of lower sulfur fuel production costs. This chapter also presents cost per ton
estimates for EGA-based NOx and fuel sulfur standards and compares these costs with
established land-based control programs.  The costs presented here are based on the compliance
with EGA standards in 2020. All costs are presented in terms of 2006 U.S. dollars.

4.1 Fuel Production Costs

       This section presents our analysis of the impact of the proposed EGA on marine fuel
costs. Distillate fuel will likely  be needed to meet the 0.1 percent fuel sulfur limit, beginning in
2015, for operation in ECAs.A As  such, the primary cost of the fuel sulfur limit will be that
associated with switching from heavy fuel oil to higher-cost distillate fuel, when operating in the
EGA. Some engines already operate on distillate fuel and would not be affected by fuel
switching costs. Distillate fuel costs may be affected by the need to further refine the distillate
fuel to meet the 0.1 percent fuel sulfur limit.  To investigate these effects, studies were
performed on the impact of the North American EGA on global fuel production and costs. These
studies, which are summarized below,  include economic modeling to project bunker fuel demand
and refinery modeling which can be used to assess the impact of the proposed U.S. Caribbean
EGA on fuel costs.

4.1.1 Bunker Fuel Demand Modeling

       To assess the affect of an EGA on the refining industry, we needed to first understand and
characterize the fuels market and more specifically the demand for the affected marine fuels both
currently and in the future. Research Triangle Institute (RTI) was contracted to conduct a fuels
study using an activity-based economic approach.1  The RTI study established baseline bunker
fuel demand, projected a growth rate for bunker fuel demand, and established future bunker fuel
demand volumes.  The basis for this work was the Global Insights economic model, which
projects international trade for different categories of commodities.  Demand for marine fuels
A As an alternative, an exhaust gas cleaning device (scrubber) may be used. This analysis does not include the
effect on distillate fuel demand of this alternative approach. It is expected that scrubbers would only be used in the
case where the operator determines that the use of a scrubber would result in a cost savings relative to using
distillate fuel.  Therefore we are only estimating the cost of compliance using distillate fuel here as we believe this is
the most likely approach.
                                           4-2

-------
was derived from the demand of transportation of various types of cargoes by ship, which, in
turn, was derived from the demand for commodities produced in one region of the world and
consumed in another. The flow of commodities was matched with typical vessels for that trade
(characterized according to size, engine power, age, specific fuel consumption, and engine load
factors).  Typical voyage parameters were 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 was thus a function of commodity projections, ship characteristics, and voyage
characteristics.

       For this analysis, total fuel costs are derived using estimated fuel consumption values and
per-tonne incremental cost projections of using lower sulfur fuel.  The fuel consumption
estimates are those developed in the inventory analysis and presented in Chapter 2.  The per-
tonne fuel cost projections were developed using the World Oil Refining Logistics and Demand
(WORLD) model, in support of the North American EGA proposal. These estimates are based
on fuel price projections estimated by the Energy Information Administration (EIA) in 2008.
We believe the use of these fuel cost estimates is appropriate for three reasons.  First, use of
these fuel cost estimates allows for a comparable analysis between the two programs. Second,
the WORLD modeling was performed recently, which is especially important given the
uncertainty associated with making projections of cost impacts in 2020.  Third, based on
sensitivity modeling performed on fuel volumes, the impact of additional distillate demand as a
result of the proposed EGA would be small on the EGA WORLD fuel cost estimates. As such,
the price pressures as a result of the proposed EGA would be negligible. This is especially true
for this analysis, given that the volume of fuel consumed by ships operating in the proposed EGA
is small (approximately 3.6 percent) relative to the North American EGA.

4.1.2 Bunker Fuel Cost Modeling

4.1.2.1  Methodology

       To assess the impacts of the proposed EGA on fuel costs, the WORLD model was run by
Ensys Energy & Systems, the owner and developer of the refinery model. The WORLD model
is the only such model currently developed for this purpose, and was developed by a team of
international petroleum consultants. It  has been widely used by industries, government agencies,
and OPEC over the past 13 years,  including the Cross Government/Industry Scientific Group of
Experts, established to evaluate the effects of the different fuel options proposed under the
revision of MARPOL Annex VI.2 The model incorporates crude  sources, global regions,
refinery operations, and world economics.  The results of the WORLD model have been shown
to be comparable to other independent predictions of global fuel, air pollutant emissions and
economic predictions.

       WORLD is a comprehensive, bottom-up model of the global oil downstream that
includes crude and noncrude supplies;  refining operations and investments; crude, products, and
intermediates trading and transport; and product blending/quality  and demand. Its detailed
simulations are capable of estimating how the global system can be expected to operate under a
wide range of different circumstances, generating model outputs such as price effects and
projections of refinery operations  and investments.
                                          4-3

-------
4.1.2.2  Assessment of the Impact of Marine Fuel Standards

       During the development of the amendments to MARPOL Annex VI, a Cross
Government/Industry Scientific Group of Experts was established, by IMO, to evaluate the
effects of the different fuel options that were under consideration at the time.  This expert group
engaged the services of EnSys to assess the impact of these fuel options using the WORLD
model. The final report from this study presents great detail on the capabilities of the WORLD
model and provides support for why the WORLD model was chosen as the appropriate tool for
modeling the economic impacts of the different fuel options.3  The following description of the
WORLD model is taken from the expert group study:

       WORLD is a linear programming model that simulates the activities and economics  of
the world regional petroleum industry against short, medium or long term horizons. It models
and captures the interactions between:

   crude supply;

   non-crudes supply: Natural gas Liquids (NGLs), merchant MTBE, biofuels, petrochemical
   returns, Gas To Liquid fuels (GTLs), Coal to Liquid fuels (CTLs);

   refining operations;

   refining investment;

   transportation of crudes, products and intermediates;

   product blending/quality;

   product demand; and

   market economics and pricing.

       The model includes a database representing over 180 world crude oils and holds detailed,
tested, with state-of-the-art representation of fifty-plus refinery processes. These representations
include energy requirements based on today's construction standards for new refinery units. It
allows for advanced representation of processes for reformulated, ultra-lower sulfur/aromatics
fuels and was extended for detailed modeling of marine fuels for the aforementioned EPA and
API studies. The model contains detailed representations of the blending and key  quality
specifications for over 50 different products spread across the product spectrum and including
multiple grades of gasolines, diesel fuels/gasoils (marine and non-marine) and residual fuels
(marine and non-marine).

       The refining industry is a co-product industry. This means that changes in production of
one product also affect production volume  and/or production costs of other products. As
necessary, the model will adjust refinery throughputs and operations, crude and product trade
patterns to ensure that a specified product demand slate is met, without surplus or deficit of any
product.

       To evaluate the impact of changes to marine fuels specifications as a result of any of the
options under consideration, the model is run with a future demand scenario for all products. The
first run,  the base case, assumes marine fuels in line the current Annex VI regulation. The second
                                          4-4

-------
run is done with marine fuel specifications in line with the option under consideration. Both runs
are optimized independently. Since the only thing that is altered between the cases is the change
in the projected marine fuels regulation, the difference between both cases is therefore a true
assessment of the actual cost and other implications of the change to the marine fuels
requirements under consideration. Thus, the incremental refining investment costs, incremental
marine fuel costs and incremental refinery/net CC>2 emissions are all directly attributable to - and
must be allocated to - the change in regulation.

       Prior to the expert group study, EnSys made updates to the WORLD model to be able to
perform the analysis of the impacts of different marine fuel options. As part of this effort, the
refinery data, capacity additions, technology assumptions, and costs were reviewed.  EnSys
reviewed relevant regulations to ensure that the WORLD model was correctly positioned to
undertake future analyses of marine fuels EGAs. In developing these updates, a number of
issues had to be considered:

    the costs of refining, including the capital expenditures required to reduce bunker fuel sulfur
    content and the potential for process technology improvements;

    likely market reactions to increased bunker fuel costs, such as  fuel grade availability, impacts
    on the overall transportation fuels balance, and competition with land-based diesel and
    residual fuels for feedstocks that can upgrade fuels;

    the effects of emissions trading; and

    the potential for low- and high-sulfur grade bunker sources and consumption to partially shift
    location depending on supply volume, potential, and economics.

       The analytical system thus had to be set up to allow for alternative compliance scenarios,
particularly with regard to (a) adequately differentiating bunker fuel grades; (b) allowing for
differing degrees to which the EGA or other standards in a region  were presumed to be met by
bunker fuel sulfur reductions, rather than by other means such as scrubbing or emissions trading;
and (c) allowing for all residual fuel bunker demand to be reallocated to marine diesel.  Beyond
any international specifications, the analytical system needed to be able to accommodate future
consideration of regional, national, and local specifications.

The primary approach taken to manage these  issues was to:

    expand the number of bunker grades in the model to three distillates and four residual
    grades;8

    allow for variation where necessary in (regional) sulfur standards on specific bunker grades;
    and
B Specifically, the following seven grades were implemented: Marine Gas Oil (MGO), plus distinct high- and low-
sulfur blends for Marine Diesel Oil (MDO) and the main residual marine fuels Intermediate Fuel Oil (IFO) 180 and
IFO 380. The latest international specifications applying to these fuels were used, as were tighter sulfur standards for
the low-sulfur grades applicable in SECAs.
                                            4-5

-------
   enable residual bunker demand to be switched to marine diesel.

       Other updates to the WORLD model included product transportation matrices covering
tanker, interregional pipeline, and minor modes were expanded to embody the additional
distillate and residual bunker grades, adjustments to the yield patterns of the residuum
desulfurization, and blocking of paraffinic streams from residual fuel blends. The details of
compliance in any particular region must be estimated external to the main WORLD model. As
discussed above, we provided our estimates of affected fuel volumes to Ensys.

4.1.2.3  Updates for ECA Analysis

       To determine the impact of the proposed ECA, the WORLD model was employed using
the same basic approach as for the EVIO expert group study.  Modeling was performed for 2020
in which the control case included a fuel sulfur level of 0.1 percent in the U.S. and Canadian
EEZs.4  The baseline case was modeled as "business as usual" in which ships continue to use the
same fuel as today. This approach was used for two primary reasons. First, significant emission
benefits are expected in an ECA, beginning in 2015, due to the use of 0.1 percent sulfur fuel.
These benefits, and costs, would be much higher in the early years of the program before the 0.5
percent fuel  sulfur global standard goes into effect.  By modeling this scenario, we are able to
observe the impact of the proposed ECA in these early years. Second, there is no guarantee that
the global 0.5 percent fuel sulfur standards will begin in 2020.  The global standard may be
delayed until 2025, subject to a fuel availability review in 2018.  In addition, the 3.5 percent fuel
sulfur global standard, which begins in 2012, is higher than the current residual fuel sulfur
average of 2.7 percent.

       In the modeling for the expert group study, crude oil  prices were based on projections
released by the U.S. Energy Information Administration (EIA) in 2006.5 Since that time, oil
prices have fluctuated greatly. Using new information, EIA has updated its projections of oil
price for 2020.6'7 In response to this real-world effect, the ECA modeling was conducted using
the updated oil price estimates.  Specifically, we used a crude oil price of $51.55 for the reference
case, and $88.14/bbl for the high price case, both expressed in real (2006) dollars.  These crude
oil prices were input to the WORLD model which then computed residual and distillate marine
oil prices for 2020. The net refinery capital impacts were imputed based on the differences in the
costs to the refining industry that occur between the Base Cases and ECA cases in 2020.

4.1.2.4  Overall Increases Due to Fuel Switching and Desulfurization

       Global fuel use in 2020 by international shipping is projected to be 500 million tonnes/yr.
The main energy content effects of bunker grade shifts were captured in the WORLD modeling
by altering the volume demand  and, at the same time, consistency was maintained between the
bunker demand figures in tonnes and in barrels. The result was that partial or total conversion  of
intermediate fuel oil (IFO) to distillate was projected to lead to a reduction in the total global
tonnes of bunker fuel required but also led to a projected increase in the barrels required. These
effects are evident in the WORLD case results. Based on our estimates, the volume of marine
fuel affected by the proposed Caribbean ECA would be about 0.14 percent of total world residual
volume. As  would be expected, since the shift in fuel volumes on a world scale is relatively
small, the WORLD model predicts the overall global impact of an ECA to also be small.
                                          4-6

-------
       There are two main components to projected increased marine fuel cost associated with
an EGA. The first component results from the shifting of operation on residual fuel to operation
on higher cost distillate fuel.  This is the dominant cost component. The WORLD model
computed costs based on a split between the costs of residual and distillate fuels.  However, there
is a small cost associated with desulfurizing the distillate to meet the 0.1 percent fuel sulfur
standard in the EGA.  Based on the WORLD modeling, the average increase in costs associated
with switching from marine residual to distillate will be $145 per tonne.0 This is  the cost
increase that will be borne by the shipping companies purchasing the fuel. Of this amount, $6
per tonne is the cost increase associated with distillate desulfurization. In other words, we
estimate a cost increase of $6/tonne for distillate fuel used in an EGA.

