Teton Dam Failure and Legacy - Updated

Dams look like they are simple, but their design and construction can be complex. When they fail after 1 year, something has gone very wrong. Teton Dam (Idaho, 1976)

Construction on the Teton Dam, reservoir, and powerhouse began in 1972 and by November 1975 the zoned earthfill embankment was essentially complete with a structural height of 305 feet and a crest length of 3,100 feet. Less than one year later, the dam experienced catastrophic failure on June 5, 1976 during its first filling. Failure of the Teton Dam and subsequent draining of the reservoir caused the deaths of 11 people and approximately $400 million in damages.

Seepage started on 3 June 1976 in the range of 1/4 cubic feet per second, and by the 5th a sinkhole had appeared and scouring had begun. Water-flow was 20 to 30 cubic feet per second of muddy water by that time. The dam failed at 11:55 am releasing 75 to 80 billion gallons of water, or 230,000 acre-feet, at a rate of more than 1 million cubic feet per second. This second link is from the Bureau of Reclamation.

There were several contributing factors. The foundation of the dam wasn't well understood by the designers. The treatment of the foundation was inconsistent. The project was behind schedule, after all.

Investigations of the Teton Dam failure attributed the catastrophe to a series of design and construction related deficiencies. These inadequacies related primarily to the foundation treatment at the dam and adherence to the overall construction schedule. It was determined the most probable physical failure mode was cracking of the dam’s impervious core due to internal erosion initiated by hydraulic fracturing of the key trench fill material.

Teton Dam was located in an area with highly permeable foundation materials. During investigation of the failure, it was discovered that proper treatment of such foundation material was not implemented. It appeared that the dam’s designers did not take the site specific geological conditions into account when developing the structure. This oversight was exacerbated by the lack of communication between the design and construction engineers about the proper preparation of the dam foundation.

As I have read about a couple of times when a new dam fails, people don't believe it. "You must be joking." And they refuse to believe it. You can hear one example of that if you watch the video below. In the end, most people did heed the warnings, and evacuate.

The first link at the top has several tabs, one is photos and another is a collection of video. The photo above is from that site, and originally (I believe) from the U.S. Department of the Interior's Bureau of Reclamation from the day of the dam failure.

This video is a Compilation of footage and audio associated with the failure of Teton Dam. It includes mostly warnings that went out over local radio stations, as well as some interviews from after the fact. It is only 3 or 4 minutes, and does include some video of the dam breach.

[Update: Commenter Steve recommends the following video (and I wholeheartedly agree): A brief History of: The Teton Disaster (Documentary). The video is 20 minutes, so plan accordingly. It is a wonderful examination of "Completion Bias." That bias in large projects (especially .gov projects?) that once you've sunk money into a project, it must complete.]

So why is this coming to my attention now? The Teton Dam wasn't the first dam on the upper Snake River. The Teton Dam replaced the Linderman Dam, which was built on almost the same spot. Linderman Dam restoration project will begin Aug. 18. But in typical bureaucratic mode, even though the remnants of the Lindeman Dam were a hazard, nothing was done between 1976 and today. That is 45 years of ignoring a problem.

Linderman Dam was largely dismantled in 1972 for the construction of Teton Dam. After the 1976 failure of Teton Dam, the river resumed its previous course, and the footings of Linderman Dam were exposed. The structural remains are located on the Teton River approximately seven miles upstream from the Teton Dam site, just downstream from the Milk Creek confluence. Successful project completion will eliminate the safety hazards; improve passage for aquatic species, including native Yellowstone cutthroat trout; rehabilitate the river to a more natural and passable elevation profile; and maintain upstream pool elevation for irrigation demands.

At least it is being addressed.

There are more videos at the first link at the top. Some of them are interesting. One is from the Bureau of Reclamation, and while it does contain a small section of good information, it is mostly a .gov propaganda piece. "We are a world-class organization who people ask for help with dams." Sure you are. If you didn't learn "valuable lessons" from a $400 million failure of an engineering project, that failed almost as soon as it was completed, and cost 11 lives, your organization should have been disbanded, and the executives shot.

Though they are right about 1 thing. This was one of the dam disasters that lead to a whole bunch of safety changes to the way dams are designed and built. It didn't fix everything, of course, but it did do some things to break up the bureaucracies.

