INTERMODAL TERMINALS DATABASE:

Concepts, Design, Implementation, and Maintenance




DAVID P. MIDDENDORF

Center for Transportation Analysis
Energy Division
Oak Ridge National Laboratory



January 1998



Prepared for
Bureau of Transportation Statistics
U.S. Department of Transportation


EXECUTIVE SUMMARY

This report describes the design, construction, and maintenance of a geographic database of intermodal freight terminals, which are places or facilities where freight is transferred between different modes of transportation. The intermodal terminals database is itself a component of the National Transportation Atlas Database (NTAD) being developed and maintained by the Bureau of Transportation Statistics (BTS) of the U.S. Department of Transportation (DOT). The NTAD consists of computer files which describe the location, topology, and attributes of the nation's highway, railroad, waterway, air, and pipeline networks. The purpose of the intermodal terminals database is to connect the various modal networks within the NTAD so that multimodal transportation analyses can be performed.

THE BASIS FOR THE INTERMODAL TERMINALS DATABASE

The recent growth of and interest in intermodalism motivates the need for a database of intermodal terminals. Unfortunately, the concept of intermodalism has often been equated with containerization or with the notion of one mode of transportation carrying the equipment of another mode, such as railcars transporting loaded highway trailers. While containerization is a technological innovation that has greatly facilitated the use of intermodalism for a wide variety of commodities, intermodalism itself is a much broader concept. It occurs whenever two or more modes of transportation meet for the purpose of exchanging cargo - either directly or through intermediate storage. In its broadest sense, intermodalism refers to "a holistic view of transportation in which individual modes work together or within their own niches to provide the user with the best choices of service, and in which the consequences on all modes of policies for a single mode are considered [FELD96]."

The rapid evolution and growth of intermodalism since the 1950s is due to technological advancements and regulatory reforms which have helped to overcome or mitigate some of the historical impediments to intermodalism. Containerization and sophisticated cargo-handling equipment have addressed some of the physical or mechanical barriers to more efficient intermodalism, while the introduction of computer and telecommunications technology is beginning to alleviate some of the often more difficult organizational and logistical impediments. The economic deregulation of the transportation industry that occurred in the United States in the late 1970s and early 1980s has had a tremendous impact on the provision of transportation services in general and the practice of intermodalism in particular.

There are many reasons why shippers and carriers engage in intermodalism. In some traffic lanes, depending on a particular shipper's own unique needs and circumstances, a combination of modes may serve the shipper better than a single mode. A shipper, for example, may utilize one mode of transportation to access another mode that provides certain advantages such as low cost, high speed, or high capacity. Shippers will sometimes resort to intermodalism to avoid being dominated by a single mode or carrier, thereby reducing their transportation costs and increasing their transportation options. Despite intense competition between modes for certain kinds of freight, carriers also make use of intermodalism when it suits their purposes. Intermodalism, for example, may help a carrier reduce its operating costs or gain new business that it might not be able to serve directly.

Government involvement in intermodalism has changed dramatically. For many years the federal government's philosophy toward regulation of the transportation industry either prohibited or discouraged any attempts by carriers from different modes to coordinate or integrate their services as well as any efforts by shippers to arrange for more coordinated or integrated multimodal transport. By 1991, however, this philosophy was completely reversed as a result of the Intermodal Surface Transportation Efficiency Act (ISTEA). This landmark piece of legislation required metropolitan planning organizations (MPOs) to consider ways of enhancing the efficient movement of freight and improving access to ports, airports, intermodal transportation facilities, and major freight distribution routes as they prepared their transportation improvement plans. It also required DOT to develop a database containing information on public and private investment in intermodal transportation facilities and services, and enabled federal funds to be used for intermodal projects. In authorizing the designation of a National Highway System (NHS), the ISTEA specified that the latter should include highways that provide connections to major intermodal terminals.

Unfortunately for government policymakers and transportation analysts, intermodalism makes policy analysis much more difficult. The relationships between modes and carriers have become so intricate that changes or problems in one mode or in one part of the country can have a profound impact on other modes and other geographic areas. Modal choice is no longer simply a matter of choosing between truck, rail, air, water, or pipeline; that is, it is not a matter of selecting between modes but a matter of selecting between services. Traditional analytical tools and databases may not be adequate to address the current multimodal and logistics reality. Satisfactory multimodal analysis will require new tools and databases.

An important function of the intermodal terminals database is to support any analysis dealing with the flow or movement of freight through a multimodal transportation network. Another important function is to support a variety of mapping and geographic information system (GIS) applications. More than anything else, the intermodal terminals database described in this report is fundamentally an inventory of existing facilities. It documents what is on the ground. It is hard to imagine how policymakers and analysts can begin to understand the impact of policy issues and initiatives on intermodal transportation without having some sense of the size and scope of the existing intermodal infrastructure.

CHARACTERISTICS AND TYPES OF INTERMODAL TERMINALS

Just as some confusion exists about the meaning of intermodalism, the notion of an intermodal terminal also conjures up different images. It is impossible, however, to design and build a consistent and coherent database of intermodal freight terminals without having a clear idea of what constitutes an intermodal terminal. How does one recognize an intermodal terminal or facility? What distinguishes one intermodal terminal from another? What distinct types of intermodal terminals can be identified?

A freight terminal is an integrated set of facilities where cargo is loaded onto or unloaded from a particular mode of transportation. An intermodal freight terminal is a special kind of freight terminal. It is a place where two or more modes of transportation meet to interchange freight, either directly or through intermediate storage. Intermodal exchange may not be the only function performed at an intermodal terminal, or even the primary one. All that is required for a freight terminal to be an intermodal terminal is that it have the necessary space and equipment to receive cargo by one mode of transportation and ship it out by a different mode. In between the inbound and outbound movement, the cargo may be consolidated with other incoming cargo of the same type, separated into smaller outbound shipments, or directly transferred between two modes as part of a seamless intermodal shipment.

The physical characteristics, complexity, and other attributes of intermodal freight terminals vary greatly. Some intermodal terminals are relatively small and almost ad hoc in nature, while others are large and involve a considerable amount of infrastructure. As an aid in identifying and classifying intermodal terminals and thereby gaining a better understanding of the concept of intermodalism in its various forms, five features or characteristics of intermodal terminals are especially useful: the pairs of modes which the terminal directly or indirectly connects, the types of cargo or the specific commodities which the terminal handles, the types of intermodal transfers for which the terminal is designed (direct, short-term storage, or long-term storage), whether the terminal is privately or publicly owned, and whether use of the terminal is open or restricted relative to either shippers or carriers.

By employing combinations of the above five distinguishing features, several common types of intermodal terminals can be found. Between the truck and rail modes, for example, containerized freight is interchanged at TOFC/COFC facilities, automobiles and other finished vehicles at vehicle terminals, dry and liquid bulk cargoes at bulk transloading facilities, and breakbulk commodities such as lumber, steel, and paper products at numerous public warehouses, distribution centers, and other reload facilities. Grain is transferred between truck and rail, truck and water, or rail and water at thousands of grain elevators and other grain-handling facilities. Petroleum products and other liquid bulk commodities such as chemicals, vegetable oils, and molasses are gathered and distributed intermodally at hundreds of tank farms and other liquid bulk terminals and storage facilities. Along the Atlantic, Pacific, and Gulf coasts, the Great Lakes, and the inland waterways can be found a variety of intermodal terminals including container-handling facilities, import and export auto terminals, grain elevators, cement terminals, rail-barge and rail-vessel bulk transloading facilities for coal, ores, and other dry bulk cargoes, and facilities for storing and transferring petroleum and non-petroleum liquid bulk commodities. Clearly, an amazing variety of intermodal terminals exists.

DESIGN OF THE INTERMODAL TERMINALS DATABASE

Designing a database of intermodal terminals is a matter of deciding what information to include and how to organize it. A long list of possible data items describing various physical and operational characteristics of intermodal terminals could easily be generated. Simply compiling all or a large subset of these items into a single record for each terminal would likely lead to a database too cumbersome and costly to build, maintain, and use. A poorly designed database, like a poorly designed bridge, will soon collapse under use. What is needed is a logical database structure based on a conceptual model of intermodal terminals and their relationships with other elements of the transportation system.

In determining the contents and organization of the intermodal terminals database, several factors were considered. Because the terminals database is a component of the broader NTAD, the specification of the latter prescribes an overall organizational structure as well as some standard file formats for the terminals database. As a result, information on terminal location is placed in a file of point-type records, while non-topological attribute data are stored in one or more files of attribute-type records. Other factors considered in deciding what information to include in the terminals database centered around the need for independence from any particular modal network database, the possible uses of the terminals database, technical problems associated with measuring or specifying terminal costs, capacity, and performance, and the availability of data and other resources for building and maintaining the database. As a further design aid, a simple conceptual model of intermodal terminals was created to highlight the important features of these facilities and their connections with the various modal subnetworks. This conceptual model revealed two important pieces of information about intermodal terminals that should be included in the database. One is the terminal's location, particularly its geographic coordinates. The other is the notion of an intermodal connector, which itself combines three other pieces of information: a pair of modes, a commodity or type of cargo, and the direction of transfer.

The resulting structure of the initial intermodal terminals database is based on the relational data model. It consists of two relational tables or files.

The first file in the database structure is the terminal location table or point file. It contains information on the identity and geographic coordinates of each intermodal terminal. The primary fields of this file are the following:

The second file in the basic intermodal terminals database structure is the intermodal connections table or attribute file. It describes the intermodal connections that can occur at each intermodal terminal. Each record in this file includes the following fields:

Because an intermodal terminal may connect more than one pair of modes or transfer more than one type of cargo between one or more pairs of modes, it may have multiple records in the intermodal connections file.

These two tables or files constitute what might be termed the basic or canonical version of the intermodal terminals database, since terminals of every type have a physical location and at least one intermodal connection. Additional files or tables of logically related attributes can be added to the canonical terminals database in the future as the need arises and as more or better data become available.

BUILDING THE INTERMODAL TERMINALS DATABASE

The process of adding records to the intermodal terminals database involved three primary steps or tasks.

The objective of the first task was to identify intermodal terminals of a particular type such as TOFC/COFC terminals, truck-rail bulk transfer facilities, or waterway terminals. The search for lists, directories, and other sources of information on the locations of intermodal terminals followed many avenues. Links or references to potential sources were revealed through a continuous monitoring of the freight transportation literature, including not only research reports and journal articles but also news publications, industry and trade magazines, and popular railroad and transportation periodicals. Other sources were identified by contacting government agencies, industry and trade organizations such as the AAR and the American Association of Port Authorities (AAPA), persons in the railroad, trucking, and transportation equipment industries with experience in intermodalism, and researchers who have recently studied some aspect of intermodal transportation. Various World Wide Web search mechanisms were also used to uncover online sources as well as references to additional published sources.

Once the presence or identity of an intermodal terminal had been discovered from a list, directory, or other source, the terminal's location had to be pinpointed for inclusion in the terminal location or point record. Various map resources were utilized in an effort to locate each intermodal terminal as precisely as the available address information would allow. These resources included printed highway maps and railroad atlases, aerial photographs, and online maps available on the World Wide Web. In many cases, the terminal operator had to be contacted by phone to obtain directions to the intermodal facility. A CD-ROM map database software package called MapExpert, published by DeLorme Mapping Company, was used to locate the position of each intermodal terminal and measure its longitude and latitude.

The third step in adding a terminal to the database was to determine its intermodal connections; that is, ascertain what types of cargo or commodities are interchanged between which modes of transportation at the terminal. For some kinds of terminals, such as TOFC/COFC facilities and auto ramps, this was a trivial matter, but for many waterway and other types of terminals, the intermodal connections were not always so obvious. Information on the kinds of freight handled at a terminal was sometimes sketchy, ambiguous, or unavailable. Some of the principal data sources did not always clearly indicate what role, if any, railroads played in exchanging freight with other modes at a terminal, or which modes of transportation were involved in the transfer of a particular commodity or type of cargo. Whenever possible, supplementary or secondary information was used to help in resolving these ambiguities. In a number of cases, the terminal owner or operator had to be contacted by phone to obtain the needed information.

Among the principal published sources of information on intermodal terminals used to build the database were the following:

Many World Wide Web sites were also mined for data, including those of several Class I railroads and a number of port authorities which provided detailed information on one or more types of intermodal terminals, including TOFC/COFC facilities; auto ramps; bulk transload facilities; lumber, steel, and paper reload facilities; and waterway terminals. Supplementing these primary sources of data were brochures and other marketing materials supplied by several motor carriers, railroads, terminal operators, and port authorities.

MAINTAINING THE INTERMODAL TERMINALS DATABASE

As of October 1997 while this report was being written, the intermodal terminals database contained records for 2865 intermodal freight terminals and 9036 intermodal connections. Despite these large numbers, the database was only partially completed. A full database could easily contain two to three times this number of records. Serious consideration must be given to the problem of keeping a database of this size up-to-date in a reasonable and cost-effective manner.

The amount of effort required to keep the database current depends on the volatility of the data, which in turn depends on how much and how fast the intermodal infrastructure is changing. Based on the large number of news items pertaining to intermodal facilities that have appeared recently in various transportation news sources, it is apparent that the intermodal infrastructure is by no means static. New terminals have recently opened, and others are either being built or in the planning stages. At the same time, existing terminals are being expanded or consolidated, and older outmoded or underutilized facilities have recently been and will continue to be closed. Changes to the intermodal infrastructure are being driven by the continued growth of containerized traffic, recent and proposed railroad mergers, the formation of ocean carrier alliances, technological advances, freight rate incentives, the availability of federal funding for intermodal projects, and many other events and factors. While the number of intermodal terminals and intermodal connections added to or removed from the transportation system over the course of a year may be relatively small compared to the number of terminals in existence, it is nevertheless significant. Systematic maintenance of the terminals database is therefore essential.

The proposed maintenance strategy has two components. The first is a monitoring activity. Various printed and online transportation news sources should be monitored on a regular basis to keep abreast of current events involving intermodal facilities. A record should be kept of each new, planned, or closed terminal and each new, planned, or discontinued intermodal connection cited in news articles, press releases, and short news items. The second component of the database maintenance strategy consists of a multi-year cycle of incremental updates. The database should be updated each year to incorporate the terminal openings and closings uncovered by the previous year's monitoring activity. Likewise, there should be no problem in keeping the terminals database current each year with regard to TOFC/COFC facilities, given the importance of this type of terminal and the availability of good data. As each revised edition of a Port Series report is released by the U.S. Army Corps of Engineers, it should be used to update the terminals database. Likewise, the auto terminals records should be updated as soon as the AAR issues a new directory of motor vehicle loading and unloading terminals. The remaining types of intermodal terminals should be considered on the basis of a multiyear update cycle, in which attention would be focused on a different type of terminal each year.

CHAPTER 1. THE BASIS FOR THE INTERMODAL TERMINALS DATABASE

This report describes the design, construction, and maintenance of a geographic database of intermodal freight terminals, which are places or facilities where freight is transferred between different modes of transportation. The intermodal terminals database is itself a component of the National Transportation Atlas Database (NTAD) being developed and maintained by the Bureau of Transportation Statistics (BTS) of the U.S. Department of Transportation (DOT). The NTAD consists of computer files which describe the location, topology, and attributes of the nation's highway, railroad, waterway, air, and pipeline networks. These files are designed for use within a geographic information system (GIS) to enable transportation analysts to study and measure the extent, usage, and performance of the U.S. freight transportation system and to determine the effects of planning and policy initiatives, capital and operational improvements, economic trends, natural disasters, and other major events on the system. [HANC94, SPEAR95].

The purpose of the intermodal terminals database is to connect the various modal networks within the NTAD so that multimodal transportation analyses can be performed. To foster a better appreciation of the need for an intermodal terminals database, this chapter provides a brief overview of the concept of intermodalism. Because different segments of the transportation industry have defined intermodalism in somewhat narrow terms, the chapter begins by defining the concept in a much broader, more general sense. It then discusses the evolution of intermodalism and the reasons why it occurs. This is followed by a discussion of why many current transportation issues, policies, and initiatives cannot be fully analysed without taking intermodalism into consideration. The chapter concludes by describing the role of the intermodal terminals database in addressing issues concerning multimodal transportation.

INTERMODALISM DEFINED

Ask a railroad or motor carrier official to define the term "intermodal" and the chances are you will receive an answer involving highway trailers and containers riding on railroad flatcars or articulated double-stack railcars. Likewise, ask a port operator the same question and you will likely hear about containerships, container cranes, container freight stations, and TEUs (twenty-foot equivalent units). This close association of intermodalism with containerization or with one mode of transportation carrying the equipment of another mode is unfortunate. As Jennings and Holcomb [JENN96] point out, it greatly "limits the research conducted in this area, and ultimately the potential to create an integrated transportation system." Davis Helberg, Executive Director of the Seaway Port Authority of Duluth, stated the issue succinctly in a recent editorial [HELB97]:

"Where is it written ... that intermodalism is the holy domain of the container? ... When iron ore was discovered on Minnesota's Iron Range in the 1890s, it began to be shipped - and still is - on specially built railcars to specially built docks for carriage by specially built ships to specially built docks for delivery to steel mills by (what else?) specially built railcars. Fast, competitive, and with minimum lost motion, it's a system that keeps Duluth-Superior the largest ore shipping port in the U.S. - and as close to seamless as one gets short of a pipeline."

Containerization is a technological innovation that has greatly facilitated the use of intermodalism for a wide variety of commodities, but it is not synonymous with intermodalism. As Helberg demonstrated, containers are not required to effect a fast, efficient transfer of cargo between modes of transportation. Intermodalism, however, is an even broader concept than Helberg's example implies for he describes a situation where freight is directly transferred between two modes. In the more general case, the cargo may be stockpiled or placed in temporary storage before being loaded into or onto another mode's equipment.

Intermodalism occurs whenever two or more modes of transportation meet for the purpose of exchanging cargo - whether directly or through intermediate storage. The BTS publication Transportation Expressions [FELD96] defines the concept as follows:

"In its broadest interpretation, intermodalism refers to a holistic view of transportation in which individual modes work together or within their own niches to provide the user with the best choices of service, and in which the consequences on all modes of policies for a single mode are considered."

By adopting this broad definition of intermodalism for the design of an intermodal terminals database, we do not limit ourselves to certain kinds of facilities such as TOFC/COFC (trailer on flatcar/container on flatcar) ramps or to certain kinds of intermodal operations such as direct transfers. Rather, we open up the database to include facilities such as grain elevators, cement terminals, bulk transfer facilities, and transloading docks at public warehouses and distribution centers as well as operations such as reloading, bulk transloading, cross-docking, consolidation, and distribution.

Note that the above definition of intermodalism says nothing about individual shipments. Some people in the transportation industry like to distinguish between intermodal shipments and multimodal shipments. The former term is used to describe a shipment involving more than one mode of transportation between origin and destination where the cargo does not have to be unpacked and reloaded when changing modes. It therefore implies either containerization or the transfer of one mode's means of conveyance, such as a highway trailer, to another mode's means of conveyance, such as a railroad flatcar. A multimodal shipment, on the other hand, is one in which the cargo does have to be unpacked and reloaded. Both types of shipment, however, involve intermodalism.

The above definition of intermodalism also does not depend on any notion of what constitutes a shipment. Instead, it recognizes that intermodalism often involves the consolidation of many relatively small shipments into a large one or, conversely, the breakdown of a large shipment into smaller ones. For example, two 100-car unit train shipments of coal may arrive at a Mississippi River dock for consolidation into a single 15-barge tow shipment. A surface freight forwarder collects small shipments from several clients, loads them according to their destination into containers, highway trailers, or boxcars at a warehouse or freight station, and turns them over to a railroad as one large shipment from an origin to a destination. A train unloads a twelve-carload shipment of lumber at a distribution center, where the cargo is then divided into a number of small truckload shipments destined to various nearby lumber yards and building supply outlets. In each case, shipments are either merged or broken down at an intermodal facility.

THE EVOLUTION OF INTERMODALISM

Although the importance and benefits of intermodalism have only recently been articulated by the federal government and incorporated into national transportation policy, the intermodal concept is perhaps as old as transportation itself. Since intermodalism occurs whenever and wherever two or more modes of transportation meet to interchange cargo, it quite naturally began when civilizations started using navigable waterways to move goods. It developed even further when they started shipping goods across the oceans. The first intermodal terminals were thus located at strategic places where land and water met.

Even the idea of one mode carrying the loaded equipment of another mode in "piggyback" fashion has been around for quite a while. Early American settlers often loaded their covered wagons onto barges and flatboats for trips down a river or canal, and horse-drawn vehicles were rolled on and off ferries, barges, and rafts. The Pennsylvania Canal between Philadelphia and Pittsburgh had both water and overland sections. In between the canals, barges carrying both passengers and cargo were hauled overland by railroad or aboard horse-drawn wagons [MULL95]. Photographs exist showing loaded Conestoga wagons sitting atop Union Pacific Railroad flatcars in Oregon during the 1880s [WOOD96]. Starting in 1885 and continuing into the 1890s, the Long Island Railroad carried farmers along with their horses and farm wagons loaded with produce into New York City [ARMS93].

Examples of other early forms of intermodalism in the history of U.S. transportation can be found in the literature. In 1847, for example, the New York, New Haven and Hartford Railroad and the Fall River Steamship Line began experimenting with the use of containers to facilitate the transfer of freight between the two carriers [MULL95]. Wood and Johnson [WOOD96] include a photograph of an early mule-to-rail log transfer operation. Road-to-rail intermodalism occurred at warehouses and team track locations where freight was taken off horse-drawn wagons and, later, trucks and reloaded onto railcars. And of course there is the previously cited example of the Duluth-Superior ore docks which began operation during the 1890s. The point is that the history of intermodalism is deeply interwoven with the history of transportation. It has only been in recent years that the significance and potential of this concept have been recognized not only by the government but by many in the transportation and logistics industry as well.

If intermodalism has always been such an intrinsic part of transportation, what accounts for the rapid evolution and growth of intermodalism since the 1950s and the recent surge of interest in the concept? The main reason lies in the fact that many of the barriers to intermodalism have either been overcome or at least mitigated. Historically, there have been many impediments to intermodalism. Among them are the following:

Much has already been written about the negative effects of economic regulation of the transportation industry on the provision of innovative transportation services. In the case of truck-rail intermodalism, economic regulation of the railroads and motor carriers impeded the spread of TOFC/COFC services for many years. The first modern COFC services appeared in the early 1920s when the New York Central Railroad (NYC) began hauling loaded steel containers owned by The L.C.L. Company. Not to be outdone by its archrival, the Pennsylvania Railroad (PRR) soon offered its own competing container service. Both services were quite successful. They were so successful in fact that the motor carriers (and some other railroads) complained to the Interstate Commerce Commission (ICC). The ICC ultimately issued a ruling in 1931 which essentially raised COFC rates to non-competitive levels. As a result of this ruling and the effects of the Depression on traffic levels, the NYC and PRR soon discontinued COFC service [BLAS96].

Interestingly, while the NYC and PRR were providing COFC service, several other railroads began providing successful TOFC services that were not challenged by other carriers. The Chicago, North Shore & Milwaukee, an electric interurban line, began hauling loaded truck trailers on flatcars between Milwaukee and Chicago in 1926. The first steam railroad to offer piggyback service was the Chicago Great Western. Other early providers of TOFC service were the Denver & Rio Grande; the Chicago, Rock Island & Pacific; the Chicago, Burlington & Quincy; and the New York, New Haven & Hartford [BLAS96]. Why were these services tolerated while NYC's and PRR's COFC services were not? The reason is that, whereas in the case of COFC service, motor carriers were the railroads' competitor, in the case of TOFC service, motor carriers were the railroads' customer. In fact, the motor carriers marketed the service.

Despite the success of these early TOFC services, growth in truck-rail intermodalism was limited for nearly three decades. Better roads, intense competition between railroads and motor carriers for long-distance traffic, and restrictive piggyback and joint motor-rail tariffs combined to prevent the widespread adoption of TOFC services. By 1953, only six railroads provided TOFC service. In that year, however, truck-rail intermodalism got a tremendous boost. The ICC ruled that railroads were not required to obtain a truck-route certificate to carry freight in their own trailers over their own tracks. Following this ruling the number of railroads offering TOFC service jumped from six in 1953 to 32 in 1955 [BLAS96].

The economic deregulation of the transportation industry that occurred in the United States in the late 1970s and early 1980s has had a tremendous impact on the provision of transportation services in general and the practice of intermodalism in particular. Among the regulatory changes affecting intermodalism are the following [MULL95]:

In intermodalism, the name of the game is efficiency. The need to move cargo between different modes of transportation is a natural impediment to intermodalism. It is an activity that shippers and carriers alike would prefer to avoid. Intermodal transfers increase transit time as well as the chances for cargo damage. Anything that facilitates cargo handling at intermodal terminals is likely to enhance the attractiveness of intermodal transportation.

In the past there were basically two types of cargo: bulk and breakbulk. Of the two, bulk cargo was generally easier to handle. Because bulk cargo is flowable and less vulnerable to loss and damage, there are many fast and efficient ways of transferring it between modal equipment. Breakbulk cargo, on the other hand, comes in all shapes and sizes. Its value per pound is usually much higher than that of bulk cargo, and its chances of being damaged rise each time it has to be handled. Breakbulk cargo somehow has to be unitized - that is, arranged or assembled into unit loads - to facilitate loading and unloading. Breakbulk cargo was usually tied into bundles, placed in small wooden crates, sacked or bagged, or stacked on flat wooden platforms or pallets.

A seemingly obvious solution to the problem of handling breakbulk cargo is to pack it into large steel boxes or containers strong enough to protect the contents. This is an old idea as precious or valuable cargo was usually placed in strong boxes in the past, at least for protection if not for ease of handling. Nevertheless, use of large steel containers did not receive serious consideration until the 1920s, and it was not until 1956 that the idea of standard-sized containers began to take off. It was during that year that trucking executive Malcolm McLean clearly demonstrated the technical and economic feasibility of land-sea container transport when he launched the first containership, a converted tanker. Why it took so long for the concept of containerized freight to gain widespread acceptance is difficult to explain. The inertia of longstanding practices and ways of thinking had to be overcome. Nevertheless, since its arrival, the so-called container revolution has had such a far-reaching impact on intermodalism that it is not surprising that many people think of intermodalism only in terms of containers.

While the maritime shipping industry was taking tentative steps toward containerization, the railroad industry was busy exploring better ways of handling highway trailers. To improve the economy and efficiency of TOFC service, three things were needed: better methods of loading and unloading trailers, better devices for securing trailers to flatcars, and lighter railcars for piggyback service [BLAS96].

For decades the conventional method of getting highway trailers on and off flatcars was by circus loading in which trailers were backed onto and pulled off flatcars by way of a ramp, usually an earthen embankment, placed at the end of a stub-end track. Although this was a simple and inexpensive procedure that could be used by any railroad, it was slow and, therefore, suitable only for low-volume loadings and unloadings. In 1960 a company named Travelift & Engineering built the first overhead crane designed to lift trailers on and off flatcars. Subsequently, Southern Pacific (SP) and its suppliers developed the first "piggypacker", essentially a large forklift designed to load and unload trailers from the side of the flatcar. Since that time the versatility and sophistication of transfer equipment have steadily improved. Currently, many of the major TOFC/COFC terminals utilize large traveling overhead cranes that can complete a transfer within 90 seconds for a rate of 40 lifts per hour. These cranes can handle either trailers or containers.

To secure a trailer firmly to a flatcar, railroads at first used a complex assortment of chocks and chains. As a result, fastening and unfastening trailers at each end of a line haul move took considerable time and manual effort. One of the first solutions was the retractable screw hitch which folded onto the deck of the flatcar. As overhead and side loading replaced circus ramps, fixed, non-retractable hitches became more common.

The third problem - the flatcar itself - was addressed in many ways. The objective was to minimize the weight of the rail equipment to reduce operating costs. Rail intermodal equipment also had to be designed to handle ever increasing trailer sizes as well as several types of domestic and international containers. Until the early 1960s, the most common type of railcar used for TOFC service was the standard flatcar. In 1961, however, Trailer Train Company (now TTX Company) introduced an 89-foot flatcar designed to handle two 40-foot trailers. Over the next two decades, this car became the standard for piggyback service. Although TTX cars could be adapted to carry containers, it was clear that more suitable and economical equipment was needed. The major breakthrough came with the development of the articulated five-unit railcar in which two containers could be stacked on top of each other in each of five wells.

While huge containerships, giant gantry cranes, and double-stack trains are some of the more visible signs of recent advances in intermodal technology, the introduction of computer and telecommunications technology may in the long run have an even greater impact on intermodalism. Containerization and more sophisticated cargo-handling equipment may have addressed some of the physical or mechanical barriers to more efficient intermodalism, but computers and communications strike at some of the thornier organizational impediments. As stated by Muller [MULL95]:

"Computerization touches every aspect of intermodal movements: rating, routing, control of containers, clearance, billing, reporting and all other functions. The container revolution produced an improvement in physical aspects, but the computer revolution makes the entire concept simple and workable, regardless of whether or not freight is containerized."

Among the telecommunications and computer technologies being employed are electronic data interchange (EDI), geographic positioning systems (GPS), and automatic equipment identification (AEI) systems. The extent to which the goal of a seamless, integrated intermodal freight transportation system is ultimately achieved will greatly depend on how well these and other information and communications technologies are designed and implemented.

Third party companies that facilitate the shipment of freight by multiple modes of transportation are another interesting facet of the evolution of intermodalism. Such companies existed even before intermodalism became a buzzword. In the past, domestic surface freight forwarders gathered the less-than-carload (LCL) shipments of several small shippers, sorted them by destination, and offered the consolidated freight to railroads. For many years the Railway Express Agency (REA), a company sponsored by the railroad industry, picked up small packages by truck, loaded them into express cars attached to passenger trains for intercity movement, and delivered the packages by truck at the destination. In 1964 REA began using 8x8x20-foot vans which were transferred by means of a side-loading device between specially-designed 85-foot flatcars and highway trailer chassis. In more recent years, hundreds of shippers' agents known as intermodal marketing companies (IMCs) have appeared on the scene. These companies purchase TOFC/COFC capacity on a wholesale multiple-carload basis and market it to individual shippers. They may pick up and deliver a customer's trailers or containers using their own equipment or they may work with local drayage firms. Using EDI and other computer and communications technology, they greatly facilitate the process of billing, reporting, and tracking intermodal shipments. Other third party companies involved in various aspects of intermodal transportation include air freight forwarders, air cargo agents and consolidators, non-vessel operating common carriers (NVOCCs), shippers' councils and cooperatives, shippers associations, transportation brokers, perishable brokers, transloaders, distribution carriers, intermodal terminal operators, customs house brokers, export management companies, insurance carriers, and third-party logistics firms [MULL95].

