Hi, I have a site with several wells, a raw water piping system, a water tower, and a potable water piping system. (Don't ask questions about the treatment and filtration of this water.. the answer is not optimal and outside the scope of this question). I believe I've set my elevations correctly, as well as my pipe directions.
I am getting a flow of -340,282,346,638,529,000,000,000,000,000,000,000,000.0 gpm at the pipe between one of the wells and its corresponding pump. I've even tried putting check valves in the pipes but I can't stop it.
Unfortunately I can't post the WTG file publicly due to the nature of the buildings in the network and the information I have on them. Any ideas as to what's happening?
Is this the flow from your well to the pump?
I am assuming you have modeled the well as a tank element (finite volume). You can check the pump curves to determine if the pump is properly setup.
Here is the relevant article for pump selection;
General Pump Selection Process
Also, I think you might have entered the pump curve values in one unit system and then changed the units. This can cause the incorrect flow values to be entered. You can check that too.
Is it possible for you to share a schematic of the setup with key elevations? We can try to recreate your model to check if we get the same result.
You can also share a sample model (removing the confidential data) and upload it securely via this link.
Hope this helps.
Bentley Technical Support
Hi Yashodhan, thanks for the reply!
This flow is going into the well from the pump - backwards from what it should be. I've checked that the pump downstream pipe is set correctly.
I've modeled the wells as reservoirs - infinite volume. For our analysis we're not interested in how much water is in the well.
I believe my pump curves are set up properly; I've included the pump curves in the upload. The model is gigantic, so look for Well #1 (ID 974) and Well #2 (ID 975).
Thanks for your help!
Also note that it is not necessary to include a check valve on the pipe directly downstream of a pump, consider removing those check valves. WaterCAD will not allow reverse flow through pumps for successful runs. A check valve on the downstream pipe is redundant and can sometimes impact the model calculations. When there is reverse flow through a pump that is a good indication that there is something else wrong with the configuration that needs to be addressed.
Hi Craig, thanks for looking into it!
The reason I have the pumps in parallel is there are 3 types of pumps - 5 pumps total in this pump house. There's a jockey pump for low flow, two pumps for normal flow, and two pumps for max flow. I will say it took me a long time to rig up all the pumps in such a way that it passed validation...
How would I best model the pump house? It seems that there is no easy way to model what's been built.
I only added the check valves later when I noticed the huge flows. I didn't think they should be necessary; they haven't been on other models. Thanks for pointing that out, I'll remove them.
As suggested by Craig, if you want to model these pump in parallel as they have different pump definitions then you better model them as VSP and not as VSPB. With the current setup results cannot be trusted.
User Notification when computing model: "Variable Speed Pump Battery speed unable to be computed"
So you could model them as 5 pumps in parallel, 2 VSP for maximum flow, 2 VSP for average flow and one jockey pump.
Bentley Technical Suppport
Answer Verified By: Didier Renault
I'd originally done that, but I was getting an error because I couldn't have the same control node for multiple VSPs. That's why I went to the pump battery. As it turns out, my pumps aren't even variable speed. So I've got everything working properly now, I think. Thanks very much for your help!