Hydraulic network analysis

Large urban water supply and distribution networks are like complex organisms. Thousands of people in cities need to be able to get water without delay in their kitchens, bathrooms, washrooms, etc. Pumps can fail or cause pressure surges in pipes. Then there are industrial facilities, which consume large amounts of water in what is sometimes a very irregular manner.

Hydraulic network analysis makes use of calibrated models which in turn are based on real measured flow and pressure data. Such analysis can improve the efficiency and performance of the water supply and distribution system and eliminate energy losses caused by inefficient feed pumps or inadequate water storage technology. Hydraulic network analysis can therefore significantly reduce operating costs for local governments and municipalities.

Network model verification

Water supply network models are like snapshots of a situation that changes as a result of the construction of new residential housing or the establishment of new businesses in a commercial district. This means that every hydraulic water supply model has to be reviewed whenever such changes occur. This is done by subjecting the real network to known operating conditions and comparing calculated and actual flow rates and flow pressure in order to detect significant errors. For example, during a low-load nighttime situation, all measured hydraulic grades should display nearly the same hydrostatic values – i.e. there should be no flow influences to speak of. Discrepancies from the calculated model would indicate errors in elevation data or large unknown leaks. In a high-load situation, the hydraulic resistances of the main network pipes are checked, whereby differences here indicate hydraulic blockages caused by defective fittings or clogs, for example.

Field measurements for calibration

Field measurements are an indispensable component of hydraulic network analyses, as they provide the most important data. A key parameter here is the pipe-roughness coefficient, or Darcy friction factor formula (λ), which must be obtained in order to calculate the pressure drop in a pipe flow. Among other things, this friction value depends on whether the flow is laminar or turbulent, as well as the type of cross-section geometry involved and the design of a pipe’s inner walls.

In order to calculate the real friction coefficient, the performance of various system components must be tested and documented. To this end, the pressure, flow and water consumption values must be recorded, whereby this physical data collection is supplemented by operation data collection. Various measurements are relevant here, including continuous 24-hour flow rate recording at selected key locations, such as water catchment sites and the main network pipes. To this is added flow data collected from other key pipes and selected nodes, such as fire hydrants, that are evenly spread over the network. The storage level (hydraulic elevation) also needs to be monitored as well as other factors that influence the measured pressure values, such as inflow and outflow data for pumping stations and booster pumps.

Network model calibration

Pressure measurement sensors and data loggers from Keller are important components for guaranteeing the intelligent management of water supply systems. The data loggers are evenly spread over the network and record synchronised pressure values in realtime, which are then used to calculate and fine-tune realistic pipe roughness values and identify punctual resistances in the pipe system. This makes it possible to align measured and calculated hydraulic grades for the purposes of roughness calibration.

A verified and calibrated network model helps detect and correct hydraulic problems in existing networks and avoid similar problems when designing future network extensions.