Efforts by the Provincial Government of the Western Cape in South Africa to improve all aspects of gravel-road construction and maintenance since 2001 have resulted in such good performance that existing gravel-road performance models are of limited value in predicting future needs and properly evaluating the need for upgrading to surfaced standards. The performance of several well-constructed road sections, representing different age groups, climates, material properties, and traffic spectra, has been monitored since 2003 and compared with the prediction of various internationally accepted roughness and gravel-loss deterioration models. A background summary of the improved processes, a brief summary of the experimental design, and a discussion of existing models are presented, with emphasis on the immediate and potential future impact of results that were obtained over a period of almost 3 years. Improvements to performance models are suggested, and the impact at the project level and on strategic-level analysis and decision making is highlighted. Recommendations are made on periodic maintenance strategies based on potential economic, social, and environmental benefits.

Efforts have been made by the Western Cape Provincial Administration in South Africa to improve all aspects of gravel road construction and maintenance. This thrust has resulted in better performance over time, as had been predicted by gravel loss deterioration models. Various internationally recognized gravel loss deterioration models were evaluated, and their predictions of gravel loss were compared with actual measurements of gravel loss. Through an extensive selection and design process and by incorporating various material property, traffic volume, environmental, and geometric parameters, 20 monitoring sections were selected for comparison. This monitoring was done for more than 5 years, starting in 2003. This paper summarizes the various available gravel loss deterioration models, provides a sensitivity analysis, discusses the improved processes, and gives a brief summary of the experimental design that was followed. The experimental monitoring and the analysis of the measured and predicted annual gravel loss as well as the comparison of the measured with the predicted gravel loss are also discussed. The impact of the sensitivity analysis and the comparison between the actual measurements and the predicted gravel loss as determined from the deterioration models are highlighted. Conclusions and suggested adjustments to the models or to the monitoring procedures are made. The two models that gave the best comparisons to the measured results and that would benefit the most by the suggested adjustments are the Technical Recommendations for Highways Manual 20 model and the World Bank Highway Development and Management model.

The Western Cape Provincial Administration in South Africa initiated an experimental study to investigate the use of various additives as an alternative to conventional regraveling construction of a gravel road. The in situ material was reworked and stabilized with nontraditional stabilizers. A control section, together with a section stabilized with a traditional stabilizer, was used to compare the performance of the various sections over time. Three sections were also sealed with a bituminous surface and monitored over time. Monitoring of the sections was done either up to the point at which the road surface broke up or for 2½ years. The performance of the different monitoring sections, representing different types of additives and varying-quality (marginal) in situ material—but under the same conditions of climate, traffic volume, and construction technique—were monitored and compared over time. Four main areas of performance were monitored: during construction, in the laboratory, over time, and in economic terms. This paper provides background on the different additives used, a summary of the experimental design, and a brief discussion of the various products. It emphasizes construction and performance monitoring over time and analysis of the results. Apart from interesting observations, conclusions are made by looking not only at performance but also by taking vehicle operating and life-cycle costs into consideration.