Most air quality reports fail on a technicality, not on bad data. A mine can measure every parameter perfectly and still be non-compliant because the report answered the wrong question. This is what an ambient air quality monitoring report has to contain to survive scrutiny.
The limits you are reporting against
India works to the National Ambient Air Quality Standards notified by CPCB on 18 November 2009. The values below apply to industrial, residential, rural and other areas alike — the separate relaxed industrial standard was deliberately dropped in the 2009 revision.
- PM10: 100 µg/m³ over 24 hours, 60 µg/m³ annual
- PM2.5: 60 µg/m³ over 24 hours, 40 µg/m³ annual
- SO2: 80 µg/m³ over 24 hours, 50 µg/m³ annual
- NO2: 80 µg/m³ over 24 hours, 40 µg/m³ annual
- NH3: 400 µg/m³ over 24 hours, 100 µg/m³ annual
For context, the WHO 2021 guidelines are far tighter: PM2.5 at 15 µg/m³ over 24 hours and 5 µg/m³ annual. A site sitting comfortably inside NAAQS can still sit outside WHO guidance, and community pressure almost always arrives on the WHO number rather than the legal one. State both in the report.
The 98% rule — the clause most sites miss
The notification does not require every measured value to comply. Values are compiled for at least 98% of the year; the remaining 2% may exceed, but not on two consecutive monitoring days. Annual figures are the arithmetic mean of a minimum of 104 measurements a year, taken twice a week at uniform intervals.
This clause changes how you plan a monitoring calendar. Two consecutive breach days are a reportable event even when the annual average looks healthy, so a station that loses data during a bad dust week can fail on completeness rather than on concentration.
What the report has to contain
- Station inventory with coordinates, height above ground and the classification of each location
- Calibration records for every continuous analyzer, with zero and span checks logged
- Completeness table per parameter per station, stating the 98% figure explicitly rather than leaving it to be inferred
- Data capture justification for every gap, with reason and duration
- Exceedance analysis naming each breach against the specific limit it breached
- Source apportionment tying each exceedance to an operational cause: haul road, drilling, crusher, blasting or handling
- Corrective action with a named owner and a date
Source apportionment decides the outcome
A report stating PM10 was 118 µg/m³ and nothing more is a compliance record. A report stating PM10 was 118 µg/m³ on the haul alignment between crusher and village, 40 m from the nearest receptor, during the 14:00 haul cycle, with suppression scheduled for week 42, is evidence that you understand the problem. Auditors read for the second kind.
Failure modes we see repeatedly
- Publishing a 24-hour average computed from fewer than 24 valid hourly values
- Mixing gravimetric and beta-attenuation results without flagging the change of method
- Publishing an AQI without naming the pollutant that drove it — AQI is the maximum sub-index across eight parameters, so one of them defines the number
- Omitting consecutive-day exceedance sequences entirely
- Missing the SPCB pre-calibration notification, which can invalidate an otherwise sound dataset
CPCB has signalled that the 2009 standards are due for revision. Confirm the current gazette notification before locking any limit into a compliance document or an EC condition.