Episode Overview

Theme:
This episode of Catalyst with Shayle Kann features Shayle in conversation with Nat Bullard, co-founder of Halcyon and renowned energy analyst, as they unpack Nat’s mammoth annual deck of 200+ slides, exploring early 2026’s most compelling energy and decarbonization data trends. Topics span China’s rapid electrification, volatility in oil vs electricity spend, gas turbine and data center booms, and the surging Texas power demand queue. This episode (Part 1 of 2) combines hard numbers, pithy banter, and sharp analysis—essential listening for industry insiders.

Key Discussion Points & Insights

1. China’s Electrification Outpaces the US

• Insight: China’s economy is now substantially more electrified than that of the United States—a major and underappreciated trend.
• Details:
  • Since 1990, the share of electricity in China’s final energy consumption surged from ~7% to 30%; the US has hovered around 20–22% over the same period.

    “China in 1990 is down at what, 7% electricity and then jumped to 30% today. So like it's a very different trajectory.”
    —Shayle Kann [07:09]

  • China’s path is notable for happening at a vast, industrial scale that dwarfs countries like Norway, whose high electrification comes from a smaller, more specialized industrial base.
  • Possible motivations include a drive for energy sovereignty; by expanding electricity generation—especially from domestic coal, renewables, and nuclear—China reduces dependence on energy imports.

    “Electricity is generated within boundaries...and so in that sense, yes, it does provide a lot more control over destiny, less exposure to market forces, to geopolitics, to everything else.”
    —Nat Bullard [09:08–10:20]

2. GDP Share: Electricity vs. Oil Spend

• Insight: GDP spent on oil is volatile and historically higher, while electricity spend has been incredibly stable—raising big questions about the future.
• Details:
  • In 1980 after the oil shock, nearly 9% of global GDP per capita was spent on oil; electricity hovered just over 3%.
  • Electricity spend remains “range bound” at 3–4% of GDP, regardless of economic shocks; oil’s share fluctuates wildly with price and dependence.

    "For all the talk...about spiking prices for electricity...it's really fascinating how range bound it is. We basically spend between 3 and 4% of GDP on electricity and that is that essentially."
    —Nat Bullard [12:09]

  • The crucial question: Will a combination of rising electricity demand (driven by AI/data centers) and efforts to keep oil prices low break electricity’s 50-year “flat line” in GDP share?

3. Gas Turbine Supply Crunch

• Insight: A dramatic under-supply in gas turbines is developing, echoing earlier “booms and busts” in the sector.
• Details:
  • After a peak order boom (~90 GW) in 2001, annual orders crashed and stayed below production for years; now demand far outstrips current annual production capacity (~60 GW), with orders for 2028 exceeding 100 GW.
  • Manufacturers remember past overcapacity and are cautious about expanding—leading to today’s bottleneck as turbine demand explodes.

    “We are the most undersupplied we have ever been. Or at least since the data starts at the beginning of the century.”
    —Shayle Kann [19:08]

  • Broader impacts: Aero-derivative turbines and even jet engine companies pivot to fill grid supply gaps, especially for data centers.

    "Everything is turbine."
    —Shayle Kann [20:18, 27:43]

4. Soaring Gas Power Plant Capital Costs

• Insight: The capital cost for new gas plants—especially combined-cycle—has nearly doubled for later-decade deliveries, driven by supply chain constraints and inflation.
• Details:
  • 2026 deliveries: ~$1,200/kW; 2030–31 deliveries: approaching $2,500/kW.

    “The cost of a combined cycle has doubled. 2026 deliveries ... $1,200 per kilowatt ... projects for 2030, 2031 ... close to $2,500...”
    —Nat Bullard [23:19]

  • Simple cycle plant costs are up 50% ($1,000 → $1,500/kW); reciprocating internal combustion engines already top $2,500–$3,000/kW.
  • Many utilities may still underestimate true future costs due to ongoing inflation, EPC shortages, and unspecified site factors.

    “I do wonder whether they're actually underestimating the total costs...really hard to find EPCs right now.”
    —Shayle Kann [25:27]

5. Tech CapEx: Data Center Buildout Surpasses Historic Booms

• Insight: Tech/data center capital expenditures are now a larger % of US GDP than even the interstate highways, the Apollo Project, or the peak broadband buildout—shaping both grid and infrastructure investment.
• Details:
  • Tech CapEx in 2025: just under 2% of US GDP, compared to peak broadband (~1.2%), Apollo and highways (~0.6%).

    “Tech capex right now is just under 2%. So basically higher than anything else ... to build compute.”
    —Nat Bullard [28:15]

  • Data center CapEx is likely not at its peak; most forecasts expect even higher numbers in coming years.

