Dawson Foundation Repair
Dawson Foundation Repair

214-234-8421 (Dallas)
713-668-2110 (Houston)

A Technical Look At Under-Slab Leaks

Under-Slab Plumbing Leaks and Damaging
Differential Foundation Movements

I offer this summary of the relevant technical issues as a focus for discussions with a homeowner’s insurance company and/or an engineer investigating a claim for foundation damage caused by under-slab plumbing leakage.

C. Randolph Riddell, P.E. (TBPE#39172STR) 22 April 2004

In those instances where the insurance company’s investigation of a claim for damages caused by under-slab leak(s) in the sanitary sewer system is not complete, two major deficiencies are usually apparent:

1. After clearly describing in a first engineering analysis that foundation movements are caused by water softening the supporting soil at the edge of the foundation, the engineer investigating the claim often excludes consideration of the effects of plumbing leakage under the slab on the damaging differential foundation movement from his published conclusions.

2. The investigating engineer’s speculative allegations about shrinkage of the supporting soil ordinarily do not establish the cause of the reported damage on a valid technical analysis. Indeed, the fact elements necessary to support such speculations typically either do not exist or have not occurred on the site. Moreover, the factual data the investigating engineer collected on the site and the additional data that can easily be contributed to the investigation refute such speculations. In lieu of facts, the investigating engineer often relies upon invalid pseudo-scientific analysis of the data and pure speculation. Therefore, the insurance company’s denial of the claim is unreasonable because it lacks a valid technical basis.

Accordingly, there are certain fundamental defects in an investigating engineer’s analysis of the data collected during an investigation.

Misplaced Focus

Often, the investigating engineer confines his attempts to correlate the site-specific data and known scientific and engineering principles to only the locations of the plumbing leaks. In so doing, he misleads the investigation away from the common sense understanding that water runs downhill. Because sewer leakage is mostly water, leakage runs downhill, too. Consequently, the damage done to the residence by the plumbing leaks must be expected to occur downhill from the plumbing leaks, not at the plumbing leaks, as the investigating engineer commonly, but mistakenly, persists. Obviously, determination of the downhill direction of the shallowest impermeable soil surface beneath the leak site is fundamental to a proper investigation.

Some investigating engineers repeatedly refer to the distinctly different strata of “(CL) low-plasticity ‘lean’ clay underlain with [sic] (CH) high-plasticity ‘fat’ clay” “that support the foundation” of the residence as if they were one homogeneous mass. In so doing, the investigating engineer misleads the investigation into overlooking the characteristically and significantly non-responsive shrink/swell behavior of lean clay to changes in moisture content. In cases where this non-responsive layer is actually supporting the foundation and is exposed to contact with the sewer leakage, its minimized potential dominates the behavior of the supported slab. However, softening by the addition of water universally weakens both lean clay and fat clay, proceeding more quickly and severely in the higher permeability of the low-plasticity lean clay. I find that this issue strongly implies the investigating engineer’s deliberate intent and explains an evident stubborn adherence to such an inappropriate mischaracterization.

Selective Emphasis

Shrinkage or swelling of expansive clay soil is a very specialized occurrence. On the other hand, softening of all kinds of soil by the addition of water is universally fundamental. By confining an investigative analysis to only shrinkage and swelling, the investigating engineer misleads the investigation away from the more universal and basic softening affect that the plumbing leakage has on the soil supporting the residence. Because of the low permeability of clay soil, water must stand on it for a long time before it soaks in far enough to soften a significant thickness of the upper layer. Consequently, downward movements of the foundation must be expected to occur more dramatically where the leakage stops running downhill, not at the leak sites where it begins its downhill run, nor along the way. Because the weight of the house persists on the upper layer of the supporting soil as it softens, those portions of the house at the locations of the softening process move down into the softened soil.

Junk Science

The arithmetic stunt of calculating an “average moisture content” for an entire 14 foot boring depth extending through at least three distinct and sharply contrasted soil strata is entirely without merit. An investigating engineer’s reliance on this bogus exercise as the basis for unwavering devotion to an invalid conclusion is not only inexcusable, respecting his engineering credentials, but a violation of the Texas Engineering Practice Act (Texas Civil Statutes, Article 3271a) (the Act). This issue is commonly pertinent anytime the investigating engineer’s career history has only recently involved foundation analysis and then dominantly in the production of similar beneficial claim investigation reports for clients.

