Model to Forecast Maximum Flows in On-Demand Irrigation Distribution Networks

Abstract

A model is developed using data from an on-demand pressurized water distribution network located in sector VIII of the Genil-Cabra irrigation district of Santaella, Córdoba (southern Spain), to simulate an irrigation season and calculate the flows that circulate in the system at any given time during the irrigation day. Using the results obtained by the model, water demand frequencies can be estimated. These results are compared to those attained by Clément’s and Mavropoulos’s formulas. This procedure enables us to determine to what extent it is possible to adjust statistical distributions to the demand obtained by both models and to verify if the hypotheses upon which these models are based are, in fact, fulfilled. Moreover, we are able to study which periods are the most appropriate for determining peak demand. Our results show that the statistical methods slightly underestimate demand because demand tends to be concentrated at two peak times during the day: one at mid-morning and another in the late afternoon. Nonetheless, the design flow obtained by the models is valid for designing the system. After studying the demand frequencies, we concluded that a better fit is achieved when a more flexible distribution such as the gamma distribution is used.