The Data Center Scapegoat: Masking Decades of Grid Underinvestment
Key Highlights
- Major blackouts like Iberian and ERCOT were caused by systemic issues, not solely by renewable energy sources.
- Aging infrastructure and regulatory barriers have delayed necessary investments in grid modernization, exacerbating reliability issues.
- The surge in digital demand from AI and data centers exposes the grid's outdated design, requiring faster, more flexible planning.
- Distributed energy solutions such as microgrids and private investments are emerging as vital components to meet modern energy needs.
- The future of reliable power lies in a combination of infrastructure upgrades, innovative technologies, and diversified energy sources, rather than blaming specific technologies.
Blaming the energy transition when the grid goes down is a knee-jerk reaction. Surely, so the story goes, it’s the fault of intermittent solar and wind when catastrophes such as the Iberian Blackout of 2025 or the ERCOT crisis of 2021, right?
Wrong. Subsequent investigation discovered that multiple systemic factors were in play during both grid meltdowns. This week, the European network of electricity transmission system operators called ENTSO-E reported that the massive outage which struck the Iberian Peninsula last year resulted from a huge voltage surge and lack of control protections in place to contain that surge.
The ERCOT power collapse during Winter Storm Uri also wasn’t the fault of renewables, but rather an extended freezing period which impacted conventional power plants, grid infrastructure and the conventional planning of the Electric Reliability Council of Texas system operators alike.
These two grid crises converge just to make a point: The causes of blackouts are rarely simple nor the fault of just one thing. In other words, grids are massive, interconnected machines made up of many parts and typically joined to act as one system with predictable frequencies, voltages and loads.
The machine has served us well, with a few exceptions, over the past 100 years. It was ready when post-war industrialization expanded across the U.S. and most developed nations. It handled new generation from renewables as well as energy efficiency technologies as load flattened out and decarbonization goals took hold in the early 21st century.
The outdated grid was falling behind before AI was a thing
So, it’s not the grid’s fault that suddenly an enormous influx of demand from artificial intelligence, data centers and industrial automation is swamping the decks of our electrical infrastructure, right?
Again, it’s complicated.
The problem with outdated and underdeveloped electric transmission and distribution systems was alerted long before we ever heard of ChatGPT or OpenAI, as many energy experts have noted. The anticipated arrival of gigawatt-scale AI and cloud-based computing is not the villain, but rather the messenger making it obvious that the grid has languished too long.
“The framing of the argument that data centers are breaking the grid or taking all of our energy away oversimplifies what is happening,” Jay Jayasuriya, principal at professional services firm Sendero Consulting, said in an emailed series of questions and answers with EnergyTech.com.
“The growth in AI and data center demand is significant, and that absolutely requires thoughtful planning and coordination,” Jayasuriya added. “The stress we are seeing on the gird existed long before AI. For years, infrastructure investment in transmission and firm generation lagged behind emerging risks, and planning processes were optimized for flat or modest load growth. Forecasts were backward looking, and capital approvals followed those forecasts.”
Past is prologue, so the saying goes, and yet it should be a guide, not an anchor, to quote another proverb. The power grid proved itself as a magnificent tool for economic growth and prosperity and comfort for all the people it touches, but utilities, regulators, politicians and system planners caught a touch of complacency and sticker shock when it comes to approving and paying for modernization on the scale required to win the AI race.
And it’s not really the grid’s fault alone.
“Utilities operate within a regulatory compact, and overbuilding generation in a flat load environment can create cost pressure and regulatory pushback, so conservative procurement was aligned with the forecasts at the time,” Jayasuriya noted.
The interconnection queue is long and full of headwind
Utilities who face elected regulatory boards must play the political game, and thus legitimately face barriers to ambitious growth projects. Ratepayers must guard their pocketbook, too, and ratepayers vote.
And there it was and here we are. Innovation in the digital infrastructure sector has created tiny processing chips which can handle supercomputing tasks. If we’re not jumping on it, the Russians and Chinese sure are.
