AI‑Driven Power Consumption May Prompt Unwilling Revival of Coal in India's Energy Mix

The unprecedented surge in computational workloads generated by artificial‑intelligence applications across banking, pharmaceuticals, and telecommunications has manifested in a measurable increase in the national electric load, compelling policy makers in New Delhi to revisit previously abandoned coal‑fired generation capacities with a gravity that belies earlier proclamations of a swift transition to renewable sources.

While the Ministry of Power, under the stewardship of the current Energy Minister, has repeatedly asserted that coal will be consigned to the dustbin of history, internal briefing papers obtained by this bureau indicate that the projected demand curve for data‑center electricity may exceed the combined output of existing solar and wind farms by as much as thirty per cent within the next five years, thereby prompting a reluctant but calculable reopening of dormant lignite pits in Jharkhand and eastern Madhya Pradesh.

Industry analysts, who have long cautioned against the romanticisation of green electricity in a nation still grappling with infrastructural bottlenecks and seasonal grid stress, note that the average data‑centre now consumes roughly eight megawatts per hour, a figure that, when multiplied by the projected thirty‑two thousand new installations slated for completion before 2030, translates into an additional load that could outstrip the incremental capacity promised by the National Solar Mission, thereby exposing a lacuna in the government's own forecasts.

Consequently, the Coal India Limited board, emboldened by a modest resurgence in export orders for metallurgical coal to East Asian markets, has submitted a revised capital‑intensity plan that envisages the refurbishment of thirty‑four aging thermal stations, a move that has been met with muted applause in parliamentary committees yet with a lingering sigh of relief among regional power distributors who fear that a wholesale reliance on intermittent renewables might jeopardise tariff stability for industrial consumers.

Environmental NGOs, while continuing to champion the cause of decarbonisation, have nevertheless issued a restrained admonition that the re‑commissioning of coal plants without the deployment of state‑of‑the‑art carbon‑capture technology may contravene India’s own climate‑change commitments under the Paris Accord, a point that the Ministry of Environment has tentatively addressed by proposing a timeline for retrofitting that many observers deem both aspirational and insufficiently financed.

Financial markets, observing the juxtaposition of a nascent AI‑driven demand curve and a potentially revived coal sector, have modestly adjusted the credit spreads of major utilities, a reaction that may be interpreted as a cautious endorsement of policy flexibility but also as an implicit warning that excessive reliance on a single, carbon‑intensive fuel source could ultimately erode investor confidence in India's broader green‑bond framework.

Given that the projected electricity consumption attributed to artificial‑intelligence training clusters may compel the Central Electricity Regulatory Commission to sanction additional baseload capacity, one must inquire whether the existing tariff adjudication mechanisms possess the requisite agility to prevent an undue transmission of coal‑related cost burdens onto small‑and‑medium‑scale enterprises that constitute the backbone of India's manufacturing sector.

Equally pressing is the question of whether the Ministry of Coal's recent allocation of funds for the refurbishment of obsolete thermal units incorporates mandatory monitoring provisions that would ensure compliance with the Emissions Standards promulgated under the National Clean Air Programme, a safeguard that, if absent, would render the revival exercise tantamount to regulatory tokenism.

A further line of inquiry must address the extent to which the projected fiscal impact of additional coal‑derived generation, when juxtaposed against the budgetary allocations for renewable‑energy subsidies, might compel an inadvertent reallocation of public resources that undermines the stated ambition of achieving a thirty‑percent reduction in coal’s share of the energy mix by 2030.

In light of the imminent commissioning of numerous AI‑focused data facilities in proximity to coal‑rich basins, does the existing environmental impact assessment framework possess sufficient granularity to evaluate cumulative emissions over a twenty‑year horizon, or does it remain constrained by a procedural myopia that privileges short‑term grid reliability over long‑term public health considerations?

Moreover, should the central government's fiscal projections for subsidising ancillary services to artificial‑intelligence data centres be reconciled with the statutory obligations to fund rural electrification schemes, one must ask whether the prevailing budgeting conventions implicitly sanction a preferential allocation of scarce public capital to technologically elite enclaves at the expense of broader socioeconomic uplift.

Finally, it remains to be examined whether the statutory powers granted to the Electricity Act's adjudicatory bodies are sufficient to compel transparent disclosure of AI‑related load forecasts by private operators, thereby enabling a democratic scrutiny that might forestall an unchecked resurgence of coal under the guise of technological progress.

Published: May 18, 2026

Published: May 18, 2026