Understanding our electricity system through modeling is crucial for future energy planning, yet these models demand comprehensive data on electricity usage. This includes the amount of electricity used by individuals, industries, and data centers, as well as the specific times and locations where this demand occurs. Unfortunately, there is a significant data gap regarding the size, location, and energy consumption of data centers. This lack of transparency is exacerbated by the confidentiality and speculative nature of data center proposals, with some owners concealing potential sites for “security and competitive reasons.”
The tech industry as a whole lacks transparency. Companies like Google, Amazon, and Microsoft often use subsidiaries to develop data centers, hiding their presence and offering little information about resource consumption. There’s no central database detailing data centers’ operational status, construction, or planning stages, and developers face no accountability, potentially exploiting this lack of oversight for profit.
For our analysis, UCS used electricity demand projections from Evolved Energy Research, which utilizes existing data center information from Baxtel. This was supplemented with news on announced projects and utility filings from public and private databases. While databases like S&P Global and BloombergNEF offer project details, they are costly and incomplete.
We also assumed which proposals might be constructed using data from an Independent System Operator (ISO) and purchased data from S&P Global. This process was challenging, time-consuming, and expensive. Eventually, our electricity demand projections aligned with other bottom-up estimates from national labs and industry studies. However, the rapidly changing data center market outdated our projections quickly, highlighting the urgent need for transparent reporting from data center developers.
Lacking precise details on data center locations and operations forces utility planners, grid operators, and agencies to rely on estimates and projections, risking inaccurate conclusions. This could lead to unnecessary investments in costly fossil fuel plants, which is not ideal for long-term investment decisions.
Errors in electricity demand data can result in poor infrastructure planning. Overbuilding leads to higher costs and emissions, while under-building causes reliability issues, both of which increase costs when corrections are needed.
Developers’ rush to construct data centers, demanding utilities meet their power needs, is problematic, particularly in states like Illinois, Michigan, Wisconsin, Louisiana, and California. Challenges include lack of state Integrated Resource Planning (IRP) processes and interstate trading that might increase fossil fuel generation.
Transparency and accountability issues in these and other states amplify planning problems: utilities may overbuild, delay coal plant retirements, or invest unnecessarily in fossil gas capacity, increasing costs and pollution while harming communities. Often, communities are uninformed about data centers’ energy use, water consumption, and pollution or receive incorrect information. Some centers are proposed in areas already suffering pollution, exacerbating problems. The lure of billion-dollar investments can cause local decision-makers to overlook risks.
State and federal policies risk increasing future reliance on fossil fuels. States often see data centers as drivers of economic growth, job creation, and tax revenue increases, offering developers tax breaks and incentives, leading to rapid growth in states with larger incentives.
However, some states are reassessing these policies. Texas faces significant revenue losses, pausing new data centers while their Public Utility Commission and ERCOT audit projects, though allowing Meta’s $10 billion project to proceed. Minnesota removed tax incentives, ending the electricity exemption from sales tax. Wisconsin eliminated financing incentives in two municipalities. New York has imposed a temporary moratorium on new data centers over 50 MW until impacts can be evaluated.
While it’s positive that these states are pausing to consider community concerns, more action is needed to ensure a clean, affordable, and healthy electricity system for impending demand. Our analysis shows the US can meet increased data center demand with clean energy, reducing fossil fuel reliance. Strong policies are crucial to support this transition, alongside best practices and regulatory changes at all governmental levels to ensure accountability and smart planning.
Increase data center transparency and accountability
State and federal policymakers should mandate transparency in data center developers’ and utilities’ contract negotiations, eliminating non-disclosure agreements. Public proceedings should be held for power purchase agreements and grid connections, with ample community notice for informed decisions. Data centers must publicly report their electricity, water, and land use, emissions, and other impacts and be held accountable for any adverse effects on reliability, environment, or public health.
National efforts to improve data center reporting include the North American Electric Reliability Corporation (NERC) requiring data centers to register as entities and developing reliability standards.
NERC has the opportunity to enact rigorous standards that protect the grid and communities that rely on it everyday — submit public comment here in support of strong reliability standards for data centers.
NERC’s draft standards aim to set data requirements, verification, and reporting procedures for data center owners, ensuring grid operators and planners have necessary site data for reliable operation and resource adequacy. While a good start, the provided information remains confidential and focused solely on grid reliability. The Energy Information Administration (EIA) is gathering data on data center energy usage. At state and local levels, efforts include banning NDAs for lawmakers and preventing utilities from connecting data centers through ex parte processes, as seen in Michigan. These practices should be standard for all proposals.
Require utilities to conduct transparent long-term planning to include data centers
States must require utilities to create long-term integrated resource plans (IRPs) that transparently report electricity demand, including projected increases due to data centers, population changes, and electrification. Currently, only 27 states are considering legislation to regulate data centers. Detailed modeling and analysis are essential for investment decisions to meet electricity demand and adhere to climate and clean energy policies.
IRP plans and their assumptions should be public. Regulators must show communities that electricity generation investments are well-planned, minimizing over- or underbuilding risks. Data center developers should cover additional electricity system costs due to increased loads, eliminating speculation to prevent ratepayers from bearing overbuilding costs.
There’s uncertainty about whether data center electricity demand will continue rising or if the AI bubble will burst. AI infrastructure investments far exceed data center revenue, concerning investors. If the bubble bursts, excess generating capacity will lead to stranded investments, raising questions about financial responsibility. As IRPs are developed, utilities should focus on near-term projects with significant financial commitments.
The data center boom is outpacing our electricity system’s response capability, already increasing electricity rates. Governments need time to assess our complex electricity system, with its aging infrastructure and growing demand. Accurate data and standardized reporting are crucial for informed investment decisions, ensuring affordable electricity and limiting climate change impacts.
Data transparency should extend to water use and environmental impact reporting. Good data is needed to develop policies encouraging renewable energy investments and keeping electricity costs affordable. Accurate data is vital for smart planning and informed decisions, providing consumers with clean, affordable power. Ensuring clean, affordable energy starts with informed decision-makers holding AI and utilities accountable for their impacts and grid investments.
Urge NERC to establish strong standards governing data centers today! Use the form here to submit your comment calling on NERC to comprehensively address the nature of the risk that computational loads such as data centers pose, ensure the proposed requirements are robust and fully vetted by stakeholders, ensure that all AI data centers are held to the same robust standards, and ensure that the process is transparent for the public and developers alike.