One of the interesting things from that propaganda video is a description of the insane management rules regarding engineers from design actually talking to engineers from construction management. They had to get management approval communicate in any way. They had to get management approval to make a long-distance call to or from the construction site.

Teton Dam Failure - Another Look

I usually don't cover an infrastructure failure more than once, but Grady, of Practical Engineering, has a new video review of the failure. It turns out that he worked on a similar dam after the failure.

Here is a reference from The US Department of the Interior, Bureau of Reclamation factsheet on the incident. Teton Dam History & Facts. (For more links, see below the embedded video.)

On June 5, 1976, Teton Dam in southeastern Idaho catastrophically failed during its first fill. Early that Saturday morning, bulldozer operators tried in vain to plug seepage holes on the downstream face of the dam. By 11 a.m., a torrent of water ripped through the dam, releasing more than 1 million cubic feet per second.

This is the Practical Engineering video The Wild Story of the Teton Dam Failure

For more from the Bureau of Reclamation, see this link. That collects a lot of info, including some interviews with people impacted by the failure, and there is a link to the final report on what went wrong from the independent panel assigned to investigate the failure.

My previous post on the Teton Dam failure is Teton Dam Failure and Legacy - Updated

Malpasset Dam Failure

Odds are, you don't know as much as you think you do. Malpasset Dam (France, 1959) Pronounced "Mal-pa-say."

The image is from Wikimedia Commons. Click for a larger view and more information.

Construction started in 1952. Filling of the reservoir started in 1954. Failure occurred December 2, 1959 just before the reservoir was completely full. 421 people were killed by the flood.

At the time of completion in 1954, it was reported as the thinnest arch dam of its height (218 feet) with a maximum thickness of 22.2 feet. The dam was equipped with one un-gated, notched spillway at the center of its 736-foot-long crest.

The dam was situated on a type of gneiss (metamorphic rock) which has proved to be impervious enough to water to make reservoirs possible. But this particular gneiss had series of very thin layers (foliation - from the Latin for leaves) which made the foundation unstable.

Geological investigations that took place after the dam’s failure revealed that it had been built on a gneiss formation with a foliation structure exhibiting a slope of thirty to fifty degrees in the downstream direction of the dam. In addition, a fault oriented perpendicular to the river was discovered just downstream of the dam.

The foliation pattern of the foundation in combination with the presence of the fault and the forces associated with the water accumulating behind the dam caused the gneiss along the left abutment to enter a compressive state in which the permeability of the formation decreased with the increasing pressure behind the dam. Uplift pressure at the abutment caused by this phenomenon increased with the filling of the reservoir

The height of the dam was higher than the terrain on the left side, and so a thrust block was added to make up the difference. Uplift caused the thrust block to move, and the dam to fail.

A more complete report on the failure is available at the following link. The traps behind the failure of Malpasset arch dam, France, in 1959 I think this was written by someone whose first language was French. Or maybe it is just a very dry technical treatise on the geology of dam foundations. I could not get through all of it.

One of the things that paper makes clear, is that the assumptions of the day lulled the engineers and builders into feeling confident that the dam would last a long time. In other words, the moral of the story is: You don't find problems that you don't know to look for.

This Movietone News footage from the UK, shows images of the aftermath. (The footage of a dam bursting is, I think, lifted from a Hollywood film.)

Michigan Dam Restoration

I haven't had an infrastructure post in quite some time. Since there is quite a bit of infrastructure being rebuilt and/or upgraded in central Michigan, it seems like the time to have one.

On 19 May 2020, at 5:46 p.m Edenville dam in central Michigan failed during heavy rains. The resulting flood waters overtopped Sanford Dam, which subsequently failed. Two other dams in the area, Secord and Smallwood dams, also had minor failures, but the Edenville and Sanford dam failures did tremendous damage. There were no deaths or major injuries reported, but there was substantial property damage.

Previous posts on that incident include the following. The Failure of Edenville Dam and the Cascade Failure of Sanford Dam, which features Practical Engineering video going over what happened on the day in question. The 2nd post is 2020 Michigan Dam Failures: Forensic Report, that links to, and highlights, some aspect of that report.

Since 2020, repairs and upgrades to the dams have begun. I tripped over this story because the one-time owner of the Edenville dam, Mueller and Boyce Hydro, has been ordered to pay $199.8 million in penalties. That company went bankrupt after the dam failure, but whatever is left will go to the state, to "mitigate enviornmental impacts." Which I think means help pay for the dam replacement/upgrades.

