Codes referenced
Aggregates measured by mass, "allowance being made for the free moisture content of the aggregates".
Field method for surface moisture by displacement; water absorption for SSD correction.
Reference oven-dry method for aggregate moisture — the standard against which probes are calibrated.
Aggregate moisture checks minimum twice per shift per aggregate type.
Why moisture probe accuracy matters
The moisture probe in an aggregate bin has one job: tell the batch controller how much water is sitting in the sand (and any wet coarse aggregate), so the controller can subtract that water from the design mix water. A 1 m³ M30 batch has 800 kg of sand carrying 6% surface moisture = 48 kg of free water sitting in the sand. If the probe reports only 5%, the controller subtracts only 40 kg — 8 kg of extra water goes into the mix per cubic metre. Multiplied across a 1000 m³/month plant, that's 8,000 kg of unaccounted water per month, with a corresponding drop in strength.
The strength relationship: water content vs strength at constant cement content is roughly linear in the practical range. Each 1% increase in effective w/c typically costs about 5–7% in 28-day compressive strength. 1% moisture over-correction = ~6% strength loss.
Conversely, under-correcting (probe reads high) reduces water in the mix, which raises strength but drops slump. Site reports "concrete is stiff today"; driver adds water at site without authorisation; strength drops; the cycle continues. Both directions of probe error have a real cost.
Probe types — strengths & limits
Four probe technologies are in commercial use at RMC plants. None is intrinsically "best"; each has failure modes the operator must understand.
| Technology | Measurement principle | Strengths | Limits |
|---|---|---|---|
| Capacitance (dielectric) | Change in capacitance between two electrodes in the aggregate mass | Low cost, robust, simple installation | Sensitive to material density (compaction variation), temperature, and probe contamination |
| Microwave (resonant cavity) | Shift in resonant frequency of a microwave cavity with water content | Less sensitive to density variation; better linearity | Higher cost; affected by metal reinforcement near the probe |
| NIR (near-infrared) | Absorption of NIR light by water molecules on aggregate surface | Surface-specific, fast response, very accurate | Lens contamination by dust; needs clear optical window; surface-only reading |
| Resistance / conductivity | Electrical resistance between two electrodes in the aggregate mass | Cheapest, simplest electronics | Heavily affected by dissolved ions in the water; salt contamination changes reading significantly |
Most production RMC plants in India use capacitance probes — low cost, well-understood. Microwave is increasingly used in new installations. NIR and resistance are rare at production scale.
Pan-dry is the only true reference
ASTM C566-19 is the international reference for aggregate moisture content. The procedure is direct: weigh a representative aggregate sample, dry in a controlled oven or on a hot plate to constant mass, reweigh. The moisture content is the mass loss divided by the dry mass.
- Sample size — at least 1 kg for coarse aggregate; 500 g for fine aggregate
- Drying temperature — 110 ± 5°C until two successive weighings at 1-hour interval differ by < 0.1%
- Calculation — Moisture % = (Wet mass − Dry mass) / Dry mass × 100
In production use, "pan-dry" is often a hot plate or a microwave oven (3–5 min cycles with stirring) rather than a laboratory oven — fast enough to give a result in 15 minutes. For the daily probe-verification check at an RMC plant, this is acceptable when the hot plate is itself calibrated against a lab oven on a quarterly basis.
The 1% deviation rule
A practical operating limit for RMC plants: the probe reading should agree with the pan-dry result within ±1.0% of moisture content.
Below ±1.0%, the resulting water dose error is small enough to be absorbed by the IS 4926 Cl 9.1.2 batching tolerance (±3% for water = ±5.1 kg per 170 kg batch) — the error does not propagate into a strength failure. Above ±1.0%, the moisture error becomes the dominant uncontrolled variable in the batch and the probe must be recalibrated before the next production batch.
A real-world example of what the 1% rule catches: a probe that reads 6.5% when the pan-dry is 5.0% looks like a 1.5% error in probe terms, but the batch controller will leave 12 kg/m³ extra water in the mix — a 7% w/c increase, a strength drop in the 6–8% range. The 1% rule is the floor below which probe errors become manageable; above it, the error becomes project-killing.
