Cold Weather Pour Standard

Purpose

Cold weather concrete is a temperature-maintenance program with measured slab temperatures, not a blanket thrown over the work at quitting time. Concrete that freezes before it reaches roughly 500 psi loses strength it never recovers, because expanding ice tears the paste apart while it is still plastic enough to tear, and the damage stays hidden until the first spring thaw takes the top eighth of an inch off. Concrete that merely gets cold has a different problem: hydration slows toward a stop below 40 F, so a slab the customer expects to drive on in a week is not ready, and stripping to a calendar drops a wall panel.

Scope

Covers residential and light commercial cast-in-place work - flatwork, footings, foundation walls and small elevated slabs - from the go/no-go call the day before through protection removal and the controlled cooldown.

It does not cover mass concrete thermal control, which has its own gradient rules; the pre-pour subgrade and steel checks, owned by their own SOPs and still required in winter; or the strip decision, which the formwork stripping SOP releases against a strength number this procedure supplies.

Roles and responsibilities

Procedure

1. Make the call the day before, against the definition. Pull the 72 hour forecast and decide whether ACI 306R's cold weather period applies: average daily air temperature below 40 F for more than three consecutive days, with air temperature not above 50 F for more than half of any 24 hour period. Acceptance: the forecast captured onto the job record with the decision and who made it. Wrong looks like a call made from a glance at tomorrow's high, which misses the third night; stop rule, if the trigger is met and protection is not already staged on site, the order is not placed. Hazard: none here, though this call drives every hazard below, so it is not delegated.

2. Stage protection before the truck, and count it. Lay out blankets, ground thaw equipment, windbreak and heaters before placement, sizing blanket coverage to the slab plus at least 3 ft beyond every edge, since the perimeter loses heat two ways. Acceptance: blanket square footage on site at or above the calculated coverage plus seam laps, counted and written down. Wrong looks like blankets ordered to the slab area exactly, leaving edges bare, which is where the corner sensor fails two nights later; stop rule, no placement until the count passes. Hazard: propane cylinders stay upright, secured and outdoors, valve closed on any not connected.

3. Refuse frozen ground and ice in the forms. Probe subgrade temperature at four points across the pour and inspect every form for ice, frost and snow. Acceptance: subgrade above 32 F at every probe, forms and steel clear of ice, no standing water. Wrong looks like a crust that reads solid to a boot and thaws under a warm slab, dropping the panel where it stood; stop rule, thaw with hydronic blankets or covered ground heaters and re-probe. Never place concrete to melt frost. Hazard: fuel-fired ground heaters produce carbon monoxide, so they are never run inside an enclosed or partially enclosed space without combustion products vented outside it, and anyone working near one carries a personal CO monitor with an audible alarm.

4. Order the mix to a temperature and verify it at the chute. Ask the plant for heated mixing water and, where the drawings allow, a non-chloride accelerator, since chloride-bearing accelerators add to the water-soluble chloride ion content ACI 318 limits by exposure class and are the wrong tool anywhere steel is embedded. Acceptance: the ticket shows the as-mixed temperature, and concrete temperature measured per ASTM C1064 at discharge is at or above the ACI 306R as-placed minimum, 55 F for any least dimension under 12 in. Wrong looks like a load leaving the plant right and arriving cold after a long queue; stop rule, reject it at the chute and record the reading with the ticket number. Hazard: fresh concrete burns skin, so waterproof gloves and sleeves stay on and contact is flushed with clean water at length.

5. Let the bleed finish before you close the surface. Cold concrete bleeds later and longer than the crew is used to, so hold off floating and troweling until the bleed water is gone and the surface takes a footprint of about a quarter inch. Acceptance: no finishing while bleed water stands, recorded as finishing start against placement finish. Wrong looks like a crew closing on schedule and sealing bleed water under a troweled skin, which delaminates in sheets the following summer; stop rule, wait, and light the work rather than finish early. Hazard: a power trowel on a slab that has not set will bog and grab, so keep hands clear of the guard ring and never run one where a boot still sinks in.

6. Hold 50 F at the slab, measured at the coldest point. Cover the slab and hold it at or above 50 F for the protection period, three days for ordinary mixes and seven where high early strength or an early strip governs, using two sensors, one at the coldest corner and one mid-slab. Acceptance: both logged twice daily, minimum readings at or above 50 F for the full period. Wrong looks like air temperature under the blanket logged as slab temperature, which reads high and hides a failing corner; stop rule, on any reading below 50 F add insulation or heat that hour, log it, and extend the period rather than counting the cold hours toward it. Hazard: unvented fuel-fired heaters inside an enclosure do two separate harms, filling the space with carbon monoxide and carbonating the fresh surface into a soft dusting floor, so combustion products are vented outside the enclosure, a personal CO monitor with an audible alarm is worn by anyone entering, and exposure is held under the OSHA construction limit of 50 ppm as an 8 hour time-weighted average at 29 CFR 1926.55.