       The above cost estimates are based on EIA's "reference case" projections  for crude oil
price in 2020. We also performed a  sensitivity analysis using EIA's "high price"  scenario.
Under this  scenario, the increase in fuel costs for switching from residual to distillate fuel is $237
per tonne.  The associated increase in distillate fuel cost is $7 per tonne.

       Table 4-1 summarizes the reference and high price fuel cost estimates with and without
an EGA. In the baseline case, fuel volumes for operation are 18% marine gas oil (MGO), 7%
marine diesel oil (MDO), and 75% IFO. In the proposed EGA, all fuel volumes are modeled as
MGO.

                             Table 4-1 Estimated Marine Fuel Costs
FUEL
MGO
MDO
IFO
UNITS
$/bbl
$/tonne
$/bbl
$/tonne
$/bbl
$/tonne
REFERENCE CASE
Baseline
$ 61.75
$ 464
$ 61.89
$ 458
$ 49.87
$ 322
ECA
$ 62.23
$ 468
$ 62.95
$ 466
$ 49.63
$ 321
HIGH PRICE CASE
Baseline
$ 102.70
$ 772
$ 102.38
$ 757
$ 83.14
$ 538
ECA
$ 103.03
$ 775
$ 103.70
$ 767
$ 82.52
$ 534
4.2 Operational Costs

       Operational costs refer to those which are incurred whenever the vessel is operating.
This analysis considers operating costs associated with both the low sulfur fuel requirement and
the Tier III NOX standards that would go into place in the proposed ECA for new vessels
beginning in 2016.
c Note that distillate fuel has a higher energy content, on a per tonne basis, than residual fuel. As such, there is an
offsetting cost savings, on a per tonne basis, for switching to distillate fuel. Based on a 5 percent higher energy
content for distillate, the net equivalent cost increase is estimated as $123 for each tonne of residual fuel that is being
replaced by distillate fuel ($200/tonne for the high price case).
                                            4-7

-------
       With respect to the low sulfur fuel requirement, we assume that all vessels in 2020 will
comply with the standards by switching to low sulfur distillate fuel when operating in the
proposed EGA.  As an alternative, an exhaust gas cleaning unit may be used.  It is expected that
this alternative equivalent technology would only be used in the case where the operator
determines that it would result in a cost savings relative to the use of distillate fuel.  To the extent
that operators choose an alternative technology, the costs may be overstated in this analysis.

4.2.1 Operational Costs Associated with the Use of Lower Sulfur Fuel

       There are two main cost components projected to increase as a result of compliance with
the fuel requirements of the proposed EGA.  The first component results from the shifting of
operation on residual  fuel to operation on higher cost distillate fuel; this is the dominant cost
component. The second is a small cost associated with further desulphurizing distillate fuel to
meet the 0.1 percent fuel sulfur standard in the EGA.  The methodology used to develop these
cost estimates is described in detail in the Technical Support Document developed for the North
American EGA proposal.  The estimated average increase in costs associated with switching
from marine residual to distillate fuel will be $145 per tonne.0 This is the cost increase that will
be borne by the shipping companies purchasing the fuel.  Of this amount, $6 per tonne is the cost
increase associated with distillate desulfurization. In other words, we estimate a cost increase of
$6/tonne for distillate fuel used in an EGA. The remaining $140 is due to switching from
residual fuel to distillate fuel.  The cost differential is modeled based on costs to the refinery and
assumes the market is in equilibrium.

       The estimated increase in operational costs associated with the use of lower sulfur fuel
was determined using the incremental cost of using distillate fuel instead  of residual fuel, the
increase in the cost of using distillate fuel, and the fuel consumption estimates provided in
Chapter 2 of this document. The change in residual fuel usage is approximately $169 million,
while the increase in cost of distillate fuel usage is estimated to be $233 million, resulting in the
total estimated increase in fuel costs in 2020 to be $64 million, as a result of this proposed EGA.

Table 4-2 Estimated Operational Costs Associated With the Use of Lower Sulfur Fuel in 2020 in the Proposed
                                           ECA
FUEL TYPE
Residual Fuel Usage
Distillate Fuel Usage
SCENARIO
Baseline (Without the ECA)
With the ECA
Baseline (Without the ECA)
With the ECA
Total Additional Fuel Costs Associated with the ECA
ESTIMATED
COST IN 2020
(MILLION)
$169
$0
$19
$252
$64
D Note that distillate fuel has a higher energy content, on a per tonne basis, than residual fuel. As such, there is an
offsetting cost savings, on a per tonne basis, for switching to distillate fuel. Based on a 5 percent higher energy
content for distillate, the net equivalent cost increase is estimated as $123 for each tonne of residual fuel that is being
replaced by distillate fuel ($200/tonne for the high price case).
                                            4-8

-------
4.2.2 Operational Costs Associated with SCR

       For vessels built on or after January 1, 2016, we assume that the engines comply with the
Tier III NOx standards through the use of SCR. We recognize that other technologies may be
used to meet the Tier III NOx standards. For instance, development work has been performed
with the goal of meeting these standards using exhaust gas recirculation and water injection
strategies. If these technologies are used, then operating costs would be lower as urea would not
be consumed in the vessel. As such, this analysis may overstate costs associated with the
proposed EGA. At the same time we consider SCR technology because, at this time, it appears
to be the most developed approach. Urea consumption for vessels equipped with SCR is
expected to be 7.5 percent of the fuel consumption.  The urea operational costs are based on a
price of $1.52 per gallon with a density of 1.09 g/cc. The cost per gallon was estimated for a 32.5
percent urea solution delivered in bulk to the ship through research completed by ICF
International for the U.S. Government, combined with historical urea price information.8'9'10'11'12
The total operational costs associated with the proposed EGA are based on the amount of fuel
consumed within the proposed EGA in the year 2020.  Fuel consumption estimates for 2020 are
presented in Chapter 2 of this technical support document including how the amount of fuel used
in this area was determined.

       The types of propulsion engines including: medium speed diesel, low speed diesel, gas
and steam turbine, were determined using the percentages that occur in the  current global fleet,
and are shown below in Table 4-3.13  These percentages were applied to the total fuel
consumption estimated for 2020, resulting in an estimate of amount of fuel  used by each engine
type. Next, the "Age Distribution" data from Chapter 2 of this document was applied to these
percentages to estimate what  percentage of each engine type would be built after 2015.  As
discussed above, both medium-speed and low-speed main propulsion engines are assumed here
to use SCR as the  Tier III  NOx control strategy.  The resulting percentage of vessels built after
2015 was then applied to the  estimated fuel consumption values per engine-type to estimate the
amount of fuel used in vessels equipped with SCR.

            Table 4-3 Percentage of Vessels by Engine Type Estimated to Use SCR in 2020
TYPE OF ENGINE
Slow-Speed Diesel
Medium-Speed Diesel
Gas Turbine
Steam Turbine
PERCENT
OF GLOBAL
FLEET
80%
17%
0.4%
2.6%
PERCENT OF EACH ENGINE
ESTIMATED TO HAVE BEEN
AFTER 201 5
TYPE
BUILT
31%
36%
95%
34%
       The result for this proposed EGA is that the operational costs associated with the use of
urea in 2020 by ships built as of 2016 are based on total urea consumption of nearly 4 million
gallons are shown in Table 4-3 and estimated to be approximately $6 million.
                                          4-9

-------
 Table 4-4 Estimated Operational Costs Associated with the Use of SCR as a Tier III NOX Control Strategy
ESTIMATED GALLONS OF DISTILLATE FUEL USED IN 2020 (MILLION)
Slow-Speed Diesel Powered Vessels
Medium-Speed Diesel Powered Vessels
Gas Turbine Powered Vessels
Steam Turbine Powered Vessels
138
29
0.67
4.5

ESTIMATED GALLONS OF DISTILLATE FUEL USED IN TIER III SCR EQUIPPED VESSELS IN 2020
(MILLION)
Slow-Speed Diesel Powered Vessels
Medium-Speed Diesel Powered Vessels
Total Gallons of Distillate Attributable to Tier III
42
10
52

TOTAL ESTIMATED UREA USAGE AND COST IN 2020 IN THE PROPOSED EGA (MILLION)
Estimated Gallons of Urea Used with Tier III Engines in 2020
Total Estimated Cost of Urea Used in the Proposed EGA in 2020
3.9
$6
4.3 Vessel Costs

       The cost analysis for the proposed EGA does not include equipment costs associated with
vessel modifications to accommodate EGA fuel for new and existing vessels or costs associated
with the Tier III NOx limits for vessels built after 2016. This is reasonable for two reasons.
First, as noted in Chapter 1 of this document, approximately 55 percent of commercial shipments
to Puerto Rico and the U.S. Virgin Islands originate in the continental United States, and
approximately 90 percent of shipments from these areas are destined to the continental United
States. All vessels that carry these goods will already be equipped to comply with the EGA
requirements, as they will operate in the North American EGA. Second, the proposed EGA
extends a maximum of about 60 nm from the baseline. Ship positional data presented in Section
7 of the Information Document suggests that there is little transit activity within the proposed
EGA, and such transit activity that occurs is likely at the outer boundary of the EGA where ships
have a lesser impact on air qualityE and where it would be possible to reroute to avoid the
proposed EGA.  It is expected that those vessels transiting through the area that do not have
Puerto Rico  and the Virgin Islands as a destination will reroute, and therefore these vessels
would also not incur equipment costs associated with EGA compliance.

4.4 Total Estimated ECA Costs in 2020

       The total costs associated with improving ship emissions from current performance to
ECA standards in 2020 include both the incremental fuel and urea operational costs presented
above. The  operational costs associated with the use of urea are estimated to be $6 million in
2020. The operational costs associated with the use of lower sulfur fuel for the proposed ECA
E See Section 5 of the Information Document for a discussion of back trajectory analysis and the impacts of ship
emissions on shore.
                                          4-10

-------
are estimated to be $64 million in 2020.  Therefore, the total costs associated with the proposed
EGA in 2020 are expected to be $70 million, Table 4-5 summarizes these costs.

                        Table 4-5 Total Estimated U.S. ECA Costs in 2020
TOTAL ESTIMATED OPERATIONAL COSTS ASSOCIATED WITH THE
PROPOSED ECA IN 2020 (MILLION)
Residual Fuel Usage
Distillate Fuel Usage
Baseline (Without the ECA)
With the ECA
Baseline (Without the ECA)
With the ECA
Total Additional Fuel Costs Associated with the ECA
Total Urea Costs Associated with the ECA
Total Additional Operational Costs Associated with the ECA
$169
$0
$19
$252
$64
$6
$70
4.5 Cost Effectiveness

       As discussed in Chapter 3, the proposed ECA is expected to bring many air quality and
environmental benefits. Sections 4.1 through 4.2, above, summarize the various costs of the
proposed ECA.  However, this does not shed light on how cost effective the proposed ECA will
be, compared to other control programs, at providing the expected emission reductions.

       One tool that can be used to assess the value of the proposed ECA is the measure of cost
effectiveness; a ratio of engineering costs incurred per tonne of emissions reduced. The U.S.
Government has compared the ECA cost effectiveness to the ratio of costs per tonne of
emissions reduced for other control programs. As is shown in this section,  the NOx, SOx and
PM emissions reductions from the proposed ECA compare favorably—in terms of cost
effectiveness—to other land-based control programs that have been implemented.

4.5.1 ECA Cost Effectiveness

       Chapter 2 of this document summarizes the inventory analyses from which the
projections of pollutant reductions are drawn. The projected emission reductions due to the
proposed ECA are presented below in Table 4-6.

                   Table 4-6 C3 Emission Inventories for Proposed ECA in 2020
EMISSION TYPE
Reference
Control
Delta Emissions
Delta Emissions (%)
ANNUAL EMISSIONS (METRIC TONNES)a'b
NOX
36,950
27,032
-9,919
-27%
PM10
3,793
512
-3,342
-86%
PM25C
3,488
471
-3,017
-86%
HC
1,509
1,509
0
0%
CO
3,609
3,609
0
0%
S02
29,568
1,075
-28,493
-96%
C02
1,797,909
1,711,452
-86,457
-5%
       Note that PM2.5 is estimated to be 92 percent of the more inclusive PMio emission
inventory for marine vessels. In Chapter 2, we generate and present PM2.5 inventories since
recent research has determined that these are of greater health concern. Traditionally, we have
used PMio in our cost effectiveness calculations.  Since cost effectiveness is a means of
                                         4-11

-------
comparing control measures to one another, we use PMio in our cost effectiveness calculations
for comparisons to past control measures.

       Using the costs associated with NOx, SOx and PM control described in sections 4.1
through 4.2 above, and the emission reductions shown in Table 4-6 above, we calculated the cost
per tonne, or cost effectiveness, of the proposed EGA.  As described above, the costs of the
proposed EGA include costs to refiners to produce additional distillate fuel, as well as costs for
reductants to reduce NOx emissions. Operational costs are incurred over time.

       The resultant cost per tonne numbers depend on how the costs are allocated to each
pollutant. We have allocated costs as closely as possible to the pollutants for which they are
incurred. The costs to apply engine controls to meet Tier III NOx standards, including reductants,
have been allocated to NOx.  In our analyses, we have allocated half of the costs of fuel
switching, to PM and half to SOX because the costs incurred for control measures to reduce SOX
emissions directly reduce emissions of PM as well.

