Google’s accelerated urgency for clean, firm power

Google needs new clean baseload power, and fast. And it’s not the only one.

• The top line: Google has quietly entered a new phase in its approach to energy, looking for options beyond solar and wind — and actively trying to play a role in getting new technologies to commercialization. As the need for clean baseload power becomes more urgent, the company plans to develop new business models and aggregate demand for technologies like advanced nuclear or long-duration energy storage via its partnership with Microsoft and steel producer Nucor. 
• The nuts and bolts: The trio is in the midst of reviewing submissions to its March request for information, which looks to bolster the development of first-of-a-kind and early commercial carbon-free energy projects. The RFI aims to zero in on just a few projects in just a few regions of the United States, namely those where Google, Microsoft, and Nucor have overlapping loads and electricity needs. It’s technology-agnostic, provided projects are both carbon-free and can supply “high-capacity factor,” i.e. essentially firm dispatchable base load.
• The current take: Maud Texier, Google’s global director of clean energy and decarbonization, said that advanced clean energy technologies have been on the Google radar for a while. But recently, the tenor has changed. “There has been a bit of an acceleration in terms of the sense of urgency and actually acquiring a lot of new clean electricity,” she told Latitude Media.

Google is aiming to run entirely on carbon-free energy by 2030. As one of the original drivers of the power purchase agreement, the company has signed contracts for at least 10 gigawatts of clean energy generation capacity since 2010. 

But as large tech companies double down on artificial intelligence and the data centers needed to power it, the path to 2030 is getting more complicated. Brian Janous, co-founder of Cloverleaf Infrastructure and former VP of energy for Microsoft, said at a recent Latitude Media event that power is now “absolutely” the biggest constraint for scaling AI.

And this new “urgency” to acquire clean baseload power isn’t just being felt within Google. Microsoft’s sustainability report last week showed a 29.1% increase in emissions since 2020 — and a 50% increase in energy intensity — that it attributes to the growth of data centers amid the AI boom. (Google hasn’t yet released its 2024 sustainability report, so it remains to be seen whether the company’s own work in AI has led to parallel emissions and energy intensity upswings.)

Industrial giants like Nucor also need more clean, firm power. The move to onshore and decarbonize heavy industries like steel requires deep investments beyond variable wind and solar.

But demand has been diffuse. And given the new urgency, and fast-approaching 2030 climate goals, Texier found that coordination was needed. 

“What we realized is that going at those technologies by ourselves was not going to be sufficient to reach the scale that we and others need by 2030,” she said. “A lot of those advanced clean energy technologies have, for sure, a bit of technology risk remaining. But…the go- to-market and the commercialization pathways…are really on the critical path right now for that success.” 

The RFI, announced in March, is an attempt to send a bigger demand signal to investors, developers, and utilities “that there will be off-takers for those types of solutions,” Texier said. They also aim to bring that same coordination to regulators and combat market design limitations on getting technologies recognized for their value.

At last month’s BNEF summit, which took place just days after the RFI closed, representatives from Google, Microsoft, and Nucor presented the initial findings. Of the 208 companies that responded to the RFI, 58 were long-duration energy storage projects and 48 were advanced nuclear. The remainder included next-generation geothermal, clean hydrogen, power, and carbon capture.

(Back in November, Texier told Latitude Media that geothermal energy could follow a similar commercialization trajectory to solar, in light of its first geothermal-powered data centers going live in Nevada.)

The team is currently in the process of whittling down the submissions to a short selection, and then will conduct more due diligence on the technologies. The trio’s goal is to sign offtake agreements with the chosen projects, and they anticipate offering additional help as needed.

But how to structure the contracts?

In reviewing those initial findings, the companies found that a lot of the potential projects will require “creative financing structures.” But what precisely that looks like is still an open question. 

Texier’s colleague Michelle Chang, Google’s decarbonization technologies program manager, said at BNEF that that could include upfront sculpting of payments, risk-sharing between various project stakeholders, and potentially government support as well. 

Texier said that in some regions these new contracts will look similar to a traditional PPA. But the “innovative tariff structures” needed in other regions are brand new, because utilities currently tend to provide incentives to just solar or wind.

As they work to develop these new commercial structures, she added, a primary goal is to de-risk projects through customization of the contracts themselves — for instance, via performance guarantees tacked onto a traditional PPA . 

“In my experience at least, performance guarantees have a certain threshold,” Texier said. “If you look at things like geothermal or advanced nuclear, there's definitely a little bit more risk from a resource perspective or construction perspective. So we need to engage…earlier in the project development value chain with those players to see how we can [take on] some of those risks, and maybe in exchange get more commercial upside.”