An evaluation of unpaved road performance and maintenance

Natural earth and gravel roads have been in use for centuries. Because their design and maintenance were based primarily on local experience problems were encountered in recent years in the transfer of technology from the U.S. and Europe to other regions, especially developing countries. Furthermore, because of scarce resources there is a need for the optimal allocation of these resources in any country. One method by which the transfer of technology and resource allocation can be achieved is the use of a systems approach, and the major objective of this research was to develop a Maintenance and Design System (MDS) for the evaluation of alternative maintenance and design strategies for unpaved roads. The evaluation is based on the total transport cost, which consists of road user costs, construction cost and road maintenance cost. Several factors such as road roughness, rutting, gravel loss, material characteristics for wet season trafficability and the minimum allowable gravel thickness hold important economic implications to either the road user or to maintenance costs. Models to predict these types of deterioration were developed in terms of road geometry, traffic and surfacing material properties. The MDS for unpaved roads evaluates the interaction of the deterioration models with the different cost components. An important feature of the MDS is that it generates and orders a number of blading and regravelling strategies in terms of total cost. This format aids in selecting the feasible least total cost strategy when non-quantifiable aspects are considered. Traffic related inputs have the greatest influence on the maintenance strategy that results in the least total cost. Factors such as road geometry, discount rate, analysis period length and gravel surfacing salvage value affect the total cost, but they do not affect the maintenance or regravelling strategies. Outputs from the MDS can also be used to investigate warrants for applying a bituminous surface treatment. Furthermore, in planning and budgeting operations a network of roads generally need to be evaluated. An example presented shows that it is possible to evaluate the outputs from the MDS using dynamic programming techniques when constraints such as limited maintenance budgets for a network exist. The MDS is thus a general program suitable for the evaluation of maintenance and design strategies of a particular link, or for a road network of unpaved roads.