This section provides an overview of available tools and methods that can be used in carrying out elements of the Analysis Framework presented in this guide. It is important to note that nearly all of the available tools and methods were designed for only one or two modes of freight transportation, though some of them can be applied together, and others can be adapted for use in analysis of multiple modes.
Overview. Transportation network and terminal performance models are a class of tools that forecast how different types of investments and different patterns of traffic affect the functional operating capacity of transportation facilities. They generally focus on capacity, speed, cost, and reliability characteristics of the facilities. Nearly all of these tools are specific to an individual mode of freight transportation (truck, rail, air, sea). The truck and rail models tend to focus on network performance, while airport and marine port models tend to focus on terminal or port facility performance. Still, they can be used to assess effects of improving a single surface mode, or to assess effects of improving access via a surface mode to an air or marine port facility.
For example, such models can address issues such as: a) how changes in road design, and shifts in the truck portion of traffic, will affect highway speed and throughput; b) how changes in truck-to-ship or rail-to-ship loading systems can affect speed and functional capacity at marine ports; and c) how changes in rail car classification yards, double-stacking and short-line interchange can impact effective railroad speed and capacity.
Highway Network Models. Many projects, even if not specifically a highway improvement, will have repercussions on the highway system. For example, adding capacity and improving service to marine, rail, and aviation facilities can reduce the number of truck trips needed and therefore will have a secondary impact to the remaining auto and truck travelers in the form of reduced highway congestion. To capture these effects and the benefits from adding capacity to the highway system, some form of highway travel demand models are most frequently used.
For highway investments, the most common models are either urban area or statewide travel demand models, with full highway network data. Separate truck models within highway network models are common to many states and MPOs. Whether a traditional four-step model of trip generation, trip distribution, mode choice, and trip assignment, or a simpler sketch-planning model, highway travel performance metrics in terms of highway volumes, speeds, travel time saved, operating cost changes, and safety effects are needed to quantify and monetize the benefits on the highway system.
Rail System Models. Average time and cost impacts on rail carriers and shippers can be calculated based on rail carrier cost and service simulation models.
The Uniform Rail Costing System (Surface Transportation Board) can estimate the changes in shipper productivity associated with rail system performance changes. The URCS model uses data on average carrier cost and performance measures to estimate the cost of providing service, so it can estimate how a change in facility capacity or speed (affecting rail cars per day) would translate into average shipper dollar savings per ton-mile.
Other models can be used to estimate how a given rail infrastructure improvement would actually change volumes, speeds, and reliability. While the data is often proprietary to the railroads, there are some generally recognized software tools. Rail Traffic Controller (Berkeley Simulation Software), RAILS 2000 (CANAC/Savage Industries), and RAILSIM (Systra) are all forms of simulation systems used by railroads to prioritize routing of trains through the network, identify conflicts and measure effectiveness. Besides the simulation systems, there has also been some work on “parametric rail capacity models” that develop capacity curves for various operating characteristics and identify areas with capacity constraints.
Terminal Models. One of the most critical issues in analyzing performance impacts of investments in nonhighway modes is the inconsistent and often illusory nature of capacity measurement in these modes. In the case of marine terminals and rail terminals/mainlines, operating practices and technology have as significant an impact on capacity estimates as do apparent physical capacity observations. There are, however, a variety of operational simulation tools (e.g., airport and marine terminal simulation models and rail operational simulations) that are used extensively in these industries to estimate capacity implications of alternative operating scenarios.
There are several variants of airport capacity models, which estimate of the capacity of runway systems and the level of delay that they present when faced with alternative demand levels. These include TAAM (Total Airport and Airspace Modeler) system, the Airfield Capacity Model (ACM) from MITRE Corp., the FAA’s Airport and Airspace Simulation Model (SIMMOD), and the LMI Runway Capacity Model from MIT. There is also the ACATS (Airport Capacity Analysis Through Simulation) model, which is an attempt to improve on the ACM framework.
