Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
92 Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 4.1 Introduction This section provides an examination of how to develop subnational commodity flow data from supplemental sources of data on local economic and goods movement activity. These data on economic activity can come from local governmental sources. These data can also come from state and local trade associations. Often these supplemental data need to be combined with other pieces of information to convert them into a standard commodity flow template, which includes origins, destinations, commodity, mode, and volumes. The Guidebook identifies the following three steps that need to be addressed in administering a commodity flow data disaggregation technique: ⢠Step 1âDetermine industries and commodities of special interest ⢠Step 2âAssemble data on local economic and goods movement activity ⢠Step 3âEstimate missing data Some of these steps are interrelated, but the Guidebook discussion of each step is ordered as shown in the above bulleted list. The description of each step is structured to focus on the following four key elements: 1. Key ConsiderationsâA brief description of the main issues encountered and tradeoffs that will need to be made for the step. 2. Implementation ProcessâA detailed description of how to implement the step. 3. ExampleâAn example of how this step has been implemented in other studies. Note that this chapter includes brief examples of each of the steps and then provides two detailed examples of estimating local commodity flows for potatoes and diesel fuel at the end of the chapter. 4. Userâs Guide Worksheet Punch ListâSimple bulleted instructions that Guidebook users can check off to ensure that they have implemented each of the major steps involved in conduct- ing a commodity flow data disaggregation. C H A P T E R 4 . 0
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 93 4.2 Step-by-Step Process for Collecting Locally Available Data Step 1âDetermine Industries and Commodities of Special Interest Key Considerations There are many potential reasons to focus supplemental data development efforts on specific industries or commodities. The industry or commodity may be one of special interest to the region of concern such as a major employer or a strategic industry from an economic development perspective. Alternatively, a transportation agency may have a preexisting commodity flow database and there is a desire to verify or adjust a specific commodity to ensure it matches with local activity. It is important to define these special industries or commodities as narrowly as possible to focus the supplemental data collection effort. Implementation Process This process begins by identifying the industry or commodity of special interest. It is impor- tant to specify whether the interest is in an industry or a commodity. In situations where a trans- portation agency is interested in the impacts of a decision on its regional economy, it is typically an industry of interest. When the goal is to improve a preexisting commodity flow database, it is typically a specific commodity or group of commodities that are of special interest. Then, it is important to understand the relevant supply chain for the industry or commod- ity. For industry-specific supply chains, it will be important to understand what commodities are produced by the industry and what commodities are used to supply the industry. Similarly, the upstream industries that produce the supplies and the downstream industries that consume the products will be important to understand. For commodity-specific supply chains, it will be impor- tant to understand the set of industries that produce the commodity along with the set of industries that consume the product. Next, it is important to develop an understanding of the supply chain to determine the types of facilities that the goods typically move through. Are the commodities extracted from specific types of locations (such as coal or nonmetallic minerals)? Are the commodities stored in ware- houses and distribution centers? Are the commodities purchased by the final consumer at retail locations or used as industrial inputs for a future processing activity? The final portion in developing the supply chain is to understand the types of modes and vehicles that are used for each leg of the supply chain. It is common for several legs of the supply chain to be serviced by multiple types of modes and vehicles with specific companies tailoring their individual supply chain to their specific set of suppliers and customers. A graphic of a sup- ply chain for soft drink beverages from NCFRP Report 14: Guidebook for Understanding Urban Goods Movement is shown in Figure 4.1 (Rhodes et al. 2012). Many specific supply chains already have been mapped through previous research and can be found through Internet research on specific industries and trade associations. Economic input-output data are a more quantitative source of supply chain information. Input-output data describe the commodities consumed and services purchased by each industry along with the commodities produced by each industry. Input-output data also can be used to quantify the percentage of a commodity that is purchased by different industries. A comprehensive reference describing input-output data and models is Input-Output Analysis: Foundations and Extensions, most recently revised in 2009 (Miller and Blair 2009).
94 Guidebook for Developing Subnational Commodity Flow Data Truck Direct Rail Pipeline Source: Rhodes et al. 2012, Exhibit 3-1, p. 18. Figure 4.1. Supply chain for soft drink beverages. A publicly available source of input-output data is the U.S. Department of Commerce Bureau of Economic Analysis (BEA). These data are comprehensive in terms of their coverage of indus- tries and commodities. However, the most recently available data from BEA are from 1997 and 2002. Therefore, the BEA data are more helpful in identifying the relationships between indus- tries and commodities than they are in quantifying the current amounts produced and con- sumed. More recent sources of these data are proprietary, but can be obtained through economic modeling and analysis companies such as IMPLAN, Regional Economic Models, Inc. (REMI), PECAS, and TREDIS. The BTS Transportation Satellite Account (TSA) data provide information on the amount of transportation purchased by industries. These data can be used as a starting point to understand the modes that are used to ship goods between suppliers and consumers of commodities. Examples Figures 4.1 through 4.3 provide supply chain examples for soft drink beverages, construction materials, and gasoline and petroleum fuels, respectively. Userâs Guide Worksheet Punch List ⢠Determine industry or commodity of interest. ⢠Determine relevant upstream and downstream industries and commodities. ⢠Identify relevant freight facilities used. ⢠Determine modes and vehicles used to move goods between relevant freight facilities. ⢠Develop supply chain schematic for industry or commodity of interest.
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 95 For-Hire Truck Direct RailBarge Source: Rhodes et al. 2012, Exhibit 3-4, p. 22. Figure 4.2. Supply chain for construction materials. Truck Ship/Barge Direct Rail Pipeline Source: Rhodes et al. 2012, Exhibit 3-2, p. 19. Figure 4.3. Supply chain for gasoline and petroleum fuels.
