Upgrading Urban Drainage Systems for Extreme Rainfall Events Using Multi-objective Optimization: Case Study of Tan Hoa-Lo Gom Drainage Catchment, HCMC, Vietnam

Ho Chi Minh City (HCMC) is the largest urban area in Vietnam where the urban inundation has increasingly become severe recently due to a rapid expansion of its urbanization and an increase of extreme rainfalls in terms of both intensity and frequency. The proper designs of the new or upgraded drainage (canal) systems based on the recent extreme rainfall data can reduce the flooding impacts. Hence, a new extreme rainfall intensity-duration-frequency (IDF) relation is first constructed for HCMC based on the most up-to-date observed data from 1982 to 2019. This rainfall IDF relation is then used as an input to investigate and select a suitable drainage routine, shape, and size for upgrading the urban drainage systems in HCMC. These optimal cross-sections as well as routine selections were determined based on an optimization model through a large number of numerical simulations. In this paper, an interfacing Visual C++ programming tool with an SWMM DLL for rainfall-runoff computation in a drainage network completes the optimization. The multi-objective genetic algorithm (MOGA) is used to find the optimum solutions by minimizing both the total flooding overflow and the installing cost as the objective functions. The developed model was applied to the case study of Tan Hoa-Lo Gom drainage basin in HCMC. Model performance was evaluated to propose the optimum solution to the drainage system upgrade. By considering the most recent extreme rainfalls and urbanization conditions, with the optimal drainage routine and cross-section identified by the model, loss flooding in Lo Gom canal after upgrading can be reduced by 57.7%. This optimal drainage model is, therefore, premise for further studies in selecting the best option of flooding prevention in urban areas.