The Seaway Economic Planning System-Computer Model Program (SEPS-CMP):
"An Essential Planning Tool"

Paper Presented at the Annual Conference of The Canadian Transportation Research Forum ,
Joint CTRF/TRF Meeting, May 1984, Jasper, Alberta


Hazem Ghonima
Senior Economist, Economics Section,
Corporate Planning Branch,
The St. Lawrence Seaway Authority, Ottawa
E. mail:


This paper examines the Seaway Economic Planning System-Computer Model Program (SEPS-CMP). The system simulates most of the economic activities related to the Seaway Transportation System and allows Seaway economists to respond efficiently to the growing requirements of the St. Lawrence Seaway Authority on matters related to the present and future economic situation of the waterway. The SEPS-CMP was developed on a user-friendly and cost-effective, micro-computer system in order to be accessible, used and modified directly by in-house economist.

The major factors which prompted the development of this system will be reviewed. The SEPS concept will be examined prior to describing its structure, major components, techniques, outputs and uses. Detailed analyses of each of the SEPS-CMP sub components be the subject of other special papers.


The Great-Lakes Seaway System is a major North American transportation artery and is part of a highly complex commercial framework with tremendous economic importance to both Canada and the U.S.A. It is mainly used for domestic as well as international shipments of bulk commodities, such as grain, iron ore, coal, stone, etc., moving in and out of the continental heartland. The St. Lawrence Seaway proper which is a system of fifteen locks is an essential part of the Great-Lakes-St. Lawrence Transportation System. It extends from Montreal to Lake Erie and includes two sections, the Montreal-Lake Ontario section, consisting of five Canadian and two U.S. locks and the eight locks, all Canadian Welland Canal section. The waterway is administered by the Canadian St. Lawrence Seaway Authority (SLSA) and its U.S. counterpart, the Saint Lawrence Seaway Development Corporation (SLSDC).

Economic Functions

The Seaway is in competition with alternative routes and modes, therefore, services and transportation charges must be maintained at a level which allow the system to be competitive, efficient and safe. The Economics Section of the SLSA is responsible for the preparation of short, medium and long range traffic forecasts. These projections are updated periodically and are used by the Authority in the preparation of its strategic planning programs. In particular, they are necessary for the estimation of toll revenues

as well as the capacity requirements of the Seaway facile ties. Moreover, impact and sensitivity analyses are undertaken in order to determine the optimum toll levels and expenditure controls that fulfill the Seaway's financial object fives without undermining the competitive position of the waterway. Economic analyses are also undertaken to deter. mine the economic impact of the Seaway system on the regional economy and vice versa.

In a rapidly changing economic environment, the. requests by the Seaway Authority for quick and sophisticated answers to the Seaway economic behaviour, have increased significantly, whereas the modest resources of the Economics Section remained unchanged. In essence, this situation required the analysis and interpretation of large amounts of data of a highly complex system in a short period of time and with limited resources. Although, it was possible to meet routine requirements, it was difficult to provide fast and sophisticated response to the growing number of "What if" scenarios of sensitivity analyses of traffic forecasts and Seaway tolls.

In our efforts to respond efficiently to these growing requirements, we developed and implemented on a micro-computer system, a comprehensive model which simulates most of the economic activities related to the Seaway transportation system and allows us to use our modest resources at optimum efficiency. Although the concept of the SEPS evolved gradually since 1974, its implementation was made possible with the advent in 1980 of the user-friendly and cost effective micro-computer systems. Before describing the SEPS-CMP, it is necessary to illustrate the Seaway traffic and tolls which are the major issues around which the SEPS was developed.

The Seaway Traffic

Cargo traffic through the waterway includes more than fifty-five (55) types of commodities which can be grouped " under five major categories as follows: grain, iron ore, coal, other bulk, and general cargo. The movement of grain is together with that of iron ore and coal, the most significant of the Seaway since they constituted about 79% of all the 1983 cargo traffic on the Welland Canal. Moreover, down- I bound movement of grains provides a back-haul for the unbound movement of iron ore. In essence, these three commodities are basic to Seaway economics. The remaining traffic is distributed over a wide range of commodities such as iron and steel products, petroleum products, coke, stone, salt, etc,.

