Codes referenced
Weighing-equipment accuracy ±2% cement/SCM and ±3% aggregate/admixture/water; tare-out allowance; automatic-control interlocks; operator display of mix-design + batch-size identification.
Batch-weight tolerance of ±1% for cement and ±2% for aggregate, admixture and water against the target recipe quantity on the printed ticket. Plant batching equipment must permit compensation for moisture.
Defines the proportions the controller must hit. Mix-design intent (Cl 9) and acceptance (Cl 15–16) are the upstream and downstream gates the controller sits between.
International parallel on batching sequence and tolerance interpretation; useful reference when reconciling tolerance language between IS and US projects.
Six-stage batching sequence
The controller drives a fixed sequence. Stage order matters: dust, segregation and weighing-accuracy considerations dictate that aggregate goes before cement goes before water goes before admixture.
Coarse first (20 mm, then 10 mm / crushed sand), fine last. Sequence prevents sand from sticking to the mixer drum and reduces inter-bin carry-over. Each fraction weighed separately against its moisture-corrected target.
Cement and fly-ash on separate scales (different densities, different absolute mass per m³). Loading mid-sequence prevents dust escape into the bin conveyor and lets the dust collector stabilise between aggregate and powder stages.
Water meter reads the added water, i.e. design water minus the SM% correction from aggregate moisture. Meter accuracy is the single biggest contributor to batch-to-batch w/c drift.
Diluted into 2–3 L of batch water and dosed into the water line. Last because admixture is a small volume and overdosing is the failure mode — sequencing last lets the controller abort cleanly if the dosing pump fails mid-batch.
Central pan mixer or drum at 12–15 rpm for 60–90 sec (per plant spec). Time-locked, not operator-discretion. Insufficient mixing → strength loss; over-mixing → aggregate breakage and slump loss.
Discharge into the truck mixer drum (pre-wetted). Ticket printed with batch number, mix-design code, actual masses, IST batch start and dispatch time, operator ID, and any logged override.
Two parallel tolerance regimes
Two standards answer two different questions about the same batch, and the controller has to satisfy both.
| Standard | Question it answers | Tolerance — Cement | Tolerance — Aggregate / Admix / Water |
|---|---|---|---|
| IS 4926:2003 | How close must the measuring equipment read to the set-point the controller asked for? | ±2% of quantity being measured | ±3% of quantity being measured |
| ASTM C94/C94M | How close must the batched mass on the ticket be to the target recipe? | ±1% of target mass | ±2% of target mass |
Most modern controllers reconcile to the tighter of the two on the ticket — that is, the operator is told the ticket-print tolerance is ±1% cement / ±2% aggregate, and the equipment accuracy is verified to be inside the IS 4926 ±2/±3% band. If the equipment drifts past the IS 4926 band, the ticket can no longer meet the ASTM C94 band, and the batch is held.
Worked example — M30, 7 m³ truck, controller in semi-auto mode
Operator selects Mix M30-Rev 3, batch size 7 m³, truck MH-12-AB-3456. SM% from this morning's probe check: sand 5.0%, 20 mm 1.0%, 10 mm 2.0%.
| Component | Per m³ (kg) | Target × 7 m³ (kg) | Controller set-point (wet, kg) | IS 4926 accept band | ASTM C94 ticket band |
|---|---|---|---|---|---|
| OPC 53 cement | 380 | 2660 | 2660 | 2607–2713 (±2%) | 2633–2687 (±1%) |
| Fly-ash | 60 | 420 | 420 | 412–428 (±2%) | — (cement-SCM combined per C94) |
| Sand (dry 720, SM 5%) | 720 | 5040 | 5292 | 5133–5451 (±3%) | 4939–5141 (±2%) |
| 20 mm (dry 660, SM 1%) | 660 | 4620 | 4666 | 4526–4806 (±3%) | 4528–4712 (±2%) |
| 10 mm (dry 440, SM 2%) | 440 | 3080 | 3142 | 3048–3236 (±3%) | 3018–3142 (±2%) |
| Added water (after SM deduct) | 198 | 1386 | 1386 − 343 = 1043 | 1012–1074 (±3%) | 1020–1064 (±2%) |
| PCE admixture | 3.3 | 23.1 | 23.1 | 22.4–23.8 (±3%) | 22.6–23.6 (±2%) |
The controller opens gates at the set-points, the load cells feed back weight, the controller cuts off when the actual reading enters the band. If the actual mass falls outside the ASTM C94 ticket band (the tighter of the two), the controller prints a flag and prompts the operator to dump or re-blend the batch. If the actual reading is outside the IS 4926 band, the controller halts production and demands load-cell verification before resuming.