Calibration frequency
Two cadences: the routine schedule, and the event-driven triggers.
| Trigger | When | What to do |
|---|---|---|
| Daily routine | Start of every shift, before first production batch | Pan-dry check vs probe reading on each aggregate bin. Deviation > 1% → adjust probe offset |
| Material source change | New aggregate supplier or new pit/ quarry source | Re-baseline probe — mineralogy affects dielectric response |
| Bin topped up | Partially empty bin filled with new material | Pan-dry check on the freshly added layer; verify probe tracks new material |
| Weather event | Heavy rain, prolonged dry heat, fog | Repeat pan-dry check; probe offset may need re-calibration |
| Mechanical event | Bin struck, probe hit by loader, cable damage | Full re-calibration; verify probe mounting & cable integrity |
| Probe replacement | New probe installed | Multi-point calibration against pan-dry samples at 3%, 6%, 9% moisture levels |
Pan-dry procedure at the plant
Collect a representative sample
From the bin, not the loader bucket
- · Use a sampling lance or long-handled scoop. Take from at least three locations in the bin — top, middle, bottom.
- · Sample mass — at least 1 kg for fine aggregate, 2 kg for coarse. Larger samples give more repeatable results.
- · Record the probe reading at the moment of sampling for direct comparison.
Weigh the wet sample
Balance reading to 0.1 g
- · Tare a clean metal pan on a 0.1 g balance.
- · Spread the wet sample in the pan, break up any clumps, weigh and record as Mwet.
Dry to constant mass
Hot plate or microwave for plant use
- · Hot plate: heat the pan with stirring at 3-minute intervals. Each cycle, weigh and stir. Stop when two consecutive weighings 1 minute apart differ by < 0.5 g.
- · Microwave: 3-minute cycles at medium power, stirring between cycles. Same end-point criterion.
- · Laboratory oven (for the quarterly calibration of the hot plate): 110 ± 5°C for 24 hours minimum.
Calculate moisture content
ASTM C566 formula
- · Weigh the dried sample as Mdry.
- · Moisture % = (Mwet − Mdry) / Mdry × 100
- · Compare to probe reading at sampling moment. Apply correction factor if deviation > 1%.
Common probe drift causes
Cement paste, oil, or admixture residue builds up on the probe face and changes its dielectric response. Clean with a damp cloth at every shift-start check. Replace the probe if the surface coating is permanently contaminated.
Aggregate flow and bin vibration loosen the probe mount over time, creating a varying air gap between the probe face and the aggregate. Tighten mount bolts monthly; verify cable gland integrity.
Wet fine aggregate can arch above the probe. The probe then reads the humidity of the air pocket below the arch — not the moisture of the material that actually discharges into the batch. Routine bin vibration or air cannons help.
UV exposure, admixture fume attack, rodent damage, and connector corrosion all shift the probe's electrical characteristics. Inspect cables quarterly; replace at first sign of cracking or insulation damage.
A new quarry has different mineralogy — different dielectric constant for the same moisture content. The probe must be re-baselined, not just re-zeroed, when source changes.
Capacitance probes in particular drift with temperature. A cold winter morning reading differs from a hot afternoon reading for the same moisture content. Some probes have built-in temperature compensation; if not, recalibrate mid-shift in extreme weather.
Cross-check with batch water balance (yield variance)
Pan-dry is the direct check; yield reconciliation is the indirect check. The two should agree.
For each batch the controller knows: cement mass (from load cell), aggregate mass (from load cell), water dose (from flow meter), admix dose (from flow meter). It also knows the design density. The actual batched density can be back-calculated from the truck's net load. If the actual density is consistently higher than the design density, the batch is over-dosing aggregate (or under-dosing water) — moisture correction is suspect. If lower, the opposite.
A yield variance consistently outside ±2% across a week, with no other obvious cause (load cell drift, water meter drift), is a strong signal that the moisture probe is misreporting in one direction. This is the most useful production-floor check for a probe that drifts between daily pan-dry calibrations.
Common deviations in Indian plants
Probe installed at commissioning, never re-verified. Drift over 6–12 months takes it far from its original offset. Daily batches are silently off by 1–2% moisture.
Exactly the days when moisture is changing fastest. The probe is least accurate when most needed.
Hot plate itself drifts. Quarterly: verify hot-plate readings against a calibrated laboratory oven on the same sample.
Quarry changes material; probe offset is now wrong; daily pan-dry still shows < 1% deviation because the new material has a different but consistent dielectric response. Catch: verify probe against two different moisture levels (3% and 9%) on the new material.
Field-checklist blocks
A short questionnaire, an inspector's checklist, and three worked examples that show how the probe-versus-pan-dry comparison resolves into a decision: batch it, recalibrate the offset, or go looking for the probe.
Questionnaire — asked at shift start, before the first production batch
- What did the probe read and what did the pan-dry read, at the same instant, for each bin? Two numbers taken together, or you have a reading and nothing to check it against.
