Open Circuit: "Grid Utilization vs Expansion: The 100 GW Debate"

Podcast: Open Circuit
Host: Stephen Lacey, Latitude Media
Guests: Caroline Golan (Chief Growth and Policy Officer at NRG, ex-Google), Brian Janis (Founder and CCO of Cloverleaf Infrastructure, ex-Microsoft)
Date: March 27, 2026
Theme: Exploring whether the clean energy transition should focus on maximizing the existing electric grid’s utilization or building new infrastructure, in light of skyrocketing power demand (especially from AI and data centers), and debating the findings from a major Brattle Group report suggesting up to 100 GW of underutilized capacity.

Episode Overview

This episode dives deep into a central tension in the energy transition: should we “build our way out” of surging electricity demand by investing in more transmission, generation, and grid expansion, or can we address much of the need by utilizing and unlocking latent capacity already built into the grid? The conversation is anchored by a new Brattle Report that claims the U.S. grid is only used about half the time—implying massive room to grow by boosting utilization rather than only expanding physical infrastructure.

The hosts and guests discuss current challenges, what it really takes to bring new loads online, local versus global planning failures, and how shifting the conversation can benefit both tech and communities. They probe regulatory, financial, technical, and social factors influencing this moment, and conclude with a “real or vibe?” take on current energy trends.

Key Discussion Points and Insights

1. Energy Demand Surges; AI and Data Centers as Driving Forces ([07:10]–[10:35])

• Brian Janis on Then and Now: Ten years ago, connecting a 100 MW data center was straightforward, with excess grid capacity everywhere and the main focus being on energy (MWh) procurement, not capacity. Now, the explosive demand from AI means capacity—not energy—is the real constraint.

  “When we were building out the cloud, there was excess capacity everywhere… no one was really thinking about capacity enablement, time to market, resource adequacy. ...By mid-decade it was clear we were going to have a problem.” — Brian Janis [07:40]

• The rapid timeline disconnect between tech and utility sectors: chips/planning move on 1-2 year cycles, while power infrastructure is built on 30-year cycles.

  “You plan power over 30-year cycles. The two just never matched.” — Caroline Golan [09:32]

2. Failure of Global Approaches and Rise of Local Impacts ([10:45]–[13:18])

• Major tech companies framed energy procurement as a global, sustainability, and innovation conversation—and failed to manage or communicate their local impacts.

• Shift over past 2–3 years to recognizing the profound local effect of data center load, and a need for tech companies to collaborate with the communities their infrastructure affects.

  “…we as a tech industry failed to get the local conversation right… we've been stuck between these two stratospheres… I’m hopeful the tip of the spear is going to be more about the local community, more about the local power impacts…” — Caroline Golan [12:11]

3. Unlocking Capacity at the Local Level ([13:38]–[15:47])

• Janis describes Cloverleaf’s approach: kitchen-table politics and engaging directly with local communities to show the slower but deeper benefits of AI/data center projects compared to other industrial development.

• Many rural communities prefer data centers as the “least disruptive” major investment; education is key.

  “A data center is perfect. It brings jobs, but not too many jobs that we have to build a new high school… The high school we have just gets better.” — Brian Janis, quoting a rural county official [14:21]

4. The Brattle Report: Are We Really at Capacity, or Squandering What We Have? ([16:36]–[21:59])

• Key stat: U.S. grid is utilized only ~50% of the time; better utilization could unlock 100 GW of capacity and save $100 billion over a decade.

• Context: U.S. grid built to serve rare peak moments; assets are idle much of the time.

• Panelists agree the potential is real but emphasize significant barriers—regulatory, technical, and market—to capturing the theoretical benefits the report identifies.

  “Our distribution and transmission system is underutilized… we are not putting in place the regulations or policy structures to think of [the grid] as an asset that needs to be delivering speed to market… needs to be utilized above 60, 70%.” — Caroline Golan [19:37]

5. "Is Utilization Thinking Too Small?" and Utility Incentives ([24:22]–[27:32])

• Pushback exists: Some argue “obsessing” over grid utilization is “small ball” when historic, bold reform/transmission builds are needed.

• Utilities are incentivized to build new assets (more capital expenditure = more profit), but optimizing utilization can allow them to serve new load (and create future expansion opportunities) faster.

  “If your answer to every new customer is… build a combined cycle plant… upgrade a 500kV line, your answer is always going to be 5 years out… And that’s not good enough for this market.” — Brian Janis [26:23]

• Tech sector’s focus should shift: Use hyperscaler load growth as a catalyst to build a better grid for everyone, not just chase “hype” demand figures.

