Saturation of clay below a water table

[kcall2]

Gents,

A site in an alluvial deposit has a water table ranging in depth from about 18 to 24 ft (based on boring logs). The soil from a depth of 10 to 30 ft is a clay, generally consistent in color (munsell), water content (25 - 35%), dry density (85 - 95 pcf) and fines content (75 - 95%). Below 30 ft is a stiff sandy lean clay. Can a clay below a water table have unsaturated zones, or at least appear unsaturated?

The adjacent property was dewatered to a depth of 40 ft during conctruction of an underground parking lot using dewatering wells. Two years following construction the adjacent site was drilled using a hollow-stem auger boring. Measured depths of ground water table (prior to dewatering and about 3-4 years later) were consistent. However, upon examing samples below the water table they did not appear saturated. Using an assumed specific gravity for each sample the degree of saturation ranged from 90 to 100%. It is odd considering the dewatering contractor on the adjacent site had problems keeping up with the rate of water flowing into the wells, and that it was dewatered to 40 ft.

Thanks for any comments.

 

[fattdad]

There is no guarantee that clay below the water table is saturated. Occluded blebs of air are often present and it takes water pressure (i.e., depth below the water table) to dissolve these blebs. Consider how high of backpressure saturation is often necessary to saturate a clay sample in the triaxial cell.

Regarding water table observations when drilling through a clay layer, you can easily remove clay faster than water can enter the boring (i.e., if you remove 20 gallons of soil in an hour and the water table only flows into the boring at 5 gallons per hour, it may appear that you encountered water in the last few feet of a boring that is well below the phreatic surface.

Consider a sand layer in the near surface: Often folks will claim a perched water table even when the sand and the clay share the same phreatic surface. Short-term field observations may indicate that the clay does not have a phreatic surface as standing water may have been seen in the sand layer. It really takes a piezometer to figure this stuff out.

Regarding the dewatering contractor: There may be water bearing sand layers or some other such condition that would provide for water flow from something other than the clay.

Good luck.

f-d

¡papá gordo ain’t no madre flaca!

 

[Ron]

As fattdad noted, clay below the water table doesn't have to be saturated...in fact it is often not saturated, particularly "fat" clays.

Groundwater levels should be check at significantly later times than 24 hours, when dealing with clayey materials.

 

[BigH]

I would say that for clayey materials - especially fat clays - that 24 hour later observations are way too short in time interval. You may have a 6 inch hole to fill up 20 ft - that's a lot of water . . . . I would take a reading, for sure, at 24 hours - but if possible, at 72 and then at a week. Plot the curve.

 

[kcall2]

Thanks for the input. I also suspect sandy layers somewhere in the profile but do not have any data to support that. Even the log for the adjacent parking garage excavation backcut mentioned no sand.

Actually the water readings were easily approximated during drilling (wet rod, sampler) - there was no option to leave the holes open for an extended amount of time - so I am confident in the location of the water surface.

So if fat clays are often not saturated then time-rate consolidation curves are not applicable. I ran several time-rates over a range of depths (including samples that appeared not to be saturated) and each curve had the classical shape with obvious transition from primary to secondary consolidation.

 

[fattdad]

So if fat clays are often not saturated then time-rate consolidation curves are not applicable.

Don't agree with that deduction. Consolidation will require that water moves around soil grains and blebs of air. Do the consolidation test and interpret the results. This should reflect field conditions.

f-d

¡papá gordo ain’t no madre flaca!

 

[Ron]

Consolidation results from a change in void ratio. Saturation is not necessary to develop a time-rate of consolidation curve.

 

[kcall2]

To determine the coefficient of consolidation, that is true. However, Terzhagi's 1-D consolidation theory is built on two basic assumptions: 1) the soil is homogenous, and 2) the sample is saturated. If the sample is not saturated you can not used the effective stress principle, which equates the change in effective stress to change in pore pressure.

 

[kcall2]

...determine the coefficient of consolidation.... I meant the compression index.

 

[BigH]

kcall2 - from memory, most consider the B value (saturation level) of 94% to be "saturated". I think that you might be getting a little too theoretical. If the "saturation" level was 99%, would the theories not then work? And how did you determine the level of saturation? Using Gs? If so, what if the Gs of the soils vary a bit from spot to spot in the stratum - 94% in one zone might be 98% in another . . . I think that if you you are under the stabilized groundwater level, you can use the theories with some confidence - especially when, as we all know, if you estimate consolidation settlement to within 30% or so, you are having a good day. If you couldn't leave the hole open, why didn't you install a standpipe? That, in our practice, was always the norm.

 

[fattdad]

At the risk of repeating myself, if you have interconnected air voids then the soil will behave as unsaturated. If the air voids are not interconnected then the change in void ratio will be affected by the permeability of the soil mass (Mv). The shape of the time curve can tell you about this, just like it can tell you whether you are in the preconsolidation range or the virgin consolidation range. If you have a time curve that shows an "end of primary" and a c-alpha, then what's the question? You have what you need to determine the time rate of consolidation in a field condition.

f-d

¡papá gordo ain’t no madre flaca!

 

[BigH]

fattdad - okay - you have interconnected air voids below the water table. What is keeping, over a long time (and I suppose it might be) the air from being pushed out ever so gently. Unless it has something to do with threshold gradients. I still think that measurement errors of Gs might have something to do with his saturation levels.

