Coal Reenters Debate Over AI-era Electricity Needs

The guest essay "The Coal Conversation" by Amy Oliver Cooke argues that continued access to dispatchable generation is necessary to meet rising electricity demand from AI, data centers, and electrification. The article cites projections that U.S. peak summer demand could reach nearly 967,000 megawatts by 2030, an 18% increase from today, and reports that gas projects surged 1,810% year-over-year, with major turbine manufacturers backlogged to 2029 or 2030. The piece says Xcel Energy has asked Colorado regulators to keep coal plants open through 2030, and that Southern Company is tracking more than 50 GW of potential new customer load. Editorial analysis: For practitioners, the article frames reliable, always-on generation as a core constraint for hyperscale data center siting and for industrial AI deployments.
What happened
The guest column "The Coal Conversation" by Amy Oliver Cooke contends that dispatchable, always-on generation is required to support growing electricity demand from AI and data centers. The article cites projections that U.S. peak summer demand could reach nearly 967,000 megawatts by 2030, an 18% increase from today. It reports that gas projects surged 1,810% year-over-year, that major turbine manufacturers are backlogged to 2029 or 2030, and that Xcel Energy has petitioned Colorado regulators to keep coal-fired plants operating through 2030. The piece also states Southern Company is tracking more than 50 GW of potential new customer demand.
Editorial analysis - technical context
Industry observers note that hyperscale data centers require predictable, low-latency, and price-stable power. Long lead times for new pipelines, generation, and turbine delivery create a mismatch between short-term load growth and available dispatchable capacity. This mismatch raises practical siting and contractual challenges for operators that commit to multi-year infrastructure investments.
Context and significance
For practitioners, the column highlights a recurrent infrastructure fault line: the need to reconcile decarbonization goals with the physical reliability requirements of compute-heavy workloads. Reporting in the piece connects supply-side constraints to potential offshoring pressure for data loads, which could affect data residency and latency-sensitive applications.
What to watch
Indicators to follow include permitting and build timelines for new generation and pipelines, queue and backlog updates from turbine manufacturers, and utility interconnection approvals for large data center loads. Observers will also watch whether utilities and regulators adjust reliability rules or contracting mechanisms for 24/7 load.
Scoring Rationale
The article highlights a practical infrastructure constraint for AI and hyperscale computing: firm, always-on power. This is directly relevant to practitioners planning deployments or interpreting latency and data residency tradeoffs, but it is an opinion piece rather than a technical or regulatory development.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems


