Water Demand Forecasting and Optimal Design of Water Distribution Networks Considering Fuzzy Random Uncertainties

Abstract

Water scarcity is a global threat due to lifestyle and climate changes, pollution of available water resources, rapidly growing population, etc. As the world population becomes more urbanized, the social, economic, and environmental vitality of the growing societies greatly depends upon the urban management of water. The development of communities has a symbiotic relationship with water, as a resource that usually goes unrecognized. Without adequate water resources and water infrastructure, urban development and redevelopment become difficult. Also, land use and development affect the use and the need for water. Hence it is recognized that water from all sources must be managed precisely to meet social, economic, and environmental requirements. newlineThe accurate forecasting of demand is essential to design various facilities of water supply scheme in general and water distribution networks in particular. The decisions on the adequate size of reservoirs, operation of pumping stations and fixing the pipe capacities etc., all depend on the design demands. Thus, in the design of a reliable water distribution network, the first and the essential step is to develop a valid and reliable water demand forecasting model, especially for assessing the peak demand. There exist three types of demand forecasting horizons, i.e., shortand#8208;term, medium-term and long-term forecasts. The short-term and medium-term forecasts provides the prediction on an hourly and monthly basis and are used for the operation and management of the system. While long-term forecast provides the prediction on a yearly basis which is responsible mainly for planning and infrastructure design. Currently, water managers produce demand estimates using longand#8208;term climate trends. However, climate change introduces uncertainties that may limit the accuracy of this method. Hence, prediction of climate-sensitive future water demand only on the basis of historical demand data will not remain much reliable. newlineForecasting of water demands play a vital role in the operations o