High Stakes at CERAWeek

It is not often – read “never” – that a CEO’s decision NOT to attend a conference is a global headline but so it was this weekend when Amin Nasser, the veteran CEO of Saudi Aramco, withdrew from CERAWeek in order to attend to matters closer to home. Mr. Nasser’s withdrawal underlines that the focus for all of us still heading to Houston will be on the Straits of Hormuz and its short, medium and long term consequences on gasoline prices.

As average American gasoline prices are destined to pass $4 gallon this week, the conversation amongst energy folks is about how quickly they will come down. Oil shock history has taught us that energy prices rise at the beginning of a crisis a lot faster than they fall at its end. And all the while, the focus on high gasoline prices at CERA this week is likely to obscure the other energy price shock lurking in the shadows which is the inexorable rise of electricity prices across the United States. Unlike gasoline prices, once retail electricity prices rise, they almost never go down.

Multiple factors go into the price of a complex commodity like electricity, but rightly or wrongly, upward pressure on electricity rates is being tied directly to the question of how we power the data center boom and who is going to pay for it. This is not an issue just for the energy industry to grapple with. How we generate and deliver the power that AI requires is one of the most important issues facing our nation. Our two political parties, which agree on almost nothing, agree that we need to do so in a manner that is affordable, durable and fair to energy consumers and communities – which means, all of us.

That is why I am in Houston this week. To talk to the data center leaders who walk the halls of CERAWeek, at what has historically been the ultimate conference of energy producers. These electricity consumers have more influence over the immediate future of our industry than is wielded by even the federal, state, local or regulatory authorities.

Not long ago, and even in some circles still today, the dominant view was that all of these new data centers would connect to the existing electricity system, commonly referred to as “the grid”. But, take it from someone who was responsible for building out of the grid at the US Department of Energy for a few years, the grid simply can’t keep up with the breadth or depth of the data center build out from a point of view of planning, permitting, physical construction or system reliability. Plus it is very hard for the data center companies to prove that they are NOT causing electricity prices for the rest of us to rise while they are siphoning out of the system vast quantities of existing electrons.

This means that the pendulum is swinging and, as we all know, pendulums never stop in the center. Data center companies are responding to calls that they supply their own power from new sources of electricity by co-locating generation with the data center. The realities of the grid, rather than the rules themselves drive this which is leading BYOG (bring your own generation) to become synonymous in industry parlance with “behind the meter”. But it is not that simple.

Power systems are not binary. The future won’t be purely grid-connected or behind-the-meter. It will be an integration of both, stitched together over time, involving physical infrastructure across multiple markets and differing regulatory frameworks. The energy systems themselves will be complex multi-billion dollar systems which trillion dollar data center companies are entirely reliant upon to conduct their businesses.

Behind the meter power is where the pendulum is swinging but the conversation needs to be about more than speed to power. It needs to encompass the use case for these power assets over their useful lives on the assumption that they will ultimately be incorporated into our national grid system. As such, reliability, certainty and affordability for all are important considerations even now.

None of this is entirely new. Large industrial energy users have always had to navigate system constraints and impacts. They’ve combined grid power access with onsite generation, demand response, energy efficiency, and long-term planning. They’ve built hybrid multi-technology solutions because they have had to. This part is not new. What is new is the size, the speed and the exacting performance requirements of the data center consumers. So, the focus going forward is on execution and coordination: who can execute to the next level of the new requirements and who can coordinate effectively across interconnection, site development, regulatory construct and community engagement. It is no time for amateurs.

Not too many years ago, the National Academy of Engineers voted electricity networks to be the single greatest engineering achievement of the 20th century. Ahead of the automobile; ahead of the airplane. Ready and reliable access to electricity underpins all of modern civilization even more so now in the age of AI than it did pre-AI. We need to all come together to make sure U.S. power grid, as supplemented and enhanced by data center-related energy assets, becomes the greatest engineering achievement of the 21st century as well.