Sinkholes in the USA: Ground Risk, Karst Geology and What to Do Before You Build
Approximately 40% of the contiguous United States sits on carbonate rock (limestone, dolomite and gypsum) that dissolves slowly in groundwater. The result is karst: a landscape of caves, sinkholes, springs and unpredictable ground that looks solid until it isn't. For developers, infrastructure operators and investors, this is not a geological curiosity. It is a material risk that can invalidate a foundation design, trigger an insurance dispute or collapse a road overnight.
Why the USA has a particular problem
Florida is the most widely cited example - hundreds of new sinkholes are reported annually, and insured losses regularly run into hundreds of millions of dollars - but karst geology extends far beyond the Southeast. The limestone plateaus of Texas, the Ozarks of Missouri and Arkansas, the Valley and Ridge province running from Alabama through Virginia and Pennsylvania, and the dolomite belts of the upper Midwest all present dissolution risk. Add to this the legacy of underground coal and mineral mining across Appalachia, the Midwest and the Mountain West, and the picture becomes clear: a very large share of US development land carries some form of ground-void risk, whether natural or man-made.
The challenge is that this risk is invisible at the surface. A site can look entirely competent - flat, dry, with no surface expression of any problem - while hosting a network of dissolution pipes or partially collapsed cavities a few meters down. Traditional investigation methods, drilling grids and cone penetration tests, sample discrete points and routinely miss the features that sit between boreholes. The cavity that causes the damage is almost always the one the borehole grid didn't hit.
The two categories of ground-void risk
Natural karst forms where acidic groundwater dissolves carbonate rock over geological time. The process is ongoing: a cavity that is stable today can become unstable if groundwater levels change, if surface loading increases (a new building, a heavy vehicle), or if a drought-then-flood cycle alters pore pressures. Florida's sinkhole season is not a myth - it correlates with dry periods that lower the water table and remove the buoyant support that was keeping a roof arch in place.
Mine workings present a different but equally serious problem. Underground coal mines were operated across Appalachia, the Illinois Basin and parts of the Mountain West for over a century, often without the detailed records that modern engineering requires. When a mine roof collapses - decades after closure - the surface above can subside suddenly or progressively. The affected area is not always directly above the workings: pillar failure can propagate laterally, and subsidence angles mean the surface expression is wider than the underground void.
What this means for development and investment
For a developer, undetected ground voids discovered after foundation design is complete mean redesign costs, schedule delays and, in the worst case, a site that cannot be built on as planned. For an infrastructure investor acquiring an existing asset - a data center, a logistics facility, a utility plant - ground risk is a CapEx exposure that should appear as a quantified sensitivity in the financial model, not as a post-closing surprise. For a lender, it is a security question: is the collateral sitting on competent ground?
The practical question is not whether the risk exists - on karst or mine-affected land it almost certainly does to some degree - but where it is concentrated, how severe it is, and whether it is stable or progressing. That is an engineering question, and it has an engineering answer.
Area imaging versus point sampling
The fundamental limitation of drilling-based investigation is that it samples points. A borehole tells you what is directly below it, and nothing about what is a meter to the side. On karst ground, where dissolution features are irregular and unpredictable in plan, a drilling grid that would be considered thorough for a uniform geology can still miss the critical cavity. The cost of a grid dense enough to give genuine confidence is often prohibitive at the pre-design stage.
Non-invasive subsurface surveys take a different approach: they image the ground across an area, producing a spatial picture of where anomalies - voids, dissolution features, soft zones, old workings - are concentrated and where the ground is competent. Anomalies between boreholes are seen, not missed. Results are available in real time in the field, so critical zones can be examined more closely in the same campaign. And because there is no drilling, there is no site disturbance and no need for investigation permits in most jurisdictions.
From survey to financial model
A ground risk survey is only as useful as the decisions it informs. For a developer, the output should feed directly into foundation design - identifying where standard foundations are adequate and where ground improvement or deeper piling is needed. For an investor or lender, the output should be structured as a quantified risk: what is the probability of encountering a void that requires remediation, and what does remediation cost? That number enters the financial model as a CapEx sensitivity, so it influences the purchase price rather than the post-closing budget.
This is the connection that most ground investigation firms cannot make: they deliver a technical report, but they do not translate it into the language of a financial model. Bridging that gap - from geology to engineering to investment risk - is where the real value lies.