Grid operators and regulators are rethinking how they forecast electricity demand from AI and data center developers, after discovering that many companies have been submitting large-load requests to multiple utilities simultaneously while they finalize site selection. That practice has inflated regional demand forecasts and distorted transmission planning, according to filings in a pending Federal Energy Regulatory Commission (FERC) rulemaking on large loads.
The core problem is what industry observers call duplicate demand. When a single data center project applies for power at several utilities at once, each utility plans as if that full load will materialize on its own system. Elizabeth K. Whitney, managing principal at Meguire Whitney, told Data Center Knowledge that this duplication skewed the regional grid’s overall demand picture, since projections stacked on top of each other rather than reflecting a single real project.
Shifting to Commitment-Based Forecasting
Google, in its FERC filing, has proposed a Capacity Commitment Framework that would require developers to make meaningful commercial commitments, such as long-term service agreements, minimum demand charges, upfront collateral, or withdrawal penalties, before a project can influence long-term transmission planning. Google also called for coordinated reviews that study large loads alongside nearby generation projects, arguing this could reduce unnecessary network upgrades and shorten timelines to energization.
Amazon, Microsoft, and OpenAI have expressed support for similar approaches. OpenAI favors expedited treatment for projects designated by the US Department of Energy as nationally significant, while Amazon has advocated using artificial intelligence and automation to speed up transmission engineering studies.
Not all proposals align on mechanism. NRG Energy has suggested replacing traditional interconnection queue priority with competitive, time-bound “open seasons” for transmission capacity, paired with requirements that large loads pay for the network upgrades they trigger.
Regional Grids Already Moving
Several regional grid operators have already begun implementing readiness requirements. ERCOT’s “Batch Zero” process requires developers to demonstrate project maturity before entering regional transmission studies. The Southwest Power Pool has developed its HILLGA framework for high-impact, large-load requests, and PJM Interconnection has its own large-load proceedings underway. According to Whitney, these efforts go beyond simply verifying demand, also addressing cost allocation, load flexibility, and co-located generation.
Industry groups including the Data Center Coalition argue that interconnection queues themselves are a symptom rather than the root cause, pointing instead to inadequate transmission infrastructure and long-term planning gaps. The coalition has called for expanded transmission planning, wider use of grid-enhancing technologies, and broader deployment of automation to accelerate engineering reviews.
For hosting operators and data center developers, the shift signals that speculative multi-site power requests may face more scrutiny going forward, with utilities increasingly requiring financial and contractual commitments before treating a project’s megawatts as real in planning models.