The concept of “state of swell” for expansive clay soil is scientifically legitimate, but the simplistic correlation of only moisture content (or the derived liquidity index value) to state of swell completely ignores the commonly overpowering influence of overburden pressure to completely suppress the potential of a moistened sample of expansive clay to achieve any swell at all. As is detailed following, the weight of an ordinary house is often sufficient to completely suppress the feeble swell pressure exerted by a shallow stratum of lean clay.

The similar alternative distortion of describing a sample with lesser moisture content as “shrunk due to dryness” completely ignores the common case of different expansive clay samples of equal volume, but different combinations of moisture content and swell-suppressing overburden pressure. By insisting on this erroneous correlation and enlisting the complicity of the superficial credibility of a geotechnical firm, the investigating engineer commits multiple violations of the Act.

Following is a presentation of the technical concepts pertinent to a comprehensive analysis of the relationship between the under-slab plumbing leaks and the damaging differential foundation movements.


The release of sewage under a foundation slab from a leaking plumbing system causes foundation damage by the action of very basic scientific and engineering principles. Prior to the onset of the leak, the foundation is supported by the soil masses that actually contact the various elements of that foundation system. As the leakage flowing from the leak(s) reaches these areas of support, this support is reduced by three major affects:

  1. auto-compaction of loose fill under the slab by wetting with sewage;
  2. redistribution of the supporting soil under the slab by erosion;
  3. softening of the supporting soil by absorbing water from the sewage.Each of these processes is presented in published works by various respected authorities in the field of soil mechanics and foundation performance.

The subject phenomenon is sinking of a weight – the house — initially sitting on soil, into that same soil as it is softened by the absorption of water from the leakage escaping from holes in the under-slab sanitary sewer system. The universality of Prudent Men understanding this phenomenon is legion. This motley crew includes: the insurance company; the engineer investigating the claim and other structural engineers; geotechnical engineers; construction workers; operators of excavation equipment; pile driving crews; plumbers; manual laborers digging in the dirt; farmers; gardeners; firemen; fishermen; soldiers; pedestrians in communities with few sidewalks; children playing in sprinklers; bathers in above-ground swimming or wading pools; and myriad other Prudent Men in the world-wide human community possessed of common sense. It is not rocket science. It is the not-so-secret recipe for mud.



For economy of effort, a proper investigation into the cause(s) of the damaging foundation movements observed at a Property begins with generalities and proceeds to relevant details. Accordingly, for foundations on expansive clay in the presence of under-slab plumbing leaks, there are only six scenarios to consider:

1. upward movement from swelling of the supporting soil as the leakage is absorbed;
2. downward movement from shrinkage of the supporting soil as the immediate-post-construction moisture is removed;
3. downward movement from settling of the supporting soil as the immediate-post-construction support surface is lowered by compressing the supporting soil;
4. downward movement from settling of the supporting soil as the immediate-post-construction support is removed by auto-compaction caused by the flow of the plumbing leakage under the slab;
5. downward movement from redistribution of the supporting soil as the immediate-post-construction support is removed by erosion caused by the flow of the plumbing leakage under the slab;
6. downward movement from penetration of the foundation into the supporting soil as the immediate-post-construction support strength is reduced by absorption of the water from the plumbing leakage.

Economy and efficiency of the investigation effort dictate an initial common-sense examination of the characteristics of each scenario. At this phase of the investigation, inappropriate or irrelevant scenarios can be eliminated.

Upward Movement from Swelling

One important engineering skill is accurately estimating complex calculations. In this type of investigation, two estimates are pivotal in assessing the pertinence of swell behavior as a valid probable cause of the damaging differential foundation movements. A simple mental calculation is all that is necessary to make a meaningful preliminary assessment of the viability of the soil swell option. Using data from a case investigated by an insurance company in Sep2000, where expansive “fat” clay (CH) supported the foundation slab:

By the ASTM Swell Test on the two samples from under the slab, the geotechnical test results show only 4.4% and 10.9% Final Vertical Swell under 300 pounds per square foot (psf) confining pressure. Significantly, the dominant influence of permeability on the speed with which expansive clay can swell is demonstrated by the ASTM Swell Test procedures; the specially prepared sample must be completely immersed in water for 20,000 minutes
(= 2 weeks) to get a meaningful reading.

