PM2.5 vs PM10: Understanding Particulate Sensors Guide

Understanding PM2.5 and PM10 particulate sensors is essential for anyone building air quality monitors in India, where particulate pollution is among the most critical environmental health issues. From Delhi’s infamous winter smog to industrial dust in manufacturing zones and Diwali fireworks nationwide, particulate matter (PM) is a constant concern for Indian makers and health-conscious citizens. This guide explains the difference between PM2.5 and PM10, how optical particle sensors work, and how to select and use the right sensor for your project.

Table of Contents

• PM2.5 vs PM10: What’s the Difference?
• Health Impact in the Indian Context
• Types of Particulate Sensors
• Popular Particulate Sensors Compared
• Wiring PMS5003 to Arduino/ESP32
• Calculating AQI from PM2.5 Readings
• Sensor Placement Tips
• Frequently Asked Questions

PM2.5 vs PM10: What’s the Difference?

Particulate matter is classified by aerodynamic diameter. PM10 refers to particles with a diameter of 10 micrometres or less, while PM2.5 refers to fine particles with a diameter of 2.5 micrometres or less. To put this in perspective, a human hair is approximately 70 micrometres in diameter—PM2.5 particles are nearly 30 times smaller.

• PM10 sources: Road dust, construction dust, pollen, mould spores, sea salt. These are coarse particles that are typically filtered in the upper respiratory tract.
• PM2.5 sources: Vehicle exhaust, industrial emissions, biomass burning, secondary aerosols formed from chemical reactions. These fine particles penetrate deep into the lungs and can enter the bloodstream.
• PM1.0: Some advanced sensors also measure particles ≤1 micrometre, which are the most dangerous as they can cross the blood-brain barrier.

From a sensor perspective, PM2.5 is harder to measure accurately because the particles are near the lower limit of optical detection. Most optical sensors use laser light scattering and can reliably detect PM2.5, but accuracy decreases in very humid conditions (above 80% RH) because water droplets mimic aerosol particles.

Health Impact in the Indian Context

India accounts for 17% of global deaths from outdoor air pollution. The Indian National Ambient Air Quality Standards (NAAQS) set PM2.5 limits at 60 µg/m³ (24-hour average) and PM10 at 100 µg/m³—both significantly more lenient than WHO guidelines (15 µg/m³ PM2.5, 45 µg/m³ PM10). Cities like Delhi, Patna, Gwalior, and Raipur regularly exceed even the relaxed Indian standards.

Health effects by PM2.5 concentration (µg/m³):

• 0–30: Good
• 31–60: Satisfactory
• 61–90: Moderately Polluted
• 91–120: Poor
• 121–250: Very Poor
• >250: Severe (health emergency)

Types of Particulate Sensors

Three main technologies are used in hobbyist and professional particulate sensors:

1. Optical Particle Counters (OPC): Use laser light scattering to count and size individual particles. The most common type used in consumer and hobbyist sensors. Examples: PMS5003, SPS30, OPC-N3.
2. Gravimetric Sensors: Weigh filter paper before and after sampling to measure particle mass directly. Gold standard for accuracy but impractical for continuous monitoring. Used in reference air quality stations.
3. MEMS Sensors: Newer technology using microscopic resonating elements. Very compact but less established. Used in some wearable air quality devices.

Popular Particulate Sensors Compared

For most Indian hobbyist projects, the PMS5003 offers the best balance of price, accuracy, and library support. The SPS30 is worth the higher price for long-term deployments requiring cleaning cycles and certified measurement performance.

Wiring PMS5003 to Arduino/ESP32

The PMS5003 uses UART (serial) communication at 9600 baud. It operates at 5V, but its UART output is 3.3V-compatible, making it safe to connect directly to ESP32 GPIO pins without level shifting.