96 Guidebook for Developing Subnational Commodity Flow Data Step 2âAssemble Data on Local Economic and Goods Movement Activity Key Considerations Sources of industry and commodity-specific information can range from regularly maintained databases to snapshots of freight activity from ad hoc data collection processes to anecdotal information from trusted industry experts. Each compo- nent of the supply chain schematic should be considered as a potential source of information for the industry or commodity of specific interest. Implementation Process The supply chain schematic developed in Step 1 can be used to identify potential supplemental sources of local economic and commodity flow data for the industry or commodity of inter- est. For each element in the supply chain, the following stakeholders should be considered for potential outreach to obtain supplemental data: ⢠Government agencies that regulate or monitor the industry or commodity. These agencies exist at the federal, state and local level, and they often maintain publicly available data on commodity movements on their web sites. Specific data requests should also be attempted with these agencies as they often make available more data to other public-sector entities. Typical types of data available include sales, production location, and permit data. Some data elements may be suppressed at more refined levels of geography. ⢠Trade organizations that represent private-sector producers in the relevant industry or commod- ity. Trade organizations also often track the amount of goods produced and consumed along with information on the importance of the industry for the broader economy. Trade organizations can include state mining associations, county farm cooperatives, or statewide industrial groups such as the Texas Petrochemical Association. Transportation agencies should be prepared to sign a nondisclosure agreement with trade organizations to obtain the most detailed data. Trade orga- nizations also can be used to confirm the accuracy of the supply chain schematic and identify specific names and locations of shippers, receivers, and carriers used for the commodity. They may also provide names of specific industry experts that have access to quantitative and anecdotal information that can be useful for understanding specific industries and commodities. ⢠Local research institutions may be operated by academic, government, or private-sector orga- nizations. They may maintain regularly updated data on the industry or commodity of interest. They may also have produced specialized reports that may be useful for understanding local commodity movements. Like trade organizations, they can validate the supply chain schematic, provide specific information about the location, size, and importance of key elements within the supply chain, and provide the names of individuals who may have useful information. ⢠There may be major companies that dominate an industry or commodity in a local region. If no specific contact has been previously identified for relevant major companies in your local region, then a logistics manager or government relations officer at the company should be con- tacted. By providing information on their own activities, they will provide insight on how the major local commodity flow movements occur. It will be important to ask them how big they are relative to the entire local market. ⢠There may also be major carriers that dominate an industry or commodity in a local region. These may include trucking firms, railroads, air cargo carriers, or shipping lines. For example, while there are several coal companies that have operations in West Virginia, only two rail- roads move the majority of the coal out of the state. The railroads could be contacted to obtain
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 97 information on the nature of these commodity flows. Similarly, FedEx and UPS dominate parcel delivery in most local areas. Therefore, they should be considered for potential outreach to understand how these types of goods are moved. ⢠Facility operators such as marine ports and airports also have information on commodity movements through their facilities. Typically, volume and commodity information can be made publicly available. Trade flows to other facility operators may be available at aggre- gated levels. Facility operators may also have collected specialized surveys providing them with information on ultimate origins and destinations at customer locations. These data may also be made available in some sort of aggregate form. ⢠Key individuals or industry experts also are useful for developing local commodity flows. They can verify the accuracy of the supply chain along with the currency of information obtained from other sources. In some instances, estimates provided by these experts may be the best data available for certain elements of the supply chain. Additionally, these individuals often have well-informed opinions about key trends that are shaping their industry in the near future. From each of the sources mentioned above, the key types of commodity flow data to capture provide information on locations of origins, locations of destinations, types and amounts of commodities produced at each location, and the modes and vehicles used to transport the goods. Additionally, information on shipment sizes, vehicle volumes, and intermediate handling loca- tions should be requested. Typically, information on volumes produced is the most readily available. This information may be available at a substate level. Information on modes and vehicles used to transport goods may be available and is very likely to be available at least at an anecdotal level. Table 4.1 provides a list of potential sources of supplemental local commodity flow data along with some specific examples of data sources from across the country. This table shows examples for select commodities, but similar organizations and associations can be found for the full range of commodities. The Examples section of this step provides snippets of some of these data sources. Often the data provided from these sources will provide a transportation agency with some, but not all, of the information needed to develop the specific commodity flow data of interest. In these cases, information from other sources may be used to fill in the missing data and transform the supplemental data into a full commodity flow database. In some instances, the supplemental data needed may be heavily connected to a specific freight mode. This can occur because the industry or commodity of interest has a component of its sup- ply chain that relies heavily on a single mode. Alternatively, this situation can arise if the trans- portation decision being considered has impacts that primarily affect a specific mode. Two of the key sources of mode-specific data that should be considered as potential supplemental sources on local goods movement activity are the Surface Transportation Boardâs (STBâs) rail Carload Waybill Sample and U.S. Army Corps of Engineers Waterborne Commerce Statistics Center. The rail Carload Waybill Sample provides detailed geographic data about loading and unload- ing points and interline locations along with information on tonnage and value. The data avail- able far exceed the detail and accuracy of CFS and FAF data on rail activity. Rail Carload Waybill data include both public use files and confidential files. The public use files provide information at the BEA level of geography. The confidential files are available at the facility level such that rail volumes by commodity and configuration type (e.g., carload, intermodal, and bulk) are avail- able for each rail line for the Class I railroads and for each major railyard in the United States. The confidential files are provided to a state agency upon request to the STB. Information must always be displayed in a manner that protects the confidentiality of the railroads and their ship- pers. Rail Carload Waybill data already are in a commodity flow format and therefore do not require the addition of missing data to be combined with other commodity flow data.
