Like the Mosul Dam, the Teton Dam was an earthen embankment structure and, like the current dangerous predicament of the Mosul Dam, the failure of the Teton Dam was the result of the cumulative effect of human error. The multiple purposes for which the dam was being built included agricultural irrigation, flood control, power generation and recreation. Investigations of the Teton River canyon for purposes of dam construction go back to the early 1930s. An impetus to undertake construction of a new dam came in the wake of a severe drought in the region in 1961 followed by a major flood the next year. The Bureau of Reclamation proposed the construction of a dam in 1963 and Congress authorized the funds the following year. After extensive surveys and core sampling throughout the 1960s, a site was chosen in the lower Teton Basin in Fremont County, Idaho. (The coordinates of the site are 43° 54′ 33” N latitude x 111° 32′ 24” W longitude) Again, I will highly recommend the use of Google Earth as a research tool. The wreckage of the dam shows up clearly as does the downstream region affected by the flood. All one needs to do is type Teton Dam in the search engine and Google Earth will zoom right in on it.
Construction of the dam began in February of 1972. Specifically, the dam was constructed of five zones of different kinds of earth material installed in compacted layers, each layer performing a different function with the central core designed to be the one that actually retained the water. As I stated, at completion the dams crest stood 305 feet above the riverbed and its width was almost 3200 feet. At the base of the dam it was 1700 feet wide. When full the reservoir would have contained 356 million cubic meters of water, equivalent to about 94 billion gallons and the lake formed above the dam would have reached upriver nearly 17 miles. One way to visualize this volume of water is to imagine it frozen as a giant ice cube. In this case the cube would be about 2350 feet on each side (disregarding an expansion coefficient for ice.) That’s a lot of water. Given that one cubic meter of water weighs one metric ton, it is obvious that the weight of the lake when full would be some 356 million metric tons. That is a considerable amount of pressure bearing on the bedrock. (By comparison, if the volume of water in Mosul Lake reservoir were frozen into a giant ice cube, it would be more than 7,300 feet on a side, or 1.4 miles.) While a number of geologists and engineers had expressed reservations about the Teton site and the unsuitability of the bedrock for dam construction, politics took precedence over prudence and the project went ahead anyway.
The main problem with the bedrock in the Teton Dam locale is that it consists of porous rhyolite and basalt. These are volcanic rocks which any geologist will advise are not the kind of foundation upon which one wants to erect a major dam with a large reservoir behind it. Interestingly, this particular rock, called by geologists the Huckleberry Ridge Tuff, was the product of a massive eruption of Yellowstone Caldera about 2 million years ago. This violent eruption was at least 6000 times more powerful than the 1980 eruption of Mount St. Helens, ejecting about 600 cubic miles of dust, ash and debris into the atmosphere. This amount of ash would have produced a severe environmental effect that would have been brutally disruptive to both the global climate and the biosphere. Here is a striking example of a juxtaposition of catastrophes, a superimposed imprinting of the effects of a modern hydraulic event upon a landscape created by a flood of ash and molten rock emanating from within the Earth. When this material cooled it formed an extensive array of fractures and fissures. This character of the bedrock is clearly displayed in the Figure 1.
So, the bedrock into which the Teton Canyon had been eroded was composed of this highly fissured and porous rhyolite and basalt. The choice, therefore, of dam site itself was the first mistake. The problem was that sites suitable for the construction of monumental dams were becoming scarce by the 1970s at the same time as pressure from multiple factions lobbying for all the presumed benefits to be derived from the new dam were considerable. I should mention that up to this point the Bureau’s record of dam building over nearly 75 years was virtually impeccable, without a single dam failure (although one close call). The engineers, geologists, contractors and other professionals in their employ were highly experienced at dam building, after all, the Bureau had built two of the great engineering wonders of the modern world, Hoover Dam and Grand Coulee Dam. But this reign of success and accomplishment became compromised as the enterprise became progressively more bureaucratized and politicized with the growth of government. Teton Dam represented the tipping point and its physical collapse parallels the economic collapse in Venezuela as that nation succumbs to total bureaucratization. As I mentioned, the failure of Teton Dam marked the end of the era of monumental dam building in America.
