Residential water demand forecasting

Abstract
Conventional methods of forecasting the future demand for municipal water supply give undue emphasis to historical rates of use, ignore many relevant factors, and fail to distinguish between components of use. This study is concerned with only the residential part of urban water use, and a forecasting model or framework for analysis is proposed. Six factors influencing future water use are regulations on the water used by appliances, type of pricing policy adopted, policy on public education, future housing patterns, cost of supply, and changes in the technology of use. Two or three outcomes are considered for each of these factors, and from combinations of these outcomes alternative descriptions of the future are formed. The cities of Baltimore and Phoenix are used as case studies to illustrate the method. For each case, a base line estimate is made that reflects past trends. The effect on the water use of each alternative future is estimated and the probabilities of each outcome are subjectively assigned. Thus a probability distribution of the future water demand is drawn.