- Is the deviation within ±1.0% of moisture content? That is the operating limit. Above it the moisture error becomes the dominant uncontrolled variable in the batch.
- Was the sample taken from at least three locations in the bin, top, middle and bottom, with a lance or long-handled scoop? A sample from the loader bucket tells you about the loader bucket.
- Was the sample at least 1 kg for fine aggregate and 2 kg for coarse? Larger samples give more repeatable results.
- Has the aggregate source changed, or has the bin been topped up with new material, since the last calibration? New quarry means re-baselining, not re-zeroing. A topped-up bin means a check on the fresh layer.
- Is the hot plate itself verified against a calibrated laboratory oven? Quarterly. A drifting hot plate is a probe you cannot trust in a different way.
- Is the probe face clean and the mount tight? Contamination and a loosened mount are the two drift causes you can clear in five minutes.
- Has yield variance been sitting outside ±2% across a week with no other obvious cause? That is the indirect check, and it catches a probe that drifts between daily calibrations.
Inspector's checklist — what to verify on site
- Pan-dry check on each aggregate bin at shift start, before the first production batch, with the probe reading recorded at the moment of sampling so the two are comparable.
- Sample drawn from at least three locations in the bin with a sampling lance or long-handled scoop, at least 1 kg for fine aggregate and 2 kg for coarse.
- Weighed on a 0.1 g balance on a tared pan, then dried to constant mass: hot plate at 3-minute cycles with stirring until two consecutive weighings 1 minute apart differ by less than 0.5 g, or microwave at 3-minute cycles at medium power on the same end-point criterion. The laboratory oven route is 110 ± 5 °C for 24 hours minimum, and that is the quarterly hot-plate calibration.
- Moisture % = (Mwet − Mdry) / Mdry × 100, computed and written beside the probe reading. A probe reading that sits 0.3–0.7% above the pan-dry result is normal and expected; an offset that size is not a fault.
- Deviation above ±1.0% handled before the next production batch: offset adjusted, probe re-checked, and the re-check recorded. The pan-dry value, not the probe value, is what goes into the correction.
- Probe face cleaned with a damp cloth at every shift-start check, mount bolts tightened monthly, cables inspected quarterly for cracking, UV damage and connector corrosion.
- Two-level check after a source change: verify the probe at 3% and 9% moisture on the new material, and re-baseline rather than re-zero if the new quarry has a different dielectric response.
- Full re-calibration and mount-and-cable inspection after a mechanical event, a strike from a falling aggregate lump, or a probe replacement.
What happens if… three worked examples
What happens. That is a 1.5 percentage-point error, and it does not stay small in the batch: the controller leaves 12 kg/m³ of extra water in the mix, a 7% increase in w/c, and the 28-day strength falls by 6–8%. The plant keeps producing and every batch looks the same, which is exactly why nobody notices.
Likely outcome. The probe is the suspect and the batcher is not. Clean the probe face, adjust the offset and re-run the pan-dry comparison before the next production batch, because above ±1.0% this is no longer an error the batching tolerance can absorb.
What happens. The shift-start comparison comes back inside ±1.0%. That leaves a water dose error small enough to sit inside the IS 4926:2003 Cl 9.1.2 water batching tolerance of ±3%, which is ±5.1 kg on a 170 kg batch (170 × 0.03 = 5.1). The error does not propagate into a strength failure, and the batch goes out.
Likely outcome. Nothing to fix on the probe, but both numbers still get logged. The value that drives the correction is the pan-dry value, and the offset between the two is what you re-apply tomorrow. Skip the logging and you have no way of telling a slow drift from a stable one.
What happens. Daily comparisons sit inside ±1.0%, yet the actual batched density keeps coming out high against the design density, and the yield variance stays outside ±2% across a week with no load-cell drift and no water-meter drift. The moisture correction is the suspect. A probe reading that is wrong in one direction will do this while still looking correct against a single reference point.
Likely outcome. Do not touch the offset on a single-point check. Run the two-level check at 3% and 9% on the current material, clean the probe face and re-tighten the mount, then re-baseline if the source has changed. If the two-level check comes back inside ±1.0% at both points while the yield variance persists, the probe is not the cause and the load cells or the water meter are where you look next.
Frequently Asked Questions
How often must moisture probes be calibrated?
What is the acceptable deviation between probe and pan-dry reading?
What causes moisture probe drift?
Related articles
Amit Haridas
Founder & Proprietor, ConcreteInfo. 25+ years experience in construction QA/QC, concrete technology, RMC plant operations, consulting and technical training across India. For moisture-probe calibration programme design, pan-dry reference procedures, or aggregate moisture troubleshooting, 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.