What will power the grid in 2035? The race is wide open

AI’s insatiable demand for energy has prompted technology companies to search for new sources of energy – a search that has fueled competition and investment in fusion and fission startups.

For many, natural gas is the easy solution for 24/7 base load power. It has been tested, inexpensive, and widely available. But the war in the Middle East has exposed the vulnerable supply chain after Iranian drone attacks destroyed much of Qatar’s natural gas infrastructure, a major exporter. At the same time, high demand has created a long waiting list for gas turbines, and today’s orders will likely not be filled until the early 2030s.

These delays pose a risk not only to technology companies, but also to the natural gas industry itself.

In the United States, 40% of the natural gas consumed today goes to electricity generation. By the time the turbine shortage subsides, the industry could be filled with a new crop of competitors. Both small modular nuclear reactor (SMR) startups and fusion energy startups plan to start connecting their first commercial power plants to the grid in the next five to seven years, about as long as it takes to source parts for a new natural gas power plant.

Nuclear threat

SMR startups may have the best chance of going without natural gas-fired power plants. In many cases, this technology modifies existing fission reactor designs, but the basic physics has proven its feasibility and has been widely used for decades.

Many SMR companies aim to have the reactors up and running before the contract expires. Kairos Power, which counts Google as a future customer, is one of them. The company has received approval to build its Hermes 2 experimental reactor in 2024, and construction is well underway. Oklo, which merged with blank check company Sam Altman in 2024, is targeting 2028 for its first commercial operations, according to its annual report.

Others hope to follow in a few years. X-energy, which counts Amazon as an investor, is aiming for the early 2030s, while Bill Gates’ TerraPower, which has a deal with Meta, plans to begin commercial operations in 2030.

To replace natural gas as the preferred source of generation, SMEs will need to scale quickly, achieving the economies of scale on which their business models depend. It won’t be easy. But technology companies seem confident enough that they are either investing in startups or signing deals with them for gigawatts worth of power.

Fusion timeline

Other tech companies are bracing for merger power. Although not as proven as fission, nuclear fusion promises to provide large amounts of energy using only seawater as fuel.

Fusion startups are also targeting the early 2030s — or earlier — to deploy their first reactors. Fusion strength

One leading company, Commonwealth Fusion Systems, is on track to flip the switch on its experimental reactor next year. Its first commercial reactor, the 400-megawatt Arc, is expected to begin generating power in Virginia in the early 2030s.

Another startup, a relative newcomer, hopes to begin building a grid-scale power plant in 2030. Inertia Enterprises has based its technology on the reactor design used by the National Ignition Facility, which was the first to prove that controlled nuclear fusion reactions can generate more energy than they consume.

But Helion may have the most aggressive schedule of them all. The Sam Altman-backed startup is racing to build Orion, the first commercial-scale power plant, by 2028 to supply electricity to Microsoft. The company is also reportedly in talks with OpenAI to deliver up to 5 GW by 2030 and 50 GW by 2035. To reach those numbers, Helion would have to build 800 reactors by the end of the decade and another 7,200 reactors in the five years after that.

If the startup can deliver energy in such quantities, it will rewrite the entire energy market. Last year, the United States added 63 gigawatts of new generating capacity across all sources. If Helion can build nearly 10 gigawatts of new capacity each year, the company alone will add more capacity than the entire natural gas industry did last year.

Price problem

The challenge facing all of these companies – including gas turbine manufacturers – is cost.

SMR startups rely on mass manufacturing to reduce costs, but this hypothesis has not yet been proven. Today, nuclear power is one of the most expensive forms of new generating capacity at about $170 per megawatt hour, according to Lazard. Fusion faces a broadly similar challenge, though it faces more unknowns. Some experts expect that the cost of one megawatt of fusion power plant could be about $150 initially.

Meanwhile, new natural gas-fired primary power plants cost about $107 per megawatt hour per Lazard, though prices have been trending higher in recent years, which could put them on a collision course with both new fission and fusion reactors.

But they may all be undermined by renewables coupled with batteries.

Wind and solar energy costs have fallen sharply over the past decade. Wind power seems to have reached a plateau in recent years, but solar prices continue to decline with no signs of stopping. Batteries have also gotten cheaper over the years, to the point where grids have installed massive amounts of them — 58 gigawatt-hours last year. Even without subsidies, prices for solar power coupled with batteries range from $50 to $130 per megawatt hour, which overlaps with fusion, fission and natural gas.

All of these numbers are relative to current chemical-derived battery technology for electric vehicles. Newer designs that directly target network connections could bring prices down even further. For example, Forum Energy recently signed a deal to supply electricity to Google from a 30 gigawatt-hour iron-air battery. Others, XL batteries, can reuse old oil tanks to store cheap organic liquids – battery size is limited only by the size and number of tanks.

Because these new batteries avoid using critical metals such as lithium, cobalt or nickel, they promise to reduce the cost of long-term energy storage so dramatically that anything else is difficult to deliver.