And the U.S. will aim to keep up, even if it’s through behind-the-meter and off-grid projects built as prime power and paid for more directly by the data center customers. The key challenge is that typical utility interconnection time frames of seven to 10 years will not match up with data center industry needs of dedicated power connection in three to five years.
“What we are experiencing now is a speed mismatch where digital demand is accelerating faster than regulated infrastructure processes can respond,” Jayasuriya said. “Data centers are not the root cause of reliability risk, but a visible signal that the system needs more agile planning, better coordination and a clearer long-term investment strategy.”
Indeed, tech giants such as Microsoft, Google, Tesla and Oracle are investing in long-term power purchase deals with utilities around all forms of generation, including nuclear. Many of them also signed on to the Trump Administration’s “Ratepayer Protection Pledge” indicating they are willing to shoulder their fair share of costs for both generation and system build-out to serve their AI factories and data centers.
And yet the AI Race compresses timeframes in a way that are uncomfortable but necessary for everyone involved.
“The rules were built for stability, and the environment is now more dynamic, requiring faster coordination across utilities, regulators, policymakers and large load customers,” Jayasuriya added.
Speed-to-power may advance grid investment, not impede
Make no mistake, on-site power is going to meet the race head-on, whether they are called microgrids, behind-the-meter or energy parks. Experts such as former U.S. Energy Secretary Ernest Moniz are predicting it, and leaders from gas-fired infrastructure firms like Williams Cos. are already creating new ventures to build it.
Microgrid developer Enchanted Rock’s CEO John Carrington said recently that his company was helping its customers accelerate faster projects than utility interconnection queues can process. Enchanted Rock is hardly alone.
Before jumping to conclusions that the power infrastructure is becoming a wild west of electricity haves and have-nots, rest assured that some believe this Speed to Power movement could prove positive in the long run.
“This infrastructure investment need existed long before AI demand accelerated,” Sendero Consulting’s Jayasuriya pointed out. “Rather than harming the grid, data center growth concentrates demand in a way that generates necessary investment momentum.
“Many data center developers are co-locating with combined cycle generation or battery energy storage systems, while some are exploring long-duration storage and small modular nuclear concepts,” he added. “This introduces private capital and innovation into parts of the value chain that have lacked investment appetite for years.”
So it may be early to read eulogies for the power grid as a whole. But it’s increasingly obvious that the 21st century overhaul may come from the distributed energy edges rather than the centralized middle out.
Blaming the data centers and Big Tech for causing this dilemma is about as jumpy as pinning the Iberian and ERCOT outages on the solar panels and wind turbines. The seeds behind both catastrophes were far more complex than that.
And the solutions to solving the power challenge of this new Industrial Compute Age are every bit as multi-faceted and diverse.
About the Author
Rod Walton, EnergyTech Managing Editor
Managing Editor
For EnergyTech editorial inquiries, please contact Managing Editor Rod Walton at [email protected].
Rod Walton has spent 17 years covering the energy industry as a newspaper and trade journalist. He formerly was energy writer and business editor at the Tulsa World. Later, he spent six years covering the electricity power sector for Pennwell and Clarion Events. He joined Endeavor and EnergyTech in November 2021.
Walton earned his Bachelors degree in journalism from the University of Oklahoma. His career stops include the Moore American, Bartlesville Examiner-Enterprise, Wagoner Tribune and Tulsa World.
EnergyTech is focused on the mission critical and large-scale energy users and their sustainability and resiliency goals. These include the commercial and industrial sectors, as well as the military, universities, data centers and microgrids. The C&I sectors together account for close to 30 percent of greenhouse gas emissions in the U.S.
He was named Managing Editor for Microgrid Knowledge and EnergyTech starting July 1, 2023
Many large-scale energy users such as Fortune 500 companies, and mission-critical users such as military bases, universities, healthcare facilities, public safety and data centers, shifting their energy priorities to reach net-zero carbon goals within the coming decades. These include plans for renewable energy power purchase agreements, but also on-site resiliency projects such as microgrids, combined heat and power, rooftop solar, energy storage, digitalization and building efficiency upgrades.