98 Guidebook for Developing Subnational Commodity Flow Data The Carload Waybill data are the source of the rail data that are used in the FHWA FAF data- base. Therefore, state DOTs and MPOs will not need to access the rail Carload Waybill Sample unless the more geographically refined data are useful. Obtaining the more detailed Waybill data is preferable to disaggregating FAF data due to the loss of accuracy that can occur in the dis- aggregation process. The Carload Waybill data do not include any forecasts, but these can be developed by associating rail activity with other economic activity for which forecasts already exist. State DOTs and other state agencies have the right to obtain the confidential files. Local agencies, such as councils of local governments and MPOs, will need to ask states to retrieve the Carload Waybill data. However, the STB may deny these pass-through requests and desire a direct request from local planning agencies, depending on the purpose and use of the data. The U.S. Army Corps of Engineers Waterborne Commerce Statistics Center data provide ton- nage, 20-foot-equivalent units (TEU), and value information by commodity for each port in the United States. However, there are no data available on trading partners or on inland origins or destinations that move through the port. All data are publicly available and can be obtained at the following link: http://www.ndc.iwr.usace.army.mil/wcsc/wcsc.htm. Commodity Type (Based on SCTG) Potential Local Sources of Commodity Flow Data Source of Examples of Supplemental Data Cereal Grains State agricultural department, U.S. Department of Agriculture, state fishery agency Data from Montana Department of Agriculture, North Carolina Farm Bureau, Idaho Pork Producers Association Other Agricultural Products Animal Feed Other Foodstuffs Natural Sands Local or state mining trade association, local or state base metal trade association Data from Georgia Mining Association, Colorado Stone, Sand, and Gravel Association, Aluminum Association of Florida Gravel Nonmetallic Minerals Nonmetallic Mineral Products Base Metals Coal Local or state mining trade association, U.S. Energy Information Association, State department of ecology West Virginia Coal Association, California Municipal Utilities Association, University of Tennessee Department of Ecology Crude Petroleum U.S. Energy Information Association, state department of ecology, state petroleum producers association U.S. EIA Petroleum Consumption Data for Washington, Louisiana Oil and Gas Association, Montana Petroleum Association Gasoline Fuel Oils Basic Chemicals Chemical manufacturers association, state and local trade associations American Chemistry Council, Ohio Rubber Group, Texas Fertilizer Association Fertilizers Chemical Products Plastics/Rubber Logs State trade associations, state forestry agency, wood products trade group Southeastern Wood Producers Association, North Carolina Association of Professional Loggers, Pellet Fuels Institute Wood Products Waste/Scrap State environmental agency, local recycling cooperative, state waste haulers association New Jersey Department of Environmental Protection, Pennsylvania Independent Waste Haulers Association, Florida Recyclers Association Table 4.1. Potential sources and specific examples of supplemental local data sources by commodity.
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 99 Example This section provides example snippets of data from the sources listed in Table 4.1. Table 4.2 shows sample data from the Montana Department of Agriculture. These data include specific crop produced, the production year, the county of production and the amount produced. For this snippet of data, it will first be necessary to convert the production unit into the desired unit for the commodity flow database. Typically, this unit will be tons. The conversion can be done using standard conversion charts for each commodity. Table 4.2 shows the county of production for the crop, which serves as the origin of the goods, but there is no information on destinations. To find information on destinations, it is useful to consult the supply chain schematic developed in Step 1 and/or to contact trade organizations that specialize in the crop being produced, major companies producing the crop, truck and rail companies that move the crop, local research insti- tutions that study crop development, and consultants to the crop industry. Figure 4.4 shows an example of a supply chain schematic that was provided through outreach to the Georgia Mining Association. The Georgia Mining Association is a trade organization representing private-sector mining interests that operate primarily in the center portion of the state and specialize in extracting kaolin. The schematic was developed by one of the associationâs member companies, but is used by the association to provide an overview of how kaolin is devel- oped. The schematic can be used to verify portions of a broader supply chain that includes the locations of facilities across the state along with other key pieces of information. Table 4.3 shows the production and consumption of energy across very specific categories for the state of Washington. These data can be used in conjunction with missing facility and origin- destination information to develop a commodity flow database across several fuel types. The example presented in the next section describes this process in greater detail. Commodity Year State County Production Prod. Unit Barley; all 2001 MT Flathead 502,000 Bushel Barley; all 2001 MT Lake 163,000 Bushel Barley; all 2001 MT Powell 82,000 Bushel Barley; all 2001 MT Ravalli 35,000 Bushel Barley; all 2001 MT Sanders 8,000 Bushel Barley; all 2001 MT Blaine 609,000 Bushel Barley; all 2001 MT Chouteau 515,000 Bushel Barley; all 2001 MT Glacier 2,137,000 Bushel Barley; all 2001 MT Hill 314,000 Bushel Beans, All Dry Edible 2001 MT Treasure 12,700 Hundredweight Beans, All Dry Edible 2001 MT Yellowstone 18,900 Hundredweight Beans, All Dry Edible 2001 MT Custer 20,000 Hundredweight Beans, All Dry Edible 2001 MT Prairie 45,200 Hundredweight Beans, All Dry Edible 2001 MT Rosebud 12,300 Hundredweight Canola 2001 MT Daniels 10,090,000 Pounds Canola 2001 MT Richland 2,827,000 Pounds Canola 2001 MT Roosevelt 3,347,000 Pounds Canola 2001 MT Sheridan 14,760,000 Pounds Source: Montana Department of Agriculture. Table 4.2. Sample data from the Montana Department of Agriculture.
100 Guidebook for Developing Subnational Commodity Flow Data Source: Thiele Kaolin Company. Figure 4.4. Sample supply chain data acquired through the Georgia Mining Association. Energy Data Units Period By Type Total Energy 2,037 trillion Btu 2010 Total Petroleum 138.7 million barrels 2010 Motor Gasoline 64.1 million barrels 2010 Distillate Fuel 25.3 million barrels 2010 Liquefied Petroleum Gases 4.2 million barrels 2010 Jet Fuel 19.3 million barrels 2010 Natural Gas 285,865 million cu ft 2010 Coal 72.7 million short tons 2010 By End-Use Sector Residential 478,794 billion Btu 2010 Commercial 380,074 billion Btu 2010 Industrial 564,920 billion Btu 2010 Transportation 612,728 billion Btu 2010 For Electricity Generation Petroleum 2,000 barrels Week of 7/1/2012 Natural Gas 3,108 million cu ft Week of 7/1/2012 Coal 1,000 short tons Week of 7/1/2012 Table 4.3. U.S. Energy Information Association consumption data for Washington.
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 101 Step 3âEstimate Missing Data Userâs Guide Worksheet Punch List ⢠Identify government agencies that regulate or track all industries and commodi- ties identified in the supply chain schematic. Make sure to research federal, state, and local agencies. ⢠Identify trade organizations for all industries and commodities identified in the supply chain schematic. ⢠Identify academic institutions and research organizations that have conducted research on the industries and commodities identified in the supply chain schematic. ⢠Assemble publicly available information from the three types of sources listed above. Contact each type of source to determine if they have or are aware of other commodity flow information for your local region. ⢠Contact specific individuals at major shippers, receivers, and carriers for your industry or commodity of interest. Confirm information collected from the types of organizations listed above and determine whether there is additional information on local commodity flows. Key Considerations The key elements in a commodity flow database are origin, destination, commodity, mode, and volume. Most supplemental data sources will be missing one or more of these data elements. Methods for estimating the missing data include making inferences from quantitative sources and anecdotal data and collecting new data. Creativity and resourcefulness are often needed to fill in these data gaps. Implementation Process There are several types of information that may be missing from supplemental data sets. In many instances, the volume of a specific commodity will be known at the desired geographic level, but the destinations of the commodity are not known. Alternatively, the origin and destina- tion information may be known, but not at the desired level of geographic specificity. A first step to filling in missing data is to reach out again to the freight stakeholders identified through the development of the supply chain schematic to determine whether there are data sources that have been overlooked and to determine, from people close to the data, a reasonable range for the values of the missing data. Another option is to consult alternative data sources, such as FHWA FAF data, to determine proxies for the data that are missing. For example, in a case where the agricultural flows from a county to external locations are desired, this information can be approximated by consulting FAF data on flows to external locations at the regional or state level. Similarly, local employment data and land use data can be used to approximate data on specific origins and destinations of commodity flows that are missing from supplemental data sets.