What follows are a few videos from Jordan Mowbray, a resident of the area, who is documenting the restoration of the Four Lakes dams. The dams are being upgraded/reinforced with things like removing old hydroelectric powerhouses, and the construction of emergency spillways to protect the dams in the event of future major rainstorms to avoid the problems seen in 2020.

First up is HUGE TRENCHER! 🔥 [Soil-Concrete-Bentonite] Cutoff Wall Install at Edenville Dam! #construction #dam. (19 minutes) Installing a cutoff wall down the middle of the embankment will improve the quality of the dam and bring it up to modern safety standards.

Next we have Sanford Dam Flooding Today! - The Lakebed is Starting to REFILL! - Drone - Dam Collapse, (12 minutes) which shows the repair progress as of April 2023.

Secord dam was either damaged during the incident in 2020, or subsequent inspections revealed repairs that needed to be done immediately. Most of the information I've seen has been about Edenville and Sanford, and the catastrophic failure. In any event some work is being done on Secord Dam. An emergency spillway is being added, together with concrete reinforcement of the dam, and the old hydroelectric power station is being removed. The video on that is Secord Dam EMERGENCY SPILLWAY Progress! - Drone - Dam Collapse. (24 minutes)

Tempe Town Lake Dam: The Best Laid Plans...

If you design something, chances are it will not work exactly as you intended. This can be especially true after exposure to the elements. Tempe Town Lake Dam (Arizona, 2010) | Case Study

The dam failed on July 20, 2010.

Tempe Town Lake Dam consisted of rubber, inflatable dams on both the upstream and downstream sides of the lake. They could be inflated and deflated to capture the spring flooding and create a lake for recreation without flooding the areas upstream of the lake. The image at the right, which is from Wikimedia Commons, is of the dam from 2009, prior to the incident in 2010. Click on it for a larger view and more info.

In Arizona it is quite hot, and they were concerned that the rubber could not stand the heat, so they designed a system that would enable water to flow over the dams to cool them, while capturing that water, so as to lose very little during the dry season.

One important feature of the dam was an innovative water capture and recycling system incorporated into the stilling basin design. In order to protect the rubber bladders from the intense heat of the desert sun (sometimes reaching 150°F at the ground surface), water was allowed to flow over the bladders continuously. This water was then captured in the stilling basin and pumped back into the lake to provide a cooling system with minimal water loss.

There was a problem however; the dams were not perfect.

Unfortunately, this unique system did not function as intended. The rubber bladders sagged at the middle which did not allow for distribution of flow over the entire bladder. As part of the regular dam inspection process, it was found that the rubber bladders were deteriorating more quickly than anticipated as a result of exposure to extreme heat. In 2007, the bladder manufacturer evaluated the condition of the rubber dams and recommended replacing the bladders after 10 years of service.

You can see in the image above that water is only flowing over a portion of the rubber bladder.

They were supposed to be replaced in early 2010. That date got pushed to July of 2010.

At 9:44 PM on July 20, 2010, less than 12 hours before construction crews were scheduled to begin work on replacing the bladders, Bladder No. 2 at the downstream dam burst and immediately deflated. Water poured out into the empty river bottom, and the lake was essentially drained by noon the next day.

There was no loss of life.

The bladders were replaced temporarily with new bladders, but in 2016 they where replaced with a steel hydraulic system.

Ultimately, the rubber dam was replaced with a hydraulically operated steel gate dam located approximately 200 feet downstream of the original rubber dam structure. Construction of the replacement structure was started in 2014 and completed in 2016.

So what are the "lessons learned?"

What are the modes of failure?

Dam designers and inspectors should understand the vulnerabilities of the specific technologies or materials that are being employed. There may be unique potential failure modes that need to be addressed or considered as well as site-specific environmental factors (e.g. extreme heat and repetitive wave motion as was the case here, but also potential for ice loading, rigorous operational conditions, debris loading or clogging, etc.).

What are the non-critical aspects of the system?

In the case of Tempe Town Lake Dam, the dysfunction of the cooling system accelerated or even led to deterioration and failure of the rubber bladder. The dam could operate without this cooling system and did so successfully for several years. However, while the cooling system was not critical to the ability to impound water, it was vital to the long-term maintenance and integrity of the structure.

What are the long-term effect of the elements? (Water always wins.)

