Understanding PM2.5 and PM10

The coarse fraction (PM10 minus PM2.5) and the fine fraction (PM2.5) come from different sources and behave differently indoors.

Coarse particles (2.5–10 µm) are mechanically generated: resuspended dust, mineral fragments, pollen, fungal spores, brake and tyre wear, large bioaerosols. They settle out within minutes in still air and are largely captured by simple filtration.

Fine particles (≤2.5 µm) are combustion-generated and secondary: vehicle exhaust, gas hob emissions, candle and incense smoke, secondary inorganic aerosol formed from NO₂ and SO₂. They remain suspended for hours, penetrate deep into the lung and follow occupants between rooms via airflow.

The PM2.5/PM10 ratio is itself diagnostic. A ratio near 0.8 indicates a combustion-dominated environment; a ratio near 0.3 indicates a mechanical-dust-dominated environment. The transition tells you which control intervention to apply first.

Outdoor sources that infiltrate indoors: road traffic (the dominant urban source), construction, biomass burning, regional secondary aerosol from agriculture and industry.

Indoor sources: gas and electric cooking (the largest indoor PM event in most homes and many workplaces), candles and incense, printing and 3D-printing, vacuum cleaning without HEPA, occupant movement resuspending settled dust, smoking and vaping.

The indoor/outdoor ratio depends on the building. A leaky pre-1980 building with no filtration runs at 0.7–0.9 of outdoor levels and adds whatever indoor sources are present. A modern airtight building with MVHR and F7 filtration runs at 0.1–0.3 of outdoor levels but is then dominated by whatever happens inside.

Almost every networked PM monitor in the UK uses optical particle counting. A laser illuminates a small sample volume of air; each particle scatters light proportional to its size and refractive index; a photodetector counts the pulses.

The raw output is a count rate per size bin. Mass is then estimated using the relationship m = (π/6) × d³ × ρ × N, where d is the bin midpoint, ρ is an assumed density (typically 1.6 g/cm³) and N is the count in that bin. The PM2.5 mass concentration is the sum across all bins ≤2.5 µm.

This estimation step is the source of most discrepancies between sensors. The assumed density and refractive index are correct for the calibration aerosol — usually polystyrene latex spheres or Arizona Road Dust — but real indoor aerosols are mixtures. Cooking smoke is lower density; mineral dust is higher; semi-volatile particles change with humidity. The result is that two well-engineered sensors can read 30–50 % apart in the same room while both are working correctly.

See particle counters for the higher-end instruments used when count distributions matter more than mass.

Indoor air quality programmes use indicative PM2.5 monitors to spot cooking events, vacuum spikes and outdoor-driven episodes — and to evaluate whether the ventilation and filtration response is adequate.

Construction and demolition sites use boundary PM10 monitors to comply with BS 5228 dust management and Local Authority Section 60/61 consents. PM10 is the controlled parameter for nuisance dust.

Urban and roadside monitoring uses PM2.5 alongside NO₂ to characterise traffic-related pollution. See air pollution monitoring.

Healthcare and cleanrooms use particle counters at HEPA filter test ports to verify filtration integrity under ISO 14644 protocols.

Indicative optical sensors have three known weak spots: humidity (particles absorb water and grow above ~70 % RH), high concentration coincidence loss (the detector cannot resolve overlapping pulses), and the calibration-aerosol problem above.

The standard mitigation is co-location correction: install the indicative sensor next to a reference instrument for a period covering the seasonal range, derive a correction function, and apply it to the indicative data. This is the basis of the MCERTS Indicative classification and the operating model behind most credible city-scale low-cost networks. We cover the methodology in how air quality sensors are calibrated.

• Why do two PM sensors in the same room give different readings?

  Optical sensors estimate mass from particle count and assumed density/refractive index. When the actual aerosol differs from the calibration aerosol — for example cooking smoke versus road dust — the conversion factor changes, and two devices calibrated to different reference aerosols can disagree by 30–50 % while both being internally consistent.

• Is PM2.5 always more harmful than PM10?

  PM2.5 penetrates further into the respiratory tract and carries more of the health-impact evidence, but PM10 includes the PM2.5 fraction and adds the coarse particles relevant to allergens, fungal spores and ultrafine resuspended dust. Both matter; the ratio between them is itself diagnostic.

• Do indoor PM levels follow outdoor levels?

  Strongly, in buildings with high infiltration and no filtration — indoor PM tracks outdoor PM with a lag of 1–6 hours and an indoor/outdoor ratio of 0.6–0.9. In well-filtered MVHR or central-air buildings the ratio falls to 0.1–0.3, but indoor sources (cooking, printing, candles) then dominate.

• Can low-cost optical sensors meet MCERTS?

  No — MCERTS Indicative for PM is currently the highest classification realistically achievable by optical sensors, and only after rigorous co-location correction. Reference-grade compliance still requires gravimetric or beta-attenuation reference samplers.