102 Guidebook for Developing Subnational Commodity Flow Data The most resource-intensive option to consider is filling in missing data by collecting new local commodity flow data. This option can include conducting an establishment survey on a specific industry, as described in Chapter 2. It also can include collecting roadside truck origin- destination surveys at the gates to an intermodal marine terminal or a private-sector truck ter- minal. Roadside truck surveys are described in Chapter 3. The survey instrument used in these kinds of data collection efforts can be much shorter than a typical survey instrument due to the very specific type of information that will be requested; this allows an increase in the number of attempted surveys and, hopefully, leads to an increase in the survey response rate. It may also be necessary to expand supplemental data to estimate missing data. For example, if supplemental commodity flow information were received regarding four of six paper mills in a local region, then data expansion could be used to approximate the commodity flows for the entire region by multiplying the supplemental data by 150 (6 divided by 4). Finally, another method that might be needed is proportioning supplemental data that were developed using a geographic scheme that does not exactly match the commodity flow database desired. The simplest way to correct for this is to adjust the supplemental data based purely on the sizes of the geographic regions that are being considered. Alternatively, proxy variables, such as employment by a specific industry, can be used to proportion supplemental data to the desired geographic boundaries. Example Sections 4.3 and 4.4 provide detailed examples of how to generate local commodity flow databases for diesel fuel and potatoes, respectively. Examples of filling in missing data along with other aspects of using supplemental commodity flow data are detailed in these sections. Userâs Guide Worksheet Punch List ⢠Identify missing data type(s) amongst the origin, destination, commodity, mode, and volume characteristics. ⢠Reach out to stakeholders identified in the supply chain to determine whether missing data or expert estimates can be identified. ⢠Identify broader commodity flow databases (such as FAF) that can be used to approximate missing data. ⢠Identify local activity variables (such as employment) that can be used to approximate missing data. ⢠If options for approximating missing data are not sufficient, collect new data to fill in missing data gaps. 4.3 Example of Developing a Diesel Fuel Local Commodity Flow Database Step 1âDetermine Industries and Commodities of Special Interest In this example, the commodity of interest is diesel fuel. The researchers were interested in the commodity flow within Washington state at a substate level of geography. The commodity flow information desired included the types of commodities, the amount being shipped by each mode, and the distances being shipped by each mode. Specifically, the interest was in the number
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 103 of truck trips on a typical day between a specific set of origins and destinations within the state of Washington. In this step, qualitative information was gathered regarding diesel distribution in Washington state, including the following reports: ⢠WSDOTâs Washington Transportation Plan Update Freight Movement (September 2008). This report contains an overview of the delivery and supply system for petroleum-based fuel in Washington state. The report summarizes the flow of refined products from the five active refineries in the state to end users at fueling stations, Seattle-Tacoma International Airport (Sea-Tac), the maritime industry, and private homes for home heating. ⢠Review of Pipeline Utility Corridor Capacity and Distribution for Petroleum Fuels, Natural Gas and Biofuels in Southwest Washington (ICF International 2007). In addition to addressing pric- ing and supply and demand patterns, the report describes the infrastructure of the diesel dis- tribution system within Washington state. The report identifies the Portland (OR)/Vancouver (WA) area as being the hub of the distribution network in the Northwest, as it receives and distributes product via pipeline, marine vessel, and tanker truck. ⢠Washington State Freight and Goods Transportation System (FGTS) 2007 Update (February 2008). This document provides an understanding of the roadway network within the state as it applies to the movement of freight. Fuel distributors and industry representatives were also interviewed to obtain preliminary information and begin to structure a general supply chain for the industry. It was determined that the major oil companies (BP, Shell, ConocoPhillips, U.S. Oil, and Chevronâdefined by ownership of their own crude oil reserves) refine crude oil into diesel and transport their product by way of pipeline, marine vessel, or tanker truck. (Tesoro produces diesel as well but was not considered a major oil company because it purchases crude oil from the major oil companies rather than sourcing it itself. Chevron does not operate a refinery in Washington state, but it imports diesel into the state via the Chevron Pipeline from Utah.) The pipelines are each owned and operated by one of the major oil companies (Olympic PipelineâBP; Yellowstone PipelineâConocoPhillips; and Chevron PipelineâChevron). Independent operators in the marine industry contract with the major oil companies to move diesel by barge or tanker ship, and the tanker trucks are operated by independent companies known as marketers. Diesel that is transported by a pipeline or marine vessel is offloaded at one of the 27 terminal locations in the state. The terminals owned and operated by a major oil company are known as proprietary terminals, and those owned and operated by an independent terminal operator are known as common terminals. For instance, BP and Shell have terminals in Seattle on Harbor Island, and ConocoPhillips has terminals in Renton and Tacoma. Common terminals are owned by NuStar in Tacoma and by Kinder Morgan on Harbor Island. From the terminals, a marketer purchases diesel and makes delivery to a cardlock location, other fueling stations, directly to fleets of vehicles, or as specified by its customers. In addition, the major oil companies can con- tract directly with the marketers to transport diesel directly from a refinery. The diesel distribu- tion supply chain has two additional features of note: 1. Marketers (these are the companies that move diesel from refineries or terminals to retail sales locations) purchase the fuel, rather than simply carry the fuel. 2. Given the volume of diesel consumed and its liquid form, it is distributed by marine vessel or pipeline whenever possible. Truck transportation is used only for last-mile distribution. Following these initial conversations, a model of the diesel supply chain actors was developed. This model is shown in Figure 4.5.
