The City's collection system of some 8,300 manholes serves a fluctuating population of 107,000 residents and 20,000 seasonal visitors. This population figure is changing rapidly as large holiday apartment blocks are replacing small retirement houses, with a consequent impact on water systems. The weather also provides a challenge - although Florida is a famously sunny location, it can also be wet, with recorded rainfalls of over 8 inches in 24 hours. This is the background against which the City's Public Works Administration are embarking on a collection network planning project with a 20-year perspective. The phases of the project were set out as follows: - Create a Base Model of the 1999 system - Calibrate and Quality Control the Base Model - Create 2010 and 2020 requirement scenarios - Use the Base Models and the scenarios as the basis for future capital expenditure and operational decisions. Wallingford Software's HydroWorks hydraulic modeling package was used for the model, and subsequently the more advanced InfoWorks package was used as soon as this was released. The calibration phase of modeling often uncovers anomalies in the operation of the system, and this case was no exception. When expected and actual flow rates were compared through field monitoring, a number of pipe obstructions and leakages were identified, and smoke testing further identified some problems with system integrity. As these operational issues were rectified during calibration, the modeling process was already delivering its first benefits. Finally, with QC completed, the model was handed over to Tampa Bay Engineering for scenario modeling. The development of 2010 and 2020 scenarios was key to the project. The assumptions made were: - population patterns taking account of all possible future changes - rainfall based on a ten-year worst case - assumptions of a city-wide rainfall event falling onto an already wetted City, using three day continuous rainfall scenarios. With the base model and the scenarios developed, the work began on simulation runs to identify how best to change the network to meet future needs. The early runs showed all the predicted SSOs, and changes in the network configuration were explored as alternative solutions to each of the forecast SSOs. The Real Time Control capability of the software -- the ability to change capacities within the model in accordance with operational rules such as the opening of valves when flow or levels reach a specific threshold -- means that the results are as close as possible to real life performance, and decisions are sounder. Engineer Robert Fahey of the City of Clearwater states, "The model was predicting a number of spots with a risk of SSOs, including a clear graphical representation of severity, and we started to look at solutions. Basically there are three options: increasing pipe sizes, increasing the capacity of existing pumps, or adding pumps at new locations. Over the years, we've run more than one hundred different solutions in the model, and it's given us not just a result to aid a specific decision, but a far deeper understanding of how our network performs. It's changed our minds on what is the best approach to capacity shortfalls. At one time we would have argued that increasing pipe sizes was the way forward. The model has shown us clearly that in fact the best value investment is in increasing pumping capacity." Robert summed up the benefits of the model as he sees them: "We use the model now to evaluate all possible network developments. We trust the results from InfoWorks, so we know it?s a good basis for decisions. We can also work with it very quickly; sometimes we hear about a new building scheme later than we should, and we need a very quick impact assessment. ?Another benefit is the reduction in SSOs, which benefit Clearwater's residents, and a reduction in customer complaints. The model outputs showing we need more pumping capacity rather than bigger pipes has saved literally millions of dollars. Although there is an upfront financial investment in setting up a model, it's far less than the cost of potential errors. I believe that, especially as programs such as CMOM move forward across the USA and identify the need for capacity improvements, a modeling approach is essential in water companies. Now, I would not want to make network decisions without the support of a model."