India's Shadow Power Grid Is Full Of Carcinogens

Delhi's power grid does not get a summer off. In 2024, the city recorded its highest-ever peak demand of 8,656 MW. In 2025, even with wetter conditions slightly moderating temperatures, the June peak still exceeded 8,400 MW, and August demand climbed to 7,050 MW, about 2% higher than the year before. The ceiling keeps rising. Between 2015 and 2025, Delhi's peak electricity demand grew from 5,846 MW to 8,442 MW, a 44% increase in a single decade, with the sharpest single-year spike of 16.4% recorded in 2024. On the hottest and most humid days, year-on-year peak demand growth runs 3.8 times higher than on moderate days.

The grid buckles. In Delhi alone, recorded instances of power cuts totalled 2,660 in January 2024 and rose to 3,278 in January 2025. Nationally, the Central Electricity Authority estimates that over 250 million hours of electricity outages occur annually across India, and these shortages cost the country approximately 1.9% of its GDP.

Every time the grid fails, something else switches on. What India has quietly built, in response to decades of unreliable supply, is a shadow power infrastructure of staggering scale. The installed capacity of captive diesel generator sets across the country is estimated at nearly 90,000 MW, roughly 17% of India's national grid, running in parallel, burning diesel, and entirely outside the view of most air quality monitoring.

India accounted for 8.8% of the global diesel generator market in 2024, generating $1,570 million in revenue. The market was valued at $1.57 billion in 2025 and is projected to reach $2.23 billion by 2031. For a country of 1.4 billion people, 90,000 MW of captive diesel backup translates to roughly one generator for every neighbourhood, every hospital corridor, every mid-rise apartment block that cannot afford to wait for the grid to return. Industrial sites held nearly half the market share in 2025, but residential demand is now the fastest-growing segment, rising at a nearly 9% CAGR as middle-class households in states like Uttar Pradesh and Bihar increasingly treat diesel gensets as essential infrastructure rather than a luxury backup.

Most conversations about air pollution picture highway traffic, factory smokestacks, or coal plants. Almost nobody talks about the diesel generator. And it is, according to new research by Professor Sailesh Narayan Behera, quietly poisoning the air in ways we have barely begun to account for.

A study published in the Journal of Hazardous Materials Advances by Professor Behera and Ph.D. scholar Vishnu Kumar offers one of the most granular accounts yet of what exactly comes out of a stationary diesel engine, and what happens to it once it enters the human body. The findings are not reassuring.

Professor Behera tracked 24 chemical elements across 10 particle-size categories, from near-nanoscale (56 nm) to particles visible under modest magnification (18 µm), at six engine loads. What they found is that diesel particulate matter (DPM) does not just rise and drift away. It concentrates. The peak mass concentration occurred at the 0.10–0.18 µm size range, submicron particles too small to see, too small to feel, but precisely sized to slip past the body's respiratory defences and settle deep in the lungs. At full engine load, concentration at that size reached 20.4 mg per cubic metre of air. That is not background reading. That is source-level exposure.

The elemental makeup of this exhaust reads like a toxicology shortlist. Sulphur, calcium, potassium, aluminium, sodium, magnesium, iron and zinc together account for over 90% of total elemental mass across all engine loads. These come largely from lubricating oil, fuel combustion and the mechanical wear of engine parts. But embedded within that bulk are the elements that matter most for human health: arsenic, chromium, cobalt, cadmium, nickel, and lead, present at lower concentrations but with disproportionate biological consequences.

Using the US Environmental Protection Agency's exposure pathway framework, Professor Behera modelled how people who spend time near these engines over a lifetime are exposed. Their measure of carcinogenic risk is the excess lifetime cancer risk, or ELCR, a probabilistic estimate of additional cancer cases per million people exposed. The USEPA's acceptable threshold sits at 1 in a million. At full engine load, even assuming the exhaust is diluted 1,000 times by the ambient atmosphere, the researchers calculated a combined ELCR of 53.7 per million. At 100-times dilution, it's still a generous assumption for someone working or living close to such a generator; at 100-times dilution, that number climbs to 537. At 10 times dilution, 5,370 per million.

Chromium alone drove 66% of total cancer risk, followed by cobalt at 19% and arsenic at 14%. Not because they were the most abundant elements, they ranked 10th, 9th, and 14th, respectively, by concentration, but because their inhalation slope factors, the measure of carcinogenic potency per unit dose, are extraordinarily high.

The non-cancer picture is more conditional but still pointed. The hazard quotient for non-carcinogenic elements exceeded the USEPA's safety threshold of 1.0 under the lowest dilution scenario at most engine loads. Workers in small workshops, guards at construction sites, residents in dense informal settlements where backup generators run through the night, these are the people absorbing the highest exposure. The MPPD model used by Professor Behera to simulate respiratory deposition found that particulate elements penetrate deepest into the pulmonary and alveolar regions, the parts of the lung from which the body has almost no mechanism for clearance, and that deposition fraction increases with every rise in engine load.

The 90,000 MW shadow grid has no emissions register. It has no health monitoring. It has no policy framework that treats it as the public health emergency the numbers suggest it is. What it does have is a growth forecast, a rising market valuation, and millions of units humming away in the gaps that official India has not yet managed to close. The generators will keep running until the grid improves. The grid will not improve without policy pressure. And policy pressure rarely arrives for a problem nobody has named yet.