U.S. Energy Information Administration - EIA - Independent Statistics and Analysis
Coal Transportation Rates to the Electric Power Sector
With Data through 2010 | Release Date: November 16, 2012 | Next Release Date: March 2014 | Correction
In this latest release of Coal Transportation Rates to the Electric Power Sector, the U.S. Energy Information Administration (EIA) significantly expands upon prior versions of this report with the incorporation of new EIA survey data. Previously, EIA relied solely on data from the U.S. Surface Transportation Board (STB), specifically their confidential Carload Waybill Sample. While valuable, due to the statistical nature of the Waybill data, much of the information had to be withheld for confidentiality reasons. With this current release, EIA augments STB data with EIA's Power Plant Operations Report (EIA-923) data. Unlike the Waybill sample, EIA-923 survey data are intended to be a full population data set. Utilizing the EIA-923 Survey's much larger data coverage enables EIA to publish data over numerous routes that were previously withheld, especially for state to state rates.
Another feature of this release is the incorporation for the first time of coal transport rates by barge and truck. As a relatively new survey field however, barge and truck data currently comprise a limited time series. Further analysis will be conducted as more data are collected; however, three years (2008-2010) are included in this initial release.
This report covers railroad transportation rates from 2001-2010 and barge and truck rates from 2008-2010. STB data covers 2001-2009 and the newly incorporated EIA data extends over the period of 2008-2010. EIA is pleased to note that, for future releases, using EIA survey data rather than solely relying on the STB Waybill will enable EIA to significantly improve the timeliness of this report, releasing data within several months of collection.
See next section — Trends, 2001-2010
Trends, 2001 - 2010
Transportation infrastructure overview
In 2010, railroads transported over 70 percent of coal delivered to electric power plants which are generally concentrated east of the Mississippi River and in Texas. The U.S. railroad market is dominated by four major rail companies that account for 99 percent of U.S. coal rail shipments by volume.
Deliveries from major coal basins to power plants by mode
The Powder River Basin of Wyoming and Montana, where coal is extracted in huge surface mines and sold at prices of around $8 to $15 per ton, is by far the largest supplier of coal for U.S. power plants. Much of this coal is transported in unit trains of 120 coal cars going from a single mine to a single power plant over distances that can exceed 1,000 miles (Interactive map and Figure 2).
East of the Mississippi—where river systems are well-developed—transport by barge often is lower cost than rail when sellers and buyers are located on or near river systems (Figure 3). Truck, used mainly for short transport distances, is the third largest (by tons hauled) transport mode for coal (Figure 4).
|Mode||East of Mississippi River||West of Mississippi River||Total U.S.|
|Barge & collier||30%||1%||11%|
|Slurry pipeline, tramway, and conveyor belt||2%||9%||7%|
|Railroad||million short tons||Share|
Nominal rates for transporting coal by rail from mines to electric generators increased 46 percent nationally from $11.83 to $17.25 per ton over 2001-2010, an average increase of 4.3 percent annually (Table 1 and Figure 5). Rates grew slowly in the early decade and then picked up steam, including a 10.5 percent spike in 2005, with relatively large increases until 2009 when the wake of the recession impacted shipping rates, causing a 2.3 percent decline in national rates. Prices quickly recovered however, growing 9.2 percent in 2010 to reach $17.25, six percent above the pre-recession annual average peak (or three percent higher in real terms).
Real rates (Chain-type Price Index, 2005) increased 19 percent nationally from 2001-2010. While rail transportation costs increased from $13.04 to $15.54 per ton (19 percent), total delivered coal prices rose from $27 to $40 per ton, almost 50 percent. Most of the increase occurred after 2004. Since coal commodity prices rose faster than transport costs, the share transportation costs comprised of total delivered cost fell from 48 percent to 39 percent. In dollar per ton terms, transport added $3 dollars while mine prices increased $11 over the 2001-2010 timeframe.
The U.S. average price masks substantial swings in real regional prices among the six major U.S. coal basins over the period 2001-2010, ranging from a 14 percent increase for Powder River Basin rates, to 149 percent for Southern Appalachia (Figure 6). Though this represents a sizable increase for coal originating from Southern Appalachia, the coal volumes are small.
