Problem

Why do I get a "network unbalanced" user notification when computing a WaterCAD/WaterGEMS model with no demands, but with a water source (reservoir or tank)? I am trying to obtain the static pressures at nodes.

Solution

With little or no flow in the pipes, the program will have a hard time converging on a balanced solution. The reason is because the relative flow change between trials will typically be large, in this case, since the numbers that it's dealing with are very close to zero. In order for the model to balance, the relative flow change between two successive trials will need to be less than the "accuracy" value, defined in the calculation options. You can view the relative flow changes of trials in the calculation summary (analysis > calculation summary) when selecting the time step in question (00:00:00 for steady state) then clicking on the "trials" tab. To check for static pressures you may need to lower the calculation options accuracy value to something like 0.02, in order for convergence to be achieved. You can also try increasing the maximum trials value which can also be found in the calculation options.

Lastly, you can try inserting a small "fake" inflow so that the flow is slightly above zero during times when it would otherwise be zero. For example in the junction downstream of the pump that turns off, add a -0.1 or -1 gpm demand, which will act as a forced inflow. If all multipliers in demand patterns are zero for a certain time, try using a small multiplier like 0.01 or 0.001 for example, to force a tiny flow and help prevent the network from becoming unbalanced.

See Also

Troubleshooting Network Unbalanced or Cannot solve network hydraulic equations

http://communities.bentley.com/products/hydraulics___hydrology/f/5925/t/97333.aspx

http://communities.bentley.com/products/hydraulics___hydrology/w/hydraulics_and_hydrology__wiki/8769.how-can-a-user-fix-the-watergemswatercad-error-message-network-unbalanced.aspx