Application of Routine Calibration of Real Water Supply Network with Adjustment of Demand Roughness Parameters Driven by Applied Pressure Real Network of Brazil

Abstract

The calibration of water distribution networks is one way to perform such procedures in hydraulic models, but several are the difficulties encountered in calibrating a real network. This work proposes the improvement of modules of the calibration method proposed by Silva (2003), where using the genetic algorithm (GA) search tool, the author calibrates a real water distribution network of a Brazilian city, adjusting parameters mainly from roughness and coefficient of leakage. The enhancement of GA is the introduction of a new decision variable, the nodal demand, which assigns demand values to nodes according to the pressure-driven demand model of Tucciarelli, Criminisi and Termini (1999). The tests of the GAs implemented are tested for this real water distribution network presented by Silva (2003). The effect of the improvement on the calibration results was significant for the network, but the need for more in-depth studies, which are of course required for the application of new algorithms in real-scale networks.