104 Guidebook for Developing Subnational Commodity Flow Data Step 2âAssemble Data on Local Economic and Goods Movement Activity The research team identified the following sources of information: ⢠Commercial Fueling Network (CFN) and Pacific Pride. The company web sites for these com- panies list the locations of cardlock facilities. These are the primary distribution locations for diesel trucks (www.cfnnet.com/ and www.pacificpride.com/). ⢠The Washington State Department of Ecology (ECY) regulates active underground stor- age tanks (USTs). The agency provided a database containing 2005 data for tank volume, fuel type, geographic coordinates, and physical addresses for 10,869 USTs, of which 2,378 were classified as holding diesel fuel. Although current information for each tank could be queried by a variety of selection options at the following ECY Web portal (https://fortress. wa.gov/ecy/tcpweb reporting/reports.aspx), the information was provided only in summary. The database acquired directly from ECY personnel was in a Microsoft Excel spreadsheet containing all locations in list format. The ECY personnel verified that the data provided in the 2005 database were substantively comparable to the current year database and that the current year database was not available in a format that could be as easily extracted as the 2005 database. ⢠The U.S. Environmental Protection Agency (EPA) regulates above-ground storage tanks. A database containing all current year above-ground storage tanks was acquired upon request. Although there was less certainty as to the specific formulation of the database in comparison to the USTs database, the above-ground storage tank database contained tank volumes, fuel types, and physical addresses for 67 identified tanks containing diesel. ⢠The Washington State Department of Revenue provided a list of active terminal locations in Washington state upon request (where fuel is distributed to trucks from refineries, barges, or pipelines). There were 27 terminal locations, including the 5 refineries. The 27 terminal racks in the state define the origins of interest. These are shown as green squares in Figure 4.6. The locations of USTs and above-ground storage tanks (ASTs) were com- pared to the complete list of cardlock locations provided by the CFN and the Pacific Pride diesel fueling network. Actual cardlock locations were obtained from each companyâs web site and matched with the locations of USTs and ASTs. In total, 376 of 433 (86.8 percent) cardlock loca- tions were matched with actual diesel tank locations. This data set provided a reliable group of known diesel distribution destination locations with tank volume data. These cardlock locations (racks) are the destinations of interest and are shown as pink circles in Figure 4.6. Data regarding fixed infrastructure and capacity, while somewhat time consuming to track down, were readily available in comparison to data on movements. Data on movements were inherently more difficult to obtain, as they vary quickly over time and space. Major Oil Companies Marketers Cardlocks/ Consumer Figure 4.5. General supply chain for diesel fuel.
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 105 The SFTA at Washington State University conducted a roadside origin-destination survey of freight trucks in 2003 and 2004. The database is categorized by UN placard number (1202 and 1203 for petroleum products), payload weight, and origin-destination (city, state) for the sur- veyed truck trips. Although this database contains empirical trip counts with pertinent associ- ated data, the trip counts for petroleum products are quite low, the type of petroleum product being moved is undefined, and expanding the trips to incorporate the entire state or a section of the state was determined to be insufficient due to uncertainties in how the survey was conducted. The FAF is an origin-destination database published by FHWA based on data collected dur- ing the 2002 Commodity Flow Survey. Report No. S6âPetroleum Products National Totals (n.d.) (http://ops.fhwa.dot.gov/freight/freight_analysis/faf/faf2_tech_document.htm) provides national aggregated movement totals for petroleum products by pipeline, water, highway, and rail. The information provided in the FAF was not found to be useful because the report does not have specific data at the substate level for the state of Washington. Additionally, there are discrepancies between the total fuel-related commodities estimated through FAF relative to other, more reliable data sources. Both the Washington State Utilities and Transportation Commission and the Washington State Office of Financial Management were contacted, but they did not possess any pertinent information. The Washington Oil Marketers Association provided a significant amount of information describing the Washington state diesel distribution network. It made note of the relationship between the marketers and the major oil companies and verified the general diesel supply chain network. It also provided names and contacts for diesel distribution marketers. Eight marketers were contacted with questions about the distribution of diesel by their com- panies. Three companies responded and provided valuable information that connected many of the missing pieces of the Washington state diesel distribution network not previously uncovered by the research. Terminals Racks (Cardlock Facilities) Figure 4.6. Location of diesel terminals and racks (cardlock facilities) in Washington.
106 Guidebook for Developing Subnational Commodity Flow Data The FMCSA was contacted to obtain information about the reporting requirements of the diesel distributors and marketers. According to federal regulations, a carrier is required to have a HAZMAT placard issued by the FMCSA and a Certificate of Registration issued by the PHMSA to conduct diesel delivery operations (these last for several years). The carrier is only required to report an individual movement in the case of a hazardous material spill. The information collected from these disparate sources resulted in a large amount of informa- tion on origins, destinations, commodities, and modes. However, volume information was still missing. The third step describes how this remaining information was estimated. Step 3âEstimate Missing Data No commodity flow data were retrieved from the sources listed above to estimate the diesel flow volume between terminals and cardlock facilities (or racks). The research team used two sources to fill in these missing data. First, the team identified the closest terminal (shortest travel time on the network) to each cardlock facility. It was assumed that the closest terminal was the one that was used as the origin of diesel fuel destined for each cardlock facility. This assumption is reasonable particularly if it can be assumed that the price for diesel was relatively equivalent at each terminal. With origins and destinations matched, the team defined the preferred service area for each terminal. These terminal service areas are shown in Figure 4.7. Information was also missing on the number and routing of truck trips between terminals and cardlock facilities. The number of truck trips could not be estimated because the volume infor- mation was missing. The routing information, while not included in commodity flow databases, was also not available from any existing sources. To fill in this missing information, the research team used nearby vehicle count data to estimate diesel flows for the state. For this commodity flow application, it was determined that the number of origin-destination pairs using each link in the road network would be a sufficient proxy for volumes. However, average annual daily traf- fic (AADT) could be used to estimate the amount of diesel consumed at each cardlock. AADT at the nearest count location possible on roadways proximal to the cardlock facility was used to Figure 4.7. Terminal service areas in Washington.
