TR-20 Att-Kin Reach Routing in Hydrocad? 1

[danmurphyIV]

I am updating a TR-20 watershed for a large watershed in HydroCAD. There are several reaches defined in the TR-20 model. TR-20 uses the Att-Kin reach routing procedure and describes these reaches with a length (L), end area coefficient (x), and an exponent (m). TR-20 uses these parameters to define a rating curve for the given reach. All of the reaches use an exponent of m = 1. No other information, including any elevation or slope informatio, is given regarding the reaches. Does anyone have any experience with this? In what manner can I model the reaches in HydroCAD? I need to model these reaches without gathering or estimating any more parameters than are included in the TR-20 model. Thanks.

 

[psmart]

HydroCAD and TR-20 use different procedures for reach routing, with different input parameters. While HydroCAD uses a physical description of the reach cross section, TR-20 uses several "reduced" parameters (as mentioned in your post) that are derived from the original geometry. In order to run the analysis in HydroCAD, you will need to estimate the original geometry based on the TR-20 parameters. I don't know of any way to do the HydroCAD routing with the TR-20 parameters alone.

 

[queque]

I recently reviewed a TR 210 output that included the
C=(90, area, discharge in cfs,and Tc.
The Tc was derived from the following equation:
Tc = ( (L^.8) *(1000/CN-9)^.7) / (1900*(S)^.5)
where Tc is in hours.
Does anyone know the reference to this equation?

Also, I used the given parameters above for the 10 watersheds and plugged them into HydroCAD with 5 rainfall Distributions.
Only the Type III gave close answers.
Diferences ranged from -11% to +15%.
I believe the engineer may have used WinTR20?
Any advice is welcomed.

 

[queque]

Clarification that was the TR 20 model.
C = 90 was the runoff coefficient.

 

[psmart]

queque,

The equation you cited is called the "lag method" or "CN method". You will find the details in NEH-4 Hydrology, in the chapter on Time or Concentration. (It is also described in the HydroCAD Owner's Manual, and in the training slides included with HydroCAD 7.1.) One of the unusual aspects of this equation is that it uses the average land slope over the entire subcatchment, rather than the slope along the flow path.

HydroCAD runoff results will match TR-20 to within 1%, as long as all the input data is the same. Other than an obvious input discrepancy, make sure the dt (main time increment) and rainfall distribution are the same. If your HydroCAD peaks are higher (especially for shorter Tc values) its propably because of the polynomial-based rainfall tables used by HydroCAD, which contain more "detail" at short durations than the linear tables used in TR-20. If you want an exact match, get the *exact* tabluar data used in the TR-20 runs and enter them into HydroCAD. I've even seen cases of TR-20 being used with incorrect rainfall tables. (In one case, alternate rainfall values had been ommited, resulting in much lower peak flows. Perhaps that's the job you're comparing!)

For a typical runoff comparision, please see the "TR-20 Sample Job #1" which is installed with HydroCAD. You will also find a detailed comparison guide at

http://www.hydrocad.net/scsresults.htm

 

[psmart]

queque, I overlooked one detail: The equation you cited gives the LAG time and NOT the Tc. To get the Tc you must divide the lag time by 0.6. For details please see the references cited above.

 

[queque]

I used the values shown on the TR20 output in a Drainage Report submitted to the County. The input includes the standard SCS precip distribution:
50 increments of 30 minute cumulative values totaling 1.
Hydrograph time increments = .0627 hours (3.67 minutes).
These are included as input and refered to as rainfall table 1 in the TR20 output, and and is the same SCS data used in most Hydrology books. The precipitation total is 3.8 inches.
The input includes the drainage area in sq.mi., CN, and Tc in hours. I converted that to acres and Tc to minutes.
I used the same CN. I get different answers.
This is what makes it confusing.

 

[psmart]

queque,

You need to compare the results in more detail as discussed in my previous reference. Let's take it one step at a time...

FIRST, Ignore the peak and compare the runoff volumes. Are they the same? For this comparison, make sure you use a long enough time span to include all the runoff, such as 0-30 hours.

The runoff volume is determined ONLY by the rainfall depth, area, CN, AMC, and the SCS runoff equation. The rainfall distribution and Tc will have NO EFFECT on the volume, so this comparison will help to isolate the reason for any differences.

Are the reported volumes the same? If not, you can calculate the runoff VOLUME be hand using the SCS runoff equation and identify the source of the problem.

After we get the volume to match, we can check the peaks...

 

[queque]

The reported and Hydrocad computed volumes are the same.
AMC =2, CN = 90 time span 0-36 hours

The Tc equation he used in the TR20 produces a Tc value that I also used in Hydrocad.
Thus, the Tc used in his TR20 produces the same volume of runoff.
That leaves the rainfall distribution selection.
I can't find the rainfall distribution tables in my Sample Hydrocad.

The rainfall distribution input into TR 20 is as follows:
0.0 .004 .008 .013 .018
0.022 .026 .031 .035 .040
.044 .048 .053 .057 .062
.066 .071 .075 .080 .093
.107 .120 .140 .170 .500
.830 .860 .880 .893 .907
.920 .924 .928 .933 .937
.942 .947 .951 .956 .960
.064 .969 .973 .978 .982
.987 .991 .995 1.000 1.000

What is this distribution named in Hydrocad?

 

[queque]

.064 .969 .973 .978 .982


oops next to last line:
first entry should read

.964 .969 .973 .978 .982

 

[psmart]

queque,

When I plot your rainfall data, it does not resemble any of the standard SCS rainfall tables. I suggest you get further information on the rainfall data, since it certainly differs from the Type III storm you were expecting.

If you do want to use this rainfall data with HydroCAD, then you can create a custom rainfall definition. For details please see Start | Programs | HydroCAD | Rainfall Info

I can also send the complete rainfall file by email.