Curing Concrete in Heat and Cold

Two clocks, one slab

Concrete does not dry. It reacts. Cement grains and water form calcium silicate hydrate, and that reaction needs liquid water present inside the paste for as long as strength is still being gained. That is the first clock — hydration — and it runs at a speed set almost entirely by the temperature of the concrete itself, not the air around it.

The second clock is evaporation. Water leaves the surface of fresh concrete because the air above it is drier, warmer, or moving faster than the slab. Once the rate of loss exceeds the rate at which bleed water rises to replace it, the surface goes into tension while it still has no tensile strength worth the name. Plastic shrinkage cracking follows, usually within the first two to four hours, usually in a pattern that looks random until you notice it runs perpendicular to the wind.

Everything a crew does on a hot day or a cold one is an intervention on one of those two clocks. Fogging, sunshades and wind breaks slow evaporation. Blankets, heated enclosures and accelerating admixtures speed hydration. Curing compound seals the surface so the water already in the mix stays in the mix. Cold weather flips the problem entirely — evaporation slows on its own, but hydration slows faster, and the danger becomes freezing rather than cracking.

Sorting the work this way changes what gets argued about on site. The question is never "should we cure?" It is "which clock am I losing, and what do I have on the truck that touches it?" A superintendent who understands that will not waste a fogger on a 4°C morning, and will not waste an insulating blanket on a 34°C afternoon with a 25 km/h wind.

Reading the evaporation clock before the truck arrives

Four variables set evaporation rate: air temperature, concrete temperature, relative humidity, and wind speed. Wind is the one crews consistently underestimate. Doubling wind speed can do more damage than a 5°C rise in air temperature, and an open slab on a bare deck with no perimeter walls is the worst case anyone routinely pours.

ACI 305R, Guide to Hot Weather Concreting, presents the relationship as a nomograph and gives a widely used threshold above which precautions against plastic shrinkage cracking are warranted. Below that threshold you are not safe, merely less exposed — mixes with low bleed rates, silica fume mixes and many fibre-reinforced mixes crack well under the classic threshold because they deliver almost no bleed water to the surface at all. For those, treat any measurable evaporation rate as a reason to fog.

Concrete temperature carries more weight in the calculation than air temperature, and it is the variable you can actually control before the pour. Cement heat of hydration then pushes it back up, so a slab placed warm on a hot afternoon may peak considerably higher in the core overnight — which matters for thermal differential, not just for surface loss.

Check the numbers the morning of, not the week before. A forecast that promised 40 percent humidity and light air is a different job when the actual is 18 percent and gusting. Reading site conditions an hour before the first truck decides whether the fog line goes out, whether the sunshade goes up, and whether the pour moves to a night placement.

The temperature you deliver sets both clocks

Delivered concrete temperature is the single lever that reaches both clocks at once. Warmer concrete evaporates faster and hydrates faster; colder concrete does neither. Which is why the batch plant conversation happens before the site conversation, and why arguing about fog nozzles while accepting a 34°C delivery is arguing about the wrong thing.

Cooling on the hot side comes from chilled mix water, ice substituted for part of the mix water, sprinkled and shaded coarse aggregate stockpiles, and — on large placements — liquid nitrogen injection. Heating on the cold side comes from heated mix water and heated aggregates. Water is the easiest component to condition because it has the highest specific heat; aggregate is the hardest because there is so much of it, which is precisely why frozen lumps in a winter stockpile are a rejection item rather than a nuisance.

Specifications usually set both a maximum and a minimum as-delivered temperature, and the numbers vary by jurisdiction and by the governing project specification — ACI 305.1, Specification for Hot Weather Concreting, and ACI 306.1, Specification for Cold Weather Concreting, are the common references, but a transport authority or a local building code may impose its own limits. Confirm which document your contract cites before assuming a number.

Haul time and drum revolutions belong in the same conversation. A long haul in the heat with the drum turning adds temperature and consumes slump; retempering with water on arrival is how a specified mix quietly becomes a weaker one. Schedule the truck spacing so the crew is not tempted, and if placement genuinely cannot keep up, stop the pour rather than water the mix.

Slowing the evaporation clock: shade, wind and water in the air

Wind breaks are the cheapest intervention and the most neglected. Snow fence, shade cloth on the perimeter scaffold, or a run of hoarding on the windward edge cuts surface air velocity and takes the sharpest term out of the evaporation equation. On a deck with no walls, temporary screens on two sides will often halve the loss rate.

Fog — not spray — is the tool for the air above the slab. A fog nozzle produces a mist that raises local humidity without landing enough water on the surface to raise the water-cement ratio at the top. Anyone who has watched a crew hose a slab "to keep it wet" has seen the soft, dusty, chalky surface that shows up six months later. Fogging goes on between screed and float, and again between float and final trowel, in the windows where the surface has no protection and no strength.

Sunshades over the working area drop the concrete surface temperature directly. On a bridge deck or a large industrial floor this is a real logistics job — frames, cloth, tie-downs that survive the same wind you are trying to block — but the alternative is a surface that sets ahead of the finishers and tears under the trowel.

Evaporation retarders sit in a category of their own. They are monomolecular films sprayed on the surface between finishing passes to hold water in until the next operation. They are not curing compounds and do not replace curing; a crew that treats them as a final coat has left the slab uncured.

