Air Quality Monitor

Scope and Equipment Families

Air quality monitoring sits under Test & Measurement, in the Environmental & Meteorological group. The defining feature of the category is breadth: a single phrase, "air quality monitor," covers instruments whose accuracy class and price differ by two to three orders of magnitude. Before any spec sheet is opened, a buyer should first locate the need in one of the equipment families below, because accuracy class and price scale with the family, not with the brand. Specifying across families — for example, expecting a consumer node to deliver compliance data — is the single most expensive mistake in this category.

There are six families a buyer will encounter. The first five are ambient air quality monitors proper; the sixth is an adjacent family listed only to disambiguate it.

1. Regulatory / reference-grade ambient gas analyzers. Single-pollutant rack-mount analyzers (one per gas) deployed in fixed ambient air quality monitoring stations (AQMS). Each is a US-EPA-designated Federal Reference Method (FRM) or Federal Equivalent Method (FEM) and forms the legal basis for NAAQS or national-standard compliance. Examples include NO/NO2/NOx by chemiluminescence, SO2 by UV fluorescence, O3 by UV photometry, and CO by NDIR (gas-filter correlation).
2. Reference / equivalent particulate monitors. Continuous PM2.5 / PM10 / PM10-2.5 monitors using beta-attenuation (BAM) or oscillating-microbalance (TEOM) principles, plus the integrated gravimetric filter sampler that is the PM2.5 Federal Reference Method (a 24-hour filter, weighed in a controlled lab).
3. Indoor air quality (IAQ) monitors. Wall-mount or desktop units for buildings, reporting CO2, PM2.5, TVOC, temperature, and relative humidity, and sometimes CO, HCHO (formaldehyde), and radon. Used for HVAC demand-controlled ventilation and green-building certification.
4. Low-cost sensor nodes / sensor networks. Compact electrochemical plus optical-particle-counter (OPC) devices for dense supplementary networks, hyperlocal mapping, and citizen science (for example PurpleAir, Clarity, AirVisual). Indicative, not regulatory.
5. Portable / handheld monitors and survey instruments. Battery-powered units for spot checks, IAQ audits, leak surveys, and occupational hygiene, often pairing a PID for VOCs with PM and CO2 channels.
6. Continuous Emissions Monitoring Systems (CEMS) — adjacent family. Stack or source monitors at the emission point (flue gas), distinct from ambient monitors. They share gas-analyzer physics (NDIR, chemiluminescence, FTIR, UV) but operate at far higher concentrations and on extracted or in-situ hot wet gas. CEMS is listed here only to disambiguate it; it is governed by separate emissions regulations and is not interchangeable with ambient instruments.

Photo: Artaxerxes, CC BY 4.0, via Wikimedia Commons

The practical consequence of this structure is that two products both correctly called "air quality monitors" may have no overlap in physics, accuracy, certification, or price. A reference NOx analyzer detecting sub-ppb concentrations and a consumer desktop unit estimating PM2.5 by light scattering are both in scope, but they answer different questions for different buyers. The reference analyzer forms the legal basis for compliance against a national ambient standard; the consumer unit provides hyperlocal awareness with no legal standing. Neither is "better" in the abstract — each is correct only for its own family of applications, and a buyer who selects across families pays for it later in rejected data or wasted capex.

A second consequence is that the regulatory families (the first two) are single-pollutant by design: a fully instrumented ambient station carries one rack-mount analyzer per gas — chemiluminescence for NOx, UV fluorescence for SO2, UV photometry for O3, NDIR for CO — plus a continuous particulate monitor and, often, a gravimetric filter sampler alongside it. The IAQ, sensor-node, and portable families, by contrast, integrate several pollutants into one compact device, trading the single-pollutant rigor of the reference station for breadth, size, and cost. The chapters that follow first enumerate what is measured, then the principles that set accuracy and cost, then the standards that decide which family a given application legally requires.

Pollutants Measured

Across all six families, the parameters an air quality monitor reports fall into a small number of groups. Knowing which parameters an application actually requires — and in which units — is the prerequisite for matching an instrument to the task. The table below summarizes the typical parameter set and the units in which each is properly reported.

Particulate matter is reported as mass concentration — PM1, PM2.5, and PM10 in micrograms per cubic metre (µg/m³). The number after "PM" is an aerodynamic size cut (2.5 µm, 10 µm), and the measured quantity is always mass per volume, never a volume mixing ratio. This unit discipline matters: particulate is the parameter most often misreported, and treating it like a gas concentration is a category error.

Criteria gases — CO2, CO, O3 (ozone), NO/NO2/NOx, and SO2 — are the gases for which most national ambient standards set limits. They are reported as volume mixing ratios (ppm or ppb) or, where a standard demands it, as mass concentration (µg/m³ or mg/m³). NOx is the sum of NO and NO2; reference instruments measure NO directly and derive NO2 by conversion, so the relationship NOx = NO + converted NO2 is built into the measurement, not assumed.

