Field permeability testing

[bighouse]

I have been performing in-situ permeability tests in the field and have sporadically been getting permeability coefficients that have been inconsistent with the soil type.

The onsite soils are generally dense to very dense glacial end moraine and outwash deposits of gravel and sand with frequent cobbles, occasional boulders, and trace amounts of silt. Some of the specific USCS classifications based on laboratory testing and the corresponding coefficient of permeability values (Kv) are listed below. The samples submitted for laboratory testing were obtained using a 3” spoon at the test depth (ie run a permeability test at 10 ft and then sample from 10 ft to 12 ft).

GP – Kv = 5.9x10-5 cm/s
SP – Kv = 9.0x10-5 cm/s
SP-SM – Kv = 7.7x10-3 cm/s
SP – Kv = 4.2x10-4 cm/s
GW – Kv = 1.2x10-3 cm/s
GP-GM – Kv = 8.2x10-5 cm/s

The permeability tests are falling head tests run through a 4” casing flush with the bottom of the hole (ie we pound the casing to 10 ft, clear then hole out to 10 ft and then run a permeability test at 10 ft). The test procedure roughly follows ASTM D6391, method B.

In the tops of some of the samples we have been noting coarse gravel and cobble pieces which I suspect may be reducing the effective opening size (ie a 2.5” piece of gravel may be obstructing the casing opening). However, I find it hard to believe that such an obstruction would have such a dramatic impact on the permeability rate (reducing it by a power of 100 or 1000) unless the casing was sitting flush on top of a cobble or boulder which we haven’t been seeing.

Is it possible that this material could actually have such a low permeability? Or has anyone had similar trouble with this test or in-situ permeability tests in general? If so, what was the problem and how did you resolve it?

Any suggestions are greatly appreciated.

 

[dBasement]

If it was higher permeability than I would be more inclined to question the result, but low permeability would more easily be explained by the mere over-consolidation of the deposit. I have seen gravelly till soils that have much lower permeabilites than the above numbers just due to the cementation of those soils. Also, is there any other cementing agent, say a bentonitic clay in the matrix? Even a clay-based fine material will have a marked effect on soil permeabilites. I have also seen low-permeability, poorly-graded sand with a trace of "fines" The fines were clay as opposed to silt (very odd material). Another really obvious question, but one that must be asked...where is the static groundwater table?

I reviewed the ASTM method you referenced - not having done that test myself. My understanding is that the only wetted surface is the soil exposed at the bottom of the casing, is that correct? Are you pre-soaking? Since you are pounding the casing, I assume you can't meet the requirement that the hole diameter is 2" larger than the casing. If so, have you considered pulling the casing back a bit after you clean the casing? I don't see a reference in the procedure to methods A or B, my procedure is '06. I do see Stage 1 and Stage 2. Do you have access to an IPI packer system? It might be worthwhile to get a test done to back up your results.

Sorry I'm a bit stumped here so I'm grasping at straws.

 

[bighouse]

The static groundwater is definitely below the test depths as we have not encountered any groundwater in our borings. Based on the hydrogeologic data we have, it is about 60 ft below where we are running the perm tests.

I will have to look more into the cementation and clay-fines possibilities. They hadn’t occurred to me. Also, we aren’t performing hydrometer testing so our labs results are based on the assumption that all fines are silts.

You are correct, the only wetted surface is the soil at the bottom of the casing and we are pre-soaking for 30 minutes. We have considered pulling the casing back to allow for permeation through a large area and hopefully minimize the effect of potential coarse gravel, cobble and boulder obstructions; unfortunately, we are bound by a client specified procedure that doesn’t permit us to do so.

I am not familiar with the ’06 procedure. We are using the ’11 procedure, which doesn’t require the hole diameter to be 2” larger than the casing, but does suggest sealing the annulus around the casing with a bentonite paste. We are not doing this, but it seems like not sealing the annulus would lead to higher permeabilities. If the ’06 is similar to the ’11 procedure, we are performing the stage 1 testing (we are not doing the stage 2 extension and testing).

I am not familiar with the IPI packet system, but I have a feeling it may be cost prohibitive on this project.

Thanks so much for the suggestions and rapid response.

 

[oldestguy]

If you are able, use an excavator and run double ring infiltration tests at the same depth, ring diameters much larger than your casing. It might tell you something about the suitability of that casing type test. I suspect considerable compaction from that driven casing.