Presenter/Author Information

Tallen CaptFollow
Shane Walker
Ali Mirchi

Keywords

Time-Series Forecast, Modeling, Urban, Water, Demand

Start Date

27-6-2018 9:00 AM

End Date

27-6-2018 10:20 AM

Abstract

Modern day water management operators make decisions based on understanding future demands in the short, medium, and long-term time frames. The application of the many sociological, environmental, and policy driven components to how water is used in a city can be disaggregated into individual components using a multi-method modeling approach. Using a mechanistic modeling method and four data sets, regional temperature, regional precipitation, historic water production values, and historic population, combined with unique constants that describe city behavior can form a universal method for characterizing and modeling city water usage. For the city of El Paso using this methodology water predictions can be made with a R2 value of .94 and .98 for predicting on a daily time scale and monthly time scale 5 years into the future. With the ability to predict water demands into the future pressure forces can be applied to the system to observe changes. System pressures include changes to amount of precipitation, temperature fluctuations, policy changes, cost of water changes, and sociological stresses. After applying these stresses, a 5 year prediction of peak day demand, daily demand, and monthly demand can be made to assist in expansion, modification, and long term planning of water treatment facilities.

Stream and Session

C10: Modeling Urban Water Demand and the Potential Impact of Water Demand Reduction Strategies

Share

COinS
 
Jun 27th, 9:00 AM Jun 27th, 10:20 AM

Urban Demand Forecast Modeling: A Systematic Approach to Modern Modeling and Forecasting in El Paso

Modern day water management operators make decisions based on understanding future demands in the short, medium, and long-term time frames. The application of the many sociological, environmental, and policy driven components to how water is used in a city can be disaggregated into individual components using a multi-method modeling approach. Using a mechanistic modeling method and four data sets, regional temperature, regional precipitation, historic water production values, and historic population, combined with unique constants that describe city behavior can form a universal method for characterizing and modeling city water usage. For the city of El Paso using this methodology water predictions can be made with a R2 value of .94 and .98 for predicting on a daily time scale and monthly time scale 5 years into the future. With the ability to predict water demands into the future pressure forces can be applied to the system to observe changes. System pressures include changes to amount of precipitation, temperature fluctuations, policy changes, cost of water changes, and sociological stresses. After applying these stresses, a 5 year prediction of peak day demand, daily demand, and monthly demand can be made to assist in expansion, modification, and long term planning of water treatment facilities.