Description

This research paper evaluates the performance and accuracy of existing gravel road deterioration models, focusing on their applicability to South African unpaved road networks. Traditional deterioration models, such as those developed in early World Bank HDM studies, often rely on empirical relationships established under specific climatic, material, and traffic conditions that do not accurately mirror the diverse conditions found across Southern Africa. The study scrutinizes key distress mechanisms in unsealed roads, including gravel loss, roughness progression, rutting, and dust generation. By comparing field-measured performance data from long-term monitoring sites against model predictions, the authors highlight significant discrepancies in gravel loss rates and roughness progression curves. The paper demonstrates that standard models frequently overestimate or underestimate gravel deterioration due to inadequate accounting for local wearing course material properties, such as shrinkage product, grading coefficient, plastic limit, and stone content. Furthermore, the paper examines the influence of climate variation, maintenance frequency (such as blading schedules), and traffic composition on model reliability. The findings emphasize that uncritical application of off-the-shelf deterioration models can lead to flawed economic appraisals, misallocated maintenance budgets, and improper re-graveling schedules. To address these deficiencies, the authors propose recalibrated model parameters and suggest updated empirical relationships tailored to local materials and climatic zones. Ultimately, the paper advocates for establishing localized monitoring regimes to continuously refine asset management tools, thereby ensuring more sustainable, cost-effective management of unpaved road infrastructure across South Africa.