I have modelled a system on WaterCAD. The issue with the model is that the pipes are all flowing at around 110-115psi. I tried accounting for all the bends, valves etc through the minor loss tab for each pipe. One thing that may be affecting the model might be the pump. I defined the pump with the help of the information from the existing hydrants, like static pressure, flow etc. I selected the 3-point option to define my pump curve. But if I define my pump curve with the 1-point option, the flow in the pipes come down to 75-85psi, which is acceptable. Isn't the 1-point option for the pump definitions inaccurate?
I also had a question regarding adding a fire demand water meter to the model. How do I add this meter. It might also help reduce the pressure in the system. Any suggestions?
You won't be very accurate if y try to "back out" a pump curve as you are attempting. You need to get the original pump cure from the manufacturer or conduct pump flow tests where you measure the pump flow, suction pressure and discharge pressure.
Simply create a new scenario with a new demand alternative where you specify the fire flow as a demand and observe the system's flows and pressure.
The more points that are defined in the pump curve the more accurate the pump definition should be as long as it's based on good field data or a pump manufacturer data sheet, as Tom suggested. The one point pump definition isn't inaccurate, however, it should be used more for estimates of pump curves such as when testing for new pumps to add to an existing system.
Adding a fire flow demand can be done by increasing or adding to a current demand in the model on junction, tank, hydrant, or customer meter element. The easiest way to do that for many elements at a time is through the demand control center.
Mark
Do you have demands (and/or storage) added to the model, and is the network balanced? (check for red timesteps with corresponding "network unbalanced" user notifications). If there are no downstream demands or storage, the pump will be operating at its shutoff head, which can be very sensitive with a 1-point curve vs. a 3-point curve.
Regarding "adding a fire demand water meter" - if you're referring to accounting for the headloss across a meter device, that can be accounted for in the headloss coefficient section of the adjacent pipe properties, or by using a GPV node element similar to modeling a backflow preventer. If you mean that you want to add the demand/outflow from opening a hydrant for fire fighting, then as Mark said you would enter it at the node that represents that outflow point (junction, hydrant, etc). In the node properties you will see a "demand collection", which opens the demand entry dialog where you can add one or more demands consisting of a base flow plus a pattern (for use during an Extended Period Simulation).
Regards,
Jesse DringoliTechnical Support Manager, OpenFlowsBentley Communities Site AdministratorBentley Systems, Inc.
Yes all the demands minor loss coefficients for bends and valves have been added in the respective pipe properties. While computing, it says that the system is balanced, but a message in the user notification says that "Negative pressures in system at one or more time steps." Source: Pressure Engine. This message is for the pump.
For the water meter, I could not find the headloss coefficient section in the pipes properties. Do you mean minor losses? I was thinking that there might be a good amount of pressure loss through the fire demand water meter, which may reduce the pressure in the whole system. I tried using the GPV node on the pipe as well, but I need to know how I should obtain the headloss curve for the GPV. Just to make sure, the values shouldn't be the same values as I entered for the pump curve, right?
Yes I had done all of that, but the pressure seems to be over 100psi for the system. When I add the demand for the furthest hydrant, the pressure in the system drops to 60-70psi. I want to reduce the initial pressure, which is over 100psi, when no hydrants are being used.