MODELING OF RAINFALL-RUNOFF FOR A STORMWATER DRAIN IN ELDORET TOWN

Abstract

Flooding in Eldoret town has been experienced during heavy downpours resulting in loss of property, life, and interruption of transportation systems. The overall objective of the research was to evaluate different scenarios of surface imperviousness and rainfall amounts on runoff generation and the influence of low impact developments (LIDs) on the generated runoff in Eldoret town. The specific objectives were: to determine the study sub-catchment characteristics (area, slope, percent impervious, rainfall, outflow, from the catchment); to calibrate and validate SWMM5 model for rainfall-runoff simulation in Eldoret; to evaluate different scenarios of rainfall and imperviousness proportions on runoff generation and to determine the influence of infiltration trenches and bio-retention cells as low impact developments (LIDs) on the study sub-catchment on stormwater runoff. On methodology, rainfall was measured using rain gauge while discharge was measured using the current meter. Digital Elevation Model of the study area was also obtained and processed. Five scenarios for analysis were formulated as follows: Maximum measured daily rainfall and increasing percentage imperviousness in tens from the actual 25% to 75%; Average measured daily rainfall and increasing percentage imperviousness in tens from the actual 25% to 75%; Minimum measured rainfall and increasing percentage imperviousness in tens from the actual 25% to 75%; The historical daily highest rainfall recorded between 2009 and 2019 with increasing percentage imperviousness in tens from the actual 25% to 75%; Historical average daily maximum rainfall recorded between 2009 and 2019 with increasing percentage imperviousness in tens from actual 25% to 75%. LIDs on stormwater runoff were also evaluated. The results showed that the catchment drained an area of approximately 696.5 hectares with a total of 23 sub catchments. The average slope was found to be 2.57% and the mean average imperviousness was 25.72%. The maximum 3-hr rainfall event observed during the study period was 32.4 mm which resulted in the maximum average discharge of 0.131m3 /s and resulted in overflow in the drain. The calibrated model had N-Imperv of 0.45, Dstore-Imperv of 2.5, and Dstore-Perv of 8. ISE values of 3.0 and 1.4 were observed for calibration and validation, respectively. NSE values of 0.97 and 0.99 were observed for calibration and validation, respectively. This meant that the model simulated well the rainfall-discharge relationship in the study area and can be used for engineering design purposes. Scenarios of percentage imperviousness and runoff indicated that impervious surfaces in urban areas are a determining factor in runoff generation and affects the average flow and total runoff positively. The results indicated a reduction of average runoff flow by 25% when infiltration trenches were used to an extent of 100% treatment of impervious area and a reduction in total runoff volume by 19.6%. Studied low impacts developments; bio-retention cells and infiltration trenches have an effect of reducing flow and total volume in the study area therefore can be used to control flooding. It was concluded that infiltration trenches are superior to bio-retention ponds in reducing flow and total runoff volume in the study area. Future study is required to calibrate the model for water quality analysis in the study area.