       The resultant estimated cost effectiveness numbers are  shown in Table 4-7. These
include costs and emission reductions that are expected to occur due to compliance with the
proposed EGA in the year 2020.
               Table 4-7 Aggregate Long Term ECA Cost per Tonne (2006 U.S. Dollars)
POLLUTANT
NOX
SOX
PM25
COST PER TONNE IN 2020
$600
$1,100
$11,000F
4.5.2 Land-Based Control Program Cost Effectiveness

       The U.S. Government has already imposed restrictions on emissions of NOx, SOx, PM
and other air pollutants, from a wide range of land-based industrial (stationary) and
transportation (mobile) sources as well as consumer and commercial products.  We have applied
a wide range of programmatic approaches to achieve significant air pollution reductions.
Regulatory regimes typically either mandate or incentivize emissions aftertreatment, cleaner
fuels or raw materials, improved practices, as well as new processes or technologies.

       Significant emission reductions of NOx and SOx in the U.S. have been achieved via
performance standards for new combustion sources and market-based programs that cap
emissions at the regional level. Since 1996, the Acid Rain Program and NOx Budget Trading
Program have been highly successful at drastically reducing both NOx and SOx from power
plants in the Eastern U.S. Since 2004, NOx, SOx and PM emissions from highway and nonroad
heavy duty trucks and equipment have been decreasing with performance and emission standards
 Converting to PM10 the cost per tonne would be 10,000. This figure is used in Table 4-8 below.
                                          4-12

-------
that will be completely phased in by 2010.  To allow technology to advance, diesel fuel for use in
vehicles in the U.S. has been reduced to less than 0.0015 percent sulfur (15 parts per million by
weight), and diesel fuel for use in off-road equipment, locomotives and domestic marine vessels
will be reduced to this level by 2012.

       Advanced technology is already required on stationary sources in the U.S., including
electricity generation produced by combustion; oil and gas; forest products (including pulp and
paper and wood products); smelting and refining (including aluminum, alumina, and base metal
smelting); iron and steel; iron ore pelletizing; potash; cement; lime; and chemicals production,
including fertilizers. On mobile sources, advanced technology to reduce NOx is fully phased in
as of 2010 for engines on heavy duty trucks and will be phased in by 2015 for engines on
harborcraft.

       Programs that are designed to capture the efficiency of designing  and building new
compliant sources tend to have better cost-effectiveness than programs that principally rely on
retrofitting existing sources. Even considering the retrofitting programs, the control measures
that have been implemented on land-based  sources have been well worthwhile when considering
the benefits of the programs.

       The cost of reducing air pollution from these land-based sources has ranged greatly,
depending on the pollutant, the type of control program and the nature of the source.  A selection
of programs and their cost effectiveness is presented in Table 4-8.  Unless otherwise noted, the
programs named in the table address newly built sources only.
                                          4-13

-------
             Table 4-8 Land-Based Source Control Program Cost Per Tonne" Comparisons
SOURCE CATEGORY14
Highway Diesel Fuel Program d
55 Fed Reg 34120, August 21, 1990
Stationary Diesel (CI) Engines °
71 Fed Reg 39154, July 11, 2006
Locomotives and Harborcraft (Both
New and Retrofits) d
73 Fed Reg 25097, May 6, 2008
Heavy Duty Nonroad Diesel Engines'1
69 Fed Reg 38957, June 29, 2004
Heavy Duty Onroad Diesel Engines d
66 Fed Reg 5001, January 18, 2001
International Shipping (U.S. ECA)
(Both New and Retrofits) d
Proposed Puerto Rico/Virgin Islands
ECAe
Light Duty Gasoline/Diesel Engines d
65 Fed Reg 6697, February 10, 2000
Fossil Fuel Fired Power Plants
(Retrofits) c
5$ Fed Reg 3590, January 11, 1993;
63 Fed Reg 57356, October 27, 1998
Other Stationary Sources
(Both New and Retrofits) c
67 Fed Reg 80186, December 31, 2002
IMPLEMENTATION
DATE
1993
2006
2015
2015
2010
2016
2016
2009
2000 to 20 10
Ongoing
NOX
COST/TONNE
-
600 - 22,000
800b
l,200b
2,400 b
2,600
600
2,800 b
3,400
4,000 - 12,000
SOX
COST/TONNE
-
-

900
6,400
1,200
1,100
6,600
300
300 - 6,000
PM10
COST/TONNE
11,000
4,000 - 46,000
9,300 (New)
50,000
(Retrofit) c
14,000
16,000
10,000
10,000
14,000

Variable
Notes:
a Units are 2006 U.S. dollars per metric ton. To convert to $/short ton, multiply by 0.907.
b Includes NOX plus non-methane hydrocarbons (NMHC). NMHC are also ozone precursors, thus some rules set
combined NOX+NMHC emissions standards.  NMHC are a small fraction of NOX so aggregate cost/ton
comparisons are  still reasonable.
0 Annualized costs of control for individual sources, except SOX for Power Plants is a typical auction price.
d Aggregate program-wide cost/tonne over 30 years, discounted at 3%, except Light Duty and Highway Fuel
aggregate costs were discounted at slightly higher rates, yielding slightly lower cost estimates.
e Estimate includes the year 2020 only.

       Another example of one of the earlier programs is the 1990 regulation promulgated by
the U.S. Government to reduce the sulfur content of highway diesel fuel.  The cost effectiveness
of PM reductions from that program varied depending on how the benefit of reduced wear on the
engines was credited. Because the cleaner fuel with 0.05% sulfur (500 ppm) lengthened the
useful life of the engines, the program could be characterized as having negative costs (with
savings up to $100,000 per tonne) if the maximum engine wear credit was attributed to the
program.  If no engine wear credit was included, the program was estimated to cost a maximum
of $11,000 per tonne of PM reduced.

       As shown above, the projected cost per tonne  of the proposed ECA falls well within the
respective ranges of the other programs.  The proposed ECA cost-effectiveness is comparable to
the cost per tonne of current programs for new land-based sources, and has favorable cost
effectiveness compared to land-based retrofit programs.
                                            4-14

-------
1 Research Triangle Institute, 2008. "Global Trade and Fuels Assessment—Future Trends and Effects of
Designating Requiring Clean Fuels in the Marine Sector"; Research Triangle Park, NC; EPA420-R-08-021;
November.  (Available at http://www.epa.gov/otaq/regs/nonroad/marine/ci/420r08021 .pdf)

2 International Maritime Organization, Note by the Secretariat, "Revision of MARPOL Annex VI and NOX
Technical Code; Input from the four subgroups and individual experts to the final report of the Informal Cross
Government/Industry Scientific Group of Experts," Subcommittee on Bulk Liquids and Gases, 12th Session,
Agenda Item 6, BLG 12/INF.10, December 28, 2007.

3 International Maritime Organization, Note by the Secretariat, "Revision of MARPOL Annex VI and NOX
Technical Code; Input from the four subgroups and individual experts to the final report of the Informal Cross
Government/Industry Scientific Group of Experts," Subcommittee on Bulk Liquids and Gases, 12th Session,
Agenda Item 6, BLG 12/INF.10, December 28, 2007.

4 EnSys Energy & Systems, Inc. and RTI International 2009. Global Trade and Fuels Assessment—Additional EGA
Modeling Scenarios, prepared for the U.S. Environmental Protection Agency.

5 Energy Information Administration, 2006. "Annual Energy Outlook 2006" (DOE/EIA-0383(2006)); Washington,
DC. (Available at: http://www.eia.doe.gov/oiaf/aeo/archive.html)

6 Energy Information Administration, 2008. "Annual Energy Outlook 2008" (DOE/EIA-0383(2008)); Washington,
DC. (Available at: http://www.eia.doe.gov/oiaf/aeo/)

7 Energy Information Administration, 2008. "International Energy Outlook 2008" (DOE/EIA-0484(2008));
Washington, DC.  (Available at: http://www.eia.doe.gov/oiaf/ieo/)

8ICF International, "Costs of Emission Reduction Technologies for Category 3 Marine Engines," prepared for the
U.S. Environmental Protection Agency, December 2008. EPA Report Number : EPA-420-R-09-008.

9 "Nonroad SCR-Urea Study Final Report" July 29, 2007 TIAX for Engine Manufacturers Association (EMA) can
be found at:http://www.enginemanufacturers.org/admin/content/upload/198.pdf

10http://www.adblueonrine.co.uk/air_l/bulk_derivery

11 http://www.factsaboutscr.com/documents/IntegerResearch-Ureapricesbackto20051evels.pdf

12 http://www.fertilizerworks.com/fertreport/index.html

13 www.sea-web.com Lloyd's Register of Ships accessed September, 2008.

14 Regulation of Fuels and Fuel Additives: Fuel Quality Regulations for Highway Diesel Fuel Sold in 1993 and
Later Calendar Years, 55 Fed Reg 34120, August 21, 1990.
Standards of Performance for Stationary Compression Ignition Internal Combustion Engines, 71 Fed Reg 39154,
July 11,2006.
Control of Emissions of Air Pollution from Locomotives and Marine Compression-Ignition Engines Less Than 30
Liters per Cylinder, 73 Fed Reg 25097, May 6, 2008.
Control of Emissions of Air Pollution From Nonroad Diesel Engines and Fuel 69 Fed Reg 38957, June 29, 2004.
Control of Air Pollution from New Motor Vehicles: Heavy-Duty Engine and Vehicle Standards and Highway Diesel
Fuel Sulfur Control Requirements 66 Fed Reg 5001, January 18, 2001.
Control of Air Pollution From New Motor Vehicles: Tier 2 Motor Vehicle Emissions Standards and Gasoline Sulfur
Control Requirements 65 Fed Reg 6697, February 10, 2000.
Acid Rain Program; General Provisions and Permits, Allowance System, Continuous Emissions Monitoring, Excess
Emissions and Administrative Appeals, 58 Fed Reg 3590, January 11, 1993; Finding of Significant Contribution
and Rulemaking for Certain States in the Ozone Transport Assessment Group Region for Purposes of Reducing
Regional Transport of Ozone, 63 Fed Reg 57356, October 27, 1998.
Prevention of Significant Deterioration (PSD) and Nonattainment New Source Review (NSR): Baseline Emissions
Determination, Actual-to-Future-Actual Methodology, Plantwide Applicability Limitations, Clean Units, Pollution
Control Projects, 67 Fed Reg 80186, December 31, 200270
                                                 4-15

-------
CHAPTER 5:  ECONOMIC IMPACTS	5-2
5.1   The Purpose of an Economic Impact Analysis	5-3
5.2   Economic Impact Analysis Methodology	5-3
5.3   Expected Economic Impacts of the Proposed ECA	5-5
APPENDICES	5-14
                                     5-1

-------
CHAPTER 5: Economic Impacts

       Chapter 4, above, provides the engineering costs associated with complying with the Tier
III NOx limits and the EGA fuel sulfur limits for all ships operating in the proposed EGA in
2020. In this chapter, we examine the economic impacts of these costs. We look at two aspects
of the economic impacts:  estimated social costs and how they are shared across stakeholders,
and estimated market impacts in terms of changes in prices and quantities produced for directly
affected markets.  All costs are presented in terms of 2006 U.S. dollars.

       The total estimated social costs associated with the proposed EGA in 2020 are equivalent
to the estimated compliance costs of the program, at approximately $70 million.  These costs are
expected to accrue initially to the owners and operators of affected vessels. These owners and
operators are expected to pass their increased costs on to the entities that purchase their
transportation services in the  form of higher freight rates.  Ultimately, these costs will be borne
by the final consumers of goods transported by ocean-going vessels in the form of higher prices
for those goods.

       We estimate that these costs added to the total  cost of shipping goods to or from Puerto
Rico or the U.S. Virgin  Islands will result in only a modest increase in the costs of goods
transported by ship. In  most  cases, ships that operate in the proposed EGA also operate in the
North American EGA and/or  the North Sea and Baltic SECAs. This means there are no
additional equipment costs associated with the proposed EGA and therefore no impacts on the
price of a vessel. With  regard to operating costs, the total costs associated with improving ship
emissions from  current  performance to EGA standards in 2020 include the differential costs of
using lower sulfur fuel,  and the  use of urea on vessels  equipped with selective catalytic reduction
(SCR) systems to meet  Tier III NOx standards. The total estimated costs incurred as a result of
using lower sulfur fuel in the  proposed EGA are US$64 million, and the total estimated cost
associated with  the use  of urea are US$6 million.  The total estimated additional  costs associated
with the Caribbean EGA are approximately $70 million in 2020.