In 1958, the year before the Seaway opened, total cargo tonnage moving through the old system amounted to 10.7 million tonnes on the Montreal-Lake Ontario section and 19.3 million tonnes on the Welland. In 1959, traffic jumped to 18.7 million tonnes on the former and 25 millions on the latter. With periodic fluctuations, Seaway tonnage trended steadily upwards with a peak of 57.5 million tonnes being recorded during the 1977 season on the MontrealLake Ontario section and 66.2 million on the Welland in 1979. Since the record levels reached in the late 1970s Seaway traffic has declined substantially, mainly due to the deep and prolonged economic recession, and by 1982 shipments on the Montreal-Lake Ontario section had fallen to 42.8 million tonnes and 49 million tonnes on the Welland. Nevertheless, the economic recovery of 1983 contributed to a slight rebound of the Montreal-Lake Ontario and Welland traffic to 45.1 and 50.1 million tonnes respectively.

The Seaway Tolls

With the opening of the Seaway in 1959, a tariff of tolls based on estimates of future traffic was established by agreement between Canada and the U.S. Its purpose was to obtain directly from the users the revenues required to cover operation and maintenance costs for this joint facility as well as interest on loans and repayment of capital over a fifty-year period. This objective was not met, and since the beginning of 1973 with the impact of inflation, revenues were no longer able to keep an even pace with the operating and maintenance costs and the operating deficit for the system escalated rapidly. Although the terms of the agreement between Canada and the U.S. permitted periodic reviews of the Tariff of Tolls in order to meet financial requirements, tolls on the Seaway remained relatively unchanged (tolls on the Welland section were suspended in 1962 and replaced by a modest lockage fee in 1967) until a revision was accepted by the two governments in 1978. The revised tariff provided for increases in tolls to be phased-in over a three-year period ending in 1980. The tariff was further revised to provide for additional revenues to cover O & M costs during the 1982 and 1983 navigation periods.

The present Seaway tolls structure is a composite charge on the cargo tonnes carried by a vessel moving via the waterway, the gross registered tonnage of the vessel (GRT) and, in the case of the Welland Canal, a lockage fee related to the number of locks the vessel goes through. The amount of charges will also vary according to the section of the Seaway the vessel is transiting, the type of cargo carried, and on whether the vessel is loaded or in ballast.

The basic objective of present tolls policy is to ensure that the two Seaway entities are able to cover the costs of providing satisfactory service to users of the system. Since this policy was inaugurated in 1978, the SLSA has remained on a self-supporting basis without resorting to government appropriations or other outside funding.

The SEPS Concept

The concept used in designing the SEPS visualized the Seaway Transportation System and its major components (waterway, locks, ports, vessels, etc.) at the centre of a huge transportation network comprising alternative routes and modes competing for various cargo flowing between different supply and demand centres. This environment is surrounded and influenced by a multitude of interrelated economic and institutional activities at various levels-establishment, regional, national, and international (see Figure below).

Therefore, the economic role of the Seaway system could be better understood and the accuracy, speed and consistency of predicting its behaviour increased if we:

Accordingly, the SEPS was developed and implemented. Although its approach focused on utilization of the Seaway, it could be used to analyse other elements of this environment.

An important characteristic of the SEPS concept is that it is developed on a user-friendly and cost effective micro-computer system rather than on a mainframe computer. This allows it to be easily accessible, used, updated, modified and expanded directly by in-house economists without the need of an intermediate programmer or the continual dependence on expensive consultants. Therefore, the experience acquired by Seaway economists on the economics of the waterway will consistently grow and be used efficiently and consequently, maximum productivity will be achieved. Nevertheless, we will still need to rely on outside organizations and consulting firms for specific data and information on various elements of this environment which need particular expertise. This however, will be less frequent and at much lower costs than in the past.

