LIBRARY OF THE
UNIVERSITY OF ILLINOIS
AT URBANA-CHAMPAIGN
510.84
X4£3c
no.\Z2rl30
AUG 51976
The person charging this material is re-
sponsible for its return to the library from
which it was withdrawn on or before the
Latest Date stamped below.
Theft, mutilation, and underlining of books
are reasons for disciplinary action and may
result in dismissal from the University.
UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN
Digitized by the Internet Archive
in 2012 with funding from
University of Illinois Urbana-Champaign
http://archive.org/details/energyintensityo127seba
INFERENCE RO
ENGINEERING LIBRARY
UNIVERSITY OF ILLINOIS
URBANA, ILLINOIS
Center for Advanced Computation
UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN
URBANA, ILLINOIS 61801
CAC Document No. 127
ENERGY INTENSITY OF BARGE AND RAIL
FREIGHT HAULING
By
Anthony V. Sebald
May, 197^
Thf> Library of r
MAY 5 19
ufitvbibiiy oi imn*
CAC Document No. 127
ENERGY INTENSITY OF BARGE AND RAIL
FREIGHT HAULING
By
Anthony V. Sebald
May, I97I+
:AC Document No. 127
ENERGY INTENSITY OF BARGE AND RAIL FREIGHT HAULING
By
Anthony V. Se"bald
Center for Advanced Computation
University of Illinois at Urban a- Champaign
Urbana, Illinois 6l801
May, 197^
This work was supported in part by a grant from the National
Science Foundation.
Introduction
In an attempt to quantify more of the total system costs associated
with transportation alternatives, studies are continuing in the area of
energy cost per ton mile for alternate freight and transportation modes.
In light of the present energy difficulties, energy efficiency is beginning
to have a significant economic impact on the various modes. Energy cost
per ton mile is also an important parameter in determining the total envi-
ronmental impact of competing transportation modes. This paper presents
results of an energy comparison per ton mile of competing rail freight vs.
inland barge freight, including the effects of circuity and the use of prob-
able competing rail lines instead of national average rail data.
The Problem
The basic underlying difficulty is that of constructing an equitable
frame of reference for comparing the two modes. Railroads haul some freight
along the barge routes and some over the continental divide. They haul in
unit trains dedicated to a single commodity (e.g. , coal) over a fixed long
distance trip (e.g., Louisville, Ky. to New Orleans) and they also haul in
mixed trains which stop and switch frequently. Finally, the railroads also
compete with the trucking industry and haul freight (in truck trailers) on
"piggy-back" systems as well as in the more conventional railcars. The
water transportation industry appears to be even more heterogeneous than
rail. Domestic water transportation includes:
1) Inland waterways (Mississippi river system and tributaries)
2) Gulf and Atlantic intracoastal waterways
3) Lakewise or Great Lakes transportation
h) Coastwise or deep sea transportation (New Orleans to New York,
Puerto Rico to New Orleans, etc.)
Even within the inland and intracoastal waterways system there is a large
number (l800 on the Mississippi-Gulf system) of barge firms ranging from
family owned tugs to large mult i- commodity freight haulers.* Barge freight
is moved on large capacity, long distance dedicated tows with the power
unit waiting for loading and unloading. It is also moved on mult i- commodity
tows in which the power unit continually moves while shore based tugs con-
nect and disconnect barges and bring supplies. Thus general, widely appli-
cable questions can be answered less precisely than specific ones.
Previous results in this area (2), (3), (M have been limited to ratios
of total domestic fuel use to total domestic ton miles. The present study
gives more precise results in that it takes into account two other impor-
tant variables. The energy intensity per ton mile calculation takes into
account the actual energy efficiencies of the most probable rail line com-
petitor of the barges on each particular haul and also includes the relative
circuities of the two modes. Circuity (defined as the modal difference in
distance travelled for an equivalent haul) is important since a ton moved
from New Orleans to Chicago will not travel the same number of miles in
both modes.
Methodology
Due to resource limitations, this study was limited to freight traffic
on the Gulf Intracoastal Waterway and the Mississippi River with all its
tributaries. Using the 1971 actual barge traffic data (5) (6), a list of
290 approximate origin-destination (OD) pairs was compiled. Data on tonnage
In 1971, 6.6% of the domestic ton mile traffic was regulated by the ICC (5)
carried for each OD pair in each of five bulk commodities (agricultural
output, lumber, petroleum, coal and chemicals) was also compiled. These
OD traffic pairs are approximate since the data in reference (5) is only
disaggregated to the regional level. Ports within the regional level were
chosen based on relative percents of corresponding traffic handled at the
major ports listed in reference (6).