 

[fattdad]

Two issues: The accuracy in which you measure saturation (to BigH's point) and at what saturation level does a sample behave as "saturated." I completly agree that B values do not reflect saturation levels (i.e., a B value of 0.95 does not indicate 95 percent saturation). I'm just saying that when you have a 1-D consolidation test that depicts an "end of primary", then the behavior of the sample is influenced by permeability.

If you take a clay sample from the ground and put it into an odometer, cover it with water and run a consolidation test, is the sample saturated? Doubtful. Will you get data that you can interpret? I think so. I think you'll see classic time curves and a classic e log p curve.

I expect folks may chime in with some experience in a desert or such and maybe then I'm off base, but in the parts of the world where I've worked, other than some unusual situation, this is what I'd do.

Not that I'm always right, that is - ha.

f-d

¡papá gordo ain’t no madre flaca!

 

[kcall2]

Fatdad - I do have good data, however I am attempting to evaluate settlements that may have been caused by dewatering as part of a forensic study - just trying to avoid taking a 1-D consolidation settlement analysis on potentially unsaturated samples (again based on visual observations only) to an expert witness.

 

[BigH]

Can I enquire about the undrained shear strength of your two clays?? What was the water content and Atterbergs of the lower stiff clay (below 30 ft). I have drilled clayey tills in the past - when you open them up, they don't appear wet by occular inspection (as opposed to tactile) - but they definitely were saturated although very stiff.

Getting back to the original post - How long was the dewatering going on for? Days/weeks/months? How far away was the dewatering wells? - deep wells? or staged well points? If you had such tight clays, why was dewatering even necessary? Did you have sand seams or lenses?? I remember a job where the contractor talked his way into well points but it wasn't really necessary as the sands were only pockets with limited amounts of water. Obviously, you are trying to determine if the dewatering caused cracking of the structure nearby?? Any condition survey before dewatering/excavation? Don't forget that you will have surface settlements due to the excavation even if you put in "tight" excavation support - see Peck's book - he has a graph of depth of excvation and settlement - three zones. I'll try to find it later. This might be the cause moreso than the effect of dewatering. One would need to know the coefficient of permeability and try to determine if the dewatering could actually lower the water table in the clay - but you have indicated your clay to be stiff which means it would be somewhat overconsolidated - then, any settlements due to dewatering would likely be within recompression. - Just some additonal things to think about - which most likely you already have.

 

[fattdad]

What did they dewater the sandy lean clay below the depth of 30 ft? If this is the case, then it would be the consolidation characteristics of that layer that could be the relavent parameter. It may be possible to get some measure of dewatering from the underlying more "sandy" layer and not realize sufficient time for the overlying "clay" to respond. As a result, you'd change the state of stress in the sandy layer, which would feel more of the burden from the overlying clay.

I am very interested in the OP's response to BigH's points as I share them. There may be elastic properties (and inadequate shoring design) to consider.

f-d

¡papá gordo ain’t no madre flaca!

 

[kcall2]

BigH - undrained strengths of 0.4-0.6 (20-27 ft) and 0.8-1.1 (>27 ft) based on mini torvane at end of shelby tube. Moisture content ranged from about 28-36% (20-27 ft) and 20-25% (>27 ft). No limits yet but suspected to have LI of about 0.5. Even some soft clay samples appeared unsaturated.

Still waiting for additional dewatering info (different consultants for underground structure). Basically point wells were used for months at 60' spacing directly behind pile-supported excavation (with tie-backs) to depth of 40 ft. 26 ft excavation ~7' from existing adjacent building. Not sure why dewatering went to 40 ft. Agreed - it does not appear that dewatering would create enough additional stress to push it over onto the virgin side of the consol curve.

Inadequate shoring also being considered. I would be interested in the Peck reference BigH.

 

[DRC1]

Dewatering is very difficult to model acurately (despite all the very colorful software currently available). That is we like to model soils as a rather unifom material, when in fact it is not. This does not have a huge effect on most geotechnical design, but dewatering is highly affected. Say the fat clay is a 10-7 permiability, yet there is a silt layer in the excavation that was 10-5. The borings may have missed it (especially if standard sampling was used) or may may have had been noticed. Clays grading in and out of silts are often missed. This would cause more infiltration tan anticipated. This layer may have been removed or altered in construction.
Generally, less permiable layers underlying more permiable layers are considered to be saturated.
My guess is that most of the recharge is from a rising water table. The material is very impermiable and the water table has not recovered from its depressed state. In tighter materials, stabilized piezometers provide reliable and usable water table data. Driller observations and bore hole measurments at completion of drilling are not usable data.

 

[BigH]

I have attached Peck's chart as found in Bell, Ground Engineer's Reference Book; and a few graphs from MJ Tomlinson's Foundation Design and Construction, 6th edition. I strongly suggest that you obtain a copy of the Tomlinson book - he has an extensive coverage of movements next to excavation - some 34 case histories he looked at and has the data given in an appendix. (Also, the book is damn fine book!) - hope I haven't offended the authors!

 

[fattdad]

to the OP:

For a dry density of 85 pcf, saturation moisture content (Gs=2.7) is 36.5 percent (i.e., pretty close to your 35 percent). For a dry density of 95 pcf, saturation moisture content is 28.5 percent.

Your data seems to indicate that you are seemingly close to saturation moisture contents.

f-d

¡papá gordo ain’t no madre flaca!