Maryland Department of Natural Resources

Foundation Engineering Problems and Hazards in Karst Terranes

Please note: While the Maryland Geological Survey can help identify whether your home is located on rocks that are likely to develop sinkholes, we cannot offer site-specific advice. Local resources for problems related to sinkholes may be found on the Sinkhole Resources page.


Just about any place where the land is underlain by relatively soluble bedrock, natural waters on and below the land surface slowly dissolve that bedrock. Dissolving is enhanced by these waters' tendency to be acidic. For example, rain is usually acidic because it contains dissolved carbon dioxide (CO2) from the atmosphere. It often becomes more acidic as it soaks into the ground and picks up more CO2 from the soil. Such a landscape in which the bedrock is shaped, or sculpted, by dissolution is referred to as karst. The most common type of bedrock in karst terranes is carbonate rock, which includes limestone (calcium carbonate), dolostone (calcium-magnesium carbonate), and marble (calcium carbonate). There are a few other kinds of rock (e.g., gypsum, which is composed of calcium sulfate) that can be involved in karst, but in Maryland karst terranes are limited to areas underlain by carbonate rocks.

Index Map of Earthquakes in Maryland
Figure 1. Map showing the distribution of carbonate rocks in Maryland. Those most associated with collapse sinkholes are the Hagerstown Valley (HV), the Frederick Valley (FV), and the Wakefield Valley (WV). To a lesser degree, collapse sinkholes are found in Green Spring Valley (gs), Worthington Valley (wo), and Long Green Valley (lg).

No single landform is common to all karst areas, but one characteristic of all karst landscapes is disrupted surface drainage due to loss of surface water to the subsurface. Two examples of karst landforms are caves and closed depressions, known as sinkholes. Maryland has approximately 100 known caves, and many more sinkholes. The carbonate rock areas of Maryland that exhibit some degree of karst development are shown in Figure 1.

Potential environmental problems in karst terranes fall into two broad categories: (1) groundwater pollution and (2) foundation engineering problems. This fact sheet discusses three main categories of interrelated foundation engineering problems: (a) differential compaction and settling due to the irregular surface between soil and bedrock; (b) soil piping, which is a type of subsurface erosion; and (c) collapse of the land surface into an underground cavity--that is, collapse sinkholes.

Differential Compaction and Settling

Building the foundation for a house or other structure involves laying a footer, usually pouring concrete into a level trench that outlines the dimensions of the house, then erecting walls on that footer. Where there is a basement, a concrete floor is generally placed on the bottom of the excavation within the confines of the footer and basement walls. As the house is built, all of the weight, or load, is carried by the outside walls and footer. In some karst terranes, that can lead to problems.

In a type of karst known as cutter-and-pinnacle karst, the contact between bedrock and soil overburden is very irregular (see Fig. 2 and 3 for example). Water preferentially dissolves bedrock along some planar feature, such as bedding, joints, or fractures, whichever is the easier path. Roughly vertical, solutionally widened joints are called cutters, or grikes. Cutters are generally filled with soil. The bedrock that remains between cutters may be reduced to relatively narrow "ridges" of rock, called pinnacles, particularly where cutters are closely spaced. Cutter-and-pinnacle karst (or simply "pinnacle karst" for short) is common in many of the carbonate valleys in Maryland (Fig. 1).

The problem develops when a building foundation lies on cutters and pinnacles. The weight of the building will compact the soil to some extent, and the building will settle. That is normal, and does not pose a problem as long as the building settles uniformly. However, in pinnacle karst, part of the foundation may be supported by a bedrock pinnacle and part may be supported by a cutter (soil-filled). The result can be differential settling of the building, which may produce cracks in the walls, foundation, and floor (Fig. 2). This may compromise the structural soundness of the bearing walls and, therefore, place the safety of the whole structure in doubt.

Large buildings (schools, shopping centers, office buildings, etc.) commonly have a detailed engineering design to avoid such potential problems, but private homes are often built with little regard to such problems. Adequate site evaluation prior to building is important; done properly, it can do much to prevent damage to a home from differential settling.

Subsurface Erosion (Piping)

Piping is subsurface erosion of soil by percolating waters to produce pipe-like conduits underground. Piping can affect materials ranging from clay-size particles (less than 0.002 mm) to gravels (several centimeters), but is most common in fine-grained soils such as fine sand, silt, and coarse clay. The resulting "pipes" are commonly a few millimeters to a few centimeters in size, but can grow to a meter or more in diameter. They may lie very close to the ground surface or extend several meters below ground.

