Can PondPack model orifices in series? I have a scenario where I have an existing 7" orifice plate mounted to an 18" culvert to control the outflow from a detention basin. I now need to restrict the flow for a 1-yr event even more with a 4" orifice at the pond discharge elevation (46.86ft). I want to allow all events more than a 1-yr to use the existing 18" culvert, so I put in a weir at an elevation slightly above the 1-yr water surface (50.25ft). All water overtopping this weir can then re-enter the 18" culvert. Somewhere I need to still control the release with a 7" orifice plate. I have tried modeling the 7" orifice plate at two different spots: at the pond discharge, and at the downstream pipe of the outlet structure containing the weir.
The attached image shows what I would like to model. PondPack enters some endless loop when analyzing this outlet structure. It seems that when I enter two orifices in my outlet structure they both are draining the pond simultaneously (in parallel) rather than in series where one orifice's discharge would then go through the other orifice.
Any suggestions on how to model this situation?
David,
I think you'd need to look into those other situations separately. Maybe in those cases one of the culverts is controlling. It seems to me that whenever you have structures in series, one of them is likely controlling so you can simplify the structure by only using that single one. Beyond that, the other options I can think of are:
1) Model it with the extra pond volume or intermediate pond as I recommended in my last post. (smaller calculation timestep likely needed in this case)
2) Adjust the culvert tolerance in addition to adjusting some of the other convergence tolerance parameters to acheive a stable rating curve.
3) If possible, come up with an equivalent culvert/orifice to represent the series of structures
4) If you already know the rating curve of the composite structure, you can enter that as a user defined outlet structure.
Jesse
Regards,
Jesse DringoliTechnical Support Manager, OpenFlowsBentley Communities Site AdministratorBentley Systems, Inc.
Jesse,
thanks for the reply. For your example it looks reasonalbe to assume that you can eliminate the culvert.
in some other runs i did with different numbers the two curves do not overlay so well which left me with dout.
david perrings
Hello David,
I looked at your model with Terry. When computing an outlet structure with multiple culverts in series, it may take a very long time to converge, as this adds a lot of complexity. Increasing the culvert tolerance will help increase run time, but the results may not be stable because it's allowed to converge with a more relaxed tolerance. If you compare the "erratic" rating curve to that of a rating curve using just the orifice, you will see that they are almost identical, except the orifice rating curve is smooth and computes almost instantly. Therefore, the orifice is controlling so the inclusion of the culverts does not appare to be necessary.
If you need to model pooling of the water inside the culvert at the entrance of the orifice, you may consider include the culvert volume in the upstream pond's elevation-volume curve or adding an intermediate "dummy" pond.
Jesse Dringoli
Bentley Technical Support
I have a similar situation. when i put a orifice plate at the downstream end of a culvert and try to model it i get a very eratic elev. flow curve which does not appear stable. The calcs do not converge even with a tolerance of .001. I submitted this question to Bentley and Terry's response was to simple model it as just an orifice plate, since the orifice plate controls. After considering this i am not comfortable with this approach. Does anyone else have any great ideas.
david perrings, pe
Hello Josh,
Yes, Pondpack can model orifices in series. Since you had also submitted your question to technical support, I took a look at your model and diagram and found the problem.Basically you need to have the 7" orifice and "upstream culvert" in series, with the 4" orifice and manhole weir in parallel, but downstream of it. Pondpack cannot handle this because you can only have one downstream ID associate with each outlet structure element. The "upstream culvert" would need to have both the manhole weir and the 4" orifice as the downstream ID. With the way you have it in your model, no outlet structure element has the manhole weir as its downstream ID, so it's as if the manhole weir is exposed to the pond headwater.
The outlet structure calculation is not going into an endless loop but is actually taking a very long time to iterate and settle on a solution. If you increase the tolerance in both culverts to 0.1, the calculation will eventually complete, but as mentioned above, it's not an accurate depiction of what you need to model.
My suggestion for modeling this case is to represent the upstream side of the manhole as a second pond. Basically the outlet structure of the real pond would be the 7" orifice and culvert in series, discharging to the second pond. The second pond would have the dimensions of the upstream chamber of the manhole and its outlet structure would consist of the weir and 4" orifice in parallel, each with the culvert as the downstream ID. With this configuration, a small calculation timestep and headwater increment is recommended.
Another option would be to take a look at our CivilStorm package, which is designed to model more complex situations. Dynamic routing can occur through conduits, which can have control structures on either end.
I hope this helps.