Mirroring a fishy Hydraflow model

[beej67]

Story -

Original engineer designs a stormwater vault using Hydraflow Hydrographs, a piece of software I am familiar with, and am not fond of. I do not have a copy of his model, but I do have a copy of his hydro study, including all his hydrograph printouts.


Situation -

Original engineer is long gone. I am hired to do the detention pond as-built certification, which in this municipality requires the engineer signing off to rerun the stormwater model. I do not own Hydraflow Hydrographs, but am using the HydroCAD trial version for now, and have a full license on order. Luckily, the important bits of the previous engineer's model are 5 nodes, so I'm good here.


Problem -

I can't get my "design" model to match his "design" model, which means the vault isn't going to meet guidelines, which is bad for my client. I need to show the municipality that what's been built meets the allowable flows of his previous hydro study, ideally without going back and fixing or amending the original hydro study first, because that's way out of my scope and my client is strapped for cash.


Source -

I have traced the source of the problem to the runoff hydrographs. I've included an attachment that has the two hydrographs compared with each other. Both are SCS Type II 24 hour, both have the same rainfall total, same basin area, and same Tc. All the runoff hydrographs in my HydroCAD are a smidge higher than his are. 484 shape factor for both. The CN difference (77 vs 76.9) is negligible, I've checked it.

The only thing I can think of is the design engineer monkeyed with the initial abstraction behind the scenes to fool his reviewers before, which I haven't checked yet. If so, that could be bad news for my client. Is there anything else you guys could think of that would account for the discrepancy? Anyone experienced this before between HydroCAD and Hydraflow? Hopefully there's a modeling parameter that I could switch in HydroCAD to make this work?



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

http://www.campbellcivil.com

 

[psmart]

The different CN values may account for some difference in the runoff volume. To check this you can adjust the CN units in HydroCAD to calculate an extra decimal place.

You should also extend the HydroCAD time span to 25 hours to match the Hydraflow study, otherwise you're truncating a small amount of runoff that occurs after 24 hours.

However, both of these adjustments seem to increase the volume discrepancy. Fortuntately, it's easy to evaluate the SCS runoff equation by hand and get the correct runoff depth of 4.9702" or 53,764CF = 1.2343AF, which matches the HydroCAD result. So the Hydraflow volume does look a bit low. I don't know if this is an issue with other parameters not shown in the report (such as the initial abstraction ratio) or a calculation problem.

The volume calculation will influence the peak flow, so resolve the volume issue first, before you evaluate the peak flow.

Any remaining difference in peak flow could be caused by:

1) Different calculation time step. Make sure HydroCAD is set to 2 minutes, to match Hydraflow.

2) Different rainfall distributions. HydroCAD uses a polynominal-based Type II curve that provides better accuracy for short Tc values, while the Hydraflow study is probably using a tabular curve. To investigate this you can download a tabular version of the Type II storm for HydroCAD.

3) HydroCAD reports an interpolated peak flow, which will be somewhat higher than the highest points on the hydrograph. To check this compare the tabular output for HydroCAD.

For a detailed guide to comparing SCS runoff results please see

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

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[beej67]

Thanks Peter.

I checked the CNs, and rounding it down instead of up doesn't really make that big of a difference. After I posted this I started fooling with abstractions, and found that I could match the peak for the hydrograph I included above by jacking the abstraction ratio up to 0.4, which is pretty ridiculous. Doing that brought some of the other hydrographs in the model closer to the right numbers, but not exactly. Unless the other engineer monkeyed with abstraction ratio for each basin individually, that's not the solution. Then again, it wouldn't be the first time I've seen this sort of thing happen.

I also adjusted the calculation time step to closer match hydroflow (0.03 ~ 0.0333333) but that only helped a little.

At this point I'm considering going to the reviewer with the hydrograph volumes as evidence that HydroCAD is doing it right and Hydraflow is doing it wrong, hoping he'll sign off on the as-built that way. Really hope he doesn't tell me to redo their hydro study after the developer has already put 1100 ft of storm drain vault in the ground.



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

http://www.campbellcivil.com

 

[psmart]

Frankly, the difference is so small that I doubt anyone adjusted the initial abstraction. In any case, I would adjust the initial abstraction only as required to match the volumes.

Once the volumes match, any remaining difference in peak flows must be due to the other factors I mentioned, including the rainfall table and peak calculation procedure. I'm pretty sure these issues will explain the remaining differences. But adjusting the initial abstraction any further to "correct" the peak flows will introduce discrepancies in the volume, so I wouldn't recommend that approach.

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[TerryScan]

beej67-

What is the criteria for the regulated flow rate? If it is based upon a pre development flow, the Hydraflow calculations for that may be similarly underestimated. Perhaps if one recalcs pre development in Hydrocad as well, your post development flows may become compliant.

 

[beej67]

Step ahead of you Terry - I checked that yesterday and it does indeed seem like I could do it fine, from a peak discharge standpoint.

That's likely the approach I'm going to have to take, but the particular municipality I'm working with have a reputation of being real bastards, and they might decide that I have to go back and retroactively fix the old busted design hydro study before I can even submit the as-built study, and my client doesn't have the time or the money for that. This particular municipality is also extremely unresponsive to engineer's questions. They're the type of people who like to make their job easier by letting their voice mailbox fill up so you can't leave them messages.

So any time up front I can spend to make this a blindly accepted punchlist sort of submittal will be worth it on the back end.


Off to download the tabular Type II storm for HydroCAD..



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

http://www.campbellcivil.com

 

[beej67]

Oh, one other thing, where is the setting that allows CNs to the tenth instead of rounding them to whole numbers? I can't seem to find it.

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

http://www.campbellcivil.com

 

[beej67]

Nevermind, found it buried under Settings-Units instead of the Calculation Settings button where I'd been looking.

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

http://www.campbellcivil.com