Exploring the Optimal Location of Pumps as Turbines Within Branched Irrigation Networks by Global Optimization

In this study a new procedure to address the optimization of the location and selection of pumps as turbines (PATs) within irrigation networks is presented in order to reduce network pressure and recover energy. The proposed procedure consists of two steps: in the first step the optimization of the location of devices is performed, then the optimal selection and regulation is investigated in the second step. Two new mixed integer non-linear models have been developed and a global optimization solver has been employed to perform both optimization procedures. The procedure has been tested on a real irrigation network in Spain. The entire 2015 irrigation season has been used to properly model the water demand within the system. According to the results, the optimal solution consists of 5 PATs, installed in the middle region of the system. The investment cost has been estimated as 169,278 € with a payback period of almost 4.2 irrigation seasons. According to the solution, the effective amount of produced energy is equal to 289 MWh if the system is properly regulated and the pressure exceeding the minimum requirements is exploited to produce energy. Compared to other studies in literature focused on irrigated agriculture, the number of PATs is not fixed but rather it results from the maximization of the energy recovery and the minimization of capital costs, simultaneously. Moreover, the proposed procedure relies on a deterministic approach, achieving promising results in terms of accuracy.