The multimodal transportation service provider is an especially interesting player on the intermodal field. United Parcel Service (UPS) is a well known example. In addition to owning and operating its own fleet of trucks, UPS has a large fleet of cargo aircraft, and is also one of the largest users of rail TOFC service. Its traffic often travels on dedicated, fast, high-priority TOFC trains. Thus, a UPS shipment may travel by truck only, by truck-rail-truck, or by truck-air-truck.

Under the current intermodal setting, a carrier may sometimes substitute one mode of transportation for another when it becomes expedient or necessary to do so. Air cargo carriers, for example, have been known to ship some freight by truck only [WOOD96]. J. B. Hunt Transport, a large nationwide truckload carrier and major user of TOFC/COFC service, will off-load its trailers and switch to long-distance trucking if a train is greatly behind schedule [WALL96]. Consequently, the shipper oftentimes may not know which mode or modes are being used to carry its freight. Indeed, the shipper may not care as long as the freight arrives at its destination undamaged and on time.

REASONS FOR INTERMODALISM

Why would a shipper want to use more than one mode of transportation to move its products to its markets given the many impediments to intermodal transport mentioned above? There are a number of possible reasons, many of them residing in the fact that each mode of transportation has its own advantages and disadvantages. The various modes differ in terms of cost, speed, capacity, and flexibility. The basic premise of intermodalism is that, under a well-integrated transportation system, the various modes work together or within their own niches to provide the user with the best choices of service. In some cases and depending on a particular shipper's own unique needs, a combination of modes may serve the shipper better than a single mode. Although some of the reasons for intermodalism may seem rather obvious, it is useful to enumerate them. Research by Holcomb and Jennings [JENN94, HOLC95, JENN96] into instances of intermodalism for non-containerized cargo has uncovered some reasons that may not be quite so obvious.

Intermodalism can have benefits for both the shipper and the carrier. Therefore, the reasons for intermodalism will be considered from the standpoint of each party separately.

From the shipper's standpoint, multimodal transportation is required if no single mode of transportation connects an origin with a particular destination. An obvious example of this occurs when a company wishes to ship its products to overseas markets. The company has a choice of going by air or by water. In most cases, however, the company's production facility will not be located adjacent to an airport or a seaport, and if it is, in all likelihood that will not be the case at the destination. Some other means of transport is necessary to get the cargo to or from either an airport or a seaport at one or both ends of the journey. Thus, virtually all overseas transportation is multimodal and requires intermodal connections.

A more interesting case is one in which a shipper wishes to take advantage of certain characteristics of a mode such as low cost, high speed, or high capacity but is not directly served by any carrier for that mode. Nearly all air cargo transport is multimodal for this reason, since very few shippers needing the high speed delivery service provided by overnight air carriers are located at an airport. As another example, a manufacturer of plastic pellets may want to ship in large volumes by rail to take advantage of low freight rates, but the nearest railroad may be several miles away. It may be more economical to ship the plastic pellets by truck to a nearby truck-rail bulk transloading facility than to build a rail spur to the plant or to use long-distance trucking.

Sometimes a shipper may have to resort to intermodalism in order to continue using a preferred mode of line haul transportation. This situation can occur as a result of railroad branchline abandonment. The shipper may decide it is cheaper to dray its product to a nearby truck-rail transloading facility rather than switch to long-distance trucking or try to maintain the rail line itself.

Another common reason for choosing multiple modes of transportation arises when a shipper has direct access to its preferred mode of transportation, but one or more of its customers do not. Missouri Mining Co., for example, recently built a barge-to-rail coal transloading facility at Warrenton, OH, in order to serve a utility whose power plant was not located next to the Ohio River [JENN94]. This bulk transfer terminal enabled the shipper to use its preferred mode of transportation - barge - and still reach a customer it might not otherwise have been able to serve.

Freight transportation often involves three distinct stages: collection or consolidation of freight at the origin, long-distance line haul shipment over a major traffic lane, and distribution of the cargo to multiple customers at the destination. The middle stage - the long-distance line haul movement - often has much different transportation requirements than the outer two stages. This alone accounts for much of the need for multimodal transportation and intermodal connections. It is one of the reasons why public warehouses, grain elevators, cement terminals, distribution centers, freight forwarders, and intermodal marketing companies exist. Railroads and barge lines, with their high-capacity equipment and relatively low rates, are generally well-suited for shipping large volumes over long distances. Trucks are also suitable for long-distance shipping, but with much lower capacity and higher rates. At the local end, however, trucking may be essential because trucks can obviously reach more places than either railroads or water carriers and because shipment sizes to individual customers may be too small to justify direct shipment by either rail or water. For the shipper, the decision of whether to use trucking alone or trucking in combination with some higher-capacity, lower-cost mode depends on the "need or ability to move enough product to justify the use of the larger capacity mode and the added handling expense required to make the modal transfer [JENN96]."

Finally, shippers will sometimes resort to intermodalism to avoid being dominated by a single mode or carrier, thereby reducing their transportation costs and increasing their transportation options, and to take advantage of special opportunities. Jennings cites two examples [JENN94]. In one case, soda ash miners in southwestern Wyoming hauled some of their product by truck to a transloading facility on the Burlington Northern Railroad (now part of the Burlington Northern & Santa Fe Railway), even though they were directly served by the Union Pacific (UP). The other case involved a coal mine in Kentucky directly served by a major railroad. The mine shipped some of its coal by truck to a second railroad because the electric utility that was buying the coal had a volume guarantee with the latter railroad.

Despite intense competition between modes for certain kinds of freight, carriers also make use of intermodalism when it suits their purposes.

In the late 1920s when modern TOFC service had its beginnings, intercity transport by rail was still often faster and safer than transport by truck because of the somewhat primitive condition of the highway system in those days. Consequently, it is not surprising that some motor carriers took advantage of the new TOFC services to ship their loaded trailers to new markets. The motor carriers, in fact, marketed these services and were the railroads' customers. In 1990, however, when J. B. Hunt Transport, one of the leading nationwide truckload carriers, formed a partnership with the Atchison, Topeka and Santa Fe Railway (now part of the Burlington Northern & Santa Fe Railway) to provide TOFC service, the transportation industry was amazed and skeptical. Why would a giant, successful trucking company like J. B. Hunt want to do business with a railroad?

There are several reasons why J. B. Hunt and, subsequently, other motor carriers such as Schneider National have resorted to using the railroads for line haul piggyback service [WALL96]. One reason has to do with the high turnover rate of truck drivers that has beset the long-haul motor carrier industry in recent years along with the high cost of continually having to train new drivers. The low unemployment rate has also made it difficult to find enough people willing to spend days on the road away from home. With more of their trailers on railcars rather than on the highways, motor carriers have seen their vehicle accident claims and liability insurance expenses drop significantly. Fuel and tire costs as well as other truck and trailer maintenance costs have also decreased due to less wear and tear. TOFC service in some cases has also reduced the amount of empty backhauls and enabled J. B. Hunt and other motor carriers to serve new markets that they might otherwise have avoided. It is also less costly to reposition empty trailers using rail. From the standpoint of the railroads, this partnership with the long-distance truckload carriers has enabled them to reclaim at least some of the merchandise traffic that they had lost to the motor carriers over the years.

Regional trucking firms have begun making greater use of TOFC service because it enables them to serve new markets with fewer drivers and attain better equipment utilization. It also enables them to compete with the transcontinental less-than-truckload (LTL) motor carriers.

Economics has also been the driving force behind rail-ocean intermodalism, in the wake of the container revolution. The development of the double-stack train concept was a major factor in inducing the ocean shipping companies to switch from all-water, port-to-port routing to the use of railroads as a "land bridge". The economics of stacking two containers per railcar in dedicated unit trains proved to be overwhelming. In addition to lower transport costs, the ocean shipping companies obtained faster transit times with ship-rail routing and faster vessel turnaround, resulting in better vessel utilization.

Railroads and barge companies have discovered that, through intermodalism, they can gain new business in two ways. First, they can serve new customers that do not have direct access to rail or barge service. The Dardanelle & Russellville Railroad, for example, established a rail-truck transloading facility to allow it to haul pipe for an oil pipeline construction project. The pipe was transferred from railcars to trucks at the transloading facility for delivery to the construction site [JENN94]. Railroads may also create a truck-rail transloading facility to serve a new off-line customer until the latter generates enough traffic to justify the cost of constructing a rail spur to the plant. The second way railroads and barge companies can gain new business through intermodalism is by opening up potential new markets for their existing customers, enabling the latter to reach markets not directly served by rail or barge. On the other hand, the creation of a truck-rail transloading facility may enable a railroad to retain the business of existing customers currently located on a lightly used branchline or industrial spur which the railroad wishes to abandon.

IMPLICATIONS OF INTERMODALISM

The above discussion shows that, while intermodalism has always been around in one form or another, under the current deregulated setting, it has become an integral part of freight transportation in the United States. Carriers and shippers alike have recognized the benefits of integrated and coordinated transportation services and have embraced the concept of intermodalism, although sometimes somewhat slowly and grudgingly. A vast industry devoted to one or more aspects of intermodalism has developed, including intermodal equipment manufacturers and leasing companies and various intermodal service marketers, providers, and other facilitators. Indeed, the whole concept of logistics and physical distribution has changed drastically within the past two decades. What does all of this mean to government officials and policymakers and to transportation analysts?

For many years the federal government's philosophy toward regulation of the transportation industry either prohibited or discouraged any attempts by carriers from different modes to coordinate or integrate their services as well as any efforts by shippers to arrange for more coordinated or integrated multimodal transport. Legislation and regulations were aimed at individual modes, and little consideration was given to modal coordination or intermodal connections. Cracks in this thick regulatory wall began to appear in 1940 when Congress issued a National Transportation Policy statement that called for "fair and impartial regulation of all modes of transportation" and "sound economic conditions in transportation and among the several carriers [MULL95]." In 1967 the diverse and disparate transportation programs of the federal government were finally collected under one executive-level agency when the U.S. Department of Transportation (DOT) was created. Nevertheless, within the DOT, each mode of transportation still had its own administration. Further erosion of the strictly single mode way of thinking appeared in the 1975 National Transportation Policy Statement which "favored elimination of unreasonable barriers to intermodal cooperation [MULL95]." In 1979 the National Transportation Policy Study Commission issued a landmark report that advocated among other things multimodal systems planning, less economic regulation, and equal government treatment among modes. Finally, in 1991, promotion of both passenger and freight intermodalism became a part of federal law under the Intermodal Surface Transportation Efficiency Act (ISTEA).

The ISTEA was a landmark piece of legislation. Although it did not define the term "intermodalism", it required metropolitan planning organizations (MPOs) to consider ways of enhancing the efficient movement of freight and improving access to ports, airports, intermodal transportation facilities, and major freight distribution routes as they prepared their transportation improvement plans. It also required DOT to develop a database containing information on public and private investment in intermodal transportation facilities and services, the volume of goods and number of people carried in intermodal transportation, and the patterns of movement of goods and people carried in intermodal transportation. The ISTEA enabled federal funds to be used for intermodal projects, although it did not create a specific category of funds for that purpose. It also authorized $191.8 million for 20 priority intermodal freight projects. The U.S. General Accounting Office (GAO) determined that, as of September 30, 1995, ten states had obligated about $35.6 million in ISTEA funds for 23 intermodal freight projects. The GAO further determined that, of the total amount authorized for the 20 priority intermodal freight projects, states had obligated only $68.4 million as of the end of 1995 [GAO96].

It is hard to find any major transportation-related issue that does not affect either directly or indirectly more than one mode of transportation. Proposed increases in the permissible length and weight of truck-trailer combinations on the nation's major highways, for example, may adversely affect the railroads by diverting some rail traffic to motor carriers. International trade agreements such as the recent North American Free Trade Agreement (NAFTA) can affect the patterns of freight flow across the country, creating bottlenecks where perhaps none existed before, and affecting not only the highway system but ports and railroads as well. Railroad mergers may enhance the prospects for increased intermodalism. For example, if the proposed division of Conrail is approved, both Norfolk Southern (NS) and CSX Transportation (CSXT) will be able to institute single-line TOFC/COFC services between northeastern and southeastern metropolitan areas in an attempt to capture some of the truck traffic moving along Interstate 95. The status of some eastern ports, especially the Port of Baltimore, may also be enhanced by the merger because of greater rail competition and better rail facilities at the port.

The relationships among the various modes of freight transportation have already become so intricate that when one carrier or mode of transportation encounters difficulties, it can affect many other carriers and modes as well. The problems of the Union Pacific Railroad following its merger with Southern Pacific are a prime example. As this report was being written, UP was experiencing severe service problems, particularly in Texas and Southern California, but also throughout its entire 36,000-mile system. These problems included shortages of equipment and train crews, resulting in idle trains, congestion at yards and terminals, and delays throughout the system. The reasons for these problems are complex and will require detailed analysis, but the consequences were immediately widespread. Congested rail facilities in Houston and other places also hampered the train movements of other railroads as they tried to move around or through these areas. Some chemical manufacturers along the Gulf Coast, desperate to get their products to anxious customers, switched to truck transport despite its higher cost. The bulk motor carriers, however, did not have enough capacity to absorb all of the potential demand, causing them to take steps to avoid having their own service problems. Meanwhile, containers offloaded from ships at the Ports of Los Angeles and Long Beach sat on the ground waiting to be hauled away by rail. Some containership companies diverted their vessels to Oakland, CA, in an effort to avoid the congestion and delays at Los Angeles and Long Beach. Thus, the problems of one carrier affected the operations of carriers within three different modes as well as the operations of numerous other businesses.

One area in which government can potentially play a significant role in promoting intermodalism is that of access to intermodal facilities. Accessibility is a key element of efficient intermodal connections. It makes no difference how many cranes, how much berthing space, or how much on-dock storage area a marine container terminal has if its rail and highway connections are deficient. Poor accessibility will greatly hamper its ability to handle large volumes of cargo economically and efficiently. Moreover, the issue of access to an intermodal terminal can extend well beyond just the immediate vicinity of the facility. A railroad, for example, cannot provide double-stack train service to a port or TOFC/COFC facility if bridge and tunnel clearances along its route to the facility are too low. Improving access to intermodal facilities can involve widening and straightening streets and highways, eliminating highway-railroad grade crossings, relocating railroad tracks, raising bridge and tunnel clearances across multi-state areas, and dredging waterway channels. These are major capital projects requiring extensive planning and financial support from both the public and private sector.

Underscoring the importance of accessibility is the inclusion of highway connections to major intermodal facilities in the National Highway System (NHS). The ISTEA authorized the designation of the NHS "to provide an interconnected system of principal arterial routes which will serve major population centers, international border crossings, ports, airports, public transportation facilities, and other intermodal transportation facilities and other major travel destinations; meet defense requirements; and serve interstate and interregional travel" [FHWA96]. In addition to the Interstate System, non-Interstate highways essential to national defense, and high-priority corridors designated by Congress, the ISTEA also specified that the NHS should include highways that provide connections to major intermodal terminals.

When DOT submitted its recommendations for the NHS to Congress in December 1993, the proposed system consisted of 161,108 miles of primary rural and urban roads, including highway connections to 148 passenger and freight terminals. However, the initial efforts by states to identify major intermodal connectors for the NHS produced inconsistent results with insufficient coverage of intermodal terminals. DOT had not provided any guidelines to the states for identifying major intermodal facilities, and some states exerted more effort than others in identifying major intermodal terminals and their access roads. DOT requested more time to identify additional intermodal connections to be included in the NHS. Consequently, the National Highway System Designation Act of 1995, signed into law on November 28 of that year, contained a provision requiring DOT to submit within 180 days recommendations for additional intermodal connections to major ports, airports, international border crossings, public transit facilities, interstate bus terminals, and rail and other intermodal transportation facilities. FHWA had already started developing procedures and criteria for states, MPOs, terminal operators, and other interested agencies and groups to follow in identifying major intermodal terminals. The primary criteria specified volume thresholds or activity levels for each terminal type. Secondary criteria included more subjective factors for determining the importance of an intermodal terminal within a specific State. FHWA issued its final guidelines in April 1995 before the National Highway System Designation Act was passed. Using the FHWA guidelines, the states identified highway connections to an additional 1,251 major intermodal terminals consisting of 194 port facilities, 167 airports, 68 Amtrak stations, 198 rail-truck terminals, 96 intercity bus terminals, 66 pipeline terminals, 377 public transit terminals, 43 ferry terminals, and 42 multi-modal passenger terminals. These highway connections added 1,925 miles of rural and urban roads to the original NHS [FHWA96]. In May 1996, DOT submitted to Congress these recommended intermodal connections to be added to the original NHS. The proposed additional intermodal connections are eligible for improvement with NHS funds on an interim basis. They will not become a formal part of the NHS until Congress enacts a law for that purpose. Legislation codifying the DOT recommendations was expected to be included in the law reauthorizing the ISTEA.

Unfortunately for the transportation analyst, intermodalism makes policy analysis that much more difficult. As noted above, the relationships between modes and carriers have become so intricate that changes or problems in one mode or in one part of the country can have a profound impact on other modes and other geographic areas. Modal choice is no longer simply a matter of choosing between truck, rail, air, water, or pipeline; that is, it is not a matter of selecting between modes but a matter of selecting between services. Shippers are becoming less concerned about the means of transportation used and more concerned about reliability, speed, shipment tracking, and total logistics costs. Traditional analytical tools and databases may not be adequate to address the current multimodal and logistics reality. Satisfactory multimodal analysis will require new tools and better databases.

ROLE OF THE INTERMODAL TERMINALS DATABASE

Due to the growing demand for geographic transportation data and the emphasis placed upon multimodal network analysis by ISTEA, modal network databases are becoming more popular with users both within and outside government. As private companies and federal and state agencies begin to take a more global view of economic development opportunities, the need for a high quality, well-maintained, and readily obtainable set of national modal transportation network databases will continue to grow. BTS has responded and continues to respond to this need by developing, maintaining, and improving the NTAD to support research, analysis, and decision making across all modes of passenger and freight transportation.

Initial work on the NTAD was directed toward the development of separate network models or databases for each of the main modes of freight transportation: highway, rail, waterway, air, and products pipeline. Each model describes the topology of a modal network in terms of links and nodes. The location of each node is denoted by its longitude and latitude. Important attributes of each link, which connects a pair of nodes, are also included in each network database. Versions of these network models - suitable for use within a GIS - have been made available to the public on CD-ROM.

These individual modal networks must somehow be connected before any multimodal analysis can be performed. In the real world, connections between different modes of transportation occur at intermodal terminals or facilities. In the absence of any data on the locations of intermodal terminals, it is possible to combine two or more of the single-mode networks in the NTAD by applying an algorithmic procedure. Any such procedure, however, will necessarily rely on some simplifying assumptions that may limit the quality of subsequent multimodal analyses. Therefore, including a database of intermodal terminal locations and attributes in the NTAD and using it to construct a realistic multimodal network should greatly improve the quality of multimodal analyses based on the modal network databases in the NTAD.

An important function of the national intermodal terminals database is to support any analysis dealing with the flow or movement of freight through a multimodal transportation network. This might involve determining and plotting a likely multimodal path taken by a freight shipment between an origin and a destination, or it might involve assigning interregional freight flows to a multimodal network. The results of such a multimodal traffic assignment would enable the analyst to:

Several actual applications of the intermodal terminals database demonstrate its utility for multimodal network analysis. An early version of the intermodal terminals database was used in a 1996 project to develop and test an analytic framework for considering the national impact of site specific freight bottlenecks. As a preliminary test of the proposed analytic framework, the project generated ocean-to-rail routes from the Ports of Los Angeles and Long Beach to 77 double-stack train TOFC/COFC terminals throughout the United States [SOUTH96]. In 1997, the terminals database was being used at Oak Ridge National Laboratory (ORNL) to determine routes and distances for multimodal shipments reported by shippers in the 1997 Commodity Flow Survey (CFS). In a future application, the terminals database will be used to develop a method for estimating region-specific truck drayage distances associated with TOFC/COFC movements.

Another important function of the intermodal terminals database is to support a variety of mapping and GIS applications. Maps depicting the locations of particular types of intermodal facilities can be a valuable aid for visualizing the connectivity of the national multimodal transportation network and the relative density of intermodal terminals in different parts of the country. GIS can be used to answer a wide variety of questions such as:

More than anything else, the intermodal terminals database described in the remainder of this report is fundamentally an inventory of existing facilities. Intermodal terminals are an important component of the nation's transportation infrastructure just like highways and railroad lines. Except for certain types of terminals such as airports and port facilities, no attempt has been made until now to try to inventory all of the places where freight can be exchanged between different modes of transportation. At a minimum, the national intermodal terminals database, when completed, will document what is on the ground. It is hard to imagine how policymakers and analysts can begin to understand the impact of policy issues and initiatives on intermodal transportation without having some sense of the size and scope of the existing intermodal infrastructure.

CHAPTER 2. CHARACTERISTICS AND TYPES OF INTERMODAL TERMINALS

The previous chapter stated that intermodalism occurs at places where two or more modes of transportation meet to interchange freight - either directly or through intermediate storage. These places represent intermodal freight terminals or facilities. Just as some confusion exists about the meaning of intermodalism, the notion of an intermodal terminal also conjures up different images. It is impossible, however, to design and build a consistent and coherent database of intermodal freight terminals without having a clear idea of what constitutes an intermodal terminal. How does one recognize an intermodal terminal or facility? What distinguishes one intermodal terminal from another? What distinct types of intermodal terminals can be identified? This chapter addresses these questions.

DEFINITION OF INTERMODAL TERMINALS

Before we can define an intermodal terminal, we need to specify what we mean by a mode of transportation. This is not as straightfoward as it may seem at first. For example, Section 171.8 of Title 49 of the Code of Federal Regulations (CFR) defines mode of transportation as any of the following transportation methods: rail, highway, air, or water. Similarly, Wood and Johnson [WOOD96] devote separate chapters to highway carriers, railroads, pipelines, domestic water carriers, and domestic aviation, implying that highway, rail, pipeline, water, and air constitute the principal modes of transportation. Note that each of these principal modes represents a distinct medium on which or through which freight can move. On the other hand, the survey form used to gather individual shipment data for the 1997 Commodity Flow Survey (CFS) identifies eight modes of transport: parcel delivery, courier, or U.S. Postal Service; private truck; for-hire truck; railroad; shallow draft vessel; deep draft vessel; pipeline; and air. These categories are a mixture of different media, types of vehicle using a particular medium, a type of service that may utilize different media to handle a shipment, and private versus for-hire carriage. In the broadest sense, "mode" refers to any specific means of transportation of interest or concern to a particular analysis or planning application. In a study of truck size and weight issues, for example, transportation analysts might consider different truck-trailer combinations as separate modes or means of transportation. The problem with this broad definition of mode of transportation is that it produces overlapping or non-orthogonal modal categories such as those used for the 1997 CFS.

For purposes of identifying intermodal terminals and designing an intermodal terminals database, a mode of transportation is defined by the medium on which or through which cargo moves. Thus, the principal modes of transportation are highway or road, rail, water, air, and pipeline. It is recognized, however, that within some of these principal modes, different sub-modes can be defined. Sub-modes can be specified on the basis of types of vehicle such as shallow draft and deep draft vessels or standard gauge and narrow gauge railcars, methods of operation such as fixed-route and flexible-route transit, types of carrier such as private versus for-hire trucking, and types of cargo such as pipelines for crude oil and pipelines for petroleum products. Although the initial version of the intermodal terminals database does not recognize any sub-modes, a case can be made for including a few of them, particularly within the water and pipeline modes, in future versions of the database.

As an aid to understanding what an intermodal terminal is, it is helpful to consider what other types of freight terminals exist. A freight terminal in general is a station, facility, or integrated set of facilities which has as one of its primary functions the loading or unloading of freight onto or off a particular transportation mode's network. It is a place where the movement of freight either begins or ends on a particular mode of transportation. It makes no difference whether the commodities being transported were produced or will be used at the terminal or whether the commodities were transferred from or will be transferred to another mode of transportation.

Freight terminals can be categorized into three types: freight traffic generators and attractors, intramodal terminals, and intermodal terminals.

Freight traffic generators and attractors are places where cargo is produced for shipment or where cargo is received for subsequent use. Because virtually any residence or place of business could be regarded as a freight traffic generator or attractor, transportation planners and analysts are generally more concerned with major generators and attractors. These may be defined as places that ship or receive such large volumes of freight that they include extensive facilities for loading, unloading, and storing cargo. These places may even have their own in-plant railroad, pipeline, or roadway networks connecting them to the intercity modal networks. Examples include coal, iron ore, potash, soda ash, and other kinds of mines; vehicle manufacturing and assembly plants; steel mills; paper mills; refineries; lumber mills; chemical manufacturing plants; power generating stations; soybean processing plants; flour mills; and so on. Clusters of smaller businesses sharing the same local transportation infrastructure such as industrial parks, regional shopping centers, and central business districts might also be regarded as major freight traffic generators and attractors. Although these terminals are clearly important for freight transportation planning purposes and probably deserve their own database, at least on a metropolitan or regional basis, they technically are not intermodal terminals because their function is not to transfer cargo from one mode of transportation to another.

The second type of freight terminal - intramodal terminals - consists of facilities where cargo changes carriers or vehicles within the same mode of transportation or between sub-modes within one of the principal modes of transportation. Many public warehouses and distribution centers are served only by trucks and, therefore, are intramodal in nature. Businesses that operate their own truck fleets and local drayage companies often handle the local pickup and delivery at these facilities, while for-hire trucking companies provide long-haul shipping to and from these locations. Some regional motor carriers have truck terminals where freight can be transferred to other regional trucking companies for interregional or transcontinental hauls. Less-than-truckload (LTL) shipments usually involve a pair of truck terminals and three separate truck moves: local pickup to a truck terminal for sorting and consolidation by destination, long-distance transport to another truck terminal, and local delivery from the second terminal to the consignee. Intramodal terminals, however, are not strictly the domain of motor carriers. Air cargo may get transferred between air carriers at an air cargo terminal. On the lower Mississippi River at and below Baton Rouge, LA, mid-stream facilities exist to transfer coal, grain, and other bulk commodities from shallow draft barges to ocean vessels. On the Columbia-Snake River System, containerized freight is loaded into barges at Lewiston, ID, Boardman, OR, and other locations for transport to Portland, OR, where it is reloaded onto containerships for export. On the West Coast, Matson Navigation Company provides a coastwise container shuttle service, picking up loaded containers dropped off by ocean-going containerships at the Ports of Los Angeles and Long Beach, taking them to Seattle and Vancouver, BC, and bringing back containers from these places to Southern California for loading back onto trans-Pacific vessels [MONG97]. Like the major freight traffic generators and attractors, intramodal terminals, especially the truck terminals, may warrant their own geographic database. Indeed, some of these intramodal terminals might be classified as intermodal, depending on what set of modes of transportation one adopts. In particular, barge-to-vessel transloading facilities might be considered intermodal rather than intramodal.

Intermodal freight terminals have equipment and facilities designed to transfer freight between two or more principal modes of transportation, either directly or through intermediate storage. Air cargo terminals, for example, provide the interface between the air and highway modes of transportation. Various kinds of port facilities accommodate the transfer of freight between water and land modes of transport, including rail, highway, and pipeline. Port terminals tend to be equipped to handle certain kinds of commodities such as containers; motor vehicles and other roll-on/roll-off (RO/RO) cargo; breakbulk cargo; dry bulk commodities such as coal, grain, ore, and sand and gravel; and liquid bulk commodities such chemicals, petroleum products, and vegetable oils. A variety of intermodal facilities exist for exchanging freight between the highway and rail modes. These include TOFC/COFC terminals; motor vehicle loading and unloading ramps; liquid and dry bulk transloading facilities; grain storage and transfer facilities; lumber, steel, and paper reload centers; and cross-dock or direct transfer facilities at certain public warehouses and distribution centers. Like intermodal port facilities, each truck-rail terminal tends to be geared towards a specific type or category of freight.

There is clearly some overlap between the three kinds of freight terminals. Major freight traffic generators and attractors, for example, sometimes have intermodal transfer capabilities. Flour mills and corn and soybean processing plants are examples. These establishments receive wheat, corn, and soybeans for conversion into flour, corn syrup or sweeteners, and soybean meal and oil. The incoming grains are generally stored in an elevator, storage bins, or cluster of silos, which are used to feed the processing plant. However, in some cases, the grain is also shipped out by some other mode of transportation. We might classify the grain elevator in these cases as an intermodal facility and the mill or plant as a major freight traffic attractor. A few oil refineries also have intermodal transfer facilities. In addition to receiving crude oil for conversion to fuel oil, gasoline, and various other refined products, a refinery may also store petroleum products received by pipeline, ocean tanker, or rail tank car, and ship it out by pipeline, inland barge, or tank truck along with the refinery's own output. Note that this last example shows how a terminal can have both intramodal and intermodal capabilities. Liquid bulk terminals, grain elevators, warehouses, and other terminals where cargo is accumulated, consolidated, mixed, or stored often have this characteristic. Liquid bulk tank farms, for example, may receive part of their supply by ocean tanker and ship by inland barge as well as other modes of transportation. A warehouse may receive freight by truck from local sources and send the goods out by long-distance motor carrier and by rail. Conversely, a major distribution center may receive cargo by long-distance motor carrier and by rail and distribute it locally by truck.

To be considered a candidate for inclusion in the intermodal freight terminals database, a freight terminal must be capable of receiving a commodity by one principal mode of transportation and shipping the same commodity out by a different principal mode. Intermodal transfer does not necessarily have to be the primary function of the terminal, but it has to be one of the functions. Freight traffic generators and attractors that have no intermodal transfer capability and freight terminals served by a single mode of transportation were therefore excluded from consideration.

