Optimal design of sewer networks including topographic criteria and drop manholes

Abstract

Optimal sewer network design is a complex problem that has been widely studied in literature. A methodology published in the literature that has shown great potential is the proposed by Duque, Duque, Aguilar, Saldarriaga. This methodology uses Mixed Integer Programming to select the layout of the network and Dynamic Programming to select the optimal combination of diameters and depths of pipes. Although the methodology showed good performance, two improvements have been implemented in later works. The first one is to consider topographic criteria when selecting the layout. This demonstrated to reduce the sewer network design’s costs. The other improvement is the addition of drop manholes in the optimization process. This allowed the methodology to comply with maximum velocity constrains in hilly regions. The present work tested the three methodologies mentioned before in a sewer network located in Bogotá, Colombia to evaluate the contributions and disadvantages of each methodology considering the cost and the hydraulic features of the resulting designs, and the computational resources required with each methodology.