Optimal Operation in Sectorized Networks with Intermittent Water Distribution

Abstract

[EN] The deterioration of water distribution infrastructure over the years, associated with an increase in demand and seasonal droughts can lead to an intermittent operation of the system, as the consumers cannot be supplied with a minimum pressure. The lack of investments to rehabilitate the system can be limited, and in this scenario, the intermittent water supply become a normal operation. This scenario difficulties even more the recovery of the continuous supply, as the expenses for water production increase and the revenue decrease. Thus, strategies to achieve an optimal operation in these conditions are fundamental to overcome this critical period in the best way possible. Two main objectives have to be set: minimize the operational costs, that can be described by the energy consumption in pumping stations and the volume of water lost in leakages, and, maximize water demand supplied to the consumers. Avoiding the supply during night periods, when the pressure remains high, naturally reduces the leakage volume lost. However, as the system is not operated 24h per day, it is expected a different pattern of consumption, with higher peaks during the supply period. This can lead to a significant increase in the power required by the pumps, as its head have to be higher to overcome the increased headlosses. Thus, this paper proposes an optimal operation of water distribution networks based on the scheduling of supply to different sectors of the network. This strategy aims to control the increase of the headlosses, as only part of the consumers will be supplied during a period of the day. Thus, the main pipes will not be overloaded and the power required for the pumping stations will remain low. The proposed procedure first divides the network into sub-systems using a k-means algorithm. Then, with the number of sub-systems defined, an optimal scheduling of their supply will be done. Each sub-system can have different time periods of supply, as bigger sub-system will require a higher water volume to be supplied. In addition, the pumps will be select to optimize the operation, and for each period, their rotational speed will be optimized to minimize the operational costs. The same number of sub-systems will be considered for the number of pumps in the pumping station, so adequate pumps can be selected to supply each sector. PSO algorithm will be used to optimize the operation.