Before we build again · The science

Failing to plan is planning to fail

When the ground you build on is karst, skipping the test isn’t a saving. It’s a bill you pay later — with interest, and sometimes with no way to undo it.

← First: how karst works

Every large building rests on an assumption about the ground beneath it. A pre-construction geologic and hydrogeologic investigation turns that assumption into knowledge — for a fraction of the project’s cost. Skip it and the ground still gets “tested,” just the hard way: during or after construction, when the fix costs far more and, in karst, may not be possible at all.

1 · The asymmetry

The cheapest part of the project is the one most often cut

A site investigation is typically a fraction of one percent of a construction budget — yet skipping it is among the largest cost risks a project carries. UK industry guidance, where the Institution of Civil Engineers and CIRIA have studied this for decades, finds about half of projects hit delays from unforeseen ground conditions, with inadequate investigation a leading, avoidable driver of overruns.

Cost asymmetry: testing first versus paying later A short green bar for the cost of a site investigation next to a long red bar for the cost of overruns, repairs, and irreversible damage when testing is skipped. Illustrative order-of-magnitude comparison. Pay a little now, or a lot later Test the ground first a fraction of 1% of project cost — known, upfront Skip it overruns, redesign, repairs to homes & wells, lost water — and damage that can’t be bought back Illustrative — order-of-magnitude comparison, not to scale or a site-specific estimate.
Figure 1. The economics that make skipping investigation tempting are exactly backwards: the small, certain, upfront cost is the one that gets cut, and the large, uncertain, downstream cost is the one that gets accepted by default.
One peer-reviewed study of construction projects found that apparent savings from inadequate site investigation were associated with cost overruns averaging on the order of 60% of project cost — the classic false economy. (Effect of inadequate site investigation on cost and time; see sources.)

2 · Why karst makes it worse

Some ground gives you a second chance. Karst often doesn’t.

On ordinary ground, a surprise means a delay and a bigger bill. In karst, the failures are different in kind: a drained conduit, a collapsed void, or a spring that runs cloudy can be permanent. The U.S. Geological Survey calls karst aquifers “highly productive but extremely vulnerable” — the same open conduits that move water for miles move sediment, contamination, and pressure changes just as fast. Once a void collapses or an aquifer depressurizes, there’s frequently no way to put it back.

Decision tree: test first versus skip testing A decision node splits into two paths. Testing first leads to voids found early, a known cost, and protected water. Skipping leads to voids found during or after construction, high cost, and outcomes that can be irreversible. Build on karst Test the ground first Voids & water paths found early → redesign, relocate, or mitigate Known cost · water protected Skip testing Voids found the hard way — during or after drilling High cost · some outcomes irreversible Both paths encounter the same ground. Only one chooses when to find out what’s in it. Illustrative — general published karst science, not a site-specific survey.
Figure 2. Testing doesn’t change what’s underground — it changes whether you learn about it before or after you’ve committed the building, the budget, and the neighbors’ water to it.

3 · Foreseeable, not freak

This has happened before — and it was documented

These outcomes aren’t hypothetical or rare — they’re in the peer-reviewed and agency record, which is exactly why they’re foreseeable and can be planned for.

Barton Springs, Austin, Texas (2018). Ten closed-loop geothermal wells were drilled into the karstic Edwards Aquifer in a residential neighborhood. A void was hit at about 240 feet. Days later, a plume of sediment surfaced at Main Barton Spring — the city’s landmark swimming hole — roughly 4,000 feet away, traced back through a karst conduit to the drilling site. The clearest published parallel to drilling closed-loop geothermal into karst.
Timeline of the 2018 Barton Springs geothermal-drilling turbidity event A left-to-right timeline: geothermal drilling into the karst Edwards Aquifer, a void encountered at about 240 feet, then a sediment plume appearing at Barton Spring about 4,000 feet away within days. Drilling begins 10 geothermal wells, karst Edwards Aquifer Void at ~240 ft cavernous porosity, just above water level Plume at the spring turbidity ~4,000 ft away, Dec 18–20, 2018 Published account; see sources. General science, not a site-specific claim about any other project.
Figure 3. In karst, a problem at the borehole does not stay at the borehole — it travels the conduit network, often surfacing somewhere else entirely.
Staufen im Breisgau, Germany (2007). Geothermal boreholes let water reach a reactive, soluble rock layer; the swelling lifted the ground up to about 30 centimeters and cracked more than 250 historic buildings. Different rock chemistry than Tennessee limestone, same lesson: drilling into soluble ground without first characterizing it can trigger large, slow, effectively irreversible damage. (Goldscheider & Bechtel; TerraSAR-X monitoring study — see sources.)

4 · The standard already exists

Other places require exactly this — for schools especially

Testing the ground before building isn’t a novel demand — it’s codified practice elsewhere. California’s Division of the State Architect requires a geohazard report for all construction on a new school site, reviewed by the state Geological Survey before the project advances. Europe’s Eurocode 7 (EN 1997-2) makes ground investigation a standard design step. The profession already treats “know the ground first” as the baseline; the only question is whether a project meets it.

The point. When a safeguard is standard practice and written into other jurisdictions’ codes, choosing not to apply it is a decision — not an oversight the science failed to warn about.

5 · What testing would catch

The tools exist, and each one retires a specific risk

A pre-construction karst investigation isn’t one test but a layered set, each matched to a hazard it finds before a drill rig arrives.

MethodWhat it findsRisk it retires
Surface geophysicsVoids, low-density zones, and the top of rock, without diggingBuilding over an unseen cavity
Borings & probingDepth to rock, clay thickness, void encountersMisjudging the load-bearing “clay bridge”
Dye tracingWhere groundwater actually flows, and how fastNot knowing whose wells and springs are downstream
InSAR / satellite monitoringMillimeter-scale ground movement over timeMissing slow subsidence until it cracks a building

These are standard hydrogeologic and geotechnical methods; a qualified professional selects and scopes them for the specific site.

The fix is simple, and it comes first

Test the ground before the borehole — not after.

Ask that a full geological and hydrogeological survey be required for any school site planned for geothermal HVAC on karst — before it’s approved. It’s the cheapest insurance a project can buy, and the only kind that works before the damage is done.

See the next public meetings & how to weigh in →

Sources