A Graph-Theory Based Algorithm to Generate Decentralized Urban Drainage Layouts

Abstract

Recently, the implementation of centralized urban drainage networks, has increasingly been questioned. The main reasons are the lack of capital especially in developing countries, environmental-ecological concerns, the risk of failure, rapid change and high uncertainty in a developing country context. Latest investigations suggest a transition from centralized urban water management to decentralized or hybrid schemes. Therefore, there is a need for methodologies to evaluate the difference in performance of all feasible systems, from purely centralized to purely decentralized. For this purpose, the current study introduces the hanging gardens algorithm based on graph theory to generate all possible layouts for optimization purposes. The focus of this study is to find the optimum configuration of the pipe networks (grey infrastructures) considering different number and location of outlets. The algorithm starts with generating a centralized layout using an arbitrarily selected outlet, than another randomly nominated outlet among all candidates is added to the current layout. In the next step, the only path between new and old outlet is systematically found. Employing a randomly generated variable a pipe in the path is chosen to cut and decentralize the layout. A real case study, the stormwater collection network of a part of Ahvaz city in Iran is designed using the proposed approach. The number and location of the outlets, the layout configuration of each part and the size of the pipes are used as optimization variables. A single-objective optimization for the cost is done and obtained results are discussed.