Oscillation when modelling Sand Filter

[SkinnyDip]

I am modeling a site and breaking it up into drainage area per BMP. I have a drainage area a little over an acre and am trying to model a sand filter treating this runoff. I have the sand filter modeled as 2 ponds. The first pond is the sedimentation chamber to allow settlement of large solids, then overflows into the second pond with the sand media allowing exfiltration. This sand filter is designed to allow storage to reduce peak flow. My problem is, I keep getting really bad oscillations with dynamic routing. Even after reducing dt all the way and increasing finer routing all the way I still get the oscillations. I get warning 87 and 80. The 80 says that pond2 exceeded pond1 by .01'. Is there any way to smooth this out? Any better ways to model a sand filter with sedimentation chamber and storage? Should I just not model a sand filter in hydrocad?

 

[psmart]

For details on modeling sand filters please see

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

If you have questions about a specific file please send it to HydroCAD support.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[SkinnyDip]

I basically have the filter portion set up like you link mentions. The problem I have is that I have another pond before it that acts as a sedimentation chamber. Both have the same invert elevations. Here is my attempt at attaching a picture to explain better. If you have a better way to model this please let me know.

 

[SkinnyDip]

How do I send a file to hydrocad support? I didn't see this option on the website.

 

[psmart]

You could model the forebay as a separate pond, with an overflow weir that feeds the sand filter.

You can send the file to support at hydrocad.net


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[SkinnyDip]

File sent. This is what I was doing.

 

[psmart]

Since the two sides of the trench are equalized by the connecting holes, they are best modeled as part of the same "pond". Just use a separate storage definition for each side of the trench. This will let you define a smaller volume for the side that is partially filled with sand.

When setting up the storage defintion for the side with the sand filter, be sure to select the option to "allow exfiltration". But deselect this option in the storage definition for the closed side of the trench.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[TerryScan]

SkinnyDip,

The image you posted is what is commonly called the "Delaware Sand Filter". I have used this device on a number of occasions. As far as modeling the hydraulics of this, I would like to offer a few notions.
1. The starting elevation in the sedimentation chamber is at the weir (there is no routing storage available below the weir as it is already full of water. There is negligable storage above the weir.)
2. The flow through the sand chamber and outlets is always more restrictive than the flow throught the weirs connecting the chambers.

When one tries to model this in detail, it creates a backwater condition between the two chambers which is causing the oscillations. Software that utilizes true full dynamic equations in routing may handle this without oscillations but that is serious modeling overkill and would unlikely yield significantly better results.

Because of this, my approach has been to ignore the sedimentation chamber and just model the sand chamber.

In the attached example:

The storage is defined for each precast chamber with a multiplier for the number of chambers. The first 0.5' of depth is gravel with a void ratio of 40%. Then next 1.0' depth is sand with a void ratio of 40%. Above that is open storage (one could use the volume of both chambers at this point).

The discharge in this example is perforated 6" pipe (0.5" holes) runing into an 8" collector/discharge pipe.