In contrast to the national picture, average coal transport rates originating from Uinta (principal coal-producing regions of Colorado and Utah, see Interactive map), Central Appalachian (CAPP), and Southern Appalachian (SAPP) Basins rose rapidly from 2001 through 2007, moderated thereafter, and by 2010 fell slightly from their earlier peak values (Figure 6). Coal originating from the Illinois Basin (ILB) experienced rising rail rates until 2005, declining rates from 2005 to 2007, and rapidly rising rates after 2007. Rail transport rates for coal originating from the Northern Appalachian (NAPP) and Powder River Basins (PRB) had a moderate upward trend over the 2001-2010 period. Annual real growth rates per short ton (2005 dollars) for each basin are indicated in the table below.
|Coal Producing Basin||Annual Rate of
|Powder River Basin||1.5%|
In spite of having the largest increase from 2008-10 (47 percent), Illinois Basin coal remained the least expensive for transportation with real rates of $9.96 per ton. Central Appalachian Basin, the highest cost on a dollar per ton basis with real rates of $19.53 per ton, traded the top spot with Uinta Basin several times from 2001 through 2010. Coal delivered from ILB and Appalachian basins to various states experienced some of the largest increases in rail transport rates between 2001 and 2010. Conversely, coal deliveries from PRB to Missouri, Texas, Illinois, and Oklahoma had some of the smallest rate increases, even on a dollar per million Btu basis. While some routes for PRB experienced above-average rate increases (deliveries to Minnesota and Nebraska), in the case of deliveries to Iowa and Kansas, PRB rates actually fell from 2001-2010.
When the national and regional rate data are further disaggregated at the state to state level, changes reveal rather dramatic disparities. While the average real rate for shipping coal from a mine to a coal power plant by rail increased 19 percent nationwide between 2001 and 2010, average rates between specific states varied widely, ranging from a decline of 23 percent (Wyoming to Kansas) to an increase of 83 percent (Virginia to Tennessee) — a 106 percentage point swing. Figure 7 shows the percent change in average real rates for those state-to-state routes for which data was able to be published over the time period. The Figure shows that 10 state-to-state pairs experienced a 50 percent or higher increase in real rates. All of these pairs, except the route from Wyoming to Minnesota, had coal origins east of the Mississippi River. Five state-to-state pairs, three with western origins and two with eastern origins, had declining real rates over the 2001-2010 period.
Barge and truck transportation costs of coal to electric generating plants for 2008-2010 averaged between $4.46 and $5.77 per ton (real 2005 dollars). Rates fell during the recession in 2009, but rebounded slightly in 2010 (Table 2). Barge and truck rates are quite variable at the state-to-state level (Tables 15 through 18).
Barge transport makes up less than 1 percent of western coal, however, significant river transportation networks in the east enable it to comprise almost one-third of coal transportation. Truck transportation over short distances comprises around 10 percent of coal transportation.
gas coal switching: A look at competing coal basins
Significant coal to gas displacement in the electric power sector has resulted in natural gas and coal generation reaching parity in the United States in 2012. This trend is often referred to as coal to gas ‘switching’, though it typically involves increased utilization of gas plants in the generating fleet rather than physically switching out coal boilers to gas burning ones in a particular plant. Nonetheless, preceding significant coal to gas switching was a different switch affecting coal markets: PRB coal displacing Appalachian coal. EIA examined the changing basin footprints in the eastern U.S. and the role transportation costs played in the outcome.
Production and delivery points were mapped by basin with their total delivered costs to determine which was lowest. While coal from all three Appalachian Basins (Northern, Central, and Southern), Illinois Basin, and Powder River Basin compete within the eastern United States, for this report EIA analyzed PRB and CAPP, the major basins in play.
In order to approximate the economically optimal coal basin for a geographic area, the eastern U.S. was divided into small cells which were assigned an actual or estimated delivery cost based on EIA-923 Survey data. Each plant receiving CAPP or PRB coal in 2007 and 2010 were mapped and their data used to estimate costs for other cells by interpolating values based on inverse distance weights, a statistical method for conducting spatial analysis. Comparing the total delivery cost in each cell allows delineation of the eastern U.S. between PRB and CAPP as the most economical basin.