Cross-section of structural damage
Figure 2.-- Cross-section sketch illustrating structural damage

Piping can become a problem in areas of cutter-and-pinnacle karst, as well as in some non-karst areas. As shown in Figure 3, what begins as piping can develop into cavities in the soil overburden. Piping tends to become accelerated when the water table is lowered by over-pumping ground water, when the amount of infiltrating water increases, or both. (The "water table" marks the top of the zone of saturation, in which all pores and voids in bedrock and soil are filled with water.)

What can cause increased volume of water that infiltrates the soil overburden? Long periods of rainfall can be a factor, but man's activities also are significant. Buildings with large roof areas, parking lots, streets and highways change the runoff and infiltration characteristics of soil by decreasing widespread, diffuse infiltration and channeling surface runoff to areas where more concentrated infiltration can occur. Figure 3 shows how runoff can be concentrated in the subsurface to create subsurface cavities. This is especially common in soil-filled cutters.

One consequence of modification in runoff and infiltration can be foundation problems similar to those mentioned in the previous section. Figure 3 shows how a leaking storm drain or water main may lead to the pipe breaking and how runoff from streets and houses may create subsurface cavities near building foundations. This, it should be added, can occur in non-karst areas too. Because the pavements and buildings act as supporting structures, the loss of soil may not be apparent until a sizable cavity has developed. At some point, structural support is lost. The result may be a relatively slow subsidence of the street or the building, during which cracks will develop in basement walls and floors, or the result may be a sudden collapse of the building or pavement. It should be noted, also, that rerouting of storm runoff from rooftops, parking lots, and streets can cause soil piping under adjacent properties.

Collapse Sinkholes

As used here, the term sinkhole refers exclusively to one type of closed depressions in karst landscapes. One type of sinkhole is the collapse sinkhole, so named because it forms suddenly when the land surface collapses into underground voids, or cavities. Collapse sinkholes are often fairly circular with steeply sloping sides. They can be so small as to be barely noticeable to 50 meters or more in width and depth. Once formed, they can also grow larger.

Cross-section of structural damage
Figure 3.-- Cross-section sketch illustrating soil piping and possible structural damage initiated by modification of natural runoff and infiltration of water (after White, 1988).

In some karst terranes, collapse sinkholes form when the roof of a cave or cavern collapses. Such is the case in some collapses in Florida (Sinclair, 1982). However, most collapse sinkholes seem associated with cavities in the soil overlying the carbonate rock. Some prefer the term cover collapse sinkhole to denote that collapse occurs in cavities in the soil overburden, or cover, rather than in the carbonate bedrock below. This is the general case for collapse sinkholes in areas of pinnacle karst in Maryland.

Other types of sinkholes form slowly by the dissolving of carbonate rock at or very near the surface. They tend to have gently sloping sides, and they seldom pose a hazard by collapsing. Like collapse sinkholes, however, they can pose environmental problems related to pollution, because they provide a point where polluted surface runoff can directly flow into the ground water.

In the United States, according to one study, the states most impacted by collapse sinkholes are Alabama, Florida, Georgia, Missouri, Pennsylvania, and Tennessee (Newton, 1987).

In Maryland, collapse sinkholes occur mainly in four areas: the limestones of the Hagerstown Valley in Washington County and the Frederick Valley in Frederick County, marble in the Wakefield Valley in Carroll County and, to a lesser degree, in marble valleys of Baltimore County (Fig. 1). Collapse sinkholes seem to be most prevalent in the Frederick Valley and the Wakefield Valley.

Cavities of various sizes tend to develop in the soil overburden where infiltrating surface waters erode the soil by piping and transport it downward through bedrock cracks, or joints, that are themselves widened and enlarged by the dissolving of the rock by the infiltrating water. This creates something like a plumbing system through which the eroded soil overburden is carried.

Infiltrating CO2-charged water dissolves more and more carbonate rock over long periods, sometimes enlarging the cracks to a meter or more in width. These solutionally enlarged joints are most effective in transporting soil when they are above the water table. A lowering of the water table, therefore, tends to increase that effectiveness. In time, this process can produce a large cavity in the soil overburden (Fig. 4).

Cross-section of structural damage
Figure 4. Cross-section sketch showing the progressive development of a cavity in the soil overburden and eventual creation of a collapse sinkhole. Once the soil "bridge" over the cavity cannot support itself, collapse occurs (after Newton, 1987).

Moisture conditions in the soil overburden are important in that moisture affects soil strength. Many factors can affect soil moisture--climate, soil texture, soil mineralogy, evaporation, transpiration (the uptake of moisture by vegetation), and sometimes depth to water table, to name a few. The interplay among the factors is complex, but the net result is that "some" moisture enhances soil strength; too much or too little diminish it.