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 107 distribute known volume of diesel produced by terminals. Each cardlock received a portion of the total diesel dispensed equal to its AADT over the sum or all AADT. This assumes that card- locks on roadways with more traffic distribute more fuel. AADT is available for all state highways in Washington through WSDOT. The final result of this analysis is the estimation of truck flows between terminals and card- lock facilities. Figure 4.8 shows the number of origin-destination pairs that use each link in the network, where origins are terminals and destinations are cardlock facilities. Some links service almost 40 origin-destination pairs. 4.4 Detailed Example of Developing a Potato Commodity Flow Database Step 1âDetermine Industries and Commodities of Special Interest In this example, the commodity of interest is potatoes and potato products (frozen potatoes, dehydrated potatoes, and potato chips). An estimate was desired for the commodity flow within Washington state and between specific origins and destinations within the state that reflect the locations of production, processing, export and consumption of potatoes and potato products. Commodity flow information includes the types of commodities, the amount being shipped by each mode, and the distances being shipped by each mode. The goal of this commodity flow database development was the number of truck trips, for each potato product, on a typical day, between a specific set of substate origins and destinations within the state of Washington. To develop a sense of the supply chain for potatoes, a meeting was arranged between the research team and the Washington State Potato Commission. The Washington State Potato Com- mission defines itself on its web site as a trade organization that has as one of its primary goals to Figure 4.8. Diesel network flow map.
108 Guidebook for Developing Subnational Commodity Flow Data âenhance trade opportunities, to advance environmentally sound production and cultural prac- tices through research, and to represent the growersâ interests in areas and issues relating to public and industry education, trade barriers, irrigation, transportation and crop protectionâ (http:// www.potatoes.com/our-commission/mission/). As such, the commission staff have good knowl- edge of who produces and consumes the commodity. In addition, the staff compile data from a number of data sources to understand industry trends and conditions. Given this extensive back- ground, the Washington State Potato Commission was considered to be a prime source of data. Following the initial conversation with the Washington State Potato Commission, a schematic of the potato supply chain was developed. This is shown in Figure 4.9. Figure 4.9. Supply chain schematic for potatoes.
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 109 Step 2âAssemble Data on Local Economic and Goods Movement Activity Several available data sources were identified by Washington State Potato Commission staff and members of the research team (prior knowledge). These include the following: ⢠U.S. Department of Agriculture (USDA) field production data. These data are publicly available via the USDAâs Economic Research Service. U.S. and state data on potato acreage, production, value, and use are provided. The USDAâs Economic Research Service also includes U.S. data on prices, price spreads, and consumer and producer price indexes; stocks of fresh and processed potatoes; and trade in potatoes and potato products. At the national and state level, no data suppression issues are present. The research team believes this data to be of high quality. ⢠U.S. Census Bureau Foreign Trade Statistics, Total U.S. Exports (Origin of Movement) via Washington state. These data are available from the U.S. Census Bureau at the state level for the top 25 commodities and are also available by country of export. These data are readily accessible via the web and thought to be of high quality, with no suppression issues. ⢠SFTA Statewide Origin and Destination Truck Surveys completed by Washington State Uni- versity in 2003 and 2007. SFTA data can be requested through the SFTA web site (http://www. sfta.wsu.edu/data_inquiries/data.htm). The research team uses statistical methods to estimate total truck flows from the sample. As the actual flows are not known, the statistical quality of the data cannot be judged. ⢠Washington State Department of Agriculture surveys conducted between 1999 and 2007. Potato acreage and production information for the state can be obtained from the Washington State Department of Agriculture. The township-range-section (TRS) level acreage data used in this study are a compilation of the results from a continuum of surveys conducted by the Department of Agriculture between 1999 and 2007. The potato production figures for each county were allocated to the TRS level by using yield and acreage information calculated for each county (this work was completed by the Washington State Department of Agriculture). These data are available by request and are thought to be of high quality. There were no sup- pression issues with this set of data. ⢠2007 Washington State Potato Commission Survey. The Washington State Potato Commis- sion conducted a survey of its members in 2007. This information is available upon request. It included destination and route. This information is thought to be of high quality. Requests can be made to the Washington State Potato Commission, who reserves the right to share or not share the information and to determine in what form information is provided. ⢠Washington State Potato Commission member data. The commission asks its members to submit monthly information on shipments of potatoes by variety, product, and destination. This information is thought to be of high quality. Requests for this information can be made to the Washington State Potato Commission who reserves the right to share or not share the information and to determine in what form information is provided. ⢠Industry experts and representatives. Although not explicitly based on data provided to the research team, more generalized industry knowledge was used to estimate some parameters. ⢠United States Census Bureau Population and Housing Unit Estimates. The Census provides city and town population estimates. This information is available on the U.S. Census Bureau web site and is thought to be of high quality. For the size of city considered in this example, there are no suppression issues. As mentioned, the research teamâs interest was in estimating daily truck trips on the stateâs road network. The Washington State Department of Agriculture had estimated Potato Produc- tion by Township and Range for 2006 as shown in Figure 4.10. These data indicate that potatoes are produced in three regions of the state: the Skagit Valley, the Lower Basin, and the Upper Basin. These three regions are considered origins of fresh potatoes. Centroids of the region are identified as the origins of truck trips. Potato production volumes for each region are estimated
110 Guidebook for Developing Subnational Commodity Flow Data by summing the production in the region, less the loss rate (6 percent). The loss rate was pro- vided by the Washington State Potato Commission. The Washington State Potato Commission also estimates the percentage of each potato prod- uct produced in each region. The Skagit Valley is estimated to produce 100 percent fresh pota- toes, the Upper and Lower Basins are estimated to produce 14 percent fresh potatoes, 73 percent frozen potatoes, 11 percent dehydrated potatoes, and 2 percent potato chips. Processing The Washington State Potato Commission provided information on the location of all potato processing facilities in Washington (as shown in Figure 4.11) and the ratio of truckloads of fresh potatoes to truckloads of processed potatoes. The ratio for fresh potatoes to fresh potatoes is of course 1:1; the ratio of fresh potatoes to frozen potatoes is 2:1; the ratio of fresh potatoes to potato chips is 4:1; and the ratio of fresh potatoes to dehydrated potatoes is 6:1. Thus, for every truckload of dehydrated potatoes leaving a processing facility, six truckloads of fresh potatoes are required. Note, however, that a truckload of fresh potatoes is assumed to be equivalent to 22.22 tons, and a truckload of frozen potatoes is assumed to be equivalent to 20 tons. No information was available regarding the location of dehydration processing for potatoes grown in the Skagit Valley. Dehydration could take place at processing facilities in either the Upper or Lower Basins. On the basis of interviews with industry experts of the Washington State Figure 4.10. Washington potato production, 2006 (hundredweight by township and range).
