Two driveways on the same street, poured the same summer by the same crew. Ten winters later, one still looks fine. The other is crumbling across the bottom third, where it meets the garage.
Same concrete, same weather, same salt truck. The difference is almost always water: where it sits, how long it stays, and what's dissolved in it. Cold does little to dry concrete. Wet concrete is another story.
What's Inside a Slab
Concrete looks solid, but it's full of empty space you can't see.
- Capillary pores. Concrete is mixed with more water than the cement needs. The extra water eventually leaves behind a network of tiny connected pores. Wetter mixes leave more.
- Entrained air bubbles. Concrete made for freezing climates gets an admixture that creates billions of microscopic bubbles, roughly 0.05 to 1.25 millimeters across, as relief space for freezing water.
- Cracks and joints. The big openings that let water reach deep into the slab.
The detail that explains most winter damage: water fills the small pores fast, in about 10 to 18 hours in a standard lab absorption test. The air bubbles fill slowly, over months to years of staying wet. A slab that gets rained on and dries out never fills its bubbles. A slab that sits wet all winter starts to.
The Number That Decides Everything
Water expands about 9% when it freezes. If the pores are partly empty, the ice and the water it pushes ahead of it have somewhere to go. If the pores are nearly full, that pressure goes into the cement paste, and the paste cracks a little.
Engineers call the tipping point the critical degree of saturation. The Federal Highway Administration puts it between 75% and 91% of the pore space, with recent research pointing to about 86%. Above that line, concrete can be damaged in a single freeze-thaw cycle no matter how much air it has. Below it, sound concrete handles winter after winter with little harm.
So the useful question isn't "how cold did it get?" It's "how wet was the concrete when it froze?"
| Condition | In the pores | What winter does |
|---|---|---|
| Sheds water, dries between storms | Small pores fill and empty; bubbles stay mostly dry | Very little |
| Damp for days at a time | Small pores stay full; bubbles slowly take on water | Slow wear at the surface |
| Soaked for weeks (a puddle, a snow pile, a crack holding water) | Bubbles fill until saturation crosses the line | Damage with every freeze |
What Air Entrainment Does, and What It Can't
When water in a pore starts to freeze, the unfrozen water gets pushed ahead of the ice. If an empty bubble is close by, the water moves into it and the pressure drops. That's why spacing matters as much as total air: the common guideline is a spacing factor of 0.008 inch or less, so no spot in the paste is far from a bubble.
For a Wisconsin driveway that sees deicers, industry guidance calls for roughly 6% to 7% air with typical 3/4" stone. Some gets lost: FHWA notes a 1% to 2% loss is common during pumping and consolidation, and overworking the surface during finishing knocks air out of the top layer, right where the salt hits.
What air can't do is make concrete waterproof. A study of freeze-thaw damage at concrete pavement joints found that once saturation passed 86–88%, freeze-thaw damage happened with or without entrained air. Air buys time. Keeping the slab from staying wet keeps you from using it up.
Why Salt Makes It Worse
FHWA's review of deicers and concrete pavement lists several ways deicers intensify freeze-thaw damage, including higher saturation and thermal shock. On a driveway, that looks like this:
- The surface stays wetter. Brine soaks into the pores and pushes saturation up.
- The surface cycles more. Salt melts ice on days the slab would have stayed frozen. The meltwater soaks in, then refreezes when the brine dilutes or the sun goes down.
- Melting pulls heat out fast. Turning ice to water draws heat from the top of the slab, and that sudden chill stresses the surface layer.
- A weak brine is often the worst. Lab studies keep finding that surface scaling peaks with a mild solution of a few percent salt, not plain water or strong brine. That's the kind of dilute slush salt leaves behind.
- Calcium and magnesium chloride can attack the paste itself. They react with calcium hydroxide in the cement to form calcium oxychloride, which expands as it forms. FHWA notes this can happen from 32°F to 122°F depending on concentration, and lab mortar soaked in these salts at 40°F cracked without ever freezing. It shows up first in joints, where brine collects.
Two rules beat any bag label: no deicer on concrete in its first winter (use sand), and never use ammonium nitrate or ammonium sulfate, which attack concrete chemically. Both are common in fertilizer, so sweep spilled lawn fertilizer off the driveway too.
How Many Freeze-Thaw Cycles a Green Bay Winter Brings
Engineering weather data for Green Bay, based on 1973–1996 records, lists an average of 66 freeze-thaw cycles a year. Across Great Lakes weather stations, GLISA researchers found a long-term average closer to 42, with counts generally falling in recent decades. Either way, it's dozens every winter.
Those counts track air temperature, though. What your concrete feels depends on where it sits.
| Where the concrete is | What it goes through | Risk |
|---|---|---|
| South-facing driveway or apron | Sun thaws it on cold days, it refreezes at night, and salt and car drip add more melt | Highest |
| Unheated garage floor | Salty slush melts off the car by day and refreezes overnight, storm after storm | High at tire tracks and the door line |
| Front stoop and steps | Exposed on several sides and salted heavily | High at edges |
| North-side walk under snow | Snow insulates it, so it cycles less, but it stays wet into spring | Lower, unless water ponds |
WisDOT notes that sun can heat blacktop into the 50s when the air is in the 20s or 30s. Concrete runs cooler, but a sunny slab still thaws, and soaks up more water, on plenty of days the thermometer says it shouldn't.
