A strength shortfall gets argued over for a month. A weighing error never gets argued over at all, because by the time it shows up in a cylinder break nobody is looking at the scale that caused it.
That is the asymmetry worth understanding before you sign a supply agreement. Every party on a project reviews the mix design. Almost nobody reviews the only equipment that reproduces it — the load cells under the aggregate bin, the cement scale, the water meter, and the moisture probe that decides how much of that water is already in the sand. A mix design is an instruction. A batching tolerance is the margin within which the plant is permitted to disobey it.
Why a batch weight tolerance is really a strength tolerance
Proportioning errors do not land evenly. An error in coarse aggregate changes workability and yield. An error in cement or water changes the water-to-cement ratio, and the water-to-cement ratio is the variable your strength, permeability and durability class all hang from.
This is why plants proportion stronger than your specified strength in the first place: the required average strength is set above the specified strength precisely to absorb production variation, which we cover in detail in our guide to required average compressive strength. Batching accuracy is one of the inputs to that variation. Tighten it and the overdesign has less work to do. Leave it unstated and you have accepted whatever the plant’s equipment happens to hold.
What an owner’s specification actually demands of the scale
Canadian public owners write this down, and their clauses are the most useful public benchmark available, because the standards that set the underlying tolerances sit behind a paywall. Manitoba Infrastructure’s concrete specification is unusually explicit.
On accuracy, Specification 830 requires that “Scales shall be designed and maintained to within 0.5% accuracy” for cement, water and aggregates, and sets the resolution as well: minimum increments “shall not exceed 2kg for the cement scale” and “10kg for the aggregate scale.”
On the batch itself: “The weight of aggregates and cement from the proportioning plant shall be within 0.5% of the designated batch weights,” with the proviso that batch weights “may be adjusted to allow for moisture variations in the aggregates.”
On who controls the equipment, the clauses are tighter than most private specifications ever get. Weighing devices “shall be tested and their accuracy approved before the construction starts,” they “may be tested with standard weights at any time,” and “Adjustments to the weighing devices shall not be made without the Department’s approval.”
The plant configuration is specified too: cement and fine and coarse aggregates proportioned by mass, separate bins or compartments for cement, fly ash and each type and size of aggregate, water and liquid admixtures proportioned by volume or mass, and dispensing equipment for liquid admixtures that is automated.
Read together, those clauses do something a tolerance number alone cannot. They make the tolerance verifiable, and they take the ability to quietly re-zero a scale out of the producer’s hands.
Where plant certification fits, and what it does not prove
Plant certification is a genuinely useful signal, and it is routinely misread as something stronger than it is.
Under the NRMCA plant certification scheme, accuracy checks on scales are performed “at least once every 6 months,” and where a public owner requires them more often, the more frequent requirement governs. Water meter accuracy is also verified at least every six months, with a required accuracy of ±1.5%. The company official agrees to “check the batching accuracy of dispensers of liquid admixtures at least every 6 months.” For moisture probes, no interval is set — the manufacturer’s directions govern.
Certification itself is time-limited: plant certification is “valid for 2 years from the date of inspection,” and delivery vehicle certification is valid for 12 months.
Then the part that should shape how much weight you place on the certificate. “The company hires the inspector and the fees established are between the company and the inspector,” and “A licensed engineer working for the company and approved by NRMCA is permitted to perform the inspections.” Delivery vehicles may be inspected in-house by company personnel, with those procedures audited during the plant inspection.
That is a documented, standardised, self-commissioned audit. It is far better than nothing, and it is not an independent regulatory inspection of your project’s concrete. The certification check list also defers the tolerance values themselves to ASTM C94 rather than restating them, which means the certificate confirms a process was followed — not which numbers your pour was held to. If you want a specific tolerance, you write it into the specification. No public Canadian source publishes the CSA tolerance table, so the governing number on an Alberta private project is whatever the project specification says it is.
Why aggregate moisture is the tolerance that moves during the day
Scales drift slowly. Aggregate moisture changes between loads.
Free moisture in the aggregate has to be taken into account when determining batch weights, and coarse aggregate matters here while fine aggregate matters more — a sand stockpile holds far more free water per tonne and varies far more across a pile. NRMCA’s quality control guidance treats probe readings as something to be checked rather than trusted, directing that probe or Speedy tester readings “should be periodically correlated to weighed and dried moisture samples.”
The arithmetic shows why. Take a cubic metre proportioned with 800 kg of fine aggregate and 350 kg of cement, and assume the plant’s moisture correction for the sand is off by one percentage point. That is 8 kg of unaccounted water per cubic metre, and against 350 kg of cement it moves the water-to-cement ratio by roughly 0.023 — on a mix specified at 0.40, a drift to about 0.42. No scale was out of tolerance. No ticket looks wrong. The mix is simply not the mix that was designed, and in a durability class where the ratio is the acceptance criterion, that is the whole argument.
This is also the mechanism behind most disputes that get blamed on water added at the site. The water the plant failed to subtract and the water a crew asked for at the chute arrive at the same place in the equation.
