Abstract
The purpose of the work described in this thesis was to establish an improved set of relationships for estimating vehicle speed and fuel consumption that can be used in road appraisal studies on rural roads in developing countries. The need for an accurate economic evaluation of various alternatives in assessing whether or not expenditure on roads will produce real economic and social benefits in the areas they serve was identified. A detailed appraisal of results from previous studies showed that an empirical approach to estimating vehicle speeds and fuel consumption for use in a decision-making framework was a more appropriate method than a theoretical approach. An experimental study of vehicle speed and fuel consumption was undertaken in the Eastern Caribbean island of St. Lucia. Regression equations were derived which related vehicle speed and fuel consumption to road and vehicle characteristics. These relationships were compared with those derived from similar studies conducted in Kenya and India. The comparison of the relationships for estimating vehicle speeds from the three studies showed large differences in predicted speeds between the different environments for all classes of vehicle. A method was developed for estimating vehicle speeds in other environments that made use of the relationships from the Kenya and St. Lucia studies. The relationships for estimating fuel consumption from the St. Lucia study were shown to be more accurate predictors than those established in either the Kenya or India studies, covering a far wider range of geometric characteristics and surface condition of the road. The modified vehicle speed relationships, established from the amalgamation of the results from the Kenya and St. Lucia studies, have been incorporated in the TRRL Road Investment Model (RTIM2), together with the fuel consumption relationships from the St. Lucia study. These relationships have improved the scope, accuracy and applicability of RTIM2 in road appraisal studies in developing countries.