Utilizing Perc Rates for BMP Design

[kevgeotech]

When an in-situ permeability rate is required for the design of a BMP, can the in-situ perc rate be utilized instead? The apparatus described in the in-situ perm test maintains a constant head. The procedure for the septic drain field perc test (that we would utilize instead) does not. Are the two interchangeable?



Kevin P. Morrow, P.E.
Senior Geotechnical Engineer
D Miller & Associates, PA

 

[fattdad]

They are not interchangable. Additinally, irrespective of local requirements, I would argue that either can lead to problems. If you take an infiltration pond that's a 50 ft by 50 ft and you have a 4 ft thick horizon of sand underlain by clay the critical aspect of the "infiltration" will be the potential for horizontal flow away from the pond. Vertical flow will quickly become irrevalent. Vertical permeability will be misleading. Horzontal flow will govern and the flow gradients will be determined (esentially) by "flow to well" equations.

This is a HUGE flaw with the way most local agencies look at the design of infiltration ponds and the like.

Drainfields don't quite have the same problem.

f-d

¡papá gordo ain’t no madre flaca!

 

[cvg]

fd - what kind of testing would you propose to determine the true infiltration rate?

 

[fattdad]

I'd do a series of borings to understand the layered sequence below the site and also the location of the static ground water table. Let's say, you have a silty sand that extends from the anticipated bond bottom elevation to the depth of 6 ft and the water table is at the depth of 4 ft. Let's say below the depth of 6 ft there is some confining layer.

I'd install a ground water monitoring/observation well and I'd then do a falling head test, which primarily mobilizes the horizontal permeability. I'd make sure that the screen interval is properly gravel packed and I'd also make sure there is a seal about 12 inches over the screen (to make sure that the pressure head in the well does not rise up the gravel pack.

I'd use the Hvorslev equations to evaluate the permeability.

I'd take a sample of the sand and do a full sieve to get the D10 and D20 values. I'd correlate these to the permeability to see what the range I should expect from the falling head test.

I'd correlate the permeability to the anticpated value of "R" used in "Dewatering and Ground Water Control", which is an "Engineering Manual" provided by the Army (available on line).

Knowing "R", permeability and the anticipated head in the pond, I'd then consider an equivalent well array to replicate the function of the infiltraiton pond. So, if the pond is 50 ft by 50 ft (i.e., 2,500 sf), I'd assign an equivalent "well" radius of 16 ft. If "R" from the permeability correlation is 30 ft, I'd use 46 ft in my calculation (i.e., Well radius plus influence radius).

Now it's a question of what's the driving head? If you have a pond with a total depth of 6 ft when innundated at the design level, I'd likely run calculations for both 6 ft and 3 ft. and see just what the "yield" would be. Bearing in mind that all well equations work for withdrawal as well as injection. For this case, it's an injection scenario.

I just want to point out one item: Much of the "thinking" from infiltraiton pond design is rooted in the design of drain fields. I don't believe this is correct. If the requirements are for a 4 ft separation between the pond bottom and the water table, with a void ration of 0.25, one foot of infiltration would fully saturate the soils above the water table. The only reason this is not the truth is if there is lateral flow. If you are trying to fully infiltrate 4 ft of water, then you are really relying on lateral flow to get the work done, which means that the system has created a "ground water mound" during the time of use. The form of a ground water mound is fundamentally related to a log curve, which is the basis for the well equations. So, using well equations to assess this problem is rational. That said, I welcome any comments.

Maybe this is better served with a sketch?

Hope this helps.

f-d

¡papá gordo ain’t no madre flaca!