Hydropower has a reputation as the set-it-and-forget-it corner of the grid. You build a dam, let a river push on a big bladed wheel, and the thing pretty much looks after itself for the rest of your life. McNary Dam, on the Columbia River between Oregon and Washington, has been doing that since Dwight Eisenhower dedicated it in 1954.<\/p>\n
But even a machine like that doesn’t last forever. On October 20, the Army Corps of Engineers pulled the rotor out of a generator at McNary called Unit 14, using a 650-ton lifting beam and a pair of gantry cranes, according to a Corps photo caption<\/a>. Unit 14 goes first, and the plan is new turbines for all 14 of the plant’s main units.<\/p>\n I went digging to see what’s going into that hole, and that’s where things got a little weird. The Corps has described the new turbines in two different ways over the last few months, and I can’t get them to match.<\/p>\n Because it’s in the way. Each unit at McNary is basically a stack. The runner (the propeller-looking wheel the river pushes on) sits at the bottom, a long vertical shaft runs up from it, and the top of that shaft spins the rotor inside the generator. If you want to change the wheel, everything above it has to come off first.<\/p>\n We saw the same routine last week at Fort Randall Dam in South Dakota<\/a>, where the rotor weighs 275 tons. The Corps hasn’t published what Unit 14’s rotor weighs, which is a little annoying for a lift it bothered to photograph. The 650-ton figure describes the beam, and I’d read that as what the beam’s built to carry, not what was hanging from it.<\/p>\n I’d still bet this one’s heavier than Fort Randall’s. McNary’s units are rated at 70 megawatts each, against 42 or 43 for the old Fort Randall units, and both plants’ machines turn at 85.7 rpm. An older Corps spec sheet also lists two 350-ton cranes inside the powerhouse, which would give you 700 tons to work with if those are the two cranes in the caption.<\/p>\n The turbines coming out are Kaplans, and their runners are about 23 feet across (280 inches, per that same spec sheet). A Kaplan’s blades can pivot on the hub, sort of like a variable-pitch propeller, so the turbine can stay efficient as the flow changes. It’s a clever setup, but it means moving parts and oil seals down in the water, which is a fairly awkward thing to have in a salmon river.<\/p>\n A fixed-blade runner skips all that. Its blades are set at one angle for life. You give up some flexibility when the river isn’t running the way the runner was designed for, but there’s less to break. “It’s fixed blade, so there are fewer components to maintain,” Paul Ocker, chief of operations for the Corps’ Walla Walla District, said in a Corps story published October 20<\/a>.<\/p>\n That story and the rotor caption both say the existing Kaplans are being replaced with stainless-steel, fixed-blade turbines. So that’s settled, right?<\/p>\n Well, back in August, the same district shot a video about the same unit<\/a>. Its description says the first installation “uses a five-bladed adjustable runner designed for McNary.” Adjustable blades are the whole point of a Kaplan, so that’s pretty much the opposite of a fixed blade.<\/p>\n So which is it?<\/p>\n I can’t tell you, and I couldn’t find anything from the Corps that explains the gap. My best guess is McNary ends up with a mix of both.<\/p>\n That’s been the plan on paper for more than a decade. When the Corps went looking for contractors in 2015, the stated goal was a combination of fixed and adjustable runners, and a few years ago its technical lead on the job, Martin Ahmann, talked about finishing the adjustable design first and then starting on the fixed one. Ice Harbor Dam, just up the Snake River, got one fixed runner and two adjustable ones.<\/p>\n But that’s me connecting dots. Neither of the Corps’ latest descriptions mentions a mix, so I wouldn’t assume all 14 will be one type or the other until the district spells out which unit gets what.<\/p>\n What the Corps does say plainly is that hardware is already on site: the powerhouse has received the major pieces for Unit 14, along with Unit 1. Saying Unit 1 was delivered is a slightly odd way to put it, and I’m assuming that means the unit’s runner or parts. The Corps hasn’t published the new runner’s diameter or weight, either.<\/p>\n Young salmon headed for the ocean have to get past this dam, and some of them go through the turbines. When the Corps signed the turbine contract, it listed fish survival first among its goals, with turbine efficiency further down the list. So I’d guess the salmon count for a lot when this job gets graded.<\/p>\n The closest thing we’ve got to a preview comes from Ice Harbor. Corps biologists tagged juvenile Chinook salmon, sent them through that dam’s new fixed-blade unit and got a 98.25% direct survival rate, along with a 2.3% drop in visible injuries compared with testing on the old adjustable turbine. Those aren’t McNary numbers, and I haven’t found any published for the new McNary design.