Water Demand Forecasting

One of the most challenging aspects in the operation and management of a water system is the development of effective long-range infrastructure and capital financing plans. Assessing infrastructure needs and developing capital improvement plans requires a thorough understanding of future water demands. The multitude of factors influencing water use such as population, housing, employment, income, and weather make it difficult even for water utilities with large planning staffs and detailed water use records. However, smaller utilities rarely have specialized staff or detailed water use records needed to conduct demand studies to assess future infrastructure needs. Ironically, the development of infrastructure and capital plans are specially important for smaller utilities where over-investment in infrastructure waste scarce capital resources and under-investment can result in antiquated systems with insufficient coverage and high maintenance costs.

The purpose of this research is to develop a water demand forecasting model that uses a multivariate approach to incorporate the climatic, socioeconomic, demographic, and water cost variables that influences water use. The model's dependent variable, average daily production, is determined from supply data provided by the participating utility. The figure below shows the variance between the predicted and actual daily production from a local water utility from 1998 to 2008.