Freight Transportation Resilience in Response to Supply Chain Disruptions (2019)

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.

70 Table 7: Pharmaceutical Tonnage and Value between Miami and Houston, Base Case and Disruption Scenarios, Both Directions (2007) Mode Tonnage (Ktons) Tonnage % Value ($M) Value % Base Disruption Base Disruption Base Disruption Base Disruption Air (truck-air) 99 N/A 33% N/A 7,139,305 N/A 39% N/A Rail (truck-rail) 11 39 4% 13% 601,631 2,539,830 3% 13% Truck 189 260 63% 87% 10,578,271 15,779,376 58% 87% Total 299 299 100% 100% 18,319,206 18,319,206 100% 100% Source: FreightSIM model Similar to the table above, Table 8 shows before and after disruption results for pharmaceuticals transported from Miami to Houston area. As expected, truck dominates the mode split before and after the disruption. Moreover, truck is favored over truck-rail to carry the cargo that could not be shipped by air due to the disruption. In both tonnage and value, truck share increases by 23 percent while truck-rail share increases by only 9 percent. Table 8: Pharmaceutical Tonnage and Value between Miami and Houston, Base Case and Disruption Scenarios, One-Way (2007) Mode Tonnage (Ktons) Tonnage % Value ($M) Value % Base Disruption Base Disruption Base Disruption Base Disruption Air (truck-air) 90 N/A 31% N/A 5,049,924 N/A 31% N/A Rail (truck-rail) 11 37 4% 13% 601,631 2,067,908 4% 13% Truck 189 254 65% 87% 10,578,271 14,161,917 65% 87% Source: FreightSIM model The model results indicate that pharmaceutical distributors will primarily use truck for shipping their products if something disrupts the air mode. A few of the shipments are sent to rail. The model outputs are consistent with the results of the industry representative interviews. In the case of pharmaceutical shipments, third-party logistics providers constantly evaluate the decision to transport shipments by truck or air daily and make decisions for mode choice based on cost, time, airport delays, weather, delayed shipments further up the supply chain, and the like. “More truck capacity” is not usually a factor; much more emphasis is given to “cost and time.” Supply chain models address long-term planning impacts of changes in the transportation system and have not been developed as short-term operational models. This is one of the primary drawbacks of using a long-term planning model for a short-term disruption, although travel demand models have been used to analyze the network routing impact of short-term disruptions in the network. While the flows that move by air are expected to move to airports, FAF regions may include many airports. The FAA only reports the total tons moving through the airports and not the tonnages by commodity. Therefore, it is necessary to either assume that the commodity flows will utilize airports in proportion to the total tonnage carried, or search for additional and difficult-to-find datasets. The FreightSIM model forecasts the allocation of freight flows among supply chains but does not report “multiple modes” as an output. The forecast in the FAF only varies the flows to and from regions by commodity. The allocation among modes and supply chains for each flow is that which is reported in the base year. From the FAF, assuming the flows reported by “mail” are moving by air, the flows are almost evenly allocated between truck and air supply chains. For the purpose of “validating” outputs, given the disruption of the pharmaceutical air supply chain between Miami and Houston, we did assume that “multiple modes and mail” is in fact “air” (see Table 9).

