Aligning the Energy Moment: It's Go Time for the Electrification Future of C&I Customers

The U.S. has endured at least half a dozen oil and gas energy upheavals in the past half century: the Arab Oil Embargo, the 1980s glut market bust, the 1990s Persian Gulf War, the 2008 housing collapse and financial crisis and the COVID pandemic of the early 2020s. The wars in Ukraine and Iran are the latest.

Those were multiple history repeats that could have inspired a U.S. national energy policy, spanning nine presidencies all harping on energy security and abundance in their own time. Yet the net result in national energy policies: zero.

Today, separate from war or geopolitics, is an energy crisis of a unique, domestic nature. The rise of artificial intelligence, cloud-based computing, energy efficiency technologies, reshored industrial capacity and electrification in manufacturing have joined forces into an irresistible force moving against an immovable (or difficult to move) object called the power grid. And the grid, according to the American Society of Civil Engineers and others, is neither adequately capitalized nor modernized enough to deal with what’s happening now and in the immediate future.

“It’s not just an AI thing,” Blue Energy CEO Jake Jurewicz, who leads a startup seeking to build natural gas-powered digital infrastructure designed to transition to carbon-free and small nuclear energy,  said during another CERAWeek session . “Before ChatGPT, (grid operators such as ) MISO, PJM and others were throwing out alerts” about grid power constraints and long interconnection queues to get projects commercially operational.

The modern market dynamic is that major customers like digital giant Microsoft and chemical manufacturer Dow want grid-scale energy now, or at least as close to now as engineering, construction and connection timelines will allow.

“There’s an extreme willingness to pay for capacity in the next few years,” Blue Energy’s Jurewicz pointed out. “And it takes a long time to build out (utility grid) transmission.”

Moniz, who also is co-chair of the Nuclear Threat Initiative and CEO of the natural gas industry-supported Energy Futures Initiative nonprofit, predicted that many more behind-the-meter projects will be necessary to meet the demand curve of AI, data centers and what some (not Moniz) call the Industrial Compute Age.

AI, for instance, is here to stay and is reshaping commercial and industrial work. It may require at least 125 GW of utility-scale or distributed or on-site power by 2030, making timing critical.

“Everybody is using it and interacting with it in a real way,” said Conner Prochaska ,director of the Advanced Research Projects Agency-Energy (ARPA-E) within the U.S. Department of Energy. “We’re doing it and we need to make sure we have the energy for it.

“It’s an exciting time but we just make sure we put them in the right places at the right time for the right applications,” Prochaska added.

Those words could apply to siting of power generation as well as AI adoption. If timing is key, so is location as hyperscalers seek energy they might finance themselves to build quickly as prime power. Later, those could interconnect back to the grid.

Currently, most projects are developed the other way around.

Companies taking greater control of their own energy levers

Some Big Tech entities such as Google are taking a straight line by acquiring their own power project developers. Others, such as Microsoft, are making long-term, baseload power purchase agreements to acquire and revive retired nuclear units.

And others are aiming off-grid, tapping into the nation’s giant reserve of natural gas as a bridge to the future while meeting the moment now. Those include well-funded startups such as Fermi America, CyrusOne, New Era and Creekstone. European developer Nscale is so bullish on the prospects of gas-fired energy to power AI that it recently acquired the Monarch Compute Campus project in West Virginia.

Industry bubbles, those periods where demand growth appears unstoppable and bullishness is recklessly off the charts, are nothing new. They often lead to busts, as any oilfield veteran could tell you after living through five or six of them.

This is not that time, by most accounts.

“What’s different about this cycle is having bankable off-takers with hyperscalers,” Eve Hanson, senior vice president of research and innovation with investor Energy Impact Partners, said at CERAWeek. “They’re dictating how quickly things are going to come online.”

Those demands and the bankability behind them could push projects from design to commissioning within three years, veritable leap years ahead of the five to 10 years in the typical utility interconnection queue.

A decade is a long time to wait for timely power

Such an accelerated timeline might require more co-located and initially behind-the-meter power projects on an unprecedented scale. Winning races in AI and industrial reshoring cannot wait, but transmission lines can be built only so quickly and safely.

A 2023 report by the International Energy Agency estimated that the average lead time for an electricity overhead transmission line project in the U.S. was slightly more than 10 years, with close to 70% of that in planning and permitting stages. Grid connection was about four years.

“It’s the scale and the timing” driving a move toward off-grid prime power in the near term, said Todd Brezler, senior vice president for digital transformation at turbine equipment manufacturer Mitsubishi Power.

“If you think of building out the grid and transmission, a transmission project takes10 years...,” Brezler pointed out. “Hyperscalers want to build something in three to five years.”

Big companies have big money and make big political donations, both directly and indirectly through capital investment in the economy. The Trump Administration certainly gets that and is taking steps to shorten regulatory hurdles, speed use of federal land  and promote a  voluntary “ratepayer protection pledge,” which companies like Google, Tesla, Oracle, xAI and others have already signed on to develop power and pay for projects in ways that, theoretically at least, won’t negatively impact energy affordability for everyday customers.

“We think Europe overregulated and stifled innovation,” said ARPA-E’s Prochaska. “We don’t want to do that.”

Prochaska was nodding toward AI guardrails and computing capacity, as well as energy. In his world, and increasingly elsewhere, it’s not only the right time for an “all-of-the-above” power generation strategy but that may be the only choice. Renewables, nuclear, geothermal, natural gas and, yes, even, coal.

Not one of them is the only answer. And yet none of them can be left behind in this historical moment.

“The grid is one of the problems,” former Energy Secretary Moniz said. “I don’t think there’s a silver bullet.”

The challenge is keeping up. Timing it is key

What this means for microgrids remains to be seen. Clearly, a potentially lucrative opportunity has arrived for project development, whether you call the project a microgrid, minigrid, energy park or co-located power plant encompassing gas, solar and battery storage.

“It’s a challenge to keep up,” Mitsubishi Power’s Brezler said. “Reliability is still a prerequisite. If you don’t have reliability, you’re not going to have economic growth.”

Nobody really asked for the moment, but it’s here. The irresistible force of digital innovation is timed just right with the immovable need for electricity, in all its form and function. One cannot flourish without the other.

Aside from war and geopolitics, this was a key theme at CERAWeek bringing together leaders from all over the globe. The world is a collaboration if you are not rivals, and competition even if you are allies.

The race to win AI and create lasting energy resiliency is under way. First isn’t always best, but aligning the timing between power and load may divide winners and losers.

And good timing is a great thing.

Interact with Us on Social Media

Follow Microgrid Knowledge on LinkedIn