Rail and barge routings were then generated for each OD pair. In the
barge case, the shortest routing was used. In the rail case, a balance of
minimum distance and minimum number of rail carriers was used in each
routing. Mileages in each case were obtained from (7) and (8). OD ton
miles (Tin) are given by the product of the tonnage and respective routing
length for each OD pair. Rail energy for each OD trip was calculated by
(a)
summing the product of energy intensity ' (Btu/Tm) and mileage for each
railroad's portion of the trip. Barge energy for each trip was assumed con-
stant as explained in the next section.
The computer program evaluated the overall intensities (Btu/Tm) for
each mode using the following weighted sums:
Rail (Btu/Tm) = -
I Tm0D
i i
£ Barge Tm .
Barge (Btu/Tm) = (f^)Avr *
^ AVG 7 Rail Tm+ .
V trip.
i i
) Barge Tnn .
r trip.
i l
The circuity weighting factor is also important m its own
T Rail Tnu .
r trip.
i l
right since changes in the estimates of energy efficiency per Ton Mile of
either the rail or barge mode can be easily included in the results of this
study by simply multiplying the barge efficiency by the above defined cir-
cuity factor. The circuity factor will remain stable until either major
traffic pattern changes or major rail or waterway construction occurs.
Results
There are two principal results of this study. First, the weighted
average energy intensity (El) of that portion of the rail industry which com-
petes with the barge lines (on the Gulf and Mississippi with tributaries)
was found to be 639 or 711 Btu/Tm depending upon whether one includes or not
the fuel used for yard switching. Both numbers are included in the compari-
son since the barge switching and tow makeup is sometimes done on contract
(e.g., tugs hitch and unhitch barges while the main tow continues moving).
It is therefore unclear how much of the switching and tow makeup fuel is in-
cluded in the barge direct EI figure quoted in Table 1. The rail El's are
weighted by 1971 waterborne commerce statistics, and include the 1971 ac-
tual energy intensities (Btu/Tm) experienced by each pertinent rail line as
explained in the previous section. The second basic result is the relative
circuity of the rail and barge modes for the 1971 waterway commerce traffic
on the Gulf and Mississippi with tributaries. On the average, barge ton
miles were 1.38 times as great as the equivalent competing rail ton miles.
Accepting for a moment that the comparison of an entire rail line's EI with
that of the average barge line is a valid one, the derived rail EI is very
(9)
accurate due both to the availability of excellent data " and the fairly
#
The actual 1971 waterborne traffic pattern was the basis for comparison
between the modes.
TABLE 1
1971 Rail vs. Barge Energy Comparison Parameters
ENERGY INTENSITIES (Btu/ton mi)
RAIL BARGE
(a)
Direct K ' 639-TH 785
Total 1330(b) l633(lD)
RELATIVE CIRCUITY (ton mi - barge ) ±^Q0 g
ton mi - rail
SAMPLE SIZE
Origin Dest. Pairs 290
Ton Miles Transported 10
Notes: (a) Includes motive fuel only, subject to the fol-
lowing clarifications:
_. ,. fuel consumed
Barge direct = — — where the fuel
ton miles
figure is fairly imprecise and includes hauling
fuel, some but probably not all switching fuel
and no maintenance fuel.
„ .. , . fuel consumed
Rail direct = — —
ton miles
Neither rail figure given includes maintenance
fuel, both include freight line haul fuel.
The smaller figure excludes switching fuel
while the larger one includes all switching fuel.
(b ) These figures are subject to fairly large un-
certainties .
large sample used (290 0D pairs and one hundred million barge Tm transported)
Although the circuity figure is subject to some uncertainty due to judgmen-
tal decisions in the choice of logical rail route, the large sample involved
would tend to reduce such uncertainty.
The above two results are combined with current estimates " of barge
freight EI in Table 1. The stated (in Table l) barge EI is the product of
Hirst's revised direct EI and the I.38 circuity explained above. Admittedly
the Hirst figure is subject to large uncertainties , but Table 1 can easily
be updated as new barge EI ratios become available. The new Table 1 barge
EI would simply equal I.38 times the new estimate of barge EI. More is said
about the barge EI estimate in Appendix A.