The hot Arizona sun and mild rocking action of the bladders worked away for more than a decade prior to failure of the dam. This progression of failure is not uncommon.

The final lesson boils down to "The Best-laid plans of Mice and Men, oft go awry." The problems were known. At least some of them were. Plans were put in place to replace the dam, but that still didn't stop a failure. In this case no one died, but infrastructure failures often do result in death, or at least a whole lot of property damage.

Taum Sauk Dam Failure

I don't usually cover the same event twice, but it has been a couple of years since I had an infrastructure post on the 2005 Taum Sauk dam failure. And I really like Practical Engineering videos, because I like the way Grady Hillhouse explains things for those of use who aren't engineers.

This is Practical Engineering's video The Wild Story of the Taum Sauk Dam Failure.

That previous post on this incident is Taum Sauk Dam Failure, at the archives. You will find the Dam Failures case study for Taum Sauk, and a link the PBS review of the event on the 10th anniversary of the failure.

A Slow Motion Dam Failure? I Guess We'll See

Starting last week, an earthen dam in Barnsdale, Oklahoma on Lake Waxhoma developed a leak. A fairly substantial leak. The powers that be declared a victory, but now it looks as though that may have been premature. (Water always wins.)

So first the story from last week. Barnsdall Mayor: Breach At Lake Waxhoma Dam Fixed Temporarily, Situation Under Control. There are a couple of news report videos at that link. The first one is less than 2 minutes and has some great images of the breach.

According to Mayor Kelley, once the hole was discovered, crews immediately started working to fill it and slow the flow of water coming through. Kelley said that crews brought 550 tons of large rock and material to build up around the mouth of the dam to help slow the water down.

The dam is rated "high hazard." That means if it did fail, there is a large potential for destruction. The lake contains about 2.5 million cubic meters or 660 million gallons of water.

After they added the rock and some dirt they declared victory.

Mayor Kelley said that due to the high amounts of rain that hit the area in the past few days, as well as the fact that the dam is designated a "high-hazard," it was a concerning situation when they first got word. However, Kelley stressed that there is no imminent danger to the area and officials do not believe the dam will collapse.

That was a week ago. This week, they are not so certain. City Of Barnsdall Pumping Water Out Of Lake Waxhoma In Effort To Relieve Pressure From Leaking Dam

Heavy rains have caused the temporary fix from last week to erode.

Mayor Johnny Kelley said that is just part of the reason why water is flowing over County Road 2331 and into a creek. Because of a weak spot in the dam, Kelley said water is flowing under the spillway, instead of over as is intended.

So they are pumping water out of the lake. Which will cause its own set of problems, as the lake appears to be a primary source of water for people in the area. And to make matters worse there is a boil order in effect for residents of the area.

The image above shows the spillway. As in other cases of dam failure, you can see vegetation growing out of the spillway. This is a maintenance issue.

Infrastructure Maintenance and Bureaucracy in the UK

This story is interesting to me, because of how it points out the universal nature of Red Tape and Bureaucracies. It also contrasts nicely with the Oroville spillway near-disaster and rebuild. Whaley Bridge: Plans submitted for permanent dam repair - BBC News

In August of 2019, the Toddbrook Reservoir in Whaley Bridge, Derbyshire, UK suffered a failure of the emergency or auxiliary spillway. The failure was very similar to what happened at Oroville Dam in California. A spillway, which had been inspected and passed as adequate, failed when high velocity water passed over the spillway. In the case of Oroville, they then elected to use the emergency spillway, which lead to severe erosion, and threatened a cascade failure the entire dam.

In the case of Toddbrook Reservoir, it could also have lead to a cascading failure of the dam. For how they dealt with the emergency in 2019, you can find some links at the bottom of this post.

While in the case of Oroville, both the main spillway and emergency spillways were both rebuilt and enhanced in about 2 years. The powers that be in the UK took more than 2 years to approve the plans.

The Canal and River Trust, which owns the dam, has applied to build a replacement for the 1970s spillway.

It estimates the project will cost between about £12m and £16m and take about two years to complete.

The plans include a new overflow side channel to prevent water levels reaching too high in the Toddbrook Reservoir.

Granted, the situation in Oroville was different, as they new they couldn't keep the reservoir empty in the face of California flood seasons. And Oroville Dam and its reservoir are key parts of the California Aqueduct system, without which Los Angeles would not have drinking water.

