Flood Routing Question

[chaa]

I'm modeling an 11 sq. mile watershed using WinTR-20 for the first time. The drainage area at an existing dam is 6 sq.mile. Approx. 1,500 feet downsream of the dam there is highway crossing A (D.A.=7 sq. miles). Approx. 6,000 feet from highway crossing A there is another crossing, let's call it B (D.A.= 8 sq. miles).
Increasing the storage area at the dam's reservoir, during a 100-year storm event a reduction of approx. 500 cfs of the peak discharge is achieved just downstream of the dam. The model's output shows that the 500 cfs reduction is carried over to crossing A but not to crossing B. At crossing B the peak discharge is approximately the same as in the existing condition (old, smaller reservoir). The schematic developed by the software looks just fine to me. I do not understand why the 500 cfs are not carried over further downstream on the watercourse.
Can anybody help? Where is the problem?

 

[francesca]

I'm having a hard time following your areas. Do you mean 11 sq mi total, 6 sq mi to the dam, an additional 1 sq mi (6+1=7) to the highway crossing, an additional 1 sq mi (6+1+1=8) to the next highway crossing? If so, (I'm lost on the remaining 3 sq mi), then you should be modeling these as separate catchments and routing the original (dam) hydrograph through the stream to the highway crossing where you add the 1 sq mi catchment hydrograph to the routed dam outflow hydrograph. This way, attenuation at the dam and in the channel will be passed downstream.

6 sq mi hydrograph --> routed through dam --> routed through 1,500 ft channel 
+ 1 sq mi hydrograph
= combined hydrograph --> routed through highway crossing --> routed through 6,000 ft channel 
+ 1 sq mi hydrograph 
= combined hydrograph 2 --> routed through highway crossing B

 

[chaa]

Thank you Francesca. The remaining 3 sq. miles is to the mouth of the watercourse. The model covers the whole watershed.
Yes, I'm modeling as you are suggesting and the schematic shows it. It's my understanding that "is quite typical for the peak flow reduction of a dam to be lost more and more as you look further downstream" of a watercourse but in my case there is only 1 sq. mile of additional drainage area...

 

[francesca]

I cannot say that I have ever heard that before and I'm trying to think of a case where it would be true. All I can come up with is where the peak from the dam lags the peak from the immediate drainage area, where that immediate drainage area is of a large enough area to exceed the dam peak.

You would have to look at your time of concentration of your 3 sq mi area and the time of concentration of the dam's 6 acres plus time taken to convey that peak to your area of interest. It could be that the peak of the dam flow arrives after your immediate watershed has peaked and reduced to such a level as the attenuated dam peak flow is lower.

I would expect this if your immediate 3 sq mi is an urban area and your dam catchment is rural. It would take longer for your 6 sq mi to peak, plus travel, than for your immediate 3 sq mi to peak and leave. In an urban environment, your 3 sq mi peak would be large with a shorter time of concentration due to the impervious, smooth surfaces of flow, whereas the dam catchment, being rural, would have significant losses (infiltration, depression storage, evaporation) and would travel more slowly over the vegetated terrain.

 

[cvg]

As Francesca suggested, I think the key is that instead of looking at instantaneous peak discharges, you need to look at the hydrographs and how they combine to determine the answer. Since you have not given that information here, we can only speculate on the hydrology of your site.

 

[chaa]

Thank you both. The situation is very close to what Francesca described. I made the condition quite unusual because I was trying to calibrate the model to some known discharges. Calibration is tricky...we have to be very carefull.