What the Damage Looks Like
| Damage | What you see | Where it starts | What's going on | How serious |
|---|---|---|---|---|
| Scaling | Top skin flakes off in patches; stone shows as it worsens | Low spots, tire tracks, the garage door line | Saturated surface breaking loose each freeze, faster with salt or a weak finish | Cosmetic early, but it spreads |
| Dark joints | Joint edges stay dark long after the slab dries | Saw-cut joints that hold water and brine | FHWA describes this darkening as the first stage of joint deterioration | An early warning |
| Crumbling crack edges | Edges round over; the crack looks wider at the top | Any crack that holds water | Crack faces soaked and freezing from both sides | Moderate, worse each winter |
| Popouts | Cone-shaped pits with a split stone at the bottom | Scattered across the surface | A porous stone near the surface soaks up water and splits | Usually cosmetic |
| D-cracking | Fine cracks running alongside a joint, often with dark staining | Next to joints and cracks | Susceptible coarse stone in the mix breaking down | Serious. It can break up concrete within a foot or two of the joint. |
| Deep spalling | Chunks breaking out, sometimes with rust stains | Edges, corners, over steel | Advanced damage, or rusting steel pushing concrete off | Serious. Get it looked at. |
D-cracking is a mix problem, more common on older streets than driveways, and no sealer fixes it. For flaking and spalling repairs, see why concrete flakes.
Worked Example: One Crack, One Winter
Take one crack at the low end of a driveway, near the garage. The numbers are illustrative; the physics is real.
The crack: 6 feet long, 3/16" wide, through a 4" slab. Silt has packed into the bottom, so it holds water like a trough.
- Water it holds: 72" × 3/16" × 4" = 54 cubic inches, a little under a quart.
- Extra volume when it freezes: about 9%, roughly 5 cubic inches that has to go somewhere.
- Concrete sitting in water: two crack faces, 72" long and 4" deep, about 4 square feet in all.
| When | What happens |
|---|---|
| Late October | Rain fills the crack. Within a day, the small pores in both faces are full. |
| November–December | Hard freezes start. The water freezes from the top down, so an ice plug traps the water below, and when that freezes it presses on the crack walls. The bubbles are still mostly empty, so damage is minor. |
| January | Salty slush runs in. The crack now holds brine, which soaks deeper. |
| February | Thaw days melt it, top it off, and refreeze it. Four months in, the faces have had the kind of long soak that slowly fills air bubbles. |
| March–April | The top edges crumble and the crack looks wider. Water that reached the soil has been feeding frost heave right under it. |
Nobody can name the day those faces crossed the critical line, but a crack that holds water from October to April is set up to cross it.
Now the same crack, sealed in September. It's cleaned out, closed-cell backer rod goes in, and flexible joint sealant goes on top. Slush runs across it and off the slab. The faces get wet now and then and dry between storms, so they stay well under the line. That's a few dollars of material.
Scale it up and you have the sunken slab: a low spot holding a half inch of slush against the garage door is that same crack, spread over 50 square feet.
Prevention, Ranked by Impact
Every step serves one goal: keep the concrete under that critical saturation level for as much of the winter as possible.
| Rank | What to do | Why it ranks here | Who does it |
|---|---|---|---|
| 1 | Get water off and away from the slab | Nothing else holds up if the slab sits in water | Homeowner; us, if a sunken slab holds water |
| 2 | Seal joints and cracks | Closes the openings that hold brine | DIY or us |
| 3 | Apply a penetrating sealer | Cuts surface absorption | DIY or us |
| 4 | Cut the brine: lighter deicer use, snow piled on the lawn | Limits extra wetting, chemical attack, and weeks of salty melt | Homeowner |
1. Drainage. Extend downspouts onto ground that slopes away, point the sump discharge away from concrete, and build up low soil at slab edges. A slab that sank toward the garage or house pools meltwater all winter. Foam lifting raises it back to slope through 5/8" holes, and you can drive on it the same day.
2. Joints and cracks. Use a flexible concrete joint sealant over closed-cell backer rod where the joint is deep, on a dry, frost-free day around 40°F or warmer. Skip rigid patch outdoors; it cracks loose. Replace failed sealant too, since FHWA notes brine gets in through failed joint seals, and torn sealant and backer rod left in place can slow drying. Our concrete crack caulk guide covers products and prep.
3. Penetrating sealer. Silane and siloxane sealers soak in and line the pores instead of forming a film, so water and brine bead up while the slab can still breathe. Industry testing gives them the best deicer-scaling resistance among common sealers, and a good one typically lasts about 5–10 years. Apply on clean, dry concrete at about 40°F or warmer, after new concrete has cured as long as the label says (often 2–4 weeks) and after joint sealant has cured. It ranks third because it can't stop water wicking up from wet soil or a puddle soaking in. See choosing a concrete sealer.
4. Less brine. Shovel first, use deicer only to break the ice's grip, keep brine out of joints, and rinse the apron in spring. Plowed and shoveled piles are loaded with salt and melt for weeks, so push snow onto the lawn, away from slab edges and the foundation.
When It's Past Prevention
Prevention slows damage; it doesn't undo it. Deep spalling, spreading D-cracking, or concrete that crumbles under a screwdriver calls for repair or replacement. We don't do replacement, but these questions will help you hire it right. The exception is the garage floor: a sound but scaled floor can usually be ground, filled, and coated, so brine sits on top instead of soaking in. That's one of our services.