Inside Glen Canyon Dam’s Dead Pool Problem

Part 2 of The Colorado River Story.

In the first piece I told you Lake Powell is sitting about 34 feet above the line where Glen Canyon Dam stops making power. This one’s about what’s underneath that line, because that’s where the story gets genuinely bad, and it’s the part almost nobody outside a handful of water nerds and environmental groups is actually talking about.

Short version: the dam has a backup way to pass water downstream once the lake gets too low to run the turbines. That backup was never built to run full-time. When they tested it a couple years back, it started chewing itself up. And the government’s own engineers wrote the problem down in a memo that basically nobody has read.

Let me walk you through it, because once you understand how this thing works you’ll never look at a “Lake Powell hits dead pool” headline the same way again.

There Are Two “Dead” Lines, Not One

Most headlines toss around “dead pool” like it’s a single cliff you drive off. It isn’t. There are two different critical levels at Glen Canyon Dam, and mixing them up is how people either freak out about the wrong thing or blow the whole thing off.

The first line is minimum power pool, 3,490 feet. That’s the level where the water sitting over the dam’s power intakes, the penstocks, gets too shallow to run the turbines safely. When the water above those intakes drops too low, they start pulling air down along with the water, same as the little whirlpool that spins up when your sink drains. That air gets dragged into the turbines and forms bubbles, and when the bubbles collapse they hit with enough force to punch pits into steel. That’s cavitation, and it’ll wreck a turbine in a hurry. So at 3,490 the eight generators have to shut off. No more hydropower. We’re about 34 feet above that right now.

The second line is true dead pool, 3,370 feet, another 120 feet down. That’s the elevation of the very last holes in the dam that can pass water by gravity, the river outlet works. Drop below 3,370 and the surface of the lake sits under every outlet in the structure. Nothing goes downstream on its own after that. The river, for all practical purposes, stops passing through the dam. We’re about 154 feet above that one.

So here’s how the levels stack up:

Look at that gap between 3,490 and 3,370. In that 120-foot band the turbines are dead and the only way to get water out of Lake Powell and down to Arizona, Nevada, California, and Mexico is those outlet works. That band is the whole ballgame, and that’s exactly where the trouble lives.

The Backup Pipes

The river outlet works, everybody calls them the ROWs, are four steel pipes, each 8 feet across, fed by two intakes in the back face of the dam. Wide open, all four together can move about 15,000 cubic feet of water per second.

They were built as a backup. A role player, not the starter. Their whole job was short stuff: flood-control releases when the reservoir got too full, the occasional experimental “high flow” burst to push sediment down into the Grand Canyon, and moving water when the powerplant was offline for a stretch. Quick jobs, then back to sleep.

That sediment job is worth a second, because it’s part of why this dam is such a headache. The Colorado carries a staggering load of silt, and Glen Canyon Dam stops it cold, so all that sediment piles up in the bottom of Lake Powell instead of rolling on down through the Grand Canyon like it did for millions of years. Down in the canyon, the beaches and sandbars that fish and native habitat and river-runners depend on slowly wash away with nothing to replace them. So every few years Reclamation runs a big deliberate release, a High Flow Experiment, to kick some sediment downstream and rebuild those beaches. It’s one of the few times they actually open up the outlet works and run real water through them. Keep that in mind, because it’s how we found out the pipes have a problem.

What the ROWs were never designed to do is run continuously, year-round, for months or years on end as the sole way to keep water flowing to 30-some million people downstream. Nobody drawing up this dam in the 1950s figured we’d ever be dry enough, or stuck enough, to lean on these pipes as the main event. But that’s exactly the corner the drought is backing us into.

And there are two separate problems with counting on them, which the environmental groups have been yelling about for years.

One is just size. Fifteen thousand cubic feet per second sounds like a lot until you remember the lower states plus Mexico are owed somewhere around 9 million acre-feet a year. Cramming that entire delivery through four backup pipes is a tight fit in the best case, and it gets worse as the lake drops, because a lower reservoir means less pressure pushing the water through. Reclamation has even said the outlet works may be functionally useless somewhere around elevation 3,394, which is above the actual 3,370 dead pool mark. Meaning the river could effectively quit passing usable water before the lake ever technically hits the number everybody points at.

The second problem is the one that should really grab you.

They Ran the Backup, and It Started Eating Itself

In April 2023, Reclamation put the outlet works to real use. They ran one of those High Flow Experiments, pushing around 3,500 cubic feet per second through the pipes for about 72 hours to move sediment and rebuild sandbars down in the Grand Canyon. Nothing exotic on paper.