Timing finishing operations also changes in the heat. Bleed water can evaporate as fast as it appears, so the surface looks ready to trowel long before the concrete underneath has stiffened. Troweling into that produces a sealed skin over a soft body, and the delamination shows up later as hollow-sounding blisters underfoot.

Sealing the slab: membrane curing and its coverage rate

Once finishing is done, the decision is water curing or membrane curing. Ponding, wet burlap and soaker hoses keep liquid water at the surface and are unambiguously the better cure for anything where surface durability governs — exposed floors, decks that will see de-icing salts, anything getting a shake-on hardener. They demand labour: burlap that dries out is worse than no burlap, because it wicks water out of the slab.

Membrane curing compound is what most jobs actually use, and its performance depends on applying enough of it evenly. ASTM C309, Standard Specification for Liquid Membrane-Forming Compounds for Curing Concrete, and ASTM C1315, Standard Specification for Liquid Membrane-Forming Compounds Having Special Properties for Curing and Sealing Concrete, define moisture-retention performance; the manufacturer's data sheet governs coverage rate for a given product, and it is not negotiable downward to make a drum stretch. Thin, streaky application produces a slab that is well cured in stripes.

Surface texture drives consumption. A broom-finished exterior slab, a tined bridge deck or a raked surface has substantially more area than its plan dimensions suggest, and coverage rates quoted for a steel-troweled floor will not cover it. Two coats applied at right angles is standard practice on textured work, and it doubles the material take-off.

Watch what the compound does to what comes next. Many curing compounds interfere with the bond of toppings, coatings, adhesives and traffic membranes. If the slab is receiving a floor covering, either specify a dissipating-resin compound the coating manufacturer accepts, or plan on mechanical removal — which is real money and real schedule. Where a curing-and-sealing compound is used to save a step, confirm compatibility before the drums arrive, not after.

Speeding the hydration clock when the air turns cold

Cold weather inverts the whole problem. Evaporation stops being the headline threat — though a low-humidity winter day with dry wind still pulls water off a slab, and heated enclosures produce some of the driest air on any site. What matters instead is that hydration slows sharply as concrete temperature falls, and stops effectively near freezing.

ACI 306R, Guide to Cold Weather Concreting, sets out the operating logic: maintain the concrete above a minimum temperature for a protection period, and never let it freeze before it has reached the strength at which freezing does no permanent damage. Concrete frozen before that point suffers expansion of the pore water that disrupts the paste permanently. The slab does not recover when it thaws. That is a demolition item, not a defect to be monitored.

Practically, the cold-weather kit is heated mix water, heated aggregates, accelerating admixtures, insulating blankets, insulated forms and heated enclosures. Non-chloride accelerators are the default where reinforcement or prestressing steel is present; calcium chloride is restricted or prohibited in reinforced and prestressed work by the governing project specification and by the applicable structural concrete code in your jurisdiction, so check what your specification actually cites before ordering an accelerated mix.

Formwork is a heat asset. Steel forms strip heat out of a column; insulated forms hold hydration heat in and can carry a member through a cold night on its own reaction. Corners, edges and thin sections lose heat from multiple faces and cool fastest — a slab edge protected by a blanket laid to the perimeter and no further will freeze while the middle is comfortably warm. Blankets need to run past the edge and down the form face.

Sudden removal of protection is its own failure mode. Concrete warm inside and cold outside develops a thermal gradient, and pulling a blanket off a warm slab into a −10°C morning cracks it as surely as any shrinkage does. Step the protection down — vent the enclosure, peel blankets progressively — so the differential across the section stays inside what the specification permits.

Direct-fired heaters carry a hazard of their own: combustion products include carbon dioxide, which carbonates a fresh concrete surface and leaves it soft, dusty and unable to take a hardener. Vent them, or use indirect-fired units where the exhaust stays outside the enclosure.

Knowing when the clock has run out

Curing duration written as a fixed number of days is a proxy, not a measurement. Seven days at 20°C and seven days at 4°C are different amounts of hydration. Specifications give a nominal period — commonly seven days for many mixes, longer for slower-gaining mixes with high supplementary cementitious material content — but the underlying requirement is a strength or maturity threshold.

Maturity methods let you stop guessing. Temperature sensors embedded at the point of interest log the concrete's own thermal history, and a maturity index calibrated against cylinder breaks from the same mix gives in-place strength at any moment. ASTM C1074, Standard Practice for Estimating Concrete Strength by the Maturity Method, covers the calibration; the calibration is mix-specific, so a change of supplier or a change of cement source invalidates it.

Field-cured cylinders remain the fallback and remain worth their trouble, so long as they are stored beside the element and protected exactly the way the element is protected. A cylinder set that spent the night in a heated site office proves nothing about the slab it came from. Acceptance cylinders — standard-cured — answer a different question entirely and should never be used to make a stripping decision.

Sensors, cylinders and blanket coverage all feed the same judgement: has enough hydration happened at the coldest, most exposed point in the pour? That point is almost never where anyone puts the probe by default. Set it at an edge, at a corner, at the top of a wall where the blanket runs short — and let the number, rather than the calendar, release the protection.