DISTINGUISHING CHARACTERISTICS

Intermodal freight terminals include a wide variety of facilities. They differ considerably in size, complexity, and functionality. Some terminals are complex facilities covering many acres of land with multiple buildings, storage areas, gates, and internal channels of freight flow. Other terminals are quite simple, sometimes even ad hoc in nature, with very little infrastructure or sophisticated equipment. As an aid to understanding what an intermodal freight terminal is and to recognizing the many kinds of intermodal terminals in operation, it is useful to consider the characteristics or features that most distinguish one type of terminal from another. Of the many characteristics that might be considered, the following five are especially suitable for classifying intermodal facilities: the pairs of modes that the facility connects, the types of cargo handled, the types of transfers that can take place, private or public ownership, and availability for public use.

Mode Pairs

Freight terminals are often classified by mode. Thus, we see various references to truck, rail, waterway, air, and pipeline terminals. This can sometimes lead to problems in multimodal analysis because it ignores the fact that, by definition, intermodal terminals involve more than one mode of transportation. For example, is a facility that sends and receives freight by truck, rail, and barge a truck terminal, a rail terminal, or a waterway terminal? It may be all three, but that does not tell us which pairs of modes are involved in cargo transfers. Does this terminal, for example, exchange freight between all possible pairs of these three modes or is it designed to transfer freight between truck and barge and rail and barge but not between truck and rail? This is important to know, because if the facility does not provide for transfers between truck and rail, then we would not want to connect the highway and railroad networks at this terminal when constructing a network for multimodal analysis. Thus, it is better to classify intermodal terminals by the pairs of modes which they connect rather than by the individual modes which serve them.

Type of Cargo

One of the most differentiating properties of an intermodal freight terminal is the kind of cargo it is designed and equipped to handle. This attribute, together with the pairs of modes involved, largely defines the purpose or function of the facility and influences the kinds of transfer equipment and storage space needed as well as the layout or configuration of the terminal. Although many intermodal terminals can and do accommodate a variety of commodities, most are designed to handle a specific type of cargo, and many only handle a specific kind of commodity.

The literature suggests several ways of classifying freight (see, for example, [MULL95], p. 3, and [STOPH94], pp. 337-338). For purposes of describing intermodal terminals, the freight typology depicted hierarchically in Figure 2-1 is helpful. At the highest level, all cargo can be classified as either containerized or non-containerized. The latter can be further subdivided into breakbulk and bulk cargo. Bulk cargo can be further categorized as either dry or liquid. The following defines each of the four basic cargo types in more detail:

These four basic types of cargo generally require different kinds of mechanical handling equipment and storage facilities for maximum efficiency. For example, although cranes designed to lift breakbulk cargo can be fitted with grab buckets to handle dry bulk materials, some type of conveyor system is usually required to transfer large volumes of dry bulk cargo between modes either directly or through intermediate storage. Intermodal terminals therefore tend to specialize on one or two of the four basic types of cargo. Some marine terminals will handle both containerized and breakbulk cargo, and others will handle a combination of breakbulk cargo such as steel and one or more kinds of dry bulk commodities. Truck-rail bulk transfer facilities often exchange both dry and liquid bulk commodities. Rarely, however, will a terminal handle more than two basic types of cargo.

The degree of cargo specialization at an intermodal facility usually goes well beyond the four basic cargo types. Many intermodal terminals for non-containerized freight are created to handle one or a few particular kinds of commodities within one of the basic cargo groups. Figure 2-1 gives some examples. Under the breakbulk category can be found terminals dedicated to the intermodal transfer of automobiles and other finished vehicles, lumber, steel, or paper products. Specialized dry bulk facilities include grain elevators, rail-to-barge and rail-to-vessel coal transloading terminals, ore docks on the Great Lakes, cement terminals, and wood chip facilities.

Examples of liquid bulk terminals are tank farms specializing in petroleum products, chemicals, molasses or vegetable oils, or some combination of these commodities.

Type of Transfer

Freight often does not get transferred directly from one mode of transportation to another at an intermodal terminal. Even containers are sometimes stacked on the ground before being hauled away. Certain kinds of intermodal terminals, in fact, operate primarily as intermediate storage facilities. Examples include grain elevators, warehouses and distribution centers, tank farms, and cement terminals. Some intermodal terminals provide equipment for direct transfer as well as facilities for storage.

Stopher et al. [STOPH94] identify three types of transfer at intermodal facilities. The first type is the direct transfer. Examples include containers lifted off a ship and placed directly on double-stack railcars, automobiles unloaded from multi-level railcars and loaded directly into auto carriers, trucks dumping sand or gravel directly into a barge, and rail hopper cars dumping coal into a conveyor system that carries it directly to a shiploading spout. The second type of transfer is the short-term storage transfer. In this case, cargo arrives on the incoming mode, is unloaded and stored for a relatively short period of time on a platform or loading dock or in a transit shed, and is subsequently loaded into and hauled away by the outgoing mode. This situation often occurs when the incoming cargo arrives at the terminal in advance of equipment for the outgoing mode. The third type of intermodal transfer involves long-term storage. This type includes two cases. In the consolidation case, the incoming cargo arrives by a mode whose cargo-carrying capacity is lower than that of the outgoing mode. For example, a large number of trucks bring paper to a warehouse where it is stored until enough product is available for loading into rail boxcars. In another example, several unit trains unload coal at an export terminal until enough is on-hand to load an ocean-going vessel. The opposite occurs in the distribution case in which the arriving mode is of relatively higher capacity than the departing mode or modes. The cargo is unloaded into a warehouse, storage tank, silo, or other storage facility and gradually distributed within the locality or larger region.

The types of transfer allowed at an intermodal terminal are somewhat related to the types of cargo handled. In general, container terminals and breakbulk facilities such as warehouses and distribution centers for general merchandise either provide for direct transfer or else strive for quick turnaround or fast throughput by means of short-term storage transfers. Truck-rail bulk transfer facilities are established specifically for direct transfers between modal equipment, although it is interesting to note that the railcars are sometimes used as a storage medium in this case. Grain-handling facilities, cement terminals, and liquid bulk tank farms, on the other hand, are usually designed for long-term storage transfers, although some of them do operate equipment for direct transfers.

Ownership and Availability

Intermodal freight terminals may be either privately owned or publicly owned. Private owners include various carriers such as motor carriers, railroads, barge companies, steamship lines, and pipeline operators; various third-party entities such as shippers' associations, farmers' cooperatives, intermodal marketing companies, warehousing companies, and independent terminal operators; and various businesses including companies engaged in coal mining, sand and gravel quarrying, petroleum refining, grain handling and processing, and steel, paper, or cement manufacturing. Public owners include municipalities; local, regional, or statewide port authorities; and federal agencies. Privately owned terminals cover virtually the entire spectrum of intermodal freight facilities, while publicly owned terminals tend to involve airports, seaports, and inland waterway facilities.

Regardless of whether a terminal is privately owned or publicly owned, it may be available for the general public to use or it may be reserved for the exclusive use of either its owner or a specific customer. Facilities owned by a carrier or a third-party entity are usually available to any shipper, but many exceptions can be found. Some TOFC/COFC facilities, for example, have been established to serve a specific customer. The serving railroad rarely advertises the existence of these facilities. Likewise, some truck-rail bulk transloading facilities serve only a single motor carrier and prohibit other motor carriers from using the facility. Intermodal terminals owned by mining, manufacturing, and refining companies are generally operated for the owning company's own use. Cement manufacturers, for example, have built their own terminals to receive large quantities of bulk cement by barge, vessel, or rail and distribute it locally or regionally by rail or truck. Once again, however, exceptions do exist. Many petroleum companies, for example, make their storage tanks available for public use to fill up excess capacity. Most publicly owned terminals are provided for public use. Often, however, the terminal may be leased to a single carrier for its exclusive use. Many publicly owned marine container terminals, for example, are leased to a particular containership company or its terminal operating subsidiary. Of course, any business can ship its freight through such a terminal by simply choosing the carrier having exclusive use of the facility. Thus, restrictions on the use of an intermodal terminal may apply to carriers, shippers, or both.

SOME COMMON TYPES OF INTERMODAL TERMINALS

Intermodal freight terminals can be grouped according to the five dimensions or properties discussed above. Because many terminals are highly specialized, particularly in the types of cargo or kinds of commodities which they handle, the number of possible groupings is quite large. Fortunately, several common or well-recognized kinds of intermodal terminals can be identified which exemplify the diversity of intermodal facilities. In fact, most sources of information on intermodal terminals deal with one of these conventional types.

Several common types of intermodal terminals are briefly described in the following subsections. Each description indicates the pairs of modes with which the terminal type is usually associated, the types of cargo or specific kinds of commodities involved, the types of transfer typically allowed, forms of ownership, and public availability. The purpose of these descriptions is to convey a sense of the diversity of intermodal terminals and to illustrate the many ways in which intermodalism can occur.

TOFC/COFC Facilities

Trailer-on-flatcar/container-on-flatcar (TOFC/COFC) terminals are places where either containers or highway trailers or both are transferred between motor carriers and railroads. The containers and trailers may be directly transferred between modal equipment or they may be set on the ground for a short period before being loaded onto and hauled away by the departing mode. Perhaps no other type of terminal outside of a port facility is more closely associated with the notion of intermodalism.

From its beginnings in the mid-1920s until the early 1980s, TOFC/COFC service primarily involved trailers riding "piggyback" on rail flatcars. Because the easiest and cheapest way of getting a trailer on and off a flatcar was by way of a ramp in the same manner as circus wagons used to be loaded and unloaded, early TOFC/COFC facilities were just "circus" ramps, usually earthen, placed or built at the end of a track. Multiple trailers were loaded in series on a string of cars through the use of fold down plates on the ends of the flatcars. This was a slow and laborious process, suitable only for low volumes of trailers. With the development of overhead cranes capable of lifting a trailer onto a flatcar and the side-loading "piggypacker" as well as the increasing use of containers instead of trailers, railroads began closing the circus ramps and concentrating their TOFC/COFC operations at a smaller number of larger intermodal hubs. The actual number of circus ramps in use during the 1970s is in dispute. One source [BLAS96] indicates that there were as many as 2100 TOFC facilities, mostly circus-loading ramps, while another source [ARMS93] cites a figure of 1500. Regardless, by the end of the 1980s, most of the one-track circus ramps were closed, following the caboose, rural freight station, and four- and five-man train crew into oblivion. The 1997 Rail Intermodal Terminal Directory produced by the Intermodal Association of North America (IANA) lists four TOFC/COFC terminals in the United States that had only circus-loading capabilities as of December 1996 [IANA97].

The number of TOFC/COFC facilities in the United States has dwindled from 1500-2100 in the 1970s to around 235 in December 1996. The smallest covered only one or two acres and had less than 25 parking spots, while the largest spanned over 250 acres and provided over 1000 parking spots. About half of the TOFC/COFC terminals covered less than 25 to 30 acres and had less than 400 parking spots. The average size of a TOFC/COFC facility was about 40 to 45 acres with approximately 700 parking spots. All but four facilities operated some type of lift equipment, including 95 terminals with at least one overhead crane. At least 90 TOFC/COFC terminals had an annual lift capacity of at least 100,000, and 28 had the capability of handling at least 250,000 trailers and containers annually. The four largest TOFC/COFC facilities in the United States had annual lift capacities of over half a million trailers and containers:

Except for the few remaining circus ramps which can only accommodate trailers, most TOFC/COFC facilities can handle either trailers or containers. A few, however, are limited to containers only. Examples include:

Two relatively recent innovations in truck-rail intermodal equipment have led to the appearance of two new types of TOFC/COFC terminal. One is RoadRailer technology, and the other is the Iron Highway.

A RoadRailer is a specially designed semi-trailer that can ride on either standard tires or on flanged railroad wheels. The idea for such a vehicle can be traced back to the mid-1950s, but it did not receive much attention until 1977 when Robert Reebie acquired the rights to RoadRailer technology and began redeveloping it. In its earlier versions, the RoadRailer carried both types of wheel. The heavy steel railroad wheels, of course, added to the tare weight of the truck-trailer combination and reduced the amount of payload that could be hauled over the highways. The RoadRailer was therefore redesigned so that it could be attached to a free standing two axle rail truck or bogie. The major advantage of RoadRailer technology is that it completely eliminates the need for any kind of railcar to haul the highway trailer, thereby greatly reducing the tare weight of a train. Moreover, RoadRailer terminals do not need cranes, lifts, or other expensive loading and unloading equipment. RoadRailers, on the other hand, are so light they cannot be operated in combination with regular rail equipment and, therefore, must move in special RoadRailer-only trains. Several railroads experimented with RoadRailer service, but all except Norfolk Southern (NS) eventually discontinued it. NS began to establish a RoadRailer network in 1986 and created a subsidiary, Triple Crown Services (TCS), to market and operate the trains. The service was highly successful and by December 1996 the TCS RoadRailer network included 12 terminals in the United States and one in Ontario, Canada. In August 1997, TCS extended its reach into Fort Worth, TX, by establishing a terminal at Saginaw Yard and entering into an agreement with BNSF to pull RoadRailer trains between there and a connection with NS at Kansas City, MO [MULL95, KEEFE89, JOC97].

The Iron Highway concept is a much more recent invention. Developed and extensively tested by CSX Intermodal, the Iron Highway is a 1200-foot continuous deck that can hold between 20 and 40 highway trailers. It can split in the middle, providing two ramps for loading and unloading. A train up to 6000-feet long, capable of hauling up to 200 trailers, can be assembled by attaching five of these platforms together. Like RoadRailer terminals, Iron Highway terminals do not require any lift equipment and can be established at virtually any track location. The concept was developed as an attempt to provide medium-haul truck-rail intermodal service in traffic lanes 300 to 700 miles long [STEPH95]. CSX Intermodal created an Iron Highway terminal on a four-acre site at Middlebelt Yard in Livonia, MI, near Detroit and another terminal on a nine-acre site in East Chicago, IN. After many delays due to technical problems, it began test-marketing Iron Highway service between Detroit and Chicago in August 1996. The service was short-lived, however. Canadian Pacific, on the other hand, continues to provide Iron Highway service between Montreal and Toronto. If the Iron Highway concept were to gain widespread acceptance, it would greatly increase the number of places where truck-rail intermodalism could swiftly and easily take place.

When a railroad operates more than one TOFC/COFC terminal in a large metropolitan area, each terminal may be dedicated to certain kinds of traffic or to specific traffic lanes. One facility, for example, may be devoted to landbridge services for containership companies, another may be dedicated to domestic container services, while a third may focus primarily on LTL traffic. Freight that is highly time-sensitive may be directed to a terminal with high-speed overhead cranes and sufficient track space for building whole trains, while less time-sensitive traffic may be sent to a smaller facility. Certain terminals may be dedicated wholly or in large part to serving major intermodal customers such as UPS, the U.S. Postal Service, J. B. Hunt, or APL. Canadian Pacific Railway's Schiller Park East terminal in the Chicago area is restricted to containerized freight moving to and from eastern Canada, while its Schiller Park West terminal handles containerized traffic moving to and from western Canada and the U.S. west coast. The primary customer for BNSF's TOFC/COFC terminal in the Chicago suburb of Willow Springs, IL, is UPS, which has a new regional sorting facility next door. CSX Intermodal operates two terminals in Memphis, one of which is for the exclusive use of Mitsui O.S.K. Lines.

The nature of TOFC/COFC freight greatly influences the location of TOFC/COFC facilities. TOFC/COFC freight is usually more time-sensitive than other kinds of rail freight. It often moves on dedicated "intermodal" trains rather than on manifest freight trains. Intermodal trains usually have the highest priority and are generally among the fastest trains on a railroad's system. Consequently, TOFC/COFC facilities tend to be located adjacent to mainlines rather than on secondary branchlines or industrial spurs. In fact, TOFC/COFC loading tracks are usually directly connected to the mainline in order to expedite the movement of intermodal trains [ARMS93]. TOFC/COFC facilities also tend to be located at major rail yards or very close to their main source of traffic. In some instances, former classification yards have been converted or expanded for either exclusive or primary use of TOFC/COFC service. Examples include BNSF's Cicero Yard near Chicago, Norfolk Southern's Landers Yard in Chicago, and Conrail's Rose Lake Yard in East St. Louis, IL. The location of BNSF's terminal at Willow Springs, IL, was dictated by UPS's plans to build a major sorting facility at that location. Because railroads are heavily involved in the movement of containerized imports and exports, it is not surprising that another prime location of TOFC/COFC terminals is next to a port container facility or in the vicinity of a port complex. Almost every major seaport in the United States either has an intermodal container transfer facility on the premises or an intermodal rail yard nearby.

Each of the Class I railroads operating within the United States owns and operates a number of TOFC/COFC terminals, except for CSX Transportation (CSXT). The facilities which CSXT serves actually belong to CSX Intermodal, a unit of CSX Corporation that is separate from the railroad itself. Several regional carriers also have their own TOFC/COFC facilities. They include Alaska Railroad; Angelina & Neches River; Arizona & California; Chicago, Central & Pacific (now a subsidiary of Illinois Central); Florida East Coast; Gateway Western (now owned by Kansas City Southern); Iowa Interstate; New York & Atlantic (which provides freight service over Long Island Railroad trackage); Massachusetts Central; Missouri & Northern Arkansas; New York, Susquehanna & Western; Providence & Worcester; St. Lawrence & Atlantic; Toledo, Peoria & Western; Vermont Railway; Wheeling & Lake Erie; and Wisconsin Central [IANA97]. In a few cases, one or more railroads utilize a terminal owned by another railroad. Southern Pacific and Wisconsin Central, for example, use Illinois Central's Moyers Intermodal Terminal at Harvey, IL, south of Chicago, and Conrail trains carrying UPS freight visit BNSF's facility at Willow Springs, IL. Kansas City Southern and I&M Rail Link jointly own a TOFC/COFC terminal in Kansas City, MO. Railroads also access another railroad's facilities by entering into through-train or haulage agreements. BNSF, for example, reaches the Hoosier Lift terminal near Remington, IN, by way of the Toledo, Peoria & Western; Louisville, KY, through a service agreement with Norfolk Southern; and Moterm in Ferndale, MI, near Detroit in conjunction with Canadian National. Norfolk Southern works with Canadian Pacific and New York, Susquehanna & Western to reach TOFC/COFC facilities in Taylor, PA, and Albany, NY, and with Florida East Coast Railway to serve terminals in Fort Lauderdale, Miami, and West Palm Beach, FL.

Railroads, however, are not the only owners and operators of TOFC/COFC terminals. They also serve facilities established by third-party companies and public entities. APL Stacktrain Services, for example, operates a facility in Woodhaven, MI, near Detroit and another in South Kearny, NJ. Another stacktrain operator, Rail-Bridge Corporation, a subsidiary of "K" Line (America), Inc., has a container facility in Elizabeth, NJ, known as the E-Rail Terminal and served by Conrail. As noted above, CSX Intermodal, an intermodal marketing company, has its own set of terminals and utilizes the facilities of other railroads at a number of locations outside of CSX Transportation's territory. Several TOFC/COFC terminals belong to port authorities. Conrail, Canadian Pacific, and CSXT serve the Ameriport Intermodal Transfer Yard, a Port of Philadelphia & Camden facility located next to the Packer Avenue Marine Terminal in Philadelphia. The Port of Tacoma operates two near-dock intermodal rail yards, both served by BNSF. The North Intermodal Rail Yard serves Terminal 7, the Husky Terminal, and the Maersk and Evergreen Terminals, while the South Intermodal Rail Yard is adjacent to the Sea-Land Terminal.

Auto Terminals

Auto terminals - also known as auto ramps or auto transloading facilities - are places where finished vehicles such as automobiles, light trucks, jeeps, and vans are transferred between different modes of transportation. Three types of auto terminal can be identified:

These are not mutually exclusive categories, however. Because of the complex logistics of motor vehicle manufacturing and distribution, assembly plant terminals may also function as auto distribution terminals and marine auto terminals may also serve as an auto distribution center for domestic vehicles.

Auto terminals are designed for either direct transfers or short-term storage transfers. At auto distribution centers, vehicles may be driven off a multilevel autorack and directly onto a waiting auto carrier, or if necessary they may be parked in a marshalling yard. Conrail employs an electronic inventory tracking system called Terminal Operations and Information System (TOIS) and a transloading procedure called Load Laning to improve the speed and efficiency of its auto terminal operations. As Conrail describes the process in its marketing literature [CRC92]:

"TOIS provides haulaway carriers with a list of inbound vehicles on the day's train scheduled to be unloaded in addition to those vehicles already parked in the facility. The haulaway carrier then tells us where to unload the vehicle, in which lane, by truck load. Vehicles are then queued up in a single lane, ready to be driven onto their assigned haulaway truck for delivery to dealerships. Drivers can just pull right up to the proper lane, load their vehicles, and depart for the dealerships."

At marine auto terminals, finished vehicles are usually driven on and off RO/RO vessels on either side or stern ramps. Export vehicles are parked in a storage area until they can be loaded on a ship. Imported vehicles may undergo inspection and further processing before being hauled away. For example, at the Port of Baltimore's Fairfield Auto Terminal, facilities include a car wash, body shop, and an accessories shop for installing air conditioning, car stereos, and other options.

The number of auto terminals depends on how they are identified. The Association of American Railroads (AAR), for example, lists 214 active auto ramps in its 1995 directory of motor vehicle loading and unloading terminals [AAR95]. Each ramp consists of a set of loading/unloading tracks and the adjacent baying area. In some locations, however, two or more auto ramps can be clustered into what might be termed an auto terminal complex. Each ramp within a complex usually handles the vehicles of a specific manufacturer or a small group of manufacturers. In building the intermodal freight terminals database, 222 auto transloading locations were identified, including some with multiple auto ramps. Out of this total, 176 terminals involved transfers between truck and rail, 53 handled transfers between truck and vessel, and 33 accommodated transfers between rail and vessel. The reason the sum of these three numbers is greater than the total number of auto terminals is the fact that some marine auto terminals connect more than one pair of modes.

Statistics on the size and capacity of auto terminals are not readily available for all facilities. However, information provided by Conrail on its own auto terminals gives some indication [CRC92]. The size of Conrail's auto facilities range from the 6.2-acre Elkhart (IN) Automotive Terminal to the 60-acre Baltimore facility. The number of bay spaces varies between 360 at Elkhart and 5685 at the Doremus Automotive Terminal in Newark, NJ. The number of railcar spots goes from 10 at the Port Jersey facility in Jersey City to 80 at the Selkirk (NY) terminal. Limited data on the size of marine auto terminals also indicates considerable variation. The Fairfield Auto Terminal at the Port of Baltimore is a 50-acre facility [MPA97], while the Blount Island Marine Terminal near Jacksonville, FL, has 300 acres for storing automobiles and other RO/RO cargo [JPA97]. Colonel's Island Terminal at the Port of Brunswick, GA, covers 345 acres, but that includes space for dry bulk materials as well as imported and exported vehicles [GPA97].

Railroads control some auto terminals, while shippers, mainly vehicle manufacturers, control others. The latter generally own the terminals that serve their vehicle assembly plants, but they may also control some of the auto distribution centers. Ford Motor Company, for example, operates a distribution facility near Ayer, MA. Shipper terminals usually handle only the products made by the owning manufacturer, while railroad-owned terminals may serve anywhere from one to many customers. Marine auto terminals are usually owned by a port authority. They may be a separate, stand-alone terminal, handling only imported and/or exported vehicles, or, as is often the case, they may be an integral part of a larger marine terminal that ships and receives other types of cargo as well.

The importance of auto terminals in the nationwide distribution of automobiles and other finished vehicles has steadily increased over time for two major reasons. First, the logistics involved in the manufacturing and distribution of motor vehicles has become incredibly complex. Although southeastern Michigan is still a major auto manufacturing center, vehicle assembly plants are now scattered throughout the country and can be found in Alabama, California, Delaware, Georgia, Illinois, Indiana, Kansas, Kentucky, Louisiana, Maryland, Minnesota, Missouri, New Jersey, New York, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas, Virginia, and Wisconsin [AAR95]. Each assembly plant usually specializes in only one or a few models, which then must be distributed throughout the country or exported. Secondly, the role of the railroad industry in the distribution of finished vehicles has also increased over the years. Railroads carry about 70 percent of the automobiles and trucks made in the United States [CSXT96]. In 1995 the Class I railroads hauled 1,374,000 carloads of motor vehicles and equipment [AAR96]. The vehicles are transported in over 38,500 specially designed and constructed multilevel autorack cars owned by TTX Company [KRAT97]. Bilevel cars can hold 10 light trucks and vans, while trilevel cars can carry 15 to 18 vehicles. Almost all of these railcars are now fully enclosed, and various other improvements have been made over the years to increase the security of the vehicles and reduce the risk of damage due to vandalism, slack action, and the elements. Most motor vehicles shipped by rail now travel in dedicated autorack trains over a well-defined distribution network. As a result, rail has become the primary mode for the long-haul transport of finished vehicles, thereby increasing the importance of auto terminals.

Truck-Rail Bulk Transloading Facilities

With the spotlight usually focused on TOFC/COFC, often overlooked in the study of truck-rail intermodalism is the direct transfer of dry and liquid bulk cargo between rail and highway vehicles. The term "transloading" is sometimes used to refer to the process of physically transferring non-containerized material from one mode's equipment to another's without having to use some type of intermediate storage [HOLC95, JENN96]. Like many other forms of intermodalism, this one has been around for some time. Before 1900 coal was dumped from railroad hopper cars directly into horse-drawn wagons. Another simple method that is still occasionally used is to spot a hopper car on a trestle and unload its contents into a truck parked underneath [MULL95]. Nevertheless, bulk transloading has also benefited from technological developments over the past 40 years. Pneumatic systems first appeared in the late 1950s, followed by the vacuum trailer in the 1970s. General America Transportation (GATX) developed the Airslide covered hopper, while New York Central built the Flexi-Flow Freightcar [MULL95]. Other equipment used to directly transfer small-grained or flowable dry bulk cargo such as cement, fertilizer, flour, and plastic pellets includes augers, blowers, and portable vacuum/air conveyors. Coarse or non-flowable dry bulk commodities such as scrap metal or metal ingots can be directly transferred using cranes equipped with grab buckets. Liquid bulk cargo such as petrochemicals, acids, caustic soda, corn sweeteners, and petroleum products is pumped directly between tank trucks and railroad tank cars.

Unlike TOFC/COFC terminals, which have evolved into medium to large hub facilities attached to mainline tracks, truck-rail bulk transloading facilities tend to be inconspicuous, widely scattered, and generally off the beaten path. They may be found virtually anywhere there is available siding space as well as space for trucks and transfer equipment, including rail yards, team tracks, secondary branchlines, and industrial spurs. Although available information about them is sketchy, incomplete, and sometimes ambiguous, their size or capacity may be as small as two and as large as 450 railcar spots. The average size appears to be around 60 car spots, but half of the truck-rail bulk transloading facilities have less than 35 to 40 spots [MBT96].

Bulk transload terminals differ considerably in the number and types of bulk commodities they will handle. For example, some terminals transfer only flour, and others only plastic pellets. One terminal may take only liquid and dry bulk foods or food-grade commodities, while another is devoted to handling only dry and liquid chemicals. Some terminals transload only dry bulk cargo, while others will accommodate only liquid bulk commodities. The larger terminals will generally handle almost any kind of flowable dry and liquid bulk cargo including foods, chemicals, petrochemicals, and petroleum products.

Although the railroads own the tracks and usually supply the railcars, they do not necessarily own the transloading equipment or control transloading operations. A variety of arrangements between railroads, bulk motor carriers, third-party terminal operators, and shippers have been implemented. CSXT has formed a subsidiary, Bulk Intermodal Distribution Services (BIDS), to manage its transload terminals. Norfolk Southern's Thoroughbred Bulk Transfer Terminals are operated under contract by either motor carriers or terminal operating companies. NS also serves a system of third-party owned and operated Independent Bulk Transfer Terminals. In November 1996, Conrail sold some of its Flexi-Flo bulk terminals to Matlack Bulk Intermodal Services (MBIS) and the remainder to Savage Industries. Some of the larger bulk motor carriers, including Bulkmatic Transport, Chemical Leaman Tank Lines, J. P. Noonan Transportation, and Truck-Rail Handling, serve not only railroad-owned terminals but their own bulk transloading facilities as well. Transloading allows for consolidation of shipments and improves the utilization of the motor carriers' equipment. Third-party transload facility operators such as MBIS and Savage not only provide transloading equipment and oversee transloading operations, but also provide other services and equipment, including truck scales, blending meters, sampling services, hot water or steam tank car heating, tank trailer cleaning, and liquid storage tanks.

A bulk transload terminal may be available to all potential users, or it may be restricted. At one extreme, a facility may be established just to serve a single shipper. Bulkmatic Transport, for example, operates a facility served by Illinois Central in Effingham, IL, to transload flour for a shipper based in nearby Teutopolis. Motor carriers often do not allow outside carriers to use their transload facilities, or they may restrict the kinds of commodities that outside carriers can transfer. Railroads likewise may reserve a terminal for the exclusive use of a particular motor carrier [JENN96]. At the other extreme, a transload facility may be made available to any bulk carrier and its customers, although certain railcar spots may be reserved for a particular carrier.

Perhaps one reason why bulk transloading facilities have lived in relative obscurity is the fact that, until very recently, they have not been widely marketed or advertised. Very few transload terminal operators have their own marketing staffs [JENN96]. The railroads, in particular, seem to rely on inquiries from shippers or motor carriers. CSXT solicits business for its BIDS terminals mainly through its own commodity marketers rather than through its terminal operators [JENN96]. By late 1997, however, some railroads, particularly CSX Transportation, Norfolk Southern, and Canadian Pacific, were beginning to market their bulk transloading services more directly by posting the locations of their bulk transloading facilities on their World Wide Web sites.

Truck-rail bulk transloading appears to be a common and viable method of intermodalism. However, very little research has been conducted on this concept outside of the recent work by Jennings and Holcomb [JENN94, HOLC95, JENN96]. More studies are needed to document existing transload facilities and to better understand the extent, economics, and operation of truck-rail bulk transloading.

