Ability to generate ditch conduit based on terrain file in OpenRoads Connect Edition

You are correct, HEC-22 does not accommodate vertical breaks in conduits. Any conduit (open or closed) is treated hydraulically as a straight slope.

Unless you are working with truly irregular ditches, I am of the opinion that a regular V or Trap shape is sufficiently accurate. in a hydraulic sense, that extracting from the terrain doesn't really add much to the analysis.  You may disagree, which is fine.

What I do is create the conduit from start to end and then compare to the ditch profile from the roadway model.  How? Insert sufficient number of bends until the path of the ditch matches (within reason) the path of the ditch as defined in the roadway model.  Why? Because this makes it easy to see if and where there are any profile breaks in the ditch.  Then use insert node (dummy feature definition) to break the ditch into multiple conduits such that each conduit has a single slope which suitably matches the physical ditch profile.  The result is multiple adjacent ditches.  

Could you use channels instead of conduits?  Yes, and I think (maybe?) that an irregular channel can (maybe?) extract from the terrain model.  However, besides the potential for lots of sections being needed, a channel can only end at an outfall, thus no channels can exist in the middle of a network. You could not have a channel which starts at a certain location and then drops into an inlet, for example.  You could employ some trickeration, I suppose, and treat the channel as an isolated network and manufacture some sort of outfall that mimics the characteristics of the inlet and.....AiYiYi.

You are correct, HEC-22 does not accommodate vertical breaks in conduits. Any conduit (open or closed) is treated hydraulically as a straight slope.

Unless you are working with truly irregular ditches, I am of the opinion that a regular V or Trap shape is sufficiently accurate. in a hydraulic sense, that extracting from the terrain doesn't really add much to the analysis.  You may disagree, which is fine.

What I do is create the conduit from start to end and then compare to the ditch profile from the roadway model.  How? Insert sufficient number of bends until the path of the ditch matches (within reason) the path of the ditch as defined in the roadway model.  Why? Because this makes it easy to see if and where there are any profile breaks in the ditch.  Then use insert node (dummy feature definition) to break the ditch into multiple conduits such that each conduit has a single slope which suitably matches the physical ditch profile.  The result is multiple adjacent ditches.  

Could you use channels instead of conduits?  Yes, and I think (maybe?) that an irregular channel can (maybe?) extract from the terrain model.  However, besides the potential for lots of sections being needed, a channel can only end at an outfall, thus no channels can exist in the middle of a network. You could not have a channel which starts at a certain location and then drops into an inlet, for example.  You could employ some trickeration, I suppose, and treat the channel as an isolated network and manufacture some sort of outfall that mimics the characteristics of the inlet and.....AiYiYi.