VOCs are reported either as TVOC (total volatile organic compounds) or, in many IAQ units, with formaldehyde (HCHO/CH2O) called out as a separate channel because of its specific health relevance. TVOC may appear in ppb (when measured by a PID) or as an indicative index (when measured by a metal-oxide sensor); the two are not equivalent, and treating a MOS index number as a calibrated absolute concentration is a frequent mistake. Speciating individual VOCs, rather than reporting a single total, is the domain of a laboratory gas chromatograph rather than a field monitor.

Comfort and context parameters — temperature, relative humidity, and barometric pressure — are reported on almost every monitor. They serve two purposes: they describe human comfort, and they feed the sensor-compensation algorithms (optical PM in particular needs RH correction, and gas-law conversions between ppb and µg/m³ need temperature and pressure), the same parameters logged by a standalone temperature and humidity recorder. Specialty parameters such as radon, NH3, H2S, methane and other combustibles, and ultrafine particle count (measured by a condensation particle counter, CPC) appear in research and industrial-hygiene contexts and are usually outside the scope of a general-purpose ambient or IAQ monitor.

Operating Principles by Technology

Accuracy and cost are set by the sensing principle, and the principle-to-pollutant mapping is the core of correct selection. Reference-grade instruments use one class of physics; low-cost and portable instruments use another. The table below summarizes the dominant principle for each pollutant at each accuracy tier; the text that follows explains how each works.

Gas analyzers (reference grade). Chemiluminescence measures NO and NOx: NO reacts with ozone to form excited NO2*, which emits light (broadband, roughly 600–3000 nm, peaking near 1200 nm) measured by a photomultiplier tube. NO2 is measured by first converting it to NO with a molybdenum or photolytic converter, so NOx = NO + converted NO2. UV fluorescence measures SO2: SO2 molecules absorb pulsed UV and re-emit fluorescence (excitation and emission in the roughly 190–420 nm region), and the fluorescence intensity is proportional to concentration. UV photometric absorption measures ozone, which strongly absorbs UV at 254 nm; concentration is derived from the Beer-Lambert attenuation between a sample cell and an ozone-scrubbed reference. NDIR (non-dispersive infrared) measures CO and CO2: the target gas absorbs a specific IR band, and concentration is computed from the difference between transmitted and reference IR intensity. Reference CO analyzers use gas-filter-correlation NDIR for selectivity.

Particulate monitors. Beta attenuation (BAM) collects particles on a filter tape that attenuates a beta-particle (electron) beam; mass loading is computed from the attenuation. The Met One BAM-1020 is a US-EPA FEM for PM10, PM2.5, and PM10-2.5 and is one of the most widely deployed continuous PM monitors worldwide. TEOM (tapered element oscillating microbalance) derives mass from the frequency shift of a vibrating tapered element as collected mass changes its resonant frequency. Gravimetric (filter-based FRM) draws air through a size-selective inlet onto a filter over 24 hours; the filter is conditioned and weighed on a microbalance. This is the legal PM2.5 reference (40 CFR Part 50, Appendix L) against which continuous methods are equivalenced. Optical / light-scattering methods — nephelometry and the optical particle counter (OPC) — measure how much light particles scatter; the scattered intensity (nephelometer) or per-particle pulse count and sizing (OPC) yields a real-time estimate, the same approach used in a dedicated dust particle meter for cleanroom counting. This is the basis of nearly all low-cost and consumer PM2.5 sensors: fast and cheap, but requiring a density or refractive-index assumption and a humidity correction.

Gas sensors (low-cost / portable). NDIR miniaturized is the standard for accurate CO2 in IAQ monitors — the same physics as reference NDIR. Electrochemical (EC) sensors measure CO, NO2, O3, SO2, and H2S: the target gas diffuses to a working electrode and is oxidized or reduced, producing a current proportional to concentration. EC sensors are compact and low power but subject to cross-sensitivity and baseline drift, and are the same sensing class used in a handheld gas detector for occupational safety. Metal-oxide semiconductor (MOS / MOX) sensors measure TVOC and reducing gases: a heated metal-oxide film changes electrical resistance in the presence of target gases. They are inexpensive and broad-spectrum but non-selective and drift-prone, reporting a relative or indicative TVOC. Photoionization detector (PID) measures TVOC at ppb: a UV lamp ionizes molecules whose ionization potential is below the lamp photon energy (common lamps are 10.0, 10.6, and 11.7 eV), and the resulting ion current is measured. A PID offers high sensitivity (ppb-class resolution, roughly 1–5 ppb in clean conditions), fast response, and broad VOC coverage, but it is non-specific and cross-sensitive to humidity and methane.

