From 1985 to 2005 global container transport grew on average by 10% per year and the growth rate of container flows from now to 2020 is still expected to be 7.5% per year. This puts high pressure on port operations and hinterland connections, and highlights the need for forecasting and decisionsupport systems on these issues. The current toolbox to support policy making (at the international, national or regional scale) consists mainly oflarge scale conventional transport models or more dedicated port forecasting models. The traditional transport models include valuable information on freight flows between regions and on hinterland transport infrastructure, which allows chain or door-to-door assignment of freight flows. An important shortcoming of these conventional transport models is that competition between ports and sector specific trends or developments (e.g. scale expansion) do not affect the choice of the ports as part of the assignment. Port forecasting models often produce detailed forecasts by commodity typeunder the assumption of a fixed hinterland. Some port forecast models do include the element of competition. However so far these models lack a detailed integration with trade flows and hinterland infrastructure needed tosimulate port competition as part of logistic chain choices. This paper describes a new approach taken by NEA and Significance, Netherlands to calculate port freight flows under different macro-economic and sector specific scenarios. The model is designed to calculate the impacts of a wide range of policy measures (e.g. infrastructure, pricing) in the port itself andits hinterland connections. Key elements in this approach are: Port competition is an explicit component of the modelling. In many regions ports dohave overlapping hinterlands and an improvement in port A reduces volumesin port B. A logistic chain approach (water access, port, hinterland) is taken in the assignment and developments in each component will affect port volume flows. Use of utility theory as a basis since this provides an integrated value for cost, time and quality factors: responses are consistent with economic theory, e.g. higher costs lead to lower volumes; the theory accounts for unobserved elements and taste variation; this will result in a spread over the options; application to detailed market segments (origin-destination, unit type, commodity group, ship vessel size category); use of a nested structure to integrate hinterland options in the port choice. The port forecasting tool is integrated within the WORLDNET database, which provides worldwide freight flows at a NSTR 3 digit level between regions (for Europe the zones are modelled at the NUTS 3 level). Sector specific scenarios are included to address changes in the industry like horizontal and vertical integration or developments in average vessel size. The approach is demonstrated for a pilot study covering the Le Havre Hamburg range. The paper will report findings from this pilot describing the sensitivity of port flow volumes to port expansion and hinterland investment plansor pricing policies. For the covering abstract see ITRD E145999
Abstract