With a bit more effort, a fairly imprecise estimate of total (direct
and indirect) system energy for both rail and barge can be obtained. The
indirect energy includes such things as electricity consumed to make loco-
motives, track and freight cars as well as the paint for the offices of the
respective companies. Using a method explained in Appendix B, one obtains
the results listed in Table 1. It must be emphasized that these are only
estimates of the total energies involved. They are useful, however, in that
they indicate the total energy consumed in providing a ton mile of rail or
barge transportation as about twice that consumed by the locomotive or tug-
boat alone.
Finally, it is worthwhile to return momentarily to the subject of the
legitimacy of comparing an entire rail line's EI to the average barge EI
It includes all domestic water transportation (coastwise, lakewise and in-
ternal) and is based on the roughest, but best available, fuel consumption
estimates. Since barge lines are numerous (l800 on Gulf-Mississippi alone),
unregulated for the most part and are exempt from fuel tax, no accurate
fuel consumption data exists.
competing with it. Rail sources argue plausibly that, by and large, barge
movements are large commodity, long, point to point hauls and therefore should
be compared to unit train movement EI, not average rail line EI. The energy
intensity of high volume dry bulk cargo is significantly lower than the line
average EI, it is argued, since:
a) The gondola cars have one of the highest net to gross weight
ratios of all rail freight cars.
b) A homogeneous train of gondola cars has a very low air resis-
tance factor when compared to boxcars and especially to piggy
back loaded flat cars.
c) Significantly less switching fuel is needed.
d) Unit trains generally travel at lower speeds than other freight
trains.
(12)
Although some unit train EI results have been published indicating a
range of 226 to 359 Btu/ton mile not including circuity but including
the emply return trip, much more data needs to be collected before any
real comparisons can be made.
Level track.
Significant grade.
Conclusions
1) The 1971 average barge circuity (ratio of barge ton miles to equivalent
rail ton miles) on the Gulf- Mississippi system was I.38.
2) 1971 rail EI (energy intensity) in Btu/Tm for lines competing for barge
traffic was 639 (excluding switching fuel) and 711 (including all switch-
ing fuel). The corresponding national average energy intensity was ap-
proximately 700 Btu/Tm.
3) The resultant energy intensity comparison including the two above men-
tioned factors and the best available barge energy intensity per ton mi
indicates that rail is from 10 to 23% less energy intensive than barge,
but such a factor is inconclusive in view of the large uncertainty asso-
ciated with the barge fuel consumption data (see Appendix A).
k) The important overall question of modal energy efficiency can only be
accurately answered if a definitive program of collection of barge fuel
consumption data is initiated. In this author's opinion, the data must
be gathered in such a way as to permit regional or national weighting by
actual traffic carried and by the circuity factors involved. This means
following all steps of the procedure used in this paper with the excen-
tion of the inclusion of the actual EI of the most probable barge line
(or average of barge lines) for each portion of each trip. The data
should also accurately reflect that portion of the barge industry asso-
ciated with high volume, bulk, long distance (over 100 miles) hauling.
5) The matter of unit train EI should also be resolved for both dry bulk
and liquid bulk traffic. To be meaningful in a national average com-
parison, these data must also be gathered in such a way that the rail
line El's used in this report's calculations could be replaced by the
equivalent unit train EI.
REFERENCES
(1) Kearney: Management Consultants, "Domestic Waterborne Shipping
Market Analysis, Executive Summary," prepared for the Maritime Admin-
istration of the U.S. Department of Commerce, February 197^, p. 10.
(2) Eric Hirst, "Energy Intensiveness of Passenger and Freight Transport
Modes 1950-1970," Oak Ridge National Laboratory, Report No. ORNL-NSF-
EP-l+U, April 1973.
(3) Richard A. Rice, "System Energy as a Factor in Considering Future
Transportation," presented at the American Society of Mechanical
Engineers Annual Meeting, December 1970. By the same author: "Sys-
tem Energy and Future Transportation," MIT Technology Review,
January 1972.
(k) William Mooz , "The Effect of Fuel Price Increases on Energy Inten-
siveness of Freight Transport," Rand Corporation, Report R-80U-NSF,
December 1971-
(5) U.S. Army Corps of Engineers, "Waterborne Commerce of the United
States," part 5, 1971.
(6) U.S. Army Corps of Engineers, "Waterborne Commerce of the United
States," part 2, 1971.
(7) Rail mileages and routes were obtained from Handy Railroad Atlas of
the United States (New York: Rand McNally & Co., 1971).
(8) Barge mileages and routes were obtained from 1972 Interstate Port
Handbook (Chicago: Rockwell F. Clancy Co., 1972).