When they went in and inspected afterward, they found cavitation damage inside the pipes. The same bubble-implosion process that threatens the turbines had started pitting and eroding the steel and the protective coating inside the outlet works themselves. And here’s the ugly twist: the physics gets worse the lower the lake drops. So the exact conditions where you’d be forced to rely on these pipes are the exact conditions where they’re most likely to tear up.

This isn’t even the first time. Cavitation and coating damage in the outlet works was first spotted way back in 1965, after a release a little bigger than the 2023 one, and it’s shown up on and off ever since. So there’s a 60-year paper trail on these pipes being touchy under heavy sustained flow, and we’re now seriously discussing running them wide open indefinitely.

Reclamation’s fix so far has been to line the insides with epoxy, a temporary patch that took months, and they’ve scheduled more relining work since. They also built a couple of small-scale physical models of the dam at their Technical Service Center in Denver, actual scale replicas, so they can run water through them and watch how the structure behaves at low reservoir levels without wrecking the real thing to find out. On top of all that they put out temporary operating guidelines whose entire purpose is to keep Lake Powell above 3,490 feet so they never have to depend on the outlet works in the first place.

Their current operating plan is basically built around not having to use the backup at all. That should tell you exactly how much confidence they have in it.

The Memo they Tried to Hide

In March 2024, Richard LaFond, the director of Reclamation’s Technical Service Center, put the problem in writing in a technical memo. It’s a dry document. It’s about steel pipes and coatings and discharge rates and maintenance schedules. It’s also, quietly, one of the more important things any federal engineer has written about this river in years, because it’s the government admitting on paper what the outside critics had been saying all along: the outlet works shouldn’t be relied on as the sole means of sustained water delivery, and running them long-term drives up maintenance and raises real operational and safety concerns.

Plain English: our backup plumbing degrades when we actually lean on it, and we know it.

Eric Balken, who runs the Glen Canyon Institute, has been about as blunt as anybody. His point is that even if everybody dropped everything today and committed to a genuine fix, like building a bypass that routes water around the dam at river level, that solution still takes something like ten years to design and build. So every year we spend arguing instead of building is a year we flat-out don’t get back. If the answer takes a decade either way, he’s asked, why are we still procrastinating?

The bypass idea, pushed by groups like the Utah Rivers Council, is exactly what it sounds like: a tunnel or channel that lets the Colorado flow around Glen Canyon Dam near river level, so water delivery no longer depends on the dam’s aging outlet plumbing at all. It’s the kind of thing that would actually solve the problem. It’s also expensive, slow, and politically radioactive, because to some people it looks like the first step toward giving up on the dam entirely. So it sits there, discussed and undone.

And the Glen Canyon Institute’s worst-case read is the one that sticks with me. Their argument isn’t just that the pipes are too small. It’s that the low-water physics could damage the outlets badly enough to make them unusable, and if that happened at the wrong moment, there might be no controlled way to release water downstream at all. Not “less water.” No good way to send it. That’s the thing sitting at the very bottom of this whole story.

What This Is, and What It Isn’t

I’m not going to sell you a Hollywood dam-burst, because it isn’t that, and anybody telling you Glen Canyon is about to crack open and wash Arizona off the map is either lying or doesn’t understand the problem. Even the Sierra Club, which is no fan of this dam, has said there’s no structural risk to the dam itself. It’s a giant plug of concrete and it’ll keep standing there whether there’s water behind it or not.

The danger isn’t the dam falling down. It’s losing control of the water going through it. And for the 30-plus million people at the end of the pipe, those two things end up feeling like the same day. Whether the river stops because the concrete failed or because the only working outlet got too chewed up to run safely, the water still doesn’t show up at the canal. It’d be some boring engineering nobody outside Denver could explain, and the result would still be dry ditches running through Phoenix.

There’s also a real chance this never comes to a head, and I’ll say that just as plainly. The states have been holding the lake up on purpose, parking water in Powell instead of releasing it, draining smaller reservoirs upstream to keep the level propped. As long as they can keep Powell above 3,490, the outlet works stay a backup nobody has to actually trust. The entire game right now is not dropping into that 120-foot danger band in the first place. But that’s buying time against a river that keeps shrinking, and one bad winter can burn through the cushion fast.

