Weekly train speeds data from the Surface Transportation Board's (STB) Rail Service Metrics. The STB began collecting service metrics from railroads in October 2014. As part of their submission to the STB, railroads provide data on the average speed of their trains (in miles per hour), broken out by commodity/type such as automotive, coal, crude oil, ethanol, grain, intermodal, manifest, etc. Railroads also report a “system average” train speed. According to the STB rulemaking, train speed should be measured for line-haul movements between terminals. The average speed for each train type should be calculated by dividing total train miles by total hours operated.

This dataset contains average weekly bids/offers ($/car) among grain shippers in the "secondary railcar auction market" for guaranteed rail freight in near and future months. Railroads auction freight in the "primary railcar auction market," and grain shippers can trade this freight among themselves in the secondary railcar auction market.

These railcar markets evolved to enable rail movements of grain to be more responsive to market pressures. Published tariff rates tend to reflect the most likely market conditions to prevail given historical precedence and future expectations; railroads adjust many of their tariff rates only once or twice per year in order to set longer term prices that account for their fixed assets and optimize their networks. Furthermore, railroads are required by law to give a 20-day notice prior to changing tariffs. Therefore, rail rates are more insulated than other modes from weekly market changes and unexpected events, including weather or transportation service disruptions. But in the short term, as new information enters the market, the optimal allocation of railcar supply with shipper demand may no longer be most efficiently allocated by the prices set by tariff rates alone. This was a characteristic of rail service prior to the late 1980’s when service was priced at the tariff rate and available on a first-come-first-served basis.

Forward-guaranteed railcar service contracts were an innovation first offered by railroads in the late 1980’s. These contracts offer guaranteed railcar deliveries within a specific time frame and serve as instruments against risk caused by unexpected events. They allow the supply of railcars to be continually reallocated among shippers through an auction bidding process as new information comes into the market, providing an alternative to first-come-first-served service purchased through tariff rates. There are two types of railcar auction markets—the primary market, in which service contracts are originally sold by railroads to shippers, and the secondary market, in which shippers resell service contracts among themselves. Sales in the primary market are administered by the railroads; sales in the secondary market are administered by third-party brokers.

For more information on this markets, see the feature article in the February 19, 2015 Grain Transportation Report (link provided below).

Weekly carloads from the Surface Transportation Board’s (STB) Rail Service Metrics. Each week, the seven Class I railroads submit originated and received carloads by 23 major commodity groups, such as grain, grain mill, primary forest products, chemicals, coal, containers, trailers, etc.

While the STB began collecting some service metrics from railroads (e.g., trains speeds, origin dwell, and terminal dwell) in October 2014, it did not start collecting carload data until March 2017 through a new rulemaking.

This data set contains rail tariff rates by railroad, commodity (corn, soybeans, and wheat), and train type, for select origin/destination pairs. These data are collected monthly from individual railroad websites.

Weekly origin dwell time data from the Surface Transportation Board’s (STB) Rail Service Metrics. The STB began collecting service metrics from railroads in October 2014. As part of their submission to STB, railroads provide "weekly average dwell time at origin for loaded unit train shipments sorted by grain, coal, automotive, crude oil, ethanol, and all other unit trains. For the purposes of this data element, dwell time refers to the time period from release of a unit train at origin or interchange location until actual movement by the receiving carrier. The data is to be reported by the receiving carrier."

The Surface Transportation Board's Carload Waybill Sample is perhaps the most comprehensive dataset available on railroad movements and trends. More technically, it is a stratified sample of carload waybills for all U.S. rail traffic submitted by those rail carriers terminating 4,500 or more revenue carloads annually. See 49 C.F.R. §§ 1244.1 to 1244.5.

Waybill data have broad applications and usage in national railroad policy and regulations, such as rate cases, costing systems, productivity studies, exemption decisions, and analyses supporting regulations. Waybill data are used by transportation practitioners, consultants, and law firms in preparing verified statements to be submitted in formal proceedings before the Board or other public agencies. Various federal agencies use the Waybill Sample as part of their informational and decision-making framework, and many states use it as a source of information for developing state transportation plans.

STB creates the Public Use Waybill file from the confidential Waybill Sample file. See the attached documents for more information. The "Reference Guide" document contains additional details on the variables and Standard Transportation Commodity Codes (STCC). In the "Creation of the Public Use Waybill Sample" document, STB provides more detail on the public use sample and how it is created. There is also a map of Bureau of Economic Analysis (BEA) Areas and a document describing the Waybill sampling instructions.

Weekly terminal dwell time data from the Surface Transportation Board’s (STB) Rail Service Metrics. The STB began collecting service metrics from railroads in October 2014. As part of their submission to the STB, railroads provide average terminal dwell time, measured in hours, excluding cars on runthrough trains (i.e., cars that arrive at, and depart from, a terminal on the same through train) for the carrier’s system and its 10 largest terminals in terms of railcars processed. (Terminal dwell is the average time a car resides at a specified terminal location expressed in hours.)

Fuel surcharge data is collected monthly from individual railroad websites. Fuel surcharges apply per mile per car. They are typically billed on top of tariff rates, but note that tariff- and contract-specific fuel surcharges may differ from the reported surcharge. For instance, in 2017 BNSF incorporated the fuel surcharge into its tariff rate and removed its explicit fuel surcharge. BNSF still reports fuel surcharge rates, which are included in the dataset.

A map where each dot is a rail yard, sized by its reported terminal dwell hours this week, and colored by an indicator of how unusually high or low dwell hours this week are for that yard.

Bar chart showing a query from the rail carloadings dataset to show the most recent week of data by railroad. It includes average weekly carloads based on the month to provide a sense of the seasonal expected level. It also includes a calculation of one standard deviation above and below that average to provide a sense of variation around the expected level.