A Freight Transport Model for Scotland

A Freight Transport Model for Scotland


M AL-Azzawi, Scott Wilson Ltd, UK


The Freight Model for Scotland is multi-modal and assesses freight by 10 different types of cargoes for more refined forecasting. In addition to the 4-stages of modelling, FMS also quantifies the network impacts on safety, economy and environment.


The Freight Model for Scotland (FMS) is a multi-modal transport model, specifically designed to model freight transportation. It covers the entire land area of Scotland, including the Scottish islands, and also has connections to major external freight trading areas such as England, Ireland, Wales and international origins / destinations.

FMS has been developed independently by Scott Wilson Ltd, originally as an in-house system to make good use of available transport data, and has been applied to many studies over the years. Like all good models it is continuously updated with new data and improvements based on its application in various projects.

All four modes of freight transport are modelled (road, rail, sea and air) and the Scottish network includes all the main road links (with some local roads), railway lines and terminals, strategic and medium-sized ports and associated routes, and airports where freight cargoes are handled. The model zone system is capable of covering all 32 local authority areas in Scotland as well as the 7 Regional Transport Partnership (RTP) regions, thereby providing the basis to test the effects of schemes and policies / strategies at all levels of Government in Scotland (local, regional and national).

A particular feature of FMS is the development of a comprehensive Base Year database of freight origin-destination (OD) movements across the country, which can produce input datasets for specific local areas, regions and corridors. The total freight market is segmented by 10 different groups of freight commodity cargoes to take into account the different characteristics of the freight market and their varying sensitivities to changes in land-uses and economic factors. This allows for a more refined and accurate modelling process and these cargo-groups can be adjusted to reflect different socio-economic scenarios when forecasting.

The model brings together a series of techniques developed by Scott Wilson Ltd and drawing on published research by others, into an integrated package using industry-standard modelling software and procedures based on Government guidance to ensure results are based on methods which are well tested and understood.

The model consists of the following applications:

- Generation and Distribution sub-model:? a Base Year OD database generates a matrix of observed flows between model zones. The outputs are freight tonnes for each of the 10 different groups of commodities for each OD movement in the model. Future changes are estimated using land-use and economic data in each zone to produce Future Year matrices;

- Mode Choice sub-model:? a series of disaggregated logit models use network skims and user-defined data (e.g. costs/charges) to adjust Future Year OD matrices by modes of transport;

- Assignment sub-model:? a detailed network-based model identifies route choice and assigns the OD matrices to a multi-modal network for predictions at the network level. For more accurate modelling, each of the 10 groups of freight commodities are assigned separately to take into account varying levels of sensitivities to factors such as value-of-time, costs, capacity, etc; and

- Evaluation sub-models:? a series of modules which estimate the impacts of assigned flows on safety (e.g. road accidents, etc), environment (e.g. emissions, noise, Sensitive Lorry Miles, etc), accessibility indices, network performance (e.g. tonnes-kms, levels of service, capacity utilisation, etc) and transport economic impacts (e.g. time savings, vehicle operating costs, revenues, carbon costs, etc). The model can be used to provide data for a variety of tests ranging from individual project / scheme evaluation to policy / strategy development.

The above features in FMS have been integrated into a complete modelling system. An important factor has been to maximise practical useability through a graphical user interface (GUI) and geographical information system (GIS) to allow presentation of results to the lay-person in easy-to-understand figures and plans. The GUI allows the user to set up and run several model batches, taking care of all the output file management, increasing productivity.

Scott Wilson Ltd has considerable hands-on experience of freight modelling and studies, and the modular approach to FMS has enabled the outputs from each sub-model to be benchmarked against identifiable features and trends in the freight market, enabling a good representation of how the freight market actually works. Model validation exercises were carried out and overall results show there is a good match between model outputs and observed flows.

The paper concludes with a review of some of the problem areas encountered by Scott Wilson Ltd in developing models of this type, including the particular issues arising from emerging changes in the freight marketplace.


Association for European Transport