Many port models have been refined by university researchers. These models typically account for both passenger and freight traffic, recognizing local differences in types of freight (bulk, break bulk, and containers), mix of ship characteristics, water depth, and wave motion, and positions of terminals. See for example: “The Seaside Port Capacity: A Synthetic Evaluation Model,” G. Malavasi and S. Ricci, University of Rome “La Sapienza,” published by Wessex Institute, WIT Transactions on the Built Environment, Volume 79. Also, “An Interactive Port Capacity Expansion Simulation Model,” C. S. Park and Y. D. Noh, Engineering Costs and Production Economics, Volume 11, Issue 2, 1987.
Overview. Modal Diversion models forecast how freight movements shift in response to changes in the availability, cost and/or time performance of available modal alternatives. Most modal diversion models used in transportation facility planning are focused on truck-rail-intermodal options because there are very real tradeoffs that shippers face when considering ground transportation options for medium and long distance travel. On the other hand, air and marine options focus more exclusively on long distance shipping and offer more distinctly different cost, performance, and availability features.
Total Logistics Cost Models predict how shippers respond to changes in the costs of modal and service alternatives. They actually estimate the total logistics cost of shipping, including direct transportation expense and inventory cost associated with modal lot sizes and service profiles. The models assume that customers (shippers) select the lowest cost option, and they depend on information about logistical factors in transportation and industry. Shipments are assigned to one mode or another, while allowing for probability uncertainty associated with inventory risk, carrier performance or unmeasured factors. Sometimes these models are based on detailed commodity-specific data. Other times, the models may be simple spreadsheet tools to estimate tons switching mode and resulting cost and travel time differences under different project assumptions.
Intermodal Transportation and Inventory Cost Model (ITIC) is a freight mode choice model from Federal Highway Administration’s Office of Freight Management and the Federal Railroad Administration. It attempts to calculate the logistics cost and decision tradeoffs seen by shipper logistics managers and then assigns the truck/rail diversion to alternatives that minimizes total logistics cost. It is based on an earlier model developed for FRA in 1995. (See Intermodal Transportation and Inventory Cost Model: Highway-to-Rail Version, U.S. DOT, FRA, and FHWA, December 2004; also Transmode Consultants, Inc., Truck-Rail, Rail-Truck Diversion Model: User Manual, developed for U.S. DOT, Federal Railroad Administration, 1995.)
Spreadsheet Logistics Model developed by MIT estimates the truck/rail mode choice for 48 typical types of customers. This is done on the basis of given customer characteristics (use rate and trip length), commodity characteristics (value/pound), and mode characteristics (e.g., price, trip time, and reliability) for rail, truck, and intermodal options. (See “Performance-Based Technology Scanning” Journal of the Transportation Research Forum Paper, 2002.)
Logistics cost models can also assess the cost savings of shipments using alternative different aircraft or ships (e.g., larger marine cargo ships that reduce costs per ton of goods). For air and marine facilities, models are most commonly simply matters of terminal volume/capacity measurement and forecasting (since obviously those modes do not have fixed right-of-way networks as exist for roads and rail).
Market Share Models are an alternative predictor of freight shipper choices. They do not estimate logistics costs. Instead, they are based on a statistical correlation between modal performance factors and traffic capture (revealed-preferences), and they then project traffic swings when relative performance changes. Stated-preference models have similar purposes but are developed statistically from structured interviews with freight transportation buyers about the tradeoffs they would make if faced with hypothetical choices. A statistical process is then applied to these responses to infer decision points and probable traffic diversions in response to changes in competitive service offerings.
For instance, the Intermodal Diversion Model from Global Insight, originally referred to as the “Reebie Intermodal Diversion Model,” estimates truck-rail diversion based on a combination of 1) the Uniform Rail Costing System, 2) TRANSEARCH commodity-flow database, and 3) a demand elasticity model calibrated from historical carrier price and volume data. The elasticities distinguish price sensitivity by traffic type, geographic region, and commodity group, and the model forecasts the specific freight flows that would likely be diverted to rail given changes in railroad or intermodal service characteristics.
