Best way to model an underdrain?

[eschelle]

I am wondering about the best way to model a pond underdrain. The set up I am thinking of would involve a perforated pipe buried in gravel at the bottom of the pond, connected to an outlet riser and culvert. There are a few ways that I could see this modeled:

1)The “model everything approach” i.e. model the perforations as orifices which are routed to the pipe, which is then routed to the final outlet structure. I am usually dealing with existing infrastructure, and the level of detail required to model this accurately is usually not available. Also I have learnt that it is generally better not to make the model any more complicated than absolutely necessary.

2)Alternately, I could assume that the perforations will not be a limiting factor and simply model the underdrain as a single small orifice in the outlet riser. This begs the question about what the size of my orifice should be - the diameter of the perforated pipe? I have seen a couple models where a reduced orifice size was used, but I am unsure about how that size was reached.

3)The most common approach I have seen, however, is to completely ignore the underdrain when modeling ponds, which is not something I am exactly comfortable with.

Your assistance is much appreciated.

 

[LHA]

1 is overkill, but will work fine.

2 and 3 are flat-out wrong, and I would reject calcs with either of those suppositions.

Does the pipe have a filter fabric around it? It should, and if it does, the permittivity of that will limit. That will be available in the filter fabric produce specs, although you will need some unit conversions.

The "gravel" - depending on its gradation - should also limit, more than the perforations. Standard permeabilities are readily available; again, will require some unit conversion.

Engineering is the practice of the art of science - Steve

 

[psmart]

You need to include whatever components might control the discharge. Setup the device routing carefully, such as:

Device#1=Culvert, routing = primary
Device#2=Orifice (top of riser), routing = device#1
Device#3=Culvert (underdrain pipe), routing = device#1
Device#4=Orifice (perforation), routing = device#3
Device#5=Filter fabric (?), routing = device#4

Note the "chain" of devices for the underdrain. 5->4->3->1

The pipe performations could be modeled as an orifice array (to allow for varying heights) or perhaps a single orifice with an appropriate multiplier.

 

[eschelle]

Thank you psmart and LHA for your comments. I agree that it is best to model all the discharge influence. I mentioned in my original post that I am dealing with existing infrastructure. More specifically I am looking at systems that somebody else designed and installed, in many cases a decade or two ago. I am trying to model these systems to see how far they are from meeting our current standards. The data I have to go on is usually comprised of some old maps and TR-55 or hand-written calculations. Although some of the plan details show that there is an underdrain, never do they give the level of detail that is really needed to model it properly. (i.e. is there filter fabric, what are the size/number/orientation of the perforations, etc.) I field check these sites, too, but I cannot see what is underground. Obviously, if I visit the system during dry weather and there is water in the bottom of the pond, I assume that the system is clogged and now effectively has a permanent pool. I was just hoping there was a work around to take into account that the water stored at the bottom of the pond is going to drain very slowly, even if I can't find all the specifics.

 

[psmart]

I'm not exactly clear what question you're trying to address with your HydroCAD model...

You can certainly make adjustments to the model to get a better simulation of what is actually observed. For example, you can simulate the standing pool by setting the outlet device above the bottom of the "pond".

 

[eschelle]

I know how to model a standing pool, but I was looking to see if there is any way to model the slow dewatering of the pond below the bottom most orifice (say on a riser) that is only affected by the underdrain. I am trying to do so as accurately as I can given the information I am reasonably able to obtain, which is not ideal. Perhaps I should just err on the conservative side and assume that the underdrain does not work, and model anything below the bottom orifice as a permanent pool. I have visited some sites where the system is still working (the bottom foot or two of the pond are below the lowest orifice on the riser, and yet the pond does drain, albeit slowly) and would like to take that into account if possible.

 

[psmart]

Can't you just use a smaller outlet at the bottom to simulate the long-term dewatering? You'll have to adjust the outlet device to calibrate HydroCAD against your on-site observations.

But this still begs the question, why do you need to model the long-term dewatering? As you point out, this is hard to predict, and may vary considerably. In terms of stormwater runoff, you can assume best case (complete dewatering between events) or worst case (permanent pool) depending on the overall design issues.

 

[eschelle]

Thanks psmart for your comments. I guess I need to step back and consider what I am really trying to accomplish with my modelling a little more and how an underdrain alters how much credit I can give these old ponds towards meeting current requirements.

 

[pjhealy]

Sure, you can do it easily. Observe the pond to see how long (time) the pool (known volume) takes to empty. If it's very slow, measure what you can: depth of drop x known surface area at the elevation/1 hour. Either way, you have volume/time=Q. Create a custom outlet or secondary outlet with this known flow rate.