Question about Subnetworks in Pressure Systems

[TSmith]

Good Evening all, 

I have been struggling to understand how subnetworks are defined in a pressure system. I have cleaned up all topology errors, and have tried setting both a PRV as a subnetwork controller for a pressure tier, as well as a pump station. 

For the PRV, it works fine but traverses more of the system than it should. I have this same behavior with the pump station (High Pressure out set as controller) where it somehow loops back on itself. 

Tried using a filter function barrier to understand where in the network it is somehow picking up the other pressure zone. 

This looks fine, however we add a bit more pipe and I just get more confused.

There is a closed valve here in green (pressure boundary valve) that the UN is traversing as part of this trace. 

I have fixed all pipes that were disconnected, ensured all boundary valves are connected and closed, I don't understand what I'm missing here. Ideally, the pump station would be the source for this newer higher pressure zone, the others are set by PRVs. I tried using a shortest path trace between this pump station and another point in the other pressure zone and got a "no path found" so what would ya'll recommend in terms of determining what is causing this behavior? I really appreciate the help!

[RobertKrisher]

Look at the terminal connections for things like your pumps, boundary valves, and you may find that the pipes connected to that valve are all connected to the same terminals.

[gis_KIWI4]

@TSmith - What @RobertKrisher  suggested and possibly another thing to check is the subnetwork definition and ensure that closed valves are set as barriers in the trace section of that tool. 

[ErikRose · Accepted solution]

If you set one flag within the subnetwork, and another to a segment outside the network that is unintentionally connected, then trace the shortest path, this can narrow down the number of transition points you need to inspect.

[TSmith]

This helped a lot- didn't think about that. Found a segment of pipe that was inadvertently connected when it shouldn't have been. 

The subnetwork definition does stop at any closed valve, but it looks like there was a loop created by an inadvertent connection between the two pressure zones. Thanks for pointing me in the right direction!