Load growth is not only changing how utilities are planning for the future; it’s also informing the strategies of solar developers that serve them.
According to Boris Schubert, the COO at solar developer Silicon Ranch, rising demand has also led to a higher need for certainty in development, particularly for hyperscalers and other data center developers looking to expand quickly.
The result? An energy-first approach to siting data centers that favors bigger projects that can get megawatts on the grid fast. To meet this demand, Silicon Ranch has embraced “pre-positioning,” partnering with hyperscalers much earlier in the project development timeline. But real acceleration, Schubert said, will also require improving the long-clogged interconnection process, via implementation of FERC Order 2023.
The Tennessee-based Silicon Ranch takes a rare approach to solar development: owning and operating all of the projects it develops and builds, and incorporating agrivoltaics and other regenerative land practices.
“The core of the DNA was, ‘how do you deploy solar in regions where you don’t have [solar renewable energy certificates], or don’t have solar roof program, or [renewable portfolio standards]?” Schubert told Latitude Media. “And from the get go, it was working with the hyperscalers and the rural electricals to create solutions that work in their specific locations. And that is different in Tennessee than it is in Georgia.”
The company was founded in 2011, and has since grown to over 180 projects across 15 states and Canada. Between 2022 and 2023, the company secured over $1 billion in equity capital. Just last month, Silicon Ranch signed a PPA with United Power for a 150-megawatt solar farm in Colorado.
The company has also worked with hyperscalers for years. Back in 2022, for instance, Meta announced seven solar projects in partnership with Silicon Ranch and two utilities in Georgia and Tennessee, together totalling 720 megawatts.
But the tenor of those conversations with hyperscalers is shifting. Proximity to fiber used to be the main driver for siting data centers. Low latency — or the time it takes for data to travel from point A to point B — requires fiber connections, so computing facilities often clustered around those connections. But Schubert said he has seen this change recently; today, the main siting consideration today is energy availability.
Strategies for speed and scale
In early 2024, Silicon Ranch announced a collaboration with the data center developer Tract, which specializes in “master-planned” data center parks. The goal was to combine shovel-ready sites for data centers in Nevada and Utah with utility-scale solar and storage sites. Silicon Ranch is handling both site acquisition and interconnection for these sites.
As the companies outlined at the time, the scale of data center demand necessitates matching it to renewable projects earlier in the development process, with benefits for “certainty and stability.” Speaking roughly a year after the partnership’s announcement, Schubert said Silicon Ranch is fielding “weekly” conversations about doing more of these pre-positioned, energy-first projects.
This is a trend reflected in Google’s recent announcement of its first-of-a-kind partnership with Intersect Power and TPG Rise Climate — an “energy-first” development model that builds data centers in places where clean power is plentiful, and brings the fiber to them. By 2030, the three aim to develop $20 billion worth of renewables and battery storage, co-located with data centers.
Project sizes are growing to meet this demand, Schubert said: “big matters.” While 500-MW or even gigawatt-scale projects have long been deployed in places like Texas, in the Southeastern communities where Silicon Ranch operates, smaller projects are the norm. But that’s beginning to change; for instance, Schubert was coming off of a visit to a 350-megawatt project in Georgia when we spoke.
“Some of the deals you see, it’s about the megawatts; that is the main driver,” he said, adding that as a result Silicon Ranch is working with its partners on getting “deployment timelines for green energy synced up with the appetite.”
But huge projects can’t be developed as fast as small ones. As a result, Silicon Ranch is adapting to the demand by taking a phased approach. A 300-megawatt project, for instance, might be composed of three 100-MW sub-projects, which are each brought online in sequence, “so that nobody has to wait until everything is built to the end, but that you can bring it up in phases.”
“That is a demand, and that is an ask that we are seeing more and more,” Schubert added.
This is an inherent benefit to solar, Schubert pointed out; modularity is built in, making expansion easier than for, say, a nuclear plant. When the priority is to get power online, getting quick access to a chunk of a larger project has value to potential customers.
Renewables developer NextEra made a similar point on its fourth quarter earnings call last month. While the renewables developer is expanding its gas generation via a partnership with GE Vernova in reaction to load growth, CEO John Ketchum emphasized that renewables are still the best option for getting new megawatts onto the grid quickly. New gas-fired generation “won’t be available at scale until 2030,” he said.
The solar interconnection hurdle
Interconnection has become the biggest hurdle for renewables developers. The U.S. interconnection process has long been dysfunctional, leaving queues clogged. One recent report found that the backlog of projects waiting for approval is 2.6 terawatts — more than all installed U.S. power capacity. It takes on average five years for a project to move from request to commercial operations.
The Federal Energy Regulatory Commission’s Order 2023 aims to fix the problem. The rules, which had a compliance deadline in summer 2024, require transmission providers to study projects in batches rather than in an individual first-come, first-served process. They also impose stricter financial readiness and site requirements for customers, in order to discourage speculative projects. However, it’s still early days; Schubert is eager to see how its promise of streamlining the process pans out.
“The key angle that we all yet have to see play out, I think, is how FERC Order 2023 will help us to have high quality interconnection queues,” Schubert said. Today, he added, the barrier to entry to the queue itself “is relatively low,” but the barrier “to high quality projects is relatively high.”
To get more quality projects into the queue, he estimates, would require prioritizing the customer-led and load-led development, and that in turn would require better analysis before projects land in the queue.
“From a regulatory perspective, we are quite confident that once FERC order 2023 has shaken out across the industry, we will still sit there with significant demand and load growth — but we will be way better positioned as the wider power industry,” Schubert said.
In order to improve certainty, Silicon Ranch has its own utility modeling team, to predict utility feedback after proposing a particular project size or location. Using their analysis, Schubert said, “leads to a higher degree of certainty for the hyperscalers” that their projects will “make it through the funnel.”
Editor’s note: This story was corrected on Feb. 11 to clarify that Silicon Ranch currently has over 180 projects in 15 states plus Canada — not 70 projects in 17 states, nationwide.