71 Table 9: FAF Mode Descriptions for Pharmaceutical Disruption Scenario Mode ID  Mode Description Remarks 1 Truck Includes private and for-hire truck. Private trucks are owned or operated by shippers and exclude personal use vehicles hauling over-the-counter purchases from retail establishments. 2 Rail Any common carrier or private railroad. 3 Water Includes shallow draft, deep draft and Great Lakes shipments. 4 Air (include truck-air) Includes shipments typically weighing more than 100 pounds that move by air or a combination of truck and air in commercial or private aircraft. Includes air freight and air express. Shipments typically weighing 100 pounds or less are classified with Multiple Modes and Mail. 5 Multiple modes & mail Includes shipments by multiple modes, parcel delivery services, U.S. Postal Service, and couriers. This category is not limited to containerized or trailer-on-flatcar shipments. 6 Pipeline Includes shipments by pipeline and from offshore wells to land. 7 Other and unknown Any mode not included within the other mode definitions and unknown modes of transport. 8 No domestic mode Applies to some intra-zonal movements of imports Source: FAF3 Documentation Having to make such assumptions is a limitation of the model as applied here, in that it can only model transportation mode shifts. Going deep into the model to change the destinations or origins of a flow would require changes to the model that would be difficult to verify. Another application of a large-scale freight model for the grain supply chain from Illinois to New Orleans case study is found in Appendix C. 5.5 EVALUATING FREIGHT FLUIDITY In addition to examining freight supply chain models, the study team examined the concept of freight fluidity, especially as it was developed as part of a recent I-95 Corridor Coalition Freight Performance White Paper (Cambridge Systematics and WSP, 2016b). Freight fluidity refers to measuring supply chain performance across multiple jurisdictions using travel time, travel time reliability and cost measures. For example, one of the case studies undertaken as part of the I-95 Corridor Coalition Freight Performance White Paper evaluated the soybean supply chain between Peoria, IL and New Orleans, LA and identified data sources and potential freight fluidity performance measures that could be used to monitor the performance characteristics of the trip. The concept of disruption risk was explicitly included in the analysis. The modeling approach from the I-95 Corridor Study was at a much higher level than the modeling described in this final report, but it did provide useful insights about the participants (agents) that would be involved in this supply chain, and the impact on performance of disruptions anywhere along the supply chain. Freight fluidity can be a measure of the supply chain performance of a single or multiple modes of freight transportation. A freight network or freight corridor serving many supply chains can also use freight fluidity as a measure of the performance. The main elements of freight fluidity are travel time and reliability, travel cost, risk and safety. The risk element of freight fluidity consists of five different types of risk---operational risk, institutional risk, disruption risk, acceleration risk and deterioration risk. The analysis described earlier focused on the disruption risk and its implications for grains and pharmaceuticals supply chains, and thus has only covered one of the five risks associated with freight fluidity. However, to illustrate how freight fluidity measures could be formulated for freight flows, the following section describes the approach used in the I-95 Corridor Case Study.

72 5.5.1 Freight Fluidity Measures In general, freight fluidity describes how well a physical supply chain performs in a freight transportation network. Different transportation performance measures are considered in order to answer questions such as: how well are the links and nodes on the network operating? Where are the bottlenecks in the supply chain or freight network? How reliable is the network in terms of transportation costs? How well do supply chains and the freight system react to disruptions? What are the chances of the cargo damage or cargo lost on the network? To answer such questions freight fluidity is measured in terms of travel time, travel reliability, travel cost, travel risk and safety. Each is discussed in more detail below. Travel Time and Travel Reliability: The I-95 Corridor Freight Performance Measurement study found that the first three of these measures, i.e., travel time, travel time reliability and travel cost, can be measured using available public and private data. Travel time is defined as how long it will take to ship cargo from origin to destination. Roadway travel time data is available through the National Performance Management Research Data Set (NPMRDS), American Transportation Research Institute (ATRI) GPS data and Google Maps. Commercial data vendors such as Transcore and RSI Logistics provide rail travel time data. Railinc also has rail travel time data but it is not a commercial data vendor. The Nationwide Automatic Information System (NAIS) is a high-quality, public source of travel time on inland and coastal waterways collected by the U.S. Coast Guard and made available through the US Army Corps of Engineers. FHWA defines travel time reliability as the extent of unexpected delay, or “the consistency or dependability in travel times, as measured from day-to-day and/or across different times of the day.” Once travel time is available, travel time reliability can be calculated using different formulas, including planning time, planning time index, buffer time and buffer time index. The use of each formula is subjective and case dependent. Travel Cost: Travel cost is an integral component of freight fluidity since it reflects not only the shipping fees but also delay, unreliability, and wasted fuel on the network. Cost data can be purchased from private suppliers. In almost all cases, the cost data provided by shippers, receivers, and freight carriers are aggregated by a third party so that confidential shipper and carrier information is protected. Truck shipping rates can be purchased from Chainalytics, which is a consortium of Fortune 100-sized shippers. Regarding rail cost data, Chainalytics and STB Rail Carload Waybill Sample provide data for purchase. The U.S. Department of Agriculture (USDA) Marketing Services can be used for barge cost data. Other publicly available data sets include: 1. A CD containing Uniform Railroad Costing System (URCS) software can be downloaded for free and used to estimate rail carload costs. Go to: https://www.stb.gov/stb/industry/urcs.html. 2. Detailed representative long-haul trucking costs (O&M, fuel, labor, etc.) are reported annually by ATRI See: http://atri-online.org/ 3. The US DOT’s Intermodal Transportation and Inventory Costing Model State Tool (ITIC-ST) is another publicly available software tool. It can be used to examine shipment details between specific origins and destinations, also considering intermodal transfers between truck and rail. See: ITIC-ST, https://www.fhwa.dot.gov/policy/otps/061012/iticst_info.cfm These may all be “model-derived” costs but are all based on large empirical datasets that can, and have been, used in modeling flows.