Micro-Computers and the SEPS

As was mentioned earlier, the concept of the SEPS evolved gradually since 1974. However, due to budgetary constraints and the fact that the system had to be accessible, used and modified directly by in-house economists, its implementation was made possible with the advent of the relatively simple and low cost micro-computer systems in 1980. The first phase of the SEPS was implemented early in 1981 on an Apple II + personal micro-computer of only 48 KB RAM (random access memory). At that time, the SEPS consisted of a smaller version of the present Seaway Tolls Model (STM), which was first put to effective use during the 1981 toll negotiations between the Canadian St. Lawrence-Seaway Authority (SLSA) and the U.S. Saint Lawrence Seaway Development Corporation (SLSDC). This system proved to be extremely helpful in providing the two entities with the necessary data required to conduct the negotiations in a competent manner.

The second phase of this project was implemented during 1982-83 on an Apple /// micro-computer system with a larger memory of 256 KB (kilobytes). This allowed us to expand the Tolls Model and complete the development of the various forecasting models including the Iron & Steel and the Grain Forecast Models. In both phases, a modular "Electronic Spreadsheet" software (VisiCalc) as well as statistical find graphic packages (Visitrend-plot and Stepwise Multiple Regression) were used to develop the various models. Moreover, a data base software (PFS) was utilized to store and analyse relevant data and information. By mid-1983 we had developed about thirty models, simulating various economic activities of the Seaway Transportation System. Although this had tremendously increased the productivity of the economics group, it had generated a new kind of problem. Besides some irritation and time loss created by moving data between these various models, the advantage of a integrated system, in which the flow of information can move directly from one model to another, which is our main objective, was not yet achieved. This was due to the limitation of the 8-bit/256KB memory micro-computer system. Nevertheless, while we were developing the SEPS, we continued monitoring the technological progress in microcomputer hardware and software in order to find the system which suited our need.

During 1983, a new software generation, which had some capabilities for integrating various applications, was introduced. This badly needed tool was available on the more powerful 32-bit/1 Megabyte RAM Apple Lisa microcomputer. This system also included a hard-disk which can store 5 Megabyte of information in addition to six integrated software application packages including spreadsheet, graphics, data management, word processing, drawing, and project scheduling.

With features such as the integrated and multi-tasking capabilities powered by a large memory, this system fulfilled, to a greater degree, our technical requirements. Moreover, since we were able to trade in our Apple III system, the Lisa was also cost-effective and met our modest budget. Consequently, it was purchased in mid-1983. Since then we have

been implementing the third phase of the SEPS on the new Lisa system. The present report, was prepared with the use of all six integrated application software packages from the Lisa system. Total expenditures to date on all hardware and software used since the introduction in mid-1981 of the Apple III system in the Economics Section totaled less than $25 000 including the Lisa system, a Dot-Matrix printer (for graphics and drawings) and a letter-quality printer.

Structure of the SEPS

The SEPS simulates the relationships between the various elements of the environment which influence the economic behaviour of the Seaway Transportation System. It is mostly used to answer questions like "What has happened" to the traffic and the economic situation of the waterway, "How and Why did it happen", and "What is likely to happen if" the situation changes. The SEPS comprises several integrated computer models, through which the flow of information can move directly. It includes the Seaway Traffic Forecast Models (STFM), the Seaway Tolls Model (STM), the Seaway Vessels Model (SVM) and their respective Impact and Sensitivity Modules. For the purpose of this paper we will focus on the STFM to demonstrate the functions of the SEPS. The specific role of other SEPS components will be discussed when used in conjunction with the STFM for SEPS applications. The logic structure of the SEPS is illustrated in the Figure below.


