Sunday, January 27, 2013

Designing and Repairing Slab Foundations built on Yazoo Clay

by Britt Maxwell P.E.



The Yazoo clay is an Eocene epoch expansive clay deposit that outcrops in a tilting band through Central Mississippi. There is an interesting cosmic connection with the Yazoo clay that I think you will enjoy, but I will leave that for the end of this blog. There are a lot of ideas about construction and Yazoo clay so I will use the most reliable information available (data subject to peer review). Note that because this blog has been very popular I have updated it (October 2017).  I start with four myths about slab foundations built on Yazoo clay.

·       If there is 7or 8 feet of select fill over the top of the Yazoo clay, it is believed that the expansive soils have been effectively mitigated. FALSE: Not even close (I explain below).

·       It’s a good idea to water your foundation. FALSE: This may work in other areas of the country, but in general it’s a bad idea with the Yazoo clay. There are a few exceptions but, in general, drainage improvements are a better strategy.

·       Underpinning your foundation will stabilize it. FALSE: Sometimes this is true, but if a foundation is heaving from expansive soils, underpinning will not stop the movement. The key point is that expansive soils can cause an upward movement (heaving). When these movements are serious, a different approach to foundation repair is needed.

·       Some foundations cannot be fixed. FALSE: Any foundation can be repaired if you are willing to spend the money for the necessary repair. A qualified structural engineer can if necessary, design a replacement foundation, if the original foundation was not suitable for the soil conditions or the site. See my Example Damage Blog with this photo: 

The first myth is a very understandable because 7 to 8 feet of fill is commonly specified (locally) for the construction of slabs for residences and small buildings. Locally, a layer of soil above the Yazoo that is not expansive is called a “buffer.” It is also called “the cover.” If the cover is manmade then it is select fill otherwise, it is just a natural deposit, usually silty clay. Based on foundations designed in Mississippi the 1950's and the 1960's that I have seen, some engineers of that era also misunderstood the real effect of the buffer. But there was very little data to make this determination. Foundation designs have definitely evolved in response to an improved understanding of the real behavior of Yazoo clay.

The buffer is very important but it does not remove the threat of Yazoo clay unless it extends much deeper. There is adequate evidence from detailed published studies of foundation failures that show this (see reference "Field Measurements of Yazoo Clay...).  The actual purpose of the buffer is to reduce the differential movements in a slab foundation created by the expansive soil to within acceptable limits. When I say a slab foundation, I mean a typical stiffened slab (with beams built into the slab) without deep drilled piers or piles. The use of a buffer is a legitimate strategy for the design of shallow foundations on expansive soils. Locally the depth of the buffer that engineers have specified has changed. In the 1970's and early 80's we typically saw 6 feet as the required thickness of the buffer. In the last 15 years or so we now often see 7 or 8 feet. For slabs built more than 18 inches above grade, a thicker buffer is needed. The same applies to slabs built below grade.

If you want to negate the primary effect of Yazoo clay you need about twice the depth of the typical buffer that we use today. The depth of buffer material needed to remove heave potential is called the active zone. At one site in Madison County, the depth of the active zone was 12.7 feet. At two other sites, it was deeper than that, because related movements were detected where the clay was at that depth. There are different ways to determine the active depth of an expansive soil, but I like to rely on field measurements of movements as opposed to theoretical methods.

An important fact is that you can not always count on a buffer of 7 or even 9 feet to keep foundation movements in an acceptable range. Do we accept a small percentage of buffer failures or do we analyze these failures and try to prevent them? I wish I could report that this issue has been completely solved, but it has not. But we are beginning to get a handle on these failures and it is clear that there is a higher risk for foundation movements when the buffer strategy will be used in the following circumstances:

·      The larger the floor footprint area of the structure, the higher the chance for a problem.

·      The more complex the floor plan, the higher the chance for a problem. This is now a part of the Post Tension Institute design procedure. The simplest plan is a square; the more complicated are long narrow wings. The more complicated floor plans will actually kick you out of the PTI design process.

·       Hillside sites that are terraced to create a level lot. The highest risk sites often have retaining walls or should have been built with retaining walls because of the steep slope. With these sites maintaining positive drainage away from the structure on the uphill end of the structure is critical. There are countless foundations failures where the surface drainage on the uphill end of these structures was inadequate. One issue often ignored is that the yard at this location may heave (degrading the slope). So if you construct minimum drainage in a few years it could be gone.

·      There seems to be a higher risk when the boundary between the buffer and Yazoo clay is natural. The alternative is to overexcavate and cut into the Yazoo and then replace the natural buffer material creating a manmade inclining boundary above the Yazoo clay that will shed percolating water.

