Creating a Statewide Commodity Flow Forecast from National FAF2 Data

As part of the planning effort leading up to the development of a statewide Freight Plan, the Oregon Department of Transportation (ODOT) developed a statewide commodity flow forecast. The methodology used to create this Oregon Commodity Flow Forecast (Oregon CFF), aimed to address the limitations of existing forecasts – inconsistent and separate databases for different modes, lack of transparency in data and assumptions, and data gaps – in a consistent methodology based on national and local data sources, and be able to meet the tight timelines. This paper will document the work done by the consulting team and the agency to create a statewide forecast that addressed these limitations. The project team decided to build on the Federal Highway Administration (FHWA) Freight Analysis Framework (FAF2) national commodity flow forecast. The FAF2 commodity flow forecast was chosen because FAF2 is national in scope, highly regarded in terms of capturing interstate and international flows, uses a relatively recent base year (2002), and provided a quick way to complete a forecast in time for the Oregon Freight Plan work. FAF2 provides freight flows in tons or dollar value between 130 FAF2 regions encompassing the US for the year 2002 plus forecasts from 2010 to 2035 in five year increments. The desired final product for the Oregon CFF was a county-county level flow forecast for truck, rail, marine, air, and pipeline modes. In order to transform the coarse FAF2 zone flows (2 zones cover Oregon) into counties within Oregon, the data was disaggregated. Since the FAF2 dataset contains the whole United States, flows with at least one trip ends within Oregon were disaggregated from FAF2 zones to Oregon counties. Each of the freight modes was disaggregated separately. In the case of truck flows, this was done based on county employment and IMPLAN inter-industry coefficients of what commodities are made and used by each industry. For rail flows, the FAF2 flows were compared to the Surface Transportation Board‘s Rail Carload Waybill data set which contains county level detail of origin and destinations. The overall numbers were found to be comparable, so the Waybill data for 2002 was used as the base, and the FAF2 ii growth rates were applied to forecast the future years. The other modes relied on local data to allocate FAF2 flows to specific Oregon facilities (rail stations, airports, marine ports, or pipeline terminals), including US Corps of Engineers Waterborne Commerce data and the Oregon Energy Report. Zones outside of Oregon were aggregated from FAF2 zones to ―Other Domestic‖ and ―Other International‖ categories. Special consideration was made for air mail and fish commodities using the knowledge of industry experts. Using the Rail Waybill data and other sources required a conversion in commodity categories, because FAF2 uses the Standard Classification of Transported Goods (SCTG)TG and other sources used the Standard Transportation Commodity Code classification. Once the data was disaggregated to represent county-county commodity flows, the FAF2 future year forecast numbers were adjusted down to account for the economic downturn that occurred after the forecast was prepared. One of the challenges of working with the FAF2 data is the inability to adjust or quantify the FAF2 underlying economic forecasts, particularly the optimistic economic conditions and low fuel price assumptions. These poses some limitations that must be taken into account. The Oregon CFF 2002 to 2035 forecast provides a basis for understanding the primary freight movements today and in the future under existing conditions. In several instances circumstances are likely to change, and the detail and transparency provided in Oregon CFF can provide a starting point for evaluating such changes.