In this way the PSV doesn't introduce losses in the main pipe, being the losses produced by the PSV valve equal to the difference of elevation between the end of the pipe and the surface level. Connect the discharge node of the valve to the tank with a short length of large diameter pipe." The elevation of the valve's nodes should equal the elevation at which the true pipe connects to the tank and should be higher than the maximum water elevation in the tank. "Add a Pressure Sustaining Valve to the end of the inlet pipe to the tank with pressure setting of 0. How do I model a tank inlet that discharges above the water surface): But if the end of the pipe becomes submerged, then the PRV will close because it can't reduce the outlet pressure by throttling.īuild a correct model for an inlet pipe discharging above the surface level is very easy, and it is described on the Help file of Epanet2 (Help Topics, Frequently Asked Questions, 8. The EPANET solution is justified because when you connect the output of the PRV to the tank, the PRV wouldn't be active while the head in the tank be lower than the PRV setting. EPANET don't give error in this case, and provide the correct solution for zero losses in the PRV, but the formulation of the problem is conceptually incorrect. So any Q value below Qo will be valid solution. In fact, infinite solutions will be possible if Qo is the flow rate corresponding to zero losses in the PRV, a reduction of the Q value will reduce the losses in the pipe and will give an excess of pressure upstream of the PRV, which will eventually drop to zero because of the valve. In effect, the PRV can introduce losses before getting the outlet end (where you force the pressure to be zero), which will be added erroneously to the pipe losses. Introducing a PRV at the end of the pipe and setting its outlet pressure to zero is an error. I really think you are making your model to simulate a tank inlet above the surface level so difficult. So just one pressure per valve had to be adjusted. In EPANET, to force the pressure difference on regulating valves in the calibration of the Valencia I have used a pressure breaker valve (PBV), a component included You can simulate real situations forcing just one pressure and trying to adjust the
#Pcswmm lav series
In large networks the problem is no so much simple, but the lack of convergence can beĭue to an inconsistency of the data like those described, when considering the global formulationĪs a summary, try to avoid connecting a PSV and a PRV both together, either in series or If you substitute the lower reservoir by a known demand, Law tries to force both flows to be equal there will be a lack of convergence due to the InĮffect, the flow rate on the first reach of the pipe will be defined by the head differenceīetween the level of the upper reservoir and the PSV setting theįlow rate on the second reach of the pipe will be defined as well by the pressureĭifference between the PRV setting and the level of the lower reservoir. This model will never converge, because there is an inconsistency in the formulation. Then you try to model this situation substituting the regulating valve by a combination ofĪ PRV and PSV in series, forcing their settings to the measured values. The properties of the pipe are known, but not the internal behavior of the control device. You have a regulating valve in the middle of the pipe to control theįlow and you have measured the pressure upstream and downstream of the valve. Let as assume you have a single pipe carrying water from one reservoir to another one,Īt a lower elevation. Perhaps you want to force the pressure upstreamĪnd downstream of a regulating device such as a throttling valve. It is no clear from your message why you are trying to combine a pressure reducing valveĪnd a sustaining valve in your model.