Truck-Rail Reload Facilities

Reload facilities are warehouses, distribution centers, and other locations where breakbulk commodities are transloaded between motor carrier and railroad. Trucks may bring the commodity to the terminal for consolidation and reloading into railcars, or one or more carloads of the commodity may be partially or fully unloaded at the terminal and reloaded onto trucks for local delivery. The cargo may be transloaded or directly transferred between modal vehicles using forklifts, cranes, or other suitable material-handling equipment, or it may be placed in an open or covered area for short-term storage in case equipment for the receiving mode has not been spotted yet. Examples of commodities handled at various reload centers include logs, lumber, steel products, canned goods, auto parts, paper products, and brick and other building materials.

Three types of reload facility can be identified. One is an open yard. This type is generally used for transloading logs and pulpwood. A second type is a specialized warehouse or distribution center whose primary function is to transfer a specific type of commodity such as lumber, auto parts, steel slabs or coils, or paper products between truck and rail. The third type is a regular public warehouse or distribution center with multiple functions, one of which is to reload a particular commodity from truck to rail or rail to truck. Many warehouses and distribution centers are now being designed or modified to provide space for intermodal transfers or to facilitate truck-rail transloading [JENN96].

The following examples give an indication of the proliferation as well as the variety of truck-rail reload facilities in operation:

Liquid Bulk Terminals

Liquid commodities such as chemicals, crude oil, petroleum products, asphalt, animal fats and oils, vegetable oils, molasses, sugars and sweeteners, spirits, and fertilizers are often shipped in bulk in large volumes. Producers, brokers, and transporters of these commodities usually consolidate them in and distribute them from liquid bulk terminals. These are "tank farms" or clusters of storage tanks. The number of storage tanks at these liquid bulk terminals can range from two or three to several hundred.

Liquid bulk terminals receive liquid commodities by one or more modes of transportation, store the products in large tanks, and ship the commodities out in one or more modes of transportation. They provide both consolidation and distribution functions and are generally intended for long-term storage transfers. However, some liquid bulk terminals are also capable of direct transfers between two or more modes of transport, such as between tank car and tank truck or between tank barge and tank car. Liquid bulk terminals which store and transfer crude oil and various petroleum products are often connected to the trunk crude oil or petroleum products pipeline networks. Facilities which handle other kinds of bulk liquids besides petroleum products tend to be located at or very near seaports and inland waterways for receipt and shipment by ocean tankers and tank barges. Short, local pipelines connect the docks with the storage tanks at these facilities.

Liquid bulk terminals can be grouped into two categories. The first group is made up of petroleum terminals. These facilities specialize in storing either crude oil or various kinds of petroleum products and petrochemicals. They are usually owned by oil producers and refiners and pipeline companies and tend to be proprietary, storing only the products of their owners. However, a few also make some of their storage space available to the public. The Petroleum Terminal Encyclopedia lists nearly 1500 petroleum terminals in operation in the United States in 1995 [PTE95]. The second group consists of what might be called multi-commodity liquid bulk terminals. These facilities store and transfer a variety of liquid products including various kinds of agricultural and industrial chemicals; various liquid edibles or food-grade products such as animal fats and oils, food acids, vegetable oils, molasses, syrups, and sweeteners; and petroleum products and petrochemicals. The smaller terminals in this category usually specialize in only one or two types of commodity, while the largest ones may accommodate almost any kind of bulk liquid product. Multi-commodity liquid bulk terminals are usually owned by commercial terminal operating companies and, in most cases, are intended for the use of oil producers, chemical manufacturers, product manufacturers, food growers and producers, utilities, transportation companies, commodity brokers, and government and military agencies. A few, however, are proprietary. In 1995, member companies of the Independent Liquid Terminals Association (ILTA), an international trade association, operated more than 400 liquid bulk terminals in the United States with a combined capacity of over 264 million barrels [ILTA95].

Grain Terminals

Grain terminals are facilities where one or more kinds of grain are either stored for subsequent shipment or where grain is directly transferred between different modes of transportation. The former type of grain terminal is by far the most common one. It usually consists of a complex of connected, reinforced concrete, cylindrical towers or silos known as grain elevators or metal frame storage bins or warehouses with compartmented interiors. A few grain-handling facilities allow the product to be dumped directly from one mode to another such as from trucks into barges. Most grain terminals are devoted solely to the storage and transshipment of this commodity, although a few also handle other dry bulk cargo such as feeds, agricultural seeds, meal, and flour and other products made from grain.

The function of a grain terminal depends on its location. Terminals located in grain-producing areas are generally used to consolidate local shipments of grain, usually made by truck, for subsequent shipment in covered hopper railcars or barges. Grain terminals located at major Great Lakes ports and seaports consolidate truck and/or rail shipments for eventual loading into lake vessels or oceangoing ships. Some of these terminals, particularly ones located along the east coast, may receive imported grain or coastwise shipments of domestic grain for local distribution by rail or by truck. Terminals in market areas generally store grain received by ship, barge, or rail and distribute it locally or regionally by truck or rail. Some grain terminals exist only to serve adjacent milling operations and are considered major freight traffic attractors rather than intermodal terminals.

An extensive infrastructure is devoted to the storage, handling, and shipment of corn, wheat, and other grains. The Official Railway Guide [ORG97A] lists over 2600 grain facilities, and this listing is incomplete. The State of Iowa alone has around 700 grain facilities [GERS97]. Before the 1930s, railroad companies owned most of the grain elevators. Today, most grain terminals are owned by cooperatives, grain marketing companies, or public agencies such as port authorities.

Waterway Intermodal Terminals

This group includes terminals having the facilities and equipment to transfer cargo between land and water modes of transportation. These terminals are located along the Atlantic, Pacific, and Gulf coasts; the Great Lakes; and navigable inland rivers, lakes, and canals.

It is important to understand that a port is not an intermodal terminal, nor is every dock. A port is an agglomeration of mostly disparate facilities that may be spread over a large contiguous area or several separate areas. The Port of South Louisiana, for example, includes a 52-mile stretch of the Mississippi River between the Port of Baton Rouge and the Port of New Orleans [PSL97]. In addition to having one or more intermodal terminals, a port may also include other types of facilities such as ship or barge building and repair yards, vessel mooring and bunkering areas, marinas and passenger boat terminals, parks and recreational areas, Coast Guard and U.S. Army Corps of Engineers facilities, seafood receiving and processing facilities, and refineries, chemical plants, steel mills, and other manufacturing establishments. A dock is a pier, wharf, or other landing place where barges or ships can be loaded and unloaded. It may or may not have an intermodal function. A dock, for example, may belong to a chemical plant, steel mill, power plant, refinery, sawmill, or other waterfront establishment that receives supplies or ships products by water. An intermodal terminal may have multiple docks, and multiple intermodal terminals may share one or more docks.

The size, layout, and other physical characteristics of waterway intermodal terminals vary considerably. In addition to having one or more docks, a waterway intermodal terminal will generally have some type of cargo storage area, cargo handling equipment, and spots for trucks and/or railcars. A terminal's storage facilities and mechanical handling equipment are largely determined by the kinds and volume of cargo it handles. Breakbulk and container terminals, for example, may have transit sheds, refrigerated warehouses, container freight stations, and outdoor storage yards along with various types of cranes, straddle carriers, yard hustlers or tractors, and forklifts. Facilities and equipment for storing and handling coal, grain, ores, and other dry bulk cargo include towers and silos, compartmented warehouses, outdoor storage areas, elevators, thawing sheds, conveyor systems, rotary railcar and truck dumpers, undertrack pits and chutes for railcar dumpers, receiving hoppers, loading towers and chutes, grab bucket systems, shipping galleries with loading spouts, loading booms and spouts, pneumatic suction loaders and unloaders, stackers and reclaimers, and various cranes. Liquid bulk terminals use tanks for storage and pneumatic and hydraulic pumping equipment, hoses, hose handling frames, mast and boom hose handling cranes, and pipelines for cargo loading, unloading, and transferring [MULL95].

Table 2-1. Number of Waterway Intermodal Terminals by Type of Cargo Handled
Cargo Type
Total
With Rail
Onlya
Containerized
207
112
63
Breakbulk
515
310
156
          Foods 53 39
          Logs & lumber 172 98
          Machinery & heavy equipment 68 51
          Metal products 238 162
          Pulp & paper products 85 66
          Vehicles 54 33
          Other & unspecified 276 164
Dry bulk
1130
545
828
          Coal 271 119 88
          Grain 411 202 283
          Ores & bulk metals 203 139 49
          Cement 105 37 86
          Wood chips 23 16 10
          Other & unspecified 565 265
Liquid bulk
731
271
615
          Petroleum 558 171 429
          Non-petroleum 291 193 87
TOTAL
2088
953
a. Number of waterway intermodal terminals that handle no other type of cargo or commodity except the one associated with the table row.

Table 2-1 presents a breakdown of existing waterway intermodal terminals by type of cargo and selected commodities. It shows the total number of terminals that handle a particular cargo type or commodity as well as the number that can send or receive that kind of cargo or commodity by rail. The rightmost column, labeled "only", shows the number of waterway terminals that handle no other type of cargo or commodity except the one associated with the given table row. This column gives an indication of the degree of cargo specialization among waterway intermodal terminals.

For a waterway terminal to be served by rail, there must be track space nearby where railcars can be spotted for loading and unloading cargo that is being transferred at the terminal. Many waterway terminals are located next to railroad tracks, but in some cases, these are running tracks and not sidings, storage tracks, or team tracks for railcar loading and unloading. Railroad facilities at a waterway terminal may be located on the dock or apron, alongside or inside a warehouse or transit shed, within an outdoor storage yard or marshalling area, or even off-site at an adjacent rail yard. As Table 2-1 shows, less than half of all waterway intermodal terminals have track space for loading and unloading railcars either on or adjacent to the terminal's premises. It is interesting that breakbulk terminals are much more likely to have direct railroad connections than either dry bulk or liquid bulk terminals.

Despite the considerable attention given to containerization, of the nearly 2100 waterway intermodal terminals in the United States, only about 10 percent transfer containerized cargo. Most of the container-handling terminals also accommodate some other type of cargo, usually breakbulk but also dry bulk. Only 3 percent of all waterway intermodal terminals are devoted exclusively to containerized freight.

Table 2-2. Number of Container-Handling Waterway Terminals by Location
Atlantic Coast: Pacific Coast:
          Portland, ME  (1)           San Diego, CA  (2)
          Portsmouth, NH  (1)           Los Angeles/Long Beach, CA  (16)
          Boston, MA  (3)           San Francisco, CA  (2)
          Port of Quonset/Davisville, RI  (1)           Oakland, CA  (10)
          Port of New York/New Jersey  (9)           Richmond, CA  (1)
          Philadelphia, PA & Camden, NJ  (7)           Stockton, CA  (1)
          Wilmington, DE  (1)           Portland, OR  (3)
          Baltimore, MD  (5)           Longview, WA  (1)
          Ports of Hampton Roads, VA  (5)           Grays Harbor, WA  (2)
          Richmond, VA  (1)           Olympia, WA  (1)
          Wilmington, NC  (1)           Tacoma, WA  (6)
          Charleston, SC  (3)           Seattle, WA  (11)
          Savannah, GA  (2)           Ports of Alaska  (51)
          Fernandina Beach, FL  (1)           Ports of Hawaii  (10)
          Jacksonville, FL  (3)
          Port of Palm Beach, FL  (1) Great Lakes:
          Port Everglades, FL  (3)           Detroit, MI  (1)
          Miami, FL  (10)           Chicago, IL  (2)
          Milwaukee, WI  (1)
Gulf Coast:
          Port Manatee, FL  (1) Columbia-Snake River System:
          Tampa, FL  (3)           Lewiston, ID  (1)
          Mobile, AL  (2)           Clarkston, WA  (1)
          Gulfport, MS  (1)           Pasco, WA  (1)
          New Orleans, LA  (7)           Umatilla, OR  (1)
          Lake Charles, LA  (1)           Port of Morrow/Boardman, OR  (1)
          Port Arthur, TX  (1)
          Houston Ship Channel, TX  (4)
          Galveston, TX  (1)
          Freeport, TX  (1)
          Corpus Christi, TX  (1)

Table 2-2 shows the geographic distribution of container-handling waterway terminals. In the lower 48 States marine container terminals can be found all along the East, West, and Gulf coasts. On the Atlantic side, the largest concentrations of these terminals are at the Port of New York and New Jersey in the New York City metropolitan area, the Port of Philadelphia and Camden, the Port of Baltimore, and the Ports of Hampton Roads covering Chesapeake, Newport News, Norfolk, and Portsmouth, VA. The Port of Miami also has a relatively large number of container-handling terminals, but most of these are small, privately owned terminals that accommodate low-capacity container vessels engaged in Caribbean and South American trade. The major locations of container terminals near the Gulf coast are New Orleans and along the Houston Ship Channel between LaPorte and Houston, TX. The leading center of containerized trade in the United States is at the twin Ports of Los Angeles and Long Beach. Other major concentrations of marine container terminals on the West coast can be found in Oakland, CA, and in Seattle and Tacoma, WA. The 51 container terminals shown in Table 2-2 for Alaska are scattered from Ketchikan and Metlakatla on the Southeast corner of the state to Adak in the Aleutian Islands. Many of these terminals are operated by companies engaged in Alaskan trade such as Alaska Marine Lines, Boyer Alaska Barge Line, and Northland Services. The Hawaiian Islands also have a relatively large number of container-handling port terminals, reflecting the fact that a considerable amount of inter-island traffic involves containerized freight. Table 2-2 shows very little container activity on the Great Lakes and virtually none on the shallow draft inland waterways with the interesting exception of the Columbia-Snake River System above Portland, OR. Containerized cargo is loaded onto barges at five locations for movement downriver to Portland where the containers are then transferred to oceangoing vessels.

Excluding the ones in Alaska and Hawaii, about three-fourths of the container-handling port terminals have rail service. In most cases the railroad connections are "near-dock" rather than "on-dock". In other words, the rail intermodal facility is off-site but is either adjacent to or very close to the marine terminal. The containers must be placed on chassis and drayed "across the street" or "down the block" to transfer them between the docks and the railroad. Only a few marine container terminals have large on-dock rail facilities capable of transferring a trainload of containers, but the number of container terminals with on-dock rail facilities is growing.

About one out of every four waterway terminals transfers one or more kinds of breakbulk commodity. The two most common types of breakbulk cargo are metal products such as iron and steel coils, plates, slabs, beams, or pipe sections and logs, lumber, or wood in the rough. About 70 percent of the terminals that send and receive breakbulk cargo also handle other types of cargo such as containerized freight and dry or liquid bulk commodities. Breakbulk terminals in general do not exhibit a great deal of cargo specialization. Two minor exceptions are terminals used to transfer logs arriving mainly by truck and terminals for transferring imported and/or exported motor vehicles. Although they may advertise the fact that they handle large volumes of steel coils or paper products, most breakbulk terminals ship and receive a wide variety of cargo. The kinds and volumes of commodities handled may change each year in response to market conditions.

Slightly more than half of all waterway intermodal terminals handle some kind of dry bulk commodity, and about 40 percent handle only dry bulk cargo. Although the majority of these terminals transfer more than one kind of dry bulk commodity, Table 2-1 indicates that a significant degree of cargo specialization does exist. In particular, it shows that, among the terminals with dry bulk facilities, over a third handle grain, and nearly one-fourth are used solely for grain storage and transfer. In fact, about one out of every five waterway intermodal terminals has facilities or equipment for transferring grain, and approximately one out of every eight is used for grain-handling only. Cement is another commodity with its own extensive network of waterway terminals. Cement manufacturing is highly regional, and the cost of transporting cement can quickly surpass its value. Therefore, the cement industry has built numerous terminals for local distribution by truck to ready-mix concrete plants, concrete products manufacturers, building materials dealers, and contractors. Many of these cement terminals are located at Great Lakes ports and along the inland waterways where bulk cement is received by barge or lake vessel. Terminals for imported cement can be found at major seaports on the Atlantic and Pacific coasts. A significant number of waterway terminals handle only building and construction materials such as sand and gravel, aggregates, or limestone for storage and local distribution.

Table 2-3. Major Rail-Barge Bulk Transloading Terminals on the Inland Waterways
Terminal Location Waterway
RR
Commodities
Big Sandy Terminal Neal, WV Big Sandy R.
NS
Coal
Peabody Coal Co. S. Carrollton, KY Green R.
PCCX
Coal
Havana Transfer Terminal Havana, IL Illinois R.
IMRR
Coal
Shipyard Terminal & Industrial Park Seneca, IL Illinois R.
CSXT
Coal, fertilizer, lime, salt
Port Amherst Coal Dock Port Amherst, WV Kanawha R.
CR
Coal
Kaskaskia Regional Port District Dock #1 Red Bud, IL Kaskaskia R.
KRPD
Coal
New Orleans Public Bulk Terminal New Orleans, LA Mississippi R. Gulf Outlet
NOPB
Coal, coke, fertilizer, minerals, ores, salt
Burnside Bulk Marine Terminal Burnside, LA Mississippi R.
IC
Coal, coke, fertilizer, minerals, ores, salt
Cora Coal Terminal Rockwood, IL Mississippi R.
UP
Coal
Consolidation Coal Co. Kellogg, IL Mississippi R.
UP
Coal
Peavey/ConAgra E. St. Louis, IL Mississippi R.
TRRA
Coal, other dry bulk
Lange-Stegmann Co. St. Louis, MO Mississippi R.
BNSF

TRRA
Coal, coke, fertilizer, grain, ores, salt
ACT Western Terminal St. Louis, MO Mississippi R.
BNSF
Coal
ORBA-Johnson Transshipment Keokuk, IA Mississippi R.
BNSF
Coal
I.E.I. Barge Services E. Dubuque, IL Mississippi R.
CC
Coal, other dry bulk
Prairie Sand & Gravel Prairie du Chien, WI Mississippi R.
BNSF

WICT
Coal, fertilizer, grain, salt
Union R.R. Coal Docks Duquesne, PA Monongahela R.
URR
Coal
USX Corp. Lower Coal Dock Clairton, PA Monongahela R.
URR
Coal, coke
USS Mining Co. Cumberland Mine Dock Alicia, PA Monongahela R.
USSX
Coal
Colona River Services Aliquippa, PA Ohio R.
CSXT
Coal
Missouri Mining Co. Warrenton, OH Ohio R.
WE
Coal
Ohio Valley Transloading Co. Powhatan Pt, OH Ohio R.
CR
Coal
Ohio River Terminals Co. Huntington, WV Ohio R.
CSXT
Coal

Of the over 1100 waterway terminals that handle dry bulk cargo, slightly less than half have rail facilities. These include a relatively small but significant number of waterway terminals whose sole or primary function is to transfer large volumes of some kind of bulk commodity such as coal, coke, or iron ore between railcars and either barges or deep draft vessels. In many cases, these terminals have space for loading or unloading unit trains. Rotary railcar dumpers or undertrack pits for bottom-dumping railcars are generally used to unload railroad equipment for rail-to-water transfers. The material is then picked up by a conveyor belt system for direct loading into ships or barges or for stacking in either an open or enclosed storage area where the material is subsequently reclaimed for vessel loading. For water-to-rail transfers, ships or barges typically unload into a conveyor belt system that moves the material to either a railcar loading tower or to a storage area from which railcars can be loaded. Table 2-3 shows the major rail-barge bulk transloading terminals on the inland waterways. Most of these are coal transloading facilities, but some transfer other bulk commodities as well. Table 2-4 lists the major rail-ship bulk transloading terminals at deep water ports on the Great Lakes and the Atlantic, Pacific, and Gulf coasts. These include coal export terminals and iron ore docks along Lake Superior and Lake Erie.

Table 2-4. Major Rail-Vessel Bulk Transloading Terminals at Deep Water Ports
Terminal Location Waterway
RR
Commodities
Greenwich Ore Pier # 122 Philadelphia, PA Delaware R.
CR
Fertilizer, iron ore
Broadway Terminal Camden, NJ Delaware R.
CR
Coal, minerals
Consolidation Coal Sales Co. Baltimore, MD Patapsco R.
CR
Coal
CSXT Curtis Bay Ore Pier Baltimore, MD Curtis Bay
CSXT
Ore
CSXT Curtis Bay Coal Pier Baltimore, MD Curtis Bay
CSXT
Coal
Curtis Bay Co. Bayside Coal Pier Baltimore, MD Curtis Bay
CSXT
Coal
Lambert's Point Coal Pier # 6 Norfolk, VA Elizabeth R.
NS
Coal
Alcoa Paradise Point Transfer Station Chesapeake, VA Elizabeth R.
NS
Alumina
Newport News Pier 9 Newport News, VA James R.
CSXT
Coal
Dominion Terminal Newport News, VA James R.
CSXT
Coal
Shipyard River Coal Terminal Charleston, SC Shipyard R.
CSXT

NS
Coal, other dry bulk
CSXT Rockport Terminal Tampa, FL East Bay
CSXT
Phosphate products
McDuffie Coal Terminal Mobile, AL Mobile Bay
BNSF

CSXT

IC, NS
Coal
Bulk Materials Handling Plant Mobile, AL Mobile R.
TASD
Coal, coke, ores, potash
New Orleans Public Bulk Terminal New Orleans, LA Mississippi R. Gulf Outlet
NOPB
Coal, coke, fertilizer, minerals, ores, salt
Burnside Bulk Marine Terminal Burnside, LA Mississippi R.
IC
Coal, coke, fertilizer, minerals, ores, salt
PABTEX Terminal Co. Port Arthur, TX Sabine-Neches Canal
KCS
Coal, coke
Bulk Materials Handling Plant Houston, TX Houston Ship Channel
PTRA
Coal, fertilizer, grain, minerals, ores, potash
AIMCOR Marine Terminal Texas City, TX Galveston Bay
TCT
Coal, coke, fertilizer
Corpus Christi Bulk Terminal # 1 Corpus Christi, TX Corpus Christi Bay
TM, UP
Ores
Corpus Christi Bulk Terminal # 2 Corpus Christi, TX Corpus Christi Bay
TM, UP
Coal, coke, ores
Kaiser International Corp. Los Angeles, CA San Pedro Bay
HBL
Ores

Over one out of every three waterway intermodal terminals deals with liquid bulk cargo, and a large majority of these terminals store and transfer only liquid bulk commodities. Petroleum products are by far the most common type of liquid bulk commodity handled at these terminals. Over three-fourths of the waterway terminals with liquid bulk facilities handle petroleum products, and nearly 60 percent of these terminals handle no other kind of commodity. About one out of every five waterway intermodal terminals is a petroleum storage and transfer facility.

Waterway terminals for non-petroleum liquid bulk commodities tend to handle other cargo as well. Liquid bulk tank farms, for example, often store and transfer petroleum products as well as miscellaneous chemicals and liquid food-grade products such as vegetable oils and molasses. In agricultural areas, waterway terminals for liquid and dry bulk fertilizers are quite common, and some grain-handling terminals also store liquid fertilizers. Rail service is also much more common at terminals with liquid bulk facilities for non-petroleum commodities. Nearly two-thirds of the waterway terminals that handle non-petroleum liquid bulk cargo have rail connections, compared to only 37 percent for all waterway terminals with liquid bulk facilities.

SUMMARY

A freight terminal is an integrated set of facilities where cargo is loaded onto or unloaded from a particular mode of transportation. An intermodal freight terminal is a special kind of freight terminal. It is a place where two or more modes of transportation meet to interchange freight, either directly or through intermediate storage. Intermodal exchange may not be the only function performed at an intermodal terminal, or even the primary one. All that is required for a freight terminal to be an intermodal terminal is that it have the necessary space and equipment to receive cargo by one mode of transportation and ship it out by a different mode. In between the inbound and outbound movement, the cargo may be consolidated with other incoming cargo of the same type, separated into smaller outbound shipments, or directly transferred between two modes as part of a seamless intermodal shipment.

The physical characteristics, complexity, and other attributes of intermodal freight terminals vary greatly. Some intermodal terminals are relatively small and almost ad hoc in nature, while others are large and involve a considerable amount of infrastructure. As an aid in identifying and classifying intermodal terminals and thereby gaining a better understanding of the concept of intermodalism in its various forms, five features or characteristics of intermodal terminals are especially useful: the pairs of modes which the terminal directly or indirectly connects, the types of cargo or the specific commodities which the terminal handles, the types of intermodal transfers for which the terminal is designed (direct, short-term storage, or long-term storage), whether the terminal is privately or publicly owned, and whether use of the terminal is open or restricted relative to either shippers or carriers.

By employing combinations of the above five distinguishing features, several common types of intermodal terminals can be found. Between the truck and rail modes, for example, containerized freight is interchanged at TOFC/COFC facilities, automobiles and other finished vehicles at vehicle terminals, dry and liquid bulk cargoes at bulk transloading facilities, and breakbulk commodities such as lumber, steel, and paper products at numerous warehouses and other reload centers. Grain is transferred between truck and rail, truck and water, or rail and water at thousands of grain elevators and other grain-handling facilities. Petroleum products and other liquid bulk commodities such as chemicals, vegetable oils, and molasses are gathered and distributed intermodally at hundreds of tank farms and other liquid bulk terminals and storage facilities. Along the Atlantic, Pacific, and Gulf coasts, the Great Lakes, and the inland waterways can be found a variety of intermodal terminals including container-handling facilities, import and export auto terminals, grain elevators, cement terminals, rail-barge and rail-vessel bulk transloading facilities for coal, ores, and other dry bulk cargoes, and facilities for storing and transferring petroleum and non-petroleum liquid bulk commodities. Clearly, an amazing variety of intermodal terminals exists.

CHAPTER 3. DESIGN OF THE INTERMODAL TERMINALS DATABASE

Designing a database of intermodal terminals is a matter of deciding what information to include and how to organize it. This is not as simple as it may seem. A long list of possible data items describing various physical and operational characteristics of intermodal terminals could easily be generated. Simply compiling all or a large subset of these items into a single record for each terminal would likely lead to a database too cumbersome and costly to build, maintain, and use. A poorly designed database, like a poorly designed bridge, will soon collapse under use. What is needed is a logical database structure based on a conceptual model of intermodal terminals and their relationships with other elements of the transportation system. In deciding what information to include and how to arrange it logically, consideration must be given to the intended uses of the intermodal terminals database, its connections with modal network databases, and such practical issues as the availability of data or the ability to measure certain properties or characteristics.

This chapter presents the design of the intermodal freight terminals database. More importantly, it discusses the basis for the design. The chapter begins by describing a simple conceptual model of intermodal freight terminals and how they are connected with the rest of the transportation system. This model reveals the fundamental elements and attributes of intermodal terminals that will need to be included in the database. Because the intermodal terminals database is a component of the much broader National Transportation Atlas Database (NTAD), its design must be compatible with the NTAD's structure. The next section therefore describes the structure of the NTAD and how its record types represent different geospatial features, including intermodal terminals. The discussion then turns to other design factors, including the intended uses of the terminals database, technical issues raised by the conceptual model, and the availability of terminals data. The final section describes in detail the structure of the initial version of the intermodal terminals database, using relational database theory as the data model.

CONCEPTUAL MODEL OF INTERMODAL TERMINALS

Chapter 2 examined the most important distinguishing characteristics of intermodal terminals and gave examples of some of the more common types of these facilities. It revealed a great deal of diversity among intermodal terminals, particularly in the kinds of cargo or commodities that are interchanged at each facility. To reduce some of the complexity of intermodal terminals and thereby aid in designing an intermodal terminals database, it is helpful to abstract the key elements of intermodal terminals and their connections with the rest of the transportation system.

Figure 3-1 illustrates schematically the inner workings of an intermodal terminal. In its simplest form, the terminal can be represented as a box with inflows and outflows. In this example, the terminal receives coal by rail, dry bulk fertilizer by barge, and lumber by truck. The coal leaves the terminal by barge, the fertilizer by both truck and rail, and the lumber by rail only. What happens within the terminal is represented by the four dashed arrows inside the box. The arrow labeled R-W(C) indicates that coal is transloaded from railcars into barges. If necessary, the coal can be stacked on the ground in case no barges are berthed at the terminal's docks. Barges unload dry bulk fertilizer into a receiving hopper served by an electric belt conveyor system which transfers the material into a storage warehouse. Much of the fertilizer is subsequently loaded into trucks for local distribution, but some of it is loaded into covered hoppers for shipment by rail to a more distant location. These transfer activities are depicted by the two dashed arrows labeled W-H(F) and W-R(F), respectively. The lumber, which is produced locally and arrives by truck, is simply reloaded onto railroad flatcars for shipment to distant cities. The dashed arrow labeled H-R(L) symbolizes this intermodal exchange. Thus, all of the intermodal activity occurring at this terminal, which is served by three modes of transportation and handles three different commodities, can be represented by four intermodal connectors.

An intermodal connector is defined by a pair of modes and a commodity or type of cargo. In Figure 3-1, there are two intermodal connectors between rail and water - one for coal and one for dry bulk fertilizer. If the terminal also reloaded steel plates from trucks to railcars, there would be a second highway-rail connector, which might be labeled H-R(S).

Another important property of intermodal connectors is directionality. Intermodal connectors indicate the direction of the intermodal transfer. In Figure 3-1, the two rail-water connectors have opposite directions. The one for coal is from rail to water, while the one for fertilizer is from water to rail. All four of the intermodal connectors in the example are unidirectional, but an intermodal connector can be bidirectional as well if the cargo type or commodity can be exchanged in either direction. At a TOFC/COFC facility, for example, containers can be transferred from railcars to trucks and vice versa. Both interchanges could be represented schematically by a single double-ended arrow.

Intermodal connections at a terminal can change over time in response to changes in the terminal's functions. This is reflected in the conceptual model by the addition and deletion of individual intermodal connectors. If, for example, the lumber reload facility at the terminal in Figure 3-1 were to close, the H-R(L) intermodal connector would disappear. Similarly, if dry bulk fertilizer were no longer shipped out by rail, the W-R(F) connector would also vanish. On the other hand, if the terminal began receiving steel products by barge for reloading into trucks and railcars, two new intermodal connectors would appear - W-H(S) and W-R(S).

