Optimized Control for Water Utilities

The rising demand for water and the need for more electricity to transport it, from remote sources to consumers, are pushing hu- manity to find ways to save water as well as energy. Thus, this paper presents a novel technique to optimize the control operations in water utilities, where the water distribution system is modeled as a Markov Decision Process to produce a control policy to minimize the energy expenses. We report experiments in a water utility that provided reduction of 39.3% in energy expenses. Moreover, it was possible to increase soundness of the system operation in order to avoid water outages.