       The economic impacts of complying with the program on ships engaged in international
trade are expected to be modest. With regard to container ships, improving from current
performance to EGA standards would increase the cost of shipping a twenty-foot-equivalent
container by about US$0.33 to US$1.35 depending on the  size of the ship and the length of the
route. This represents an increase of less than one percent in the cost of shipping a 20-foot
container. The price impacts on oil tanker services are also expected to be small, with a price
impact of less than US$0.002 per barrel.  With regard  to cruise ships, we estimate that the price
impacts of the proposed EGA on a large cruise ship that operates from the U.S. East Coast
throughout the Caribbean would be  approximately US$0.40 per passenger per day for a 14-day
cruise; this represents a less than one percent increase  in the price  of a cruise. The price impacts
on a medium sized cruise ship that operates a route between the U.S. and Puerto  Rico will be
approximately US$0.60 per passenger per day for a 5-day  cruise; this represents  a less than one
percent increase in the price of the cruise.  The impacts on a small cruise  ship that spends nearly
one-quarter of the time  in the proposed EGA is estimated to be approximately US$1.30 per
                                          5-2

-------
passenger per day for an 8-day trip; this represents a less than one percent increase in the price of
the cruise.

       It should be noted that this economic analysis holds all other aspects of the market
constant except for the designation of the proposed EGA. It does not attempt to predict
equilibrium market conditions for 2020 or the impacts of any other programs or economic
conditions that may affect marine transportation.  This approach is appropriate because the goal
of an economic impact analysis is to explore the impacts of a specific program; allowing changes
in other market conditions would confuse the impacts due to the proposed regulatory program.

       The remainder of this chapter provides detailed information on the methodology we used
to estimate these economic impacts and the results of our analysis.

  5.1   The Purpose of an Economic Impact Analysis

       An Economic Impact Analysis (EIA) is prepared to provide information about the
potential  economic consequences of a regulatory action. Such an analysis consists of estimating
the social costs of a regulatory program and the distribution of these costs across stakeholders.

       In an economic impact analysis, social costs are the value of the goods and services lost
by society resulting from a) the  use of resources to comply with and implement a regulation and
b) reductions in output.  There are two parts to the analysis. In the economic welfare analysis,
we look at the total social costs  associated with the program and their distribution across key
stakeholders. In the market analysis, we estimate how prices and quantities of goods directly
affected by the emission control program can be expected to change once the program goes into
effect.

  5.2 Economic Impact Analysis Methodology

       Economic impact analysis is rooted in basic microeconomic theory. We use the laws of
supply  and demand to simulate  how markets can be expected to respond to  increases in
production costs that occur as a result of the new emission control program. Using that
information, we construct the social costs of the program and identify how those costs will be
shared  across the markets and, thus, across stakeholders. The relevant concepts are summarized
below and are presented in greater detail in Appendix 5A to this chapter.

       Before  the implementation  of a control program, a market is assumed to be in
equilibrium, with producers producing the amount of a good that consumers desire to purchase at
the market price. The implementation of a control program results in an increase in production
costs by the amount of the compliance costs. This generates a "shock" to the initial  equilibrium
market conditions (a change in supply).  Producers of affected products will try to pass some or
all of the  increased production costs on to the consumers of these goods through price increases,
without changing the  quantity produced. In response to the price increases, consumers will
decrease  the quantity  they buy of the affected good (a change  in the quantity demanded). This
creates surplus production at the new price.  Producers will react to the decrease in quantity
demanded by reducing the quantity they produce, and they will be willing to sell the remaining
production at a lower price that does not cover the full amount of the compliance costs.
                                          5-3

-------
Consumers will then react to this new price.  These interactions continue until the surplus is
removed and a new market equilibrium price and quantity combination is achieved.

       The amount of the compliance costs that will be borne by stakeholders is ultimately
limited by the price sensitivity of consumers and producers in the relevant market, represented
by the price elasticities of demand and supply for each market.  An "inelastic" price elasticity
(less than one) means that supply or demand is not very responsive to price changes (a one
percent change in price leads to less than one percent change in quantity).  An "elastic" price
elasticity (more than one) means that supply or demand is sensitive to price changes (a one
percent change in price leads to more than one percent change in quantity). A price elasticity of
one is unit elastic, meaning there is a one-to-one correspondence between a percent change in
price and percent change in quantity.

       On the production side, price elasticity of supply depends on the time  available to adjust
production in response to a change in price, how easy it is to store goods, and the cost of
increasing (or decreasing) output.  In this analysis  we assume the supply for engines, vessels, and
marine transportation services is elastic: an increase in the market price of an engine, vessel or
freight rates will lead producers to want to produce more, while a decrease will lead them to
produce less (this is the classic upward-sloping supply curve). It would be difficult to estimate
the slope of the supply curve for each of these markets given the global nature of the sector.
However, it is reasonable to assume that the supply elasticity for the ocean marine transportation
services market is likely to be greater than one.  This is because output can more easily be
adjusted due to a change in price.  For the same reason, the supply elasticity for the new
Category 3 engine market is also likely to be greater than one, especially since these engines are
often used in other land-based industries, especially in power plants.  The supply elasticity for
the vessel construction market, on the other hand,  may be less than or equal to one, depending on
the vessel type, since it may be harder to adjust production and/or store output if the price drops,
or rapidly increase production if the price increases. Because of the nature of this industry, it
would not be possible to easily switch production to other goods, or to stop or start production of
new vessels.

       On the consumption side, we assume that the demand for engines is a  function of the
demand for vessels, which is a function of the demand for international shipping (demand for
engines and vessels is derived from the demand for marine transportation services). This makes
intuitive sense:  Category 3 engine and ocean-going vessel manufacturers would not be expected
to build an engine or vessel unless there is a purchaser, and purchasers will want a new
vessel/engine only if there is a need for one to supply marine transportation services. Deriving
the price elasticity of demand for the vessel and engine markets from the international shipping
market is an important feature of this analysis because it provides a link between the product
markets.

       In this analysis, the price elasticity of demand is nearly perfectly inelastic. This stems
from the fact that, that, for most goods, there are no reasonable alternative shipping modes. In
most cases, transportation by rail or truck is not feasible, and transportation by aircraft is too
expensive.  Approximately 90 percent of world trade by tonnage is moved by ship, and ships
provide the most efficient method to transport these goods on a tonne-mile basis.   Stopford
notes that "shippers need the cargo and, until they  have time to make alternative arrangements,


                                           5-4

-------
must ship it regardless of cost... The fact that freight generally accounts for only a small portion
of material costs reinforces this argument."2 A nearly perfectly inelastic price elasticity of
demand for marine transportation services means that virtually all of the compliance costs can be
expected to be passed on to the consumers of marine transportation services, with no change in
output for engine producers, ship builders, or owners and operators of ships engaged in
international trade.

       The economic impacts described below rely on the estimated engineering compliance
costs presented in Chapter 4. These include the expected increases in  operating costs for vessels
operating in the EGA.  These increased operating costs include increases in fuel costs, and the
use of urea for engines equipped with SCR, as well as  a small increase in operating costs for
operation outside the EGA due to the fuel price impacts of the program.

  5.3 Expected Economic Impacts of the Proposed ECA

5.3.1  Engine and  Vessel Market Impacts

       The market analysis explores the impact of a regulatory program on the prices and
quantity of goods produced in directly affected markets.  In this case, the vast majority of vessels
that operate in the proposed ECA also operate  in the North American ECA and/or the Baltic or
North Sea SECAs. The equipment costs associated with ECA compliance are already incurred
as a result of those programs, and therefore the proposed ECA would not be expected to result in
any change to the prices of affected marine diesel engines or vessels, or the quantities of vessels
produced.

       Table 5-1 and Table 5-2 present the estimated price impacts for a sample of engine and
vessel combinations that were developed for the North American ECA, for medium speed and
slow speed engines,  respectively. These tables are provided to show the expected engine and
vessel impacts for the limited number of vessels that would not already operate in any other
ECA. These new engine and new vessel costs are unlikely to be incurred, however, as owners of
vessels that do not operate in any other ECA would be expected to find ways to redistribute their
fleets to avoid them. These price impacts reflect the impacts of the costs that will be incurred
when the most stringent ECA standards are in  place in 2020. These estimated price impacts are
small when compared to the price of a new vessel.

 Table 5-1 Summary of Estimated Market Impacts - New Medium Speed Engines and Vessels (2020; $2006)
SHIP TYPE
Auto Carrier
Bulk Carrier
Container
General Cargo
Passenger
Reefer
RoRo
AVERAGE
PROPULSION
POWER
9,600
6,400
13,900
5,200
23,800
7,400
8,600
NEW VESSEL ENGINE
PRICE IMPACT (NEW
TIER III ENGINE
PRICE IMPACT)3
$573,200
$483,500
$687,800
$450,300
$952,500
$511,000
$543,800
NEW VESSEL FUEL
SWITCHING
EQUIPMENT PRICE
IMPACTb
$42,300
$36,900
$49,200
$34,900
$65,400
$38,500
$40,500
NEW VESSEL
TOTAL PRICE
IMPACT
$615,500
$520,400
$736,000
$475,200
$1,107,900
$549,500
$584,300
                                          5-5

-------
1 Tanker
Misc.
6,700
9,400
$492,800
$566,800
$37,400
$41,900
$530,2001
$608,700 |
   a Medium speed engine price impacts are estimated from the cost information presented in Chapter 4 using the
   following formula: ((10%*(Mechanical Fuel Injection to Common Rail)+(30%*(Electronic Fuel Injection to Common
   Rail))+(T3 Engine Modifications)+(T3 SCR))
   b Assumes 32 percent of new vessels would require the fuel switching equipment.
     Table 5-2 Summary of Estimated Market Impacts - Slow Speed Engines and Vessels (2020; $2006)
SHIP TYPE
Auto Carrier
Bulk Carrier
Container
General Cargo
Passenger
Reefer
RoRo
Tanker
Misc.
AVERAGE
PROPULSION
POWER
11,300
8,400
27,500
7,700
23,600
10,400
15,700
9,800
4,700
NEW VESSEL ENGINE
PRICE IMPACT (NEW
ENGINE PRICE
IMPACT)3
$825,000
$672,600
$1,533,100
$632,900
$1,385,300
$781,000
$1,042,100
$744,200
$453,600
NEW VESSEL FUEL
SWITCHING
EQUIPMENT PRICE
IMPACT5
$48,000
$42,700
$63,900
$41,000
$61,200
$46,500
$53,900
$45,300
$32,000
NEW VESSEL
TOTAL PRICE
IMPACT
$873,000
$715,300
$1,597,000
$673,900
$1,446,500
$827,500
$1,096,000
$789,500
$485,600
   a Slow speed engine price impacts are estimated from the cost information presented in
   following formula:  (5%*(Mechanical Fuel Injection to Common Rail))+(15%*(Electronic Fuel
   Rail))+(T3 Engine Modifications)+(T3 SCR))
   b Assumes 32 percent of new vessels would require the fuel switching equipment
Chapter 4 using the
Injection to Common
       A selection of new vessel prices that were developed for the North American EGA is
provided in Table 5-3, and range from about $40 million to $480 million. The estimated price
increases range from about $600,000 to $1.5 million.  A price increase of $600,000 to comply
with the EGA requirements would be an increase of approximately two percent for a $40 million
vessel.  The largest vessel price increase noted above, for passenger vessels, is about $1.5
million;  this is a price increase of less than one percent for a $478 million passenger vessel.
Price increases of this magnitude would be expected to have little, if any, effect on the quantity
sales of new vessels, all other economic conditions held constant.  Again, these impacts are
presented for illustration only; most vessels that operate in the proposed EGA will have incurred
these costs as result of the North American EGA and/or the North Sea and Baltic Sea SECAs.

        Table 5-3 Newbuild Vessel Price by Ship Type and Size, Selected Vessels (Millions, $2008)
VESSEL
TYPE
Bulk Carrier
Container
VESSEL SIZE
CATEGORY
Handy
Handymax
Panamax
Capesize
Feeder
Intermediate
SIZE RANGE (MEAN)
(DWT)
10,095 - 39,990 (27,593)
40,009-54,881 (47,616)
55,000-78,932 (69,691)
80,000-364,767(157,804)
1,000-13,966 (9,053)
14,003-36,937 (24,775)
NEWBUILD
$56.00
$79.00
$97.00
$175.00
$38.00
$70.00
                                              5-6

-------
VESSEL
TYPE

Gas carrier
General
cargo
Passenger
Reefer
Ro-Ro
Tanker
VESSEL SIZE
CATEGORY
Panamax
Post Panamax
Midsize
LGC
VLGC
Coastal Small
Coastal Large
Handy
Panamax
All
All
All
Coastal
Handymax
Panamax
AFRAmax
Suezmax
VLCC
SIZE RANGE (MEAN)
(DWT)
37,042-54,700 (45,104)
55,238-84,900 (67,216)
1,001-34,800 (7,048)
35,760-59,421 (50,796)
62,510-122,079 (77,898)
1,000-9,999 (3,789)
10,000-24,912 (15,673)
25,082-37,865 (29,869)
41,600-49,370 (44,511)
1,000-19,189(6,010)
1,000-19,126(6,561)
1,000-19,126(7,819)
1,000-23,853 (7,118)
25,000-39,999 (34,422)
40,000-75,992 (52,300)
76,000-117,153 (103,112)
121,109-167,294 (153,445)
180,377-319,994 (294,475)
NEWBUILD
$130.00
$165.00
$79.70
$37.50
$207.70
$33.00
$43.00
$52.00
$58.00
$478.40
$17.30
$41.20
$20.80
$59.00
$63.00
$77.00
$95.00
$154.00
Sources: Lloyd's Shipping Economist (2008), Informa (2008), Lloyd's Sea-Web (2008)
5.3.2 Fuel Market Impacts

       The market impacts for the fuel markets were estimated through the modeling performed
to estimate the fuel compliance costs for the coordinated strategy. In the WORLD model, the
total quantity of fuel used is held constant, which is consistent with the assumption that the
demand for international shipping transportation would not be expected to change due to the lack
of transportation alternatives.

