Can somebody help me with the output above? If you look at the junction, water is flowing outwards without any apparent source of water. I was investigating the error in the model and looking at the dead ends to find out whats the issue.
Aadesh,
What is the actual flow value as seen in the pipe properties? It could be a very small "ghost flow" from numerical convergence noise.
If not, check the junction to see if there is a negative demand (inflow) added or possible an emitter coefficient with negative pressure.
If this does not help, please provide a copy of the model (File > Save to Package) for review: Sharing Hydraulic Model Files on the OpenFlows Forum
Regards,
Jesse DringoliTechnical Support Manager, OpenFlowsBentley Communities Site AdministratorBentley Systems, Inc.
Hi Jesse,
Thank you for your response.I am investigating this model for a valve closing exercise. I have run into issues such as above where significant flow is coming from the customer side in a dead end. The input demand in the model is all based on customer meter demand data. Ideally, the model should throw up an error saying "there is no source into the network" if all the valves are closed. Is it not so?
If demand is cut off from supply, water will be forced through the closed valves as the fixed demands impose a "known" in the equation and must always be solved as such (similar to the known HGL imposed at reservoirs for example). In this case you will indeed receive a user notification about disconnected demands as seen in this article: Disconnected Demand Nodes user notification when computing model
However. I am not quite sure if the issue you are describing is due to this. You appear to be showing a case where flow is in reverse aware from a dead end demand. First, check the start/stop orientation of the pipe to make sure that the dead end node is the stop node. If you need to fix this, re-compute the model and check again. Customer meter demands are associated with the adjacent node as described here.
If this does not help, check your user notifications and calculation summary to see if the network is unbalanced at the times during which you see the significant reverse flow. Also, check the "accuracy" calculation option. If a value larger than the default (0.001) is used, it can cause a higher magnitude of numerical noise in cases where flow should be closer to zero.
If you still need help, please provide a copy of the model and identify any specific elements or steps to reproduce: Sharing Hydraulic Model Files on the OpenFlows Forum
This is the HGL variation at a high point in the model. The HGL of the junction drops massively when the pump turns off. I anticipated some negative pressure at a high point but this is a ,massive drop(less than -5000m) in HGL. Could you please elaborate on this? Thank you.
Is this a separate model/issue from the other one discussed further above? If so I will split this post to a new thread for better tracking.
Without seeing the model it is difficult to tell why the pressure becomes so low when the pump is off. Here are a few things to check:
If this does not help, please provide a copy of the model for review: Sharing Hydraulic Model Files on the OpenFlows Forum