What has to hold, and how often it is checked
| Item | What it has to hold | How it is verified |
|---|---|---|
| Cement, water and aggregate scales | 0.5% accuracy; increments not exceeding 2 kg (cement) and 10 kg (aggregate) under Manitoba 830 | Tested and approved before construction; standard weights at any time; no adjustment without the owner’s approval |
| Batch weights | Within 0.5% of designated batch weights, adjusted for aggregate moisture | Percent deviation calculated between targeted and actual weight for each ingredient |
| Water meter | ±1.5% under the NRMCA scheme | Accuracy verified at least every six months |
| Liquid admixture dispensers | Batching accuracy within the scheme’s limits; automated dispensing under Manitoba 830 | Batching accuracy checked at least every six months |
| Moisture probes | Manufacturer’s stated accuracy | Readings periodically correlated to weighed and dried samples |
| Plant certification | Full check list compliance | Valid two years from inspection date |
| Delivery vehicles | Check list compliance | Valid twelve months; in-house inspection permitted, audited at plant inspection |
Two of those rows are the ones that get skipped in practice: the moisture correlation record, because it is a comparison rather than a certificate, and the batch deviation calculation, because it requires somebody to compare target against actual rather than glance at a printout.
What the batch record shows, and what it cannot
Recording equipment has to register the quantities of ingredients batched and allow the particular batch to be identified against the corresponding delivery ticket. Manitoba requires a delivery ticket issued at the batch plant for each load, carrying at minimum the contract number, truck number, the date and time loaded or the time of first cement-aggregate mixing, and the volume of the load.
That makes a ticket traceable. It does not make it an accuracy report, and the two are often confused. Reading the ticket tells you what was ordered and when the clock started — the subject of our breakdown of the numbers on a concrete batch ticket. Proving the tolerance requires the recorded actual weights set against the target weights, batch by batch. Nor does any of it speak to whether the drum is still mixing what it received, which is a separate question answered by the mixer uniformity test.
Where Alberta’s own public specifications draw their lines
Alberta Transportation’s supply specification for Portland cement concrete is worth reading even on a private job, because it shows which controls a sophisticated Alberta owner treats as non-negotiable.
Concrete temperature at discharge “shall be between 10°C and 25°C.” On-site addition of water “may be permitted by the Consultant,” but only where “no more than 60 minutes have elapsed from the time of batching,” and slump or air adjustments have to be made “within 90 minutes from batching of the concrete.” Sampling, testing and mix design must be done by “a CCIL/CSA certified and qualified concrete testing laboratory that has a permit to practice in the Province of Alberta.” Quality control testing is the contractor’s responsibility; the consultant takes samples and performs quality assurance testing and inspection.
Manitoba sets the mixing and delivery envelope in comparable terms: concrete in agitating trucks “shall be discharged completely within 90 minutes after the cement was added to the aggregates,” non-agitating trucks within 30 minutes, stationary mixers of one cubic metre or less get a minimum mixing time of one minute with 20 seconds added per additional cubic metre, and “all water shall be in the drum by the end of the first one-quarter of the specified mixing time.”
None of these are exotic requirements. They are what a specification looks like when the owner intends to be able to prove something later.
Four lines to put in the specification before the first truck
- State the tolerance and the method of verification. A tolerance with no stated verification method is a preference. Name the percentage and require the target-versus-actual deviation calculation per ingredient.
- Require dated calibration records. Scales, water meter and admixture dispensers, with the dates of the last accuracy check, supplied before the first pour rather than after a dispute.
- Require the moisture correlation record. Probe readings correlated against weighed and oven-dried samples are the only evidence that the batch water is being corrected against reality.
- Say who may adjust the equipment, and when. Public owners reserve approval of any adjustment to weighing devices. Private specifications rarely do, and it costs nothing to add.
Send us the project before the mix is locked
Most batching arguments we are asked into were decided at the specification stage, when a durability class was written against a water-to-cement ratio and nothing was written about how that ratio would be proven. The cheapest time to settle it is before the first truck is scheduled.
Send us your project details — pour date, mix type, PSI, volume — and our team will get back to you within 1 business hour with a quote and delivery slot. For mixes proportioned to a specific durability or strength requirement, see custom mixed concrete in Calgary, or read how our volumetric mixers work for projects where the proportions are set at the site rather than at a central plant.
FAQ
What batching tolerance applies on a private Alberta project? Whatever the project specification states. The underlying standard tables are not published publicly, and plant certification check lists defer the values to ASTM C94 rather than restating them, so a private job with a silent specification inherits the plant’s own practice.
Is a certified plant held to a tighter tolerance than an uncertified one? Not necessarily tighter — more verifiably. Certification confirms the equipment has been checked on a schedule, scales at least every six months, water meters to ±1.5%, admixture dispensers at least every six months, and that certification lasts two years from the inspection date.
Who performs a plant certification inspection? The producer hires the inspector, and a licensed engineer employed by the company and approved by the scheme may perform the inspections. Treat the certificate as a documented self-commissioned audit rather than an independent inspection of your project.
Can a plant be within tolerance and still deliver the wrong water-to-cement ratio? Yes, and it is the most common version of the problem. An uncorrected change in aggregate moisture adds water without any scale leaving its tolerance — roughly 8 kg per cubic metre for each percentage point of error on 800 kg of sand.
What single record is most worth requesting? The target-versus-actual batch weight deviations for your own pours, alongside the dated calibration certificates. Those two together answer the question a delivery ticket cannot.
Sources
- Manitoba Infrastructure, Specification 830 — Concrete
- National Ready Mixed Concrete Association, Plant Certification Guidance
- National Ready Mixed Concrete Association, Quality Control Guide
- Alberta Transportation, Supplemental Specification 5.5 — Supply of Portland Cement Concrete
Clause wording above is quoted from the published owner specifications and certification guidance cited. Tolerances in referenced consensus standards are not reproduced here; confirm the governing values against your own project specification.
