Designing a Comprehensive Integrated Database for Long-Term Demand Forecasting


  • Lisa Krentz - Hazen and Sawyer

Tampa Bay Water, a regional water supply wholesaler, provides water to six Member Governments serving a population of approximately 2.5 million people in the Tampa Bay region. In support of its 2018 long-term water supply plan, Tampa Bay Water is currently undertaking the redevelopment of its long-term water demand models. The model and forecast development process will involve estimation of econometric models of water consumption by sector, designed to capture the influence of trends in socioeconomic, weather, and climate conditions, as well as end use modeling and accounting procedures to capture passive water efficiency.

To support the estimation of econometric and end use water demand models, a comprehensive set of databases and tracking tools will be developed to (a) promote flexibility for model specification and exploratory data analysis, (b) ensure that acquired information can be maintained through time to support future demand-side evaluations, and © standardize analytical routines can be replicated and updated efficiently through time. Specifically, this will involve the development of automated procedures involved in model data development and forecast generation within a SQL Server database. Tampa Bay Water’s water consumption database, which contains monthly, account level billing data dating back to 1998 for each of its Member Governments will serve as the foundation of the he basis of the integrated database design. This data represents over a half-million water use accounts across the region, geocoded to local property appraisers data through a parcel identification number. The integrated database development process will improve the alignment of billing records with the array of explanatory variables that could be specified in the econometric modeling and end use accounting processes. In particular, this increase alignment will standardize and enhance the number of alternative spatial units for modeling, as well as the ability to sub-classify residential and nonresidential sectors, for empirical analysis.

This presentation will discuss the primary dimensions, components, and functions of database and illustrate how they are integrated to support water demand model development, long-term forecasting, and Tampa Bay Water’s on-going water supply planning processes.

For more information, please contact the author at

Hear about new publications with our email newsletter

We will never share your details with anyone else.


Newsletter Newsletter

Horizons Fall 2017 (pdf)

Horizons showcases significant water, wastewater, reuse, and stormwater projects and innovations that help our clients to achieve their goals, and can help you achieve yours. Articles are written by top engineers and process group leaders, demonstrating and explaining the beneficial application of a variety of technologies and tools.

View previous issues »