Demand for Electricity Soars

The future ain’t what it used to be

“It’s difficult to make predictions — especially about the future.”

This wise quote has been attributed to everyone from Yogi Berra to Danish physicist Niels Bohr. Climate people likely agree.

Take demand for electricity, for example. It’s growing fast and the outlook is nothing short of extraordinary. But demand for power in the US had been flat over the past decade. Could anyone really have seen this coming? 

So how did we get here?

First came the climate strategy of electrify everything. Using electricity to heat homes, charge cars, cook, decarbonize industrial activity, and more is now widely understood to be the best way to lower greenhouse gas emissions — especially when the electricity is sourced from renewable power. Of course, switching from fossil energy to electricity ramps up demand for the latter. This surge in demand for electricity was clearly in the cards for a while.

Next came the drive to onshore manufacturing. See, for example, President Biden’s IRA plan and its emphasis on domestic jobs. Political pressure for more onshoring continues. This was a less foreseeable addition to the electricity demand.  

And now, seemingly out of the blue, here comes AI with its insatiable appetite for power. According to Goldman Sachs, AI will cause data center power demand to grow 160% by 2030.

Who needs to do what now?

For environmentalists

If we want to achieve our top goals, we’ll need to get better at balancing the demand for more infrastructure with our concerns about climate and nature. Big infrastructure projects like data centers and renewable power/transmission always create challenging tradeoffs. There will of course be some occasions where the right answer is to just say no to proposals for new infrastructure. But more often, in my view, the question should not be whether but how. How can we balance the need for new infrastructure with acceptable outcomes for communities and nature?

There are proven ways to answer tough questions like this. My former organization, TNC, deploys a process they call “Development by Design.” The idea is to use the so-called mitigation hierarchy — avoid, reduce, and offset — to manage such trade-offs in science-based ways. I’m sure other organizations have good ideas on how best to do this.

The key thing is to improve our skills at reaching prudent compromises with diverse parties who prioritize different outcomes. This is never easy, but it is doable.

And for you activists out there — don’t worry, we still need you too. It will always be important for activists to campaign against bad actors. That’s sometimes the only thing that brings people to the negotiating table.

For tech companies 

A few years back, tech giants like Microsoft and Apple made ambitious commitments to reach net zero by 2030. They have subsequently acknowledged it’s turning out to be more difficult than they had expected. And now AI is going to make their challenge exponentially more difficult. 

To their credit, the tech companies have not backed off their goals (at least not yet), and I don’t think they will. That means we can keep pushing some of the biggest and most powerful companies of all time to figure out as fast as possible how best to: 

1. accelerate the buildout of clean energy facilities (and transmission);  

2. make data centers cleaner and more community-friendly; and 

3. take carbon removal to scale. 

All of this is critical. These are the same challenges that society as a whole has no choice but to address.

I also think these companies realize they not only have an obligation to develop AI in an environmentally responsible manner, but they also have the opportunity to use that AI to advance climate solutions. But we’ll get into that in another newsletter. 

In the meantime, there’s one more area where tech companies can help — leveraging the substantial existing additional electricity and storage capacity we currently leave on the table. 

For example, we often generate more energy than we can use (like when it’s very windy or sunny). We can do more to incentivize power users with some flexibility to use more energy when it’s abundantly available and less when it’s not. Likewise, we have storage capacity we don’t use. (I’m thinking of the fully charged Tesla sitting idle in my garage right now). Innovative companies (see this fascinating article about what Octopus is doing in Texas) are trying to make this happen. Tech companies can help in all kinds of ways on these fronts.

For entrepreneurs 

I anticipate that opposition to new data centers will intensify significantly and slow new projects. Local communities and conservation organizations in areas rife with data centers feel like they have been treated very poorly to date. 

What if some data center developers organize themselves as public benefit corporations (PBCs)? Unlike regular corporations, PBCs are mandated by their charters to prioritize more than outcomes to shareholders. They can, for instance, prioritize environmental outcomes such as clean energy use and conservation without risk of shareholder pushback. They could emphasize positive outcomes for the communities where they plan to build new facilities like training and new jobs. We could finally have an ESG strategy that really delivers. 

The AI revolution is underway and its energy needs are massive. But it might be just the catalyst we need for responsible, climate-, nature-, and community-friendly growth and innovation. I hope all of the diverse players engaged here can collaborate and promote smart development to the benefit of all. 

Onward,