       The expected price impacts of the  coordinated program are set out in Table 5-4.  Note
that on a mass basis, less distillate than residual fuel is needed to go the same distance (5 percent
less). The prices in Table 5-4 are adjusted for this impact.

       Table 5-4 shows that the coordinated strategy is expected to result in a small increase in
the price of marine distillate fuel, about 1.3 percent. The price of residual fuel is expected to
decrease slightly, by less than one percent, due to a  reduction in demand for that fuel.

                  Table 5-4 Summary of Estimated Market Impacts - Fuel Markets
FUEL
Distillate
Residual
Fuel
Switching
UNITS
$/tonne
$/tonne
$/tonne
BASELINE
PRICE
$462
$322
$322
CONTROL
PRICE
$468
$321
$468
ADJUSTED FOR
ENERGY DENSITY
N/A
N/A
$444
% CHANGE
+1.3%
-0.3%
+38.9%
                                           5-7

-------
       Because of the need to shift from residual fuel to distillate fuel in the EGA, ship owners
are expected to see an increase in their total cost of fuel. This increase is because distillate fuel is
more expensive than residual fuel.  Factoring in the higher energy content of distillate fuel,
relative to residual fuel, the fuel cost increase would be about 39 percent.

5.3.3 Marine Transportation Market Impacts

       We used the above information to estimate the impacts on the prices of marine
transportation services. This analysis, presented in Appendix 5B to this chapter, is limited to the
impacts of increases in operating costs due to the fuel and emission requirements of the
coordinated strategy.  Operating costs would increase due to the increase in the price  of fuel, the
need to switch to fuel with a sulfur content not to exceed 1,000 ppm while operating in the EGA,
and due to the need to dose the aftertreatment system with urea to meet the Tier III standards.

       Estimates of the impacts of these increased operating  costs were performed using a
representative fleet, estimated fuel consumption, actual operational  parameters, and sea-route
data for three types of ocean going vessels:  container, tanker, and cruise liner. Data obtained in
2010 from Lloyd's of London for ships that call on the proposed EGA were used to develop a
representative range of ships for this analysis.  The characteristics used to develop these
representative ships include: gross tonnes (GT), engine power (kilowatt - hour (kW-hr)), cruise
speed,  cargo and passenger capacities, and ship call data for each vessel type. Additionally, to
develop a representative sea-route for our price estimations, we created theoretical trips for both
cruise ships and for cargo carrying vessels.  Three different hypothetical cruises were developed
based on actual cruises that visit the proposed EGA;  these routes reflect travel between the U.S.
and Puerto Rico, as well as a route that travels exclusively inside the Caribbean.  The container
vessel routes developed are between the U.S. and Puerto Rico, and Singapore and Puerto Rico,
while the tanker vessel route developed is between La Guaria, Venezuela and San Juan, Puerto
Rico. All hypothetical routes and their respective representative vessels and are shown in Table
5-5 below, more detailed information is included in Appendix 5B of this chapter.

                            Table 5-5 Summary of Vessels and  Routes
TYPE OF
SHIP



Cruise Ship






Cruise Ship


ENGINE
SIZE
(KW)


22,000






53,000


ROUTE




San Juan, Puerto Rico; St. John U.S.V.I.;
Basseterre, St. Kitts; Pointe-A-Pitre,
Guadeloupe; Fort-de-France, Martinqiuqe;
St. Georges, Grenada; Bridgetown,
Barbados; St. John's, Antigua;
Frederiksted, St. Croix U.S. V.I.; San Juan,
Puerto Rico.
Fort Lauderdale, Florida; San Juan, Puerto
Rico; Matthew Town, Bahamas, Fort
Lauderdale, FL.
CARGO




800 Passengers






2,000
Passengers

NAUTICAL
MILES IN
THE
PROPOSED
EGA
300






100


TOTAL
NAUTICA
L MILES
OF THE
TRIP
1700






2,000


                                           5-8

-------
TYPE OF
SHIP
Cruise Ship
Container
Vessel
Container
Vessel
Container
Vessel
Tanker
ENGINE
SIZE
(KW)
72,000
5,000
15,785
36,540
10,000
ROUTE
New York, NY; Turk Islands; San Juan,
Puerto Rico; St. Thomas, U.S.V.I.; Fort-
de-France, Martinique; St. Georges,
Grenada; Oranjestad, Aruba; Ocho Rios,
Jamaica; Cozumel, Mexico; Key West,
Florida; New York, New York.
Miami, Florida; San Juan, Puerto Rico
Miami, Florida; San Juan, Puerto Rico
Singapore to San Juan Puerto Rico
La Guaria, Venezuela; San Juan, Puerto
Rico
CARGO
3,000
Passengers
600 TEU
1,400 TEU
6,600 TEU
780,000
Barrels
NAUTICAL
MILES IN
THE
PROPOSED
EGA
200
100
100
100
130
TOTAL
NAUTICA
L MILES
OF THE
TRIP
5,500
930
930
12,500
530
       To estimate the increase in operational costs that may be incurred as a result of this
proposed EGA, we determined the amount of fuel that would be used for each of the theoretical
routes and representative vessels shown in Table 5-5.  We then estimated what the fuel costs
would be if these vessels operated using residual fuel only, and then again if they used distillate
in the proposed EGA.  This estimation was performed assuming that the vessel would continue to
operate on residual fuel when outside of the EGA, and that approximately 33 percent of these
vessels would also use an exhaust aftertreatment technology that would require urea usage.

       The overall price differences for each of these hypothetical trips were  obtained by
subtracting the residual fuel operational costs from the calculated EGA operational fuel / urea
costs. Table 5-6 summarizes these cost increases as they relate to goods shipped and Table 5-7
summarizes these per-passenger impacts.
                                           5-9

-------
           Table 5-6 Estimated Economic Impacts of PR/VIECA for Cargo Ships (US$2006)
VESSEL
TYPE
Container
(600 TEU)
Container
(1,400 TEU)
Container
(6,600 TEU)
Tanker
(115,000
DWT;
780,000 bbl
crude)
ROUTE
Miami FL - San Juan, PR
(930 nm; 100 nm in ECA)
Miami FL - San Juan, PR
(930 nm; 100 nm in ECA)
Singapore - San Juan, PR
(12,500 nm; 100 nm in
ECA)
Venezuela - San Juan, PR
(540 nm; ISOnminECA)
PRE-ECA FUEL
COST PER TRIP
$14,900
$47,100
$1,432,000
$16,700
POST-ECA
FUEL COST
PER TRIP
$15,500
$49,000
$1,434,000
$18,200
PRICE
INCREASE PER
TEU OR BARREL
$1.00
(0.25%)
$400 base cost
$1.35
(0.34%)
$400 base cost
$0.33
(0.04%)
$800 base cost
$0.002/barrel
(negligible %)
       For these commercial vessels, the expected cost increase of shipping goods to or from
Puerto Rico, as measured by the increase in costs per TEU or per barrel of fuel, is expected to be
small, at significantly less than one percent.  We estimate that a container ship that travels
between the U.S. and the proposed ECA and operates part of the time in the ECA would see an
increase in operating costs of US$1.00 to US$1.35 per TEU, depending on the size of the ship
and the length of the route.  This represents an increase of less than one percent in the cost of
shipping a 20-foot container. A container ship operating between Singapore and Puerto Rico is
expected to see an increase in operating costs of about US$0.33 per TEU, or less than one
percent of the cost of shipping a 20-foot container. The price impacts on oil tanker services are
also expected to be small, with an estimated price  increase of less than  US$0.002 per barrel.
                                          5-10

-------
           Table 5-7 Estimated Economic Impacts of PR/VIECA for Cruise Ships (US$2006)
VESSEL
AND ROUTE
TYPE
Small Cruise
Ship (32,000
GT and 800
passengers)
Island Tour
Medium
Cruise Ship
(80,000 GT
and 2,000
passengers)
Direct Trip to
Puerto Rico
Large Cruise
Ship (120,000
GT and 3,000
passengers)
Long Tour of
the Caribbean
from the U.S.
East Coast
ROUTE
San Juan, Puerto Rico; St. John
U.S. V.I. ; Basseterre, St. Kitts;
Pointe-A-Pitre, Guadeloupe; Fort-
de-France, Martinqiuqe; St.
Georges, Grenada; Bridgetown,
Barbados; St. John's, Antigua;
Frederiksted, St. Croix U.S. V.I.;
San Juan, Puerto Rico.
Fort Lauderdale, Florida; San
Juan, Puerto Rico; Matthew
Town, Bahamas, Fort Lauderdale,
FL.
New York, NY; Turk Islands; San
Juan, Puerto Rico; St. Thomas,
U.S.V.I.; Fort-de-France,
Martinique; St. Georges,
Grenada; Oranjestad, Aruba;
Ocho Rios, Jamaica; Cozumel,
Mexico; Key West, Florida; New
York, New York.
PRE-ECA
FUEL
COST PER
TRIP
$123,000
$298,000
$987,000
POST-ECA
FUEL
COST PER
TRIP
$131,000
$303,000
$1,002,000
PRICE
INCREASE
PER
PASSENGER
PER DAY
$1.30 ($10 per
trip for the 8-day
trip)
$0.60 ($3 per trip
for the 5-day
trip)
$0.40 ($6 per trip
for the 14-day
trip)
       For similar sized cruise vessels, the expected cost increase of carrying passengers to or
from Puerto Rico, as measured by the increase in costs per passenger per cruise, is expected to be
small, at less than one percent. We estimate that a cruise ship that operates part of the time in the
ECA would see an increase in operating costs of US$0.40 to US$1.30 per passenger per night,
depending on the size of the ship, the length of the route, and the number of passengers. This
represents an increase of less than one  percent in the cost of a stateroom per night. A large cruise
ship operating between New York and Puerto Rico is expected to see an increase in operating
costs of nearly US$6 per passenger per cruise. The price on a small cruise ship cruising from
and returning to San Juan, Puerto Rico is expected to see an increase in operating costs of about
US$10 per passenger per cruise. The price impacts on a medium sized  cruise ship operating on a
nearly direct route between Fort Lauderdale,  Florida and San Juan, Puerto Rico are also expected
to be small, with an estimated price increase of less than US$3 per passenger per cruise. The
estimated increase in costs per trip per passenger incurred as a result of this proposed ECA are
substantially less than the average fuel charge currently charged to passengers if the price  of oil
per barrel exceeds a certain threshold, this surcharge can range from US$5 to US$10 per
passenger per day.
                                           5-11

-------
       Our analysis also suggests that increases in operational costs of the magnitude expected
to occur for vessels operating in the EGA are within the range of historic price variations for
bunker fuel. This is illustrated in Figure 5-1. This figure is based on variation in fuel price
among the ports of Singapore, Houston, Rotterdam, and Fujairah.
           $675
           $575
           $475
           $375
           $275
           $175
Baseline Value (Cheapest)
Most Expensive Fuel
3% Increase due to EGA
                                                        
-------
and out of U.S. foreign trade zones, the 50 states, the District of Columbia, and Puerto Rico was
about $1.4 trillion. Of that, about $1 trillion was for imports.3
                                           5-13

-------
       Appendices

                                    APPENDIX 5A

       The methodology used in this Economic Impact Analysis (EIA) is rooted in applied
microeconomic theory and was developed following U.S. EPA's recommended procedures.4
This appendix describes the economic theory underlying the analysis and how it was applied to
the problem of estimating the economic impacts of the proposed EGA on shipping engaged in
international trade.

The Economic Theory Used to Estimate Economic Impacts

       The approach used to estimate the economic impacts of the proposed EGA relies on the
basic relationships between production and consumption in competitive markets.

Multi-Market, Partial-Equilibrium Approach

       The approach is behavioral in that it builds on the engineering cost analysis by
incorporating economic theory related to producer and consumer behavior to estimate changes in
market conditions. As Bingham and Fox note, this framework provides "a richer story" of the
expected distribution of economic welfare changes across producers and consumers.  In
behavioral models, manufacturers of goods affected by a regulation are economic agents who
can make adjustments, such as changing production rates or altering input mixes, which will
generally affect the market environment in which they operate.  As producers change their
production levels in response to a new regulation, consumers of the affected goods are typically
faced with changes in prices that cause them to alter the quantity that they are willing to
purchase. These changes in price and output resulting from the market adjustments are used  to
estimate the distribution of social costs between consumers and producers.

