Industrial dust detectors in 2026 fall into three sensing families: electrostatic/triboelectric probes for in-duct and stack monitoring, laser backscatter units for ambient and workshop measurement, and optical PM counters for cleanroom-grade PM1.0/PM2.5/PM10 readings [S2][S4][S5].
Available measurement ranges stretch across four orders of magnitude, from 0-100 mg/m³ for low-emission duct work to 0-6000 µg/m³ (6 mg/m³) for indoor air quality, with stated accuracies between ±10% and ±15% relative error depending on the principle and the OEM [S3][S4][S6]. Selecting the wrong family is the single most common spec error: a triboelectric bag-house probe has no resolution to alarm on a 0.5 mg/m³ cleanroom drift, and a PM2.5 optical counter cannot survive a cement-stack duty cycle.
Three sensing principles, three duty envelopes
Triboelectric (electrostatic) probes detect particles striking a sensing rod and translate the charge transfer into a 4-20 mA or relay output, making them the default pick for bag-house broken-bag detection in cement, power, steel, and fertilizer plants [S5][S7]. Laser backscatter monitors, exemplified by the AGA series at 24 VDC supply and ≤2.5 W power draw, measure duct or workshop dust at 0-100/500/1000/4000 mg/m³ ranges with ≤10 s response and ≤2% F.S. drift [S4]. Optical PM counters (PM0.3/PM0.5/PM1.0/PM2.5/PM10) using laser scattering at (650±20) nm with successive-wave power density below 5 mW/m² target cleanroom and IAQ use cases [S3][S4].
The principle dictates the install location: triboelectric probes sit inside the duct or stack, laser backscatter heads sit 0.7-5 m away in open environment or 0.7-20 m down a pipe, and PM counters typically pull an isokinetic sample into a chamber [S4][S5]. Mixing these up, for example bolting a PM-counter head into a 200 °C cement flue, is the most common cause of sensor failure within the first quarter.
Range, accuracy, and drift: the spec numbers that matter
Range brackets cluster around four canonical spans: 0-100 mg/m³ for low-emission stacks, 0-500/1000 mg/m³ for general industry, 0-4000 mg/m³ for cyclones and bag houses, and 0-6000 µg/m³ for indoor air quality where the unit is µg/m³ not mg/m³ [S3][S4]. Accuracy specs on currently listed industrial units run from ±10% (high-sensitivity PM LCD display units) to ±15% (general-purpose high-precision dust concentration detectors) [S3][S6].
Drift matters as much as accuracy: laser backscatter units such as the AGA6050 hold ≤2% F.S. drift, a figure that lets a plant extend the calibration interval to 12 months in benign duty and is one of the more credible spec numbers in this category [S4]. For triboelectric probes the dominant error source is coating build-up, not drift, so plants that run high-humidity or sticky particulate should budget for a self-cleaning purge-air option or accept quarterly manual cleaning [S5]. A 24 VDC operating supply with 15 V to 32 V tolerance is the de-facto industrial rail, and any unit that demands a regulated 24 V ±5% will nuisance-fault on long cable runs [S2].
Hazardous-area certification: ATEX/IECEx or walk away
Any dust detector installed in a Zone 20/21/22 area (combustible dust cloud or layer) must carry the matching ATEX 2014/34/EU equipment marking or the IECEx equivalent, and the AGA6050 holds double certification for both gas and dust explosion-proof duties [S4]. Explosion-proof electrostatic pipeline monitors, such as the YF-8500-PM pattern, are explicitly sold for dust-removal pipeline monitoring with alarm output [S7].
Plants that try to save money by installing a non-certified PM optical counter in a grain elevator or flour mill tend to discover the issue at the insurance audit, not at commissioning. The minimum viable marking for a typical coal-handling or bag-house service is dust ignition protection by enclosure (typically 'tb' or 'tc') on a Zone 21 boundary, with the gas-side rating added if any solvent vapour can reach the head. Buyers should also verify the IP rating: IP66 is the current norm for outdoor and washdown locations, and a unit offered at a sharp discount with only IP54 is almost always an indoor-only product mis-positioned for industrial duty [S4].