73 Travel Risk: Quantifying travel risk is more challenging. Risks affecting product integrity, loss, damage and theft, as well as incorrect item counts are routinely monitored by freight shippers, receivers, carriers and their insurers, but this data generally are not made public and may not be compiled outside the respective organizations. Five different types of risks are identified: Operational Risk: this category encompasses immediate challenges to daily supply chain performance brought about by traffic congestion, weather, work zones, customs hold-ups, truck shortages, etc. Institutional Risk: This category covers risk to supply chain performance from uncertainties in the implementation of improvements, brought about by delayed action by federal, state, and local agencies. Disruption Risk: This category describes infrequent but possibly serious supply chain disruptions, such as those caused by major storms, tornadoes, earthquakes, infrastructure failure, labor disputes, political actions and wars. Analytical tools were used to forecast the reaction of grains and pharmaceutical supply chains to this type of risk. Acceleration Risk: This category encompasses conditions that may grow much worse rapidly. The typical example is a phase transition or state change in traffic flow, where a roadway incident turns slow moving traffic into gridlock. Deterioration Risk: This category covers conditions that gradually grow worse, causing performance to decline over time. Worsening congestion on roadways is the obvious example, imposing steadily lower reliability and higher buffering costs. Safety: Safety data are for the most part not readily accessible. The federally-supported Fatality Analysis Reporting System (FARS) data is the most accurate and readily accessible data on transportation-related fatalities. State DOTs maintain records of fatalities, injuries and property damage on state and local roads, but the data are usually reported by class of roadway or by type and severity of accident, rather than by highway corridor or roadway segment. 5.5.2 Freight Fluidity Case Study The case study in the I-95 Corridor Report is very similar to the cereal grain corridor modelled between Chicago and New Orleans (found in appendix C). While the scenario simulated with the CMAP freight models was for cereal grains (excluding soybeans) between northeastern Illinois and New Orleans LA, the case study in the I-95 Report examined soybean transportation between Peoria, IL and New Orleans, LA. In the case study, the factors of freight fluidity measured were travel time, travel time reliability, and travel cost. Due to data limitations, travel risk and safety were not measured. Despite the minor difference in the type of grain, and the focus on a farm in central Illinois rather than Northeastern Illinois, the corridors are otherwise identical. The freight fluidity calculated in the case study can be used as a proxy for understanding freight fluidity for cereal grains between Chicago and New Orleans. The path examined in the I-95 Corridor Report case study was, 1) truck delivery between a farm in El Paso, IL and a river port in Peoria, IL with access to the Illinois River, and then 2) by barge along the Illinois and Mississippi Rivers to a New Orleans export terminal. For the truck segment between El Paso and Peoria, travel time was calculated with data from Google Maps. Where available by Google Maps, the peak/off-peak or historical travel data can be used to estimate travel time reliability. Since this data was not available for the trip between El Paso and Peoria, a travel time index published in the Urban Mobility Report by the Texas Transportation Institute (TTI) was used to approximate the Truck Travel Time Reliability (TTTR) index with a travel-time multiplier.9 According to the Urban Mobility Report, for a small metropolitan area, the 95th percentile travel time is typically 11 percent more than free-flow travel time. The 11 percent converted into an index multiplier of 1.11 and was applied to the travel time calculated with Google Maps 9 Texas Transportation Institute, “2012 Urban Mobility Report,” http://d2dtl5nnlpfr0r.cloudfront.net/