Overview. Given direct travel performance impacts and mode switching impacts estimated by the two preceding types of models, it is a relatively straightforward process to assign dollar values to the changes in travel time, cost of shipping by mode, operating costs, schedule reliability, and other logistics factors. In addition, it is also possible to estimate the value of improvements in access from ports or terminals to markets.
State or Regional Highway Models calculate the dollar value of time, cost, safety, and reliability improvements associated with changes in highway system performance. Surface Transportation Efficiency Analysis Model (STEAM) is a benefit analysis tool for sketch planning of roadway transportation programs. It does not include roadway network features, but rather calculates traveler benefits at the regional or corridor levels. It calculates the economic value of benefits in travel time, accidents, nonfuel operating costs, and fuel costs. It can also distinguish eight commodity types and four vehicle types, and also account for air quality benefits.
ITS Deployment Analysis System (IDAS) is a related system that assigns values of time, and parameters for operating costs, emissions, and accident rates. IDAS also estimates benefits due to improved reliability based on traffic volumes and a reduction in nonrecurring delay. For freight movement in heavy traffic volume areas, reliability benefits can be a key component of the analysis. IDAS functions through direct interaction with a travel demand model network (see State or Regional Highway Network Models below) so the analysis can be conducted for specific links or zones in the model allowing the user much flexibility in identifying specific facilities, corridors, or regions for conducting the analysis. IDAS benefits analysis can be conducted by allowing IDAS to conduct the traffic assignment using network and zone inputs from an external model or it can use the results of a traffic assignment conducted by an external travel demand model and input into IDAS as a fully loaded network for benefits analysis.
State or Regional Highway Network Models operate at the more detailed link and node (network) level to calculate time and cost savings associated with changes in specific highway network inter-connectivity features or major improvement in connections between highways and special generators such as ports or intermodal rail terminals. Results of highway models can be translated into dollar values using values as shown in the AASHTO Red Book, or using broader factors that are discussed more fully in the Caltrans Benefit/Cost Guide (discussed later in Section 10.5).
Railroad Models likewise calculate the dollar value of improvements in rail system performance. RAILDEC – Railroad Decision Model from the Federal Railroad Administration is a family of software tools designed to evaluate the economic benefits from proposed rail-related infrastructure benefits. It is designed for use by metropolitan planning organizations (MPOs) and state DOTs to conduct benefit/cost analyses for railroad-related projects and highway-rail interactions (including grade crossings). It also operates within a risk/uncertainty analysis framework.
Rules of Thumb on highway and rail freight improvements can also be used to assign money values to freight transportation benefits. Sources include:
Market Access Models estimate how business activity (generating freight demand) can shift among locations when new freight routes and ports open up or improve access to areas that were previously not attractive as freight generators. For instance, a new international gateway (such as an international marine port, airport, or border facility) or a new link between ports and markets (such as a rail line or highway link) can bring enhanced productivity at a national level and economic growth to depressed areas that would not otherwise see such growth. Whereas network performance and modal diversion models estimate cost savings for freight between fixed origins and destinations, a market access model offers a complementary measure of additional net income growth that may be beyond the cost savings.
The Local Economic Assessment Package (EDR-LEAP) estimates the magnitude of potential opportunities for regional business expansion and attraction resulting from highway or rail projects that affect a community or region’s market access and connections to outside areas. This may include access to customer/supplier delivery markets, transportation terminals (including airports, marine ports, and intermodal rail transfer facilities), international borders or industrial centers. These types of business access benefits are in addition to the simple travel time and cost savings benefits that are traditionally recognized in transportation planning models. A predecessor model that focused on highway connectivity was called Highway Economic Opportunities Model (ARC-Opps or HWY-Opps). This type of tool can be used as an adjunct to REMI, REDYN or Global Insight models to forecast long-term economic impacts of connectivity improvements. It is also included as part of the TREDIS framework for benefit/cost evaluation, discussed later.