The Cargo, GRT and Vessel Transit forecasts, which are prepared periodically with the help of the STFM, are automatically fed to the STM where they interact with the Seaway toll structures and are converted to corresponding toll revenues. The latter estimates will then be correlated with major Seaway expenditures data in the financial module of the STM (see Figure below). The final output is a detailed statistical report about the financial situation of both Seaway entities (SLSA and SLSDC), including their surplus or deficit situations as well as their cash f low positions. The outputs of the STFM are also fed to the SVM - which contains a data base and toll modules for vessels moving via the Seaway by size, type, age, owner, traditional route and cargo carried etc. - in order to project future fleet mix.



If a new financial goal is required, the STM and its Impact and Sensitivity modules could be used in conjunction with both the STFM and the SVM to determine the optimum toil structure and levels which achieve the desired financial objectives of both entities without undermining the competitive position of the waterway vis--vis alternative routes or modes, i.e. the SEPS-CMP will take into account the interests of the various Seaway users in conjunction with any change in the Seaway tolls. If however, this new objective cannot be achieved solely by a change in tolls, then the SEPS can also be used to determine the impact of an optimum change in Seaway expenditures. Moreover, the comprehensive market framework, developed within the STFM, can be used to identify potential traffic which could move via the Seaway and contribute to the achievement of the desired financial goal. Finally, the above three options can be applied separately or in conjunction with each other.

The Seaway Traffic Forecast Model (STFM)

The STFM is designed to satisfy the requirement for regular updating of Seaway traffic forecasts. It consists of twenty-four (24) modules (22 for Cargo Traffic Forecasts and one for each Vessel and GRT Traffic Forecasts, (see Figure below) which simulate the impacts of alternative economic, technological and institutional conditions on major cargo traffic moving via the waterway. The STFM provides conditional forecasts to pretest the effect of modified assumptions on world macro-economics, regional and modal levels for Canada and the U.S. It is mainly used to produce short, medium and long range Seaway traffic forecasts for twentytwo groups of commodities which are subsequently used to generate Vessels Transits (lockages) as well as GRT forecasts. Moreover, the model was designed to interact with the other two major SEPS components, namely the Seaway Tolls Model (STM) and the Seaway Vessel Model (SVM). In conjunction with these two models, it can be used to perform impact and sensitivity analyses related not only to the Seaway traffic but also to the Seaway tolls, to the vessels moving via the waterway and to other components of the Seaway Transportation System.

The modules of the STFM are constructed to allow maximum flexibility in updating, forecasting and modifying. It is possible to specify a set of exogenous factors which are routinely available from any number of government agencies or private forecasting organizations such as DRI, Chase Econometric, etc.

The approach used in the STFM follows the general SEPS concept, described earlier, with some slight modifications to suit the different types of commodities. In general, world demand and supply for major commodities transported through the Seaway are first analyzed and projected. Then the North American balances between domestic supply and demand are estimated, and the exports, imports and domestic shipments, which might move via the Seaway, are then segregated. Finally, the weight of the combined factors (wcf) which influence the selection of the mode(s) and route(s) through which the cargo will move, is applied in order to estimate the Seaway share of this movement. Transportation cost differentials are used instead of the (wcf) only when available on a consistent basis and from reliable sources.

The methodology applied to derive the (wcf) as well as the traffic forecasts combines both qualitative and quantitative techniques. Correlation and regression analyses which are used to obtain these estimates consist of using constant and weighted exogenous variables to derive a set of equations that forecast traffic. Validation and testing the models and their equations consist of running a correlation of estimates with past movements. The degree of accuracy was found to be between plus or minus 3 to 8% depending on the type of commodity and the length of the forecasted period.

The results of the various regression analyses performed for the development of the SEPS are very encouraging. Generally, the coefficients of determination (R2) are high averaging 0.9, indicating that the factors selected explain 90% of the past traffic variations. The F-tests, which test against the null hypothesis of no correlation, are greater

than 10, indicating that correlation exists between selected variables. The Durbin-Watson statistic (D-W), a test of autocorrelation, is between 1 and 2.5, indicating the absence of auto-correlation between the selected independent variables.

