Freeze–thaw and soil risk in residential concrete

What freeze–thaw cycling does to saturated concrete

Concrete is porous. Water moves into it, and when that water freezes it expands. The expansion happens inside the material, against the surrounding paste, and it happens again on the next cycle. The damage is cumulative rather than dramatic — which is why it usually shows up as a surface gradually coming apart rather than as a sudden failure.

The critical variable is not how cold it gets. It is how often the concrete crosses the freezing point while it is wet. A climate that freezes hard in December and stays frozen delivers fewer of these events than one that swings back and forth around freezing through the whole season. Concrete that is dry when it freezes is largely unbothered.

Two things make the same climate much harder on one slab than another. De-icing chemicals lower the freezing point of water on the surface, which increases the number of freeze and thaw events the surface goes through, and they draw more moisture into the concrete. And how the concrete was finished and cured determines how dense and resistant its top layer was to begin with — a surface worked while bleed water was still present, or one that dried too fast because it was never properly cured, is weaker before its first winter.

This is the mechanism behind scaling and much surface spalling. It is a durability problem in the wearing surface, and on its own it says nothing about whether the slab has moved.

Why drainage and water exposure matter more than anything else

Water does three separate jobs on a concrete problem, and it is worth separating them.

It saturates the concrete, which is what makes freeze–thaw damage possible. It transports soil, carrying fine material out from under a slab edge and leaving a space behind. And it changes soil volume, swelling some soils as they wet and letting them shrink as they dry.

All three are driven by where water goes on a property, which is usually decided by a handful of unglamorous things: where downspouts discharge, whether gutters are clear, which way the ground falls away from the house, and whether any part of the surface holds water instead of shedding it.

This is why a contractor who looks at downspouts and grading before quoting is usually the one taking the job seriously, and why concrete work done without addressing water has a countdown attached. It is also why the cheapest genuine fix on many properties is not concrete work at all — extending a discharge line so it releases well clear of a slab costs very little and addresses a real cause.

• Watch what happens during heavy rain, not after it. Ten minutes outside in bad weather tells you more than any dry-day inspection.
• Note where roof water lands and where it goes from there.
• Note anywhere water still stands a day later.
• Note anywhere water disappears into a joint or crack rather than running across the surface — that is water going somewhere you cannot see.

Expansive soil versus poorly compacted fill

These are two different problems that both produce movement, and they behave differently over time.

Some soils change volume with moisture content — swelling when wet, shrinking when dry. Where that happens, the movement is cyclical rather than progressive. It tends to track the seasons, it can partly reverse, and a slab over it may rise and fall rather than simply descending. Vegetation matters here too, because a large tree drawing moisture out of the ground changes the volume of the soil around it.

Poorly compacted fill is a different story. Material placed during construction and not fully compacted continues to consolidate under its own weight and whatever is on top of it. That process is progressive and largely one-directional: the ground goes down and stays down. It also tends to be self-limiting — it finishes eventually.

The practical distinction for a homeowner is whether the movement looks like it is finishing or continuing, and whether it seems to come and go. Progressive settlement that appeared in the first years after construction and then stopped points at consolidation. Movement that changes with the seasons points at soil volume. Movement that follows heavy rain and keeps going points at water transporting material.

None of that identifies which soil is under your property. It identifies which behaviour you are observing, which is the more useful thing anyway.

Settlement versus heave

Settlement is concrete going down because the ground under it lost volume or was carried away. Heave is concrete going up because something under it gained volume or pushed.

They are easy to confuse because both produce an uneven surface and a step at a joint. The tell is which side moved relative to what has not. A slab that has dropped away from a fixed reference — a door threshold, a step, a wall — has settled. A slab sitting higher than something that has not moved has heaved.

Heave has a small number of common causes: soil swelling with moisture, tree roots growing under a slab, and in freezing climates frost action lifting a slab where water in the ground beneath it froze and expanded. Root heave in particular is worth identifying, because it changes the options completely — lifting is irrelevant to a slab that is already too high, and replacing a section without addressing the root produces a new section that lifts in turn.