A formal calculation of the distribution of the total as-built weight of the entire house is not necessary for a meaningful preliminary assessment of the viability of swelling of the expansive clay soil. The weight of the 12″x15″ concrete grade beam the engineer investigating the claim measured alone imposes 182 psf on its supporting soil. The slab, walls, brick, ceiling, and roof add approximately 688 psf dead load to this real, as-built confining pressure. The total real confining pressure is approximately 870 psf. Accordingly, the swell test results for only 1/3 the confining pressure, at 300 psf, must be seen to be greater than any expectable actual swell performance under the almost three-times greater imposed real weight.

Nonetheless, generously rounding the maximum tested swell to 10%, the depth of necessary moisture change and “minimum wetting time” may be estimated using the maximum surveyed elevation differential of 4-3/8″ and the pertinent average measured permeability of 12″ in 40 years. Dividing 4.375 by 0.1 and then by 12 and multiplying the quotient by 40, the estimated time to achieve 4-3/8″ of swell by penetration of water from the soil surface is at least 145 years.

If, in a similar case, the upper stratum supporting the foundation is lean clay, the swell pressure would be less and the average permeability would be greater, but the result would likely be equivalent.

Clearly, swelling is not an appropriate concern. A more correct diagnosis would find that the residence is suffering from differential downward foundation movements.

Five Causes of Downward Movement

Because the pattern of differential elevations surveyed on the upper slab surface and the height of voids under the bottom of the slab varies greatly, credible causes of downward differential movements must be “localizable” to relatively small areas of the overall slab. Accordingly:

1. Localized shrinkage of supporting expansive clay soil by moisture loss, when it occurs, is only localized to persistent sources of unusual heat, otherwise it extends over very large areas appropriate to natural drying influences.
2. Consolidation of local “pockets” of soft supporting soil would produce local differential movements.
3. Auto-compaction is a process that causes physical removal of the supporting contact between the soil surface that served as the form when the wet slab and grade beam concrete was originally placed.
4. Erosion is another common process that physically removes supporting soil.
5. Similarly, a localized loss of enough strength in the supporting soil by absorption of water while maintaining contact produces the same downward foundation movements as physical removal of the soil.

Shrinkage Due to Moisture Removal

Localized shrinkage of supporting expansive clay soil by moisture loss, when it occurs, is characteristically localized to persistent sources of unusual heat, such as furnace rooms and trenches carrying steam lines. The absence of these peculiar heat sources under a residence also eliminates soil shrinkage as a cause for localized damaging downward differential slab movements.

Because all three elements of a foundation system – if piers, grade beams, and slab — must move together to achieve an observed drop in elevation of the slab surface, the supporting soil beneath the bottoms of the piers must lose moisture, if shrinkage is the cause. Because of the long time – 100+ years — required to remove moisture from the low permeability fat clay soil mass supporting the bottoms of the piers, shrinkage is quickly eliminated as a reasonable cause.

Arborists agree that the feeder root portion of the root system that removes moisture from the soil is found in the uppermost 2′ of soil. Because the bottoms of piers are likely at least 6′ below the ground surface, any loss in moisture attributable to tree roots occurring below the bottoms of the piers requires that moisture move up through the low permeability clay soil mass at least 4′. At the permeability rate discussed earlier, such movement would take at least 160 years. This eliminates soil shrinkage as a cause for recent damaging slab movements at the residence.

Consolidation of Supporting Soil

Inelastic compression of supporting soil under load is termed “consolidation”. Post-construction consolidation is recognized as occurring in two stages. The “primary” consolidation occurs within a few hours or days of the imposition of the load, depending upon the specific soil properties and conditions, typically attaining a fraction of an inch maximum value. The immediately ensuing “secondary” consolidation proceeds very slowly in small fractions of an inch over geologic time. Consequently, new major differential movements occurring after several decades of stable foundation performance are not attributable to this process.

