March 20, 2017, 10:44 AM
No Fire Flow Design Value
I am running a fire flow design simulation and am noticing that a couple of my fire flow nodes are not producing a value within the design flow column. I have examined these nodes separately within a steady-state (no fire flow) simulation and it does not appear to make sense that these nodes would exhibit a design fire flow of zero. Does an absent value within the design flow column indicate zero or is there something else that is causing the design flow to produce a no value result?
March 20, 2017, 02:46 PM
It may be easiest to discuss this over a WebEx session with us at Technical Support. I will send you an email and see when we can meet.
March 21, 2017, 02:25 PM
Nice to speak with you today. As discussed there are a few things that can cause the software to show the design flow as blank.
1) The model cannot converge to a solution where the node in question can get down to 20 psi. Example one system in a closed loop pump station could not converge to 20psi as the HGL associated with 20 psi was at an HGL below the suction pressure of the pump station. The software could not find a way to reduce the suction pressure to achieve 20 psi to calculate the "Design Flow" in that particular instance as even if the pumps are off it could not get down to 20 psi. Another possibility is that the model can sometimes reach a point where it becomes unstable under certain flow conditions (such as causing PRV's to fight for control) that could only happen under extreme circumstances like when calculating the design flow. This is generally rare, but can occur. If this is occurring, the model cannot solve the hydraulic equations to get an answer at 20 psi and leaves the value as blank.
2) Another possibility is when there is a "kink" in the hydrant curve for a given node such as what can occur and the design fireflow answer occurs near the "kink". Due to the change in slope the software can bounce around the solution and fail to find the direct answer. The Kink can occur if certain pumps come on at higher flows or when a fireflow downstream of a PRV occurs and the change in slope on the curve occurs when the PRV goes fully open as the headloss per flow changes once the PRV reaches this point. Increasing the simulation option accuracy to 0.0001 and the fireflow accuracy in the design fireflow section also to 0.0001 generally helps the model find an a solution when that is occurring. The higher accuracy shortens the estimate for the next flow value and keeps the program from over estimating the next "guess" due and thus "bouncing" on either side of the kink where the slope changes.
Lastly the best thing to do when this occurs is to look at the fireflow report and first decide if you have good results for the required fireflow. This means that at the required Fireflow if all other key junctions are at 20 psi or higher, then the system can satisfy the required fireflow and if you don't know the "Ultimate" fireflow it can supply and maintain the design requirements it is not too problematic. Running either a manual fireflow using a standard run or by using the multi-fireflow tool with a single junction is another way to see what is occurring in the system under fireflow conditions to help you examine what specifically is occurring when the design fireflow returns as a blank value.
You can also generally glean information from the fireflow run report (make sure your standard tab report options use a full report and generate warning messages) to see what was occurring during the fireflow iterations as this can also at times help clarify what exactly may have occurred to cause difficulty in calculating the design fireflow" value. The combination of these different methods is generally sufficient to provide sufficient answers for the few times the design fireflow returns as initially blank.
Hope this helps you in your modeling efforts.
If you have difficulty with resolving a remaining fireflow issue please contact us at email@example.com and send us your model and we can help further.
Last edited by Patrick Moore; March 23, 2017 at 01:28 PM.