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 111 Potato Commission, it was estimated that 25 percent of these potatoes would be dehydrated in the Upper Basin, and 75 percent would be dehydrated in the Lower Basin. Potatoes grown in the Upper Basin were assumed to be processed in the Upper Basin, and potatoes grown in the Lower Basin were assumed to be processed in the Lower Basin. Mode It was assumed that 25 percent of the frozen potatoes processed in the state leave the state via rail. This assumption was supported by conversations with an industry expert at one of Washingtonâs largest frozen potato producers, ConAgra. It also was assumed that 11 percent of fresh, dehydrated, and chipped potatoes are shipped out of the state on rail. Typically, these potatoes are destined for regions east of the Mississippi River. Potato Consumption and Export Potatoes grown and processed in Washington are either consumed in Washington or exported. The percentage of potatoes to major destination by region (estimated by the 2007 Washington Potato Commission Survey) is shown in Table 4.4. Potatoes destined to international locations are exported via the Port of Seattle. According to the 2007 Washington Potato Commission Survey, these shipments use the routes shown in Table 4.5. Figure 4.11. Location of Washington potato processors.
112 Guidebook for Developing Subnational Commodity Flow Data Potatoes destined for Eastern Washington are assumed to be distributed to Moses Lake, Spo- kane, Kennewick, Warden, Yakima, and Grandview, proportional with population. Destinations in Western Washington are Seattle, Tacoma, Stanwood, and Auburn according to population. Destinations in Oregon and California are served via I-5, I-205, I-82, and Highway 97 equally. Destinations in Idaho are served via I-90, Highway 2, Highway 12, I-82, and I-84 equally. Other destinations in the United States are served via I-90 and I-82 equally. Destinations in Canada are served via I-5 or Highway 9 equally. Destinations in Mexico are served via I-5, I-205, I-82, or Highway 97 equally. Truck Trips To convert short tons to truckloads of potatoes, it was assumed that fresh and dehydrated potatoes weigh out at 22,000 pounds. For frozen potatoes, the research team assumed that a truckload can carry 20,000 pounds because of the refrigeration unit necessary. For potato chips, a truckload can carry only 5,000 pounds as the trucks meet volume constraints before weight constraints. Table 4.6 shows the number of truck trips per day between each origin and each destination for each product type estimated using this procedure. Major Destinations Lower Basin Skagit Valley Upper Basin Eastern Washington 12.48% 2.03% 6.22% Western Washington 14.29% 6.81% 6.40% Oregon 2.31% 4.35% 1.25% California 14.58% 40.72% 11.85% Idaho 0.00% 0.00% 34.33% States West of Mississippi 22.01% 13.30% 12.76% States East of Mississippi 24.26% 23.58% 11.99% Canada 8.85% 7.04% 2.91% Mexico 0.14% 1.96% 0.25% Other International 1.09% 0.20% 12.03% Table 4.4. Percentage of potatoes to major destinations by region. Major Destinations Lower Basin Skagit Valley Upper Basin Eastern Washington I-90, I-82, Highway 12, 14 I-5, I-90, Highway 2, 405 I-90, I-82, Highway 17 Western Washington I-90, I-82, I-5, 240, 395 I-5, 405, 167 I-90, Highway 17 Oregon I-90, I-82, I-84, Highway 97, 395, 597 I-5 I-90, I-82, I-84 California I-90, I-82, I-5, Highway 97, 395 I-5 I-90, I-82, I-5, Highway 17, 395 Idaho I-90, I-82, I-84, SR 17, 395 States West of Mississippi I-90, I-82, 395 I-90, I-80, I-5, I-84, 405 I-90, I-82, I-5, I-84, SR 17, 395 States East of Mississippi I-90, I-82, 395 I-90, I-80, I-5, 405 I-90, I-82, I-5, I-84, SR 17, 395 Canada I-90, I-82, I-5 I-5, I-90 I-5, I-90 Mexico I-82, I-5, Highway 97 I-5 Table 4.5. Major destinations by route of exported Washington-produced potatoes.
Total Production Fresh Frozen Dehydrated Chips Destinations Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin East Washington 0.17 0.06 0.12 2.90 1.71 3.96 0.00 4.17 9.67 0.00 0.24 0.53 0.00 0.27 0.63 Moses Lake 0.01 0.00 0.01 0.13 0.07 0.17 0.00 0.18 0.42 0.00 0.01 0.02 0.00 0.01 0.03 Spokane 0.08 0.03 0.06 1.40 0.83 1.92 0.00 2.02 4.68 0.00 0.12 0.26 0.00 0.13 0.30 Kennewick 0.03 0.01 0.02 0.50 0.30 0.69 0.00 0.72 1.68 0.00 0.04 0.09 0.00 0.05 0.11 Warden 0.01 0.00 0.01 0.13 0.07 0.17 0.00 0.18 0.42 0.00 0.01 0.02 0.00 0.01 0.03 Yakima 0.02 0.01 0.02 0.37 0.22 0.50 0.00 0.53 1.23 0.00 0.03 0.07 0.00 0.03 0.08 Grandview 0.02 0.01 0.02 0.37 0.22 0.50 0.00 0.53 1.23 0.00 0.03 0.07 0.00 0.03 0.08 West Washington 0.06 0.07 0.14 0.99 1.99 4.53 0.00 6.49 14.76 0.00 0.28 0.61 0.00 0.32 0.72 Seattle 0.02 0.02 0.04 0.28 0.56 1.28 0.00 1.83 4.15 0.00 0.08 0.17 0.00 0.09 0.20 Tacoma 0.01 0.02 0.03 0.23 0.47 1.06 0.00 1.52 3.46 0.00 0.07 0.14 0.00 0.07 0.17 Stanwood 0.01 0.01 0.03 0.20 0.40 0.92 0.00 1.32 2.99 0.00 0.06 0.12 0.00 0.06 0.15 Auburn 0.02 0.02 0.04 0.28 0.56 1.28 0.00 1.83 4.15 