Tampa Bay Water, Florida's largest wholesale public water supplier, has historically relied on
groundwater to meet approximately 70 percent of the region's public water supply needs of
240 million gallons per day (mgd). The state has mandated a reduction in groundwater use
because of environmental effects observed near the wellfields. The 2000 through 2001 drought
magnified the importance of a reliable water supply in the Tampa Bay, Florida, region to meet
current and future municipal water supply needs. To meet current and future demands and to
diversify their water supply, Tampa Bay Water has developed new surface water supplies as an
alternative source. Tampa Bay Water has also developed other sources from desalinization of
seawater and through system interconnections.
New surface water sources were made operational beginning in 2002 to replace traditional
groundwater. Unlike groundwater supplies, the natural variability in surface water flows has
caused new challenges in developing long-term plans. Tampa Bay Water's Enhanced Surface
Water System (ESWS) is composed of withdrawals from two rivers and a 15-billion-gallon offline
storage reservoir. The surface withdrawals are operational and the reservoir is currently
under construction.
This paper describes a combination of statistical methods that were applied for evaluating future
water supply availability. These statistical methods included long-term trend analysis, seasonal
analysis using Fast Fourier Transforms (FFT), time series analysis using Box-Jenkins
autoregressive, integrated, moving-average (ARIMA) modeling, and Monte Carlo simulations.
Mathematical models of the 25-year forecast for the ESWS were developed, including error
bounds. The availability or non-availability of the large reservoir was also analyzed to determine
its impact on reducing variability. The reservoir dramatically altered the expected supply time
series characteristics; however, these statistical approaches were still applicable in this situation.
This work was used by Tampa Bay Water to help plan for near-term and long-range capital
improvements. Collectively, these models provide a thorough approach to analyze and represent
a highly variable and complex water supply system. Includes reference, tables, figures.