• THE GRID
Record Power. Still Not Enough.
I pulled up the EIA's capacity tracker this week. One number stopped me cold.
Eighty-six gigawatts.
That's how much new power the U.S. plans to add this year. A record. More than any year in modern history.
Last year we added 53 GW. That was the most since 2002. This year blows past it by 62%.
Solar makes up 51%. Battery storage is 28%. Wind takes 14%. Gas fills the gap.
|
Planned 2026 Capacity
86 GW
|
Added in 2025
53 GW
|
|
|
Battery Storage 2026
24 GW
|
Utility Capex 2026
$259.1B
|
|
Battery storage is the breakout. The EIA expects 24 GW of new batteries this year. Up from 15 GW in 2025. U.S. battery capacity has grown exponentially in five years, with more than 40 GW added.
And utilities are spending to match. S&P Global puts U.S. utility capex at $259.1 billion this year. Up from roughly $200 billion in 2025. The five-year total? Nearly $1.3 trillion.
Crews are pouring concrete. Stringing wire. Building substations. The pace has never been this fast.
So you'd think the power crunch is fading.
It's not.
NTT Data's latest global data center outlook makes the case. The conclusion: power shortages could still slow AI expansion in major markets, particularly in the United States and Europe.
"AI demand is accelerating faster than many parts of the underlying infrastructure system can respond."
Data centers use about 1.5% of global electricity. That sounds small. But they cluster.
In the biggest hubs — Northern Virginia, Dublin, Singapore — one campus can eat 20% to 30% of local power.
We're adding megawatts at a record clip. But demand keeps landing in the same few zip codes. Those zip codes can't absorb it.
The generation exists on paper. Getting it to the right site takes substations, transformers, and transmission lines. Each one takes years to build and permit.
Builders can buy land in weeks. Raise capital in months. But a grid connection? Two years. Three. Sometimes more.
The supply is national. The bottleneck is local.
Europe faces the most strain. NTT Data gives it the highest infrastructure stress score of any region it tracks. The U.S. has more room — but every new AI campus needs hundreds of megawatts.
Record supply. Record demand. And demand is winning. The trade of 2026 is everything that closes the gap — transformers, substations, lines, and the builders behind them.
This is where Elon Musk is housing an AI technology that Jeff Brown believes will help power the next monster IPO on Wall Street.
You see, while everyone was distracted by the SpaceX IPO…
Elon Musk quietly started backing a NEW AI startup that has been called…
"The fastest-growing business in the history of capitalism."
And Jeff will also show you how to claim a stake for as little as $50.
• VOLTAGE
The Grid Workaround
A year ago, Bloom Energy traded below $25.
In June, it touched $351.28. A 52-week high. More than 1,300% in twelve months.
The catalyst? Oracle.
In April, Oracle committed to an initial 1.2 GW of Bloom's solid-oxide fuel cells, with an intention to procure up to 2.8 GW. That's enough to power a mid-size city. Oracle wants them for AI training clusters — the kind that eat megawatts for breakfast. And 1.2 GW is already going in.
Bloom's Q1 backed it up. Revenue hit $751 million — up 130% year-over-year. Product sales surged 208%. Net income reached $70.7 million.
Management raised full-year guidance to $3.4–$3.8 billion. AEP signed a $2.65 billion deal earlier this year. Brookfield committed $25 billion in financing. The market cap crossed $74 billion.
Every hyperscaler needs power now. Not in 2029. Now. And the grid can't deliver fast enough.
Bloom's fuel cells ship on trucks. They make power in weeks. No grid connection. No years-long permit wait. Oracle gets capacity today. Bloom locks in long-term revenue.
The broader fuel cell market tells the same story. Analysts peg it at $16.8 billion this year. By 2034, they see $139 billion. That's a 30% annual growth rate — driven almost entirely by data center demand.
Q2 earnings land July 28. Analysts expect EPS of $0.35 — up 250% from last year.
Fuel cells aren't replacing the grid. They're routing around it. And the companies doing it fastest will own the next decade of power.
• WIRED IN
Three Moves Worth Watching
X-energy is advancing a 6-GW nuclear partnership with UK energy giant Centrica — the biggest international SMR deal on the table. Centrica owns 20% of Britain's operating nuclear fleet. X-energy's Xe-100 reactors could break ground there by decade's end. This is where SMR ambitions go global.
Staying in nuclear: Nano Nuclear Energy just acquired Secured Transportation Services, a nuclear materials hauler with two decades of experience. The microreactor developer can now move its own fuel and reactors. Few companies in the world hold those transport licenses. Going vertical in nuclear is rare — and smart.
On the storage side, T1 Energy (formerly FREYR Battery) is buying KORE Power for $32 million. KORE makes cells for grid-scale battery projects. T1 wants to aim them at AI data center backup power — a fast-growing niche where reliability is everything.
Nuclear going global, nuclear going vertical, and batteries going behind the meter. The smart money is building the supply chain, not waiting for the grid to catch up.