The Part That Should Actually Make You Mad

Here’s the political shitshow. All that post-2026 rulemaking I covered in Part 1, the giant federal process deciding how the whole river gets run after this year? Reclamation has said flat out that it’s “not an infrastructure EIS.” Translation: the process rewriting the rules for the entire river is specifically not looking at fixing the dam’s plumbing. They drew a box around the infrastructure problem and set it off to the side.

Meanwhile the Lower Basin states, the ones whose water literally comes through these pipes, have straight-up told Interior to study an overhaul of Glen Canyon Dam. Environmental groups have been asking for the same thing for years, at conferences, in the press, in direct letters to the agency. And the fix that would actually settle it, some version of that river-level bypass, is a ten-year build nobody has committed to starting.

So put it all together. A known problem with a paper trail back to 1965. A 2023 test that proved the damage is real and current. A 2024 memo where the government admits it in writing. Outside groups and downstream states all pointing at the same thing. And the official move is to keep the fix out of the main planning process and bet that the lake stays high enough that we never have to find out what happens down in that danger band. None of that is an accident of engineering. It’s a decision to look away from the expensive problem and hope the weather cooperates.

What This Means For You

I’m not going to redo the whole water-prep rundown from Part 1 here, but the plumbing angle changes one thing worth spelling out.

Most people picture the water crisis as a slow, orderly thing. Lake gets low, cuts get announced on the news, everybody grumbles and rations, life goes on. The outlet works problem is the reason it might not play out that clean. If Powell drops into that danger band and the pipes start acting up, deliveries could get throttled or interrupted for emergency repairs on infrastructure that’s underwater and under pressure, and that kind of interruption doesn’t send you a polite 90-day notice. That’s the difference between a shortage you watch coming for months and a delivery hiccup that lands on some random Tuesday.

Which is the whole reason I keep beating the same drum. Store real water and enough of it, at least a gallon and a half per person per day, two weeks minimum and a month if you can manage it, in food-grade containers you actually rotate. Full walkthrough’s here: Long-Term Water Storage. And the power dies before the water does, so back that up too: Cheap Off-Grid Power: Real Solar Setups Under $1,000. The point of prepping this stuff is so a bad Tuesday is an annoyance instead of an emergency.

Where This Goes Next

So that’s the dam. Two dead lines, a 120-foot danger band, a set of backup pipes that were never built for the job and started coming apart the one time we really used them, a memo where the feds admit it, and a fix nobody’s started that takes ten years no matter when they begin.

Next in The Colorado River Story I’m getting into the fight I flagged in Part 1: the upper states versus the lower states, and the legal knot underneath the whole thing. Who’s actually required to give up water when there isn’t enough, why the 1922 deal makes that almost impossible to answer, and what happens when the feds quit waiting for a handshake and just impose an answer. That’s where this stops being about engineering and starts being about who wins and who dries up.

Get your water handled before that one drops.

Sources and Reports Worth Reading Yourself

• Circle of Blue, “Glen Canyon Dam Faces Its Existential Moment”: https://www.circleofblue.org/2026/supply/glen-canyon-dam-faces-its-existential-moment/
• High Country News, “The coming failure of Glen Canyon Dam”: https://www.hcn.org/articles/the-coming-failure-of-glen-canyon-dam/
• Sierra Club, “How Much Longer Can Glen Canyon Dam Last?”: https://www.sierraclub.org/sierra/how-much-longer-can-glen-canyon-dam-last
• Glen Canyon Institute, Lower Basin states ask Interior to study a dam overhaul: https://glencanyon.org/lower-basin-states-tell-interior-to-study-overhaul-of-glen-canyon-dam/
• Missoula Current / Utah News Dispatch, on the Reclamation memo and outlet works: https://missoulacurrent.com/glenn-canyon-dam/
• Coyote Gulch, review of the river outlet works controversy: https://coyotegulch.blog/2026/02/13/unpacking-the-controversy-over-glen-canyon-dams-river-outlet-works/
• U.S. Bureau of Reclamation, Glen Canyon Dam operations and 24-Month Study: https://www.usbr.gov/uc/water/crsp/cs/gcd.html
• U.S. Bureau of Reclamation, Post-2026 Colorado River Operations: https://www.usbr.gov/ColoradoRiverBasin/post2026/index.html

The Reclamation technical memo on the river outlet works was authored by Richard LaFond, director of the agency’s Technical Service Center, in March 2024. Lake levels in this piece reflect mid-July 2026 Bureau of Reclamation data. Check the live USBR trackers for current numbers before you rely on them.