    “Most of the estimates based on what companies themselves are saying for their estimated capex have a higher number for next year.”
    —Nat Bullard [30:09]

  • These figures don’t even include all the related power/water/transmission investment being driven by the same demand.

6. Texas: Unprecedented Surge in Load Interconnection Requests

• Insight: A “mad rush” for data center power in Texas has caused the interconnection queue for large loads to jump more than 5× in under two years.
• Details:
  • January 2024: ~42 GW of large-load requests; November 2025: 226 GW—compared to current Texas peak load of ~85 GW.

    “Those are stupid high numbers...going from 41 to 226 in Texas alone in two years in the queue.”
    —Shayle Kann [32:42]

  • Many of these sites are speculative; similar to supply-side queues for generation, but now driven by developers seeking to flip sites or lease them.
  • Contrast to other sectors: Unlike hospitals, which are tied to local needs, data center demand is highly fungible/flexible.

    “What I think we're not used to is a demand side interconnection queue that has some of those same speculative elements.”
    —Nat Bullard [34:38]

7. Dueling Forecasts: Texas Power Demand

• Insight: There’s a massive divergence between grid operator ERCOT’s and Transmission Service Providers’ (TSP) forecasts for Texas load growth to 2030—differing by more than the total electricity use of many countries.
• Details:
  • ERCOT’s forecast: ~1,000 TWh by 2030 (100% growth); TSPs: ~1,600 TWh (240% growth). Gap = ~500 TWh, >10% of total US annual demand.

    “So there's a difference between these two forecasts of about 500 terawatt hours. ... more than 10% of all US electricity demand as just the delta...”
    —Shayle Kann [40:34]

  • ERCOT’s forecast discounts demand that is less certain; TSPs forecast based on all requests, incentivized to overstate for capital project justification.

    “ERCOT has the reason for demand, supply balancing and upkeep... the transmission service providers have every incentive to go as big as possible, because that's how they get paid.”
    —Nat Bullard [37:59–40:20]

8. Global Perspective: Data Centers Are Not the Only “Big Demand” Story

• Insight: On a global scale, other sectors—especially industrial electrification and electric vehicles—are still bigger growth drivers for electricity demand than data centers (for now).
• Details:
  • IEA data: 2024–2030, about 30% of worldwide incremental demand comes from industrial electrification; transport and appliances are also ahead of data centers (~8%).

    “Electrification of industry is going to be like 30%... even appliances...space cooling...10%...data centers are right around 8%.”
    —Nat Bullard [43:18]

  • Expect these rankings to shift, possibly rapidly, as data center boom accelerates worldwide and electricity supply becomes more constrained or zero-sum.

    “I will be very interested to see what this print looks like a year from now, two years from now, three years from now.”
    —Nat Bullard [43:18]

Notable Quotes & Memorable Moments

• "Everything in China is bigger when it comes to energy... it's a huge industrial economy... also moving at a much more rapid pace."
  —Nat Bullard [07:55]

• “For all the talk...about spiking prices for electricity...how range bound it is. We basically spend between 3 and 4% of GDP on electricity and that is that essentially.”
  —Nat Bullard [12:09]

• “If you're in charge of building gas turbines now, you probably have the institutional memory of the early 2000s.”
  —Nat Bullard [17:08]

• “Every new data center is a new electrified industry facility that isn’t going to happen probably.”
  —Shayle Kann [45:16]

• “No one knows anything. That’s my new slide for today’s market.”
  —Nat Bullard [37:59]

Timestamps for Key Segments

• [05:03] — Introduction to China’s electrification leadership
• [10:20] — Why electrification = energy sovereignty for China
• [12:09] — The historic steadiness of electricity’s GDP share vs oil’s booms/busts
• [17:08] — Gas turbine market history and current supply shortage
• [23:19] — Capital cost escalation in gas plant construction
• [28:15] — Data center (“tech”) CapEx vs historic US infrastructure booms
• [32:42] — Explosive growth of large-load interconnection requests in Texas
• [37:59] — ERCOT vs TSPs: wildly different Texas electricity demand forecasts
• [43:18] — Global: Industry leads electricity demand growth, not data centers (yet)

Summary

This episode unpacks early 2026’s defining energy transition trends with wit and verve. China’s electrification push quietly surpasses the US, historic volatility in oil prices and the unshakable stability of electricity spending are put in context, and the unprecedented surge in gas turbine demand reveals grid vulnerabilities. The US, led by data center demand, is in the throes of a power plant and data infrastructure buildup not seen since highway and space race days, with Texas as a case study in power planning uncertainty. Expect Part 2 for further slides, sectors, and speculation.