* — \
*1 PROt^
Memorandum: Disaggregation of Category 1 /
Category 2 Commercial Marine Vessel
Emissions for 2011

-------

-------
EPA-454/B-20-025
November 2013
Memorandum: Disaggregation of Category 1 / Category 2 Commercial Marine Vessel
Emissions for 2011
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Research Triangle Park, NC

-------
VERG
www.erg.com
MEMORANDUM
TO:	Laurel Driver/US EPA
FROM: Roger Chang, Heather Perez and Richard Billings/ERG
DATE: November 20, 2013
SUBJECT: Disaggregation of Category 1 / Category 2 Commercial Marine Vessel Emissions
for 2011
1.0 Introduction
The Emission Inventory and Analysis Group (EIAG) annually produce the National
Emission Inventory (NEI). The NEI compiles comprehensive emissions data for criteria and
HAPs for mobile, point, and nonpoint sources. For this project ERG is revising the 2011 version
of the commercial marine vessel (CMV) component of the NEI to provide a more detailed
emission estimates for marine vessels equipped with Category 1 and 2 propulsion engines. This
memo specifically addresses the disaggregation of the Category 1/2 emissions into vessel type
data. Were additional details could be obtained allocation profiles for categories such as ferries,
tugs and towboats and deepwater vessels. This memorandum is a deliverable for Task 7 of Work
Assignment 3-01 for EPA Contract No. EP-D-11-006, Work Assignment 3-01.
2.0	Methodology
2.1	Disaggregation into vessel types
EPA's Office of Transportation and Air Quality (OTAQ) provided 2011 C1/C2 emissions
estimates used only two source classification codes (SCCs). The goal of the current project was
to disaggregate OTAQ's emissions into individual vessel types including deepwater, ferries,
fishing, government, Great Lakers, support (offshore and research) vessels, and tugs.
The report entitled " Category 2 Vessel Census, Activity, and Spatial Allocation
Assessment and Category 1 and Category 2 In-port At-sea Splits," (Census Report) February 16,
2007, developed activity profiles by vessel type that were allocated to underway shapes
throughout the United States. That report served as the basis for allocation of emissions by
activity data (i.e., kw-hrs) by vessel types in the 2011NEI version 1. Table 1 summarizes the
reallocated emissions data by vessel type.
1

-------
Table 1. Comparing Emission Estimates from Previous to Revised 2011
Emission Inventory
Vessel Type
Pollutant
2280002100
2280002200
Port
Underway
Previous 2011 Emissions (TPY)
Total
NOx
85,214
640,143
Disaggreg
ated 2011 Emissions (TPY)
Deepwater
NOx
575
56.819
Ferries
NOx
20,489
11,032
Fishing
NOx
3,677
69,793
Government
NOx
18,274
12,722
Great Lake
NOx
300
29,692
Support (Offshore & Research)
NOx
12,892
318,571
Tugs
NOx
29,007
141,513
Total
NOx
85,214
640,143
This approach also allows more in-depth investigation in emission patterns across the
United States between different vessel types, as shown in Figure 1 (note revised spatial
allocations for tugs, ferries, and deep water Category 1 and 2 vessels are presented in the sections
to follow).
Fishing Vessels
Government Vessels
Great Lakes Vessels	Support Vessels
Figure 1. 2011 Volatile Organic Chemical (VOC) Emissions by CI /C2 Vessel Type
2

-------
2.2 Ferries
The report entitled " Category 2 Vessel Census, Activity, and Spatial Allocation
Assessment and Category 1 and Category 2 In-port/At-sea Splits," (Census Report) February 16,
2007, developed activity profiles by vessel type that were allocated to underway shapes
throughout the United States. This report served as the basis for the 2011 CI / C2 CMV
component of the NEI, this included allocating activity by vessel type to the underway and port
shapes currently within EIS.
However, later review indicated that the ferry-related emissions were allocated to a very
limited number of ports, resulting in artificially high emissions at some ports and no ferry
emissions at other ports. The Bureau of Transportation Statistics' National Census of Ferry
Operators (NCFO 2011) database provides ferry operation data from a nationwide census of
ferry operators along with other sources of ferry data such as the U.S. Coast Guard and the U.S.
Army Corps of Engineers. This relational database format provides for the reporting of the
information at various levels, such as by operator, route segment, terminal, or vessel. For this
effort, the ferry emissions for underway and port were reallocated to a greater number of ports
based on the number of terminals at each port which ferries visit.
The NCFO provides data by individual terminals and includes a terminal name, city, and
coordinates to further identify their locations. These terminals were matched to ports within EIS
via name matching, GIS mapping, and manual investigation. This process succeeded in assigning
1,370 of the 1,458 terminals total (94%) to a port within EIS. The 88 terminals that remained
were further investigated for inclusion in this inventory effort. Some of them lacked important
identifying information such as city or coordinate locations, while others had complete location
information but appeared to have no activity according to trip count data in the NCFO.
The terminals were assigned to 105 new ports and 113 existing ports. The additional 105
ports were mapped and assigned new ShapelDs for inclusion in EIS. These additional ports were
mapped as small circles with a radius of 0.25 mile using NFCO coordinates as the centroid of the
port. They were then spatially adjusted as needed to avoid overlapping with shipping lane areas
or other port shapes. Ultimately, 218 ports were selected for use in the ferry emission
reallocation, a significant improvement over the 34 ports used in the Census report.
Once the universe of ferry ports was determined, ERG reallocated national emissions to
these ports based on the number of terminals at each port. The 218 ports used in this allocation
effort represented 1,370 terminals total. National ferry emissions were allocated to each port
based on the port's terminal count with respect to the national terminal total of 1,370, as seen in
the following equation.
Sample Calculation for VOC Port Ferry Emissions for Cheboygan, MI
Port Emissions = National Emissions x Terminals at this port / Total Terminals
Nationwide
0.68 tons = 468.71 ton x 2 / 1,370
3