Auto-compaction of Wetted Soil

Auto-compaction of loose under-slab fill upon wetting with plumbing leakage removes the supporting contact at any area where sufficient wetting occurs. In the ordinary sequence of residential construction, the wet slab concrete is laid on top of a layer of newly deposited and smoothed fill soil that is usually not well compacted. Before and as the concrete hardens, the weight of the slab continues to press down on the top of this fill soil, establishing and maintaining contact between the concrete and the soil. After the concrete hardens, it is able to bridge across any relatively small gaps that might develop in the future in this originally uniform supporting contact.

The original contact between the supporting soil and the foundation slab is broken by auto-compaction of the loose fill soil that is commonly placed to form the bottom of the slab. This basic scientific process is seen where loose dry material, such as sandy dirt, is wetted and “packs down” under its own weight wherever it gets wet. When this happens under the hardened slab, an empty space is formed between the bottom of the slab and the top of the wetted fill soil.

Erosion of Wetted Soil

Removal of soil by erosive transport in the flow of under-slab plumbing leakage similarly removes supporting contact in highly localized areas. Erosion is the process of carrying soil particles in moving liquid. The sub-microscopic clay flakes are suspended in water by the scientific principles of chemistry, electricity and molecular physics. When the water runs downhill, the suspended clay flakes go downhill with it. Any tiny spaces between large soil particles that this water moves through that are also bigger than the suspended sub-microscopic clay flakes will also allow the clay particles to pass through. Because the clay flakes are suspended in the water, this erosive transport proceeds at even very slow water flow rates.

Wherever empty spaces are created by the auto-compaction process, larger scale erosion can also carry soil particles into these spaces. The eroded soil that fills such empty spaces is carried downhill by the flowing water from locations uphill along the flowpath. The resulting change in the pattern of empty spaces under the slab appears as though the original downhill space moves uphill as the eroded soil fills the space, similar to the movement of a bubble in liquid under glass.

Weakening of Supporting Soil

As the supporting soil absorbs water from the leakage, the soil softens, becoming weaker. When the soil supporting any of the foundation elements absorbs enough water from the plumbing leakage and becomes too weak to continue holding that element in its original immediate post-construction position, the weight carried at that weakened site will press the foundation element down into the wet and weakened soil, squeezing it out from under the foundation element.

This weakening process begins and progresses most severely at the leakage accumulation site, as it does at the bottom of a puddle. More commonly, this process is the not-so-secret recipe for mud.

As puddles are commonly seen to form in the low spots on the ground, so the leakage accumulates in low spots under the slab. The lowest spots are at the bottom of the holes that were drilled to form the piers. Directly above and connected to these pier holes are the trenches that were dug in the soil to form the bottom of the grade beams. Above and directly adjacent to these trenches are “dimples” in the uneven sloping surface of the native clay soil under the slab. Any path leading to these low spots that the leakage may travel along allows the leakage to flow to the soil supporting these elements of the foundation system: piers; grade beams; slab.

Because the supported weight of the house does not relent, this weight pushes the foundation elements down into the softened supporting soil as it softens. If the softening effect is not uniformly distributed across the entire foundation system, this downward movement will not be uniformly distributed. The foundation will tilt and twist as a result. Brittle construction finishes, such as interior sheetrock and exterior brick veneer, will crack as their supports become distorted by these differential foundation movements.

The Persistence of the Combined Effect

By acting together, the under-slab voids create an “allowance” for downward movements and transfer the supported weight to adjacent points of support where the weakening is progressing. Ultimately, the foundation drops at these localized areas. As long as the plumbing continues to leak, the cause persists and the downward movements persist in either severity or extent or both.

Thus, by the process of elimination, the destabilizing effects of plumbing leakage on the foundation are established as the probable cause of the observed damaging downward differential movements.


Locations of Foundation Failure and Soil Sampling

The location where the soil strength must be measured by a soil study is at the thin layer actually contacting the foundation elements, as I have described. The insurance company and their investigating engineers and geotechnical firms usually take no samples from such locations.