       This is also a multi-market, partial equilibrium approach.  It is a multi-market approach
in that more than one market is examined: the markets for marine engines, vessels, and
international shipping transportation services.  It is a partial-equilibrium approach in that rather
than explicitly modeling all of the interactions in the global economy that are affected by
international shipping, the individual markets that are directly affected by the EGA requirements
are modeled in isolation.  This technique has been referred to in the literature as "partial
equilibrium analysis of multiple markets."6

       This EIA does not examine the economic impact of the proposed EGA on finished goods
that use ocean transportation services as inputs. This is because international shipping
transportation services are only a small part of the total inputs of the final goods and services
produced using the materials shipped. A change in the price of marine transportation services on
the order anticipated by this program would not be expected to significantly affect the markets
for the finished goods. So, for example, while we look at the impacts of the program on ocean
transportation costs, we do not look at the impacts of the controls on gasoline produced using
crude oil transported by ship, or on manufactured products that use petroleum products as inputs.
                                         5-14

-------
       It should also be noted that this EIA estimates the aggregate economic impacts of the
control program at the market level. This is not intended to be a firm-level analysis; therefore
compliance costs facing any particular ship operator may be different from the market average,
and the impacts of the program on particular firms can vary significantly. The difference can be
important, particularly where the rule affects different firms' costs over different activity rates.

Competitive Markets

       The methodology used in this EIA relies on an assumption of perfect competition. This
means that consumers and firms are price takers and do not have the ability to influence market
prices. Perfect competition is widely accepted for this type of analysis and only in rare cases are
other approaches used.7 Stopford's description of the shipping market and how prices are set in
this market supports this assumption.8

       In a perfectly competitive market at equilibrium with no externalities, the market price
equals the value society (consumers) places on the marginal product, as well as the marginal cost
to society (producers).  Producers are price takers, in that they respond to the value that
consumers put on the product.  It should be noted that the perfect competition assumption is not
primarily about the number of firms in a market.  It is about how the market operates: whether or
not individual firms have sufficient market power to influence the market price. Indicators that
allow us to assume perfect competition include absence of barriers to entry, absence of strategic
behavior among firms in the market, and product differentiation.A'9 Finally, according to
contestable market theory, oligopolies and even monopolies will behave very much like firms in
a competitive market if it is possible to enter particular markets costlessly (i.e., there are no sunk
costs associated with market entry or exit). This would be the case, for example, when products
are substantially similar (e.g.,  a recreational vessel and a commercial vessel).

Intermediate-Run Impacts

       This EIA explores economic impacts on affected markets in the intermediate run. In the
intermediate run, some factors of production are fixed and some are variable.  A short-run
analysis, in contrast, imposes all compliance costs on producers, while a long-run analysis
imposes all costs on consumers. The use of the intermediate run means that some factors of
production are fixed and some are variable, and illustrates how costs will be shared between
producers and consumers as the markets adjust to the new compliance program. The use of the
intermediate time frame is consistent with economic practices for  this  type of analysis.

       Short-Run Analysis

       In the very short run, all factors of production are assumed to be fixed, leaving producers
with no means to respond to the increased costs associated with the regulation (e.g., they cannot
adjust labor or capital inputs). Within a very short time horizon, regulated producers are
constrained in their ability to adjust inputs or outputs due to contractual, institutional, or other
A The number of firms in a market is not a necessary condition for a perfectly competitive market. See Robert H.
Frank, Microeconomics and Behavior, 1991, McGraw-Hill, Inc., p 333.


                                           5-15

-------
factors and can be represented by a vertical supply curve, as shown in Figure 5-2. Under this
time horizon, the impacts of the regulation fall entirely on the regulated entity. Producers incur
the entire regulatory burden as a one-to-one reduction in their profit.  This is referred to as the
"full-cost absorption" scenario and is equivalent to the engineering cost estimates.  Although
there is no hard and fast rule for  determining what length of time constitutes the very short run, it
is inappropriate to use this time horizon for this type of analysis because it assumes economic
entities have no flexibility to adjust factors  of production.  Note that the BAF is a way to avoid
this scenario. Additionally, the fact that liner price schedules are renegotiated at least annually,
and that individual service contracts may be negotiated more frequently, suggests that a very
short-run analysis would not be suitable.
             Price
                                                      Q
                       Figure 5-2 Short-Run: All Costs Borne by Producers
Output
       Long-Run Analysis

       In the long run, all factors of production are variable, and producers can be expected to
adjust production plans in response to cost changes imposed by a regulation (e.g., using a
different labor/capital mix). Figure 5-3 illustrates a typical, if somewhat simplified, long-run
industry supply function.  The supply function is horizontal, indicating that the marginal and
average costs of production are constant with respect to output. This horizontal slope reflects
the fact that, under long-run constant returns to scale, technology and input prices ultimately
determine the market price, not the level of output in the market.
                                           5-16

-------
 Price
Increase
r
                                                                  ^ With Regulation
                                                       >
                                                        Unit Cost Increase
                                                                 3 • Without Regulation
                                                                   Output
                         Figure 5-3 Long-Run:  Full Cost Pass-Through
       Market demand is represented by the standard downward-sloping curve. The market is
assumed here to be perfectly competitive; equilibrium is determined by the intersection of the
supply and demand curves. In this case, the upward shift in the market supply curve represents
the regulation's effect on production costs and is illustrated in Figure 5-3.  The shift causes the
market price to  increase by the full amount of the per-unit control cost (i.e., from P0 to PI). With
the quantity demanded sensitive to price, the increase in market price leads to a reduction in
output in the new with-regulation equilibrium (i.e., Qo to Qi). As a result, consumers incur the
entire regulatory burden as represented by the loss in consumer surplus (i.e., the area Poac PI). In
the nomenclature of El As, this long-run scenario is typically referred to as "full-cost pass-
through."

       Taken together, impacts modeled under the long-run/full-cost-pass-through scenario
reveal an important point:  under fairly general economic conditions, a regulation's impact on
producers is transitory. Ultimately, the costs are passed on to consumers in the form of higher
prices. However, this does not mean that the impacts of a regulation will have no impact on
producers of goods and services affected by a regulation.  For example, the long run may cover
the time taken to retire today's entire capital equipment, which could take decades. Therefore,
transitory impacts could be protracted and could dominate long-run impacts in terms of present
value. In  addition, to evaluate impacts on current producers, the long-run approach is not
appropriate. Consequently a time horizon that falls between the very short-run/full-cost-
absorption case and the long-run/full-cost-pass-through case is most appropriate for this EIA.

       Intermediate Run Analysis

       The intermediate run time frame allows examination of impacts of a regulatory program
during the transition between the very short run and the long run. In the intermediate run, there
is some resource immobility which may cause producers to suffer producer surplus losses.
                                           5-17

-------
Specifically, producers may be able to adjust some, but not all, factors of production, and they
therefore will bear some portion of the costs of the regulatory program.  The existence of fixed
production factors generally leads to diminishing returns to those fixed factors.  This typically
manifests itself in the form of a marginal cost (supply) function that rises with the output rate, as
shown in Figure 5-4Error! Reference source not found..
                                                                    : With Regulation
                                                                 Cost Increase
     Price
    Increase
: Without Regulation
                                            Qi          Qo

                     Figure 5-4 Intermediate-Run: Partial-Cost Pass-Through
      Output
       Again, the regulation causes an upward shift in the supply function. The lack of resource
mobility may cause producers to suffer profit (producer surplus) losses in the face of regulation;
however, producers are able to pass through some of the associated costs to consumers, to the
extent the market will allow. As shown, in this case, the market-clearing process generates an
increase in price (from PO to PI) that is less than the per-unit increase in costs, so that the
regulatory burden is shared by producers (net reduction in profits) and consumers (rise in price).
In other words, there is a loss of both producer and consumer surplus.

Economic Impacts of a Control Program - Single Market

       A graphical representation of a general economic competitive  model of price formation,
as shown in Figure 5-5  (a), posits that market prices and quantities are determined by the
intersection of the market supply and market demand curves. Under the baseline scenario, a
market price and quantity (p,Q) are determined by the intersection of the downward-sloping
market demand curve (DM) and the upward-sloping market supply curve (SM). The market
supply curve reflects the sum of the domestic (Sd) and import (Sf) supply curves.
                                          5-18

-------
                                                       =  p
            Domestic Supply
     Foreign Supply

a) Baseline Equilibrium
                                                                          Q
                             Market
       pO

       P
pD

P
               qDd   qd

            Domestic Supply
   qOf    qf

Foreign Supply
                              b) With-Regulation Equilibrium
                                      QD Q

                                  Market
                   Figure 5-5 Market Equilibrium Without and With Regulation

       With the regulation, the costs of production increase for suppliers. The imposition of
these regulatory control costs is represented as an upward shift in the supply curve for domestic
and import supply by the estimated compliance costs.  As a result of the upward shift in the
supply curve, the market supply curve will also shift upward as shown in Figure 5-5(b) to reflect
the increased costs of production.

       At baseline without the new standards, the industry produces total output, Q, at price, p,
with domestic producers supplying the amount qd and imports accounting for Q minus qd, or qf.
With the regulation, the market price increases from p to p , and market output (as determined
from the market demand curve) decreases from Q to Q .  This reduction in market output is the
net result of reductions in domestic and import supply.

       As indicated in Figure 5-5, when the new standards are applied the supply curve will shift
upward by the amount of the estimated compliance costs. The demand curve, however, does not
shift in this analysis.  This is explained by the dynamics underlying the demand curve. The
demand curve represents the relationship between prices and quantity demanded.  Changes in
prices lead to changes in the quantity demanded and are illustrated by movements along a
                                          5-19

-------
constant demand curve. In contrast, changes in consumer tastes, income, prices of related goods,
or population would lead to change in demand and are illustrated as shifts in the position of the
demand curve.B'10 For example, an increase in the number of consumers in a market would
cause the demand curve to shift outward because there are more individuals willing to buy the
good at every price. Similarly, an exogenous increase in average income would also lead the
demand curve to shift outward or inward, depending on whether people choose to buy more or
less of a good at a given price.

Economic Impacts  of a Control Program - Multiple Markets

       The above description is typical of the expected market effects for a single product
market considered in isolation (for example, the ocean transportation service market). However,
the markets considered in this EIA are more complicated because they are linked: the market for
engines is affected by the market for vessels, which is affected by the  market for international
marine transportation  services. In particular, it is reasonable to assume that the input-output
relationship between the marine diesel engines and vessels is strictly fixed and that the demand
for engines varies directly with the demand for vessels. Similarly, the demand for vessels varies
directly with the demand for marine transportation services. A demand curve specified in terms
of its downstream consumption is referred to as a derived demand curve. Figure 5-6 illustrates
how a derived demand curve is identified.
                        Price
                      Equipment
                        ($/Q)
                                          ] QE
                                                             Q .Equipment
                       Price
                       Engines
                       ($/Q)
                                                      nit Cost Increase
                                                             Derived
                                                             Demand
                                          B Qen0

                                        D QE = 0 Qe,
                                                         Q - Engines
                          Figure 5-6 Derived-Demand Curve for Engines
B An accessible detailed discussion of these concepts can be found in chapters 5-7 of Nicholson's (1998)
intermediate microeconomics textbook.
                                           5-20

-------
       Consider an event in the engine market, such as a new technology requirement, that
causes the price of an engine to increase by APeng. This increase in the price of an engine will
cause the supply curve in the engine market to shift up, leading to a decreased quantity (AQeng).
The change in engine production leads to a decrease in the demand for equipment (AQs). The
difference between the supply curves in the equipment market, S'E  - SE, is the difference in price
in the engine market, APeng, at each quantity. Note that the supply and demand curves in the
equipment market are needed to identify the derived demand in the engine market.

       In the market for vessels and engines, the derived demand curves are expected to be
vertical.  The full costs of the engines will be passed into the cost of vessels, and the cost of
vessels will be passed into the cost of ocean transportation.

Using Economic Theory to Estimate the Social Costs of a Control Program

       The economic welfare  implications of the market price and  output changes with the
regulation can be examined by calculating consumer and producer net "surplus" changes
associated with these adjustments.  This is a measure of the negative impact of an environmental
policy change and is commonly referred to as the "social cost" of a regulation. It is important to
emphasize that this measure does not include the benefits that occur outside of the market, that
is, the value of the reduced levels of air pollution with the regulation. Including this benefit will
reduce the net cost of the regulation and even make it positive.

       The demand  and supply curves that are used to project market price and quantity impacts
can be used to estimate the change in consumer, producer, and total surplus or social cost of the
regulation (see Figure 5-7).
                                          5-21

-------
                        $/Q
                                               Q2 Q,
                                (a) Change in Consumer Surplus with
                                         Regulation
Q/t
                        $/Q
                                               Q2 Q,
                                (b) Change in Producer Surplus with
                                         Regulation
                                                             Q/t
                        $/Q
                                               Q2 Q,
                              (c) Net Change in Economic Welfare with
                                         Regulation
                                                             Q/t
     Figure 5-7 Economic Welfare Calculations:  Changes in Consumer, Producer, and Total Surplus

       The difference between the maximum price consumers are willing to pay for a good and
the price they actually pay is referred to as "consumer surplus."  Consumer surplus is measured
as the area under the demand curve and above the price of the product.  Similarly,  the difference
between the minimum price producers are willing to accept for a good and the price they actually
receive is referred to as  "producer surplus." Producer surplus is measured as the area above the
supply curve below the  price of the product.  These areas can be thought of as consumers' net
benefits of consumption and producers' net benefits of production, respectively.
                                            5-22

-------
       In Figure 5-7, baseline equilibrium occurs at the intersection of the demand curve, D, and
supply curve, S.  Price is PI with quantity Qi.  The increased cost of production with the
regulation will cause the market supply curve to shift upward to S .  The new equilibrium price
of the product is P2. With a higher price for the product there is less consumer welfare, all else
being unchanged. In Figure 5-7(a), area A represents the dollar value of the annual net loss in
consumer welfare associated with the increased price. The rectangular portion represents the
loss in consumer surplus on the quantity still consumed due to the price increase, Q2, while the
triangular area represents the foregone surplus resulting from the reduced quantity consumed, Qi
-Q2.