The more complex the intermodal terminal or the more functions it performs, the greater the number of intermodal connectors it is likely to have. The terminal depicted in Figure 3-1, for example, is relatively complex, although not atypical. It functions as a rail-to-barge coal transloading facility, a fertilizer storage and distribution warehouse, and a truck-to-rail lumber reload center. The four intermodal connectors reflect these three functions. A simpler intermodal terminal, such as an auto ramp, that transfers only one type of cargo or commodity between two modes would have only one intermodal connector. Note also that, even though the example terminal is a waterways facility, not every intermodal connection involves the water mode of transportation. In general, there does not have to be an intermodal connection between every pair of modes serving an intermodal terminal.

Intermodal connectors are a powerful abstraction. They distill the essence of the intermodal transfer process while concealing most of the details. The R-W(C) connector in Figure 3-1, for example, represents the entire coal transloading process, including all the yard space, facilities, personnel, and equipment involved - rotary car dumpers, electric belt conveyors, stackers, reclaimers, barge loading towers, equipment operators, docks, outside storage area, railcar storage tracks, and so on. Intermodal connectors represent the time, space, equipment, and processes involved in transferring freight either directly or indirectly between two modes of transportation.

Figure 3-1 depicts only one part of the conceptual model of intermodal terminals. There remains the part of connecting the terminal with the modal networks serving it..

The mathematical construct most commonly used to represent a transportation network is the graph. A graph is a collection of vertices and edges (in the field of mathematics) or nodes and links (in the field of transportation planning). Links represent continuous portions of a travelway such as sections of highway, railroad trackage, trunk pipelines, or navigable river channels. Each link is bounded by a pair of nodes. Nodes in a transportation network usually represent intersections or junctions where traffic can enter or leave the network or change direction. They can also represent locations where one or more link attributes change resulting in changes in traffic speed or flow capacity. In more detailed representations of transportation networks, nodes have also been used to denote bridges, tunnels, and jurisdictional boundaries.

Graph theory in general does not impose any geometric or spatial structure. Transportation networks, however, represent spatial features. The shape of the network is important for conducting accurate spatial analysis. At a minimum, the shape of a transportation network is determined by the geographic coordinates assigned to its nodes. In more detailed network representations, shape points between nodes are used to display the horizontal alignment of individual links.

Currently available transportation network databases consist of individual modal subnetworks. Each subnetwork was developed independently, often by a different organization. More importantly, the various modal subnetworks are not connected with each other in the network database. A major technical challenge is to construct an analytically tractable intermodal freight network from these disparate spatially referenced subnetworks [SOUTH97]. The logical place to join modal subnetworks is at intermodal terminals.

Figure 3-2 shows the intermodal terminal of Figure 3-1 within its environs. In addition to the terminal, the figure shows nearby links of the highway, railway, and waterway subnetworks. Because coal is transferred from railcars to barges and fertilizer is transferred from barges to railcars at this terminal, we need to connect the rail and waterway subnetworks at this location. Likewise, we need to connect the highway and waterway subnetworks to simulate the multimodal movement of fertilizer and the highway and rail networks to model the multimodal movement of lumber.

To make these modal subnetwork connections, we first need to know the location of the intermodal terminal. In other words, the coordinates of the terminal must be specified. Ideally, the same coordinate system used to fix the location of nodes in the various modal subnetworks should also be applied to the intermodal terminal. If that is not possible, then a function for converting between the different coordinate systems will be necessary.

The next step is to add access links from each of the modal subnetworks to the intermodal terminal. In Figure 3-2, link a-b is the access link from the highway network. It represents the local streets trucks would follow to get from the main roadway network to the intermodal facility. Similarly, link c-d is the rail access link. It represents a rail spur, siding, branchline, or industrial track leading to the terminal from the rail mainline network. Link e-f connects the terminal with the waterway network. It may be regarded as the path a vessel or barge would take between the main channel and the terminal's docks.

Now that we have simulated access to the intermodal terminal from each modal subnetwork, we have to simulate the intermodal transfer of freight. This, of course, is the purpose of the intermodal connectors described earlier. In Figure 3-2, the dashed line b-d inside the box representing the terminal corresponds to intermodal connector H-R(L) in Figure 3-1. It represents the reloading of lumber between trucks and railcars. Similarly, line d-f matches intermodal connector R-W(C), the transloading of coal from rail to barge; line f-b with connector W-H(F), the transfer of fertilizer via intermediate storage between barge and trucks; and line f-d with connector W-R(F), the transfer of fertilizer via intermediate storage between barge and rail. Thus, the path a-b-d-c may be regarded as the movement of lumber between the highway and rail networks, path c-d-f-e as the movement of coal between the rail and waterway networks, and so on.

The intermodal terminal illustrated in Figures 3-1 and 3-2 is served by only one railroad. What if, as is sometimes the case, the terminal is served by two or more rail carriers? Each carrier might have its own set of tracks leading into the terminal, or one carrier might reach the terminal using trackage rights over one of the other carrier's lines. Either way, how does this situation affect the conceptual model? Because many network routing applications must consider each railroad's network individually, the conceptual model of intermodal terminals, which will be used to design the intermodal terminals database, should be extended to reflect this need. The easiest way to do this is to include railroad-specific access links and intermodal connectors. Suppose, for example, that some of the dry bulk fertilizer arriving by barge at the terminal in Figures 3-1 and 3-2 is hauled away by either of two rail carriers - Railroad A and Railroad B. To indicate this situation, we would replace the single intermodal connector W-R(F) with two intermodal connectors - W-RA(F) and W-RB(F). In addition, we would include a separate access link from the terminal to each rail carrier's network. If only Railroad A, however, was involved in the truck-rail lumber reload operation, then the H-R(L) intermodal connector in Figure 3-1 would be replaced by the carrier-specific connector H-RA(L).

It is important to emphasize again that each intermodal connection at a terminal is commodity-specific. The connection between the highway and rail networks at the terminal illustrated by Figures 3-1 and 3-2 is applicable to the transportation of lumber only. Likewise, the connections between the rail and waterway networks apply only to the movement of coal in one direction and fertilizer in the other. If we were interested in routing the shipment of liquid bulk vegetable oils by barge and rail, we would not be able to connect the two networks at this terminal, because it is not designed or equipped to handle the intermodal transfer of this type of commodity or cargo.

The conceptual model described above has revealed two important pieces of information about intermodal terminals that should be included in the database. One is the terminal's location, particularly its geographic coordinates. The other is the notion of an intermodal connector, which itself combines three other pieces of information: a pair of modes, a commodity or type of cargo, and the direction of transfer.

STRUCTURE OF THE NTAD

BTS has developed a standard set of file formats for the various geospatial databases that constitute the NTAD [SPEAR95]. These file formats currently define six distinct record types: link, node, point, area, geography, and attribute. Each record type defines a file of fixed length records composed of fixed length fields. The first four fields are the same for each record type. They contain a record type identifier, a version number, a revision number, and a modification date. The number and definitions of the remaining fields are unique to each record type. Combinations of these six record types are used to describe the three major types of geospatial features represented in the NTAD: networks, points, and areas.

Intermodal terminals are regarded as point features in the NTAD. They are described by the point and attribute record types. In addition to the four common fields, point type records have the following six fields:

The other record type used to describe intermodal terminals in the NTAD is the attribute record. Besides the four common fields, only one other field is specified for this record type. The FEATURID is required so that an intermodal terminal's attribute record can be linked with its point record. The BTS NTAD specification also allows for multiple attribute files for intermodal terminals and the other types of geospatial features. Each attribute file can have a different number of fields. The only requirements are that each attribute record and each field within an attribute record must have a fixed length.

The structure of the NTAD recognizes that intermodal terminals and other spatial features of the transportation system have topological as well as non-topological attributes. Most GIS software packages also recognize this distinction. The topological attributes of a transportation feature include its location and its connections with neighboring transportation features. Non-topological attributes can include a wide variety of characteristics, which vary greatly between the different types of features. Therefore, it makes sense to separate a transportation feature's topological attributes from its non-topological ones. The design of the NTAD accomplishes this by including only topological information in the link, node, point, area, and geography record types and placing the non-topological information in the attribute record type. Moreover, the NTAD structure allows for multiple attribute files for each type of spatial feature. This is critical for designing and building a modular and extendable intermodal terminals database.

The design of the NTAD imposes some organizational structure on the intermodal terminals database, but still leaves a great deal of discretion as to what information to include and how to assemble it into one or more attribute files. The only information which the NTAD specification explicitly requires is the terminal's location, using longitude and latitude as the coordinates, and its identity, denoted by a unique FEATURID as well as a facility name or description. The conceptual model presented in the previous section also revealed the need for locational information. Under the NTAD specification, this information is placed in the point record file. Information on intermodal connections at a terminal would be stored in an attribute file.

OTHER DESIGN FACTORS

Several other technical considerations and factors arise in deciding what information to include in the terminals database. These issues concern the need for network independence, the intended uses of the terminals database, the availability of data, and technical questions about how to measure certain attributes such as terminal capacity and costs or impedances. The following discussion addresses each of these technical issues.

Need for Network Independence

It is important that none of the information stored in the terminals database depend on or be associated with any particular transportation network database or collection of modal subnetwork databases. If the terminals database is even loosely tied to a specific network database, then every time the latter is replaced or upgraded, the terminals database in all likelihood would also have to be replaced or significantly modified. For example, if the terminals database contained information that tied it to a 1:2,000,000-scale highway network, then that information would most likely have to be modified extensively if the network were replaced by a 1:100,000-scale representation. Moreover, integrating the terminals database too closely to a specific network database could unduly limit its usefulness to potential users. A regional planner, for example, may wish to use all or part of the terminals database in connection with a more detailed local or regional transportation network database. This might require a considerable amount of extra work if the terminals database is too closely associated with the transportation networks included in the NTAD.

This issue arises when deciding whether or not to include information on network access in the terminals database. Referring to Figure 3-2, we might want to include information on access links a-b, c-d, and e-f or the location of points a, c, and e in the terminals database. These access points and links, however, depend on a specific representation of the highway, railroad, and waterway networks. These network models could be changed in several ways that would affect any topological information on terminal access. For example, additional links might be added to the network databases that would change the location of access points or shorten the length of access links. Using the techniques of network thinning and chaining, the analyst may also alter the network representation to better suit certain applications. These alterations would also affect any information on access points or links already embedded in a terminals database. Network access points and links, therefore, are not simply attributes of the intermodal terminal itself. They depend on the specific representation of the modal subnetworks. Moreover, they can be generated algorithmically on an ad hoc basis for different analytical applications. To keep the terminals database independent of any specific network database or representation, it would be better not to include any spatial information on network access. Rather, users of the terminals database should be given the analytical tools needed to attach the terminals database to any network database or representation available or appropriate for a particular analysis.

Intended Uses of the Database

The kinds of information stored in the terminals database and the way in which this information is organized will depend on the intended uses of the database or, alternatively, the analytical requirements of its intended users. One of the primary purposes of the NTAD is to assist transportation analysts and policymakers in studying various multimodal transportation issues at the national or multi-state regional scale. The requirements for terminals data at this high geographic level of analysis are much different than the requirements for a more localized analysis. A local planner or traffic engineer studying ways to improve highway access to an intermodal facility or the possible impact of terminal improvements on local traffic patterns will need much more detailed terminal and road network data than can or should be provided by a national database. Likewise, an intermodal terminal operator interested in improving capacity or efficiency will need and, more importantly, will have ready access to much more detailed information about the terminal and its operation than can be economically gathered, stored, and maintained in a national-scale database.

Although the NTAD regards intermodal terminals as a point feature, they can and often do cover a considerable amount of space. Therefore, it might be useful to include information on terminal boundaries or area as well as information on the internal structure, operations, and physical components of intermodal terminals. Such information would be needed for any study or application in which it is necessary to simulate or analyze the movement or circulation of freight either within the confines of the terminal or between the terminal's gates and points of access to the primary modal networks. This kind of analysis, however, is at a scale that is much too microscopic to be easily and economically supported by a national-level terminals database. Detailed information on size, shape, storage capacity, physical plant, equipment, and internal configuration is available for some types of intermodal terminals. These data, however, are too expensive to gather and maintain on a timely basis for every facility that might be included in the terminals database.

Most transportation policy applications of the terminals database are likely to involve mapping and other GIS functions, multimodal freight routing, or analysis of national, interregional, or corridor freight flows. Terminals data needed to support these activities have already been mentioned. At a minimum, the terminals database requires data on where each terminal is located, which pairs of modes it connects, and what types of cargo or commodities get transferred between each pair of modes.

It is unlikely that all of the possible uses of the terminals database or the NTAD itself will ever be fully articulated or anticipated. The database, therefore, must be designed in such a way that it can easily incorporate new information as it is needed and when it becomes available. Fortunately, the NTAD structure is flexible enough to support a modular and extendable terminals database.

Measuring Terminal Costs and Performance

To improve the quality or effectiveness of multimodal routing and freight flow analysis, it would be helpful if the terminals database included information on how much freight a terminal could handle, the amount of time required or the cost involved in effecting an intermodal transfer, or other such measures of terminal capacity, cost, impedance, or performance. Because it takes time, manpower, and equipment to unload, consolidate, sort, separate, store, transfer, and reload cargo, intermodal terminals are impediments to the steady, uninterrupted progress of freight. They are sources of congestion and delays. Intermodal transfer costs influence the degree to which shippers and carriers engage in intermodalism. Many terminal operators spend considerable time and money in a continual effort to make their facilities highly efficient or to reduce delays as much as possible. Because of the importance of this inherent characteristic of intermodal terminals, the terminals database ought to include some measures or indicators of intermodal transfer costs, impedance, or performance at each facility.

Unfortunately, there is no standard measure or definition of impedance, either on modal network links or at intermodal terminals. One of the more commonly used measures is transit time, although some attempts have been made to develop link performance functions which compute link impedance as a function of multiple variables such as transportation cost, transit time, and energy use. In the case of freight transportation, another factor affecting impedance is the type of commodity being shipped or, more precisely, the characteristics of the commodity such as its unit value, weight or density, handling requirements, packaging, quantity, and so forth. Thus, transit time is more critical for high-valued, time-sensitive freight than it is for low-valued commodities shipped in bulk quantities. The definition of terminal impedance also depends on the type of transfer involved. Transloading or reloading time and cost are more important at terminals intended for direct transfers or short-term storage transfers, while at terminals designed for long-term storage transfers, storage capacity and costs are more relevant. Intermodal transfer cost or impedance is a multidimensional variable. No single value or measure can be assigned to any given intermodal terminal or even to a particular intermodal connection at a terminal.

Likewise, no single measure or indicator of terminal performance has been widely accepted. Like intermodal transfer costs or impedance, terminal performance is a variable. At any given terminal, it changes over time in response to numerous interacting factors. Some of these factors are internal, including the following:

Other factors affecting terminal costs and performance are external. They include the following:

The implication is that neither terminal cost nor performance is something that can be easily measured once for each terminal and stored in some record in the terminals database. Some research on terminal performance measures has begun [STOPH94], but much more is needed before a standard set of terminal cost and performance indicators can be specified for the terminals database. Most likely different sets of indicators will be needed for different types of terminals or for different types of intermodal connections within a terminal. This issue further demonstrates the need for a modular database structure in which new data modules or files can be added and obsolete or outmoded data modules replaced without requiring a massive overhaul or restructuring of the entire terminals database.

Availability of Data and Other Resources

Regardless of the kinds of information analysts would like to see included in the terminals database, the actual content will depend on how much of this information is available as well as the cost of acquiring it and incorporating it into the database structure. There are two aspects to this issue. One involves whether or not the desired information has already been collected and exists in some readily usable form. The other aspect involves whether restrictions have been imposed on the use or dissemination of existing information that might be included in the terminals database.

While detailed data exist for some types of intermodal terminals, it often resides outside the public domain and is unlikely to become readily available for BTS or transportation analysts to use. The type, quality, format, level of detail, and extent of coverage of available data vary by type of terminal, particularly between publicly and privately owned terminals. Consequently, it is possible to obtain considerable if not comprehensive information for some facilities and only a minimal amount of information for many others. In particular, data on freight volumes or terminal throughput, terminal costs and capacity, and other measures of terminal performance are often sketchy, outdated, unreliable, inconsistent, or simply unavailable for many intermodal terminals. Where this information does exist, it is often proprietary with severe restrictions placed on its use or dissemination. Even the most basic information about a terminal such as its physical location, the modes of transportation which it connects, and the types of cargo or commodities it handles is difficult to obtain for some types of terminals. Chapter 4 identifies the sources of information used to build the initial version of the terminals database and describes the strengths and limitations of each source.

Historic concerns about computer memory requirements, execution times, and data storage requirements were not a major factor in designing the terminals database. The cost of computer-based resources continues to decline even while the functionality and capabilities of these resources continue to improve dramatically. Of greater concern were the requirements for skilled personnel needed to build, maintain, and use the terminals database as well as the cost of developing and maintaining the potentially large volumes and variety of data that might be included in it.

RELATIONAL DATA MODEL

After considering each of the design factors discussed above and reviewing available sources of information on intermodal terminals, it was recommended that the terminals database be developed in discrete stages, starting with more basic, publicly available data on terminal location and intermodal connections and gradually developing a broader dataset as research findings, new or better data, and more resources for gathering data become available. This appproach requires a database structure that is modular and flexible. This simply means that, instead of packing all information about intermodal terminals into a single file or record type, the information should be carefully partitioned into linked files of logically related data items which can be easily inserted, replaced, or removed from the database. The principles of relational database design constitute a data model that is especially well suited to this modular approach.

Several models or ways of logically organizing data have been developed and applied since the earliest days of automated data processing. These include the inverted list, the hierarchical, the network, the relational, and most recently, the object-oriented data models. Of these, only the relational model has a firm mathematical foundation. Introduced by Dr. E. F. Codd in 1970, the relational database model has become very popular and is the basis of many of today's most popular database management systems (DBMS) that run on personal computers and engineering workstations. The relational database model does have its limitations. For example, it does not represent certain kinds of information very readily, such as the information stored or embedded in images, text documents, and engineering drawings. This is one reason why much attention has been devoted in recent years to object-oriented database systems. Nevertheless, the relational model can suitably represent most aspects of intermodal terminals. Although the NTAD may be distributed as a collection of flat, fixed-format, fixed-record-length files, this does not preclude the application of relational database design theory to the specification of the terminals database structure.

Relational database theory is the subject of entire books (see, for example, [DATE86]). A detailed description, therefore, is beyond the scope of this report. However, two important concepts underlying relational database structure and design are worth noting here.

First, data in a relational database are logically organized into tables of a special type known mathematically as relations. Each table consists of rows and columns that store information about some aspect of a particular type of entity or a relationship between entities. In the terminals database, for example, one table could contain information on the name and geographic coordinates of each terminal, a second table could include data on each terminal's intermodal connections, a third table might deal with certain performance measures, and so on. In relational database theory, a table's rows are called tuples and its columns are known as attributes. A row or tuple describes the attributes of a particular instance of an entity or relationship such as the name, longitude, and latitude of an intermodal terminal, or the modes of transportation and type of cargo involved in an intermodal connection at an intermodal terminal. A relational table or relation contains no duplicate tuples or rows and all attribute values are atomic. Relational algebra is a set of eight unary and binary operations for manipulating relational tables. These operations are used to form new relations or tables by extracting specified rows or columns from a single table or by combining two tables through the union, intersection, difference, Cartesian product, join, and divide operations. Relational algebra is the basis of commonly used database query languages such as SQL (Structured Query Language). One of the reasons why Codd proposed the relational model was to inject these rigorous mathematical principles into database design and management.

The second important contribution of relational database design is that of normalization theory. The objective of database normalization is to reduce the amount of data duplication and the number of dependencies between attributes and sets of attributes within and between the various tables or relations of a database. A poorly designed database will contain unnecessary data redundancy and data dependencies which can make it extremely difficult to maintain the integrity of the database. Because of this redundancy, revisions or updates in one part of the database may not be made in another part, resulting in inconsistent records which can easily make the database useless or highly unreliable at best. There are several levels of normalization. Each level seeks to eliminate a certain type of data redundancy or data dependency. By definition, a relational database must be in first-normal form because each attribute value in a row or tuple is single-valued or atomic; in other words, the value of an attribute cannot itself be a tuple or list of values. Beyond first-normal form, the degree of normalization required for good database design depends on the complexity of the database, which in turn reflects the complexity of the real world entities and relationships which the database is attempting to model. In general, a database should at least be in third-normal form. This is obtained when each row or tuple in the table consists of a primary key value that identifies some entity such as an intermodal terminal, together with a set of mutually independent attribute values that describe the entity in some way. An attribute is mutually independent of all other non-key attributes when its value does not depend on the values of any of the other non-key attributes. It should be noted that third-normal form is usually a desirable goal and not an absolute requirement of sound database design.

The principal advantage of the relational data model, aside from its solid mathematical basis and widespread acceptance, is the fact that it fosters modular database design. The relational model enables the database developer to divide a complex object such as an intermodal freight terminal into its component parts and interrelationships and to logically group information about each part or relationship into a separate table. Modularity is further enhanced by applying normalization theory to the database design to reduce or eliminate unnecessary data redundancy and functional dependencies between the fields of a record or columns of a table. Thus, as new or better data about intermodal terminals become available, additional attribute-type tables (in the terminology of the NTAD specification) can be easily added to the terminals database and existing tables can be easily modified without causing the whole logical structure of the database to collapse.

INTERMODAL TERMINALS DATABASE STRUCTURE

The initial intermodal freight terminals database consists of two tables or file types. The first contains information on the identity and location of each intermodal terminal, while the second contains information on each intermodal connection at a terminal. The two tables can be linked or joined by a common field indicating the terminal's FEATURID or identifier. These two tables or file types represent the canonical form of the terminals database. All intermodal terminals, regardless of what type or how complex, have a location and at least one intermodal connection. This information constitutes the minimum necessary to implement the conceptual model of intermodal terminals described earlier. Additional tables or file types can be added to the terminals database as needed or as the requisite data become available for public dissemination. It is very likely that many of these additional tables may apply to only certain types of intermodal terminals. Alternatively, enhanced versions of the terminals database may include a table or file of "pointers" to other independent datasets or sources of information about selected terminals in the canonical database. This would be a simple way of expanding the NTAD terminals database by linking it to other terminals-related databases, both public domain and proprietary.

The following sections describe in detail the two tables making up the canonical terminals database.

Terminal Location or Point File

Each row or record of the terminal location table or point file identifies a particular intermodal terminal and its location. This table is exactly the same as the point file defined in the NTAD file format specification for transportation point facilities. Table 3-1 shows the record layout for this file.

Table 3-1. Format of the Terminal Location Table or Point File
Field

No.
Field

Name

Field

Type
Field

Length
Start

Position
End

Position
Field

Description

1
RECTYPE
char
1
1
1
Record type (always 'P')
2
VERSION
char
2
2
3
Version number
3
REVISION
char
2
4
5
Revision number
4
MODDATE
num
8
6
13
Modification date (mmddyyyy)
5
POINTID
num
10
14
23
Point ID
6
FEATURID
char
10
24
33
Feature (terminal) ID
7
LONGITUD
num
10
34
43
Longitude (signed integer with 6 implied decimal places)
8
LATITUDE
num
10
44
53
Latitude (signed integer with 6 implied decimal places)
9
DESCRIPT
char
35
54
88
Terminal name or identification
10
STFIPS
num
2
89
90
State FIPS code:

00 if in multiple States

99 if not in United States

The terminal location table or point file consists of ten fields or columns. The first four are the standard record header fields common to all files or record types in the NTAD. They include:

The remaining six fields identify each terminal and its location. Although they were presented earlier in the discussion of the NTAD's overall structure, they are repeated again here with a few additional details:

For some types of intermodal terminals, a standard set of codes already exists for uniquely naming or identifying individual terminals. Every airport, for example, has a unique 3- or 4-character airport code. Most types of intermodal terminals, however, do not have any such official identification numbers or codes. Thus, the following convention was adopted to create a unique FEATURID for each terminal in the database. The intermodal terminal FEATURID has the following format: SSCCC-NNNN. SS is the two-digit FIPS code of the State where the terminal is located and CCC is the three-digit county FIPS code. Following the concatenation of the two FIPS codes is a hyphen and then a four-digit sequential number, NNNN, which begins at 0001 for each SSCCC combination.

Intermodal Connections Attribute File

Each record in the intermodal connections table describes a specific intermodal transfer that can be made at an intermodal terminal. As the conceptual model of intermodal terminals presented earlier in this chapter demonstrated, an intermodal terminal may provide for multiple intermodal connections. In the language of database design, this is known as a one-to-many relationship. Thus, if a terminal allows for more than one intermodal connection, it will have multiple records in the intermodal connections table or file, each record describing a specific intermodal transfer. The terminal depicted in Figure 3-1, for example, would have four intermodal connections records associated with it. Each intermodal connection at a freight terminal is defined by the pair of transport modes involved and by the type of cargo or commodity being transferred. If one of the modes is rail and multiple rail carriers serve the terminal, a separate intermodal connection record is created for each railroad. This is because each railroad operates over its own network of tracks.

The information contained in the intermodal connections table can be used to generate commodity-specific bimodal transfer links that join two or more modal subnetworks to create a multimodal network. In the conceptual model of intermodal terminals presented earlier, intermodal connectors or transfer links are notional or artificial links whose purpose is to represent the complex and often lengthy process of transferring cargo between two modes of transportation. This process may involve long-term storage, consolidation of small incoming shipments, subdivision into smaller outgoing shipments, repackaging, and reloading. At the discretion of the analyst, intermodal connectors may also represent modal access to the terminal, or the analyst may add specific terminal access links to each mode's network representation. Either way, there is no need to store intermodal transfer links in a file of link-type records. A computer program designed to build a connected multimodal network can use the information in the intermodal connections table to generate these links automatically. Even if one or more modal network databases are replaced or modified, the records in the intermodal connections table, along with the terminal location records in the terminal point file described above, can be used to connect the modal subnetworks.

Table 3-2. Format of the Intermodal Connections Attribute File
Field

No.
Field

Name

Field

Type
Field

Length
Start

Position
End

Position
Field

Description

1
RECTYPE
char
1
1
1
Record type (always 'T')
2
VERSION
char
2
2
3
Version number
3
REVISION
char
2
4
5
Revision number
4
MODDATE
num
8
6
13
Modification date (mmddyyyy)
5
FEATURID
char
10
14
23
Feature (terminal) ID
6
MODE1
char
1
24
24
First mode in the connection:

A = Air

H = Highway

P = Pipeline

R = Railroad

W = Waterway

7
MODE2
char
1
25
25
Second mode in the connection (same codes as for MODE1)
8
RRID
char
4
26
29
AAR reporting mark of rail carrier if railroad connection; otherwise, left blank
9
CARGO
num
3
30
32
Cargo type (codes shown in Table 3-3)
10
DIRECTN
num
1
33
33
Direction of intermodal transfer:

1 = one-way:

(MODE1 MODE2)

2 = two-way:

(MODE1 MODE2)

The intermodal connections table is a terminal attribute file. It corresponds to a type T or attribute record type file in the NTAD specification. Table 3-2 shows the layout of each intermodal connection record. Each record contains ten fields, the first four of which are the standard NTAD record header fields. The remaining six fields are described as follows:

Table 3-3. Classification of Cargo for the Intermodal Terminals Database
Cargo Code
Description
100 CONTAINERIZED CARGO: cargo shipped in various international and domestic containers and cargo shipped intermodally in highway trailers
200 BREAKBULK/UNITIZED CARGO: non-containerized cargo shipped in bags, barrels, boxes, bundles, cartons, crates, drums, and other types of unit loads or as individual items
210
FOOD PRODUCTS
220
LUMBER & WOOD: e.g., logs; lumber; mahogany; plywood; pulpwood; timber; and wood posts, poles, and piling
230
METAL PRODUCTS: e.g., iron and steel bars, beams, coils, pipe sections, plates, sheets, slabs, & tubes; aluminum and other non-ferrous metal products
240
MACHINERY: e.g., engines & turbines, farm machinery, construction & earth moving equipment, industry machinery, electric power generators, military equipment, & other heavy equipment
250
MOTOR VEHICLES: automobiles, buses, trucks, vans
260
PULP & PAPER PRODUCTS: e.g., fibreboard, paperboard, or pulpboard; linerboard; newsprint; paper rolls; recycled or waste paper; wallboard; & wood pulp
270
CLAY, CONCRETE, GLASS, & STONE BUILDING MATERIALS: e.g., clay bricks & blocks, concrete products, flat glass, gypsum board
280
LIVESTOCK
290
OTHER BREAKBULK CARGO: appliances, auto parts, cotton, electronics, rubber products, and other non-containerized goods not elsewhere classified
300 DRY BULK CARGO: loose, granular, or free-flowing dry cargo shipped in bulk rather than packaged, bundled, palletized, or unitized form
310
COAL
320
GRAIN: barley, corn, oats, rice, rye, sorghum, soybeans, wheat
330
METALLIC ORES & BULK METALS: misc. ferrous & non-ferrous ores, ferroalloys, pig iron, scrap metal
340
WOOD CHIPS
350
CEMENT
390
OTHER DRY BULK CARGO
391 CHEMICALS: e.g., fertilizers, plastic granules & pellets, resin powders, synthetic fibers
392 DRY EDIBLES: e.g., alfalfa pellets, citrus pellets, feeds, flour, meal, peanuts, raw sugar, seeds, starches

The cargo classification shown in Table 3-3 evolved as data on the location and intermodal connections of different types of intermodal facilities were added to the terminals database. It was necessary to invent this taxonomy because existing commodity classifications such as the Standard Transportation Commodity Classification (STCC), developed by the AAR in the early 1960s as a tariff mechanism, and the Standard Classification of Transportable Goods (SCTG), developed for use in the 1997 CFS, are too detailed for the purposes of the terminals database. In addition, existing commodity classifications tend to be based more on industry groups than on the physical characteristics of freight.

