* — \
*1 PROt^
Memorandum: Development of 2011 Railroad
Component for National Emissions Inventory

-------

-------
EPA-454/B-20-024
September 2012
Memorandum: Development of 2011 Railroad Component for National Emissions Inventory
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Research Triangle Park, NC

-------
fcERG
www.erg.com
MEMORANDUM
TO:	Laurel Driver/US EPA
FROM: Heather Perez, Susan McClutchey, and Richard Billings/ERG
DATE: September 5, 2012
SUBJECT: Development of 2011 Railroad Component for National Emissions Inventory
1.0 Introduction
As part of Work Assignment 5-07 under EPA Contract EP-D-07-097, entitled "Mobile
Source Emission Inventories - FY12," ERG developed growth factors for Class I and Class II
and III railroads that were applied to the 2008 emission values developed for the National
Emission Inventory (NEI) to approximate emission levels in 2011. The emissions were allocated
to line haul shape IDs and yard locations based on 2008 allocations. ERG provided the EPA with
the 2011 estimated railroad emissions as an Access database for inclusion into EIS staging tables
by the EPA WAM.
This report documents the development of the growth factors (Section 2), application of
these growth factors to 2008 data are discussed in Section 3 along with a summary of the 2001
emissions estimates. Lastly A listing of references used in this study are presented in Section 4.
2.0 2008/2011 Railroad Growth Factors
Railroad freight traffic data were obtained from a variety of sources including the
Department of Transportation's Bureau of Transportation Statistics (BTS) and Surface
Transportation Board, The Department of Energy's Annual Energy Outlook (AEO), the
American Association of Railroads (AAR), and the American Short Lines and Regional Railroad
Association.
Initially growth rates were reviewed as reported for the AEO's 2012 reference case.
These rates were developed relative to billions of ton miles traveled. Data that specifically
covered the period from 2008 to 2011 were not included in the data table, so earlier reports
(2008-2011) were compiled and reviewed. When the separate reports were evaluated together,
the actual and projected annual growth rates were inconsistent. This observation suggested that
the commercial rail freight market had a high degree of uncertainty associated with it for the
study period.
1

-------
This observation about the volatility of the market was substantiated when aggregated
quarterly rail traffic data from the BTS were reviewed. Their data showed a peak in 2008
followed by a significant decline in activity for multimodal rail traffic. Gradually the rail freight
market was returning to the 2008 activity level, though the 2011 data point suggests that
activities were slightly under the 2008 peak. Note that the rail traffic data presented in Figure 1 is
for intermodal freight traffic and is only provided as a general indicator of rail activities.
Billions of revenue ton-miles
480
440
400
360
320
280
240 4	
2O02Q1
2004 Q1 2006 Q1
Figure 1. Intermodal Rail Traffic 2002 to 2012
20C8Q1
2012 Q1
2010 Q1
ERG compiled railroad freight traffic data from the 2008 and 2011 R-l reports submitted
by all Class I rail lines to the Surface Transportation Board. The R-l data are more
comprehensive than the BTS's intermodal study as it includes all freight shipped by Class I
railways. For the most part, the R-l data follow the Trends noted in Figure 1 with one exception:
the Soo Line saw an increase in traffic of over 45%. The Soo Lines business activities were
investigated to understand the anomaly in their freight traffic data. The Soo Line acquired the
Dakota, Minnesota and Eastern Railroad and the Iowa, Chicago & Eastern Railroad in late 2008.
It is believed that this merger is the cause of the increased activity reported by the Soo line.
When all the rail traffic data are aggregated, the Soo line's freight traffic has a relatively
small impact on the overall trend as it is the smallest of the Class I railways. The percent change
from 2008 to 2011 for total ton-miles (including the Soo line) is -2.479.
2

-------
Table 1. Class I Railroad Ton Miles Activity for 2008 and 2011

Total Revenue Ton-miles
of Freight (thousands)
Total non-Revenue
Ton-miles of Freight
(thousands)
Total Revenue and non-
Revenue Ton-miles of
Freight (thousands)
% change
from 2008
to 2011