0.00 0.08 0.17 0.00 0.09 0.20 Oregon 0.03 0.01 0.02 0.57 0.28 0.73 0.00 0.70 1.79 0.00 0.04 0.10 0.00 0.05 0.12 via I-5 0.01 0.00 0.01 0.14 0.07 0.18 0.00 0.17 0.45 0.00 0.01 0.02 0.00 0.01 0.03 via I-205 0.01 0.00 0.01 0.14 0.07 0.18 0.00 0.17 0.45 0.00 0.01 0.02 0.00 0.01 0.03 via I-82 0.01 0.00 0.01 0.14 0.07 0.18 0.00 0.17 0.45 0.00 0.01 0.02 0.00 0.01 0.03 via Hwy 97 0.01 0.00 0.01 0.14 0.07 0.18 0.00 0.17 0.45 0.00 0.01 0.02 0.00 0.01 0.03 California 0.35 0.12 0.15 5.81 3.42 4.63 0.00 8.34 11.29 0.00 0.48 0.62 0.00 0.54 0.73 via I-5 0.09 0.03 0.04 1.45 0.85 1.16 0.00 2.09 2.82 0.00 0.12 0.16 0.00 0.14 0.18 via I-205 0.09 0.03 0.04 1.45 0.85 1.16 0.00 2.09 2.82 0.00 0.12 0.16 0.00 0.14 0.18 via I-82 0.09 0.03 0.04 1.45 0.85 1.16 0.00 2.09 2.82 0.00 0.12 0.16 0.00 0.14 0.18 via Hwy 97 0.09 0.03 0.04 1.45 0.85 1.16 0.00 2.09 2.82 0.00 0.12 0.16 0.00 0.14 0.18 Table 4.6. Potato truck trips per day. (continued on next page)
Total Production Fresh Frozen Dehydrated Chips Destinations Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin Skagit Valley Upper Basin Lower Basin Idaho 0.00 0.36 0.00 0.00 10.26 0.00 0.00 25.03 0.00 0.00 1.45 0.00 0.00 1.63 0.00 via I-90 0.00 0.07 0.00 0.00 2.05 0.00 0.00 5.01 0.00 0.00 0.29 0.00 0.00 0.33 0.00 via Hwy 2 0.00 0.07 0.00 0.00 2.05 0.00 0.00 5.01 0.00 0.00 0.29 0.00 0.00 0.33 0.00 via Hwy 12 0.00 0.07 0.00 0.00 2.05 0.00 0.00 5.01 0.00 0.00 0.29 0.00 0.00 0.33 0.00 via I-82 0.00 0.07 0.00 0.00 2.05 0.00 0.00 5.01 0.00 0.00 0.29 0.00 0.00 0.33 0.00 via I-84 0.00 0.07 0.00 0.00 2.05 0.00 0.00 5.01 0.00 0.00 0.29 0.00 0.00 0.33 0.00 West of Mississippi 0.11 0.08 0.22 1.84 2.28 6.98 0.00 5.56 17.05 0.00 0.32 0.94 0.00 0.36 1.11 via I-90 0.05 0.04 0.11 0.92 1.14 3.49 0.00 2.78 8.52 0.00 0.16 0.47 0.00 0.18 0.55 via I-82 0.05 0.04 0.11 0.92 1.14 3.49 0.00 2.78 8.52 0.00 0.16 0.47 0.00 0.18 0.55 East of Mississippi 0.20 0.13 0.24 3.40 3.70 7.70 0.00 9.04 18.79 0.00 0.52 1.03 0.00 0.59 1.22 via I-90 0.10 0.07 0.12 1.70 1.85 3.85 0.00 4.52 9.39 0.00 0.26 0.52 0.00 0.29 0.61 via I-82 0.10 0.07 0.12 1.70 1.85 3.85 0.00 4.52 9.39 0.00 0.26 0.52 0.00 0.29 0.61 Canada 0.06 0.03 0.09 0.99 0.85 2.81 0.00 2.09 6.85 0.00 0.12 0.38 0.00 0.14 0.45 via I-5 0.03 0.02 0.04 0.50 0.43 1.40 0.00 1.04 3.43 0.00 0.06 0.19 0.00 0.07 0.22 via Sumas (Hwy 9) 0.03 0.02 0.04 0.50 0.43 1.40 0.00 1.04 3.43 0.00 0.06 0.19 0.00 0.07 0.22 Mexico 0.02 0.00 0.00 0.28 0.00 0.04 0.00 0.00 0.11 0.00 0.00 0.01 0.00 0.00 0.01 via I-5 0.00 0.00 0.00 0.07 0.00 0.01 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 via I-205 0.00 0.00 0.00 0.07 0.00 0.01 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 via I-82 0.00 0.00 0.00 0.07 0.00 0.01 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 via Hwy 97 0.00 0.00 0.00 0.07 0.00 0.01 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 Other International 0.00 0.13 0.01 0.00 3.70 0.35 0.00 9.04 0.84 0.00 0.52 0.05 0.00 0.59 0.05 Port of Seattle 0.00 0.00 0.00 0.00 3.70 0.35 0.00 9.04 0.84 0.00 0.52 0.05 0.00 0.59 0.05 Table 4.6. (Continued).
Developing Subnational Commodity Flow Data Using Supplemental Sources of Local Economic Activity 115 Step 3âEstimate Missing Data The process of estimating missing data was implemented in several of the interim steps of this process. For example, no information was available regarding the location of dehydration processing for potatoes grown in the Skagit Valley. Dehydration could take place at processing facilities in either the Upper or Lower Basins. On the basis of the expert knowledge of the Wash- ington State Potato Commission, it was estimated that 25 percent of these potatoes would be dehydrated in the Upper Basin, and 75 percent would be dehydrated in the Lower Basin. Potatoes grown in the Upper Basin were assumed to be processed in the Upper Basin, and potatoes grown in the Lower Basin were assumed to be processed in the Lower Basin. ConAgra staff also was used as industry experts in other cases in which missing data had to be estimated. 4.5 Next Steps This chapter provides a detailed description of several aspects of how to develop subnational commodity flow from supplemental sources on local economic activity. It might be useful at this point to identify some local commodity flow data sources. Many of these sources can be identi- fied by conducting an Internet search using publicly available search engines. Additionally, it might be helpful to consider private-sector participants. Consider what organizations they may belong to and how they might be able to help identify subnational commodity flow data sources for future freight planning efforts. Also, consider how these local sources of data might com- pare to commodity flow data captured from other methods such as establishment and roadside intercept surveys. Refer back to the Playbook section to identify the next portion of this Guidebook that will be most relevant to a particular stage in the data collection process.
