Public Utilities
How Does it Flow?
How do We Know?

How do you study an extremely complex water system when all of its pipes are hidden from view? Portsmouth Public Utilities uses a computer model to simulate the flow of water from our water treatment plant through a maze of pumping stations, storage tanks and over 350 miles of underground pipe.  The model is significantly improving the way Portsmouth manages its water system.

Portsmouth's "Cybernet ®" model is used to:

• Predict the time it takes for water to move from the water treatment plant through the distribution system. Travel time is an important predictor of water quality.
• Determine the size of new pumps, storage tanks and water mains as the system is expanded or rehabilitated.
• Evaluate the amount of stored water and water flow needed to fight fires and develop new industries.

Practical applications have included:

• Site selection and sizing of a new water pumping station.
• Selection of the sites for 300 new water quality sampling stations throughout Portsmouth's neighborhoods using water travel time and other water quality data.
• Determining the sizes for replacement of 35,000 feet of water pipe in the Parkview neighborhood.
• Evaluation of water tank storage to determine more efficient pump operation.

How a computer model recreates Portsmouth's water system.

The computer uses a mathematical model to recreate our water system and simulate the movement of water. The first step in the creation of the model is to gather all the available information about the system. This includes the size, age, length, and materials of all the pipes; the height and volume of all the water storage tanks; the capacity of all the pumps; flow data for fire hydrants and water usage data from across the city. This data is then input into the computer to build the model. The model's output is then compared to real-world conditions and then the model is adjusted or calibrated to reduce error. For example, Portsmouth used the natural fluoride in its water as a "tracer" to calibrate the model to predict real-world flows and travel times. Finally the model must be updated as flow or other conditions change.