This distinction decides the method. Settlement can often be corrected by filling the space underneath and raising the slab. Heave cannot: there is nothing to fill, and the answer is either removing the high material or addressing whatever is pushing.

Warning signs a homeowner can observe

None of these is a diagnosis. They are observations worth writing down, because the pattern across several of them is far more informative than any single one.

• A step where two sections meet, and which side is higher.
• A gap where a slab has pulled away from a step, a wall or the house.
• A hollow or drummy sound when a slab is tapped or walked on in hard shoes — that suggests space underneath.
• Water pooling where it did not used to, or running back toward the building.
• Soil visibly washed away or eroded at a slab edge.
• Surface flaking, pitting or breaking away, with or without the stones showing through.
• Whether anything has visibly changed in the last year — and whether it changed after heavy rain, after a dry summer, or steadily.

Why two identical cracks can have completely different causes

A crack is a symptom with a long list of possible causes, and the crack itself carries very little information about which one applies. Two cracks of identical width, in identical concrete, can mean entirely different things.

Concrete shrinks as it cures and moves with temperature, and it has little tensile strength, so it relieves stress by cracking. A great deal of residential cracking is that and nothing more. Control joints exist precisely because cracking is expected — they are deliberate weak lines so the inevitable crack happens somewhere chosen.

The same-looking crack can also be the visible end of a slab losing support underneath, a slab that was loaded beyond what its thickness and reinforcement suit, a slab restrained from moving where it needed to, or a surface layer that was weak from the day it was finished.

What separates them is not the crack's appearance but its behaviour and its context. Is one side higher than the other? Is it still changing? Does water run into it? Is there settlement visible nearby? Does it follow a control joint, in which case the joint is doing its job? Those four questions distinguish more than measuring the width ever will.

This is also why photographing a crack next to something for scale, marking its ends with a dated pencil line, and looking again a month or two later is the single most valuable thing a homeowner can do. It costs nothing and it answers the question that decides the fix.

What a concrete contractor can actually evaluate

More than a homeowner and less than people expect. A competent contractor on site can sound a slab systematically to map where it is unsupported, probe the ground at an exposed edge to judge how soft the material is and how far down it stays soft, read the drainage pattern across the property, and on larger jobs drill a small test hole and look.

They can also tell you what they think caused the problem and what evidence they are basing that on — which is the most informative question you can ask any of them, and the one whose answers vary most.

What they generally cannot do is characterise the soil profile, quantify how much more movement to expect, or make a determination about the building's structure. Those are different disciplines, and a contractor who claims otherwise is overreaching.

When an engineer may be the right call instead

For residential flatwork — driveways, patios, walkways, garage floors, steps — a concrete contractor is usually the right first call, and involving an engineer would be disproportionate.

That changes when the concern involves the building rather than the flatwork. Cracking in an interior or basement wall, a wall that is leaning or bowing, doors or windows that have recently started sticking, floors inside the house that slope or sag, cracking in the foundation itself: those belong with a licensed structural engineer or a qualified foundation professional, and they belong there before anybody prices surface concrete.

It also changes where a repair has already been done and failed, where movement is large and clearly ongoing, or where a contractor tells you the problem may be structural. In that last case, take the answer seriously rather than continuing to get opinions until somebody says something more convenient.

An engineer's report costs money and produces a document that does not fix anything by itself. What it produces is a defensible account of what is happening, which is worth having before committing to work that assumes an answer.

Looking up the ground under your own property

Everything above describes how ground behaves in general. None of it says what is under your slab, and that gap is where most concrete advice on the internet quietly turns a regional map into a claim about one address. What follows is the honest alternative: the public datasets that exist, what each one actually reports, and — more usefully — what none of them can tell you.

The USDA Natural Resources Conservation Service publishes the Web Soil Survey, a free public tool that reports mapped soil properties for an area you draw on a map. Among the properties it carries are drainage class and linear extensibility, which is the measure soil scientists use for how much a soil changes volume as it wets and dries. That is the property behind the phrase “expansive soil”.