For example, imagine a soil cavity exists a few meters below the surface. Following a period of much rain, soil moisture content may become high enough that it effectively reduces the strength of the soil, allowing collapse to occur. In a few cases, the weight of the extra water from infiltrated rain may even be enough to trigger a collapse. On the other hand, in a period of drought, drying can reduce soil strength. Drying tends to cause shrinking, which causes cracking, which in turn can lead to spalling of soil from the roof of the cavity, eventually resulting in collapse. Thus, some soil moisture is good because it increases the strength and reduces erodibility of the soil, but too much or too little moisture generally reduces the strength of the soil and makes collapse more likely.

The water table and soil moisture fluctuate naturally during the year in response to patterns of precipitation, evaporation, and transpiration. Ground water is not static, or stationary; it moves through pores and cracks in the bedrock along a gradient, or "slope," from higher elevations to lower, eventually discharging into streams. However, pumping a well or group of wells (residential, commercial, municipal, or industrial wells) forms a cone-shaped depression in the water table around the well(s). If a cone of depression becomes large enough due to prolonged pumping, possible consequences include occurrence of a collapse sinkhole, an increase in soil piping, loss of water in nearby wells, dried up springs or streams, ground subsidence, or even reduction of support beneath foundations of buildings. In cases such as these, one often hears the terms, dewatering and overpumping.

Dewatering due to overpumping is not the only manmade cause of collapse sinkholes. The same kinds of alterations to drainage discussed previously are often involved too. A study of collapse sinkholes occurring over a nearly fifty-year period in Missouri (Williams and Vineyard, 1976) showed that about half of collapse sinkholes were natural and about half were man-induced. Altering drainage conditions was found to be the chief manmade "cause" (Table 1). These observations apply to Missouri only, but they do show that there are several possible triggering mechanisms.

A natural collapse is the product of a process that can span many thousands of years. Human activities (such as those listed in Table 1) or unusually wet or dry weather can trigger collapse where the natural process has set the stage for transport of soil from a growing void. However, correctly determining the immediate cause, or trigger, of a collapse sinkhole is often very difficult.

Table 1.-- Frequency of collapse sinkholes in Missouri on the basis
of "immediate" cause, as derived from records kept from 1930 through
the mid-1970s (after Williams and Vineyard, 1976).
Collapse Sinkholes in Missouri
Number by natural causes 51
Number man-induced 46
Altered drainage 24
Water impoundments 10
Dewatering 7
Highway construction 3
Blasting 2

Accurately predicting collapse sinkholes is also very difficult. Conditions must be such that collapse is imminent. In other words, any of a number of contributing processes must have operated for some time. Often, there is no warning that a cavity in the soil has developed until a collapse occurs.

In areas of known collapse sinkholes, frequent observations have sometimes proven useful. Sinclair (1982) lists the following possible precursors of sinkhole collapse:

It is important to realize that each of these things can occur for other reasons--and in non-karst areas. However, if any of these observations are made in an area of known sinkholes and if roads or buildings are potentially at risk, the property owner should consider acquiring the services of an engineering geologist or a foundations engineer. If you find a newly formed or forming collapse sinkhole, please report it at once. • If on State highway right-of-way, call the State Highway Administration at 410-321-3107 (or call the district office listed in the blue pages of your telephone book).

A final word of caution: Never climb into a collapse sinkhole, and never go into any visible opening at the bottom of a sinkhole--especially if you are alone. Always exercise caution when walking around sinkholes.


Newton, J. G., 1987, Development of sinkholes resulting from man's activities in the eastern United States: US Geological Survey Circular 968, 54 p.

Sinclair, W. C., 1982, Sinkhole development resulting from ground-water withdrawal in the Tampa Area, Florida: U.S. Geological Survey Water-Resources Investigations 81-50, 19 p.

White, W. B., 1988, Geomorphology and Hydrology of Karst Terrains: Oxford University Press, New York, 464 p.

Williams, J. H. and Vineyard, J. D., 1976, Geologic indicators of subsidence and collapse in karst terrain in Missouri: Presentation at the 55th Annual Meeting, Transportation Research Board, Washington, D.C.

Prepared by James P. Reger
Permission is granted to reproduce this Fact Sheet so long as proper credit is given to Maryland Geological Survey. Compiled by the Maryland Geological Survey, 2300 St. Paul Street, Baltimore, MD 21218 This electronic version of "Fact Sheet No.11 " was prepared by R.D. Conkwright, Division of Coastal and Estuarine Geology, Maryland Geological Survey.