Here is a decent 11 minute video from Juan Browne of Blanclirio on what happened at the time of the failure.

For what I said about this incident at the time, you can see this link.

Rebuilding Infrastructure

After the failure of the Edenville and Sanford dams in Michigan in May of 2020, Jordan Mowbray took a page out of Juan Browne's book and started documenting the rebuilding efforts.

Secord Dam was determined to be at risk of the same kind of failure, and so a lot of rebuilding was required. The work on the Secord dam was halted in 2024, after completion of the emergency spillway, because of a lawsuit filed against Four Lakes Task Force, the organization created to own the 4 dams after the disaster. (2 dams failed, 2 were either damaged or deemed to be in need of repair. There may be more dams in the mix.) That lawsuit was completed, and the work on rebuilding the main spillway was restarted.

This is Jordan Mowbray's video Secord Dam WORK has RESTARTED! HUGE Progress! - Drone - Dam Collapse

Here are a few links to prior posts on the whole situation.

• The Failure of Edenville Dam and the Cascade Failure of Sanford Dam
• 2020 Michigan Dam Failures: Forensic Report
• We Do a Fair Job With Infrastructure When We Don't Ignore It

State's Actions in Michigan Dam Failures Are an Issue

The Edenville Dam failure is a subject of lawsuits. Of course. Judge allows 25 lawsuits against state of Michigan over 2020 dam failure to proceed

The Edenville dam failed in 2020, and the resulting flood took out a 2nd dam downriver, the Sanford Dam, and did serious damage to towns in the path of the water.

This was what I always considered to be the crucial point in the story.

Residents had argued in their lawsuits that the state not only failed to push dam owner Boyce Hydro to improve its spillway capacity, but that the state also pressured Boyce to raise lake levels during the winter out of concern for a native mussel.

Despite the fact that the dams were in bad shape. Despite the fact that new studies had SHOWN the spillways were incapable of handling a severe flood, the state refused to allow the owner to completely drain the lake, and "encouraged them" to raise water levels ahead of the spring rainy season.

Someone needs to be held accountable for those bone-headed decisions, though I doubt anyone will.

Infrastructure Rebuild in Michigan

When we build infrastructure, we don't do too bad as a technology-using species. We shouldn't wait for a disaster to fix things, but it often seems like that is all we do.

On 19 May 2020 at 17:46 EDT following a heavy rain storm, the earthen embankment of the Edenville Dam failed by way of static liquefaction. (See this excerpt from a video by Practical Engineering on static liquefaction; it is queued up to the relevant portion of the video.) The dam failure released enough water to overtop the Sanford Dam 16 kilometers downstream. While that dam was damaged it did not breach.

After some legal wrangling, bureaucratic finger pointing, and other nonsense, plans were put in place to replace the Edenville Dam, the Sanford Dam, the Smallwood Dam, and the Tobacco River Dam. A couple were damaged, all had been rated as being in very bad shape before the disaster, but politicians don't like to deal with infrastructure. It isn't flashy, and doesn't really help them buy votes from anyone. And so it is all too often replaced or repaired only after a disaster.

It is too complicated to relate the entire history of these dams, but the full video from Practical Engineering, not just the portion highlighted above, isn't a bad place to start.

This is Jordan Mowbray's video Edenville Dam Emergency Spillway Work Rapid Progress- Earthen Embankment - Dam Drone - Dam Collapse. Shows some of the current state of the rebuild of the dam.

Hidroituango Dam Failure

What happens when schedule and cost try to supersede engineering reality? Nothing good. (I haven't had an infrastructure post in a while; I thought it might be good to get back into that.)

Today we have the story of an engineering disaster from 2018. The Ituango Hydroelectric Dam in Colombia. Hidroituango, failed engineering, wrong decision-making, or simple ambition?

Note: Colombia is the name of a South American country. Columbia is the name of a record club from the 1970s that could get you cheap vinyl, but was really difficult to quit. For this reason, most of the articles on the subject are in Spanish. Most of the those that have been translated into English have some problems.

The project in question is a hydroelectric dam on the Cauca River near the city of Ituango in Antioquia Department, Colombia. There are 32 departments in Colombia, and while it is tempting to think of them as states, they are just administrative divisions.

Three diversion tunnels were dug to route the river around the site of the dam and powerhouse. The original design called for there to be a sluice gate system for regulating the flow through the tunnels. Those gates were not constructed.

