90% of the sulphur dioxide drifting unchecked into Delhi's air comes from coal plants without pollution-control equipment. 81% of that traces specifically to plants the Centre itself exempted from ever installing it.

The Rule Didn't Fail on Enforcement

• It Failed by Being Redefined Out of Existence The commitment, stated precisely: in 2015, India gave every coal-fired power plant two years to install flue gas desulphurisation systems to cut sulphur dioxide emissions. The data, a decade later: only about 12% of India's coal fleet has FGD installed and as of July 2025, 78% of that fleet - Category C plants - will never have to install it at all. Notice what actually changed between those two facts. A deferral, which happened four times between 2015 and last year, keeps an obligation alive while pushing its deadline forward - the plant still owes compliance, just later. What happened in July 2025 was categorically different: the 2021 locational split into Category A (near Delhi-NCR or million-plus cities), B (near critically polluted areas) and C (everywhere else) was used not to phase compliance but to permanently remove Category C plants from the rule's scope. A deferred rule and an exempted rule look similar on paper. They are not the same failure and conflating them misses where this actually broke. Where the chain held: Category A plants, the ones sitting closest to political and media scrutiny, largely complied. Dadri and Mahatma Gandhi thermal power plants, both fitted with FGD, generated between 3,400 and 4,000 million units of electricity annually - output comparable to the region's largest emitters - while releasing only around 2,000 tonnes of SO2 each. FGD works exactly as designed when it's actually installed. This wasn't a technology failure or an unaffordable-capital-cost story at the plants where it mattered most politically. Where the chain broke: Rajpura's two units, sitting in Category C and therefore permanently exempt, released an estimated 20,851 and 22,690 tonnes of SO2 respectively - each unit alone emitting more than some entire uncontrolled plants combined. Talwandi Sabo, Rajiv Gandhi TPS, Guru Hargobind TPS, Harduaganj and Ropar sit in the same exempted category, none fitted with FGD. Sulphur dioxide does not respect a 300 km administrative boundary; it drifts, forms secondary particulate matter and lands in Delhi's air regardless of which category a plant was assigned. The specific link that broke isn't monitoring, funding or technological capacity - Category A proves all three can work. It's that the categorisation criterion itself, proximity to major cities rather than actual emission volume or downwind reach, gave regulators a definitional exit ramp instead of an enforcement mechanism. For an aspirant, that's the transferable insight: environmental rule dilution doesn't always show up as weak enforcement of a rule still on the books. Sometimes it shows up as the rule's boundary being redrawn until most of what it was meant to cover simply falls outside it.