Background material on methods for measuring and valuing market access effects of truck transportation improvements are included in NCHRP Report 456 (Guide for Assessing Social and Economic Effects of Transportation Projects) and in NCHRP Report 463 (Economic Implications of Road Congestion). Discussion of highway connectivity to airports, marine ports, and intermodal terminals is discussed in reports of the Appalachian Regional Commission (Handbook for Assessing Economic Opportunities from Appalachian Development Highways, Economic Development Research Group with Cambridge Systematics, 2001).
Overview. Economic impact models are frequently used to convert direct economic effects into broader regional/macroeconomic impacts on measures such as employment by industry, gross regional/state product, and personal income. This can include direct benefits from savings in business costs for current shipping patterns, and/or economic growth benefits from improved market access. The most frequently used models are dynamic, time-series economic simulations. Sometimes, static input-output models are applied in conjunction with price elasticity response calculations to accomplish these same results. Application of these various types of models for multimodal transportation projects are summarized in NCHRP Synthesis Report 290, Current Practices for Assessing Economic Development Impacts from Transportation Investments, NCHRP Synthesis of Practice, TRB, 2000. Links to various economic simulation models are offered on the web site of the TRB Committee on Transportation and Economic Development (http://www.tedcommittee.com).
Economic Forecasting Models have been developed by both commercial vendors (e.g., Global Insight, Economy.com) and noncommercial economists (often universities) to project future economic trends either for entire regional economies or individual industries. Various OECD and UN products also provide industrial analysis forecasts at a more global level. These projections can be useful to project future demand for transportation services and help identify potential capacity constraints, and they can also be used to show historical relationships between transportation infrastructure/investment (capital stock) and economic growth (or industrial performance). However, these latter types of models do not provide the necessary “levers” to be sensitive to the nuances of modal service changes (e.g., scheduling, reliability, capacity, or accessibility) associated with individual freight transportation projects.
Input/output (I/O) models such as IMPLAN and RIMS-II are essentially variations of accounting tables that track the buying/selling interrelationships between industries within given regions. They reflect forward and backward linkages in the flow of money, associated with business suppliers and consumer spending. They can thus capture the full economic impacts (including multiplier effects) derived from changes in demand or output in a given industry. However, traditional I/O models alone are not designed to easily estimate how impacts vary over time or capture the business cost effects of transportation improvements. The I/O models have been used together with industry-specific cost response or logistics models to calculate the broader growth effects that transportation projects can have on various industries.
Regional Simulation Models include both “General Equilibrium” and “Structural Economic Simulation” models such as the REMI, REDYN, Global Insight, Inforum, FAIR, and REAL models. They combine features of input/output models with the long-term elements of forecasting models, to forecast the economic growth trajectory of industries between multiple regions under baseline conditions and alternative scenarios. These models can and have been used to forecast the spatial restructuring of business activities resulting from changes in comparative business costs among regions. Regional Simulation Models still require travel model results to determine the direct transportation cost effect, and they also require exogenous analysis to capture the full range of changes in market access and associated economies of scale that can also result from major transportation projects. Of note, these models forecast changes in regional growth as shares of a closed national economy, so they do not allow for changes in international trade (which is important for projects serving ports/borders).
Production Function Models encompass a class of industry-specific equations and tools that forecast how businesses evaluate supply chain and production options to optimize size, locational dispersion of siting, production processes, and distribution channels. They are typically sensitive to relative changes in market prices and costs of capital, labor, and transportation.
Overview. Project evaluation has several major elements. It is desirable to estimate costs and benefits of any major freight transportation project over a suitable timeframe and compare them to alternative plans of action, considering financial and social, public and private benefits and costs. However, techniques differ in breadth of coverage and consideration of incidence of these factors.