6. Barriers to Change: Planning, Market Signals, and the "Frankenstein Grid" Risk ([28:29]–[34:16])

• Most interconnection solutions haven’t fundamentally changed in decades.

• Rapid, patchwork, “behind-the-meter” solutions may create a “Frankenstein grid” (uncoordinated, inefficient, capital-wasting).

• Many mega-projects may not be financeable because hyperscalers don’t (or can’t) sign the twenty-year offtake contracts needed to anchor such infrastructure.

  “Not to say none will get built, but it’s a lot harder than people think… it’s really hard to build big things in this country.” — Brian Janis [32:34]

7. Distributed Resources (DERs), Virtual Power Plants (VPPs), and Market Design ([35:58]–[41:58])

• Critique: Not all “unlocked” grid capacity is interchangeable—load must be matched locally.

• The lack of clear market signals, planning sophistication, and regulatory support for DERs and VPPs is a barrier—this is a “market design” and “visibility” problem, not a technical one.

  “There’s a difference between real electrons and paper electrons… part of the problem… right now is there are real electrons out there, but they don’t exist on paper because of the way we do load planning, the way we do resource adequacy…” — Caroline Golan [38:22]

8. “Real or Vibes?”: Quick Takes on Energy Headlines ([46:07]–[54:02])

Is Coal “Back”? [46:19]

• Vibes—Coal is hanging on longer at some sites, but no new coal is really being built.

Golden Age of Gas? [46:50]

• Gas is expanding, especially for short-term, peaking, and flexibility—but filling with renewables where possible is preferable.

Solar and Batteries Aren’t “Real” Capacity? [48:55]

• Disagree—Recent projects (Google's Minnesota site, OpenAI Oracle Stargate) rely heavily on renewables for true capacity and flexibility.

Small Modular Reactors (SMRs) are the Only Solution? [49:56]

• SMRs are a piece of the puzzle, but not essential or the only option.

LCOE (Levelized Cost of Energy) Irrelevant Now? [50:39]

• Cost still matters, especially for non-hyperscaler customers. Speed is key for hyperscalers, so they’ll pay more for “bridge” capacity; but 30-year, high-cost contracts are not sustainable.

  “Speed is the imperative… where [companies are] price-insensitive is around short-term capacity. But… they’re not going to be signing up for $150/MWh power for 30 years.” — Brian Janis [51:04]

Have We Lost the Ability to Build Big? [54:11]

• U.S. has lost the “muscle” to build at scale—workforce skills and permitting friction are big obstacles.
• Capital isn’t the issue; it’s the lack of skilled labor.

Notable Quotes

• On the grid’s promise:

  “If you add load in the right places at the right times or shift demand… you can spread those fixed costs across more usage and that puts downward pressure on rates.” — Stephen Lacy, [16:36]

• On cost pressures:

  “With the current capital cost and inflation impacts… there is no way, if we keep utilization the same, that we’re not going to drive rates up… the only answer to affordability is improving utilization.” — Brian Janis [19:03]

• On workforce constraints:

  “If there’s a will, there’s a way. I would say if there’s a workforce, there’s a way. …Capital’s not a problem, but you don’t have enough skilled labor out there that actually knows how to do a lot of this.” — Caroline Golan [55:01]

Key Timestamps

• [07:10] – The market evolution: from excess capacity to acute constraints
• [10:45] – Tech companies’ global vs. local disconnect and planning failure
• [13:38] – The role of communities in project siting and benefits
• [16:36] – The Brattle report and the debate over grid utilization vs. expansion
• [19:17] – Regulatory and cost pressures; who pays for grid upgrades?
• [24:22] – Critiques: is focusing on utilization “small ball”?
• [28:29] – The risk of a “Frankenstein grid” and limits of mega-projects
• [35:58] – DERs, VPPs, and why regulatory/market design matters
• [46:07] – "Real or Vibes?"—rapid fire on energy industry narratives

Tone and Takeaways

The episode is fast-paced, frank, and laced with insider banter. The hosts and guests leverage deep careers at the intersection of tech and energy to demystify both hype and hidden complexities in “the 100 GW debate.”
Their consensus:

• The U.S. grid is underutilized and there’s vast opportunity in smarter capacity management.
• New builds are inevitable, but a blunt build-everything approach is wasteful and risky.
• True progress requires regulatory, market, and planning reform, as well as a renewed focus on workforce development and community partnerships.

Bottom line: The “right answer” is a blend of smart utilization and selective expansion. Focusing on utilization is not “small ball”—it’s essential to affordability and avoiding grid waste.