#include <HardwareSerial.h>

HardwareSerial pmsSerial(2); // Use UART2

struct PMS5003Data {
  uint16_t pm1_0;
  uint16_t pm2_5;
  uint16_t pm10;
};

bool readPMS(PMS5003Data &data) {
  if (pmsSerial.available() < 32) return false;
  
  uint8_t buf[32];
  if (pmsSerial.read() != 0x42) return false;
  if (pmsSerial.read() != 0x4d) return false;
  
  for (int i = 2; i < 32; i++) buf[i] = pmsSerial.read();
  
  data.pm1_0 = (buf[10] << 8) | buf[11];
  data.pm2_5 = (buf[12] << 8) | buf[13];
  data.pm10  = (buf[14] << 8) | buf[15];
  return true;
}

void setup() {
  Serial.begin(115200);
  pmsSerial.begin(9600, SERIAL_8N1, 16, 17);
}

void loop() {
  PMS5003Data data;
  if (readPMS(data)) {
    Serial.printf("PM1.0: %d PM2.5: %d PM10: %d µg/m³n",
                  data.pm1_0, data.pm2_5, data.pm10);
  }
  delay(1000);
}

Calculating AQI from PM2.5 Readings

India’s Central Pollution Control Board (CPCB) uses a piecewise linear formula to convert PM2.5 concentration (µg/m³, 24-hour average) to Air Quality Index (AQI). For real-time monitoring (instantaneous readings), use the NowCast algorithm:

// Simplified AQI calculation for India (CPCB standard)
int calculateAQI_PM25(float pm25) {
  if (pm25 <= 30) return (int)((50.0 / 30.0) * pm25);
  if (pm25 <= 60) return (int)(50 + ((50.0 / 30.0) * (pm25 - 30)));
  if (pm25 <= 90) return (int)(100 + ((100.0 / 30.0) * (pm25 - 60)));
  if (pm25 <= 120) return (int)(200 + ((100.0 / 30.0) * (pm25 - 90)));
  if (pm25 <= 250) return (int)(300 + ((100.0 / 130.0) * (pm25 - 120)));
  return (int)(400 + ((100.0 / 130.0) * (pm25 - 250)));
}

Sensor Placement Tips

Where and how you place a particulate sensor dramatically affects reading quality:

• Height: Mount at 1.5–2.5 metres above the floor for representative ambient readings
• Away from sources: Keep at least 1 metre from cooking areas, printers, and air conditioning units
• Ventilation: Ensure airflow around the sensor; do not enclose it in a sealed box without passive ventilation
• Orientation: Most optical sensors have a preferred orientation (inlet facing down or sideways) per datasheet
• Humidity correction: In monsoon months (70%+ RH), raw PM2.5 readings can be 50–200% higher than actual values; apply humidity correction using Laulainen or Malm formulas

Frequently Asked Questions

Is PM2.5 or PM10 more dangerous to health?

PM2.5 is significantly more dangerous than PM10. The smaller particles penetrate deeper into the respiratory tract—PM2.5 reaches the alveoli in the lungs, PM1.0 can cross into the bloodstream. Short-term exposure to high PM2.5 causes respiratory irritation, and long-term exposure is linked to lung cancer, cardiovascular disease, and premature death. This is why WHO guidelines are much stricter for PM2.5 than for PM10.

Why does my particulate sensor read very high values during India’s Diwali?

Fireworks produce massive amounts of fine particulate matter, metal oxides, and sulphur compounds. PM2.5 levels in Indian cities regularly reach 500–1,000 µg/m³ (Severe+ category) during Diwali night. Your sensor is correctly detecting these extreme levels. Some optical sensors saturate at very high concentrations—check your sensor’s maximum measurable range in the datasheet.

How often do I need to clean my particulate sensor?

Most optical particulate sensors (PMS5003, SDS011) use an internal fan to draw air through. The fan and optical chamber accumulate dust over time, causing readings to drift upward. Clean the sensor every 6–12 months by gently blowing compressed air through the inlet. The Sensirion SPS30 has an automated fan cleaning cycle built-in that runs weekly by default.

Can I trust cheap ₹200–₹300 dust sensors from local markets?

Low-cost sensors like the GP2Y1010 and DSM501 provide only a voltage output proportional to particle density without sizing particles into PM2.5 and PM10 categories. They cannot distinguish between PM2.5 and PM10, have poor calibration, and significant temperature and humidity sensitivity. For meaningful air quality data in Indian conditions, invest in a PMS5003 or SDS011 which provides calibrated mass concentration readings in µg/m³.

What is the lifespan of a particulate sensor?

The PMS5003 and SDS011 are rated for approximately 8,000 hours of use (just under one year of 24/7 operation). Running the sensor intermittently—say, 5 minutes every hour—can extend lifespan to 5–7 years. The SPS30 has a rated lifetime of 10 years with its automatic cleaning cycle, making it the best choice for long-term fixed installations.

Shop Environmental Sensors at Zbotic →