Inside Digital Infrastructure: Spring Lane Capital’s Rob Day on Power Strategy, Microgrids, and What It Takes to Win the AI Race

Step into the full digital infrastructure ecosystem with our Q&A series featuring innovators who are redefining what’s possible across data centers, fiber networks, broadband, and next‑generation connectivity. Hear from investors, lenders, developers, and operators on what’s changing across power, capital, and risk — and what it takes to scale at speed.

Rob Day, Partner and Co-Founder at Spring Lane Capital, joins Sahir Surmeli, Co-chair of Mintz’s Sustainable Energy & Infrastructure Practice, to discuss how the race for power capacity is reshaping investment strategy, project development, and the economics of behind-the-meter generation across the data center sector.

In this Q&A, Rob breaks down why the “bring your own capacity” model is driving a stampede of investor interest into microgrids and grid flexibility and how interconnection bottlenecks and equipment constraints are sorting the market by scale. He explains why he believes the market’s fixation on natural gas is a long-term strategic mistake and what it will take for battery-powered microgrids to unlock behind-the-meter viability for industrial loads. Rob also shares a playbook, drawn from Spring Lane’s investment in Soluna, for how developers can start consuming power today while building the AI infrastructure of tomorrow.

Why the “Bring Your Own Capacity” Mandate Is Reshaping Data Center Investment

Sahir: What’s one change you're seeing in digital infrastructure or data centers that’s materially affecting how you do business today?

Rob: The need for power capacity is already having a huge ripple effect within the investor community. The phrase everyone is using now is "bring your own capacity," the idea that if you want to place a data center, you also need to be able to add that capacity onto the grid. All of these companies are pledging they'll do it.

What that's meant is it has really juiced the microgrid sector. Battery and storage developers, natural gas-fired on-site generation companies, and renewable project developers are all seeing accelerated interest. But the effect extends well beyond data centers themselves. Data center developers and the broader developer community are taking what would ordinarily have been a grid-facing project and instead doing it behind the meter for this customer set. Even companies providing grid flexibility with no direct tie to a data center, such as demand response operators and virtual power plant aggregators, are seeing a surge of investor interest, especially for anything that delivers flexible capacity to utilities.

Some data centers are now tying their power strategies directly to virtual power plant and demand response programs, which is opening up a whole new category of opportunity at the intersection of digital and energy infrastructure.

The flip side is that this is a really marked change, because it has pulled investor interest away from other areas significantly. Having seen a few cycles of this kind of herd behavior among investors, it rarely ultimately ends well. But right now, the herd is stampeding.

Interconnection Queues, Equipment Bottlenecks, and the Scale Divide in Data Center Development

Sahir: Where do you see the greatest friction in this space right now, whether legal, commercial, or operational? And does size create a tension between smaller and mid-size developers versus the players building at massive scale?

Rob: There are several friction points, and scale matters differently in each one. 

Interconnection is first and foremost. The queues are long and getting longer, and it is very tough for anybody to achieve anything close to what is being forecast in terms of both data center build-out and capacity additions. That is going to put a fine point on permitting reform, but there is no silver bullet. This is a physical grid that is very difficult to manage, with a bunch of different stakeholders who are also themselves very hard to manage. 

Scale disparity compounds the problem. Smaller developers have fewer resources to elevate themselves in the queue; bigger corporations are the ones with deeper budgets for stakeholder engagement.

Equipment supply chains are a big holdup, and also size-driven. If you are bigger, you can go cut a deal with the people who have the transformers. If you are smaller, you are having to be super scrappy. I have seen developers adopt unproven transformer manufacturers, get a big failure rate, and then have that affect their ability to get financing.

Geography may be the most underappreciated friction point. I went to a Power and Projects conference recently where everybody viewed data center development as a real estate play. To my mind, it is going to end up being a power play. Data is very fungible; for a lot of applications, latency is not a concern, and for a lot of this market, power costs are going to be king. Everybody is trying to build close to or inside urban cores, which puts pressure on sizing. A few places like Loudoun County, Virginia, offer proximity and scale, but those are few and far between and driving up prices. Something has got to give. People will start realizing what needs to be close to demand and flock to bigger, more flexible data centers further out.

Natural Gas vs. Renewables: Why Long-Term Data Center Power Strategy Is Being Mispriced

Sahir: How realistic is on-site generation and behind-the-meter power with current or near-term technologies for mid-size or larger data centers, and are there strategic pitfalls you see in the current approach? 

Rob: There are a bunch of long-term strategic mistakes being made right now, because everybody is focused on natural gas. You are making a 20-year-plus lived asset and placing a 20-year bet on natural gas prices being consistent. If there is one thing we have never seen in the US natural gas market, it is consistent prices. Everybody kind of fooled themselves because there was a period of relatively low and relatively stable prices, but I am old enough to remember gas popping to $8 a therm, and people making bad strategic decisions at a narrow point in time that then reverted. You are tying your operations to something that is hugely variable.

The thing about renewables is that the fuel costs do not vary, and they match that 20-year timeline. More people need to be thinking about how to purposefully tie renewables to data centers, along with battery storage for bring-your-own capacity, along with tying into demand response programs. There are ways to stage in over time. The role that natural gas plays is probably near-term, portable power: deploy it, then move it to the next data center being built.

Speed to build matters here, too. As an investor in Soluna projects, one of the things I have learned is that crypto mining and AI build-out can work in tandem. It takes a couple of years to build out an AI center but about six months to build out a Bitcoin mining center. So, you start with crypto mining, begin consuming power while you build out the AI on the other half, and then migrate. Right now, everybody is focused on the real estate side. Whoever thinks through the power side of this is going to win.

How Close Battery Microgrids Are to Unlocking Behind-the-Meter Viability for Data Centers 

Sahir: Given the cost curve for batteries and generation, how far away are we from 24-hour behind-the-meter coverage at a cost that pencils out? 

Rob: I do not think we are very far away from it from a battery cost standpoint. Prices continue to drop on the basis of scale and innovation, and because this is behind the meter, you are not having to pay transmission costs or utility margin for the most part. That itself reduces some of the costs.

If you can marry it with more software-driven capacity, such as demand response, ancillary services, and any kind of curtailment ability in your data center, the economics improve further. 

The real problem is that it gets complicated. Right now, people are trying to move very quickly, and they do not want complicated. What will really unlock microgrids behind the meter for industrial purposes is when you can set up all of those things in tandem to create, essentially, an on-site utility: dialing up and down air conditioning, turning on the battery, running on-site rooftop solar, and being tied to the grid, all managed together as one system.

One thing worth understanding: within the microgrid developer community, the word got out that you make more money by serving the grid with your batteries than by serving your on-site customer. The barrier has been more bureaucratic and market structure inertia rather than a cost thing. The costs already work at today's battery prices.

Subscribe To Viewpoints