Locations of Plumbing Leaks vs Leakage Accumulation

The physical locations of the sites of accumulation of the plumbing leakage are not revealed by tests of the under-slab sanitary sewer plumbing system. Only the physical locations of the leaks in the under-slab plumbing pipes are determined by such plumbing tests. However, the range of locations for the accumulation sites is established by scientific and engineering principles so basic they are usually referred to as “common sense”:

Ø Plumbing leakage is mostly water;
Ø Water runs downhill;
Ø Water collects in low spots.

Because water runs downhill, the direction of slope of the downhill-sloping soil surface away from each leak location must be determined. In addition, this sloping soil surface must be relatively water-resistant; that is, a low permeability soil, such as clay. The loose fill soil, previously discussed, is deliberately chosen for its high permeability, not low. Consequently, water soaks down through the fill much more easily than it can penetrate the underlying clay. For these very basic scientific and engineering reasons, leakage flows through the loose fill soil, down the slope of the clay soil surface and away from the leak locations. Water can only move through the mass of the clay soil at the extremely slow speed described by its permeability.

Locations of Soil Borings

Soil samples taken from boring locations away from supported foundation elements only provide useful descriptions of soil properties, but not of soil conditions relevant to the weakening of soil support at the foundation elements. This is one of the critical flaws in an insurance company’s typical investigation of a claim.

In the analysis of a damaged building with a distorted foundation, it is relevant and necessary to examine the condition of the supporting soil masses that actually contact the various foundation elements. Although secondarily relevant for soil property data, it is not sufficient to examine only non-supporting soil masses remote from the locations of supporting soil contact. For this reason, the physical location or position of the examined sites is critical to a proper, complete, and meaningful investigation of a damaged building.

Significantly, the physical locations the insurance company, the investigating engineer, and the geotechnical firm typically select for the soil borings and sampling are remote from supported foundation elements. Consequently, necessary data describing soil conditions that are necessary to determine the affects of the plumbing leakage on the soil supporting the foundation elements are not provided in the resulting geotechnical report.

The data describing soil properties that are necessary to make these determinations can credibly be provided by tests of soil samples taken from borings made through the slab. In the typical claim investigation, these relevant data are provided in the geotechnical reports from the boring samples. Even though those reports only present data derived from samples taken from non-supporting soil masses in remote locations, these remote data can provide valid insights into the affects of the plumbing leakage on the supporting soil, but only for circumstances tightly correlated to the sampled stratum properties and conditions.

Distance vs Time and Permeability

The speed with which water moves through soil is permeability. Once the permeability is known for a particular soil stratum, the amount of time required for water to move a distance of one foot through that soil may be easily calculated. In one claim investigation, for instance, the boring samples taken from under the residence slab showed a range of permeability from “fastest” — at about 0.080″ to 0.050″ per month to — “slowest” — at about 0.005″ to 0.007″ per month. A “medium” speed (permeability) of 0.025″ per month requires 40 years for water to move one foot through such soil. Consequently, the distance separating sampled non-supporting locations from un-sampled supporting locations is a primary factor in determining the relevance of the sampled non-supporting moisture content and affected strength to the moisture content and strength of the un-sampled supporting soil.

Specific to this typical claim investigation, soil samples taken only one foot away from an accumulation site of under-slab plumbing leakage would not show increased moisture content and weakening affects until 40 years had passed after the leakage had weakened the supporting soil at the un-sampled accumulation site. As discussed, such accumulation sites are the bottoms of the piers, the grade beam trenches, and low spots under the slab.

Significant to insurance companies and one of the causation theories common to their investigating engineers, this same period of 40 years is required for a plant root to cause shrinkage by drawing water from soil only one foot away. For this reason, the relevant technical literature routinely, repeatedly, and reliably restricts any such soil shrinkage attributable to the action of plant roots to prolonged dry periods and droughts. Clearly, rain provides greater quantities of more readily available moisture to the shallow depth plant roots than does the geologically trapped moisture in the soil mass to any deeper roots.

C. Randolph Riddell, P.E. (TBPE#39172STR) 22 April 2004

Return to Engineering Information Page

Read More about Plumbing

See Testimonial Letter #254 – an under slab plumbing leak and foundation repair