Relationship between density and permeability

[TigerGuy]

Is there an empirical relationship between a soil's permeability at a range of densities? I can't find one and am hoping somebody out there could point me in the right direction.

I have a soil with permeability 1.5E-8, at a dry density of 121 pcf and 14% moisture. Liquid Limit = 37, Plastic Limit = 13, 35.5% sand, 35.5% silt, 29% clay. In place density tests at 111-115 pcf.

So, without running additional tests, is there a correlation to indicate what the in place permeability might be?

 

[oldestguy]

Look up the revised editions of the USNavy Design manuals. My old copy of NAVFAC DM-7, March 1071 in Figure 8-5 show for both 25% silt and also for 25% clay curves at about 10E-8 ft./.min. Table 9-1 give about 10E-7 ft./min for densities ranging from 70-120 pcf dry, depending. One ft./min is 0.5 cm/ sec. You would be plenty safe, since what you have is darn tight.

Try US Dept of Interior Bureau of Reclamation. They do a lot in this category and their manuals may help. My old "Earth manual" is vague.

 

[EireChch]

With such a minor change in density, I wouldn't bother looking in to the possibility of increasing the permeability due to the soil being less dense (if thats what you are attempting). That is provided that you have undertaken a suitable number of appropriate tests to get a representative permeability of that layer.

For me one test would only be suitable for preliminary design.

If you estimated your permeability using different tests (i.e. soakage pit, packer test, falling head in a borehole etc.) you would definitely get a range of permeability. This is where you need some sound engineering judgement.

 

[fattdad]

permeability of compacted soil is influenced by the saturation of the sample at compaction. So, a sample compacted at 95 percent dry density on the dry side of optimum has a greater permeability than a sample compacted on the wet side of optimum. The difference can be 2 or 3 orders of magnitude! Please recognize that the line of optimums is not a vertical line. Please also recognize that for any compacted point you can scale the degree of saturation from the Proctor curve. We'd typically specify 95 percent compaction at or greater than 90 percent saturation.

Refer to Mitchel, Hooper and Campenella, ASCE, Vol 91., No. SM-4, July 1965, pp. 41-65.

f-d

ípapß gordo ainÆt no madre flaca!

 

[TigerGuy]

The pond is completely cut, no compacted fill. They are allowed 1E-7 cm/sec. That's where my question originates. Since the shelby tube samples gave such low results for permeability, and now the in-place soil is exhibiting slightly reduced density (equipment traffic shearing the surface layer, I think), I need to know whether they need to scarify and recompact or just let it ride.

Of course, we have a reviewing agency that needs the "box checked" before they will approve filling the pond. I didn't anticipate the reduction in density in cut, compared to the boring samples. So, here I am trying to reverse investigate, ugh.

Permeability was determined by ASTM D5084, flex-wall. Other samples had results around 4E-9, with nearly identical density and grain size curves.

I think I'll start looking closely at the saturation values, and checking the suggested references.

 

[fattdad]

Please also recognize that consistent values require constant B values during triaxial saturation. Blebs of air tend to reduce the permeability. So, if the B value is not at or near 1, you'll get lower values for permeability.

Also, the testing gradient can be a factor. May want to check the gradient?

f-d

ípapß gordo ainÆt no madre flaca!

 

[MEM1]

What area are you located in, and what is the pond for? In Kansas we have to prove the pond will pass seepage requirements by pulling Shelby tubes post construction.

 

[TigerGuy]

F-D: We had B values between 0.95 and 1 on all perm tests, with 3 psi gradient thru the sample. After multiple runs, k20 values were consistent and repeatable.

MEM1: We are neighbors. Project site is in Missouri.

I've found a conversion formula from DW Taylor in "Fundamentals of Soil Mechanics" published in 1948.

log k = log k0 - ((e0 - e)/(0.5 e0))

I think by using the Shelby tube sample void ratio, and the calculated void ratio based on the in field density and moisture, I can utilize this formula to satisfy everybody.

It's one of those situations that you know it will work, just by looking at the material. However, providing the supporting documentation simplifies everybody's life.

 

[fattdad]

everything correlates in log-log relations!

f-d

ípapß gordo ainÆt no madre flaca!

 

[TigerGuy]

Isn't that why the log scale was invented?

Thanks again for the input

 

[GeoEnvGuy]

Density has nothing to do with hydraulic conductivity for that type of soil. Density only influences the porosity which is used to estimate linear velocity. This document can be downloaded and has equations on how to use KC to nonplastic and plastic soils based on index tests.


PREDICTING THE COEFFICIENT OF PERMEABILITY OF SOILS USING THE KOZENY-CARMAN EQUATION

January 2003

 

[fattdad]

GeoEnvGuy,

Please refer to my citation of Mitchell, Hooper and Campenella, read it then make your claim again.

It's quite likely that two compacted identical soils at the same density have completely different permeabilities. I don't argue that Kozeny-Carman don't give you a basis. I'm just reacting to the certainty of your statement. It doesn't tell the whole story.

f-d

ípapß gordo ainÆt no madre flaca!

 

[GeoEnvGuy]

Hey Fattdad, i looked over the paper and it deals with compacted clays and not natural clay deposits of varying density, I believe the problem above deals with a natural clay deposit using complete cut and not compacted clay. The paper does show the variation in the lab permeameter test of compacted clays at various moisture contents below saturation. But I don't see how a natural clay deposits variation in density affects the overall hydraulic conductivity more so than atterberg limits or grain size variation.

 

[theCorkster]

I recall running flexible wall permeameters on Bay Mud some years ago. The notion that the permeability of a clay is not influenced by the void ratio of the clay is incorrect. The perception may be that the relative change is small (over of magnitude) and therefore not significant.

 

[fattdad]

Okay, the OP talked about dry density and I extrapolated that to some understanding of compacted properties. In the realm of in-situ permeability, I'd run borehole tests.

That's a different and also fun conversation - the best method for borehole tests!

f-d

ípapß gordo ainÆt no madre flaca!

 

[TigerGuy]

This is relatively undisturbed soil. I agree that borehole tests are the best, but how do you find clients willing to pay for them and wait for results?

We pulled samples during the investigation and tested them in the lab. Using the results, recommendations were given for construction. After excavation, not all density readings were as high as expected.

That's the challenge I found in front of me.

On a side note, I have a horrible time hitting target moisture and density on recompacted samples in the lab, but that's a different conversation