7. Prove 500 psi before the first freeze. Break field-cured cylinders made and cured per ASTM C31 alongside the slab, or use a maturity method correlated to that mix, before any exposure to freezing. Acceptance: a documented strength at or above 500 psi, the ACI 306R threshold for resisting a single freezing cycle, with date and hour. Wrong looks like a crew assuming three days under blankets did it, on a mix whose accelerator was omitted at the plant; stop rule, keep the protection on and re-break, because none of it comes off on an assumption. Hazard: the cylinders live in the slab's protection, not in a truck cab.

8. Take the protection off gradually and measure the drop. Remove blankets in stages, loosening rather than pulling, and measure surface temperature against ambient before and after. Acceptance: surface temperature drop in the first 24 hours after removal no greater than 50 F for a section under 12 in least dimension, the ACI 306R limit for that size. Wrong looks like blankets pulled at noon on a bright day and the surface crashing 60 F overnight, which cracks the top in a pattern the customer reads as bad concrete; stop rule, lay the blankets loosely back on and stage removal over another day. Hazard: frozen blankets are heavy and slick, so roll rather than drag them and stay off the slab edge while carrying one.

The record this produces

One protection log per pour. Fields: job and pour identifier, the forecast and go/no-go decision with the decider's name, staged blanket area against calculated coverage, four subgrade probe temperatures, batch tickets with as-mixed and measured discharge temperatures per load, placement finish and finishing start times, sensor locations, every logged reading with date and hour, any excursion with its value, action and hour, field-cured cylinder identifications with break age and result, and removal date with before and after surface and ambient temperatures.

It lives on the job record with the batch tickets. The office reads it to release the strip and set the customer's real date. The engineer reads it if a strength question opens, and a continuous sensor log is the difference between a defensible pour and an argument. The estimator reads it next winter, because a shop logging corner excursions on every November pour is short on blankets, not on luck.

One pass, with a step that failed

A 26 ft by 40 ft attached garage slab, 5 in thick, ordinary mix, late November. Slab area is 26 x 40 = 1,040 sq ft. Step 2 coverage: 3 ft beyond each edge gives 32 by 46 ft, or 1,472 sq ft. At 150 sq ft per blanket that is 9.8, so 10 minimum, and the crew staged 12 for seam laps.

Step 1 passed: forecast average daily air temperatures of 31, 28 and 33 F on three consecutive days, all below 40 F, so the period applied. Step 3 passed with subgrade probes at 39, 41, 38 and 40 F after a night under hydronic blankets. Step 4 passed: batched at 68 F, discharge measured 61, 60 and 58 F across three loads, all above the 55 F as-placed floor. Step 5 held finishing until 15:10 against a placement finish of 12:35.

Step 6 failed on night two. The mid-slab sensor read 58 F but the northwest corner read 41 F against the 50 F floor, ambient 19 F, wind across an open side of the garage. The crew leader logged the excursion at 06:40, added a second hydronic line under the corner, doubled the blanket over the outer 6 ft and hung a tarp windbreak. The corner read 53 F four hours later and 56 F that evening. Because the cold hours do not count toward protection, the three day period was extended by one day and the office moved the customer's date.

Step 7 passed on the extended schedule: two field-cured cylinders broken at day three read 1,120 and 1,190 psi, averaging 1,155 psi, well above the 500 psi threshold. Step 8 passed too. At removal the surface read 62 F against 24 F ambient, a 38 F differential, inside the 50 F first-day limit, so the blankets came off in two stages over the afternoon.

What the corner costs unmeasured is specific. A slab edge at 41 F is not freezing, so nothing dramatic happens that night, but hydration in that outer band nearly stalls while the middle keeps gaining. The perimeter goes into the first hard freeze materially weaker than the core, and it surfaces the following March as scaling along the door line, in the strip the customer sees daily.

When the forecast turns mid-pour

Temperature drops faster than forecast during placement: keep placing while the concrete still meets the as-placed minimum at the chute, and cover behind the finishers rather than waiting for the whole slab. The customer wants blankets off for a delivery: the answer is the strength value or the differential limit, and the call names the number and the date it will be met. Where the site cannot hold 50 F at all, the honest answer is not to pour, because cold weather concrete without a maintainable protection plan is a warranty claim scheduled for spring.

References

• ACI 306R, Guide to Cold Weather Concreting, for the period definition, minimum temperatures by section size, the protection period, the 500 psi freeze threshold and the maximum drop after protection, in the edition your specification names. See related: the ACI cold weather pour trigger card in this library.
• ACI 318 for water-soluble chloride ion limits by exposure class, which govern accelerator selection where steel is embedded, in the edition your building department has adopted.
• ASTM C1064 for discharge temperature and ASTM C31/C31M for field-cured cylinders. See related: the test cylinder sampling and handling SOP.
• 29 CFR 1926.55 for the carbon monoxide limit in construction, with the heater manufacturer's venting and clearance instructions.
• See related: the formwork stripping and shoring removal SOP, which consumes this strength value.