-------
In lieu of better underway activity data, the ferry underway emissions followed an
allocation method similar to the original census report: underway activity was allocated to the
underway shapelD closest to each port using the same proportion derived from the port
reallocation above. Activity and fuel were also multiplied by this proportion to ensure activity
remains consistent with emissions. Figure 2 shows the revised spatial allocation for ferry
activities.
Figure 2. 2011 C1/C2 Ferry Allocation (VOC emissions)
if, J
M


V'
\

1 *
1 V
<



% 1
>

4
"7^
2011 VOC Emissions (tons)







< 0.6 tons
> 7 tons
4

-------
2.3 Tugs and Towboats
Previous tug spatial allocation methods assigned emissions to navigable waterways based
on total cargo tonnage as a surrogate for vessel traffic. Additional review indicated that, because
this method includes all vessel types, total cargo tonnage is not an appropriate surrogate in some
cases. For example, in the Great Lakes cargo tonnage is predominantly from other vessels such
as lakers; therefore, the use of total cargo tonnage for spatially allocating underway tug activity
nationally assigns activity and emissions in locations where there are no tug operations and under
reports activity and emission where there are activities.
Data from the 2011 Waterborne Commerce of the United States (WCUS) (USACE,
2011) has cargo tonnage broken out by three vessel types: tugs, dry cargo, and tankers. The
dataset is primarily composed of tug activity data and serves as a better surrogate for spatially
allocating national tug underway emissions to waterway segments. Again, the national emissions
values and the port emissions do not change; this revision targets only spatial reallocation of tug
underway emissions.
¦H




J
\ f

A? .


A,


• /
V a ri \


V A
I
„ J*s
- \
V

* 864 trips
,242,
,34 «ps
Figure 3. 2011 Waterborne Commerce of the United States Trip Count Data
The WCUS dataset provides vessel trip count per waterway segment. National emissions
were assigned to waterway segments based on the proportion of total trips represented by each
segment, as shown in the following equation:
5

-------
WWEm = NatEm x (WWTrips / NatTrips)
Where:
WWEm = Waterway underway emissions
NatEm = National tug underway emissions
WWTrips = Trips associated with the waterway segment
NatTrips = Total trips Nationwide
5.46 tons of VOC = 3,237 Tons of VOC x (9,346 trips on Detroit River, MI waterway/
5,536,686 national trips)
Waterway segments are associated with one or more underway shapelDs available in
EIS. To apportion the waterway4evel emissions to individual shapelDs, the waterway segments
were spatially intersected with the underway shapes, and the length of each segment portion was
calculated. Waterway underway emissions were calculated as follows:
ShapeEm = WWEm x WWLength / WWLengthiotai *
Where:
WWEm	= Waterway underway emissions
ShapeEm	= ShapelD Emissions
WWLength = Waterway length per segment within ShapelD
WWLengthiotai = Total Waterway length
Table 2. Sample Allocation of Waterway Segment-level Emissions to Individual EIS
ShapelDs



Lengt
h

Waterway
Segment
Waterway
Name
EIS
ShapelD
(Miles
)
Percent of
Waterway Emissions
3301
DETROIT RIVER MI
2498
12.91
53.37%
3301
DETROIT RIVER MI
2496
9.92
41.01%
3301
DETROIT RIVER MI
2497
1.34
5.54%
3301
DETROIT RIVER MI
2494
0.02
0.08%
Total
24.19
100.00%
2.91 tons of VOC for ShapelD 2498 = 5.46 tons of VOC * (12.91 miles / 24.21 miles)
The allocation percentages derived for emissions were also applied to activity and fuel
data to ensure that the emissions data were consistent with these activity elements at the
individual underway shapelD level. After processing, emissions and activity were summed to the
shapelD level to provide total emissions per underway shape. The result of this process is a
6

-------
national allocation that closely resembles the original waterborne commerce data and represents
reasonable activity patters across the United States.
Wr
PL /

