Water utilities rarely get to test changes on the live network. Closing a boundary valve to split a district into a separate pressure zone, or lowering the setting of a pressure reducing valve, can leave homes without water, push pressure spikes into old pipes, or reverse flow in places nobody expected. Crews usually rely on experience and a few pressure loggers, which is not enough when a change affects thousands of connections.
A hydraulic digital twin starts from the utility's GIS records: pipe lengths, diameters, materials, elevations, pump curves and valve settings. Demand patterns are added for each zone and the model is calibrated against SCADA pressure and flow readings from the same period. Calibration is the critical step. If the model does not reproduce known pressures at night and at the morning peak, its predictions will mislead the people who rely on it.
Consider a realistic case. A utility wants to close a valve between two zones to reduce night pressure in an older district where repeated bursts have been logged. In the twin, the team runs a full 24-hour demand cycle and checks the minimum pressure at the highest buildings, the time needed to refill storage tanks, and which streets would now be fed through a longer, narrower path. The result shows whether the change solves the burst problem or simply moves the risk elsewhere.
The same model should also test failure cases. What happens if the reducing valve sticks, a pump trips during the evening peak, or a fire hydrant draws heavy flow in the affected zone? Does fire flow stay above the required threshold? How quickly does a valve closure create transient pressure? The output is a short list of proposed changes, each with the properties and hydrants affected and the margin left before pressure falls below the agreed minimum.
The model is only as good as its data. Many older networks contain unrecorded connections or pipes whose diameter is uncertain. Good practice is to field-check the proposed valve locations, log every discrepancy, and roll out changes in stages: change one valve, watch SCADA for a defined window against acceptance criteria written beforehand, then continue. Platforms such as TwinUp's Clavus City can display these hydraulic results on the 3D street and building layer, so operators can see which addresses are affected before a crew leaves the depot.