A big concrete dam looks like it was carved out of one enormous block of stone. Up close, it’s more like a stack of pancakes. The concrete goes in one layer at a time, and each layer is supposed to bond so well to the one below it that the whole structure behaves like a single chunk of rock.<\/p>\n
Priest Rapids Dam, a 950-megawatt hydroelectric plant on the Columbia River in central Washington, has a spot where that bond gave up. So the utility that owns it is doing something that sounds pretty drastic. Grant County Public Utility District is drilling 38 holes down through the dam’s spillway into the bedrock underneath, and it’s using them to strap the concrete to the rock with tensioned anchors.<\/p>\n
You shouldn’t read this as breaking news, though. Grant PUD’s commissioners awarded the contract back in December 2025, construction got going this year, and the job isn’t scheduled to wrap up until 2028. What’s new is the budget. On November 6, the Columbia Basin Herald<\/a> reported that the utility’s proposed 2027 budget sets aside $33.7 million for next year’s share of the spillway work. That figure only covers 2027.<\/p>\n A lift joint is the seam between two pours of concrete. Dams go up in horizontal layers called lifts, and the flat surface where one lift sits on the next only holds as well as the bond between them.<\/p>\n Grant PUD found out it had a problem in March 2018. Inspection drilling turned up leaks inside the spillway monoliths (the giant concrete blocks a spillway is built from), and the utility traced them to a disbonded lift joint<\/a>. A paper presented to the Association of State Dam Safety Officials<\/a> a year later put it more plainly. One internal joint had separated, and reservoir water was getting into the gap to varying degrees.<\/p>\n The utility declared what’s known as a non-failure emergency and lowered the reservoir by about three feet. Crews drilled drainage and inspection holes from the grout gallery (a tunnel that runs through the inside of the dam), and that drilling took water pressure off the joint. They pulled about 230 core samples along the way. Then they wired the monoliths with instruments to see whether anything was moving.<\/p>\n So should anyone living downstream be worried?<\/p>\n Grant PUD said at the time that the leak never posed a safety risk to the public or its employees, and the dam kept generating power the whole way through. The emergency condition was later called off, along with the limit on the reservoir level. Still, a separated joint you have to keep an eye on with sensors isn’t really something a utility wants to live with for decades, and it looks like Grant PUD came to the same conclusion.<\/p>\n There’s a second reason for the job, too. Grant PUD says the anchors will “bring the dam up to modern seismic standards.” Construction on Priest Rapids started in the mid-1950s, and it’s safe to assume earthquake engineering has moved on a fair bit since then.<\/p>\n The basic idea is simpler than it sounds like it should be. Crews drill a hole down through the concrete and keep going into solid rock. They lower a cable into the hole and lock its bottom end into the bedrock with cement grout or something similar. Then they pull the top end tight and fill the rest of the hole. Once it’s tensioned, the cable works a lot like a ratchet strap in a pickup bed. It clamps everything between its two ends together.<\/p>\n Which means the concrete on either side of the separated joint gets clamped together, and the spillway gets pinned down onto its foundation. According to Kuney Construction, which is doing the work in a joint venture with Nicholson Construction, the 38 anchors run through the ogee portion of several spillway monoliths and down into the rock beneath them. The ogee is the curved, S-shaped slope that water slides down when the spillway gates open.<\/p>\n Priest Rapids’ spillway isn’t a small target, either. That same ASDSO paper describes 23 monoliths, each one 52 feet wide. Between them, they carry 22 radial gates that can pass up to 1.34 million cubic feet of water per second.<\/p>\n What I can’t tell you is how long each anchor is or how much force it holds. I couldn’t find either number in anything Grant PUD or its contractors have published, and Kuney only says the anchors cross multiple monoliths.<\/p>\n Wanapum Dam is Grant PUD’s other hydroelectric plant, just upstream, and its spillway turned out to be cracked in February 2014. Divers found a crack about 2 inches wide running all the way across one spillway monolith, and the top of the pier above it had shifted downstream. The utility drew the reservoir down, and Grant PUD says the Priest Rapids job is similar to the work it did at Wanapum.<\/p>\n The Wanapum fix used 35 post-tensioned tendons, each one a bundle of 61 strands, according to Engineering News-Record. Some of those tendons ran up to 260 feet long. Crews added 69 solid bar anchors on top of that, and the whole repair was done by April 2015.<\/p>\n I wouldn’t assume the Priest Rapids anchors look exactly like that, since it’s a different dam with a different problem. But Wanapum is arguably the best clue we’ve got. And frankly, a separated joint that’s been monitored for years sounds a lot less dramatic than a fresh crack with a pier already leaning downstream.<\/p>\n The awkward part is where all of this happens. The spillway sits right in the Columbia, so the joint venture has to design and install its own working platforms, and it’s bringing in cranes and barges to support the drilling. When it recommended the contractor, Grant PUD said the plan would allow the dam to keep operating during the work. It also said the contract requirements were written with input from the Wanapum people, the dam’s operators and its environmental staff.<\/p>\n That’s a pretty big ask. You’ve got heavy drilling gear perched over a big river with a hydro plant running right next to it. As you’d expect, it isn’t a quick job. Grant PUD’s plan had construction starting around April 1, 2026, with completion expected by July 2028.<\/p>\n Then there’s the contract. Grant PUD sent the job out to 27 prequalified contractors and got one bid back, from Nicholson-Kuney, at about $58.7 million. That came in under the utility’s own engineering estimate of $64 million, and commissioners approved the $58.683 million contract on December 9, 2025. One bidder out of 27 seems a little thin to me. Then again, I can see why plenty of firms would pass on two-plus years of drilling through a working dam.<\/p>\n Old concrete dams tend to need some pretty strange surgery as they age. We’ve covered a Canadian plant whose swelling concrete has to be sawed apart with diamond wire<\/a> to keep its turbines lined up. And the Columbia has had a busy fall in general, with the Army Corps of Engineers hoisting a generator rotor out of McNary Dam<\/a> to start a turbine swap. The Priest Rapids fix is a lot less photogenic, since most of it ends up buried inside the concrete.<\/p>\n Grant PUD’s proposed 2027 budget puts $33.7 million toward the spillway work next year, and its schedule calls for the job to be done by July 2028.<\/p>\n","protected":false},"excerpt":{"rendered":" A big concrete dam looks like it was carved out of one enormous block of stone. Up close, it’s more … <\/p>\nSo what’s a lift joint?<\/h2>\n
How do you bolt a dam to the ground?<\/h2>\n
Grant PUD has done this before<\/h2>\n
Drilling over a working river<\/h2>\n