The cargo classification devised for the terminals database is hierarchical. The top level represents the four basic types of cargo: containerized, breakbulk, dry bulk, and liquid bulk. The second level, where appropriate, identifies important commodities or commodity groups. Metal products, for example, were singled out under the breakbulk category because several Class I railroads serve facilities for reloading steel coils or other steel products between truck and rail. Likewise, a category for pulp and paper products was included because of the significant number of paper distribution centers where paper products are transloaded between truck and rail. A third level was added in a few cases to further stratify some of the commodity groups. The resulting cargo classification shown in Table 3-3 captures the various degrees of cargo specialization exhibited by existing intermodal terminals. Thus, the existence of auto distribution centers resulted in a category for motor vehicles, coal export terminals and transloading facilities in a category for coal, grain elevators in a category for grain, Great Lakes ore docks in a category for metallic ores, cement terminals in a category for cement, and petroleum tank farms in a category for petroleum products. The taxonomy in Table 3-3 is about as detailed a classification of cargo as available data on intermodal terminals will generally support.

SUMMARY

In determining the contents and organization of the terminals database, several factors were considered. Because the terminals database is a component of the broader NTAD, the specification of the latter prescribes an overall organizational structure as well as some standard file formats on the terminals database. As a result, information on terminal location is placed in a file of point-type records, while non-topological attribute data are stored in one or more files of attribute-type records. Other factors considered in deciding what information to include in the terminals database centered around the need for independence from any particular modal network database, the possible uses of the terminals database, technical problems associated with measuring or specifying terminal costs, capacity, and performance, and the availability of data and other resources for building and maintaining the database. As a further design aid, a simple conceptual model of intermodal terminals was created to highlight the important features of these facilities and their connections with the various modal subnetworks.

The resulting structure of the initial intermodal terminals database is based on the relational data model. It consists of two relational tables or files. The terminal location table or point file contains information on the identity and geographic coordinates of each intermodal terminal. The intermodal connections table or attribute file describes the intermodal connections that can occur at each intermodal terminal. Each intermodal connection represents the ability to transfer cargo of a particular type or commodity group between a specified pair of modes at a terminal. Because an intermodal terminal may support multiple intermodal connections, it may have multiple records in the intermodal connections file. These two tables or files are linked by a common field specifying the terminal's FEATURID or unique identifier. They constitute what might be called the canonical version of the terminals database, since terminals of every type have a physical location and at least one intermodal connection. Additional files or tables of logically related attributes can be added to the canonical terminals database in the future as the need arises and as more or better data become available.

CHAPTER 4. BUILDING THE INTERMODAL TERMINALS DATABASE

The task of adding terminal location and intermodal connections records to the intermodal terminals database was not a trivial exercise. First of all, the fact that there are thousands of intermodal freight terminals that could be included in the database made it a time-consuming project. Secondly, because of the many different types of intermodal terminals, no single comprehensive source of information on these facilities has ever been compiled. Consequently, one of the first challenges to be overcome was locating directories, lists, and other published information covering each of the various kinds of intermodal facilities: TOFC/COFC terminals, auto ramps, truck-rail bulk transload facilities, truck-rail reload centers, waterway terminals, liquid bulk storage and transfer terminals, and so on. Thirdly, the reliability, currency, and level of detail of terminals data varies considerably between and, in some cases, within sources of data. As a result, it was fairly simple to pinpoint the location of some intermodal terminals, while others required a phone call to the terminal operator or, if that was not possible, an "educated guess" as to the exact location of the facility. Likewise, determining the intermodal connections at a terminal was trivial in some cases and exceedingly difficult in others. Thus, the process of populating the terminals database was not simply a matter of transcribing or transferring information from a printed source to an electronic file. Considerable knowledge of geography and the freight transportation system was required in order to detect errors or problems in the data and to make reasonable judgments when needed information was ambiguous, insufficient, or completely lacking.

The main purpose of this chapter is to describe how the initial version of the terminals database was built. The discussion focuses on the sources of information used to identify and locate intermodal terminals and determine their intermodal connections. The simplest way to do this is by type of terminal, since most directories and other sources of information on terminals are specific to one or a few terminal types. Each source also tends to present its own difficulties relative to locating terminals or determining their intermodal connections.

It should be noted that, at the time this report was being prepared, the initial version of the terminals database was by no means complete. Although the database contained records for over 2860 intermodal terminals as of October 1997, some terminal types were only partially represented. For example, the database contained nearly all known TOFC/COFC terminals, auto ramps, and terminals with connections to the coastal, intracoastal, Great Lakes, and inland river waterways. It also contained a large number of truck-rail bulk transload facilities, although there may be many more not in the database due to the lack of a comprehensive directory of these facilities. On the other hand, the database contained only a few public warehouses served by rail or truck-rail reload centers for lumber, steel, paper, or other breakbulk freight. Likewise, only grain elevators, cement terminals, petroleum tank farms, and liquid bulk storage and transfer terminals with waterway connections were included in the database. Consequently, the following discussion covers only the sources of information that had been used through October 1997. As work continues on building and maintaining the terminals database, new or improved sources of data may be uncovered and utilized.

GENERAL PROCEDURE

The process of building the terminals database involved three primary steps or tasks:

The first task was basically one of discovery. Its objective was to uncover the existence or presence of intermodal terminals within a locality. Unfortunately, one cannot simply consult the Yellow Pages of telephone directories for a comprehensive list of intermodal facilities as the term "intermodal terminals" is not a widely recognized business category, nor are intermodal terminals generally advertised as such. Many intermodal terminals are part of a larger facility such as a major rail yard or are buried within a complex of facilities. The transfer of freight between modes may be only one of many activities taking place at these locations. Consequently, one of the better ways to isolate intermodal terminals is to understand the kinds of places where intermodal transfers are likely to occur and to search for sources of information on the locations of these places.

The search for sources of information on the locations of intermodal terminals followed many avenues. A few sources such as The Official Intermodal Guide, The Official Railway Guide, and the Port Series reports of the U.S. Army Corps of Engineers were already well known. Links or references to other potential sources were revealed through a continuous monitoring of the freight transportation literature, including not only research reports and journal articles but also news publications, industry and trade magazines, and popular railroad and transportation periodicals. Other sources were identified by contacting government agencies with interests in a particular commodity such as grain or coal or in some other aspect of freight transportation, industry and trade organizations such as the AAR and the American Association of Port Authorities (AAPA), railroad equipment suppliers such as the TTX Company, persons in the railroad and trucking industries with experience in intermodalism, and researchers who have recently examined some aspect of intermodal transportation. Within the past two years, an ever increasing number of sources have appeared on the World Wide Web (WWW), and the various Web search mechanisms have also uncovered references to additional published sources. In general, a great deal of painstaking detective work was required to find available data on the locations of various kinds of intermodal terminals, and more than one source was discovered serendipitously. Most likely, there are still some important, but relatively obscure, sources of information on intermodal facilities waiting to be stumbled upon.

Once the presence or identity of an intermodal terminal had been discovered, its location had to be pinpointed for inclusion in the terminal location or point record. Rarely does the source indicate the exact longitude and latitude of a terminal (although the U.S. Army Corps of Engineers Port Series data come close by reporting the longitude and latitude of individual docks in some cases). In fact, many sources provide only limited or incomplete physical address information. Compounding the problem is the fact that many intermodal terminals do not have a physical mailing address simply because no mail is ever sent directly to the terminal. Often the reported street address for a terminal corresponds to the location of the terminal owner's or operator's offices and not to the actual location of the terminal facility itself.

Various map resources were utilized in an effort to locate each intermodal terminal as precisely as the available address information would allow. The primary paper map resources included street and highway maps obtained from the American Automobile Association, county road maps for a few States, the Atlas & Gazetteer series published by DeLorme Mapping Company, and available volumes of the Comprehensive Railroad Atlas of North America published by Steam Powered Publishing. Two sites on the WWW were also useful in locating the street addresses of some terminals: Excite Travel by City.Net at http://www.city.net/indexes/top_maps.dcg and Yahoo Maps at http://maps.yahoo.com/yahoo/. For locating waterway terminals, the aerial photographs accompanying the U.S. Army Corps of Engineers Port Series reports were an invaluable aid.

The quality and utility of these map resources varied considerably. The aerial photographs of port and waterway facilities were indispensable for locating waterway terminals. Street and highway maps and atlases were generally, but not uniformly, helpful in determining the general vicinity of a terminal if not its exact location. Their major weakness, with a few exceptions, was their tendency to either exclude railroad and other non-highway transportation facilities or to display these features in much less detail than streets and roads. For example, road maps and atlases typically do not delineate railroad yards, and many only show the right-of-way of mainline tracks, leaving out industrial spurs, branchlines, and other secondary tracks where certain types of truck-rail transfer facilities are often situated. In addition, most paper maps and other map resources do not indicate city block numbers. Knowing a terminal's street number often was not as helpful as knowing the nearest street intersection. Consequently, in a relatively small but significant number of cases, the only recourse was to phone the terminal operator and ask for directions to the facility.

After determining as closely as possible where the terminal is located, the next step was to assign a longitude and latitude. Unfortunately, some degree of arbitrariness in making this assignment cannot be avoided. Although the NTAD regards intermodal terminals as a point feature, they nevertheless occupy some amount of space. Intermodal terminals in fact come in assorted shapes and sizes. Their boundaries are often ill-defined, nor is there always a fence around them. Many truck-rail intermodal facilities occupy a portion or one side of a larger railroad facility such as a classification yard whose outlines may not appear on any map. Still other truck-rail transfer facilities are located along a siding or team track that likewise does not show up on any map. In the absence of any aerial photographs, drawings, or diagrams of the area, a considerable amount of judgment was often required in deciding where to fix the geographic coordinates of the terminal. Even when the boundaries of the terminal are plainly visible on a map or aerial photo, there is still some arbitrariness in assigning coordinates. Although the NTAD specification provides for six decimal places of precision in recording the latitude and longitude of a terminal (0.000001 degree of latitude is approximately 4.4 inches), precision is not the same as accuracy. In many cases, the hope was that the assigned coordinates placed the terminal within a mile of its actual location or at least in the proper location relative to nearby arterial streets and roads, waterways, or major geographic barriers.

A CD-ROM map database software package called MapExpert, published by DeLorme Mapping Company, was used to locate the position of intermodal terminals and measure their longitude and latitude. MapExpert displays map areas at 14 different levels of resolution. At the highest levels, it reveals all major and minor streets and roads, railroad lines, rivers and streams, shorelines, State and county boundaries, and various points of interest. At all times the software displays the longitude and latitude of the point directly under the cursor, reporting the coordinates at 0.01 of a second at the highest resolution or magnification. The procedure was to place the cursor over the point deemed to be the approximate center of the terminal, record the displayed values of longitude and latitude to the nearest second, and convert the values from degrees-minutes-seconds to decimal degrees. This information was then entered into a terminal point record along with the facility's name and its State FIPS code.

The final step in adding a terminal to the database was to determine its intermodal connections; that is, ascertain what types of cargo or commodities are interchanged between which modes of transportation at the terminal. For many kinds of terminals, such as TOFC/COFC facilities and auto ramps, this was a trivial matter. For waterway and some other types of terminals, however, the intermodal connections were not always so obvious. Information on the kinds of freight handled at a terminal was sometimes sketchy, ambiguous, or unavailable. Some of the primary data sources did not always clearly indicate what role, if any, railroads play in exchanging freight with other modes at a terminal, or which modes of transportation are involved in the transfer of a particular commodity or type of cargo. Whenever possible, supplementary or secondary information was used to help in resolving these ambiguities. In a few cases where available data were inconsistent or extremely vague and where considerable doubt existed about what actually took place at the terminal, the terminal operator was contacted by phone. The resulting conversations sometimes revealed a situation even more complex than suggested by the published information. In general, however, if the available information implied that a particular intermodal connection might be possible at a terminal, then it was assumed to exist and was recorded in the database.

SOURCES OF INFORMATION

For convenient reference, Table 4-1 summarizes the principal sources of information used to build the canonical terminals database. Most of these sources are not comprehensive; that is, most of them do not include all of the terminals of a particular type. However, they were highly useful either because they were the only known source of information available for a particular type of terminal or because they supplied additional or more recent information for a significant subset of terminals also covered by other sources. In addition, most of these sources are updated or reissued periodically, some on a regular basis such as annually and others irregularly or on an occasional basis. Note that some of the primary sources are located on the World Wide Web. None of these online sources existed when work began on the design of the terminals database in late 1994. Each of the principal sources is described in more detail under one of the sections below dealing with the type of terminal covered by the source.

Table 4-1. Principal Sources of Information on Intermodal Terminals
Source Terminal Types Comments
American Association of Port Authorities,

Seaports of the Americas: 1996 AAPA Directory

Waterway Lists mainly publicly owned terminals at deep water ports; limited data on types of cargo and rail connections
Arkansas Industrial Development Commission, Oklahoma Department of Commerce, Directory of Public and Private Ports and Terminals on the McClellan-Kerr Arkansas River Navigation System in Oklahoma and Arkansas Waterway Data include terminal owner, river mile, type of cargo, and rail connections; more recent than corresponding U.S. Army Corps of Engineers Port Series data; more useful when used in conjunction with Port Series reports
Association of American Railroads, Motor Vehicle Loading & Unloading Terminals, March 1995 Auto ramps Specifies ramp street address, serving railroad(s), and direction of transfer; includes marine auto terminals served by rail
Burlington Northern & Santa Fe Railway, http://www.bnr.com/ TOFC/COFC Gives street address of each BNSF-served TOFC/COFC terminal; includes sketch map showing terminal location relative to major streets and roads
Canadian Pacific Railway,

http://www.cprailway.com/

TOFC/COFC, truck-rail bulk transfer & breakbulk reload Gives mailing address of CP-served TOFC/COFC, bulk transfer, and lumber, paper, and steel reload facilities in U.S.
Emap Business Communications, Ltd., Containerisation International Yearbook, 1995 Waterway container Lists name and location of port terminals with container facilities; indicates presence of rail facilities
G. R. Leonard & Company, Leonard's Guide National Warehouse and Distribution Directory, 6th Edition, 1996/1997 Warehouses and distribution centers Data include facility owner, mailing address, products handled, and rail carrier (if any); not all facilities are intermodal
Independent Liquid Terminal Association, Bulk Liquid Terminals and Storage Facilities, 1995 Directory Bulk liquid Data include mailing address, modes served, and commodities handled; no indication of mode pairs; limited to facilities of ILTA members
Intermodal Association of North America, 1997 Rail Intermodal Terminal Directory TOFC/COFC Data include facility address and serving railroad(s); most comprehensive list of TOFC/COFC terminals
K-III Directory Corp., The Official Railway Guide, bimonthly publication TOFC/COFC, bulk transfer, grain elevators, lumber reload centers Data include facility address, commodities handled, and serving railroad(s); data not always reliable or current; more reliable TOFC/COFC data available from newer sources

Besides the primary sources listed in Table 4-1, a number of supplementary sources were also used. These sources can be grouped into the following categories:

Information from sources in the first two categories was used to corroborate data in the principal sources, resolve ambiguities in the primary data, update or confirm older information, or fill in some small gaps in the primary sources. The World Wide Web business directories such as BigBook (http://www.bigbook.com/) and Big Yellow (http://www1.bigyellow.com/) were often helpful in confirming the existence of a terminal or in determining a terminal's physical location. Many of the newest intermodal terminals were discovered and included in the database because of news items found in transportation-related news sources such as the Journal of Commerce Online, Traffic World, and various railroad and transportation periodicals. As was the case with the online primary sources, many of the online secondary sources appeared well after the terminals database project was underway.

Many of the intermodal terminals included in the initial version of the terminals database were found in more than one source of information. This was especially the case for terminals with waterway intermodal connections. Sometimes it was obvious when a previously discovered terminal was found in an additional source. In many cases, however, it was not immediately apparent, usually because one of the sources contained only partial or rudimentary information on intermodal connections. After several duplicate terminals were accidentally detected in the database, it became clear that the danger of inadvertently including the same terminal more than once in the database was not insignificant. A constant awareness of and alertness to this problem was necessary to prevent it from happening.

TOFC/COFC Terminals

TOFC/COFC terminals were among the first to be entered into the terminals database. When work on building the database began in early 1995, the only extensive source of information on this type of intermodal facility was The Official Railway Guide (ORG), a bimonthly publication of the K-III Directory Corporation. The ORG lists the company address and key contact personnel of each national, regional, and shortline rail carrier in the United States, Canada, and Mexico. For most of the national and regional carriers, it also includes other information such as route maps or profiles, interline connections, and whatever other information a particular railroad provides the publisher. Each issue is divided into six sections:

The focus of Section D differs for each issue. Information on TOFC/COFC facilities appears once or twice a year when the Intermodal Connection issue is released.

Each description of a TOFC/COFC facility in the ORG Intermodal Connection section indicates the city or town where the terminal is located, the name of the railroad that either owns or uses the facility, and the name and phone number of the terminal manager or some other contact person. Most, but not all, entries also give a street address or street location. Other information may also be provided, such as days and hours of operation, handling equipment, and number of parking spaces or acres of storage area. A terminal may have more than one entry if it is served by more than one railroad.

The ORG Intermodal Connection list of TOFC/COFC facilities has several significant limitations and problems. First of all, it does not include all TOFC/COFC terminals, particularly ones not owned by railroads. For example, it does not include terminals controlled by stacktrain service companies or any of the Triple Crown Services RoadRailer terminals. Also missing are some of the port-controlled TOFC/COFC facilities. Secondly, the list is not always current. Entries for new terminals often do not appear even a year after their opening, while entries for closed terminals sometimes continue to appear or even reappear more than a year later. A third problem is the inclusion of several spurious entries. The July/August 1997 ORG, for example, includes entries for the Hoosier Southern Railroad in Tell City, IN; the Belfast & Moosehead Lake Railroad in Unity, ME; the New Hampshire Northcoast in Ossipee, NH; the New York, Susquehanna & Western Railway in Rochelle Park, NJ; and the Pend Oreille Valley Railroad in Usk, WA. After contacting these railroads, it was learned that these carriers have never operated or served a circus ramp or any other TOFC/COFC facility at these locations.

Other more recent sources of information on TOFC/COFC terminals have superseded the ORG Intermodal Connection listings. The principal one is the Rail Intermodal Terminal Directory, first published in January 1997 by the Intermodal Association of North America (IANA). It includes nearly every TOFC/COFC terminal in operation in the United States and Canada in late 1996, including all Triple Crown Services RoadRailer terminals and facilities controlled by stacktrain service companies, port authorities, and other non-railroad entities. Each description of a terminal provides a street address and the name of the railroad owning or using the facility to aid in locating the terminal on a map. Other data include the name and phone number of a contact person, days and hours of operation, annual lift capacity, number of acres and parking spots, and the number of pieces of lift equipment by type. Since none of this information can be reproduced in any format without IANA's written permission, the IANA Directory was only used to locate each terminal on a map in order to assign a longitude and latitude.

Four Class I railroads now provide online information on their TOFC/COFC terminals. Burlington Northern & Santa Fe, Canadian Pacific, Norfolk Southern, and Union Pacific maintain World Wide Web pages giving street addresses, days and hours of operation, and telephone contacts for each terminal which they either own or serve. The BNSF and UP pages also include helpful sketch maps showing the location of each terminal in relation to major nearby streets and roads. The UP listings even give detailed driving directions and can be downloaded in the Adobe Acrobat PDF-file format.

Several recently opened TOFC/COFC terminals, such as the Kansas City Southern facility near Jackson, MS, and the RoadRailer terminal at Saginaw Yard north of Fort Worth, TX, were discovered in news articles appearing either in print or on the World Wide Web. Those same news sources also revealed a number of new TOFC/COFC facilities either being developed or in the planning stages as well as several existing facilities that have closed since the beginning of 1997.

Even though street addresses were given for most TOFC/COFC terminals, these facilities were not always easy to find on a map. Much depended on the quality of the map, including whether or not it displayed every street, indicated block numbers, and showed railroad tracks and yards. Sometimes clues to the location of a facility came from a magazine article. Fortunately, the street addresses were usually indicative of the terminal's general vicinity if not its exact physical location. Moreover, many TOFC/COFC facilities are located at or near rail yards, which also helped in finding them. Nevertheless, in some cases it was necessary to call the terminal manager to get detailed directions to the facility.

One issue which arose in connection with TOFC/COFC terminals concerns the so-called "near-dock" rail facilities, which are adjacent to and serve one or more port facilities. Among the many examples are the Ameriport terminal next to the Packer Avenue Marine Terminal at the Port of Philadelphia, the Intermodal Container Transfer Facility (ICTF) next to the Seagirt Marine Terminal at the Port of Baltimore, and the North and South Intermodal Yards at the Port of Tacoma. In addition to transferring containers and highway trailers between truck and rail, these facilities also handle containers to and from one or more nearby port terminals. In fact, some of these rail facilities are controlled by the ports rather than the railroads. The question was whether these facilities should be treated as separate TOFC/COFC terminals or whether they should be considered part of the adjacent waterway terminals. The decision was to include the rail facilities in the database as separate TOFC/COFC terminals with only a truck-rail intermodal connection. For the nearby marine terminals that use these rail facilities, however, bidirectional truck-water and rail-water intermodal connections for containerized cargo were entered into the intermodal connections file. Many marine container terminals have their own intermodal rail yards. These "on-dock" rail facilities were not regarded as separate TOFC/COFC terminals because they normally do not make truck-rail transfers and because their operations are an integral part of the marine terminal's. The distinction between "on-dock" and "near-dock" rail connections can be somewhat nebulous. The deciding factor is whether the sole purpose of the rail facility is to transfer containers between railcars and containerships or whether it also handles truck-rail transfers.

Auto Terminals

The primary source of information used to locate auto terminals was the March 1995 edition of AAR's Motor Vehicle Loading & Unloading Terminals directory. This publication lists all locations in the United States, Canada, and Mexico where motor vehicles are either loaded onto or unloaded off multilevel railcars. It includes auto ramps at assembly plants, auto distribution centers, and auto import and export facilities at port terminals. Each entry identifies the street address of the auto ramp, the serving rail carrier or carriers, and whether the ramp is used to load motor vehicles onto multilevel railcars, unload them, or both. Additional information includes the makes of vehicles handled at the ramp and the names and phone numbers of contractors involved in loading, unloading, haulaway, and inspection. The AAR auto terminals directory, however, does not provide any information on facility capacity such as amount of track space and vehicle storage space or throughput such as annual volume of vehicles.

Some auto terminals consist of more than one auto ramp. This was revealed by the fact that some auto ramps listed in the AAR directory had the same street address. In several other cases, two or more auto ramps were found next to or very close to each other with addresses located on either intersecting or adjacent parallel streets. Each ramp usually handled a different mix of vehicle makes. Some ramps within a cluster were used for unloading railcars while others were used for loading. Rather than include each auto ramp as a separate facility in the terminals database, clusters of two or more adjacent or nearby auto ramps were combined into a single auto terminal complex.

As noted above, the AAR directory of motor vehicle terminals includes import/export facilities connected with marine terminals. The AAR directory, however, does not indicate this explicitly. Clues that an auto ramp may be associated with a port facility came from a street address which placed the ramp at or close to a port complex along with an indication that the vehicles involved were either imports or exports or were foreign makes. Whenever it was suspected that an auto ramp or a cluster of auto ramps might be connected to a waterway terminal, the appropriate U.S. Army Corps of Engineers Port Series report, described later in this chapter, was consulted to determine the exact port facility involved. In fact, several auto terminals had already been discovered as waterway terminals were added to the database using U.S. Army Corps of Engineers data. This is just one of many examples where intermodal terminals of overlapping types were discovered in separate sources of information. As noted previously, a constant awareness of and alertness to this possibility was required to prevent the same terminal from being entered into the database more than once.

As was true for TOFC/COFC terminals, street addresses given in the AAR directory were indicative of the general location of auto terminals. Nevertheless, these facilities were often difficult to locate on available maps, because they tend to be located in outlying areas beyond the coverage of many city road maps. In many cases, the location had to be approximated using railroad atlases and small-scale topographic maps. In the case of Conrail-owned auto terminals, a handy supplemental source was the Conrail Premium Service Automotive Network Guide, which included sketches of the layout and boundaries of each terminal as well as colored aerial photographs of the facility and its surroundings. The aerial photographs accompanying the U.S. Army Corps of Engineers Port Series reports were especially helpful in locating auto terminals connected with port facilities. Given the rural or urban fringe nature of many auto terminals, it was also difficult to get adequate directions from terminal operators.

The AAR's motor vehicle terminals directory is the only extensive source of information on this type of intermodal terminal. Fortunately, it appears that this source will be updated on at least an occasional if not a regular basis. The AAR released an updated edition in October 1997 as this report was being written.

Truck-Rail Bulk Transfer Facilities

Truck-rail bulk transfer facilities were among the most difficult intermodal terminals to identify and locate. Although several principal sources covering this type of intermodal terminal were utilized, all of them were deficient in one or more ways. Consequently, a number of secondary sources were also examined, and many phone calls were made to terminal operators to request directions or to clarify published information.

One of the first sources used to find truck-rail bulk transfer facilities was The Official Railway Guide. As mentioned in the discussion of TOFC/COFC terminals, the focus of Section D of the ORG differs with each edition. The Bulk Connection edition usually appears twice a year. It contains two types of entries. One is for transfer facilities and the other is for bulk trucking services. Each facility entry gives the name of the facility owner or operator, a mailing address (which in many cases is not the physical location of the transfer facility), and the phone number and usually the name of a contact person. Other data include commodities handled, handling equipment, storage capacity, hours of operation, and serving railroads, although one or more of these items may be missing for an entry, and the amount of detail given on each of these items varies between entries. Under commodities handled, for example, the entry may only indicate that either dry bulk or liquid bulk or both types are transferred, or it may provide a list of specific commodities. Entries pertaining to trucking services include the name and address of the motor carrier, a contact name and phone number, insurance information, area served, equipment available, and many of the same data items listed under the facility profiles.

At first glance, the ORG Bulk Connection edition appeared to be an excellent source of information on truck-rail bulk transfer terminals. Subsequent use, however, revealed numerous problems. Transfer facility addresses are often given as Post Office boxes only. Where street addresses are reported, it was subsequently discovered that they often correspond to company headquarters, truck dispatching offices, or truck terminals rather than to the actual location of the transfer facility. Conversely, it was also subsequently discovered through phone calls to bulk motor carriers that some of the trucking services listings actually represent locations where truck-rail transfers occur. Another problem is inconsistent or ambiguous information. For example, while an entry might indicate that only dry bulk cargo is handled at a facility, it might also list some liquid bulk commodities as specific examples. Some transfer facilities have multiple listings with dissimilar information on certain data items. Attempts to contact some of the terminal operators revealed that a number of the transfer facilities listed in the ORG were either closed or were under the control of a different company. After several weeks of encountering these kinds of problems, it was abundantly clear that the information in the ORG Bulk Connection edition was too unreliable to use in building the terminals database.

Efforts to find an alternate source lead to the discovery of the trade publication Modern Bulk Transporter. Since 1985, the December issue of this monthly publication has included a directory of American and Canadian facilities where bulk commodities can be transferred between tank and dry bulk trucks and either the rail or water modes. Thus, the MBT directory includes truck-rail and truck-water bulk transfer facilities, although the vast majority of listings in the December 1996 directory belonged in the former group. Each facility profile may include some or all of the following data: name of the company - usually either a bulk motor carrier or an independent terminal operator - that either owns, operates, or serves the facility; mailing address; phone number of a contact person; products handled; serving rail carriers; number of railcar spots; whether truck-water transfers are possible; services and equipment available; methods used to transfer dry bulk commodities; and whether the facility is off limits to outside carriers. Eight types of bulk products are identified: acids, liquid chemicals, dry chemicals, asphalt, liquid foods, dry foods, dry plastics, and petroleum products. Fortunately, these commodity groups are compatible with the cargo classification developed for the terminals database as shown in Table 3-3.

The MBT directory is a more extensive and current source than the ORG Bulk Connection catalog. However, it too suffers from some of the same problems. Chief among these is the lack of specificity on the actual location of the transfer facility. Sometimes the only address information given is a street name without a block number, and other times only a Post Office box number is provided. As was the case with the ORG, street addresses in the MBT directory often represent the location of a motor carrier's truck terminal or dispatching office and not the location of the actual transfer facility. In fact, the transfer facility may be located in a different community. For example, a facility listed as being in Teutopolis, IL, is actually located in nearby Effingham. The MBT entry for this facility indicates that it is served by the Illinois Central Railroad. Since the IC has neither tracks nor trackage rights into Teutopolis, questions were immediately raised. A phone call to the facility operator - a bulk motor carrier - revealed that the transfer facility was located next to IC's mainline through Effingham, while the motor carrier's base terminal was located in Teutopolis.

Phone calls to the terminal operators were often necessary to locate the truck-rail bulk transfer facilities listed in the MBT directory. The resulting conversations revealed several other idiosyncrasies in the MBT listings. For example, a motor carrier might serve several transloading terminals in an urban area, but be listed only once in the directory. Conversely, a single terminal might be served by multiple motor carriers with an entry for each carrier appearing in the directory. Because address information was often vague or missing, it was not easy to detect the latter situation without contacting either each carrier or the terminal operator. Another data problem that surfaced during phone conversations with motor carriers involved the kinds of commodities transferred at a facility. In a number of cases, what the carrier reported over the phone differed from what was shown in the directory. Like the ORG, the MBT listings also included a few facilities that appeared to be closed or out of business, although not as many.

Several major bulk motor carriers provided marketing materials with information on truck-rail transfer facilities which they either own, operate, or serve. These included Bulkmatic Transport Company, Chemical Leaman Tank Lines, J. P. Noonan Transportation, and Truck-Rail Handling, Inc. The information these firms provided was a valuable supplement to the data contained in the MBT directory.

Until recently, railroads did little or no marketing of their bulk transfer services or facilities. Fortunately, there are signs that this is now changing. Three Class I railroads now provide on their World Wide Web sites some locational information on the bulk transfer facilities they either own or serve. The information provided by Canadian Pacific Railway is limited to the name of the terminal operator and a mailing address. Norfolk Southern, however, provides detailed information on each of its own Thoroughbred Bulk Terminals as well as the Independent Bulk Terminals it serves. Each facility profile includes data on the name of the firm operating the terminal, street location of the facility, phone number of a contact person, number of railcar spots, the serving rail carrier if not NS, available equipment and services, and products handled, including acids, liquid chemicals, dry chemicals, liquid edibles, dry edibles, petroleum, plastics, sand, soda ash, minerals, and other bulk. Each facility page also includes a sketch map showing the location of the facility relative to access roads and nearby major highways. As this report was being written, CSX Transportation also upgraded its WWW site with detailed information on each of its TransFlo (formerly known as BIDS) terminals. Each facility profile includes directions to the terminal from a nearby airport; types of products handled such as acids, asphalt, caustics, dry chemicals, dry edibles, liquid chemicals, liquid edibles, minerals, phosphates, and plastics; and types of transfer equipment available and ancillary services provided. It is hoped that other railroads will soon follow the example of NS and CSXT and display information about their bulk transloading facilities on the World Wide Web.