Reference: IS 4926:2003 §10 measuring accuracy; ASTM C94/C94M-24a batch-weight tolerance; IS 456:2000 Cl 9 mix proportioning intent.
Load cell → controller → discharge interlocks
A load cell is a strain-gauge transducer that outputs a millivolt signal proportional to the mass it supports. The controller samples this signal at 50–200 Hz, low-pass filters it, and converts to kg via a calibration factor. Three interlock rules keep the chain honest:
- Zero check before each batch. Empty hopper reading must be within ±0.5 kg of true zero. If it has drifted (build-up, temperature, residual material), the controller applies an automatic tare. IS 4926:2003 explicitly allows a tare adjustment up to 10% of nominal scale capacity, performed at least once per day.
- Cut-off tolerance. The gate closes when the mass reading enters the band, not when the controller opens the gate. With free-fall material above the gate (cement screw, aggregate conveyor), the in-flight material continues to land in the hopper after the gate closes — the controller pre-emptively cuts off early to allow for this "free-fall overshoot", calibrated per material.
- Discharge interlocks. The weigh hopper will not discharge into the mixer until all component set-points have been satisfied within tolerance. If one ingredient fails to reach set-point, the controller will hold the batch and alert — it will not let an under-cemented batch proceed.
Operating modes — manual, semi-auto, fully auto
Operator reads load-cell dial, opens and closes gates by hand. Used for short runs, special mixes, breakdowns. Highest dependency on operator skill — and the most common source of weight-band violations.
Controller opens the gates, reads load cells, cuts off automatically; operator monitors and prints ticket. Tolerances enforced by software; audit trail automatic. The pragmatic middle for 60–90 m³/h plants.
Order ticket (or ERP order) triggers recipe selection, batching, ticket printing, dispatch. Moisture probe auto-feeds the controller. Operator becomes a supervisor, not a button-pusher. Highest throughput, lowest human-error floor — requires reliable probes and disciplined calibration.
Recipe management — the live mix-design library
A controller in production carries tens to hundreds of mix designs at any time: M20, M25, M30, M40, M50, a few SCC mixes, a pavement mix, a few customer-specific variants. Each entry is a record with:
- Recipe code + revision (M30-R3). Revisions are immutable once issued — superseded revisions are archived, not deleted, so the controller can prove what was in production on any past date.
- Per-component target mass (cement, SCM, each aggregate fraction, water, each admixture).
- Accept bands — IS 4926 ±2/±3% internally; ASTM C94 ±1/±2% on the ticket.
- Free-fall overshoot per material — calibrated to the gate type.
- Mixer cycle time and agitator rpm.
- Approval signature — the QA manager's sign-off that this revision is released for production.
When a customer calls for "M30, 100 mm slump, pumped", the dispatcher selects the corresponding code-revision. The operator cannot select a non-released revision. IS 4926:2003 requires the plant operator to be provided with "a clear display of the quantities of materials to be batched for each mix and batch size and with information identifying the display to be selected for each designed and prescribed mix" — i.e. the controller must show what the operator is batching, and the operator must be able to prove which mix was selected.
Audit trail — the per-batch record
IS 4926:2003 §10 calls for a production record. The controller enforces the minimum data set; the ticket is the customer-facing face of that record.
| Field | Source | Why it matters |
|---|---|---|
| Ticket number + IST timestamp | Controller, sequential | Uniqueness; ties batch to cube test |
| Customer, site, PO | Dispatch entry | What was the customer promised? |
| Mix-design code + revision | Operator / dispatch | Which recipe was active? |
| Truck + driver | Truck ID at gate | Delivery chain |
| Batched masses (cement, SCM, agg, water, admix) | Controller print | What actually went in (vs. target) |
| Moisture correction applied | Controller log | Probe value used at batch time |
| Batch start + dispatch time (IST) | Controller clock | Drives transit-time calculations |
| Override log (operator ID, reason, value) | Override screen | The trail when something went off-spec |
A controller that retains the load-cell voltage trace at start, mid-fill and final cut-off — not just the final weight — is what makes a load-cell drift investigation possible after a 28-day cube fails. Without it, you are arguing from the printed ticket; with it, you can plot the cut-off profile and see whether the cut-off fired early, late, or normally.
Common deviations seen on Indian sites
Operator enters the dose into the water-line entry instead of the admixture-line. Pump never runs; concrete arrives with zero plasticiser — site complaint of stiff mix within 15 minutes of batching.
Controller configured to print recipe targets on the ticket instead of measured values. ASTM C94 ticket-tolerance rule is meaningless; load-cell drift is invisible.
Aggregate bin hangs up; controller sees 0 kg. Operator presses "skip ingredient" instead of clearing the hang-up. Cement over-dose → flash set in the drum.
