Catchment outflow node to ditch

Hello Isolde,

The thread that you replied to was from over two years ago. I have split your question into a new thread, referenced the old thread and included the screenshot in question.

The screenshot provided is of the "Office Project" example model, which is included with StormCAD, SewerGEMS and CivilStorm. I'm not sure on the background of the choice of outflow node for the catchments, but the model is just for illustrative purposes. One possible interpretation is that the contour of the land slopes from the top-left corner of "catchment 9" to the bottom right corner, and all of the runoff exits a common discharge point in the bottom right corner, hence the outflow node is set to the cross section node on the right side (CS-SW3). From what I have seen, this is the typical interpretation of a catchment - the area that drains to a common outflow point. Now, if that common outflow point spans a length of channel and if the exact outflow node is a concern, the modeler may want to subdivide their catchments and split their channels/conduits into more sections to allow for a finer level of detail as far as where the flow enters the drainage system. You could even perform a study to test the impact on the hydraulic results between such a fine level of detail, and a less detailed version.

To answer your question, runoff from catchment 9 could potentially still have an impact on channel CH-5, in that the runoff that enters the channel will add to the other upstream flow and could influence the depth of flow in the channel. This can then effect the upstream channels due to the backwater analysis, though this may typically only occur in profiles that are more mild/flat.

You mentioned the word "design", though - note that the automated design feature (in the StormCAD GVF Rational and SewerCAD GVF Convex solvers) sizes pipes based on theoretical, mannings-based capacity, in which case the catchment 9 flow entering downstream would not effect automated sizing in the upstream pipes. (if you were to model them as conduits and transitions per the first screenshot in the old thread - automated design only works on conduits at the moment).

You may want to take a look at our Stormwater Conveyance Modeling and Design book for further information on modeling practice for catchments/drainage areas.

Dear Jesse,

I would just like to come back to this question for a particular point: considering what you said, how then modelize with StormCad a ditch / a trench running along a street for example?

Please see on the following drawing (sorry for bad quality) what kind of structure I mean.

I would design this trench with a "hand calculation" following these summarized steps:

- set a first length (L);

- compute the time of concentration, addition of the overflow time on the street, and of the time in the gutter, which is supposed (through a presumed velocity) in the first place;

-Then, find the right size of the gutter so as to accommodate the runoff computed (considering an area of W x L) ;

-adjust the velocity in the gutter, and the time of concentration;

-and finally recompute with the proper time of concentration.

With Stormcad, as the time of concentration is used-defined, is it necessary to adjust it manually with the gutter velocity computed by the software or is there a better solution? Moreover, if the catchment area is set to the upstream node of the gutter, the hydraulic profile is "upside down", as it consider that all the water comes in one single node: the flow slowly reduces along the ditch. That is not representative of the reality, is there a solution for this?

Many thanks again in advance for your help.

Best regards,

I may not be fully understanding what you're doing, but I will try to address your questions:

Re: "With Stormcad, as the time of concentration is used-defined, is it necessary to adjust it manually with the gutter velocity computed by the software or is there a better solution?"

I'm not sure I understand why the time of concentration (Tc) would be adjusted based on gutter velocity. The Tc pertains to the catchment runoff, which discharges from the catchment into the gutter. In other words, the runoff has already left the catchment area when it is in the gutter. The Tc should be fixed, based on the characteristics of the land. In some situations an average velocity may need to be used (velocity of the flow across the catchment surface before it reaches the discharge point). You can see several Tc calculation methods available when the "Tc Input Type" is set to "Composite", such as SCS Lag, TR-55 Sheet flow, etc.

Re: "Moreover, if the catchment area is set to the upstream node of the gutter, the hydraulic profile is "upside down", as it consider that all the water comes in one single node: the flow slowly reduces along the ditch. That is not representative of the reality, is there a solution for this?"

This refers back to the original question - see my previous response.

If the length between catchbasin elements is short enough, the assumption of a single discharge point for the catchments should likely be acceptable, as there would likely not be a notable difference between the results from this assumption and a real system where runoff is distributed evenly over that length.

If the length is long enough for this to be an issue, then you may need to split it into multiple links. For example, Cross Section > Channel > Cross Section > Channel, and so on, with the outflow node of the catchment for the corresponding drainage area at each segment set to the cross section.

Regarding "flow slowly reduces along the ditch": I assume you're referring to the attenuation effect in StormCAD's GVF-Rational solver due to the recalculation of Q=CIA as the system time increases when moving downstream. This happens once for each element though, based on the travel time in the link (gutter in your case).

One could consider this assumption conservative, in that the full runoff enters the upstream end of the ditch/gutter and therefore it needs to be sized to accomodate that entire flow.

I hope this helps,