Over the Labor Day weekend, one of us stood at a dam with the family, watching an enormous amount of water pour over a structure that has been sitting in that river for decades, and asked a simple question: this thing is already here, already holding back a whole river — why isn't it making any electricity? It turns out that for the overwhelming majority of America's dams, the answer is that it isn't. And that is a surprisingly large, surprisingly cheap opportunity that we are mostly leaving on the table.
We want to talk about an idea that is almost boring in how sensible it is: identifying the dams that are good candidates to be retrofitted for hydroelectric power, and getting the best of them online. Not every dam can or should be converted. But we should study the ones that can be, and move forward on the ones that make sense.
A Power Plant Already Built
Here is the fact that surprises most people. The United States has more than 92,000 dams. Only about three percent of them generate electricity. That leaves roughly 89,000 "non-powered" dams — structures built over the last century for flood control, navigation, irrigation, water supply, and recreation — that hold back rivers every single day without turning any of that energy into power.
According to the Department of Energy and Oak Ridge National Laboratory, which have studied this closely, retrofitting the suitable subset of those non-powered dams could add up to about 12 gigawatts of electricity to the grid — enough to power on the order of nine million homes. That is a meaningful amount of clean, domestic power, and the most expensive, most disruptive part of a hydro project — the dam itself — is, in these cases, already built and paid for.
Why This Is Almost Too Sensible to Ignore
Line up the advantages and it is hard to see the downside.
It is cheaper than building new. The dam, the reservoir, the river-blocking structure — the parts that cost the most and take the longest — already exist. A retrofit adds a powerhouse and turbines to infrastructure that is already there, rather than pouring a massive new structure from scratch.
It is faster to get online. Less construction and, for genuinely low-impact retrofits, a more manageable path to approval mean electricity flowing years sooner than a greenfield power plant of any kind.
It is clean, renewable, and dependable. Hydropower is zero-emission and, unlike wind and solar, it can be dispatched and provides steady baseload generation — exactly the kind of reliable clean power a modern grid needs.
It adds little new ecological harm. This is the part we find most compelling. The reservoir and the dam already reshaped that stretch of river decades ago; the environmental cost was paid long ago. Adding generation to an existing dam avoids the enormous ecological impact of damming a new, free-flowing river. To be clear, a retrofit still deserves real environmental review — fish passage, downstream flows, and local input all matter — but the footprint is a fraction of what a new dam would impose.
And there's a bonus. Beyond the non-powered dams, many of the roughly 2,300 dams that already generate power are running on decades-old turbines. "Uprating" them — swapping in modern, more efficient equipment — squeezes more clean electricity out of hydro plants we already operate. It is one of the easiest clean-energy wins available, and it barely gets discussed.
The Timing Could Not Be Better
We have written about the energy crunch bearing down on this country. The build-out of AI data centers alone is projected to push electricity demand up dramatically over the next few years, and we are already seeing rising energy costs land on families and businesses alike. When demand is surging and prices are climbing, the last thing a sensible country does is ignore power that is cheap, clean, and mostly already constructed.
Twelve gigawatts will not, by itself, solve a national energy shortage — we should be honest about that. But it is real, it is domestic, it is clean, and it comes largely from concrete we have already poured. In an energy crunch, you take the affordable, low-impact megawatts first. You do not leave them sitting in the river.
Being Clear-Eyed About It
We are not pretending this is a magic wand, and we will be honest about the limits. Not every dam is a good candidate — plenty are too small, poorly situated, or structurally unsuited to generation. Some aging dams should be repaired for safety, and a few should actually be removed to restore a river rather than retrofitted. Retrofits still cost money, still take time, and still require genuine environmental and community review. None of that is free.
But here is the thing: the hard analytical work of identifying the strong candidates has already begun. The Department of Energy and its national labs have mapped the potential and flagged the challenges. What is missing is not the research or the technology — it is the will to fund the studies, prioritize the best sites, streamline the permitting for genuinely low-impact retrofits, and actually build them, instead of letting good projects languish for a decade in process.
A Common-Sense Energy Win
This is exactly the kind of pragmatic, un-ideological policy this movement exists to champion. It is clean, renewable energy that an environmentalist can love, delivered at a lower cost and faster timeline that a fiscal conservative can love, without damming a single new river. It does not require anyone to pick a side in the exhausting energy culture war. It just requires competence and follow-through: study the candidates, clear the sensible ones through a streamlined process, upgrade the turbines on the dams already running, and get the power flowing.
We have tens of thousands of dams sitting in our rivers right now doing everything except the one thing that would help most in this moment — making electricity. Let's study the ones that can be retrofitted, uprate the ones already generating, and turn infrastructure we have already built and paid for into the clean, affordable power we badly need. Sometimes the smartest new energy source is the one we already have.