Figure 8 shows the estimated optimal economic area between CAPP and PRB for 2007 and 2010: CAPP is strong in light shaded areas, PRB in darker. Costs for CAPP include the sum of mine prices and transport costs. For PRB, the same variables apply but, in addition, EIA added a $1/million Btu charge on top to approximate additional costs of having a lower heat content coal (i.e. larger boilers, bigger storage and handling areas, lower efficiency). Note: this $1 adder was determined based on research by the U.S. Department of Energy and was only incorporated into the GIS analysis below; it is not in any other elements of this report. See Methodology for greater detail.
The resulting assessment shows a stark deterioration in the regional competitiveness of CAPP coal compared to PRB over the timeframe (2007-2010). CAPP delivered coal prices in 2010 ranged from $2.13 to $7.69 per million Btu while PRB costs (including assumed $1 adder) ranged from $1.80 to $5.16 per million Btu (both in Btu weighted annual average, nominal dollars). In 2007, prior to subsequent price increases, CAPP was price competitive over the wider Ohio River Valley, the northeast, and most of the southeast with the exception of Mississippi and Alabama. Only three years later, CAPP had been displaced over most of the Midwest outside of the Ohio River and virtually all of the Southeastern United States. Outside of its immediate production area, CAPP maintained its advantage over the northeastern United States.
The decline in competitive market area for CAPP can be attributed entirely to mine price. From 2007-2010, delivered coal prices per million Btu increased, on average, by 53 percent for CAPP compared to 22 percent for PRB. Upon further examination of the data, this divergence was in spite of changes in rail transportation costs. Rail transport costs for CAPP coal increased by less than half a percent from 2007-2010. PRB rates for coal transportation increased by an average of 18 percent over the same period, effectively slowing the erosion of CAPP's relative competitiveness to PRB. Rather, mine prices account for virtually all of the change in final CAPP coal delivered prices between 2007 and 2010. In only three years, these higher prices for CAPP steam coal (especially in light of recent low natural gas prices) have eroded much of its domestic market potential.
See next section — Methodology
EIA uses the confidential version of the STB Waybill data, which includes actual revenue for shipments that originate and terminate at specific locations. The STB Waybill data are a sample of all rail shipments. EIA's 2011 report describes the sampling procedure. EIA aggregates the confidential STB data to three different levels: national, coal-producing basin to state, and state to state. EIA applies STB withholding rules to the aggregated data to identify records that must be suppressed to protect business-sensitive data. Also, EIA adds additional location fields to the STB data, identifying the mine from which the coal originates, the power plant that receives the coal, and, in some cases, an intermediate delivery location where coal is terminated by the initial carrier but then transported to the ultimate customer under a new waybill. These latter multiple-leg shipments are excluded from the data used to calculate the reported STB transport rates. Otherwise, there is the possibility that the calculated rates to states such as Missouri and Illinois—which have key intermediate junction points—could be skewed.
Form EIA-923 data contains data fields to identify coal supplier down to mine name, mine state and mine county, coal user down to power plant name and state location, transport mode, total delivered cost, coal cost at the mine, and tons shipped. From data representing the full population, EIA eliminates all the records that either do not have any data for calculating transport rates or records with imputed transportation data. For each origin-destination pair, there can be multiple records within a year accounting for different mines, power plants and by-month data. EIA examines the data at the mine-power plant level to identify and eliminate outliers. EIA also applies withholding rules to the EIA-923 data to protect business-sensitive data.
The two data sets — STB and EIA — differ in terms of purpose, transport modes for which data are gathered, and the sample sizes of the respective surveys. The STB Waybill is a billing-year accounting mechanism which only surveys rail transport whereas the EIA surveys all modes transporting coal. As stated above, the STB data are a sample survey (http://www.eia.gov/coal/transportationrates/ ) whereas the EIA-923 data are intended to be a full population data set or census. EIA estimates transport cost as the difference between total delivered cost and commodity cost. In 2008, EIA revised Form EIA-923 to include commodity cost. When new elements appear in a data survey form, a typical outcome is that a few respondents provide no data because internal accounting may be difficult or missing, or respondents provide inconsistent or nonsensical data (which is not included in data reported by EIA).
