Evaluating the performance of the rational method in estimating peak flow: case study [of] Makerere University

Abstract

Runoff estimation forms the basis for the design of storm water planning and management systems. The Rational method has been among the widely used peak flow estimation methods due to its simplicity. Runoff as a parameter depends on a number of parameters including ground permeability, rainfall duration, catchment area characteristics like the land uses etc. However, the Rational method does not take account of all these factors which could cause errors in the estimation of the peak runoff leading to oversizing of storm water management systems. The runoff coefficient values have not been adjusted to match the local conditions in Uganda which contributes to the largest percentage of errors in the estimation of peak flow using the Rational method. Therefore, this study focused on assessing the effectiveness of the Rational method in estimating peak flows with focus on the runoff coefficients. To achieve the objectives of the study, the storm runoff and rainfall received in the selected watershed were established over a period of two months. The storm runoff at the study area site was measured using a weir installed at one of the drainage channels whereas the rainfall data was obtained from the Uganda National Meteorological Authority (UNMA). The observed discharges were used for calibration and validation of the rainfall-runoff model (HEC-HMS) so as to obtain simulated discharges whereas the Rational method was used to obtain empirical estimates of the discharge. The calculated values were then compared with the simulated discharges from models (simulated observed values from HEC HMS) so as to obtain a factor to be applied to the runoff coefficients. The key findings of this study were that the Rational method overestimates the design peak flow as suggested by literature (Rossimiller, 1980). The Rational method estimates of the peak discharge ranged between 1.089 to 2.897 m3/s whereas those from a model calibrated using observed flows ranged between 0.32999m3/s to 1.36416 m3/s for the five return periods, that is, 5-, 10-, 25-, 50-, and 100- year return periods. It was also noted that the simulated discharges were reliable given that the calibration and validation processes yielded NSE (Nash-Sutcliffe efficiency) of 0.764 and 0.747 respectively which are above the acceptable of 0.5. By comparison of the Rational method peak discharges with those of the HEC HMS model, a correction factor of 0.455 (approximately) was developed to adjust the runoff coefficients. Therefore, it was vital to conduct this research so to improve the Rational method peak runoff estimate by localizing the runoff coefficients for a sub urban catchment using a correction factor of approximately 0.5. As such, designers are advised to take into consideration the findings of this study to cut down on high-cost designs resulting from oversizing of storm water management systems.