(9) The energy intensity of each rail line was calculated by dividing the
line's total diesel fuel consumption in the freight and yard switching
categories (Personal communication with Mr. H. Wolf, U.S. Interstate
Commerce Commission, April 197^ ) by the total revenue ton miles carried
by the line (U.S. Interstate Commerce Commission, "Transport Statistics
in the United States," part 1, 1971, pp. 1^2, l68, 19U, 220, 2^6, 272,
and 298).
(10) Telephone conversation with Mr. Harry N. Cook, National Waterways
Conference, Inc. , April 197^.
(11) A list of research results in the area is given in Table Al of Appendix
A. Dr. Hirst's revised results (Ref. (2), Appendix A) were chosen here
since:
(a) Dr. Hirst's and Dr. Mooz's research are independent national
average estimates of total actual fuel consumption and traffic.
(b ) Although both used the same fuel consumption estimates,
10
there appears to be some double counting of barge ton miles
in Dr. Mooz's results (see Ref. (2), p. 38).
(c) Of the two, Dr. Hirst's results are the most recent and apply
to the year of this study (1971 )•
(12) Telephone conversation with Mr. George Anderson, Western Railroad
Traffic Association, Chicago, Illinois, April 197*+.
(13) Letter from Mr. Harry N. Cook, Executive Vice-President, National
Waterways Conference, Inc. to Dr. Eric Hirst, FEO, March 7, 197*+.
11
APPENDIX A. Barge Freight Energy Intensity
Without any doubt, the most uncertain piece of data in the entire area
of Barge vs. Rail energy efficiencies is that of barge fuel consumption.
The research results given in Table Al indicate the uncertainty involved.
The commonly held opinion is that more data must be collected. Although
the information in existing data has been fully extracted and been found to
be insufficient, gathering new fuel data compatible with a ton mile weighting
similar to the approach used in this paper is not a trivial task due to the
large number (l800 on the Gulf- Mississippi system) of mostly unregulated
barge lines which must be queried.
TABLE Al
RESULTS OF RESEARCH ON WATERWAY ENERGY INTENSITY (Btu/Tm)
Author and Applicable Data Year Btu/tm
Hirst ( , 1970 680
(2)
Hirst v , 1971 570
Moozv , ca. 1968 500
(h)
Moozv , 1970 512
Brinegar , 1973 h62
National Petroleum Council , 1973 510
(7)
National Waterways Conference, Inc. ,
1968 Ul5 (Lowest Sample 217 Btu/Tm)
Notes:
(a) These figures do not include circuity and should be compared with the
rail result from Table 1 of the main report (after correction) which
is U63 to 515 Btu/Tm.
f 8)
(b) The Western Railroad Traffic Association has published results of
10,000 net ton coal train movements which range from l6h Btu/Tm (level
track) to 260 Btu/Tm (significant grade) after correction for barge cir-
cuity. These figures are most logically comparable with the National
Waterways Conference data given above.
12
References — Appendix A
(1) Oak Ridge National Laboratory, Report No. ORNL-NSF-EP-UU .
Domestic (including coastwide, barge, and lakewise) national average
Fuel source: Ref. (3)
Ton mi source: Transportation Association of America, "Transportation
Facts and Trends."
(2) Telephone conversation with Dr. Hirst.
Domestic (including coastwise, lakewise and barge) national average
Fuel source: James J. Mutch, Rand Corporation Document R1391-NSF,
December 1973.
Ton mi source: Transportation Association of America, "Transportation
Facts and Trends."
(3) "The Effect of Fuel Price Increases upon the Energy Intensiveness of
Freight Transport," Rand Corporation Report R-80U-NSF.
Domestic (including coastwise, barge and lakewise)
Fuel source: U.S. Bureau of Mines, "Minerals Yearbook" and "Mineral
Industry Surveys — Crude Petroleum, Petroleum Products
and Natural Gas Liquids."
Ton mi source: Interstate Commerce Commission and American Waterways
Operators, Inc.
(h) An update of (3) based on 1970 actuals. Letter to J. Feeney from Dr.
Wm. E. Mooz, August 8, 1973.
(5) Statement by Claude S. Brinegar, Secretary of Transportation, Before
the House Appropriations Subcommittee on Transportation, March 5, 197^+.
Includes "Freight Transportation by Water."
Fuel and ton mi source: Department of Interior estimates.
(6) "Interim Report Phase I, Transportation Task Group of the National
Petroleum Council's Committee on Energy Conservation.
Tug and barge operators only.