2008 2011
2008 2011
2008
2011

BNSF
664,384,072 648,431,637
5,997,398 6,117,197
670,381,470
654,548,834
-2.362
CSXT
248,121,469 228,394,651
347,234 1,216,165
248,468,703
229,610,816
-7.590
GTC
53,452,403 51,253,084
624,848 518,201
54,077,251
51,771,285
-4.264
KCSR
29,624,261 30,485,863
6,077 1,338,343
29,630,338
31,824,206
7.404
NS
195,343,113 191,712,562
273,331 1,267,931
195,616,444
192,980,493
-1.348
Soo
23,681,180 34,581,354
241,414 333,090
23,922,594
34,914,444
45.948
UP
562,629,694 544,397,317
5,187,410 5,485,720
567,817,104
549,883,037
-3.158
Total
1,789,913,904
1,745,533,115
-2.479
ERG also tried to obtain rail freight trend data from the Association of American
Railroads (AAR), but their current posted data only extend back to 2009. It was noted that the
AAR data are used in the in the BTS's National Transportation Statistics and are similar to the
data in Table 1.
Because Class II and III rail operations are often affected differently by changes in the
economy than Class I railways, data were obtained from the American Short Lines and Regional
Railroad Association to assess their growth rate for the study period. Unfortunately, freight
traffic data in terms of ton-miles was not available, so information regarding employee hours for
2008 and 2011 were evaluated, quantifying a decline in activity of 8.37 percent. It is possible
that this decline overstates the actual change in Class II and III rail traffic, as employee
efficiency may also change during periods of economic uncertainty.
Lastly, it should be noted that growing the 2008 data using these measures does not
account for improved locomotive efficiency. Because the price of railroad fuel increased over the
study period and because fuel usage is such a large component of rail finances; when demand
declines, railways often use their newer, more efficient locomotives and retire the older engines
to reduce their system-wide fuel consumption. Under these conditions, less fuel would be needed
to move cargo, suggesting that actual 2011 emissions may be slightly less than those estimated
for this project.
3.0 Emissions Estimate Summary
The railroad component of the 2008 NEI was provided by ERTAC. ERTAC revised their
data since the 2008 NEI was posted. Prior to scaling the 2008 data to represent 2011, the 2008
NEI data set was amended to include the updates in the latest version of the ERTAC data. Note
there were no changes to the Class I line haul operations which represent the largest rail emission
source. The latest 2008 ERTAC data set for Class II/III line haul operations contained additional
state and railroad provided data as well as an updated fuel use factor. These revisions were re
applied to the county level and re-aggregated to match the latest ERTAC data set. The yard
3

-------
engine data also had slightly different emissions and record counts. The new yard emissions
were summed and reallocated to new yards which were assigned unique IDs and EIS/GIS point
locations.
The growth factors developed in Section 2 were applied to the updated 2008 railroad
emission estimates using the following equation to approximate 2011 emissions:
EE2oiiij = EEaoogij x (1+ GF; /100)
Where:
EE2onij	= 2011 railroad emission estimate for operation type i and pollutant j
(Tons/Yr)
EE2oo8ij	= 2008 railroad emission estimate for operation type i and pollutant j
(Tons/Yr)
GF;	= 2011/2008 growth factor for operation type i (Class 1 railroad and all
yard operations = -2.475percent and Class 2 and 3 operations =
-8.37 percent)
i	= Rail operation type (Class 1 line haul, Class 2 and 3 line haul, and yard
operations)
j	= criteria pollutant and regulated HAPs.
The 2011 emissions using the above approach are presented in Table 2 along with the
updated 2008 values for each locomotive category. HAP emissions for each locomotive category
are presented in the Appendix of this report.
Table 2. 2008 and 2011 Annual Emission Estimates by Locomotive Category (Tons)
Pollutant
Class I
Class II/III
Switch
Total
2008
2011
2008
2011
2008
2011
2008
2011
CO
110,969
108,218
4,631
4,244
9,231
9,002
124,830
121,463
nh3
347
339
14
13
28
27
389
379
NOx
754,433
735,731
47,035
43,100
74,431
72,586
875,899
851,417
PM10
25,477
24,846
1,158
1,061
2,105
2,053
28,740
27,960
pm25
23,439
22,858
1,065
976
2,042
1,991
26,546
25,826
SO.
7,836
7,642
327
300
624
608
8,787
8,550
voc
39,952
37,000
1,829
1,676
5,125
4,998
46,905
43,674
4