       In addition to the changes in consumers' welfare, there are also changes in producers'
welfare with the regulatory action.  With the increase in market price, producers receive higher
revenues on the quantity still purchased, Q2. In Figure 5-7(b), area  B represents the increase in
revenues due to this increase in price. The difference in the area under the supply curve up to the
original market price, area C, measures the loss in producer surplus, which includes the loss
associated with the quantity no longer produced. The net change in producers' welfare is
represented by area B - C.

       The change in economic welfare attributable to the compliance costs of the regulations is
the sum of consumer and producer surplus changes, that is, -(A) + (B-C). Figure 5-7(c) shows
the net (negative)  change in economic welfare associated with the regulation as area D.

How the Economic Theory Applied in This EIA

       In the above explanation of how to estimate the market and social welfare impacts of a
control action, the price elasticities of supply and demand were nonzero.  This was reflected in
the upward-slope of the supply curve and the downward slope of the demand curve.  In the
derived demand analysis, a nonzero price elasticity of demand in the vessel market yielded a
nonzero price elasticity of demand in the engine market.

       However, the price elasticity of demand in the international  shipping market is expected
to be nearly perfectly inelastic (demand curve with near-infinite slope - a vertical demand
curve). This is not to say that an increase in price has no impact on quantity demanded; rather, it
means that the price increase would have to be very large before there is a noticeable change in
quantity demanded.

       The price elasticity of demand is expected to be near perfectly inelastic because there are
no reasonable alternatives to shipping by vessel for the vast majority of products transported by
sea to the United States and Canada.  It is impossible to ship goods  between these countries and
Asia, Africa, or Europe by rail or highway. Transportation of goods between these countries and
Central and South America by rail or highway would be inefficient  due to the time and costs
involved. As a result, over 90% of the world's traded goods are currently transported by sea.11
While aviation may be an alternative for some goods, it is impossible for goods shipped in bulk
or goods  shipped in large quantities. There are also capacity constraints associated with trans-
continental aviation transportation, and the costs are higher on a per tonne basis.
                                          5-23

-------
       A nearly perfectly inelastic price elasticity of demand simplifies the analysis described
above. Figure 5-8 reproduces the relationships in a multi-level market but this time with a nearly
perfectly inelastic demand curve in the international shipping market. The relationships between
this market and the markets for vessels and engines means that the derived demand curves for
engines and vessels are also expected to be nearly perfectly inelastic.  Specifically, if demand for
transportation services is not expected to be affected by a change in price, then the demand for
vessels will  also remain constant, as will the demand for engines.
                                (a) The vertical demand curve for
                                   ocean transportation market
                      O.ship
                                           '-"'o.ship
                                (b) The vertical demand curve for
                                      ocean vessel market
                     P1.
S1,eng

s,
                                             Q0,eng
                                  (c) The vertical demand curve for
                                         C-3 engine market

             Figure 5-8 Market Impacts in Markets with Nearly Perfectly Inelastic Demand

       As indicated in Figure 5-8, a change in unit production costs due to compliance with the
engine emission and fuel sulfur requirements in the  proposed EGA shifts the supply curves for
engines, vessels, and ocean transportation services.  The cost increase causes the market price to
increase by the fall amount of per unit control cost (i.e. from P0 to PI) while  the quantity
demanded for engines, vessels, and transportation services remains constant. Thus, engine
manufacturers are expected to be able to pass on the full cost of producing Tier III compliant
                                           5-24

-------
engines to the vessel builders, who are expected to be able to pass the full cost of installing the
engines and fuel switching equipment on to the vessel owners. The vessel owners, in turn, are
expected to be able to pass on these cost increases, as well as the additional operating costs they
incur for the use of SCR reductant (urea) and low sulfur fuel while operating in the EGA.

       Note that the fuel and urea costs affect the ocean transportation services market directly,
but affect the vessel and engine markets only through the derived demand curves.  That is, the
equilibrium prices and quantities for vessels and engines will change only if the quantity of
ocean transportation services demanded changes due to fuel and urea costs. Because the changes
in fuel and urea prices are expected to be too small to affect the quantity of ocean transportation
services demanded,  the markets for vessels and engines are not expected to be affected by fuel
changes.

       The sole exception for the assumption of nearly perfectly price elasticity of demand is the
cruise market. Clearly, the consumers in that market, tourists and holiday-makers,  have
alternatives available for their recreational activities.  If the cost of a cruise increases too much,
they may decide to spend their vacation in other activities closer to home, or may elect to fly
somewhere instead. As a result, the costs of compliance for the cruise industry are more likely to
be shared among stakeholders. If the price elasticity of demand is larger (in absolute value) than
the price elasticity of supply, ship owners will bear a larger share of the costs  of the program; if
the price elasticity of demand is smaller (in  absolute value) than the price elasticity of supply,
consumers will bear a larger share of the program. Similarly, the vessel builders and engine
manufacturers will also bear a portion of the costs.  If the quantity demanded for cruises
decreases, the derived quantity demanded for vessels will decrease, as will the derived quantity
demanded for engines. If the supply curves for these industries are not perfectly elastic (i.e.,
horizontal), then the downward-sloping derived demand curves will lead to shared impacts
among the sectors.

       As described in section 5.3.3 of this  chapter, the impacts on the cruise market are
expected to be small, with total engine and vessel costs increasing about one percent and
operating costs increasing between 1.5 and 6 percent. These  increases are within the range of
historic variations in bunker fuel prices.  The impact on the cruise market,  then,  may be similar
in effect to the market's response to those changes.

       Finally, it may be possible for cruise ships to offset some of these costs by advertising the
environmental benefits of using engines and fuels that comply with the EGA requirements.
Many cruise passengers enjoy this form of recreational because it allows them a personal-level
experience with the marine environment, and they may be willing to pay an increased fee to
protect that nature.  If people prefer more environmentally friendly cruises, then the demand
curve for these cruises will shift up. Consumers will be willing to bear more of the costs of the
changes. If the demand shift for environmentally friendly cruises is large enough, both the
equilibrium price and quantity of cruises might increase.
                                           5-25

-------
                                     APPENDIX 5B

Estimation of Transportation Market Impacts

       The U.S. has submitted a proposal to IMO to designate an emission control area in which
ships would need to comply with stringent fuel sulfur limits and Tier III NOx standards.  To
characterize the increase in vessel operating costs due to the proposed EGA, and therefore the
impacts on transportation market prices, calculations were performed for three types of ocean
going vessels, including: container, tanker, and cruise liner. Our estimates were developed using
typical vessel characteristics, projected fuel and urea costs, and worst case sea-route data. This
appendix presents the methodology used for these calculations.

Container and Tanker Vessels

       A series of representative container and tanker vessels were derived using data obtained
from the Lloyd's of London Sea-Web Database and Army Corps of Engineer (ACE) data.12'13
The ACE database is composed of port entrances and exits and was used to identify actual ships
that have visited the proposed EGA.  Lloyd's database was used to identify the characteristics of
these ships and to provide information on existing vessels in the world fleet including: vessel
size (Gross Tonnes (GT)), main and auxiliary engine power (kilowatt - hour (kW-hr)), number
of TEUs or barrels carried, etc. Theoretical routes were developed that these ships could travel
based on shipping lane data presented in Chapter 2 of the Technical Support Document.
Distances traveled in each route were estimated from either www.nauticaldistance.com or
Google Earth.  Table 5-5 summarizes the modeled vessel characteristics and route information.

       Operating  costs include those associated with switching from residual fuel to 0.1% sulfur
distillate fuel and urea consumption for vessels equipped with SCR.  The fuel and urea costs are
based on projections that are presented in the EGA proposal. These fuel cost estimates are
$322/tonne for residual fuel and  $468/tonne for 0.1% sulfur distillate fuel. We use a urea
consumption rate of 7.5% that of the fuel consumption rate, with a urea price estimate of
$1.52/gallon.

       To develop representative cruise ship routes for our price estimations, we looked at Army
Corps of Engineer data to find the actual makeup of the fleet of cruise ships that have visited
Puerto Rico and the U.S. Virgin  Islands, from there we researched the actual routes these vessels
take and used these routes to develop hypothetical routes. We also used the characteristics of
these actual vessels obtained from Lloyd's Sea-Web Database to develop  representative ship
configurations and numbers of passengers aboard.14

       Baseline Operating Costs

       In order to estimate the increase  in the cost to operate over select routes, we needed to
establish the fuel usage and costs for our baseline route  (i.e. the price of the route operating on
residual fuel). We determined average operational values for our hypothetical vessel by
selecting the mid-point of the operational ranges used today on  cargo vessels and tankers.
Baseline estimations of the fuel used for the routes and ships were determined by multiplying the
engine power of the average sized containership (in kilowatts (kW)) by the average estimated
                                          5-26

-------
engine efficiency, and the appropriate brake-specific fuel consumption (BSFC) value consistent
with the inventory analysis in Chapter 2 of this document (see Equation 5B-1 below). This value
was then multiplied by the distance of the trip, and divided by the average vessel speed to find
the total fuel consumed over the trip, see Equation 5B-2. As average values are represented here,
it is possible that these values could vary slightly from actual measured values depending on a
vessel's speed, engine efficiency, and specific fuel consumption, but we believe that these
estimates provide a reasonable forecast of container vessels in operation today with similar
characteristics as those modeled here.

       Equation 5B-1


Engine _ Power (kW] x Engine _ Efficiency x BSFC (^residA.ur     ) = Fue^ - Consumption _ Rate (^resid
                                                     .ur   /,        -            _
       Equation 5B-2

 „  ,  _        .     „   ,0-  ,  A*     Distance(nm)          tonne     „  ,   _       ,,        ,
rue I  Consumption  Katel^™"1/,  )* - - - -. — x - = rue I  Consumed (tonne  .,)
                             /hr   Vessel _Speed(*nots/hr)  l.OOO.OOOg
       Total fuel usage for each leg of the trip was multiplied by the price of the fuel in 2006
U.S. dollars per tonne ($/tonne) which provides the baseline cost of fuel for each leg. These
costs were then summed to produce an aggregate estimation of fuel cost for the entire trip.  This
analysis shows a per trip fuel cost of nearly $15,000 for a small container ship traveling a direct
route between Miami, Fl and San Juan Puerto Rico. This analysis also shows a per trip fuel cost
of over $1.4 million for a large container ship traveling between Singapore and San Juan Puerto
Rico.

       Operating Costs with an ECA

       Operating cost increases due to an ECA  are due to increased fuel costs and urea
consumption within the ECA. Operating costs are assumed to remain unchanged outside of the
ECA. In addition, the ECA is assumed to have no impact on the route travelled for vessels
visiting the proposed ECA.

       Increased Fuel Costs

       To determine  the estimated fuel usage and increase in fuel costs incurred as a result of the
proposed ECA for representative vessels traveling their respective theoretical routes, we used the
same methodology as in our baseline analysis with the appropriate distillate fuel properties.
Since distillate fuel will most likely only be used in the proposed ECA, the remainder of the trip
is assumed to continue to operate using residual fuel which is reflected in this analysis.  Equation
5B-3 provides the approximation of the amount of distillate fuel used per hour given a ship's
engine power and fuel consumption. Due to the chemical properties of the two marine fuels,
there is approximately a five percent (5%)  increase in energy, on a mass basis, when operating
                                          5-27

-------
on the distillate fuel instead of the residual fuel, and this increase is accounted for in Equation
5B-3. Equation 5B-2 was then used to estimate the actual tonnes of distillate fuel used.