While the principal sources of data on truck-rail bulk transfer facilities usually identified the kinds of products or commodities handled at each terminal, they hardly ever indicated the direction of the transfer. Based on phone conversations with motor carriers and terminal operators as well as information contained in news items reporting on the opening of a new bulk transfer facility, it appears that transfers at these facilities are often one-way. In the absence of any information indicating directionality, however, it is difficult to guess accurately whether a particular commodity at a particular terminal is transferred from truck to rail or vice versa. Therefore, unless specific information on the direction of bulk transfers at a facility was available from some source, it was assumed that the intermodal connections were two-way.

Truck-Rail Reload Facilities

Railroads and trucking firms exchange breakbulk cargo at a wide variety of facilities. These include public warehouses and distribution centers as well as more specialized facilities dedicated to a particular product such as lumber, steel, or paper. As this report was being written, only a few truck-rail reload centers were included in the initial terminals database. However, several significant sources of information on these facilities were identified, and some experience was gained in utilizing these sources.

The Leonard's Guide National Warehouse and Distribution Directory is an extensive, although not a complete, listing of public warehouses and distribution facilities throughout the United States and Canada. The 1996/1997 volume represented the sixth edition of this publication. Each entry includes some or all of the following data items: name and address of the warehouse company, name and phone number of a contact person, warehouse classification, storage area, a description of the facility's structure, truck dock capacity, serving rail carriers if any, security and fire protection systems, insurance coverage, distribution services, areas served, products handled, available handling equipment, bank references, association memberships, year established, number of employees, and other information. The directory does not classify the products handled. It simply repeats whatever information the warehouse companies provided to the publisher. It does, however, distinguish warehouses that handle food products. Since some of the facilities listed in this directory do not have rail sidings, not all of them are intermodal terminals. However, a cursory perusal of the listings indicated that a large majority of the facilities are served by at least one rail carrier.

Although only limited use of this directory has been made so far, a few idiosyncrasies and shortcomings have been noticed. Once again, the street address may not always correspond to the location of the warehouse. This is often the case when the warehouse company operates multiple facilities within the urban area. There may, however, be only one entry for the company in the directory that summarizes information for all of the facilities. Fortunately, many of the listings refer to advertisements which often provide additional information, including the locations of individual warehouses owned or operated by the company within the locality. The directory also does not indicate which products handled at a warehouse are interchanged between truck and rail or the direction of the transfer. In the absence of any information from some other source, it can only be assumed that any of the products could be reloaded in either direction. Finally, despite the fact that the alphabetical listings of warehouses by State are spread over nearly 500 pages, the directory does not include all truck-rail reload facilities. In particular, it does not include some of the more specialized terminals devoted to the handling of specific products such as steel, lumber, and paper. Despite these problems, the Leonard's Guide National Warehouse and Distribution Directory appears to be a fairly reliable and extremely useful source of data on truck-rail breakbulk transfer facilities.

Besides the Leonard's Guide, only a few other sources of locational information on truck-rail reload centers have been found. Usually once a year, the ORG includes a Lumber Connection section, listing a few dozen truck-rail reload facilities. Data on each facility include company name, address (often only a Post Office box number), name and phone number of a contact person, commodities handled, handling equipment, storage space, security and fire protection systems, hours of operation, and serving railroads. In the National Carrier Services & Facilities section of the ORG, the only railroad providing locational information on its reload facilities is Conrail. It lists the street addresses of its Steelnet terminals, and also identifies its Paper Connection and Lumber Transfer Distribution terminals, but does not indicate their street location. On the World Wide Web, only the Canadian Pacific and CSX Transportation currently identify the reload facilities which they serve. Information on Canadian Pacific-served facilities is limited to the name of the facility operator and a mailing address (sometimes only a Post Office box number). Each listing also indicates whether the facility handles steel products or forest products (either lumber or paper products). On its Web site, CSX Transportation lists the names and locations of the paper and forest products distribution centers and reload facilities which it serves as well as its HiRail metals distribution terminals where steel coils, pipe, rods, slabs, sheets, beams, plates, and other metals products are transloaded between trucks and railcars. CSXT's WWW site also includes profiles of its TransFlo warehouse facilities. Each profile gives directions to the facility; indicates the kinds of products handled such as canned goods, grain products, and other foods, alcohol, and scrap paper; and specifies the number of track spots and truck doors, types of equipment, and other systems and services available at the facility. Other railroads mention on their Web pages that they serve lumber, steel, or paper distribution centers, but as of yet have not provided any additional information.

Liquid Bulk Terminals

The principal source of information on petroleum terminals is the Petroleum Terminal Encyclopedia (PTE) published periodically by Stalsby/Wilson Press. Each listing includes some or all of the following data: name of terminal operating company, name and phone number of the terminal manager or other contact person, mailing address, a description of the physical location of the terminal, days and hours of operation, inbound and outbound modes of transportation, total storage capacity, specific kinds of petroleum products stored at the terminal, and other miscellaneous information such as important features, physical limitations, and loading restrictions. For marine petroleum terminals, additional information includes high and low water depths, berth length, and location of the waterway.

The Petroleum Terminal Encyclopedia was used in conjunction with several other sources. First, the BigBook and Big Yellow business directories on the World Wide Web were used to generate independent lists of gasoline, fuel oil, and other petroleum products storage facilities in each State. These lists revealed some terminals not included in the PTE as well as some changes in terminal operating companies that have occurred since the publication of the PTE's Eighth Edition. Second, information on physical location was used in conjunction with various map resources to derive a longitude and latitude for each terminal. The address information itself was not stored in the terminals database, since there are no fields for this kind of data in the database files. Third, the information on modes of transportation was used to infer the intermodal connections at each terminal. Because petroleum products are usually transferred between modes indirectly via intermediate storage, it was assumed that an intermodal connection existed between each inbound mode and every other outbound mode. For marine petroleum terminals, modal information was also available from the U.S. Army Corps of Engineers Port Series reports.

For information on liquid bulk terminals that store and transfer other products besides petroleum, a major source was the Bulk Liquid Terminals and Storage Facilities directory published regularly by the Independent Liquid Terminals Association (ILTA). It contains information on the tank storage and pipeline facilities operated by ILTA member companies in the United States and in 21 other countries. Each facility profile identifies the terminal operating company and its mailing address, name and phone number of the terminal manager, modes of transportation serving the facility, number of storage tanks and total storage capacity, and commodities handled. Address information in many cases is insufficient to locate the terminal on a map. Oftentimes only a Post Office box number or a highway name is given instead of a complete address. Moreover, as is the case with so many other sources of data on intermodal terminals, the mailing address may refer to the location of the terminal operating company's business offices rather than the location of the terminal itself. Modal information generally does not identify the inbound and outbound modes, making it difficult to determine the intermodal connections. In some cases, additional information is provided on truck facilities and rail spurs from which it is possible to infer some if not all of the possible intermodal connections at the terminal. In contrast to the modal data, information on commodities handled is often quite detailed. The commodities, however, can be easily assigned to the broader cargo groups created for the terminals database. In summary, the ILTA Terminal Directory provided an extensive, though not a comprehensive, list of liquid bulk terminals. Supplemental sources, such as the U.S. Army Corps of Engineers Port Series reports, were needed because of the limitations mentioned above.

The only liquid bulk terminals included in the terminals database at the time this report was being written were those located along the coastal, Great Lakes, and inland waterways. They were mainly identified and located using information from U.S. Army Corps of Engineers Port Series reports described in the next section. The Petroleum Terminal Encyclopedia and the ILTA Terminal Directory have so far only been used to supplement the Port Series and other data sources on waterway terminals. They will become more important when liquid bulk terminals with no waterway connections are added to the terminals database.

Waterway Intermodal Terminals

The U.S. Army Corps of Engineers conducts extensive field surveys of all the docks located along the coastal, Great Lakes, and inland waterways. Each port or section of waterway is surveyed about once every ten years on average. The Navigation Data Center compiles the survey data into a series of reports known as the Port Series. Table 4-2 shows the 56 Port Seriesreports that were available as of October 1997, including the ports or stretches of inland waterway that each one covers and the most recent date of publication. Field data in a report are usually about one year older than the date of publication.

Table 4-2. U.S. Army Corps of Engineers Port Series Reports as of October 1997
Report

No.



Ports or Waterway Segments Covered
Publication

Date
1
Portland and Searsport, ME, and Portsmouth, NH
1985
3
Boston, MA
1994
4
Ports of southern New England (Bridgeport, CT; New Haven, CT; Connecticut River; New London, CT; Providence, RI; Fall River, MA; New Bedford, MA; Fairhaven, MA)
1994
5
New York, NY, and NJ, and ports on Long Island, NY
1988
6
Albany and ports on Hudson River, NY
1996
8
Philadelphia, PA; Camden, NJ; Wilmington, DE; and ports on Delaware River
1984
10
Baltimore, MD
1991
11
Ports of Hampton Roads, and ports on the James and York Rivers, VA
1993
12
Wilmington and Morehead City, NC
1997
13
Charleston and Georgetown, SC
1997
14
Savannah and Brunswick, GA
1990
15
Jacksonville and Fernandina Beach, FL
1997
16
Miami, Port Everglades, Palm Beach, and Port Canaveral, FL
1991
17
Ports of Tampa and Port Manatee, FL
1996
18
Mobile, AL
1986
19
Panama City and Pensacola, FL; Pascagoula and Gulfport, MS; and ports on the Apalachicola, Chattahoochee, and Flint Rivers
1986
20
New Orleans, LA
1990
20A
Mississippi River ports below and above New Orleans, LA
1991
21
Baton Rouge and Lake Charles, LA
1990
22
Port Arthur, Beaumont, and Orange, TX
1988
23
Galveston and Texas City, TX
1996
24
Houston, TX
1989
25
Corpus Christi, TX
1993
26
Freeport, Point Comfort/Port Lavaca, Brownsville, and ports along Gulf Intracoastal Waterway, TX
1991

A large portion of each Port Series report is devoted to a table describing each pier, wharf, and dock at the ports or along the waterways covered by the report. Each facility profile contains a combination of descriptive and statistical information, including the following items:

It is important to note that the Port Series data pertain to individual piers, wharves, and docks and not to intermodal terminals. Docking facilities and intermodal terminals are not synonymous. First of all, not every docking facility is involved in an intermodal exchange of freight. Some, for example, are only used to moor vessels for storage, fueling, or maintenance. Secondly, not every docking facility is associated with an intermodal terminal. A dock may be part of an oil refinery, chemical plant, steel mill, electric power generating station, lumber mill, seafood processing plant, sand and gravel quarry, coal mine, ready-mix concrete plant, or other manufacturing or mining establishment that receives supplies or ships products by water. In other words, the pier, wharf, or dock may be associated with a freight traffic generator and attractor rather than with an intermodal terminal. Thirdly, an intermodal terminal may have multiple docking facilities. The North Locust Point Marine Terminal at the Port of Baltimore, for example, has six docking facilities, according to Port Series Report No. 10. Conversely, multiple intermodal terminals may share the same docking facility. The Westway Trading Corporation bulk liquid terminal at the Port of Jacksonville, for example, uses one of the docks at the Jacksonville Port Authority's Talleyrand Marine Terminal. At Port Everglades, fifteen petroleum terminals have pipelines extending from five nearby docking facilities. In using the Port Series data to build the terminals database, it was therefore necessary to identify the intermodal terminals among the often hundreds of individual piers, wharves, and docks described in each of the reports.

Deciding whether a docking facility was part of an intermodal terminal was usually straightforward. In many cases it could be quickly determined from the description of the facility's purpose. Facilities not used to ship or receive any kind of cargo were immediately eliminated from consideration. If receipt or shipment of commodities was one of the purposes of the facility, further study was needed to determine whether the facility was associated with some kind of freight traffic generator or attractor. This entailed a careful examination of the complete facility description for references to electric power generating stations, refineries, chemical plants, quarries, mines, steel mills, various other manufacturing or processing plants, shipbuilding or ship repair yards, and offshore oil drilling facilities. Clues as to the nature of the docking facility could often be found in the facility name, owner, operator, purpose, description of mechanical handling facilities (which often described the operation of the docking facility in detail), railway connections, and supplementary remarks. If the facility profile clearly indicated or strongly implied that all cargo passing through the dock was produced by or intended for the use of an establishment located at or near the dock, then the facility was considered to be part of a freight traffic generator or attractor rather than an intermodal terminal.

Once a dock was determined to be connected with an intermodal terminal, the next step was to identify the terminal itself and any other piers, wharves, or docks associated with it. This was usually a simple matter, since the name of the docking facility nearly always specified the name of the terminal. Moreover, in most cases involving a terminal with multiple docking facilities, the docks were contiguous, and their facility profiles appeared next to each other in the Port Series report.

The possibility that the dock served more than one terminal also had to be considered. In particular, docks with either multiple owners or operators were always examined more closely to determine if more than one terminal was involved. In some cases, the description of the dock's purpose provided a clue. It might indicate, for example, that one operator received or shipped a certain commodity such as bulk cement, while another operator received or shipped a different commodity such as vegetable oil, thereby suggesting the possible existence of a cement terminal and a liquid bulk terminal. Further evidence of multiple terminals could also be gathered from the description of mechanical handling facilities, railway connections, and any supplementary remarks. These data items might reveal, for example, the presence of pipelines extending from the dock to a set of liquid bulk storage tanks as well as pneumatic pipelines connecting the dock with several bulk cement storage silos. Other suitable sources such as the Petroleum Terminal Encyclopedia and the ILTA Terminal Directory were checked to see whether they included a terminal at or near the location of the dock. In this manner, it was usually possible to identify each of the intermodal terminals connected with a particular pier, wharf, or dock in the Port Series data.

Once the terminal was identified, the task of locating it in MapExpert and determining its longitude and latitude was greatly facilitated by the large aerial photographs of port areas that accompany each Port Series report. Each pier, wharf, and dock has a map reference number which is shown on an aerial photograph. The aerial photos also have labels for access roads, nearby highways and rail lines, bridges, and certain other landmarks which make it easier to match the photos with paper maps and MapExpert's digitized maps. Port Series reports issued since 1992 also indicate the longitude and latitude of each pier, wharf, and dock. These coordinates were plotted in MapExpert as a further guide to the location of the terminal. Waterway terminals in urban areas were generally much easier to locate using these devices than were terminals located in rural settings, because of the limited number of streets and other landmarks in the latter locations. Nevertheless, as a result of the aerial photos, the docking facility coordinates, and detailed descriptions of facility location included in the Port Series data, the ability to locate waterway terminals accurately was much better than it was for any other type of intermodal terminal.

Determining the intermodal connections at a waterway terminal was often a greater challenge than determining the terminal's coordinates. The description of the purpose for which the dock is used clearly indicated which commodities arrive at the terminal by water and which commodities leave by that mode. The main problem was in deciding what role trucks and especially railroads played in hauling cargo to and from the waterway terminal.

For some types of waterway terminals, the intermodal connections were fairly easy to ascertain using data from the Port Series and, in some cases, from other sources. Grain elevators are one example. Each Port Series report contains a separate table of waterfront grain elevators showing for each terminal which modes are unloaded and which are loaded. It was assumed that a storage transfer connection existed between each inbound mode and each different outbound mode at a grain elevator. If the waterway terminal was a liquid bulk storage facility, either the Petroleum Terminal Encyclopedia or the ILTA Terminal Directory was consulted for additional modal information. The former was particularly useful because it specified the inbound and outbound modes. In addition, the descriptions of railway connections and mechanical handling facilities in the Port Series dock profiles were examined for references to tank car loading or unloading racks. Major coal, ore, and other bulk transloading terminals were also easy to spot in the Port Series dock profiles. These are primarily if not exclusively rail-water or water-rail bulk transfer operations. The descriptions of railway connections and mechanical handling facilities at these terminals included detailed information on railcar space, rotary car dumpers, undertrack pits for railcar dumping, stackers, reclaimers, electric belt conveyors, barge or vessel loading towers, railcar loading spouts, and other intermodal facilities and equipment. Major marine container terminals with access to either on-dock or near-dock rail facilities were also relatively easy to identify, either from the Port Series descriptions of their docking facilities or from other sources of information on container terminals.

For other kinds of waterway terminals, particularly the ones which handle a variety of bulk and breakbulk commodities, the problem of determining the role of trucks and railroads was often quite difficult. Several items in the profiles of all the docking facilities associated with a terminal were carefully studied for clues. The first item to be examined was always the description of railway connections. If none of the docks associated with the terminal had any nearby railroad facilities either on the apron, in or next to the transit sheds or other enclosed storage facilities, or in the open storage areas, then the question of rail intermodal connections at the terminal was quickly answered. If, however, rail facilities were present at or near one or more of the docks serving a terminal, then considerably more effort was usually required to determine which commodities were handled by rail and whether the rail carrier or carriers brought the commodities to the terminal or hauled them away. The descriptions of railway connections and mechanical handling facilities as well as any supplementary remarks were carefully reviewed for references to truck and railcar loading and unloading equipment or facilities. For example, the docking facility profile might refer to railcar loading spouts at a bulk cement storage silo, undertrack pits for dumping the contents of railcars, truck receiving or loading hoppers, railroad tracks serving an open storage area dedicated to a particular kind of bulk commodity, or railroad tracks serving a transit shed where steel, paper, or other breakbulk commodities were stored for either consolidation or local distribution.

The amount of truck and rail information included in each Port Series facility profile varies considerably. To the extent possible, other sources were consulted for additional modal information, and in some cases, terminal operators were contacted by phone. Several things were learned from these informal telephone interviews. First, just because a waterway terminal is served by rail does not necessarily mean that there are any rail-water intermodal connections. Some waterway terminals are primarily local distribution centers for one or more kinds of commodity. A portion of their incoming cargo arrives by water and a portion by rail. The cargo is then distributed locally from the terminal by truck. Secondly, the rail and water modes may not handle the same kinds of commodities at a waterway terminal. For example, the water carrier may unload one kind of commodity while the rail carrier unloads another. Finally, a waterway terminal may ship and receive a wide variety of bulk and breakbulk commodities, but the serving railroad may haul only one or two commodities to or from the terminal. Rarely do the dock profiles in the Port Series reports reveal these situations explicitly. Ultimately, a considerable amount of "engineering judgment" had to be applied in many cases in order to specify the intermodal connections. As a result, the terminals database may contain a few too many rail-water connections and not enough truck-rail connections at waterway terminals.

The Port Series reports cover nearly all stretches of navigable waterways and channels and document nearly every port facility in the United States. The biggest limitation of this source is the considerable age of some of the reports in the series. As Table 4-2 indicates, a few reports have not been revised since 1984 or 1985, and of the 56 reports in the series, 17 have not been updated since 1990, and 34 contain data that was collected before 1993. Consequently, other sources of information on waterway terminals were utilized as a supplement to or as a substitute for some of the Port Series reports. Although none of these additional sources are as comprehensive as the Port Series set of publications, nevertheless, some of them cover a significant subset of waterway terminals and are therefore summarized in Table 4-1. The following paragraphs briefly describe these sources and how they were used in building the terminals database.

The Directory of Public and Private Ports and Terminals on the McClellan-Kerr Arkansas River Navigation System, published jointly by the Oklahoma Department of Commerce and the Arkansas Industrial Development Commission, was used to update the portion of Port Series Report No. 68 covering dock facilities on the Arkansas and Verdigris Rivers. Although undated, it was believed to be very recent. The introduction indicated that this was the first edition of this directory, suggesting that it may be revised on occasion in the future. The directory provided the following information on each terminal: name and mailing address of the terminal owner or operator; location on the river; type of freight handled if any as well as other dock services such as towboat refueling, barge cleaning and repairs, and barge fleeting; mechanical appliances such as cranes, belt conveyors, pipelines, forklifts, and lift trucks; storage facilities; rail connections; and additional remarks. Comparison of this source with Port Series Report No. 68 revealed very few new terminals. Most of the changes that had occurred since the last U.S. Army Corps of Engineers survey involved dock closings and new owners or operators of existing facilities. A few changes in the kinds of commodities handled and in the availability of rail service were also noted.

In a similar manner, the Directory of Terminals on the Tennessee River Waterway, published in June 1996 by the Tennessee River Valley Association, was used to update U.S. Army Corps of Engineers data in Port Series Report No. 64. Data provided on each terminal include: name and mailing address of the terminal operator, location on the river, public or private ownership, public or private use, commodities handled if any, storage facilities, serving rail carriers if any, and the nearest highway. Commodities are specified according to the following categories: chemicals; coal and coke; forest products; grain and grain products; iron and steel; petroleum products; rock, sand, gravel, and products; and general commodities. The directory proved to be a valuable resource as it revealed a number of changes since the U.S. Army Corps of Engineers' survey of the Tennessee River in 1985. There were several new terminals, a few terminal closings, and some new operators of existing terminals. Most of the changes were concentrated in Paducah, KY; Decatur, AL; and Knoxville, TN. Since the June 1996 edition replaces an earlier issue, it appears that this source may again be updated at some time in the future.

The Inland River Guide is an annual publication of The Waterways Journal, Inc. Among its many sections is a directory of public use terminals on the Mississippi and Ohio Rivers and their navigable tributaries. Each entry provides the following information: name of the terminal; river milepost; name, address, and phone number of the terminal operator or provider of stevedoring services; products handled; storage space; cargo handling equipment; number of barges that can be handled; days and hours of operation; nearby rail carriers; availability of switch boats; and other services available. Interspersed with the terminal profiles are advertisements by terminal operators that sometimes provide additional or more detailed information. The IRG has several limitations that preclude it from being used as a stand-alone source. It excludes intermodal terminals intended for private use only and also leaves out many public use terminals. It is therefore not a complete directory of intermodal terminals on the inland waterways. The degree of specificity of the information on products handled varies from a detailed list of commodities to such broad claims as "all commodities" or "all cargoes". The fact that a railroad is listed does not necessarily mean that there are rail facilities at the terminal; rather, it may only mean that rail service exists within the terminal's market or service area. Nevertheless, the IRG was useful in verifying and updating some of the older Port Series reports covering the inland waterways. As a result of its use, several relatively new terminals were identified, and a few other problems in the Port Series data were resolved.

Seaports of the Americas is a directory of ports belonging to the American Association of Port Authorities (AAPA). Each port profile lists the commissioners and staff, provides some overall statistics on annual tonnage, and identifies the primary inbound and outbound cargoes at the port. From the standpoint of the terminals database, the most important piece of information was a listing of the port's principal public use terminals and facilities. These listings are not always complete. Some of them, for example, only include publicly owned terminals. Besides the name of the terminal, information may include the name of the operator or provider of stevedoring services, length and depth of berthing space, types of cargo, availability of rail service, storage facilities, and cargo handling equipment. The amount of detail provided is generally much less than that found in the Port Series reports. In particular, information on types of cargo handled and available rail facilities is usually minimal or sketchy. The main value of this source was in updating some of the older Port Series reports. The AAPA Directory was especially useful for identifying terminals that had opened since the last survey of the port by the U.S. Army Corps of Engineers. Where extensive changes had occurred, the port staff was contacted and asked to send more detailed information.

One other useful source of published information on waterway terminals was the Containerisation International Yearbook, produced annually by Emap Business Communications, Ltd. This publication provides information on all of the major container-handling terminals not only in the United States but throughout the world. Each entry contains detailed information on a particular marine container terminal, including its name and the name of the terminal operating company; the terminal's street address; the number, length, and depth of container berths; number, types, and lift capacity of container cranes; the ocean carriers that directly call at the terminal; container storage space; yard equipment such as front-handlers, yard chassis, and yard tractors; container freight station facilities; computer systems; rail facilities; and days and hours of operation. In some cases, statistics on annual volumes of loaded and empty containers are also provided. This source was primarily used to update some of the older Port Series reports and to ensure that all of the important marine container terminals had been identified and were included in the terminals database. It also helped in some cases in determining whether a terminal had on-dock rail service or direct access to nearby rail facilities.

Besides the above principal sources, many other supplementary sources of waterway terminals information were utilized. Each of these was connected with a specific port. These additional sources involved both online information and printed materials.

Many port authorities now have home pages on the World Wide Web. The better Web sites provide descriptive information on the major terminals located at the port. This information was often extremely useful in updating Port Series data, especially for identifying new terminals and determining intermodal connections.

In comparing the data in some of the Port Series reports with online information as well as information in some of the other principal sources described above, it became clear that extensive developments and changes had occurred at a number of ports since they were last surveyed by the U.S. Army Corps of Engineers. The staff at these ports were contacted and asked to provide more recent information on terminals and other port facilities. Without exception, the contacted ports readily responded to these requests by sending voluminous folders or packets of printed materials including maps, port directories, newsletters, magazine articles, annual reports, and other marketing materials. The following port authorities supplied information:

The information provided by these port authorities was invaluable in identifying new intermodal terminals, determining the intermodal connections at these facilities, and improving the accuracy and currency of the terminals database.

CHAPTER 5. MAINTAINING THE INTERMODAL TERMINALS DATABASE

As of the beginning of October 1997, the terminals database point file contained records identifying and locating 2865 intermodal freight terminals. The associated attribute file contained 9036 records identifying individual intermodal connections at these terminals. Despite these large numbers, the terminals database was only partially completed. It included virtually all TOFC/COFC, auto, and coastal, Great Lakes, and inland waterway terminals. It also included a large number of truck-rail bulk transfer facilities and a few truck-rail reload facilities for breakbulk cargo. Still to be added were the remaining truck-rail bulk transfer facilities, warehouses and distribution centers with both truck and rail docks, air cargo terminals, grain elevators, and petroleum and other liquid bulk storage terminals, except those with waterway connections. The complete intermodal freight terminals database could easily be two to three times larger than its current size. Given the considerable amount of time and effort required to build the initial version, serious consideration must be given to the problem of keeping a database of this size up-to-date in a reasonable and cost-effective manner.

Maintaining the terminals database involves correcting errors, adding records for new terminals, removing or earmarking records of closed terminals, and modifying the records of existing terminals as their intermodal connections change. The amount of effort required to keep the database current depends on the volatility of the data, which in turn depends on how much the intermodal infrastructure changes over a period of time. If significant numbers of intermodal freight terminals are constantly opening or closing, then frequent revisions of the database may be necessary. If, on the other hand, relatively few terminals appear or disappear in a year's time, then limited maintenance may be all that is needed. In either case, some mechanism for keeping tabs on intermodal developments is needed to get a sense of how quickly and in what ways the intermodal infrastructure is changing.

This chapter is divided into two parts. The first section attempts to provide some insight into the current pace of change in the intermodal infrastructure. It lists some of the more recent terminal openings and closings as well as those planned within the next few years. Given these prospects for change, the second section offers a cost-effective strategy for updating the terminals database in a timely manner. The key principles behind the proposed maintenance strategy involve keeping abreast of new developments and prioritizing updates to the database.

RECENT AND PLANNED DEVELOPMENTS

It is difficult to say whether the number of intermodal freight terminals is increasing, decreasing, or remaining fairly constant or whether the rate of change is accelerating or decelerating. We do know that the number of TOFC/COFC terminals has dropped considerably during the past two or three decades, from 1500 or more in the 1970s to less than 250 by 1997. The numerous small circus-loading ramps have largely been abandoned in favor of a relatively small number of large to medium-sized TOFC/COFC hub centers. Is this trend continuing or has it run its course? Have similar trends occurred for other types of intermodal terminals or are they about to occur? For example, are there fewer but larger liquid bulk terminals today than there were five, ten, or twenty years ago? Will smaller grain elevators be abandoned in favor of larger ones with greater railcar-loading capacity in order to take advantage of lower rail rates for 100-car unit grain trains? Or is the opposite trend occurring for some types of intermodal terminals? For example, is the number of truck-rail bulk transfer facilities increasing? Are more warehouses and distribution centers providing truck-rail reload facilities? These are interesting and important questions that could be answered more easily over a period of time with a well-maintained intermodal terminals database. In the meantime, the following discussion shows that changes in the number of certain types of intermodal terminals are occurring and are likely to continue to occur into the foreseeable future.

Nearly a dozen TOFC/COFC terminals have been closed in the last twelve months because of low utilization, the UP-SP merger, and other reasons. In the latter part of 1996, BNSF closed its ramp in Missoula, MT, which was handling only six to eight intermodal railcars of business per day. More recently, BNSF closed its terminals in Galesburg, IL, and Springfield, MO, because of low and declining traffic volumes. BNSF also closed its off-line TOFC/COFC ramp at the Paducah & Louisville Railway (PAL) Kentucky Street Yard in Louisville, KY, in favor of Norfolk Southern's Louisville terminal. BNSF had been using the PAL as an interline carrier to reach Louisville, but with the change of terminals, BNSF and NS began offering joint intermodal service between Louisville and the West Coast. Union Pacific recently had a TOFC/COFC facility in North Las Vegas, NV, but this terminal does not appear on UP's World Wide Web site directory of TOFC/COFC facilities, implying that it has been closed. In May 1997, UP closed the TOFC/COFC terminal and the classification yard at Roseville, CA, in order to begin a $125 million rebuilding project at the former Southern Pacific facility. All of the other recent TOFC/COFC ramp closures have been a consequence of Union Pacific's acquisition of the Southern Pacific railroad in 1996. UP closed SP's facility in Pine Bluff, AR, and shifted its traffic to UP's North Little Rock terminal. In Denver, SP's North Yard ramp was closed on July 1, 1997, in favor of UP's 36th Street Yard facility. SP's intermodal traffic in East St. Louis, IL, was switched to UP's terminal in nearby Dupo, and its TOFC/COFC facility in Kansas City, KS, was shut down in favor of UP's Neff Yard ramp in Kansas City, MO. The opposite appears to have occurred in the Reno-Sparks, NV, area where SP's TOFC/COFC terminal in Sparks is still being used while UP's facility in Reno has been closed. In the Chicago area, SP used Illinois Central's Moyers Intermodal Terminal in Harvey, IL. Most of that business has now been shifted to UP's Yard Center TOFC/COFC facility in nearby Dolton, IL. UP also planned to transfer SP's Forest Hill TOFC/COFC operations in Chicago to UP's Canal Street facility.