Benefit/Cost Analysis (BCA) is fundamentally a comparison of all of the positive and negative impacts of a project expressed on a consistent basis in terms of net present values. While this is clearly an attractive methodology, one of the major criticisms of BCA is that some of the impacts are not measurable in dollar terms. As a result, BCA studies frequently estimate the total monetary value of benefits and costs for travelers and transportation agencies, leaving out other positive and negative impacts that are not measurable in money terms (which are then dismissed as immeasurable “externalities”). Some studies have attempted to convert other impacts into monetary terms through surveys that derive “willingness to pay” for impacts (“stated-preference”) or observations of actual choices as reflected in property values.
Another limitation of BCA is that it is designed to aggregate all benefits and all costs, without regard to their incidence. In the case of major freight projects, this means that it ignores the different roles of public agency and private investment functions, which need to considered in evaluating opportunities for “win-win” propositions in public-private partnerships. However, it is possible to evaluate the incidence of benefits and costs by conducting BCA separately for subgroups that may include private carriers, various industries, and the general public. Such an approach is recommended for the analysis framework in this report.
Multimodal BCA. There are several forms of multimodal models for benefit/cost analysis. TransDec: Transportation Decision Analysis Software, developed as part of NCHRP 20-29 (2), is designed for evaluation of transportation investment decisions spanning multiple modes of ground transportation. This package is notable because it is explicitly concerned with freight as well as passenger transportation. It structures a process of evaluating transportation investments on the basis of multiple objectives, such as improved accessibility, connectivity, cost-effectiveness, resource impact, and economic growth. The system is designed for comparing multiple alternatives, with minimum performance thresholds.
TREDIS: Transportation Economic Development Impact System is a web-based evaluation system that is most notable for its distinctions between both freight and passenger benefits, and its simultaneous coverage of roadway, railroad, aviation, and maritime transportation. This economic analysis system evaluates how changes in transportation costs and accessibility relate to the operating requirements of various industries and resulting productivity and growth. It works with the REMI model, REDYN model or IMPLAN model with cost response factors, and then processes results to show benefits and costs from alternative perspectives.
Three web sites that contain guidelines for conducting multimodal transportation benefit/cost analysis are:
Internal Rate of Return. The internal rate of return (IRR) is the discount rate that makes the Net Present Value (NPV) of all cash flows equal zero. It is particularly useful for investments that require and produce a number of cash flows over time. Technically, IRR is a discount rate: the rate at which the present value of a series of investments is equal to the present value of the returns on those investments. As such, it can be found not only for equal, periodic investments but for any series of investments and returns. This makes IRR an attractive approach in the private sector. However, this method is problematic, as it assumes that all of the intermediate cash flows can be discounted/reinvested at the IRR. This is particularly unrealistic when the IRR is very high. This method is also sensitive to the sequencing and timing of investments and returns.
Cost-Effectiveness Analysis (CEA) differs from BCA in that it does not seek to simultaneously evaluate all positive and negative impacts, and it does not require that all positive and negative effects be boiled down to a common measure of dollars. Rather, CEA compares the effectiveness of project alternatives in achieving various individual indicators of desired benefits (such as reducing congestion and improving air quality and freight flow). However, CEA is limited as it examines single dimensions of impact that may affect different parties (travelers, shippers, or transportation providers), and it still does not differentiate coincidence of costs.
Multiple Criteria Appraisal (MCA) is most popular in Europe as a more comprehensive alternative to the use of BCA. It provide a means of considering the wider issues of qualitative and quantitative benefits and costs, as well as distribution and equity of their incidence, in a unified framework based on rating criteria.