       Equation 5B-3


                                           R<\Fr(8'distil/       )                                   .
                                           -LJkJ-L \^ y    /T T-jr   J  J                                   /
Engine _ Power (kW] x Engine _ Efficiency x	      ~    = Fuel _ Consumption _ Rate (^disti1/ )
                                                 1 H U.UO
       Urea Costs

       Switching to a distillate marine fuel will achieve reductions only in sulfur and particulate
emissions. In order to meet the required Tier III Nitrogen Oxides (NOx) emission reductions,
new vessels built as of 2016 may be equipped with SCR.C Using an SCR  system requires dosing
exhaust gases with urea, which adds some additional costs to the operation of the vessel.  Urea
consumption for vessels equipped with SCR is expected to be 7.5 percent of the fuel
consumption.  The urea operational costs are based on a price of $1.52 per gallon with a density
of 1.09 g/cc. The cost per gallon was estimated for a 32.5 percent urea solution delivered in bulk
to the ship through research completed by ICF International  for the U.S. Government, combined
with historical urea price information.15'  '17'18'19 The estimated cost of using urea is based on
an estimated dosing rate of seven and a half percent (7.5%) per gallon of distillate fuel used.
Subsequently, to estimate the volume of urea required for our routes, we multiplied the distillate
quantity determined above  by the estimated urea consumption value. As we expect these costs to
be incurred several years in the future, we used the analysis performed for the EPA by EnSys
which predicted that in 2020, 33.2% of the fuel used in EGAs will be on vessels equipped SCR.20

       Total Increase in Operating Costs

       To estimate the total increase in the operating costs of a vessel incurred while operating
in the proposed EGA, we then multiplied the fuel and urea quantities used by their corresponding
prices ($322.48/tonne for residual, $467.92/tonne for distillate, and $1.52/gal for the urea).  In
order to estimate how the increase in operational costs may affect the price per TEU, we divided
the increase in cost by the number of TEUs each representative ship would carry (or in the case
of a Tanker  Vessel - the number of barrels of oil).

Cruise Ship

       We also conducted  an analysis to determine the estimated increase in operating costs for
different cruise ships that may visit the proposed EGA. To conduct this analysis, we used ship
c As an alternative, an exhaust gas cleaning device (scrubber) may be used. This analysis does not include the
effect on distillate fuel demand of this alternative approach. It is expected that scrubbers would only be used in the
case where the operator determines that the use of a scrubber would result in a cost savings relative to using
distillate fuel. Therefore we are only estimating the cost of compliance using distillate fuel here as we believe this is
the most likely approach.


                                            5-28

-------
characteristics and route data from actual vessels that travel to Puerto Rico and the U.S. Virgin
Islands. Three cruise ship vessels were developed with representative vessel characteristics
including: engine power, GT, number of passengers, vessel speed, and fuel consumption rates.
A separate hypothetical route was developed for each representative ship type. A hypothetical
route that a small cruise  ship may take was developed based on actual routes and ports visited by
cruise ships today.14 The itinerary includes: San Juan, Puerto Rico; St. John U.S.V.I.; Basseterre,
St. Kitts; Pointe-A-Pitre, Guadeloupe;  Fort-de-France, Martinqiuqe; St. Georges, Grenada;
Bridgetown, Barbados; St. John's, Antigua; Frederiksted, St. Croix U.S. V.I.; San Juan, Puerto
Rico, and an example is  shown in Figure 5-9 below.
                  Figure 5-9 Hypothetical Route Developed for a Small Cruise Ship

       The hypothetical route that a medium sized cruise ship may take was also based on actual
routes and ports visited by cruise ships today.14 The route was developed to model a nearly
direct trip between Puerto Rico and Florida and includes the following stops: Fort Lauderdale,
Florida; San Juan, Puerto Rico; Matthew Town, Bahamas, Fort Lauderdale, FL; an example is
shown below in Figure 5-10.
                                           5-29

-------
              Figure 5-10 Hypothetical Route Developed for a Medium Sized Cruise Ship

       A hypothetical route that a large cruise ship may take was developed based on actual
routes and ports visited by cruise ships today.14 This route was developed to represent a long
cruise taken from the East Coast of the U.S. throughout the Caribbean. The itinerary includes:
New York, NY; Turk Islands; San Juan, Puerto Rico; St. Thomas, U.S.V.I.; Fort-de-France,
Martinique; St. Georges, Grenada; Oranjestad, Aruba; Ocho Rios, Jamaica; Cozumel, Mexico;
Key West, Florida; New York, New York and an example is shown in Figure 5-11 below.
                                          5-30

-------
                 Figure 5-11 Hypothetical Route Developed for a Large Cruise Ship

       In order to estimate the amount of fuel used during these hypothetical routes, the mileage
during each leg of the journey was estimated, and used in conjunction with average main and
auxiliary engine power, average cruise speeds, and brake specific fuel consumption. The
average cruise speed for each representative ship was derived from data on similar sized vessels
that visit the Caribbean. The brake specific fuel consumption values used were from the
inventory chapter of this document (Chapter 2) where 195 g/kW-hr was used for large slow-
speed diesel engines such as those found in large cruise ships, and 210 g/kW-hr was used for
medium-speed diesels found in the small and medium sized cruise ships and also used for all
auxiliary engines. The required power estimation used here was developed for the "2005-2006
BC Ocean-Going Vessel Emissions Inventory" and was shared with several cruise ship operators
for their input and validation.21  This relationship was developed to approximate effective power
given cruise ships' diesel-electric operation. The auxiliary engines reported within the Lloyd's
of London 'Seaweb' database are presumably operated independently of the vessel's main
diesel-electric power generation, and are assumed to operate at an average of 50% power for the
entire voyage.
                                          5-31

-------
                       Table 5-8 Representative Cruise Ship Characteristics
VESSEL
TYPE
Small
Cruise
Ship
Medium
Cruise
Ship
Large
Cruise
Ship
ROUTE

San Juan, Puerto Rico; St. John U.S.V.I.;
Basseterre, St. Kitts; Pointe-A-Pitre,
Guadeloupe; Fort-de-France, Martinqiuqe; St.
Georges, Grenada; Bridgetown, Barbados; St.
John's, Antigua; Frederiksted, St. Croix U.S.
V.I.; San Juan, Puerto Rico.
Fort Lauderdale, Florida; San Juan, Puerto Rico;
Matthew Town, Bahamas, Fort Lauderdale, FL.

New York, NY; Turk Islands; San Juan, Puerto
Rico; St. Thomas, U.S.V.I.; Fort-de-France,
Martinique; St. Georges, Grenada; Oranjestad,
Aruba; Ocho Rios, Jamaica; Cozumel, Mexico;
Key West, Florida; New York, New York.
MAIN
ENGINE
POWER
22,000
kW

53,000
kW

72,000
kW

Auxiliary
Engine
Power
4,100 kW

l,500kW

2,000 kW

Gross
Tonnage
32,000

80,000

120,000

Vessel
Maximum
Speed
(knots)
22

23

24

Number of
Passengers
800

2,000

3,000

The methodology used above to estimate fuel and urea costs (see Equation 5B-1 through 5B-3)
were also used here. Additionally, the operational cost increases for the fuel used by auxiliary
engines were estimated as well as the cost increases incurred as a result of dosing the engine
exhaust with urea using the same methodology as for main propulsion engines. The total
estimated price increase for the cruise was divided by the length of the  cruise to estimate the
increased cost per day.

       To put the estimated price increases in perspective, we also developed the percent
increase for the various stateroom types available on the vessel.  The estimated stateroom prices
used for the different hypothetical cruises are shown in Table 5-9.

        Table 5-9 Representative Cruise Liner Stateroom Prices and Estimated Increase in Prices
CRUISE SHIP TYPE
Large Cruise Ship
Medium Cruise Ship



Small Cruise Ship



STATEROOM TYPE
Interior
Ocean View
Balcony
Suite
Interior
Ocean View
Balcony
Suite
Interior
Ocean View
Balcony
Suite
ORIGINAL AVERAGE PRICE
PER NIGHT ($)
$100
$130
$150
$220
$100
$140
$200
$240
$200
$230
$290
$450
PERCENTAGE
INCREASE
0.4%
0.3%
0.2%
0.2%
0.6%
0.4%
0.3%
0.3%
0.6%
0.6%
0.4%
0.3%
                                            5-32

-------
1 Harrould-Koleib, Ellycia.  Shipping Impacts on Climate:  A Source with Solutions.  Oceana, July 2008. A copy of
this report can be found at
http://www.oceana.org/fileadmin/oceana/uploads/Climate_Change/Oceana_Shipping_Report.pdf

2 Stopford, Martin.  Maritime Economics, 3rd Edition. Routledge, 2009. p. 163.

3 Census Bureau's Foreign Trade Division, U.S. Waterborne Foreign Trade by U.S. Custom Districts, as reported by
the Maritime Administration at
http://www.marad.dot.gov/library_landing_page/data_and_statistics/Data_and_Statistics.htm , accessed April 9,
2009.

4 U.S. EPA.  "OAQPS Economic Analysis Resource Document."  Research Triangle Park, NC:  EPA 1999. A copy
of this document can be found at http://www.epa.gov/ttn/ecas/econdata/6807-305.pdf: U.S. EPA "EPA Guidelines
for Preparing Economic Analyses." EPA 240-R-00-003. September 2000. A copy of this document can be found
at http://yosemite.epa.gov/ee/epa/eed.nsf/webpates/guidelines.html

5 Bingham, T.H., and T.J. Fox.  "Model Complexity and Scope for Policy Analysis."  Public Administration
Quarterly, 23(3), 1999.

6 Berck, P., and S. Hoffman. "Assessing the Employment Impacts." Environmental and Resource Economics
22:133-156.  2002.

7 U.S. EPA "EPA Guidelines for Preparing Economic Analyses."  EPA 240-R-00-003. September 2000, p. 113. A
copy of this document can be found at http://yosemite.epa.gov/ee/epa/eed.nsf/webpates/guidelines.html

8 Stopford, Martin.  Maritime Economics, 3rd Edition. Routledge, 2009. See Chapter 4.

9 Robert H. Frank, Microeconomics and Behavior, 1991, McGraw-Hill, Inc., p 333.

   Nicholson, W., Microeconomic  Theory: Basic Principles and Extensions, 1998, The Dryden Press, Harcourt
Brace College Publishers.

   UN Conference on Trade and Development (UNCTAD), Trade and Development Report, 2008, Geneva.

12 http://www.ndc.iwr.usace.army.mil/wcsc/wcsc.htm

13 www.sea-web.com  Lloyd's

14 The following websites were accessed to collect cruise ship route data: (1) Holland American Line, Cruise
Destinations. Retrieved May, 2010, from www.hollandamerica.com, (2) Royal Caribbean International, Plan a
Cruise: Destinations. Retrieved May, 2010, from http://www.royalcaribbean.com, (3) Carnival Cruise Lines,  Find A
Cruise. Retrieved May, 2010, from www.carnival.com, (4) Celebrity Cruises, Destinations. Retrieved May, 2010,
from.www.celebritycruises.com, (5) MSC Cruises, Our Destinations. Retrieved May, 2010, from
www.msccruises.com, (6) Crystal Cruises, Destinations. Retrieved May, 2010 from, www.crystalcruises.com, (7)
Ferries Del Caribe, M/S Caribbean Express.  Retrieved May 2010, from, www.ferriesdelcaribe,  (8) Happy Cruises,
Happy Cruises - 2010 Season.  Retrieved May, 2010, from www.happycruises.eu, (9) Azamara Club Cruises,
Destinations. Retrieved, May 2010, from www.azamaraclubcruises.com, (10) Princes Cruises, Find  Cruises.
Retrieved, May 2010, from  www.princess.com, (11) Regent Seven Seas Cruises, Find a Cruise.  Retrieved, May
2010, from www.rssc.com,  (12) Silversea Cruises, Plan a Voyage. Retrieved May, 2010, from, www.silversea.com,
(14) Windstar Cruises, Cruises.  Retrieved May, 2010, from, www.windstarcruises.com, (15) Cunard Line,
                                                 5-33

-------
Destinations. Retrieved May, 2010, from, www.cunard.com, (16) Costa Crociere S.p.A, Destinations. Retrieved
May, 2010, from, www.costacruise.com, (17) Disney Cruise Line, Cruises & Destinations. Retrieved May, 2010,
from, www.disneycruise.disney.go.com, (18) Norwegian Cruise Line, Destinations. Retrieved May, 2010, from,
www2.ncl.com, (19) P & 0 Cruises, Find a Cruise. Retrieved May, 2010, from, http://www.pocruises.com, (20)
The Yachts of Seabourn, Destinations.  Retrieved May, 2010, from, www.seabourn.com, (21) SeaDream Yacht
Club, Destinations.  Retrieved May, 2010, from, www.seadreamyachtclub.com

15ICF International, "Costs of Emission Reduction Technologies for Category 3 Marine Engines," prepared for the
U.S. Environmental Protection Agency, December 2008. EPA Report Number: EPA-420-R-09-008.

16 "Nonroad SCR-Urea Study Final Report" July 29, 2007 TIAX for Engine Manufacturers Association (EMA) can
be found at:http://www.enginemanufacturers.org/admin/content/upload/198.pdf

17 http://www.adblueonline.co.uk/air_l/bulk_delivery

18 http://www.factsaboutscr.com/documents/IntegerResearch-Ureapricesbackto20051evels.pdf

19 http://www.fertilizerworks.com/fertreport/index.html

20 EnSys Navigistics, "Analysis of Impacts on Global Refining & C02 Emissions of Potential MARPOL
Regulations for International Marine Bunker Fuels," Final Report for the U.S. Environmental Protection Agency, 26
September 2007,

21
  http://www.cosbc.ca/index. php?option=com_docman&task=doc_view&gid=3&tmpl=component&format=raw&It
emid=53
                                                5-34

-------