Key spec parameters and units. The indicative ranges below frame what to expect on a spec sheet. PM2.5 / PM10 measuring range is typically 0–500 or 0–1000 µg/m³ (consumer to ambient). CO2 range is 400–2,000 ppm for IAQ and roughly 400–5,000 ppm extended. CO range is 0–50 or 0–500 ppm depending on class. Reference O3 analyzers offer low-ppb resolution with full scale to about 0.5–1 ppm. Reference NOx and SO2 analyzers reach sub-ppb detection with full scale to hundreds of ppb up into the ppm range. TVOC is ppb-class on a PID and an indicative index on a MOS sensor. The lower detection limit (LDL) of a reference gas analyzer (chemiluminescence, UV-fluorescence, UV-photometric) is sub-ppb, on the order of 0.5 ppb. Accuracy is a few percent of reading for reference grade; a typical CO2 NDIR IAQ accuracy is ±(50 ppm + 3% of reading); low-cost PM is indicative only. Response time (t90) is seconds to tens of seconds for reference analyzers and near real-time for optical PM. Data interval and averaging follow the governing standard (1-h, 8-h, 24-h, annual). Outputs commonly include 4–20 mA, RS-485/Modbus, MQTT, Ethernet, and cellular. Calibration of reference instruments requires traceable span gases and periodic multipoint calibration; sensors drift and need recalibration.

Units discipline matters. Gases are reported in ppm or ppb (volume mixing ratio) or in µg/m³ or mg/m³ (mass concentration); the two are convertible only with known temperature and pressure (and molar mass). Particulate is always mass per volume (µg/m³). Mixing these up is the most common spec error in the category.

Governing Standards and Frameworks

Which equipment family an application legally requires is decided by the governing standard, not by preference. The frameworks below are grouped by their purpose: health-based guidelines, US regulatory methods, indoor and building ventilation, and the China standards most relevant to this platform's procurement audience.

Health-based guidelines. The WHO Global Air Quality Guidelines (2021) set PM2.5 at 5 µg/m³ annual mean and 15 µg/m³ 24-hour mean, tightened from the 2005 values of 10 and 25 µg/m³. The 2021 update also lowered the NO2 annual guideline to 10 µg/m³ (from 40). The 2021 AQG covers PM, O3, NO2, SO2, and CO. These are health-based reference values, not legally binding limits in themselves.

United States — regulatory. The NAAQS, set under the Clean Air Act, define the criteria-pollutant limits. Compliance monitoring must use FRM/FEM methods designated under 40 CFR Part 53 and operated per 40 CFR Part 58. 40 CFR Part 50, Appendix L defines the PM2.5 Federal Reference Method (gravimetric, 24-hour); continuous monitors such as BAM and TEOM qualify as Federal Equivalent Methods (FEM). The practical takeaway: where NAAQS compliance is the goal, the instrument family is fixed by these regulations, and low-cost optical sensors do not qualify.

Indoor and building ventilation. ASHRAE Standard 62.1 governs ventilation for acceptable indoor air quality. Note that 62.1 has not contained an indoor CO2 limit for roughly three decades; the ~1,000 ppm figure is a common rule-of-thumb, not a 62.1 cap, and the 2025 revision emphasizes ventilation that adjusts dynamically to occupancy and pollutant levels. Green-building and IAQ certifications carry the explicit thresholds: LEED awards enhanced IAQ points at CO2 below 800 ppm and PM2.5 below 12 µg/m³, whereas WELL v2 uses its own values of CO2 around 900 ppm (basic) / 750 ppm (enhanced) and PM2.5 15 µg/m³ (basic) / 10 µg/m³ (enhanced), while RESET Air mandates continuous measurement of five parameters (PM2.5, TVOC, CO2, temperature, RH), with thresholds such as PM2.5 below 35 µg/m³ (acceptable) or below 12 µg/m³ (high performance) and CO2 below 1,000 ppm (acceptable) or below 600 ppm (high performance).

China. GB 3095-2012 (Ambient Air Quality Standards) sets Grade I PM2.5 at 35 µg/m³ (24-h) and 15 µg/m³ (annual), and Grade II PM2.5 at 75 µg/m³ (24-h) and 35 µg/m³ (annual). Grade I applies to special and protected areas, Grade II to general areas. Air quality is termed "excellent" below 35 µg/m³ (24-h), "good" or favorable from 35 to 75, and "polluted" above 75 µg/m³. GB/T 18883 (Indoor Air Quality Standard) is the indoor counterpart to the ambient GB 3095. The table below consolidates the key numeric thresholds across these frameworks.

Cross-Cutting Selection Criteria

Selection in this category is a sequence, and most mistakes are premature decisions made before the accuracy class is fixed. Apply the criteria below in order; the first two decide the equipment family and the sensing principle, and the rest refine the choice within that family. These eight steps can serve as a fixed RFQ template.

