An Approach to Demand Pattern Estimation: Monte Carlo Simulation and Fractal Analysis.

Day to day activities of urban population in normal routine conditions are largely affected by the operation of the urban water supply system. In addition, the working hours, pattern of shift duties and living standards in a society decide the quantum of water use at different times in a day. This demand pattern is an important parameter for water supply agencies in order to ensure safe and reliable water supply. In the present study, short-term (hourly and daily) water demand is estimated for drinking as well as various other household activities. BIS standard of 135 lpcd and socioeconomic survey data are considered as the basis to derive a time series distribution of water consumption for different activities. Monte Carlo simulation (MCS) is performed to generate the random distribution of consumption pattern. The hourly water consumption for various activities shows persistent behavior with Hurst coefficient in the range of 0.592 to 0.837 and the peak flow factor of the estimated demand pattern is 5. This pattern is useful as a basis for design of service reservoir as well as pump scheduling. It is also an important parameter in the model study in performance assessment of water distribution system (WDS).