For example some respondents in 2008 provided the same amounts when they reported free-on-board commodity cost and delivered cost. EIA implemented follow-up with respondents to clarify instructions and, consequently, the coverage of the EIA-923 survey has tended improve from the levels first experienced in 2008.
In the tables that accompany this report, EIA has calculated coverage rates for both STB and EIA data, according to the level of aggregation (national, basin to state, or state to state) for the data.
For example (see last two columns of Table 1), at a national level, in 2008, out of all coal shipments by rail to electric power plants reported on the EIA-923 survey 83 percent of the shipments (by tons) had acceptable free-on-board commodity and final delivered costs. By 2010, 90 percent of shipments (by tons) had acceptable cost data. The coverage rates for the EIA-923 data should improve as more problems are resolved in subsequent years. In comparison, the coverage rate of the STB data (aggregated to a national level) has remained around 33 to 36 percent for the period 2001-2009 (Table 1). Unit trains—at a 50% sampling rate—are the most densely sampled shipments in the STB data. As more coal is shipped on unit trains, the coverage rates for the STB data can increase but can never exceed 50 percent.
Under existing confidentiality policies, STB and EIA rates are withheld from publication when the number of respondents is so low that some respondents can back-calculate costs of other respondents. At lower levels of aggregation—for example state to state compared to basin to state—it is usually necessary to withhold more rates to assure data confidentiality. Therefore, because of the relatively high number of shipments included in each origin-destination pair of EIA-923 data, more EIA rates than STB rates normally can be published. This can be seen in a comparison of Tables 23 and 27, for example.
Even though there usually are fewer observations in the STB data set than in the EIA-923 data set for each origin-destination combination, there is a high degree of similarity between STB and EIA-923 average rates. At a national level, EIA-923 average rates are within about 2 percent of STB average rates in the two years—2008 and 2009—when they can be compared (see percent difference reported in Table 1). Differences are expected to be larger at the basin to state and state to state levels according to how the STB sampling procedure over- or under-samples various segments of the waybill population. Also, because EIA-923 data are not always a full census, biases can be introduced. Nevertheless, differences are modest for the most part (see percent differences reported in Tables 3 through 8 and Tables 19 through 22). It is also worth noting that, overall, there are about the same number of positive and negative differences between STB and EIA rates, indicating that there is no systematic bias in the differences, or, approximately, that the expected values of the differences between the STB and EIA average rates are zero. EIA plans to continue analysis of differences between these rates.
The EIA-923 survey collects data for all transport modes including railroad, barge, truck, Great Lake colliers, coal-slurry pipeline, tramways, and conveyor belt. Data confidentiality rules preclude the publication of EIA-923 rates for coal-slurry pipeline, tramway, and conveyor belt below the national level. In this report, EIA is providing average rate estimates for truck and barge at the national level (Table 2) and at the state to state level (Tables 15-18). Truck and barge transportation occurs mainly in states east of the Mississippi River. Consequently, truck and barge rates by basin to state do not add much additional information and were not calculated by EIA.
EIA calculated the $1 per ton adder for subbituminous coal compared to bituminous coal based on studies by the National Energy Technology Laboratory (NETL), which has provided data on cost differentials between power plants burning different coal types. Subbituminous power plants have an approximate factor of 1.4 versus a bituminous plant on equipment volumes to account for the requirements of handling larger volumes of coal to obtain the corresponding heat content. For example, larger capacity is needed for boilers, ducts, pollution control equipment, waste steam pipes and coal handling equipment and stockpiles. The factor 1.4 is the ratio of bituminous coal heat content (12,300 Btu/lb.) to subbituminous coal heat content (8,800 Btu/lb.). NETL optimizes the combination and size of equipment and arrives at a cost addition ranging from $0.60 to $1.20 per million Btu in 2010 dollars. EIA assessed a value of $1 per million Btu in this report in the market area section assessing PRB and CAPP market area competitiveness.
This $1 adder was only included in the Coal to
gas coal switching: A look at competing coal basins section and is not included in any of the data tables or other parts of this report. EIA acknowledges that this analytic technique does not take into account factors specific to individual plants or other contractual or special considerations which are important in determining actual coal purchasing economics and choices. Rather, this analysis is intended to provide on-average cost information and demonstrate the significant shifts in regional coal market dynamics over the period.
Email: Vlad Dorjets