Say “Detroit” on a car site and everybody’s head goes to the same place. Ford, GM, Woodward Avenue, and a lot of strong opinions about the Lions.<\/p>\n
So let me save you some confusion. This Detroit is in Oregon, and it got the name because, according to Oregon State Parks<\/a>, the people who first settled there came from Michigan. About 45 miles outside Salem, it’s also home to a 463-foot concrete dam. And on Wednesday, a federal judge in Eugene refused to stop the Army Corps of Engineers from taking the lake behind that dam well below its usual fall low.<\/p>\n The Corps said Thursday<\/a> it plans to start in late November. Marion County says it’s going straight to the Ninth Circuit to try to block it.<\/p>\n Lowering a lake is usually about the most boring thing the Corps does. It happens every fall at Lake Pend Oreille in Idaho<\/a>, and it happens at this lake too. So why’s this one in federal court?<\/p>\n I went through the judge’s 20-page opinion<\/a>, and it pretty much comes down to mud.<\/p>\n Detroit Dam blocks the North Santiam River, and the USGS calls it<\/a> the tallest dam in the Willamette River basin. It’s a concrete gravity dam, which basically means it holds the water back by being too heavy to push over. It runs 1,523.5 feet across and has been there since 1953, with two generators rated at 100 megawatts combined.<\/p>\n I’d love to give you the Corps’ number for what it weighs, but I couldn’t find one. An old company history from Utah Construction puts the dam’s volume at 1,357,000 cubic yards. At roughly 2 tons a cubic yard for typical concrete, my back-of-the-envelope math lands somewhere around 2.7 million tons, so treat that as an estimate.<\/p>\n At full pool, the USGS says the lake behind it holds 455,100 acre-feet. An acre-foot is a foot of water spread over an acre, roughly 326,000 gallons, which by my math makes a full lake about 148 billion gallons.<\/p>\n Every fall, the Corps has historically dropped the lake to 1,450 feet to leave room for winter rain. This fall, it’s going to 1,425. The Corps says that could take up to two weeks, and it’ll then hold the lake there for two more, unless a storm means it has to start refilling sooner.<\/p>\n The late November start is the Corps’ own wording this week. Its environmental review, as the judge’s opinion describes it, leaves room for early December too, so I wouldn’t set a calendar alert just yet.<\/p>\n After that, the plan is to go lower each autumn until the lake hits 1,395. You may have seen this described as the lowest the lake has ever been. That’s true of 1,395, which state Rep. Ed Diehl’s office says would be about 30 feet lower than the lake’s ever been. It isn’t true of this year’s number, since the lake reached 1,425 in 2015 and again in 2019.<\/p>\n You may also be wondering how anyone lowers a lake to 1,425 feet behind a dam that’s only 463 feet tall. Those are elevations, so they’re measured from sea level, and they have nothing to do with how deep the water is. The top of the dam sits at about 1,580 feet, which by my math puts the foundation near 1,117. So even at 1,425, there’d still be roughly 300 feet of dam below the waterline.<\/p>\n Young Chinook salmon and steelhead find their way to the ocean by following the current. A big, slow reservoir mostly kills that current, so they can spend a long time wandering around in there with predators and disease. And once they do reach the dam, they’ve got three ways out: the turbines, the spillway, or the regulating outlets, which are basically gated pipes low in the dam.<\/p>\n You can probably guess which of those goes badly for a small fish.<\/p>\n It’s the turbines, by the way. And these fish make them harder to avoid, because they stick near the surface. A 2021 federal ruling quoted in this week’s opinion said they’re unlikely to find the outlets unless the lake’s surface is within 20 to 25 feet of the opening. So if you want them to use the outlets, you’ve got to bring the lake down to them.<\/p>\n Now look at the elevations above. At 1,425, the lake’s still sitting above the power intakes. At 1,395, the surface would be below their centerline. I can’t find anywhere the Corps spells that out as the reason, but it looks to me like a pretty big deal for keeping fish out of the turbines.<\/p>\n The 2024 federal Biological Opinion behind all this, a document fisheries scientists write on what a project needs to do to keep a protected species going, describes 1,395 as no more than 50 feet above the upper outlets. That’s about twice the 20 to 25 feet in the older ruling. I’m no fish biologist, and I couldn’t find anything explaining the gap.<\/p>\n So this year’s 1,425 looks to me more like a dress rehearsal than a fix. The Corps calls it an important step for the fish, and it says it’ll use each year’s muddy-water data to decide how much lower to go next time. (The Corps’ release also calls them “endangered.” They’re actually listed as threatened, though they’re protected under the Endangered Species Act either way.)<\/p>\n Conservation groups sued the Corps in 2018, a federal court found it had violated the Endangered Species Act, and interim orders brought deep drawdowns to some other reservoirs in the system. So the Corps got sued for not doing enough, and now it’s getting sued for doing it. That’s a rough gig, frankly.