QA releases M30-Rev 4; controller is left on Rev 3 because no one updated the live production library. Two grades of M30 are in the field under the same label.
Field-checklist blocks
A short questionnaire, an inspector's checklist, and three worked examples that show how a controller that looks right on the panel can still be producing concrete nobody can defend.
Questionnaire — ask before the batch is run
- Which recipe revision is live in the controller, and does it match the revision the QA manager released?
- Is the moisture correction in force for every aggregate fraction — sand 5.0%, 20 mm 1.0%, 10 mm 2.0% on the morning's probe check?
- Is the added water set after the SM deduction, so the water meter delivers design water minus the free moisture carried in the aggregates?
- Have all hoppers been emptied and the scales zeroed before batching commences, and is the empty-hopper reading within ±0.5 kg of true zero?
- Is the ticket printing the actual measured masses rather than the recipe targets?
- Is the mixing cycle time-locked at the plant spec — 60–90 s at 12–15 rpm — rather than left to the operator's judgment?
- Is the operating mode the right one for the run — manual only for short runs, special mixes and breakdowns, semi-auto as the default on a 60–90 m³/h plant?
Inspector's checklist — what to verify on the controller and on the ticket
- Zero check on every hopper before each batch, with the tare adjustment performed at least once per day and kept within the 10% of nominal scale capacity IS 4926 allows.
- Actual batched masses printed on the ticket — cement, SCM, each aggregate fraction, water and admixture.
- Both tolerance regimes checked: the measuring equipment inside the IS 4926 ±2% / ±3% band, and the printed ticket inside the ASTM C94 ±1% / ±2% band.
- Moisture correction logged with the probe value used at batch time.
- Batch start and dispatch timestamps in IST on the ticket, because they drive every transit-time calculation downstream.
- Override log checked for operator ID, reason and value, with supervisor-level authorisation and a re-printed ticket carrying the actual quantities.
- Load-cell trace retained at start, mid-fill and final cut-off, not only the final weight.
What happens if… three worked examples
What happens. The morning probe reads sand 5.0%, 20 mm 1.0% and 10 mm 2.0%. The cement target of 380 kg/m³ becomes 2,660 kg for the batch, inside both bands. The sand target of 5,040 kg becomes a wet set-point of 5,292 kg, and the added water of 1,386 kg drops to 1,043 kg after the 343 kg moisture deduction.
Likely outcome. The controller opens the gates, cuts off when the reading enters the band rather than on a timer, and prints a ticket whose actual masses satisfy the tighter ASTM C94 limits of ±1% on cement and ±2% on aggregate, admixture and water.
Correction. None on the set-points. Keep the probe value on the batch record — that is the number that explains the water figure when somebody asks about it at 28 days.
What happens. The controller is configured to print the recipe targets rather than what the load cells read.
Likely outcome. The ASTM C94 ticket-tolerance rule becomes meaningless and load-cell drift stays invisible. A cement batch that was actually heavy still prints 2,660 kg, and when the 28-day cube fails there is nothing on the record to argue from.
Correction. Print the measured values. A ticket nobody can trust turns the audit trail into decoration, and it is the only customer-facing face of the production record.
What happens. The bin hangs up and the controller reads 0 kg. The operator presses “skip ingredient” instead of clearing the hang-up.
Likely outcome. The batch proceeds with cement over-dosed into a mixer that never received its aggregate, and the concrete flash-sets in the drum. The override is logged, but by then the truck is loaded.
Correction. Treat “skip ingredient” as a batch-destroy, not a batch-fix. Clear the hang-up, dump the batch, and record why the bin hung — because it will hang again on the next run.
Frequently Asked Questions
What is the batching tolerance for cement at an RMC plant?
Can the operator override the batch recipe at the controller?
How is a batch logged for traceability?
Related articles
RMC Plant Quality Management — Five Control Zones
Where batching sits inside the broader plant QA chain — materials, batching, water/admix, dispatch, lab.
Mix Design Masterclass — IS 10262 Walkthrough
Where the recipe the controller executes actually comes from.
IS 456 Cube Test Acceptance — Cl 16.1 / Table 11
The downstream gate every batched cube must clear — and how a batch ticket ties back to its cube result.
Amit Haridas
Founder & Proprietor, ConcreteInfo. 25+ years experience in construction QA/QC, concrete technology, and RMC plant operations. NRMCA Certified Trainer (USA) and ISO Lead Auditor. For batch-controller audits, recipe-revision reviews, or ticket-tolerance reconciliations, contact amit@concreteinfo.in.
Technical updates, not sales emails
One email when a new article goes out on concrete technology, QA/QC and construction quality. No promotions, and your address is never sold or passed to advertising networks. Subscribing is separate from making an enquiry.