This paper describes a new approach given to the Optimal Power Use Surface (OPUS) methodology, which consists of the decomposition of a water distribution system (WDS) into an open tree-like structure (a spanning tree). Once the sumps in the model are identified, integer linear programming (ILP) is used to accelerate the design process, calculating the diameter of every node in the tree. This is achieved by focusing on setting up efficient ways to dissipate energy and distribute flow. The tree structure is built starting from the water sources. Then, the rest of the tree is assembled adding adjacent pipe-node pairs, one at a time. The methodology is tested on three benchmark problems (Hanoi, Balerma, and Taichung). When compared with results obtained through other methodologies, this new approach stands out for allowing designs with construction costs very similar to those obtained in previous works but requiring a number of iterations several orders of magnitude below. The methodology proves that following hydraulic principles and applying ILP is an excellent choice to obtain low-cost WDS designs, with very little effort, and provides an alternative path to the tiresome search process undertaken by metaheuristics.