Fuel source: Estimate based on propulsion efficiency, annual HP hours
of propulsion and fuel efficiency.
Ton mi source: Not given in the interim report.
(7) "A Waterways Fuel Tax: Measurements of the Menace," Washington, D.C.,
May 1970. Based on a 32. h precent sample survey of inland waterway
carriers conducted by National Waterways Conference, Inc. in 1969.
Includes tug and barge operators only.
(8) Telephone conversation with Mr. George Anderson, Western Railroad
Traffic Association, Chicago, Illinois, April 197^.
13
APPENDIX B. Derivation of Total Energy Efficiencies
A reasonable way of estimating total energy impacts of transportation
is by the use of a Leontief Input Output inverse matrix, whose elements
(I - A). . by definition are the total (direct and indirect) output ($) of
industry i needed per dollar of output of industry j . These matrix elements
can easily be converted to energy units.
The basic method and 1963 results for the rail industry are given in
reference (l). The same document also extrapolates 1963 data to 1971 but
due to a difference in the groundrules of comparison, the 1971 results in
Table lb of (l) are not applicable here. Only the method (Section IIB) is
useful.
1971 total rail freight EI, X, can be estimated via the product:
X = A • B • C
where
A = locomotive and switching fuel used per ton mi in 1971.
B = ratio of total refined petroleum used per ton mile in 1963
to the locomotive and switching fuel used per ton mi in 1963.
C = ratio of direct energy used per ton mi by the railroads in
1963 to the refined petroleum per ton mi used by the railroads
in 1963.
D = ratio of the total (direct and indirect) energy per ton mi
used by the railroads in 196~3 "to the direct energy used by
the railroads in 196" 3.
Backup data for (l) gives the following values:
*
Including a correction for capital purchases such as rolling stock,
Ik
B = 1.08U
C = 1.129
Table la of (l) gives the value of D:
D = 1.70
Table 1 of the present report gives the value of A:
A = 711 Btu/ton mi
Therefore, 1971 total rail freight EI is given by
639 Btu/ton mi * 1.08U x 1.129 * 1.70 = 1330 Btu/ton mi
If the same values for B, C, D are assumed valid for the barge case, 1971
total barge freight EI is given by
785 Btu/ton mi x 1.08U * 1.129 x 1.70 = 1663 Btu/ton mi
Note that these total energies add nothing to the comparison between rail
and barge, they simply estimate how much total energy is spent per ton mi
in each case.
References — Appendix B
(l) Anthony Sebald and Robert Herendeen, "The Direct and Indirect Dollar,
Energy and Employment Impacts of Air, Rail and Automobile Passenger
Transportation," Energy Research Group, Center for Advanced Computa-
tion, University of Illinois, Urbana, Illinois, October 1973.
BIBLIOGRAPHIC DATA
SHEET
1. Report No.
UIUC-CAC-DN-7U-127
3. Recipient's Accession No.
4. Title and Subtitle
ENERGY INTENSITY OF BAEGE AND RAIL FREIGHT HAULING
5- Report Date
May, 1974
6.
7. Author(s)
Anthony V. Sebald
8. Performing Organization Rept.
N°" CAC 127
9. Performing Organization Name and Address
Center for Advanced Computation
University of Illinois at Urbana-Champaign
Urbana, Illinois 61801
10. Project/Task/Work Unit No.
11. Contract/Grant No.
NSF GI 35179X
12. Sponsoring Organization Name and Address
National Science Foundation
1800 G Street
Washington D.C. 20301
13. Type of Report & Period
Covered
Research
14.
15. Supplementary Notes
16. Abstracts
In an attempt to quantify more of the total system costs associated with
transportation alternatives, studies are continuing in the area of energy
cost per ton mile for alternative freight and transportation modes.
This paper presents results of an energy comparison per ton mile of
competing rail freight vs. inland barge freight, including the effects
of circuty and the use of probable competing rail lines instead of
national average rail data.
17. Key Words and Document Analysis. 17a. Descriptors
Energy
Intensity
Freight
Rail
Barge
Waterways
17b. Identifiers /Open-Ended Terms
17c. COSAT1 Field/Group
18. Availability Statement No restriction on distribution.
Available from National Technical
Information Service, Springfield Va.,
2211?].
19. Security Class (This
Report)
UNCLASSIFIED
20. Security Class (This
Page
UNCLASSIFIED
21. No. of Pages
Ik
22. Price
-ORM NTIS-35 [REV. 3-72)
USCOMM-DC 14952-P72