Sign Up for News Updates
I was sifting through a stack of questions piled up in my email from the holiday break asking about 2026 predictions. “What will be the big energy stories?” a reporter I was meeting with asked. Another friend asked me, “Will the electric grid fail as disastrously as some are predicting?” And not unexpectedly, a colleague in the industry asked me, “What’s up ahead for the propane industry in 2026?”
I’ll admit that my crystal ball can be cloudy at times, but I do have some ideas of where I think things are headed. At the same time, I’m remembering the caution Peter Drucker provided about predicting when he said, “Trying to predict the future is like trying to drive down a country road at night with no lights while looking out the back window.”
I’ve decided to leave true predicting to the weather forecasters. They never seem to get in trouble for not getting it right, and I want to avoid any blowback for saying something crazy, so instead of predicting, I’ll simply provide a point-of-view on the energy and power landscape for 2026 using a few questions from my inbox.
Q1: Will The U.S. Electric Grid Fail?
Yes, it’s very likely to do so in isolated regions. Strain on the grid due to increased demand load will be one factor but so too will severe weather crashing into aging infrastructure.
In late 2025, a massive electrical substation fire in San Francisco caused a power outage that affected roughly 130,000 customers, disrupting the city for days during peak holiday shopping. As traffic lights went dark and congestion worsened, Waymo’s driverless cars in the affected areas became confused, parked in the middle of roads while awaiting instructions. The company ultimately suspended service, leaving tow truck drivers busy hauling the stranded vehicles across the city.
This incident may be amusing to anyone who wasn’t involved but make no mistake, while power failures may not be disasters in-and-of-themselves, disasters will come. Severe weather, wildfires, even earthquakes are the most obvious culprits, and California always seems to be ground zero.
The state would do well by “islanding” its grid into small sections to reduce cascade-failure probabilities. California has strung together approximately 25,000 miles of transmission lines and 240,000 miles of distribution lines, of which approximately 147,000 miles are overhead. To protect from severe weather, the Public Utilities Commission estimates that burying the state’s overhead lines would cost ratepayers $559 billion, but that price tag is likely a better investment than continually dealing with blackouts and exasperated customers who pay the second highest rates in the nation for power (Hawaii is #1).
Q2: What Issue Will Affect the Grid Most?
Transformers. Transformers step electric current up and down between circuits. The average lead time to get a new one is now around 120 weeks and climbing. Transformer demand rose by 274% from 2019 to 2025 in concert with growing electricity demand. The same issue plagues gas turbine generators; GE Vernova has an 80-gigawatt backlog that stretches until the end of the decade. Without these heavy equipment pieces, grand designs for grid build-out are impossible.
Q3: What’s the “Venezuela Effect” for Energy?
Mixed. Venezuela claims to have more than 300 billion barrels in the ground, roughly 17% of all global oil reserves, but it only produces about 900,000 barrels a day. The volume implications to oil markets and refineries are not substantial, especially for the U.S. At the end of last year, U.S. imports from Venezuela were just 135,000 barrels a day.
Most reports tell us that U.S. refineries could take in as much as 1 million barrels a day from the country, so upside potential does exist. However, returning Venezuela’s crude production to 3M barrels of oil a day as was done in the past is estimated to require more than a decade of work and nearly $200B in investment. That means we’ll see only incremental effects on global oil prices.
Q4. Will AI Data Center Spending & Grid Strain Continue?
Probably. The AI race will continue to expand as big tech moves on from large language models and training to more sophisticated models that will manage tasks, make decisions, and negotiate deals machine-to-machine. At the Consumer Electronics Show in early January, NVIDIA CEO Jensen Huang said, “The ChatGPT moment for physical AI is here — when machines begin to understand, reason, and act in the real world.”
All of this next-generation AI movement tells me that data centers are going to stay hot. At the same time, they’re going to have trouble. Today, the total number of planned data centers is 2,733. The number under construction, however, is 624. A variety of challenges complicate construction, not the least of which are angry neighbors revolting against the development of these mammoth facilities in their neighborhoods.
The other big constraint is connecting to the grid, which can take years to approve. This is where Bring Your Own Power (B.Y.O.P.) can make a difference. Propane-powered generators made by companies like Rolls Royce can deliver 2.5 MW of co-located power. Eight strung together can provide firm power to a 20MW data center. Designers can add solar and battery energy storage systems to an on-site microgrid to avoid any dependence on traditional grid hookup. With propane as the base power, such a system would be much cleaner than one powered by diesel.
Natural gas will be the dominant fuel for power generation to meet new electric demand in the U.S., while LNG will serve that role worldwide. In 2025, the U.S. exported a record 111 million metric tons of liquefied natural gas, making it the world’s largest LNG supplier. Europe was the largest buyer as it continued replacing Russian gas.
Q5. What Will Be the Breakout Energy Story?
Several candidates. Nuclear, for example, is going to be talked about more with small modular reactors leading the way. The Nuclear Regulatory Commission approved license renewals for 13 reactors last year, and reactor stations like Holtec Palisades in Michigan are being brought back online.
Fusion is another topic we’re likely to hear more about. Skeptics like to say, “fusion is the energy of the future and always will be,” but that doesn’t ring as true as it once did. DOE is investing more into this technology and private investment is expanding. The fusion industry raised $2.64 billion in funding in the 12 months leading to July 2025, so the tokamak may be back and more visible than ever in 2026.
Virtual Power Plants (VPPs) are a solid contender as well. VPPs are comprised of hundreds or thousands of households and businesses that pool potential assets like thermostats, electric vehicles, appliances, batteries, and solar arrays. These devices can be flexibly charged, discharged, or managed to meet grid needs and in so doing, can provide many of the same energy services as a traditional power plant. As more homeowners adopt battery energy storage systems through services like Base Power, I can imagine utilities putting more emphasis on developing these into local VPPs, especially because they have yet to bring utility-scale battery energy storage to its full potential.
Geothermal is a long-shot contender. I doubt that it will be a breakthrough story because it is still a niche technology. It is getting attention, though. At the start of the 2020s, less than $2 billion flowed into geothermal annually. By 2030, that figure could hit nearly $9 billion.
If one thing is certain, it’s that uncertainty will continue to cloud the future of energy. That’s my one 2026 prediction you can take to the bank.