<\/p>\n Turbidity. That’s just how cloudy water is, and it’s measured in Nephelometric Turbidity Units, or NTUs, which is a mouthful nobody says out loud. When you pull a lake down, you expose banks that have been sitting under water, and the river starts carrying that dirt away. The Corps’ own environmental review acknowledges turbidity will go up.<\/p>\n That matters because Salem’s Geren Island plant draws from the North Santiam and supplies about 220,000 people. It uses slow sand filtration, which is about as low-tech as it sounds. River water trickles down through layers of sand, and a living layer on top catches what’s in it, which works great until you start feeding it mud.<\/p>\n Salem’s pilot tests had the filters lasting 67 days at 10 NTU and just 8 to 13 days at 50 NTU. The North Santiam normally runs below 10. Salem told the court the drawdown would likely push the river far past 50, and that it may have to shut the plant down once turbidity tops 10.<\/p>\n Marion County puts the number of people downstream at over 230,000. The city of Stayton, which gets all its water from this river, joined the case, and so did the Santiam Water Control District, whose 114-mile irrigation system serves more than 17,000 acres.<\/p>\n They’ve also got a precedent they keep pointing to. In November 2023, the Corps dropped nearby Green Peter Reservoir 142 feet in a month. The muddy water disrupted treatment plants in Lebanon and Sweet Home, and thousands of kokanee salmon died. The Corps’ next Green Peter drawdown, the following fall, produced less turbidity, which the opinion ties to a later start and less loose sediment. Still, I get why people in Salem aren’t just taking the Corps’ word for it.<\/p>\n The Corps’ answer is that 1,425 isn’t new territory. It says the 2015 and 2019 trips to that level came with no adverse turbidity reported downstream, and that it doesn’t expect drinking water to be affected. Going lower in stages is supposed to limit how much old lakebed gets exposed in any one year.<\/p>\n Then there’s refilling. Marion County points to a very strong El Ni\u00f1o forecast and warns the lake might not come back in time, and it says that in 2015, docks had to move and there wasn’t enough water for fishing on the state’s free fishing weekend. The Corps says decades of data show the lake refills reliably. Drought can do plenty of damage on its own, as Blue Mesa in Colorado<\/a> has shown this year.<\/p>\n U.S. District Judge Mustafa Kasubhai agreed that losing drinking water would be irreparable. He just didn’t think anyone had shown it was likely. He wrote that Green Peter and Detroit differ in three big ways: Detroit will uncover a lot less sediment, it’s going to a level it has reached twice before, and it’ll happen later in the year.<\/p>\n Kasubhai also wrote something I found pretty striking: the record gives him no way to tell how muddy the river will actually get, or for how long. So there’s no number in that record. Salem couldn’t prove the bad outcome, and as far as I can tell from the opinion, the Corps hasn’t put an exact good one in there either.<\/p>\n His legal point was that the size of the potential harm doesn’t lower the bar. “‘Possibility’ is not ‘likelihood,'” he wrote. He gave the refill argument about the same treatment, including the county’s claim that the lake falls short one year in four.<\/p>\n Under the Corps’ emergency plan, an affected city contacts the Oregon Health Authority, and the Corps halts only if the governor’s office asks or there’s been a finding of a drinking water emergency. The judge noted the plan doesn’t say how long stopping would take, though he still counted it as another reason to deny the request.<\/p>\n And there’s the report. After Green Peter, Congress told the Corps to study high turbidity from these drawdowns, and that report was due January 4. According to the opinion, the Corps hasn’t delivered it, has no estimated completion date, and says finishing it depends on funding. That isn’t a great look, and I’d expect to hear about it again on appeal, even if it didn’t change this ruling.<\/p>\n Probably, but it isn’t locked in. “We are extremely disappointed in today’s decision, but this fight is far from over,” Colm Willis, who chairs the Marion County Board of Commissioners, said in a statement, adding that the county would seek an emergency injunction from the Ninth Circuit. As I write this, I can’t find any sign the appeals court has ruled.<\/p>\n Kasubhai also left the door open. His denial lets the plaintiffs come back during this drawdown or before next year’s if they’ve got new data specific to Detroit. The underlying lawsuits are still alive, and they matter more for the deeper drops, because that’s when the lake goes somewhere it’s never been.<\/p>\n If you drink Salem’s water or fish that lake, the Corps is holding virtual public sessions on November 18 to go over the schedule and its plans for drinking water, with times due within the week. If the Ninth Circuit doesn’t step in first, the Corps plans to start the drawdown in late November, with Detroit Reservoir due at 1,425 feet up to two weeks later.<\/p>\n","protected":false},"excerpt":{"rendered":" Say “Detroit” on a car site and everybody’s head goes to the same place. Ford, GM, Woodward Avenue, and a … <\/p>\nSo what’s the Corps planning to do?<\/h2>\n
Why does a salmon care how full the lake is?<\/h2>\n
The towns downstream are worried about mud<\/h2>\n
The judge wasn’t convinced, at least not yet<\/h2>\n
So is this actually happening?<\/h2>\n