Guidance on Methodology for Multimodal Studies (GOMMs) is a tool that implements the UK Treasury’s Green Book for appraisal of alternatives for public sector funding as it applies for transportation projects regardless of mode. The tool lays out all the various considerations of accessibility, economic, environmental and distributional impacts through use of Appraisal Summary Tables (ASTs). There is a worksheet for rating “Transport Economic Efficiency” from the perspective of consumers, business, private sector providers, and developers. There are also separate worksheets for rating “Public Accounts” from the perspective of Local and Central Governments.
Scottish Transport Appraisal Guide (STAG) is a variant of GOMMS used in Scotland. It also builds on the concept of Appraisal Summary Tables, with a concise rating form for assessing project alternatives in terms of seven factors: 1) social and economic context, 2) planning objectives and measures of performance along them, 3) project rationale, 4) fit with land use and other policies, 5) implementability, 6) efficiency for conventional transport user benefits and costs, and 7) economic impacts in terms of employment and GDP (gross domestic product – a measure of economic output).
NCHRP 8-42, Guidebook for Assessing Rail Freight Solutions to Roadway Congestion, Global Insight and Economic Development Research Group for the Transportation Research Board, 2006.
Internet Guide to Transportation Benefit/Cost Analysis, Caltrans, 2005, http://www.dot.ca.gov/hq/tpp/offices/ote/Benefit_Cost/index.html.
NCHRP Research Results Digest 258, Development of a Computer Model for Multimodal, Multicriteria Transportation Investment Analysis, Texas Transportation Institute for the Transportation Research Board Washington, D.C., 2001, http://onlinepubs.trb.org/onlinepubs/nchrp/nchrp_rrd_258.pdf#search=%22NCHRP%20Report%20258%22.
NCHRP Synthesis Report 290, Current Practice for Assessing Economic Development Impacts from Transportation Projects, Economic Development Research Group, Inc for the Transportation Research Board, Washington, D.C., 2000, http://www.edrgroup.com/pages/pdf/synth290.pdf.
Transportation Research Circular 477, Assessing the Economic Impact of Transportation Projects: How to Match the Appropriate Technique to Your Project, Weisbrod, G., and B. Weisbrod, Transportation Research Board, Washington, D.C., 1997, http://gulliver.trb.org/publications/circulars/circular477.pdf.
Scottish Executive, Scottish Transport Appraisal Guidance, 2003, http://www.scotland.gov.uk/Topics/Transport/integrated-transport/stag.
United Kingdom Department of Transport, Labor and the Regions (DTLR), Guidance on the Methodology for Multimodal Studies (GOMMS), London, England, 2003, http://gulliver.trb.org/publications/circulars/circular477.pdf.
European Commission, Guide to Benefit/Cost Analysis of Investment Projects, prepared for Evaluation Unit, DG Regional Policy, 2003, http://ec.europa.eu/regional_policy/sources/docgener/guides/cost/guide02_en.pdf.
Federal Aviation Administration, FAA Airport Benefit/Cost Analysis Guidance, Office of Aviation Policy and Plans, Federal Aviation Administration, U.S. DOT, 1999, http://www.faa.gov/airports_airtraffic/airports/aip/bc_analysis/media/faabca.pdf.
Federal Railroad Administration, GradeDec.Net Benefit/Cost Analysis Tool, U.S. DOT, 1998.
NCHRP 8-36, Return on Investment on Freight Rail Capacity Improvement, Cambridge Systematics, Inc. and Reebie Associates, Inc. for the American Association of State Highway and Transportation Officials (AASHTO), 2005, http://www.transportation.org/?siteid=30&pageid=1399.
Federal Highway Administration, Benefit/Cost Forecasting Toolbox for Highways. U.S. DOT, Washington D.C., 2001, http://gradedec.fra.dot.gov/.
Federal Highway Administration, Freight: Benefit/Cost Methodology, FHWA Freight Management and Operations, U.S. DOT, FHWA-HOP-05-060, 2005, http://ops.fhwa.dot.gov/freight/intermodal/cost_bene_meth/cost.htm.