-------
4.0 References
U.S. Environmental Protection Agency, Documentation for Locomotive Component of the
National Emissions Inventory Methodology (NEI), May 3, 2011,
http://www.epa.gov/ttn/chief/net/2008inventory.html.
Department of Transportation / Bureau of Transportation Statistics (BTS), Multimodal
Transportation Indicators, February 2012.
http://www.bts.gov/publications/multimodal transportation indicators/february 2012/html/rail r
evenue ton miles.html
Surface Transportation Board, Complete Annual R1 forms (2008-2011).
http://www.stb.dot.gov/econdata.nsf7f039526076cc0f8e8525660b006870c97QpenView
The Department of Energy, Annual Energy Outlook (AEO), June 2012.
http ://www. eia.gov/forecasts/aeo/data, cfm
American Association of Railroads (AAR), Class I Railroad Statistics, May 10, 2012.
http://www.aar.Org/~/media/aar/Industry%20Info/AAR-Stats-2012-05-10.ashx
Timmons, Richard; American Short Lines and Regional Railroad Association. Short Lines
Today-Employee Hours (2006-2011), February 22, 2012.
http://www.aslrra.org/images/news file/AASHTO SCORT Feb 22 2012.pdf.
5

-------
Appendix A - 2011 HAP Emissions by Locomotive Category

-------
Table A-l. 2011 HAP Emissions by Locomotive Category (Tons per year)
Pollutant
Class I
Emissions
Class II/III
Switch
Total
1.3 -Butadiene
113.899
4.866
9.186
127.950
2,2,4-Trimethylpentane
82.974
3.758
11.208
97.940
Acenaphthene
0.738
0.032
0.060
0.830
Acenaphthylene
10.413
0.445
0.850
11.708
Acetaldehyde
659.298
28.164
53.178
740.640
Acrolein
109.648
4.684
8.845
123.176
Ammonia
338.587
13.276
26.958
378.820
Anthracene
2.461
0.105
0.201
2.767
Arsenic
0.009
0.000
0.001
0.010
B enz [a] Anthracene
0.395
0.017
0.032
0.444
Benzene
90.721
3.875
7.317
101.913
Benzo[a]Pyrene
0.070
0.003
0.006
0.079
Benzo [b]Fluoranthene
0.157
0.007
0.013
0.176
Benzo [g,h,i,]Perylene
0.078
0.003
0.006
0.088
Benzo |k|Fluoranthcnc
0.128
0.005
0.011
0.144
Beryllium
0.696
0.030
0.058
0.783
Cadmium
0.696
0.030
0.058
0.783
Carbon Dioxide
—
1,617,263.311
3,283,729.797
4,900,993.109
Carbon Monoxide
108,217.732
4,243.692
9,001.821
121,463.245
Chromium (VI)
0.049
0.002
0.004
0.056
Chromium III
0.096
0.004
0.008
0.108
Chrysene
0.292
0.012
0.024
0.329
Ethyl Benzene
74.000
3.351
9.996
87.348
Fluoranthene
1.840
0.079
0.151
2.070
Fluorene
3.417
0.146
0.279
3.842
Fonnaldehyde
1,519.044
64.891
122.519
1,706.454
Hexane
203.501
9.216
27.490
240.207
Indeno[ 1,2,3 -c,d]Pyrene
0.067
0.003
0.006
0.075
Lead
2.088
0.089
0.173
2.350
Manganese
0.051
0.002
0.004
0.057
Mercury
0.696
0.030
0.058
0.783
Naphthalene
63.268
2.702
5.193
71.163
Nickel
0.163
0.007
0.013
0.183
Nitrogen Oxides
735,730.789
43,099.979
72,586.140
851,416.909
Phenanthrene
13.805
0.590
1.127
15.521
PMio Primary (Filt + Cond)
24,845.831
1,060.928
2,052.942
27,959.702
PM2 5 Primary (Filt + Cond)
22,858.165
976.001
1,991.354
25,825.519
A-l

-------
Table A-l. 2011 HAP Emissions by Locomotive Category (Tons per year)
Pollutant
Class I
Emissions
Class II/III
Switch
Total
Propionaldehyde
225.701
10.222
30.489
266.411
Pyrene
2.591
0.111
0.213
2.914
Styrene
77.700
3.519
10.496
91.715
Sulfur Dioxide
7,641.578
299.635
608.406
8,549.619
Toluene
118.400
5.362
15.994
139.757
Volatile Organic Compounds
37,000.155
1,675.721
4,998.128
43,674.003
Xylenes (Mixed Isomers)
177.601
8.043
23.991
209.635
A-2

-------
United States	Office of Air Quality Planning and Standards	Publication No. EPA-454/B-20-024
Environmental Protection	Air Quality Assessment Division	September 2012
Agency	Research Triangle Park, NC

-------