There are other aspects but all of these risks have a connection with each other. That is, with each one of these situations there is an increased risk that water at some point around the foundation will work its way down to the expansive soil. This is clearly the critical issue. Once water infiltrates down to the clay there becomes a potential for upward movement or heave at that location. In the worst case a perched water table develops at that location. A natural surface for the top of the Yazoo makes this more likely because there can be subsurface pockets for water. A perched water table is to Yazoo clay like a match is to gasoline. A perched water table is a water table that develops in the buffer material above the Yazoo clay. The permeability in the clay is very low, so water will stand on the surface of the clay like water in a pond. Unfortunately this condition could exist on a site and not be discovered by the soil borings. If the soil above the Yazoo also has a low permeability then water from the water table might not flow into a boring hole until it has been left open and monitored. The situation is complicated because the surface of the Yazoo clay is an ancient eroded surface that is uneven and might contain old buried stream channels. If you have a serious foundation problem there is a very good chance it is related to a perched water table. As a result we have to look closely at all the complex elements that produce these conditions.

Field measurements made so far have given us a hint of the kinds of movements that are possible with Yazoo clay where there is an established water table. Unfortunately there is not enough solid data for us to understand how much variability exists. We recently measured 17 ½ inches of differential heave in a slab that was constructed during the mid 1970's. In this case the Yazoo clay was only about 4 feet from the top of the slab and there had been apparent problems with plumbing leaks for years. The historic Manship House was releveled in1980 but by 2010 it was 13 1/2 inches out of level. At another site we were able to compare the rate of heave over 5 years over a large area where the depth of the clay varied from about 7 feet to about 14 feet. The rate of heave at 7 feet was just over ¾ inch per year using regression on the data. The rate of heave when the buffer was 9 feet was ½ inch per year. If we apply this movement at a single spot to a residence and we define a failure to occur when it is 5 inches out of level, then failure occurs at about 6 ½ years with 7 foot buffer and failure occurs in 10 years with a 9 foot buffer. In these scenarios there is the assumption that there is only one point under the structure where water gets to the Yazoo clay. Fortunately this is rare, but it does happen. This shows us how critical a plumbing leak or poor drainage can be.


 RECOMMENDATIONS FOR HIGHER RISK STRUCTURES:

1. Make at least 3 soil borings roughly inline along each segment or wing of a structure to see how much variability occurs in the geology. With only 2 borings we assume a steady transition between the borings and we never see the real transition.

2.  If the expansive clay is found above the active zone in a boring, 1 or 2 borings need to be left open at the low elevations of the site (but in the footprint of the structure) for a long enough period to determine if a natural water table exists. Some of the local geotechnical engineers already do this.

3.  Constructed surface drainage can severely degrade due to expansive soil movements, erosion and other factors. This means that you should compensate for this with more initial runoff slope in your drainage system. We already knew this but there is added emphasis.


Now it’s time for my cosmic connection. The picture on the left is the south polar region of Mars, taken by the Mars Global Surveyor. The picture on the right is Google Earth image of a field 6.7 miles northeast of Waxahachie, Texas. The origin of the Martian polygons are still being debated, the Texas features are gilgai landforms also called hogwallows that formed in expansive clay. There is a big difference in scale of the pictures. The Martian photo shows an area of 1.9 miles where as the Texas photo shows an area of about 300 feet. So the whole Texas photo would easily fit inside a single Martian polygon. On Mars we are looking at a very large fracture pattern, on Earth we are seeing a change in relief of the surface. The darker patterned areas are depressions. Each depression is surrounded narrow polygonal ridge area. These landforms on Earth only occur in expansive soil deposits, like the Yazoo clay. Note the linearity of the gilgai that is not seen in the Mars image. So clearly there are different processes at work. If you want to find this image on Google Earth the image date is 2/27/2001 and it's located at 32°26'34.64"N  96°45'14.11W. The primary clay mineral in Yazoo clay is Smectite and this mineral has also been discovered on Mars in areas like the Mawrth Vallis region using the CRISM instrument on the Mars Reconnaissance Orbiter. So maybe someday future residents of the Mawrth Vallis region will also need a good foundation engineer.

One last thought: From a strictly scientific point of view it would seem that polygonal landforms are more common in the universe than life. Maybe the Curiosity rover on Mars will change that in the near future. Good Luck Curiosity!

 REFERENCES

Maxwell, Britt.(1994) “Influence of Horizontal Stresses on Gilgai Landforms”  Journal of Geotechnical Engineering, Vol. 120, No. 8,  pp. 1437-1444. http://dx.doi.org/10.1061/ (ASCE)0733-9410(1994)120:8(1437).

Maxwell, Britt.(2011) “Field Measurements of Yazoo Clay Reveals Expansive Soil Design Issues” Journal of Performance of Constructed Facilities, Vol. 25, No. 1, pp.18-23. http://dx.doi.org/10.1061/(ASCE)CF.1943-5509.0000069

Taylor, Angela C.(2005) "Mineralogy and Engineering Properties of the Yazoo Clay Formation, Jackson Group, 
Central Mississippi." Mineralogy and Engineering Properties of the Yazoo Clay Formation, Jackson Group, Central Mississippi. Mississippi State: Mississippi State University.

Bishop, J. L., et al. (2012), Mineralogy and morphology of geologic units at Libya Montes, Mars: Ancient aqueously derived outcrops, mafic flows, fluvial features and impacts, J. Geophys. Res., doi:10.1029/2012JE004151.