Data Centers, AI, and the Coming Energy Crunch
Imagine the incongruity: gleaming, hyper-modern data centers—the backbone of artificial intelligence and advanced cloud computing—powered by the decaying infrastructure of America’s old coal plants. This unexpected pairing is at the heart of a new partnership between Babcock & Wilcox (B&W) and Denham Capital, aiming to confront the mushrooming power demands of data centers in both the U.S. and Europe. Their solution: rapidly repurposing shuttered coal-fired plants into natural gas-powered facilities, providing a cost-effective stopgap to satisfy tech’s insatiable power appetite.
The scale of the need is staggering. According to industry analyses, data centers—especially those underpinning intense AI workloads—are expected to demand up to 65 gigawatts (GW) of power globally between 2025 and 2028. The U.S. alone faces a 45 GW shortfall above existing supplies. Compounding matters, a five-year median wait for new power projects to connect to the grid leaves few options on the table for a quick fix. In this bottleneck, existing sites with grid infrastructure already in place offer a rare pragmatism no one can ignore.
B&W brings over 160 years of engineering prowess, particularly in coal-to-gas conversions, while Denham Capital touts its experience managing and developing power assets across six continents. This coupling, announced with typical fanfare, is a gambit to deploy “transitional” energy sources for the data economy’s near-term future.
The Science and Politics of Transition Fuels
The headline claim: natural gas emits about half the carbon dioxide per unit of energy that coal does. But the reality, as climate scientists hasten to point out, is far less comforting. While switching from coal to gas might seem a triumph for expediency and emissions reduction, the process remains fraught with trade-offs. Methane leaks from extraction and distribution make natural gas far from a clean solution—just a slightly less egregious one.
This is where the notion of “transitional fuel” enters the debate. Harvard climate expert Dr. Leah Stokes describes natural gas in these terms: “It’s a bridge, not a destination.” Undeniably, rapid decarbonization is needed. Yet, political realities, the ambitions of Big Tech, and legacy grid challenges continually conspire to prioritize quick wins over sustainable transformation. In fact, Pew Research Center studies suggest that while 69% of Americans support phasing out fossil fuels, energy policy remains stubbornly incremental.
“We should not mistake the repowering of coal plants as progress if it cements dependence on yet another fossil fuel. Every dollar spent this way is a dollar not invested in truly clean solutions.” — Dr. Leah Stokes, Harvard University
European regulators are taking note, with the European Commission pressing for stranded coal assets to be revamped for renewables rather than fossil-based stopgaps. Yet the private sector, especially in the U.S., retains wide latitude—often favoring the low-hanging fruit of natural gas given the financial and regulatory inertia. Decision-makers at Denham and B&W argue that their partnership isn’t a retreat from climate commitments, but rather a strategic necessity in a world where data—and thus energy—can’t wait.
The Opportunity and the Ongoing Moral Hazard
A closer look reveals a powerful tension. Yes, these conversions may keep grids stable and data flowing for the next wave of technological leaps, but what precedent does this set? If we build data center booms atop natural gas, do we risk entrenching more fossil fuel infrastructure right when the world most needs to pivot?
History offers a cautionary tale. In the last century, the notion of “clean coal” was touted as a panacea to bridge the U.S. economy into a green future. Billions in subsidies later, the emissions cuts proved marginal, and entrenched industries stalled the shift to renewables. Now, the repurposing of coal plants for natural gas teeters on the edge of repeating that cycle, absent strong policy guardrails and commitments to eventual full decarbonization.
Denham is reportedly poised to launch a dedicated portfolio company focusing purely on these conversions, with ambitions to streamline engineering and environmental retrofits. Yet for the data center sector—so often cast as future-forward and visionary—reliance on old technologies feels more like a compromise than an innovation.
Wouldn’t the true leap forward be repurposing these sites for solar, wind, or batteries? The interconnection queues for renewables might be clogged now, but advocates argue fast-tracking their approval would represent public investment in real long-term climate solutions.
Ultimately, the B&W and Denham partnership presents a dilemma progressives cannot ignore. Solving for today’s growth cannot come at the expense of tomorrow’s survival. Policy, investment, and activism must pressure these new ventures to be a temporary measure—a stepping stone to a truly sustainable data economy—not the final chapter in fossil-fueled power.
