Delhi plans mesh of sensors to monitor pollution air hot spots

Pollution is one of the biggest problems dogging Delhi, and the city experiences a public health emergency every winter with the air quality deteriorating to hazardous levels

The national capital’s 40 real-time ambient air quality monitoring stations (CAAQMS) account for a network that roughly covers every 37 sq km of Delhi’s 1,484 sq km area, if these stations were spread out at equal distance from each other across the city.

Vehicles drive through smog on Thursday morning on NH-24 near Akshardham temple in Delhi. (Raj K Raj/HT Photo)

While this network is considerably denser than the four stations functional each in Gurugram, Ghaziabad, Noida and Faridabad, and the two stations in Greater Noida, the Delhi Pollution Control Committee (DPCC) is now considering deploying multiple low-cost sensors to enhance the coverage further.

The idea is to install these sensors at the neighbourhood level to create a spatial grid alerting the anti-pollution body of changes in air quality levels in different parts of the same area, to enable local interventions in places where spikes are reported, DPCC officials aware of the matter said. If approved, the plan is start by using these sensors in Delhi’s 13 pollution hot spots, where urgent corrective action in needed, they added.

“These sensors function in a similar way to an automatic real-time stations present in Delhi, but they will also be able to identify local sources of pollution, which otherwise cannot be detected by a real-time station as it covers a substantially larger area. Depending on which sensor is showing high-levels of pollution, we can take corrective action, for example acting on dust in its vicinity or any kind of garbage burning,” said a senior DPCC official who asked not to be named.

Pollution is one of the biggest problems dogging the national capital, and the city experiences a public health emergency every winter with the air quality deteriorating to hazardous levels. Between November and December last year, for example, the city witnessed 17 “severe” days, and even “very poor” and “poor” days on the Air Quality Index (AQI) rating scale are considered periods of respite. To be sure, this problem is not restricted to winter. Delhi’s average PM2.5 level in 2021 was 87 micrograms per cubic metre, 17 times higher than safety levels prescribed by the World Health Organization (WHO).

In 2018, DPCC identified 13 pollution hot spots, which have a substantially higher annual PM 2.5 concentration than the city’s mean average. These are Jahangirpuri, Anand Vihar, Ashok Vihar, Wazirpur, Punjabi Bagh, Dwarka Sector 8, Rohini Sector 16, RK Puram, Bawana, Mundka, Narela, Okhla Phase II and Vivek Vihar.

Initially, it is these areas that DPCC wants to target with the network of low-cost sensors.

While an automatic monitoring station costs around ₹1-1.5 crore, a low-cost sensor costs only ₹50,000, if it is to measure both PM 2.5 and PM 10 particulate matter. For each additional sensor to measure gases such as SO2, O3 and NOx, the cost goes up by ₹25,000 per filter.

In Delhi, the 40 real-time stations have been installed by the Central Pollution Control Board (CPCB), the India Meteorological Department (IMD) and DPCC, each measuring PM 10, PM 2.5, SO2, NOx, NH3, CO, O3 and benzene.

The low-cost sensors are relatively less accurate but experts said that the data can be used very effectively for analysis and intervention.

Ronak Sutaria, founder, Respirer Living Sciences, who specialises in gathering local-level data from low-cost sensors, says while a real-time station influences area of 3-5 sq km, low-cost sensors could cover a radius of 1 sq km easily, creating both a spatial and temporal grid that can be analysed by DPCC.

“If we have to assume each pollution hot spot in Delhi is spread over an area of 5 sq km, five such low-cost sensors can be installed covering a kilometre each and you can then identify in which grid the pollution levels are higher. The denser the grid, the more accurate the readings will be,” he says, stating while costing nearly 30-40 times lesser than an ambient air quality station, these low-cost sensors still have an accuracy that is only around 10-15% apart from a real-time station.

“If the real-time station is showing “severe” air, then these sensors will show”severe” air too, and the actual concentration may be 10-20 micrograms apart, but it will still give a good idea of the concentration levels in the area,” he added, pointing out that local data also gives a better idea of spike of bad air spikes during different times of the day.

Professor SN Tripathi from IIT Kanpur, who is part of a committee formed by CPCB to look at ways to utilise both low-cost sensors and ambient air quality monitoring stations together, says Delhi could benefit with a low-cost network in addition to the existing stations. “While the existing network of 40 stations in Delhi is good enough for policy-level decisions, low-cost sensors will be able to give an idea of certain locations at the neighbourhood-level which have a higher concentration level. Based on this, you could strategically act on the sources of pollution around that sensor,” he said, adding that Beijing is uses a network of over 3,000 low-cost sensors.

Anumita Roychowdhury says the approach to use low-cost sensors to tackle hots pot pollution can help coordinate ground-level action better. “Only once we know the source, can we act on it. Ground-level action such as high readings from burning of waste or dust can be easily captured,” she said.