<\/p>\n Then there’s the steel. At Ice Harbor, the Corps said stainless holds up better against cavitation and corrosion than the original carbon steel. Cavitation is what you get when fast water drops so low in pressure that vapor bubbles form and then collapse against the metal. Keep that up for decades and it pits the blades, which is the same problem that came up with the seven-blade runner on the Cumberland River<\/a>.<\/p>\n You’d figure a plant with 14 units could lean on the other 13 while one’s in pieces. McNary can’t, really. The Corps calls the powerhouse a bottleneck on the Columbia because the river routinely brings more water than the turbines can pass, so generation drops significantly any time one of the main units goes down.<\/p>\n Picture a 14-lane toll plaza with traffic backed up for miles. Close one booth and the other 13 can’t pick up the slack, because they’re already going flat out. As far as I can tell, the cars in this case are water that ends up going over the spillway instead of making electricity.<\/p>\n That’s expensive water. Back when it signed the contract, the Corps put McNary’s power revenue at $150 million to more than $300 million a year, or a bit more than $16 million per unit. By my math, that’s about $44,000 a day for an average unit, on old numbers. The plan back then was to install two units at a time, which would take 140 megawatts out of the plant’s 980 during the work. The Corps hasn’t said how long Unit 14 will be down.<\/p>\n More water through the machines, mostly. The Bonneville Power Administration, the federal agency that markets McNary’s power and pays for this work, says the project should let each unit pass more water and use it more efficiently. Unit 14 is also one of three getting rewound, along with Units 5 and 13. A rewind means replacing the copper windings in the stator, the stationary ring the rotor spins inside.<\/p>\n BPA says those three will end up at 100 MVA to match the rest of the plant. MVA isn’t quite the same thing as megawatts, but it’s in the same neighborhood, and the Corps still lists the plant at 980 megawatts, or 70 per unit. So it looks to me like the generators got ahead of the turbines a while ago, and the new runners are partly a catch-up job. What I can’t find is the new per-unit rating.<\/p>\n A few figures floating around deserve an asterisk. The October story says the 14 generators have been in service for 72 years, which is fair for the plant as a whole, even though the first power came online in December 1953 and the last unit didn’t start up until February 1957. The August video calls them 75-year-old turbines, which is harder to square, since McNary wasn’t making any power 75 years ago.<\/p>\n The line about nearly 700,000 homes is also a rule of thumb. The Corps’ own fact sheet says about 686,000, which is just 980 megawatts times its estimate of 700 homes per megawatt, and that assumes full output.<\/p>\n The lock number’s the one I’d really question. That same October story credits McNary’s navigation lock with more than 51 million tons of cargo a year and a $25 billion economy. A separate Corps history of the dam puts the lock at more than 5.6 million tons in fiscal 2021.<\/p>\n The bigger number looks an awful lot like the Pacific Northwest Waterways Association’s count for all cargo on the lower Columbia: 51.1 million metric tons worth $25.3 billion in 2023. So I’d treat it as a figure for the river, not for this one lock.<\/p>\n The Corps awarded the turbine contract eight years ago to Alstom Renewable US, a GE company at the time, for $321.3 million, the biggest the Walla Walla District had ever handed out. The plan was to wrap up in about 14 years. The Corps’ announcement also figured the old turbines would have to keep going for about seven more years on average until their replacements showed up.<\/p>\n That was eight years ago.<\/p>\n BPA’s latest status report that I could find<\/a>, for the fourth quarter of fiscal 2025, shows $533.1 million authorized for the runners and rewinds, a $500.2 million forecast, and $58.2 million spent at that point. So the money looked fine on paper.<\/p>\n The schedule’s another story, with an authorized in-service date of October 7, 2033, against a forecast of December 7, 2037. I wouldn’t be shocked if that one moves again, frankly, since nobody’s put a new runner in the water here yet.<\/p>\nSo why pull the rotor at all?<\/h2>\n
So will the new blades tilt or not?<\/h2>\n
Why do salmon get a say in this?<\/h2>\n
One unit down hurts more than you’d think<\/h2>\n
What does McNary get out of this?<\/h2>\n
Some of these numbers don’t quite line up<\/h2>\n
Don’t hold your breath for the finish line<\/h2>\n