Book Description

The Mackenzie River shipping corridor is one of the most important transportation corridors in the Northwest Territories (NWT), an area of Canada that is rich in natural resources. The corridor is the main means of transporting goods to many communities in the NWT. It is also considered to be a potential shipping route for delivering heavy equipment to Alberta's oil sands, one of the largest hydrocarbon deposits in the world. The route, however, presents uncertainties and challenges in the delivery of goods due to shallow and variable water levels, and navigational hazards. Moreover, the route's capacity to move goods has not been realized due to low demand in the area. Better understanding of the route's capacity and reliability may enable greater utilization of this transportation corridor. The current study designs a network representation of the transportation corridor that can be used to understand route capacity and reliability issues. This research addresses two questions: i) what data are required to build a shipping network representation for the Mackenzie River corridor, and ii) how can the network representation be built using available data sources? To answer the first question, datasets used in inland and maritime freight transportation literature were identified, and then the datasets relating to the Mackenzie River inland water transportation system were gathered and organized. The data were taken from different published and unpublished reports, as well as other data sources, such as Water Survey of Canada (WSC)'s water level data and GNWT Geomatics's shape files. The data include spatial features of the Mackenzie River freight transportation system, water level, freight operators and their operations, and freight demand. Spatial features of the freight transportation system consist of the Mackenzie River and its adjacent channels, landing locations at communities, danger zones, navigational hazards, and other intermediate river locations. Shape files provided the locations of spatial features. Other attributes of these features were obtained from documents and other data sources such as river miles from Canadian Coast Guard (CCG)'s danger zone information and Mackenzie River's distance chart, and hydrometric station IDs from WSC's hydrometric database. Abstract information about freight operation (i.e. speed, transit time, and loading/unloading times) and landing facilities could be obtained. However, available information satisfies the data requirements for a strategic-level freight problem. The second part of this study describes building a network representation. This process involved identifying different node types (terminals, hazardous locations, and intermediate points), and link types representing freight operations on the Mackenzie River system (loading/unloading at communities, tug and barge operations on normal river segments, and tug and barge operations on hazardous river segments). Network nodes were prepared using shape files of the Mackenzie River spatial features in GIS, and other node attributes were also coded. Each node has six attributes: a unique ID, location (longitude and latitude), neighbouring (connected) node information, location type (i.e. node type), information on the nearby hydrometric station, and its river mile. Then, links were built from the node information by applying an algorithm that was written using simple logic to calculate link length and to assign link type, mean speed, and water level. Links have five attributes: start and end nodes, length, link type, mean speed, and water level. Furthermore, a path generation algorithm was written to find all the paths between any OD pair in the network. Other network data include tug and barge information and community cargo demand. All these data were stored conveniently in matrix form, in order to be used in computation software (e.g. MATLAB or OCTAVE) for application later. Network visualization was mostly performed in GIS. Nodes and links were checked manually, and the path generation test was conducted to verify that the network was coded correctly and that the path generation algorithm was working properly. Future work will involve formulating a mathematical model based on the network representation to estimate freight flow on the Mackenzie River system under different supply-demand and climate change scenarios.