Hi guys,
I have a pretty large model, in where I connected my hydrants with batch pipe split. The mains have user-defined length. Some hydrants have a lateral pipe (with calculated length), with a junction placed over the main. After splitting thousands of pipes, some of the new pipes (~100) had negative or zero length.
I made a query to filter these and group edited the length to 0.1m, to get through validation.
Have anyone faced this problem before?
Is there a numeric limit of the minimal distance between two junctions? - This might be the source of the problem.
Best regards,
Hello Attila,
I'm having some difficulty understanding your problem. I experimented with this and was not able to reproduce the problem, so I may not be clearly understanding the process you're using. Are you able to share a small example with steps to illustrate the problem?
You mentioned that you have some hydrants already connected to a lateral pipe, connected to a junction on the main line, which would mean that the batch pipe split is not needed. If the distance between those hydrants and the junction on the main (that they are already connected to) is less than the Tolerance field in the Batch Pipe Split tool, you could end up with some unexpected results. I'm unclear if this is what is happening, though in my testing in the latest version (CONNECT Edition) it did not result in zero or negative lengths.
There is an option under Tools > Options > Hydraulic Model to "Round Length to Nearest" but from your description I do not believe this is related.
Regards,
Jesse DringoliTechnical Support Manager, OpenFlowsBentley Communities Site AdministratorBentley Systems, Inc.
Attila Bibok
Technical Development Engineer
Budapest Waterworks
Email: attila.bibok@vizmuvek.hu
Web: www.vizmuvek.hu
Hi Attila,
Thanks for the details. After spending some additional time attempting to reproduce the issue, I believe I know what may be going on here. I suspect there are cases where the junction that overlaps the main pipe is actually overlapping another junction already connected to the main line, at the same coordinates. Consider the below case, where J-4 is connected to P-1 and P2, and J-2 is connected to P-3. The two junctions overlap (I used the same exact X and Y coordinates), with no pipe between them.
If you run the batch pipe split tool on this network, it indicates two splits occur, one for P-1 and one for P-3
In the pipe flextable, you can see that a few new pipes were created, two of which are connected between the overlapping nodes, with a length of zero since the nodes have the same coordinates
Here's the result in plan view:
And here's what we see when zoomed in and with the node dragged to the side a bit. I added a bend so you can easily see the two parallel pipes:
This situation likely would not occur if the model was built with Modelbuilder where all the pipes were imported at once and the tolerance option was used to connect pipes based on spatial proximity.
If you can confirm that this is what has happened in your model, you could delete the two resulting parallel pipes, right click on the end of one of the pre-existing main pipes, choose Reconnect, attach it to the junction connected to the lateral pipe, then delete the left-over junction.
One possible option for detecting this situation before running batch pipe split would be to run the "Nodes in close proximity" query in Network Navigator (under the Network Trace category), with a very small tolerance set. Use the highlight button to highlight any occurrences, zoom to each one, then manually connect by dragging the junction that's connect to the lateral pipe to the side, right click on the end of the lateral pipe, choose "Reconnect", attach it to the node on the main pipe, then delete the leftover junction.
Please let us know if the above explains the problem you have occurred and we'll look into improving the batch pipe split in the tool in cases like this, and document the workaround above in our Wiki.
Answer Verified By: Attila Bibok