Over the basalt bedrock of the region there has accumulated a very interesting and unique type of soil material called loess (pronounced ‘luss’). This mostly volcanically derived soil is extremely fertile and is widely distributed throughout the Pacific Northwest. All that is needed for loess to support abundant biological activity is the addition of water, but much of the area over which it is found in Idaho and Washington is quite arid, hence, the numerous hydrological projects undertaken throughout the northwest by the Bureau of Reclamation that occasioned the building of dams, canals, aqueducts and irrigation systems. The annual precipitation in the headwaters of the Snake is less than 20 inches per year and most of this comes in the form of snow. I will have more to say about this curious loess material as well as the Yellowstone caldera eruptions in future contributions, but for now we will note that due to the local abundance and accessibility of loess, which happens to be a very friable substance, it was chosen as the core material for the Teton Dam. In fact, the origin of the term loess is a German word that means loose and from which the English word derives. The decision to use abundantly available loess material was the result of cost saving efforts which was in turn due to the insufficiency of available funds to construct the dam correctly. I will comment further on this matter below. The choice of loess for the core material was mistake number two.
In the course of site preparation contractors began grouting operations. As I explained in the article about Mosul Dam in Iraq, grouting is a method of injecting fluidized concrete under high pressure into voids and fissures in the rock surrounding a dam for the purposes of preventing seepage. When the grout hardens it forms an impervious barricade to water penetration, or at least that is the working assumption. In the course of excavation it was confirmed that the bedrock adjacent to the Teton Dam site was porous, containing large voids and cavities, some of which consumed enormous quantities of grout. In some cases the voids were never actually filled. Some of the grout holes extended to a depth of more than 300 feet. Large fissures and caverns were especially prevalent around the right abutment. More than 1.1 million cubic feet of grouting material was injected into the Swiss cheese like bedrock with no end in sight. This was more than double the amount of grout originally estimated. The decision was made to suspend grouting operations due to the accruing cost overruns. This was mistake number three.
In early June of 1973, the contractor, Morrison-Knudson, completed the diversion tunnel and rerouted the Teton River on June 8. Emplacement of the embankment material commenced in October, 1973. The embankment, as I said, consisted of five layers with the central core layer, the one that was to serve as the barrier to water penetration, composed of friable loess. Over the loess core were layers of rock and clay extracted from the site and from borrow pits nearby. At its completion the Teton Dam consumed some ten million cubic yards of earth and rock material.
By the beginning of 1976 the dam was, for the most part, complete. Officials had already begun filling the reservoir in October, 1975. The approved filling rate for a large dam reservoir was generally not permitted to exceed a maximum rise of about 1 foot per day. This allows the site to adjust slowly to the substantially increased pressure imposed by the weight of a large body of water. However, in the spring of 1976 two occurrences conspired against this preferred filling rate. First, there was an unusually heavy snowpack on the Teton Mountains that winter, and second, the month of May was unusually warm. The warm weather triggered a rapid melting of the copious volume of snow which had accumulated and that caused a rapid rise in river level. The decision was made to capture the increased amount of meltwater, thereby accelerating the filling of the reservoir and hence, the completion date, for the project was significantly behind schedule due to the numerous delays arising as a result of the highly problematic site preparation. So the reservoir was allowed to fill at its natural rate rather than a controlled rate. This led to a water level rise over four times greater than the recommended rate. A process that should have taken up to three years was virtually completed in that single spring of 1976. What compounded the problem of the accelerated lake level rise was the fact that the outlets by which the water level could be drawn down were only partially operational. So this was major mistake number four.
In the first few days of June, 1976, as the reservoir reached about 280 feet in depth behind the dam, about 30 feet below the dam crest and about 15 feet from the overflow spillway, several springs showed up downstream from the dam, indicating that water was leaking through its bedrock base. Early in the morning of June 5, a Morrison-Knudson worker arrived at the site to find a leak in the dam itself, about 15 feet above the stream bed close to where it abutted the right canyon wall as one was looking downstream. The measured rate of leakage was about 20 to 30 cubic feet per second. The time of the discovery of this leak was between 7:30 and 8:00 a.m.
This was the beginning of the end for the Teton Dam.
Around 9 a.m. the project manager, Robert Robison, made a closer inspection of the leak and discovered that its’ flow rate had increased to nearly 50 cubic feet per second. He also noticed a second smaller leak issuing from the abutment rock near its contact with the embankment, about 130 feet below the crest. Sometime between 9:30 and 10:00 a.m. a wet spot appeared on the downstream face of the dam no more than about 20 feet from the abutment. The wet spot quickly turned into active seepage and a hole appeared. At this point it was clear that action must be taken. As the hole was growing larger two bulldozers were dispatched to push boulders and material into the hole but it continued to erode faster than they were able to fill it.
