Using weighted Curve Numbers for Basin Routing 1

[Yakman256]

Okay guys....

I just got off the Phone with a guy that is the Township Engineers Stormwater reviewer and all stormwater comments in the Engineers review are generated by this reviewer.

He just told me that I'm not allowed to use weighted curve numbers for a Basin Routing using TR-55. I'm told that I have to generate a separate hydrograph for each type of Land cover (lawn, Impervious, Woods and Meadow) ... Then combine the inflow hydrographs and use the combined hydrograph for the routing. I asked what the time of concentration to use when modeling the roads and he told me to use the same time of concentration for all the land covers. He said that a hydrograph is a volume and that runoff volumes are to be calculated separately.

I told Him that I have a copy of TR-55 and it shows the calculation of hydrographs using a weighted Curve number and he told me TR-55 is wrong.

He is citing this section of the Township code:

"Weighted averaging of runoff coefficients shall not be used for manual computations or input data for water quality and runoff volume calculations."

He says that because a hydrograph is a runoff volume, you can't use weighted curve numbers for basin routings.

I'd love to hear your thoughts.....

 

[TerryScan]

I believe this thread may be relevant and explain some things:

http://www.eng-tips.com/viewthread.cfm?qid=335186

 

[psmart]

An update to the referenced thread (now closed):

An option to calculate separate runoff for each CN (Weighted-Q) was added in HydroCAD-10 build 8, released on March 1, 2013. Doing this with other programs (TR-55, TR-20, etc) requires that you create a separate subcatchment for each CN.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[Yakman256]

Just for the record I modeled the watershed as directed using hydraflow. When routed though the basin, the peak flows from the basin were lower than using the non-weighted curve number so it would seem that using the weighted curve number, at least in this situation, is more conservative.

 

[Drew08]

I think the rule and TR-55 is more accurate than you give it credit. Albeit it is an oversimplification as most regulations are in order to be dummy-proof, and the reviewer doesn't know quite how to articulate it.

TR-55 states: "The watersheds must be hydrologically homogeneous, that is, describable with one CN". If you have multiple land use types, by definition, it is not homogeneous. In a typical situation you can claim that land use CN's are all close enough to be homogeneous, or land uses are uniformly distributed where grass sheet flows to meadow which sheet flows to forest. The problem emerges with impervious area, which is usually parking lots and roads with their own drainage systems that never sheet flow onto a pervious surface (home gutters you might be able to claim do).

If impervious area is a relatively small component, then it can be lumped into an area-weighted CN without much difference, otherwise it should be separately modeled. As comments at the previous post suggests, I've use a volume-weighted CN rather than an area-weighted CN to get around this and still be able to model with a single watershed. Or, if the model you're using has a %impervious parameter it does the same thing (HEC-HMS, SWMM, many others), in which case the CN should only represent pervious surfaces.

 

[beej67]

This is a regional thing. Most of the country just uses a weighted CN, but I believe some areas in the northeast do it by Q weighting. I'd be very interested to read some material comparing the results of the two methods for a wide variety of watershed cases.

Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[gbam]

Wow, I thought lumped parameter modeling was the norm everywhere in the USA. I have never heard of this approach.

 

[beej67]

LincolnPE talks a bit about it in the other link. I gather he's from the Northeast, if I am properly extrapolating on the context clues. If I used Q based weighting instead of area based weighting on anything I do in the south the reviewers would pan it.



Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[psmart]

beej67 - CN weighting is still the standard in the Northeast. Q-weighting is basically unheard off. Until Q-weighting is supported by the NRCS TR-55 and TR-20 software it's not really an option for engineers or reviewers. I understand the LincolnPE has recommended the procedure to NRCS, so I'll be interested to see what develops.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[Yakman256]

TR-55 states that the Graphical Method requires homogeneous watersheds. Unless I've missed something, the Tabular method has no such limitation. I've modeled the site using both the weighted and non-weighted curve numbers and find that it only makes a marginal difference and does not affect any of the basins that I have already designed. TR-55 also makes assumptions on rainfall distribution that I doubt occurs during a natural rain event. When was the last time anyone saw a storm event that lasted 24 hours? There is also research papers that show that the flows by predicted by TR-55 can be off as much as 30% based on measured stream gauge data. That being said.... Why wouldn't weighting curve numbers be an acceptable approach to modeling the functions of basins especially when the difference is less than 10%?

 

[civilman72]

Yakman, It might be an appropriate time to break out Voodoo Hydrology.

 

[sam74]

For what ever reason the City in the Upper Midwest I am living in has an engineering firm who acts as a reviewer of drainage calculations and other things and they require that the weighted Q method is used. I have no idea why.

I did some research to try an understand the method and located the reference mentioned earlier, NEH 4 Chp. 10, and it mentions that it will essentially produce the same results but weighted Q is more difficult to calculate and technically is more correct.