Two other outcomes of the UP-SP merger have resulted in the transfer of TOFC/COFC facilities to another railroad. In the New Orleans area, UP operated a TOFC/COFC terminal in Westwego before the merger, while SP had a facility in nearby Avondale. As part of the huge trackage rights agreement UP negotiated with BNSF to obtain the latter's approval of the merger, UP sold 192 miles of SP trackage between Iowa Junction, LA, and Westwego to BNSF along with UP's Westwego TOFC/COFC facility. BNSF moved into the Westwego terminal in February 1997 while UP shifted its New Orleans TOFC/COFC operations to the former SP facility in Avondale. BNSF also obtained trackage rights on SP's line between Denver and Salt Lake City as well as the right to use SP's Roper Yard TOFC/COFC terminal in the latter city. BNSF began providing TOFC/COFC service to Salt Lake City in February 1997, while UP shifted SP's TOFC/COFC traffic to UP's Salt Lake City North Yard facility.

Offsetting some of the recent closures of TOFC/COFC facilities are a few new terminals. Conrail opened a new TOFC/COFC facility in October 1996 at Pitcairn Yard in Wall, PA, east of Pittsburgh. This terminal replaced Conrail's Island Avenue facility within the city of Pittsburgh. On November 8, 1996, Guilford Rail System opened a new TOFC/COFC terminal in Waterville, ME. This facility is part of a new service called DownEast Express, a joint venture of Guilford Rail System and Conrail designed to provide intermodal service connecting Maine industries with major markets in Atlanta, Chicago, Kansas City, and St. Louis. In December 1996, Kansas City Southern opened a new facility near Jackson, MS. More recently, CSX Intermodal opened a new doublestack terminal at Leewood Yard in Memphis, TN. This facility, dedicated to handling containers for Mitsui O.S.K. Lines, will be expanded from five to eight doublestack car spots. In September 1997, Triple Crown Services, Inc., the intermodal company jointly owned by Norfolk Southern and Conrail, opened a new RoadRailer terminal at Saginaw Yard north of Fort Worth, TX. BNSF operates the RoadRailer trains between the new facility and a connection with Norfolk Southern at Kansas City, MO. The newest TOFC/COFC terminal is located in Decatur, IL, where Norfolk Southern just recently began operating direct doublestack train service to the eastern ports of Norfolk, VA, Charleston, SC, and Jacksonville, FL, as well as connecting service to all West Coast ports and eastern Canadian ports. The Decatur container terminal is operated by Pacific Rail Service.

In addition to these four new TOFC/COFC terminals, a few more are currently under construction, and several others are being planned or given serious consideration. Iowa Interstate Railroad (IAIS) redesigned its yard at Council Bluffs, IA, to make room for a TOFC/COFC facility. IAIS currently uses UP's terminal across the Missouri River in Omaha, NE. On April 1, 1997, UP began constructing a new TOFC/COFC terminal at Ebony, AR, near West Memphis. Scheduled to open in September 1998, it will replace UP's crowded facility at Sargent Yard across the Mississippi River in Memphis. UP has also announced plans to build new TOFC/COFC terminals just east of Reno, NV; in Texarkana, AR/TX, to serve the northeast Texas area; and in the vicinity of West Colton Yard in the Los Angeles area. The last of these, known as the Inland Empire terminal, would replace the former SP facility in the City of Industry. UP is also interested in building a new TOFC/COFC terminal next to its yard in West Chicago, IL, on land currently leased by the Du Page County Airport Authority. In the Bronx, NY, the Harlem River Intermodal Yard is currently being constructed on a 77-acre site. This project has become controversial because of a proposal to divert about 50 of these acres for a paper recycling plant. Two proposed TOFC/COFC terminals have also generated some controversy of the "not in my backyard" variety. One is in Fairburn, GA, about 20 miles southwest of Atlanta, where CSX Intermodal planned to build a 500-acre facility that would take some of the pressure off its existing terminal at Hulsey Yard near downtown Atlanta. In response to opposition from local residents, CSXI shifted the location of the facility and scaled back its size, but that has not quelled all of the protests. The other proposed terminal arousing local concern is one which Kansas City Southern plans to construct along with an industrial park on a 550-acre site next to Lavon Lake, just east of Lavon Junction, near Wylie, TX. KCS wants to build this facility to replace its small, four-track, stub-end terminal at Zacha Junction in Dallas. One other planned TOFC/COFC terminal is the Sweetwater Lift on Norfolk Southern's line near Austell, GA. NS announced plans for this second TOFC/COFC yard in the Atlanta area back in December 1993 when it acquired 800 acres of land for the facility. However, no word on the terminal's progress or its status has been issued recently.

New TOFC/COFC terminals may also appear if the Surface Transportation Board (STB) approves the partitioning of Conrail between CSX Transportation and Norfolk Southern. Both railroads have touted single-carrier intermodal service between the Northeast, Midwest, and Southeast as one of the important benefits of the proposed double merger. NS has indicated a desire to operate Triple Crown Service RoadRailer trains over Amtrak's Northeast Corridor. The carrier proposes to build a new RoadRailer terminal in Bellevue, OH, to replace an existing facility in Crestline, OH, as well as new RoadRailer terminals in Baltimore, MD; Charlotte, NC; and Morrisville, PA. CSX Intermodal plans to close its Snyder Avenue TOFC/COFC terminal in Philadelphia and replace it with a new facility at Greenwich Yard. In Chicago, CSXI is considering a major new TOFC/COFC terminal at the abandoned 59th Street Yard that once belonged to the Pennsylvania Railroad. Both railroads have also announced plans to increase the capacity of some of the Conrail TOFC/COFC facilities which they will acquire as well as some of their own existing terminals [WOLF97A; WOLF97B].

Several new auto terminals are under construction or being planned. BNSF recently began construction of a new auto terminal in National City, CA, to handle imports of Honda vehicles. The facility is expected to accommodate 100,000 automobiles per year. Union Pacific is currently building new vehicle distribution terminals in San Antonio and Amarillo, TX, and in Santa Rosa, NM. It is also expanding existing vehicle terminals in Denver, CO; Galena Park, TX, near Houston; Las Vegas, NV; Mesquite, TX, near Dallas; Phoenix, AZ; and Tulsa, OK. During August 1997, Norfolk Southern began operating its new vehicle loading yard in Vance, AL. The facility was built to serve a new Mercedes-Benz plant that produces sport utility vehicles. A month later NS completed construction of a new 38-acre auto distribution facility in Jacksonville, FL, with track space for 36 multilevel autorack cars and ground space for 2515 vehicles. By the end of 1997, NS plans to complete another auto distribution terminal in Titusville, FL, as well as the four auto mixing centers which it is building for Ford Motor Company in Chicago; Fostoria, OH; Kansas City; and Shelbyville, KY. The latter facilities are designed to receive new vehicles from Ford assembly plants, sort the vehicles by destination, and ship them to their destinations in dedicated unit trains. NS also plans to build new auto terminals in the Baltimore and Philadelphia areas if its proposal to acquire parts of Conrail is approved by the STB.

Various transportation news sources have cited several new or planned truck-rail bulk transfer facilities. In June 1997, Canadian National Railway opened a new CargoFlo bulk commodities distribution center at the site of its former TOFC/COFC terminal in Chicago. The 35-acre facility handles a variety of commodities including dry and liquid chemicals and food products. Consisting of eight tracks that can accommodate 120 railcars, it is operated by Chemical Leaman Tank Lines. Canadian Pacific Railway planned to build a bulk commodities facility next to its TOFC/COFC terminal at its former Shoreham Yard car shops in Minneapolis, MN. The facility was scheduled to open in late 1997 for limited operations. Norfolk Southern appears to be particularly active in the area of truck-rail bulk transfer [NSPP97]. The railroad is expanding its network of Thoroughbred Bulk Transfer terminals with new facilities under construction in Charlotte, NC, and Miami, FL, the latter on the Florida East Coast Railway. Both terminals are scheduled to open during 1998. In 1997, NS opened two new rail-to-truck coal transfer terminals in South Carolina. The Columbia facility will transfer 500,000 tons of coal per week for local delivery by truck. The Branchville facility was created to serve South Carolina Electric & Gas Company's Cope plant. In September 1997, NS began hauling liquid corn syrup and dry bulk sugar to a new rail-to-truck transfer facility owned by Cargill, Inc, in Chattanooga, TN. The products are transferred to trucks for delivery to customers in northern Alabama, northwest Georgia, and Tennessee. NS, however, recently lost an Independent Bulk Transfer terminal when Shenandoah Bulk Service Corporation ceased operations at its Front Royal, VA, facility.

News about other types of truck-rail transfer facilities is relatively scarce, but several recent news items suggest there may be considerable activity in this area as well. CSX Transportation recently opened a new log loading facility at its railyard in Jackson, KY. Operated by Frederick & Lewis Enterprises, the facility ships loads of rough cut logs transferred from trucks to railcars and receives carloads of lumber to be distributed by truck to area businesses. In Norfolk, VA, Southern Commodities Services, Inc., opened a new warehouse with intermodal capabilities. The 150,000-square foot facility, located in the Norfolk Industrial Park, has 12 truck docks and seven rail doors for its spur on the Norfolk Southern Railway. Once again, NS appears to be particularly active in developing and serving truck-rail reload centers [NSPP97]. In September 1997, the railroad began serving a new paper distribution center built by Verst Group Logistics in Richwood, KY, near Cincinnati. The facility is a 253,000-square foot warehouse that receives paper and paper products by rail and distributes them throughout the Cincinnati and northern Kentucky area by truck. NS expects to serve a new metals distribution facility currently being built by Western Intermodal outside Charlotte, NC. The fully enclosed facility will have 100,000 square feet of dock space for trucks and railcars, and its overhead cranes and forklifts will allow for either direct transfer between truck and rail or indirect transfer through intermediate storage.

Recent efforts by the large western railroads to encourage shipment of grain in longer unit trains may soon have an effect on numerous small and medium-sized grain elevators [UNR97]. Both UP and BNSF have been offering rate incentives for loading 52-car, 75-car, and 100-plus car grain trains. Only a minority of elevators have enough capacity to load a 100-car train, and many small and medium-sized elevators would not be able to load a train half that size. The railroads are beginning to bypass the small local elevators. It is too early to tell whether large numbers of elevators will merge or consolidate. However, it is interesting to note that Mason-Logan Counties Co-op is currently building a 100-car grain loading facility at Allen Station between South Pekin and Barr, IL, along UP's Madison Subdivision [JOHN97].

Many of the more highly publicized intermodal developments have been occurring along the waterfront. This is particularly the case where planned and newly opened container facilities are concerned. At the Port of Los Angeles, a new container terminal on Terminal Island was officially dedicated on May 12, 1997. The 235-acre facility, known as Global Gateway South, features four ship berths, 12 gantry cranes, a 26-acre on-dock rail yard, and automated truck entry and exit gates. American President Lines (APL) abandoned its 129-acre terminal at the Port to move into the new facility. Yang Ming Marine Transport subsequently took over APL's former terminal and combined it with its own adjacent 54-acre container facility. The Port of Los Angeles also opened a new near-dock rail yard to serve the container terminals of Nippon Yusen Kaisha (NYK Line) and the Evergreen Group. At the neighboring Port of Long Beach, Hanjin Shipping Company began operating out of its new 170-acre container terminal in October 1997. Because the new terminal was not quite finished, Hanjin was also still using its old 57-acre facility at the Port. When completed, the new terminal will include an on-dock rail container transfer yard. Farther north, the Port of Olympia, WA, opened the Landing Sunmar Terminals container facility on May 23, 1997. The terminal includes a 76,000-square foot container freight station and a 10-acre container yard. On August 13, 1997, the Port of Tacoma, WA, broke ground for the construction of a new container terminal to be used by Hyundai Merchant Marine. In Seattle, APL's existing container facility at Terminal 5 is being expanded from 83 to 160 acres with on-dock rail facilities that will allow for two full trains to be assembled at the terminal. This project is scheduled for completion in the fall of 1998.

The trend is toward larger marine container terminals and on-dock rail transfer facilities. It is being driven by the development of increasingly larger containerships and by the evolution of the ocean carrier industry into a small number of large alliances and merged carriers. Some analysts consider 150 acres to be the minimum size required for a container terminal under the emerging environment, although 200 to 250 acres may be necessary [TIRS97]. Consequently, many ports, especially on the West Coast, have plans for new, expanded, or consolidated terminals and on-dock rail facilities. The Port of Los Angeles has announced plans to build a 315-acre container terminal on landfill in San Pedro Bay. The terminal, which will not be ready until 2001 at the earliest, will have 16 container cranes and an on-dock rail yard. In the meantime, the Port plans to build a near-dock rail yard to serve the area around APL's former terminal, now occupied by Yang Ming. The Port of Long Beach has been attempting to develop a 275-acre container terminal and on-dock rail transfer facility for China Ocean Shipping Company. However, the project, which would be located on land once occupied by the Long Beach Naval Shipyard, has been stymied by environmental and historical preservation issues. At Oakland, plans to consolidate some of the existing container terminals, which range in size from 21 to 82 acres, have been put on hold. Also in limbo are plans to develop a joint rail intermodal facility, originally announced in 1995. Currently, the Port of Oakland is attempting to acquire the 500-acre Naval Supply Center next to existing port property. The Port of Seattle plans to expand the existing 106-acre Terminal 18 container facility on the east side of Harbor Island to a 199-acre terminal served by at least two on-dock rail yards with a minimum capacity of 28 cars each. Also planned is a loop-track stretching the length of the island that would allow a 9000-foot long unit train to be assembled at the terminal. On the Atlantic side, several ports also have plans regarding container terminals and on-dock rail facilities. The Port of Charleston, SC, has acquired 1300 acres on Daniel Island where it plans to build a new terminal with twelve 1000-foot long berths and on-dock rail links. Port Everglades near Fort Lauderdale, FL, has been authorized to acquire 271 acres of undeveloped land next to its Southport Container Terminal to develop a rail container transfer yard. At the Port of Boston, which has been experiencing a decline in traffic, the Massachusetts Port Authority recently announced its intentions to concentrate all container operations at the Conley Marine Terminal in South Boston and to dedicate the other container facility, the Moran Terminal in Charlestown, to handling imported vehicles.

Recent and planned developments at seaports and on the inland waterways are by no means limited to container facilities. Bulk terminals have also been in the news lately. The Los Angeles Export Terminal, a new coal and coke export facility located on Terminal Island at the Port of Los Angeles, shipped its first load of coal to Japan on October 14, 1997. Coal arrives at the terminal in unit trains delivered by Union Pacific. The previous coal export terminal, operated by Kaiser International, is no longer allowed to accept coal at its facility. Along the Mississippi River in St. James Parish between Baton Rouge and New Orleans, LA, a new 180-acre bulk transfer facility is scheduled to open in January 1998. Served by the Illinois Central Railroad, the facility includes berths for both ocean-going vessels and barges; cranes, stackers, reclaimers, and conveyors for handling bulk material; and a railroad loop track with rotary car dumper. Initially, the terminal will import iron ore and export coal. It may also handle liquid bulk cargo, coke, and fertilizers. At the Port of Pascagoula, MS, the Jackson County Port Authority plans to develop a multi-commodity bulk terminal on the site of a closed grain elevator. Along the Illinois River north of Pekin, IL, a new rail-to-barge grain transloading facility is under construction next to the Peoria & Pekin Union Railway's mainline. On the other hand, an existing grain facility was recently closed. The Port of Mobile closed its public grain elevator in September 1997, although the port authority was negotiating with a major grain company to take over operation of the newer portion of the facility.

Two recent developments involve brand new inland waterway ports. The Port of Shreveport-Bossier began operating in April 1997. The port is located at the head of the Red River Waterway in northwestern Louisiana near the border with Texas and southwestern Arkansas. It includes a liquid bulk terminal for petrochemicals and also handles limestone, scrap materials, coal, and coiled steel [HALL97]. In October 1997, ground was broken for the new Port of Leetsdale, located on the Ohio River about 15 miles northwest of Pittsburgh. The first phase of the project involves a barge berthing facility, cranes, and a crane runway as well as improvements to nearby roads and railroads and to an existing building that will become a terminal. A food-grade warehouse will be constructed in a later phase. The port plans to receive rice and paper shipped up the Mississippi River to the Ohio River and transfer the products directly to highway trailers or railcars [COAT97].

Two other developments illustrate how new intermodal connections may be created at existing waterway terminals. In the first example, Cargill recently started to import Brazilian soybeans through the Port of Brunswick, GA. During July 1997, Norfolk Southern began hauling the soybeans out of Brunswick in unit trains. The result was a new water-rail intermodal connection at the Port [LIND97A]. The second example involves a proposed movement of iron ore by a combination of Great Lakes vessel and rail. Iron ore concentrate will be shipped by vessel from Quebec to the Lake Erie docks at Sandusky, OH. From there, Norfolk Southern will transport it to a new iron production facility in Butler, IN. Operations are expected to start before the end of 1998. Currently, the docks at Sandusky are used to transfer coal from railcars to trucks and Great Lakes vessels. Thus, a new water-rail intermodal connection involving a new commodity will be created once operations begin [LIND97B].

The above discussion provides clear evidence that numerous changes in the intermodal infrastructure are being made and will continue to be made in the foreseeable future. While the number of intermodal terminals and intermodal connections added to or removed from the transportation system over the course of a year may be relatively small compared to the number of terminals in existence, it is nevertheless significant. Systematic maintenance of the terminals database is therefore essential. In fact, deferred maintenance could easily derail it.

SUGGESTED MAINTENANCE STRATEGY

Approaches to maintaining the terminals database form a spectrum. At one extreme, we could attempt to keep the entire database current on a continuous basis, adding new terminals and intermodal connections as soon as they open and removing existing ones as soon as they close. At the other extreme, we could completely rebuild the database after a period of years has elapsed, say every five years when the Commodity Flow Survey is conducted. Neither of these strategies is very appealing. The first approach, in addition to being expensive, may not be very practical since many sources of information on intermodal terminals are not updated on a frequent basis. The sheer size of the terminals database also makes this strategy difficult to implement. The second approach could require almost as much time and cost as the initial effort to build the database. An intermediate solution would therefore seem to be appropriate.

The proposed maintenance strategy has two components. The first is a monitoring activity. It recognizes the importance of keeping abreast of the latest developments in intermodal terminals and services. The second component consists of a multi-year cycle of incremental updates to the database. It inherently involves setting priorities on each year's modifications. Each of these components is discussed further in the following subsections.

Monitoring Current Events

Regardless of how often the terminals database is revised, various transportation news sources should be monitored on a regular basis to keep abreast of current events involving intermodal terminals. A record should be kept of each new, planned, or closed terminal and each new, planned, or discontinued intermodal connection cited in news articles, press releases, and short news items. This record does not have to be formal. It could, for example, consist of a folder of news clippings, a set of notes, or both. Some degree of organization, however, is desirable such as filing the information by type of terminal and keeping notes on planned facilities separate from notes on actual terminals. As point and intermodal connection records are added to, deleted from, or changed in the terminals database, the corresponding information in the informal file can either be discarded or else archived for possible future reference.

The monitoring activity is not as onerous as it may seem. The most useful daily sources of news pertaining to transportation can usually be perused in an hour or less time. Two to four days a month could be devoted to reviewing weekly, monthly, and aperiodic sources. Generally being on the alert for references to intermodal terminals or connections while reading any transportation-related material is another aspect of the monitoring activity.

Useful sources that often contain news about intermodal terminals or services fall into one of the following categories:

The monitoring activity is not likely to uncover every terminal opening and closing. For example, it may not yield much information on privately owned terminals intended only for private use. The terminals that it does identify will generally be large public use terminals. The fact that their opening, closing, or planned development was considered newsworthy is a strong indication of their importance. Many of the news sources mentioned above were monitored while the initial version of the terminals database was being built. They proved to be valuable resources that can be a tremendous aid in maintaining the database.

Multiyear Update Cycle

Rather than update the entire terminals database each year or wait several years between updates, the proposed maintenance strategy would update parts of the database each year. Under this incremental approach, annual revisions would be based on a set of priorities and the expected availability of certain principal sources of terminals data. Thus, terminals of the highest priority would be considered each year. Terminals at the next level of priority would be considered as new listings or directories become available. Terminals at the lowest level would be considered once over a cycle of several years.

The following is a suggested schedule for updating the terminals database incrementally on an annual basis with a multiyear update cycle. It indicates the order in which different types of terminals should be considered in making the annual revisions.

There is no special significance attached to the above chronological ordering of terminal types. In fact, it could be adjusted to take advantage of new or revised terminals directories as they become available so that updates are always made with the latest information.

REFERENCES*

  1. [AAR95] Association of American Railroads, Motor Vehicle Loading & Unloading Terminals, Washington, DC: Damage Prevention and Freight Claim Section, Association of American Railroads, March 1995.
  2. [AAR96] Association of American Railroads, Railroad Facts 1996 Edition, Washington, DC: Policy, Legislation and Economics Department, Association of American Railroads, September 1996.
  3. [ARMS93] Armstrong, John H., The Railroad: What It Is, What It Does, 3rd Edition, Omaha, NE: Simmons-Boardman Books, Inc., 1993.
  4. [BLAS96] Blaszak, Michael W., "Intermodal: Railroading's Indispensable Adaptation," Pacific RailNews, Issue 394, September 1996, pp. 26-37.
  5. [CN97] Canadian National Railway, http://www.cn.ca/cn/english/about/corporate/facilities2.html/, July 1997.
  6. [COAT97] Coates, Claudia, "Ohio Gives Berth to a New Port," Journal of Commerce, October 22, 1997.
  7. [CPR97] Canadian Pacific Railway, http://www.cprailway.com/tools/entree/, October 1997.
  8. [CRC92] Consolidated Rail Corporation, Conrail Premium Service Automotive Network Guide, Philadelphia, PA: Consolidated Rail Corporation, 1992.
  9. [CSXT96] CSX Transportation, http:// www.csxt.com/docs/, July 1996.
  10. [DATE86] Date, C. J., An Introduction to Database Systems, Volume I, Fourth Edition, Reading, MA: Addison-Wesley Publishing Company, 1986.
  11. [FELD96] Feldman, Richard, and Marilyn Gross, Transportation Expressions, Washington, DC: Bureau of Transportation Statistics, U.S. Department of Transportation, 1996.
  12. [FHWA96] Federal Highway Administration, Pulling Together: The National Highway System and Its Connections to Major Intermodal Terminals, Report to the Committee on Transportation and Infrastructure, U.S. House of Representatives, Washington, DC: Federal Highway Administration, U.S. Department of Transportation, May 1996.
  13. [GAO96] U.S. General Accounting Office, Intermodal Freight Transportation: Projects and Planning Issues, Washington, DC: U.S. General Accounting Office, July 1996.
  14. [GERS97] Gersten, Alan, "New Grain Storage System Advocated," Journal of Commerce, July 17, 1997.
  15. [GPA97] Georgia Ports Authority, http://www.gaports.com , 1997.
  16. [HALL97] Hall, Kevin G., "Shreveport-Bossier Seeks Its Place on Barge Map," Journal of Commerce, October 1, 1997.
  17. [HANC94] Hancock, Kathleen L., Spatial Data and Geographic Information Systems Within the Bureau of Transportation Statistics (BTS), Paper prepared for the Bureau of Transportation Statistics, U.S. Department of Transportation, October 1994.
  18. [HELB97] Helberg, Davis, "Wading Through the Acronym Muddle: A Container is a Container is a Container," Journal of Commerce, September 2, 1997.
  19. [HOLC95] Holcomb, Mary C, and Barton Jennings, "Intermodal Freight Transportation, Transload Option," Transportation Quarterly, 49:2, Spring 1995, pp. 31-40.
  20. [IANA97] Intermodal Association of North America, 1997 Rail Intermodal Terminal Directory, Greenbelt, MD: Intermodal Association of North America, 1997.
  21. [ILTA95] ILTA, Bulk Liquid Terminals and Storage Facilities, 20th Edition, Washington, DC: Independent Liquid Terminals Association, 1995.
  22. [JENN94] Jennings, Barton E., "An Investigation of Transload: The Use of Non-Containerized Multi-Modal Bulk Shipments Within the U.S. Freight Carrier Industry," Proceedings of the 36th Annual Meeting of the Transportation Research Forum, Arlington, VA: Transportation Research Forum, November 1994, pp. 307-320.
  23. [JENN96] Jennings, Barton, and Mary Collins Holcomb, "Beyond Containerization: The Broader Concept of Intermodalism," Transportation Journal, 35:3, Spring 1996, pp. 5-13.
  24. [JOC97] Journal of Commerce Staff, "Triple Crown Boosts Its Service to Texas," Journal of Commerce, August 22, 1997.
  25. [JOHN97] Johnson, Tim, Brian Jennison, and Vic Neves, "Railroad News: Union Pacific," CTC Board Railroads Illustrated, November 1997, p. 11.
  26. [JPA97] Jacksonville Port Authority, http://www.jaxport.com, 1997.
  27. [KEEFE89] Keefe, Kevin P., "A Horse of a Different Color: Norfolk Southern's Triple Crown Service Leads the RoadRailer Revolution," Trains, 49:8, June 1989, pp. 28-37.
  28. [KEEFE96] Keefe, Kevin P., "NS Makes a Success of a 1990's Freighthouse," Trains, 56:1, January 1996, pp. 112-113.
  29. [KRAT97] Krattinger, John, "Conrail's Autorack Fleet," Conrail Technical Society, http://www. rrhistorical.com/crts/autorack.html, undated.
  30. [LIND97A] Lindsey, Scott, "Railroad News: Norfolk Southern," CTC Board Railroads Illustrated, September 1997, p. 17.
  31. [LIND97B] Lindsey, Scott, "Railroad News: Norfolk Southern," CTC Board Railroads Illustrated, November 1997, p. 17.
  32. [MBT96] "MBT Directory of Bulk Truck Transfer Facilities," Modern Bulk Transporter, 59:6, December 1996, pp. 66-89.
  33. [MONG97] Mongelluzzo, Bill, "Competition Heats Up on LA-Seattle Corridor," Journal of Commerce, August 15, 1997.
  34. [MPA97] Maryland Port Administration, http://www.mpa.state.md.us/, 1997.
  35. [MULL95] Muller, Gerhardt, Intermodal Freight Transportation, 3rd Edition, Lansdowne, VA: Eno Transportation Foundation and Intermodal Association of North America, 1995.
  36. [NS97] Norfolk Southern Corp., http://www.nscorp.com/nscorp/html/, October 1997.
  37. [NSPP96] Norfolk Southern Corp., Point-to-Point, 4:1, http://www.nscorp.com/nscorp/html/pt_to_pt/ptp9641.html#1/, March 1996.
  38. [NSPP97] Norfolk Southern Corp., Point-to-Point, 5:3, http://www.nscorp.com/nscorp/html/pt_to_pt/ptp97-3.html#tbt/, October 1997.
  39. [ORG97A] The Official Railway Guide, Hightstown, NJ: K-III Directory Corp., Vol. 129, No. 6, May/June 1997.
  40. [ORG97B] The Official Railway Guide, Hightstown, NJ: K-III Directory Corp., Vol. 130, No. 1, July/August 1997.
  41. [PSL97] The Port of South Louisiana Handbook and Directory, La Place: LA, The Port of South Louisiana, 1997.
  42. [PTE95] Petroleum Terminal Encyclopedia, Eighth Edition, Houston, TX: Stalsby/Wilson Press, 1995.
  43. [SOUTH96] Southworth, Frank, Demin Xiong, Michael S. Bronzini, and David Middendorf, A Framework for Considering National Trade and Transportation Impacts of Site Specific Freight Bottlenecks, Report by Oak Ridge National Laboratory for the Investment Analysis Branch, Office of Policy Development, Federal Highway Administration, June 1996. (DRAFT)
  44. [SOUTH97] Southworth, Frank, Demin Xiong, and David Middendorf, "Development of Analytic Intermodal Freight Networks for Use Within a GIS," Proceedings of the Geographic Information Systems for Transportation Symposium (GIS-T), Washington, DC: American Association of State Highway and Transportation Officials, 1997, pp. 201-215.
  45. [SPEAR95] Spear, Bruce, National Transportation Atlas Data Dictionary and Database Formats, Washington, DC: Bureau of Transportation Statistics, U.S. Department of Transportation, September 1995.
  46. [STEPH95] Stephens, Bill, "CSX Intermodal Launches Iron Highway," Trains, 55:10, October 1995, pp. 14-15.
  47. [STOPH94] Stopher, Peter R., Charles Apffel, Asaf Ashar, Eric Kalivoda, Ram Pendyala, and Radhika Reddy, "Defining Performance Measures for an Intermodal Management System," Proceedings of the 36th Annual Meeting of the Transportation Research Forum, Arlington, VA: Transportation Research Forum, November 1994, pp. 321-341.
  48. [TIRS97] Tirschwell, Peter, "Oakland Could Pay Price for Inaction," Journal of Commerce, May 14, 1997.
  49. [UNR97] Unrein, John, "Railroads Going Against the Grain of Small Shippers," Journal of Commerce, August 9, 1997.
  50. [WALL96] Wallace, Rich, "Why Does J. B. Hunt Ride the Rails?", Pacific RailNews, Issue 394, September 1996, pp. 38-47.
  51. [WOLF97A] Wolf, Hal, "The New Thoroughbred," CTC Board Railroads Illustrated, September 1997, pp. 52-55.
  52. [WOLF97B] Wolf, Hal, "The New CSX Transportation," CTC Board Railroads Illustrated, October 1997, pp. 52-55.
  53. [WOOD96] Wood, Donald F., and James C. Johnson, Contemporary Transportation, 5th Edition, Upper Saddle River, NJ: Prentice Hall, 1996.

* Some of the Web URL addresses might be obsolete because this documentation was finished in 1998.

LIST OF ABBREVIATIONS AND SYMBOLS