1. Required accuracy class first. Compliance or legal use demands FRM/FEM-designated instruments; building optimization can use good IAQ-grade NDIR plus EC; mapping and awareness can use low-cost sensor nodes. Do not specify a low-cost optical PM sensor where a BAM is legally required.
2. Match principle to pollutant. CO2 to NDIR (never MOS for accurate CO2); reference NOx to chemiluminescence; reference SO2 to UV fluorescence; reference O3 to UV photometry; TVOC at ppb to PID; broad indicative VOC to MOS; reference PM to BAM, TEOM, or gravimetric; real-time indicative PM to optical.
3. Detection limit, range, and resolution appropriate to the expected concentrations. Ambient ppb work and occupational ppm work are different instruments.
4. Cross-sensitivity and interferences. Electrochemical and PID sensors respond to non-target gases; PIDs are strongly affected by humidity and respond to methane. Verify interference specs against the actual gas matrix.
5. Calibration and drift / total cost of ownership. Reference analyzers need traceable span and zero gases and periodic multipoint calibration; low-cost sensors drift and need recalibration or collocation correction. Budget for consumables (filter tape, span gas, sensor replacement) and field maintenance, not just capex.
6. Environmental compensation. Confirm built-in temperature and RH compensation, especially for optical PM (hygroscopic growth at high humidity inflates readings) and for EC and PID sensors.
7. Averaging and data handling aligned to the governing standard (1-h, 8-h, 24-h, annual) and the connectivity required: Modbus or 4–20 mA for a BMS; cellular or MQTT for distributed networks; an integrated data logger and QA flags for regulatory work.
8. Siting and ingress protection for outdoor stations: sampling inlet height, shelter or enclosure, IP rating, and power and comms. A fixed ambient station is usually co-located with meteorological instruments such as an anemometer on a mast, or a full weather station, so that pollutant data can be interpreted against wind and atmospheric conditions.

The recurring theme across these eight steps is that accuracy class and standard come first and everything else follows. An instrument selected on price before the accuracy class is fixed will, more often than not, turn out to be the wrong family entirely — indicative data where regulatory data was required, or an over-specified reference analyzer where an indicative node would have sufficed. The cost of that error is rarely the instrument price; it is the rejected dataset, the failed audit, or the rebuilt network.

Manufacturers, Sourcing and Pitfalls

The supplier landscape mirrors the equipment families: reference-grade analyzer makers, particulate and IAQ instrument makers, building and HVAC integrators, low-cost network and consumer brands, and the OEM sensor-component suppliers whose parts sit inside many finished monitors. Knowing where a vendor plays helps a buyer avoid mismatching family and need.

• Reference ambient gas and particulate analyzers: Thermo Fisher Scientific (iQ-series gas analyzers; Partisol, TEOM, and beta-attenuation particulate), Teledyne API (T-series analyzers), ENVEA (formerly Environnement S.A.), HORIBA, Ecotech (Acoem), and Met One Instruments (BAM-1020).
• Particulate / aerosol and IAQ instruments: TSI Incorporated, Met One, Thermo Fisher, Testo, and Aeroqual.
• Indoor / building IAQ and HVAC integration: Honeywell, Siemens, Emerson, 3M, Aeroqual, Testo, Kaiterra, and Airthings.
• Low-cost sensor networks / consumer: Aeroqual, Clarity, PurpleAir, IQAir (AirVisual), Oizom, and Airveda.
• Sensor component suppliers (OEM): Sensirion, Alphasense, Ion Science (PID), Figaro, Winsen, and Renesas.

When the family and the principle are settled, sourcing comes down to verifying that the specific instrument is designated for the required standard (FRM/FEM listing for regulatory work, RESET Air certification for that program), that consumables and calibration gases are locally available, and that the data interface fits the destination system. The same diligence applies whether the buyer is procuring a single reference NOx analyzer or a hundred-node low-cost network.

Common pitfalls. The mistakes below recur often enough that they are worth treating as a pre-purchase checklist of what not to do:

• Treating a light-scattering consumer PM2.5 reading as regulatory data — it is indicative and humidity-sensitive.
• Using MOS TVOC index numbers as if they were calibrated absolute concentrations.
• Comparing CO2 readings to a non-existent ASHRAE 62.1 "limit"; 1,000 ppm is guidance or rule-of-thumb, not a code cap.
• Ignoring the averaging period when comparing a reading to a standard — a 1-minute spike is not a 24-hour mean.
• Confusing ambient air quality monitors with CEMS — different concentrations, regulations, and instruments.

One last commonly overlooked dimension is data lifecycle. For regulatory and certification work the dataset, not the instrument, is the deliverable: QA flags, traceable calibration records, correct averaging, and a defensible chain from inlet to reported value are what survive an audit. Choosing the right family is necessary but not sufficient; operating it under the discipline the standard expects is what makes the data usable.