Bioretention Oscillations

[bassfshng8]

I am modeling an existing watershed to compare the volume of water from pre to post. I have the watershed subdivided into 3 drainage areas. The reason I need to subdivide the watershed is because of proposed changes to the existing bioretention and an additional proposed bioretention. I need accurate water surface elevations and peak flows for the existing bioretention.

The problem is the 12" terracotta pipe that the bioretention outlets into. During large storms the 12" pipe reaches its capacity. I have attempted every reach/pond routing method and have failed to remove the oscillations.

If I use the Sim-Route method, the oscillations are gone but I get multiple errors related to tailwater conditions.

Any help will be greatly appreciated! Thanks

Frank Miller
Century Engineering

 

[psmart]

Several comments/suggestions:

1) Since you have tailwater effects, you need to use DSI routing instead of SI. Do not use SimRoute except for very special purposes. For details see

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

2) If the finer routing doesn't solve the problem, set it back to the default of 1 and explore other solutions.

3) Oscillations can be triggered by abrupt changes in a rating curve. (stage-storage or stage-discharge) In this model, both curves jump abruptly as the water rises above the stone.

Please review your approach to modeling the bioretention. The outlet setup may be the issue. For example, what does the 34x18 orifice represent? Is there a soil layer in the pond, or just stone with 40% voids?

For further details please see

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

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[bassfshng8]

I have attempted the DSI method, oscillations still occur. I have attempted the Sim-Route method, oscillations still occur. I have attempted to adjust the finer routing using 1, 2, 3...all still producing oscillations.

The 34x18 orifice is the top of a catch basin with a trash rack. This is acting as an emergency spillway for large storm events. The 6" underdrain is at the bottom of the catch basin acting as the low flow orifice.

The bioretention is filled with a sand/peat/mulch mixture only. I was told by the reviewing agency to assume 40% void space. At the bottom of the layer is the 6" underdrain that is surrounded by stone, I did not include this area in the storage because of the small size.

This is an existing system so I cannot change any orifice or grate sizes.

I believe I am going to get reverse flows. I have the 12" TC pipe flowing into a 24" RCP. This is causing the 12" pipe to get backed up during larger events. Again these are existing conditions, I cannot change the pipe sizes. Because of possible reverse flow I have tried the Sim-ROute method as well with the same results.

 

[psmart]

Please stick with my routing suggestions. Messing with this randomly will just waste your time.

40% is the maximum voids you can expect from clean stone. The bioretention mixture will be much less - maybe 20% at best.

Since you have a soil (media) layer above the stone, the media is likely to be the outlet control, rather than the underdrain itself. This means that (1) You should use an exfiltration "outlet" to estimate the flow through the media, rather than (just) the 6" orifice, and (2) your storage can only include the volume that lies above the media (since this is the outlet control).

You will still have the overflow and final 12" pipe, which needs to be part of the same pond node, and NOT a separate node: Here is the outlet setup:

Device#1=12" Culvert outlet barrel, Routing=Primary
Device#2=6" orifice, Routing=Device#1 (to the 12")
Device#3=Exfiltration, Routing=Device#2 (to the underdrain}
Device#4=Overflow orifice, Routing=Device#1 (to the 12")

This is significantly different than your original model, in which you had the overflow going through an 8" pipe, rather than directly to the 12".



Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[bassfshng8]

The existing outlet structure is a 34"x48" precast catch basin. There is a 6" underdrain that is at the bottom of the bioretention basin. All water that enters the bioretention area will reach this underdrain. Large storm events fill the bioretention area and then enter the top of the catch basin. There is an 8" pipe that leaves the catch basin and is 56' until reaching the 12" pipe. The 12" pipe was an old system that a juction box was installed when the bioretention was created.

I don't think I can remove the 8" orifice. I also agree with you about the void space, but the direction from the reviewing agency was to use 40%.

I don't understand why my storage can only include the area above the media? If we agree there is void space within the media layer, then there will be storage correct?

 

[psmart]

When you use a pond routing, you're assuming a level pool of water controlled by specific outlet devices. So you need to clearly define the storage volume and the outlet(s). This means you cannot include any storage that is downstream of the outlet(s). Since the flow to this void space is controlled by the flow rate through the media, it cannot be part of the level pool. Hence my recommendation that the pond storage should end at the media layer.

I suggest you recheck what the 40% applies to. This is a maximum value for clean uniform stone, so when you add any mixture of material the value will only decrease. In any case, since the media layer will be an outlet control point, the storage in the stone voids below shouldn't be part of the model.

If the entire structure will discharge through the 8" pipe and then to a 12", you can adjust the outlet setup to include both pipes:

Device#1=12" Culvert outlet barrel, Routing=Primary
Device#2=8" Culvert, Routing=Device#1
Device#3=6" underdrain orifice, Routing=Device#2 (to the 8")
Device#4=Exfiltration, Routing=Device#3 (to the underdrain}
Device#5=Overflow orifice, Routing=Device#2 (to the 8")


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[beej67]

Personally, I wouldn't include any volume OR flow through the media in the routing analysis. There's no guarantee it'll be dry from the previous storm when the next one comes, given how these bioretention things have a tendency to clog, and given how few people maintain them properly. I'd model the whole thing as a surface pond and be done with it.

Down my way we typically don't try to claim any sort of detention from the things, just water quality treatment, and use simple spreadsheets to prove the design. Then do our detention in "proper" ponds.


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

http://www.campbellcivil.com

 

[psmart]

I tend to agree with beej67. Even if you allow for flow through the media, you'll often find that it's too small to have any significant effect on peak flows. Sure, you'll need to make sure you can handle peak flows through the structure in an acceptable way, but the media will serve mostly for long-term draw-down rather than peak management.

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[TerryScan]

Frank,

I see a couple of things in your model that need attention:
1. The calculation time span extends too long for no reason. 30 hours seems to cover everything.
2. What is Device #2 on SD 20? It appears useless in the model.
3. The oscillation problem:
I explored your model to isolate the oscillations. They existing even if one disconnects BIO from everything. I noticed your outlet is below storage. Since the storage is 0% at the bottom anyway, I changed it's invert to match the culvert (96.07 instead of 97.0). It appears that elimnated the harsh oscillations.
4. What is the top elevation of the CB structures? You show a 200ft flood elevaton that appears arbitrary. When elevations exceed the top (100yr?) your elevation shoots up a column rather than spreading over the ground surface. This can create false high head and promote other oscillations in the system.

 

[beej67]

One trick I've used to account for dead storage in multiple bioretention facilities within the same overall watershed, is pro-rate the initial abstraction fraction to include the capture of the bioretention facilities. Basically shaving some volume off the front end of the hydrograph.

Definitely clear that sort of approach with a municipality first though, because that's a modeling knob you're usually not supposed to fiddle with.

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

http://www.campbellcivil.com