DEVELOPMENT OF ROAD REHABILITATION MODEL IN FLOOD-PRONE AREAS: A CASE STUDY OF REEVES STREET LEKKI, LAGOS

Abstract

This study presents a novel predictive framework for road rehabilitation in flood-prone urban environments. The research integrates precipitation analysis, geotechnical investigation, and probabilistic modeling to develop a comprehensive road deterioration prediction system. The precipitation data analysis from 2014-2024 revealed extreme rainfall events up to 198.5 mm daily. Geotechnical testing showed that the coefficients of uniformity (Cu) of the subgrade ranged from 5.05 to 6.11, while coefficients of curvature (Cc) varied between 2.10 and 2.66, field moisture content varied between 5.29% to 10.32%, maximum dry density values ranged from 671.80 kg/m³ to 715.14 kg/m³, with corresponding optimum moisture content values between 7.07% and 8.65%. the California Bearing Ratio (CBR) exceeding 10% across all test locations, though falling short of the 30% requirement for sub-base applications using FMWH (2013) guidelines. Community surveys of 120 residents revealed that 58% rated post-flood road conditions as "poor," with 91% experiencing vehicle or property damage. A dual-model approach combining Storm Drain analysis and Markov Chain simulation was developed to predict pavement deterioration. The Markov model demonstrated that under normal conditions, roads deteriorate from 100% "Good" condition to 19.69% over 10 years, while high precipitation scenarios accelerate this to just 2.82% remaining in "Good" condition.