Abstract

The accurate forecast of residential water demand is critically important for arid and oil rich countries such as Saudi Arabia which depend on costly desalination plants to satisfy the growing water demand. Achieving the desired prediction accuracy is a challenging task since the forecast model should take into consideration a variety of factors such as economic development, climate conditions and population growth. The task is further complicated given that the water sector in the country is characterized by high levels of unaccounted-for-water and heavy government subsidies, which lead to an artificially high water demand. This study presents a model for forecasting the long term water demand for Riyadh city, the capital of Saudi Arabia. The proposed model used historic records of water consumption to calibrate an econometric predictive model for per capita water use. The explanatory variables included household income, persons per household and maximum monthly temperature. Both the effects of unaccounted-for-water and conservation measures were also included. The calibration results indicate that socio-economic factors and weather temperature are equally important for water demand. The results also predict considerable savings if a comprehensive policy for water conservation and unaccounted-for-water management is implemented.

Key words: Water demand forecast, predictive model, water demand management, Saudi Arabia.