New Orleans Becomes Ground Zero for a Nuclear Reactor Left for Dead in 2017

Nine years ago, the mPower reactor was a corpse. Babcock & Wilcox had poured federal money and engineering hours into the small modular design, collected more than $100 million in support from Washington, and then watched the financing dry up in 2017. The project went into the drawer of ideas whose time had not yet come. This week, Louisiana Economic Development announced that a startup called Applied Atomics is pulling it back out, and doing it from a headquarters in the Bywater.

The numbers are modest by the standards of the megaprojects Louisiana has grown used to announcing. Applied Atomics plans to invest $8.2 million in an 18,500-square-foot headquarters and technical center, and it expects to create 41 jobs averaging $121,000 annually, Louisiana Economic Development announced Wednesday. Compare that to the billions flowing into LNG terminals along the river or the gigawatt-scale data center campus Meta is building in Richland Parish, and this looks like a rounding error. It is not. What Applied Atomics is buying, through rights secured from BWX Technologies earlier this year, is a fully engineered 195-megawatt reactor module that already has a decade of licensing work behind it. That is the kind of head start that normally costs a startup years and hundreds of millions of dollars it does not have.

The company's timing is the real story. Ben Kellie, Applied Atomics' CEO and co-founder, put it plainly: "There's a lot of power demand and power needs, but there are also a lot of opportunities for low carbon power here with data centers coming in." He is not wrong, and he is not alone in reading the market that way. Louisiana's own Nuclear Strategic Framework, released this year, names data centers and other large industrial loads as the anchor customers the state needs to make new nuclear generation pencil out. Meta has already put out a solicitation for between 1 and 4 gigawatts of new nuclear capacity to serve its Richland Parish campus starting in the early 2030s. Stack a handful of mPower modules together and you get exactly the kind of hundreds-of-megawatts-to-gigawatt block that a hyperscale data center operator wants, with the added selling point that it runs around the clock and does not depend on wind blowing or sun shining.

John Fleig, chief innovation officer at Louisiana Economic Development, framed the deal as a race against the clock rather than a slow-build economic development win. "I think it's a speed-to-market thing for start-ups in this space. They've got to prove that they can deploy their technology and get rolling quickly, they're on the clock," Fleig said. He also tied it to the state's longer game: "In Louisiana, if we're going to continue to have an energy economy for the next 100 years, we have to make sure that we have new players in place, whether it's these small modular reactors." That is the right instinct. Louisiana built its economy on being the place that moves and refines the energy the rest of the country depends on. Nuclear, especially small modular reactors sized to serve a single industrial customer, is the next version of that same trade, and there is no reason that work should happen anywhere but here.

Nobody should mistake the New Orleans facility for a power plant. There is no reactor going into the Bywater, no radioactive material on site. What Applied Atomics is building is engineering offices, a component assembly operation and a training center, plus an outdoor rig called the "Plexiplant" that will use steam and electricity to simulate plant operations for equipment testing and operator training, expected to come online in early 2027. That is the unglamorous, necessary scaffolding that has to exist before anyone pours concrete for an actual reactor. Fleig's broader pitch, that Louisiana has to be "open and welcoming" to companies like this and support them with capital, coaching and connections to large load users, whether a data center or an industrial operation along the river, is the correct posture for a state sitting on this much existing energy infrastructure and this much appetite from Big Tech for firm power.

The skepticism, and there should be some, is that mPower already failed once, and a change of ownership does not repeal the physics or the financing challenges that sank it in 2017. Small modular reactors have a long history of slipping timelines and blowing budgets across the industry, not just at Babcock & Wilcox. What is different this time is the buyer side of the market. Data center operators are not waiting on federal loan guarantees, they are writing checks for firm, carbon-free power right now, and that changes the math for everyone chasing this technology. The Plexiplant fires up early next year. That is the first real test of whether Applied Atomics can move at the speed Fleig says the industry demands.