Near the end of construction, 2 of the 3 diversion tunnels were sealed by the construction company. It isn't clear to me why that was done. During construction of the powerhouse, the third tunnel was blocked by a landslide and the reservoir began to fill. (Some reports call that landslide an internal collapse of the tunnel; either would be bad.)

Further landslides followed and with water levels rising and all outlets now blocked, EPM decided to flood the dam's turbine rooms on 10 May to release the river's increasing pressure on the structure.

While the move caused irreparable damage to transformers and other systems which had already been installed in the turbine room, it did reduce water levels.

But only two days later, one of the sealed tunnels naturally unblocked, causing water to gush through it and creating flash floods further downstream.

Since then studies indicate that there is a risk of more landslides, which could block both the diversion tunnels and the intake for the powerhouse. If that happens there will be no way to drain the reservoir, and it will eventually overtop the dam. This would ultimately result in a catastrophic failure of the dam with a full reservoir.

Why were the controls on the diversion tunnels not built? Why were the diversion tunnels sealed? Why was the geology not better understood? (Landslides and failed seals?) The answer seems to be budgets and timelines. Though most of the reports are in government-speak that has been translated from Spanish.

Here is a key paragraph from another source. Hidroituango or: The greatest project nobody wants to take responsibility for

It is evident by now that local stakeholders made changes in the designs of the project in order to finish it on time and that these changes resulted in a series of engineering mishaps that caused the current emergency. In their desire to see the project through and their ambition to demonstrate the prowess of Antioquian engineering they forgot to incorporate lessons learned from past projects. They have caused incalculable environmental damage to the Cauca river basin, displaced local communities, endangered the finances of the region (and possibly the country). They have also brought about energy uncertainty. Although investigations into possible wrongdoing have begun, it is very unlikely powerful people will be held accountable for the damage they have caused.

This is the best report I've found, though it is not "official" in the way most reports are. EPM’s Hidroituango: New Report By Independent Geological Engineering Expert Challenges Existing Narrative. It seems to highlight some of the problems both with the design and the construction of the diversion tunnel.

This is a short video that captured the moment (just after) the seals on the diversion tunnels failed. (The video I wanted to use has disappeared, but this one isn't bad...) Fail Hidroituango Colombia Dam Hydroelectric

Maintaining Infrastructure

What it takes to maintain and rebuild infrastructure. Not sexy or exciting, just necessary for modern life to be maintained. Politicians love to ignore it, and then are shocked to discover that it doesn't last forever.

In May of 2020 during a substantial storm, the Edenville Dam in Michigan breached. This caused the Sanford Dam to overflow, and also be damaged. Several other dams in the area were also damaged in the storm, including the Forest Lake Dam near Omer, Michigan. While the dam held, the spillway was extensively damaged, and had to be rebuilt.

This is a relatively small dam, as dams go. It is 48 feet high, 800 feet long with a storage capacity of 7,440 acre-feet of water.

The video below is long at 25 minutes, but the first 5 or 10 minutes shows the difference in construction of the spillway after the repairs, and as originally built, and there is a section at about 3 minutes in that shows the damage to spillway after the 2020 storms. I find the upgrades to slow the energy in the flowing water to be fascinating, but then I like this kind of engineering.

This is Jordan Mowbray's video Forest Lake Dam Complete! - Dam is Repaired and Forest Lake is Full!

Most of the news articles from 2020 that I had linked in various posts are either gone, or moved behind paywalls. Here is a post that includes an engineer's video of what happened on 19 May 2020. The Failure of Edenville Dam and the Cascade Failure of Sanford Dam

Another Dam, Another Failure - This time of a spillway gate

It hasn't been a catastrophic failure. But a failure nonetheless. Lake Gonzales’ dam spill gate failure brings uncertainty for residents, wildlife and environment

Officials with the Guadalupe Blanco River Authority(GBRA), which manages regional waterways, say a large tree became lodged in the spill gate on Aug. 3. The weight of the tree and the resulting water pressure may have led to the spill gate’s failure.

The lake is 12 miles west of Gonzales, or about an hour’s drive east of downtown San Antonio.

And there is a statistic that is worth taking note of.

A sobering line from the magazine’s investigation: “Of the approximately 300 dam failures in Texas since 1910, half have occurred in the last nine years.”