Figure 4. 2011 C1/C2 Tug VOC Emissions
2.4 Deepwater Vessels
Previous deepwater spatial allocation methods assigned emissions to navigable
waterways based on total cargo tonnage as a surrogate for vessel traffic. Additional review
indicated that, because this method includes all vessel types, total cargo tonnage is not an
appropriate surrogate in some cases. For example, in the Great Lakes cargo tonnage is
predominantly from other vessels such as lakers; therefore, the use of total cargo tonnage for
spatially allocating underway tug activity nationally assigns activity and emissions in locations
where there are no tug operations and under reports activity and emission where there are
activities.
Army Corps of Engineers' Entrance and Clearance (E&C) data (USACE E&C, 2011) has
individual vessel movements for each port visited in the U.S.; this included 1,859 foreign ports
and 199 U.S. ports. To streamline the process, trips were grouped by origination and destination
(O/D) pairs; and for foreign trips, the busiest port in each country was used as surrogate for all
traffic to and from that country. Canada and Mexico-bound trips were aggregated into east coast,
west coast, and Great Lakes routes.
For this C1/C2 allocation of deepwater vessels emissions and activities the E&C data
were matched with vessel characteristics data to remove Category 3 vessels. The remaining
2011 VOC Emissions (tons)
7

-------
C1/C2 0/D pairs were mapped using network analyst tools in a GIS to determine route paths
along navigable shipping lanes from the USAGE waterway network (USACE, 2013). For foreign
trips, ERG extended the waterway network to generate paths to major foreign ports. A portion of
the route network is shown in Figure 5.
Figure 5. C1/C2 Routes Derived from 2011 Entrance and Clearance Data
Again, the national emissions values did not change; this revision targets only spatial
reallocation of deepwater underway emissions. The E&C dataset provides vessel trip count per
route for both domestic and foreign routes. The domestic route trip counts are adjusted for
double counting by dividing vessel trip counts by two. Due to the nature of the E&C dataset,
each domestic route appears in the dataset twice: once as a departure from the originating port
and again as an arrival at the destination port. Each foreign route, however, only appears in the
dataset once, with a domestic port of origination or a foreign destinati on. The adjustment to the
domestic trip counts per route corrects the discrepancy between domestic and foreign routes. It
should be noted that the E&C data is being used as a surrogate for C1/C2 traffic patterns and
may not include of all possible C1/C2 vessel trips.
Routes are associated with one or more underway shapeEDs available in EIS. The routes
were spatially intersected with the underway shapes, and the length of each segment portion was
calculated. The total distance traveled along each route and shapelD segment was calculated by
multiplying the number of vessel trips along each route by the length of each route. National
deepwater underway emissions were assigned to routes based on the proportion of the national
distance traveled represented by each segment, as shown in the following equation:
8

-------
REm = NatEm x ((LengthRiSj x TripRiSj) / E (* LengthRiSj x TripsRiSj))
Where:
REm
NatEm
TripRiSj
LengthRiSj
Underway emissions by shapelD (tons/year)
National deepwater underway emissions (tons/year)
Trips associated specified Route i and segment shapelD j
Length of segment for shapelD j for specified Route I (km)
Example: The segment of the route from Juneau, AK to Skagway, AK that intersects with
shapelD 152 is 50.6 kilometers long. C1/C2 deep water vessels traveled this route 20 times in
2011. The total national kilometer-trips traveled in 2011 is 2,422,139 kilometers. The proportion
of the national deepwater underway VOC emissions assigned to this segment is calculated
below:
0.54 Tons of VOC = 1,299 Tons of VOC x ((50.6 km x 20 trips) / 2,422,139 km-trips)
The allocation percentages derived for emissions were also applied to activity and fuel
data to ensure that the emissions data were consistent with these activity elements at the
individual underway shapelD level. After processing, emissions and activity were summed to the
shapelD level to provide total emissions per underway shape. The result of this process is a
national allocation that represents reasonable activity patterns across the United States, shown in
Figure 6. Total C1/C2 VOC emissions for all vessel types are shown in Figure 7.
9

-------
2011 Deepwater VOC Emissions
Figure 6. 2011 Deepwater C1/C2 Underway VOC Emissions
10

-------
Total 2011 C1C2 VOC Emissions
Figure 7. Total 2011 C1/C2 Underway VOC Emissions
3.0 References
U.S. Army Corps of Engineers, Vessel Entrances and Clearances (E&C) Calendar Year 2011,
http ://www. navi gati ondatacenter. us/d ata/datacl en. htm.
U.S. Army Corps of Engineers, National Waterway Network, 2013,
http://www.navigationdatacenter.us/data/datanwn.htm.
U.S. Army Corps. Of Engineers, Waterborne Commerce of the United States (WCUS), 2011.
http ://www. navi gati ondatacenter us/data/ datawcus. htm. March 2013.
U.S. Department of Transportation/Bureau of Transportation Statistics: National Census of Ferry
Operators (NCFO) 2011.
http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/subiect areas/ncfo/index.html. March
2013.
U.S. Environmental Protection Agency/ Office of Transportation and Air Quality; Category 2
Vessel Census, Activity, and Spatial Allocation Assessment and Category 1 and Category 2
In-port At-sea Splits, February 16, 2007.
11

-------
United States	Office of Air Quality Planning and Standards	Publication No. EPA-454/B-20-025
Environmental Protection	Air Quality Assessment Division	November 2013
Agency	Research Triangle Park, NC

-------