Output protocols and integration to PLC/DCS
Standard output is a 4-20 mA analog loop scaled to the chosen range, with relay alarm contacts for threshold tripping, and most current-generation units also expose RS-485 Modbus for PLC/DCS/IPC integration [S2][S4]. Higher-end NB-IoT or Wi-Fi variants are aimed at remote dust monitoring rather than dense plant networks and coexist with, rather than replace, the analog loop [S3].
A common buyer mistake is assuming a wireless NB-IoT dust detector will slot into a legacy DCS without a gateway. It will not: the 4-20 mA output remains the only universally accepted signal into a safety-rated PLC, and the digital bus is best treated as a parallel data path for trending rather than a control path. For a working bag-house interlock, hard-wire the relay output to the trip coil and use the analog loop for trending, a layout the AGA-series manuals explicitly support [S4].
Selection map: which detector fits which duty
Bag-house broken-bag detection on cement, power, or steel duct work: triboelectric probe with 4-20 mA and relay, ATEX dust certification, 24 VDC, expect to set the alarm at 10-50 mg/m³ [S5]. Low-emission stack after a bag filter at sub-20 mg/m³: laser backscatter unit at 0-100 mg/m³ range, IP66, ≤2% F.S. drift, 24 VDC ≤2.5 W [S4]. General workshop IAQ where you need PM2.5/PM10 read-out on an LCD: high-sensitivity optical unit at 0-6000 µg/m³ with ±10% accuracy [S3]. High-range cement or metal-machine workshop: 0-1000/4000 mg/m³ laser backscatter with 0.7-20 m pipe range [S4].
Who should NOT buy the cheap PM counter: anyone needing a SIL-classified trip, anyone with stack gas above 50 °C without a sample conditioner, and anyone with combustible dust in the measurement zone. The optical counter family is built for HVAC and cleanroom service, and the sensor heads will not survive abrasive or sticky particulate service even though the spec sheet looks similar. Plants running wood, grain, or any combustible organic dust should also note that a dust mask is the PPE layer, not a substitute for engineering controls, and the detector should be backed by a working gas detector where solvent vapour can co-exist with the dust hazard.
Installation, calibration, and service intervals
Online calibration without disassembly is offered on the AGA-series laser backscatter units, which is the single biggest maintenance-cost lever in this category and worth paying a 10-15% premium for at purchase [S4]. Triboelectric probes typically require a zero/span check every 6-12 months, with the probe head physically cleaned at the same interval in dirty service [S5]. PM optical counters usually carry a 12-24 month recommended service interval, with the laser diode and fan as the dominant wear items.
Spare-parts strategy is a measurable spec question: a detector that uses a generic 24 VDC industrial supply and a standard M12 connector is dramatically cheaper to keep running than a proprietary sealed head that must go back to the OEM. For buyers comparing like-for-like, factor in the cost of two PM dust particle meter calibration checks per year plus one probe-cleaning service event, and divide by the expected 8-10 year field life to get a true cost-per-year rather than a purchase-price comparison. Plant engineering teams running multiple bag houses often consolidate on a single OEM family so that calibration gas, probe rods, and mounting hardware are interchangeable, an approach that also simplifies training.
Sourcing signals worth tracking into late 2026
Two shortlist criteria to apply on every quote: the unit must publish a written drift figure in % F.S. (not just "high accuracy"), and the hazardous-area marking must name the actual zone (Zone 20/21/22) rather than a generic "explosion-proof" label [S4][S7]. Compare linear guide and crossed-roller guide sourcing trends tend to track the same supply-chain tightness in 2026, so lead time on imported detectors has been a useful leading indicator for the broader industrial-instrument market.
Trackable signals into the next two quarters: the OEM's stated calibration interval versus the warranty terms, and whether the manufacturer's published 0-4000 mg/m³ range comes from a single span or multiple user-selectable spans (the multi-span design avoids the "out of range" lockout that kills probes in variable-duty bag-house service) [S4].
Background reading: Cobalt raw material sourcing: a 2026 spec-first buyer's field guide.