Every dam in the world needs maintenance. And eventually they will probably all need replacement, or at least very serious work. Nothing lasts forever.

Rebuilding Oroville Dam After the Near-disaster

Though I posted on Oroville Dam in past years, the story of how it happened and how the problem was solved never stops being fascinating. At least to me.

On February 7th, 2017 while lowering the water level of the reservoir, the main spillway experienced a failure. This presented as an "unusual" disturbance pattern in the water down the spillway. The gates were closed to see what was going on and it was clear that the spillway was failing. I want to say catastrophically, but they did manage to avoid a full scale catastrophe of a dam failure. I think it was a near miss.

While they were trying to understand what was going on with the main spillway, the water level rose to above the level of the emergency spillway. On February 11th water poured over the emergency spillway. That was the first time that had happened in the history of the dam. The land in front of the emergency spillway began to erode immediately. 200,000 people were evacuated, and eventually the main spillway was reopened. You can find more on the near disaster at the following link. The Oroville Dam Near-disaster: The Cost of Ignoring Infrastructure from the 357 Magnum Archives. It was posted in 2019

Anyway, Practical Engineering, a YouTube channel I quite like, recently published a good, 20 minute video on the rebuilding of the main spillway and the emergency spillway.

This is the video from Practical Engineering on Rebuilding the Oroville Dam Spillways.

More Infrastructure Being Updated

Usually you only read about dams when one has failed. For a change, here is a dam that is being replaced. New Ossipee Lake Dam celebrated in Freedom

The existing dam is 135 years old, and showing signs of failure.

The upgrade was about 20 years in the making.

[DES Water Division Director Thomas] O’Donovan recalled the major flood of 1998 which he said caused $1 million worth of damage around the lake, which he said was due to the “inability to manage the impoundment correctly.”

Plans for the upgrade didn't start in 1998, but did start shortly after that in 2002. It took that long to get the design done, and approved and funding in place.

The new dam has a safer, modern spillway, has a larger flow capacity to reduce flooding and has automated spillway gates.

The old dam relies on stop logs which have to be added and removed manually, though they are currently jammed, making it impossible to manage the water level.

I can't find a decent photo of the existing dam that is in the public domain. Click the link at the top of this post for a view of the dam under construction.

New Hampshire has a couple of dozen "high hazard" dams that need attention. High Hazard refers to the potential of loss-of-life and economic impact in the event a failure, not a statement on the condition of the dam.

Lake Dunlap Dam To Be Rebuilt

The old dam collapsed, on camera (see the video below) in 2019, draining the lake. The infrastructure will be rebuilt. That is almost amazing today. Construction to begin on new Lake Dunlap Dam any day.

And now those dreams are finally a reality, a new dam for Lake Dunlap with an estimated price tag of $35 million.

Doug Harrison is President of the Lake Dunlap Water Control and Improvement District and said the funding for the dam came from the state.

“GBRA has taken out of $40 million loan from the state at virtually zero percent interest that we'll pay back over 30 years,” said Harrison.

The people impacted by the failure will pay to replace the dam. That seems fair, though I'm not sure what the real interest rate should be today. The news story says that they will pay "virtually zero percent." The new dam will also create hydroelectric power, which will contribute to repaying the loan.

The news story linked above has a video that includes an artist's rendering of the new dam, and a view of the current state of the lake. It is a bit more than 2 minutes, though more videos will play if you let them.

Here is video of the collapse. Video shows moment dam gate collapsed at Lake Dunlap. The gate collapse lowered the water levels by 14 feet in the lake.

2020 Michigan Dam Failures: Forensic Report

While we're on the subject of dam failures... Michigan Dam Failures Caused by Rapid Liquefaction of Downstream Sand Embankments | 2021-09-14 | Engineering News-Record

First, let's refresh your memory.

The Edenville and Sanford Dams along the Tittabawassee River in north central Michigan failed on May 19, 2020, after two days of heavy rain, resulting in flooding and property damage.

There was an ongoing dispute between the dams owner, the federal government (in the form of the regulating body) with some interaction with the local community. Oh, and the state environmental agency was also involved, if memory serves me.

