Factors Influencing Traffic Demand Forecast for Privately Owned Toll Schemes

Chapter 1 – Introduction 1 Chapter 1 Introduction 1.1 Description of the Problem Traffic forecasts studies have long played an important role in the planning and financing of toll facilities throughout the world. The performance of the toll scheme depends critically on the estimates of the traffic that will use the scheme in the years ahead. Therefore, reasonable forecasts of traffic demand are a critical element in the evaluation of project feasibility. However, forecasting is an inexact science and traffic forecasts are subject to considerable uncertainties. These uncertainties arise both from the data and from the forecasting procedures used. Uncertainties in the values of the variables used within the forecast model and the assumptions that have been made in producing the forecasts are likely causes of forecast error. Furthermore, the choice of the analytical model that is used in predicting traffic volumes may be a cause of forecast error. Both these aspects are discussed in detail in chapter two. The demand for a new toll scheme comes from different potential users, each with their own travel movements and each making an individual decision on whether or not to pay a toll. Furthermore, in most cases the opening year of the project will be several years after the study is done. Traffic forecasts also depend on future forecast of economic activity, which is further subjected to the short-term economic cycles. 1.2 Thesis Objective The aim of the thesis is to review of forecast traffic demand for toll roads, shadow toll roads and public transport schemes, in order to investigate the accuracy of traffic forecasts and to determine whether there are systematic biases affecting the forecasts. For this purpose a number of schemes in different countries where the forecasts were different to the real traffic flow after the opening date will be examined. Previous studies (DETR, 1999) showed that demographic and socio-economical factors as well as the model used within the forecasts are potential causes of inaccuracy in the predictions of car-ownership and travel demand. The thesis will also investigate other aspects such