At 10:33 a.m. Teton Dam project officials alerted the Sheriff’s offices of both Madison and Fremont Counties to begin preparing for the onset of a devastating flood in the downstream area. At about 11 o’clock a.m. one of the dozers started to slide into the hole. The second dozer operator worked frantically to pull it out but his effort was futile, for the hole enlarged as fast as the first dozer could be pulled back. Finally both dozers slid into the expanding hole and were swept away into the torrent.
Sometime between 11 and 11:30 Project officials called the Sheriff’s offices and informed them that people living in the flood path must be evacuated immediately. As the call was being made a whirlpool appeared in the reservoir just upstream of the swiftly expanding leak. Two more bulldozers made a frantic and futile effort to shovel rip-rap into the whirlpool but it was sucked down as fast as they could dump it in. The whirlpool quickly expanded from a few feet in diameter to about 20 feet as the hole in the dam face continued to grow.
At about 3 minutes before noon a large section of the dam crest gave way and a massive wall of water poured through the opening. The roiling torrent of some 80 billion gallons carried away about 4 million cubic yards of embankment material that now mantles the floor of the canyon for miles downstream. The pumping station and power house at the base of the dam were completely obliterated and buried by the mass of water and mud.
Time magazine for June 21, 1976 ran a story recounting the experiences of a family who witnessed the entire collapse from start to finish. Dale Howard was a geography professor at Minot State College in North Dakota who visited the site that Saturday morning with his family. While he was snapping pictures from an observation platform he noticed “that darn hole started growing—quite slowly at first—forming a small waterfall down one side. It still looked like just a minor leak.” As Howard continued shooting the collapse began to unfold before his eyes and he was able to record the color pictures documenting the progressive failure as shown in Figures 5 through 9. When the two bulldozers were sucked into the hole it became apparent that Howard and his family were witnesses to something truly extraordinary. He commented “My wife was excited and my kids were crying because they thought the world was coming to an end. It was really frightening. If I had had a weak heart, maybe it would have stopped.” He continues his account: “The hole was enormous, and huge chunks were breaking off . . . By this time you could see daylight through the hole. It was almost like a natural bridge. Then the whole thing fell, and it was a raging torrent.” Howard goes on to describe that when this raging torrent hit the power plant “it just disintegrated. The water picked up a huge oil tank like a cork and away it went. There was a beautiful grove of cottonwood trees down below, and they were snapped off like matchsticks. Later I could see the water out on the plain. It was almost like a surrealist picture.”
Figure 10. Aerial view of the dam breach at peak discharge Source: https://www.flickr.com/photos/waterarchives/5811723977
As the torrent of water surged downriver residents of the towns lying in its path – Sugar City, Teton and Newdale – scrambled to evacuate, having to leave in such a rush that most of their belongings were left behind. Luckily, Teton and Newdale were on higher ground and spared the worst ravages of the flood wave. Another town, Wilford, did not fare so well, receiving the full force of the flood it was literally wiped off the face of the map. When the surging flood wave reached Sugar City at about 1 o’clock it was 15 feet deep. The largest city in the path of the flood was Rexburg. As the flood swept into Rexburg it struck a logging mill on the outskirts of town and hundreds of massive logs were picked up by the powerful currents and became battering rams that smashed and destroyed buildings by the hundreds.
It took about 8 hours for the reservoir to empty. 35,000 people had to be evacuated. Hundreds of families had their homes obliterated. 14 people lost their lives either directly or as a consequence of the flood. Several thousand were injured. Irrigation canals that brought water to farms and crops were choked with debris and rendered useless; highways, railroad lines and power lines were damaged or destroyed; seven bridges were destroyed; cars, trucks, trailer homes and whole houses were swept away; an estimated 400,000 acres of farmland was inundated. An estimated 13,000 to 20,000 head of cattle were drowned and in the aftermath their corpses littered the flood pathway. The floodwaters cut a swath of destruction for some 80 miles along the Teton and Snake River valleys, from the dam site to American Falls Reservoir. Some estimates place the value of the flood damage as high as 2 billion dollars.
The devastation was extreme, as the following series of images demonstrates. Most of these images, unless otherwise noted, are part of the great collection of Teton Dam Flood photographs preserved by Brigham Young University, Idaho. Most of these images are from the Rexburg and Sugar City area and the flood path above and below these towns. Please take the time to peruse this gallery in an effort to get a sense of what this event was like from the perspective of those who experienced it and to appreciate the extraordinary power and force of water as well as the destruction that can follow in its wake.
Click on left arrow to advance photos in chronological order from right to left.
– Randall Carlson