Iron Ore and Steel Traffic Forecast Model

As can be seen from Figure below, the Iron & Steel Traffic Forecast Model simulates most of the activities involved,from the production of iron ore in the Canadian Supply Center of Quebec-Labrador mines to its consumption in the U.S. Demand Centers. Between these two points, the demand for the Quebec-Labrador iron ore traffic through the Seaway and its competitive position vis--vis alternative sources, modes or routes, are influenced by the interaction of several economic, technological, and institutional factors existing at various levels of this worldwide commercial environment. The Iron & Steel Model simulates most of these relationships and produces forecast figures for the iron ore traffic as well as the iron and steel products imported to the U.S. via the Seaway. Traffic forecasts for coke and coal (used by the steel industry) are also produced as by-products of this model.


Grain Traffic Forecast Model

The figure below shows the World Grain Supply-Demand Model which was developed as an integral part of the Seaway Grain Traffic Forecast Model in order to estimate demand for Canadian and U.S. grain by major importing regions. It simulates the interaction between the factors of supply and demand of major world grain producing and consuming regions as well as the grain trade flows between them. The regions are linked together through the import-export activities, the size of which is detemmined by many factors, including the forces of supply and demand and the dynamic market share distribution between the regions. The information obtained from this model is fed into the Canadian and U.S. grain forecast models where they interact with other domestically related factors to derive grain shipments (for exports and domestic uses) by domestic transportation routes and modes. Subsequently, future grain traffic through the Seaway is estimated.


Potential Uses

Following are some examples of prospective future uses of the SEPS system.

A) Assuming that, due to weather conditions, the wheat production in the U.S.S.R. in 1985 should decrease by 20% below the 1984 level and that all major wheat exporting countries are able to maintain their market shares. What will be the impact of this development on the Seaway Grain and Total Traffic? on the Seaway Toll Revenues and Financial Sitaution?

Once the relevant data is fed into the World Grain Supply-Demand module of the STFM Model, we will be able to forecast instantly the following information:

The output from the STFM including the new 1985 cargo traffic as well as its corresponding Vessel Transits and GRT forecasts will be fed to the STM which in turn will generate the new 1985 Seaway Toll revenues by sections as well as by entities i.e. the SLSA and SLSDC. The new financial situation (Surplus or deficit) of both entities will also be provided.

B) What if changes occur in more than one factor at the same time? One such change could be a variation in the market shares of the wheat exporting countries. The model will take this variation into account when computing items 1 through 7 and will present new results for each of these items. A drop in the U.S. or Canadian market share would be a relevant information at this stage.

C) What if the U.S.S.R. wheat production increases by 20% in 1983? This situation will be the reverse of A) and B). A decrease will be noticeable in most of items 1 to 7 and the financial position of the Seaway could be negatively affected. Accordingly, it might be decided that a change in toll levels and/or structure is required in order to achieve a new financial objective. Therefore the SEPS will be used to determine the optimum toll structures and levels which fulfill this new goal without undermining the Seaway competitive position. Moreover, the system will also be used to assess the impact of these new toll changes on the various vessels using the Seaway System.


Although the third phase of the SEPS-CMP has been implemented, the system is by no means completely developed; continual research and modifications are required to bring the system to an operational and advanced stage which will allow it to respond efficiently to the requirements of the future. Moreover, although the SEPS is a comprehensive system covering a wide scope of the economic environment of the Seaway Transportation System, it should not be regarded as the "answers producer" to all of the Seaway economic problems, but rather as a tool which can help Seaway economists analyse the economic behaviour of the Seaway system.


The author gratefully acknowledges the assistance provided by S. Cullen-Naubert and S. Miguel.

Back to Top

Copyright 1998 TAF Consultants™. All rights reserved.
Information in this document is secured from sources believed reliable and due care is taken in preparation, but accuracy and completeness cannot be guaranteed. The opinions expressed in this paper are not necessarily the opinions TAF Consultants. Other products and companies referred to herein are trademarks or registered trademarks of their respective companies or mark holders.
Revised: January 09, 2005