The U.S. Geological Survey publishes national mapping of karst — the soluble carbonate bedrock that produces sinkholes and sudden loss of support — and separately publishes material on land subsidence and aquifer-system compaction, which is the slow, broad settlement that groundwater withdrawal can cause. NOAA’s National Centers for Environmental Information publishes the U.S. Climate Normals, the 30-year station averages that include how often a location crosses the freezing point.

Between them those four cover the three mechanisms this guide has described: volume change in the soil, loss of support beneath the slab, and freeze–thaw exposure at the surface. Searching for any of them by name will find the publisher directly, and each is free.

Here is the part that matters more than any of it. Soil survey mapping is drawn at a scale that describes an area, not a parcel, and the map unit under your house may not describe your back garden. None of these datasets knows that the previous owner filled a low spot, that the driveway was poured over a trench, or that a downspout has been discharging against one corner for fifteen years. Those are the things that actually decide what your slab is doing, and no public dataset contains them.

So the reasonable use of all of this is to arrive at a contractor conversation better informed rather than more certain — to know that shrink–swell is plausible in your area and worth asking about, not to conclude that it is your answer. A probe in the ground still settles the question, and it is the only thing that does.

• USDA NRCS Web Soil Survey — mapped drainage class and linear extensibility for an area you select.
• USGS karst mapping — where soluble bedrock, and therefore sinkhole risk, is present.
• USGS land subsidence material — the mechanism behind broad, slow regional settlement.
• NOAA NCEI U.S. Climate Normals — station averages including days crossing freezing.
• None of the above — undocumented fill, a buried trench, or one badly aimed downspout.

Maintenance that may reduce future damage

None of this reverses damage already done, and none of it is a substitute for addressing a cause. What it can do is stop making things worse.

• Keep roof water away from slabs. Clear gutters, and extend downspout discharge so it releases well clear of concrete rather than beside it.
• Correct low spots where water collects against a slab edge.
• Keep joints and any stable cracks filled so water runs across the surface rather than into the ground beneath it. In freezing climates this is worth doing before winter rather than after.
• Reduce de-icing chemical use on vulnerable exterior concrete. Sand provides traction without the chemical effect.
• On new concrete, ask the contractor what they recommend for its first winter — a surface exposed to freezing and salt in its first season is being tested before it is ready.
• Photograph anything you are watching, with a date. A record is what turns "I think it has got worse" into something a contractor can act on.

Common questions

Sources

Referenced but not yet verified

These are the authoritative sources this guide draws its general framing from. None has been retrieved and confirmed by a person, so none is presented as a citation and no link is published for it — a source nobody has checked should not carry the weight of one. No statement in this guide depends on a figure from any of them.

• NOAA National Centers for Environmental Information — U.S. Climate Normals (1991–2020). Would support: Station-level winter temperature normals used to describe how often a metro crosses the freezing point. Status: not retrieved.
• NOAA National Weather Service — Climate records and local climate pages. Would support: Counts of days crossing the freezing point, used to compare freeze–thaw exposure between metros. Status: not retrieved.
• USDA Natural Resources Conservation Service — Web Soil Survey. Would support: Parcel-level soil mapping used to describe drainage class and shrink–swell behaviour under residential flatwork. Status: not retrieved.
• U.S. Geological Survey — Karst map of the conterminous United States. Would support: Distribution of soluble carbonate bedrock, used to describe shallow limestone and karst features in Middle Tennessee. Status: not retrieved.
• USDA Natural Resources Conservation Service — Soil Survey Manual (USDA Handbook 18). Would support: Definitions of shrink-swell behaviour and drainage class, used to describe how soils change volume with moisture in general terms. Status: not retrieved.
• U.S. Geological Survey — Land subsidence and aquifer-system compaction. Would support: General mechanism of consolidation and subsidence in unconsolidated material, used to describe how uncompacted fill settles over time. Status: not retrieved.