Problem

How do you model an open channel where the Manning's n coefficient varies based on the depth or flow?

Solution

The Implicit and Explicit (SWMM) numerical solvers in CivilStorm and SewerGEMS enable you to model a depth vs. roughness coefficient or flow vs. roughness coefficient table for a conduit. The Roughness Type in a Conduit can be set to "Roughness Depth" or "Roughness Flow", where you can vary the Manning's n coefficient based on depth or flow, respectively:

If you change the Roughness Type to "Roughness Depth" or "Roughness Flow", you can enter a Depth or Flow versus Mannings's n value curve and the program will adjust the roughness value accordingly based on the rating table. Meaning, as the depth or flow in the channel or conduit changes, the overall roughness coefficient will change accordingly, to reflect the change in material/roughness as the water level changes.

The "Roughness-Flow" Roughness Type is useful when describing a conduit or channel whose roughness changes with flow. This could occur for example in a natural channel, in which higher flows reach sections of the channel that have a different roughness (grassy vs. sandy/rocky, etc) and thus the overall effective roughness coefficient changes. If you instead know how the overall Mannings n changes with flow depth, use the "Roughness Depth" Roughness Type instead.

When modeling an Irregular Channel, you also have the option to define changes in roughness using the Bank Channel Roughness Type. With this selected, the Irregular Channel section editor allows you to enter the separate Mannings n beyond the bank stations. You would also choose the Channel Weighting Method (default is Pavlovski) which is used to determine the overall Mannings n as water flows beyond the banks.

StormCAD and SewerCAD

Note: with StormCAD and SewerCAD (or the GVF-Rational or GVF-Convex solvers in SewerGEMS and CivilStorm), currently the Roughness Type field is only available for irregular channels and only the Bank Channel method is available.

FlowMaster

Currently the pipe and channel worksheets do not have the ability to vary the roughness coefficient with depth or flow. So, you would need to observe the depth, adjust the friction coefficient, look at the new depth, adjust it again, and so on until you settle on the right friction factor. Or, if you have SewerGEMS or CivilStorm, you can use the set up a model with a single pipe (manhole with inflow > conduit > outfall) and use the "roughness depth" or "roughness flow" table as mentioned further above.