I found the following wiki article (communities.bentley.com/.../modeling-a-connection-to-an-existing-system) for recommended approaches to modeling a connection to a system. We typically set our model up with the fictitious pump curve method based on hydrant flow results when a full system model is not available.
However under the assumptions and limitations it says "The hydrant flow test is useful for predicting changes in pressure when downstream demands change, but not for evaluating other types of system changes such as the addition of new pipes, or operational alternatives such as fire pumps starting up." My question is, can the ficticiou pump curve be set up using a minimum 20 psi rather than the recorded residual flow to mimic maximum head available to the system?
As an example the following flow results were recorded: Static Pressure = 90 PSI, Residual Pressure = 80 PSI & Flow Hydrant = 980 gpm
As recommended in the article the min and max points on the pump curve would be set at the static pressure (90 psi) and residual pressure (80 psi) with a median point somewhere in between calculated using the flow formula to simulate the water network (i.e. P1=90 psi -> Q1=0 gpm, P3=80 psi -> Q3=980 gpm). This limits the available flow to 980 gpm and According to the article this setup is only acceptable for evaluating normal demands which makes sense. For modeling fire flows, is it valid to use the static pressure as the min condition and the lowest acceptable system pressure (20 psi) as the max condition to simulate flows in the network (i.e. P1=90 psi -> Q1=0 gpm, P3=20 psi -> F3=2803 gpm Max)?
Hello Alex,
If you are simulating connection to an existing system, I would recommend to follow the tips provided in the article you referred, Modeling a Connection to an Existing System by using the residual flow and pressure values to mimic the system head.
In this system then you can try running automated fire flow analysis to ascertain if the fire flow demands are met. Rather than using higher flow and pressure values to define the pump curve, define the residual values from hydrant flow tests and then test the network for your fire flows. Even if the pump curve operates beyond the defined limits it will give you an idea how the pump curve needs to be modified based on the system head curve generated.
Let me know if this helps.
Regards,
Yashodhan Joshi
Thank you Yashodhan. I previously tried doing what you have suggested and the program would not allow me to operate the pump beyond the defined pump curve. I got negative pressures in the system and received the following message: "PMP-1 exceeds the maximum operating point specified for the pump curve."
Based on the tips provided in the article, the pump curve is set per the flow test (90 PSI static pressure, and 980 gpm at 80 PSI residual pressure). I am trying to achieve a demand flow of 1,500 gpm which is representative of a fire flow. I believe what the warning message is telling me is my pump is not capable of providing more that 980 gpm because that is the max operating point on my pump curve. My residual pressure only dropped to 80 PSI however so there is obviously more pressure available in the system than what my fictitious pump is providing.
It seems that the recommended setup does not accurately simulate what the system is capable of when a fire truck is pumping from it as the system pressure will be drawn down to ~20 PSI which should supply far more flow than what my pump is limited to based on the flow test. I feel like I'm missing something. There is simply no way the available flow can max out at 980 gpm when that was recorded at a residual pressure of 80 psi.
This link has some information on the user notification that you are seeing: User Notification: "Pump exceeds maximum operating point specified for the pump curve." The demands will still be satisfied, but the pump is operating beyond the maximum operating point you specified.
Where are the low pressures occurring in the system? The fire flow is higher than the flow from the hydrant test, but if the headloss in the pipes is sufficiently high, you may end up with low pressures elsewhere in the system. You may want to review the system to make sure there are no unusual headlosses in the system if a higher flow is applied to the model.
The workflow for modeling a connection to an existing system is an approximation. It is possible that if a fire occurs in the system that additional jockey or fire pumps not accounted for during the hydrant flow test may turn on. Since this is an approximation, it may be best to try to model this back to the actual source, if possible. Even a skeletonized system with the actual pump data may give you better results.
Scott