An interim report from the Association of State Dam Safety Officials reports that a clear factor in the failures of two dams built in the 1920s was static liquefaction (flow) instability of saturated, loose sands in the downstream section of their embankments. [SNIP]

Static liquefaction has rarely been observed in dam failures and is more commonly associated with earthquakes, according to the forensic team. It occurs when the mobilized shear strength in saturated, loose sand decreases rapidly to values significantly less than the applied static shear stresses, resulting in a force imbalance. The sand quickly liquefied, became mud and then accelerated the velocity of the already fast-moving water with its mass.

There is more at the link above. The complete interium report can be found at the following link: INVESTIGATION OF FAILURES OF EDENVILLE AND SANFORD DAMS. It is a 42 page PDF, so you may not want to download it to your phone. It does contain a number of photos of the dam at various points leading up to the failure, and as the failure unfolded.

The Failure of Edenville Dam and the Cascade Failure of Sanford Dam

It has been a while since the dams in Michigan failed, affecting a lot of Michigan, and destroying a lot of homes and businesses. It was May 19, 2020, at 5:46 p.m. due to flooding from a large rain storm, a portion of the earthen embankment underwent static liquefaction. I hadn't heard of that before this incident. Liquefaction is not uncommon after an earthquake - there are some amazing/horrifying videos out there if you look. As the name implies, what was solid ground suddenly starts behaving like a liquid. As you can imagine this is bad news for the things built on top of that dirt, or built out of dirt, like an earthen embankment dam. All of which is explained in the video below.

One of the things that has been noted, and is mentioned again in the video, is that the people who mostly enjoy the benefits of these old dams, the people who own property around the man-made lakes, are not the people paying for their upkeep. Not usually. There have been a couple of instances where the property owners upstream have accepted financial responsibility for the dams, but usually only for rebuilding after a disaster. At least those are the instances I've seen. In some cases the dams are owned and maintained by municipalities, but that is not usually the case for these old hydroelectric dams.

After the license to generate power for the Edenville dam was revoked by the Federal Energy Regulatory Commission (FERC), there was a project underway to transfer ownership to a local governmental agency so that tax dollars could be used to maintain the dams and the lakes. That was supposed to transfer ownership in 2022.

This is the Practical Engineering video What Really Happened At Edenville and Sanford Dams?

And as I have mentioned before, we need to do a better job of maintaining these dams, or mitigating the potential problems. This is an extreme case. And is was made worse by the fact that the Michigan department in charge of environmental protection wouldn't let the dam owner drain the lake to preserve some habitat created when the dam was built, and to protect some species of muscle or snail or whatever. So, how'd that work out?

Liulan Reservoir Dam Failure in China

China apparently has quite a few dams that were built during the 1950s, that are not up to modern standards. Watch: Typhoon Maysak breaks dam wall in China, kills 2 and forces thousands to evacuate

Tropical Storm Maysak, initially brought days of extreme rainfall across the country's southern Guangxi Zhuang Autonomous region. But now, the typhoon has triggered a massive flood that claimed two lives, affecting about 55,000 people in Nanning and forcing around 48,000 of the city's residents to evacuate

The dam was breached on 6 July.

Maysak has triggered multiple reservoir emergencies, increased river levels above warning levels and flooded several villages across Guangxi.

Some info on the dam. Liulan Reservoir: A medium-sized reservoir located in Hengzhou City, Nanning.

The construction of the Liulan Reservoir hub project began in August 1958 and was largely completed by 1960. A hydropower station was added at the dam's toe in 1974. In August 2024, the Nanning Municipal Government Service Bureau issued an approval regarding the implementation plan for the standardized management creation of Liulan Reservoir in Hengzhou City . In 2025, the Liulan Reservoir Irrigation District Continuation and Modernization Renovation Project was also accelerating its progress, with a total estimated investment of 382.53 million yuan for the project.

Translation: They realized that the dam need improvements, and began work. That work was scheduled to be complete in 2027, but they didn't make it.

I can't find the reference now, but I read somewhere that China has a large number of infrastructure projects done in the early days of communist regieme, that were not really engineered. That process continued for a long time. They are now dealing with infrastructure, in some parts of the country, that are not what we would consider safe.

Guangxi Zhuang Autonomous region is in the south of China, with a coast on the Beibu Gulf (off the South China Sea). It shares a border with Vietnam

NOW: A devastating situation is unfolding as Liulan Reservoir collapses following heavy rainfall from Typhoon Maysak in Hengzhou, Nanning, Guangxi, China. pic.twitter.com/vyuQn5nuyl

— Weather Monitor (@WeatherMonitors) July 6, 2026