How long are dams like Hoover Dam engineered to last? (2006) | Hacker News Reader

How long are dams like Hoover Dam engineered to last? (2006)

straightdope.com

79 pointsmonort39 comments

For those who wanted the answer to the title in these comments, paraphrased it was: with proper maintenance, indefinitely. The "with proper maintenance" phrase should be pointed out to management next time they think something we build should last forever; even these remarkable structures that appear formidably unchanging undergo continuous, hands-on, expert maintenance.

There is a tangentially-related Straight Dope thread [1] asking "Is the Hoover Dam concrete still curing?", because if you take the tour (highly recommended for hackers with an affinity for industrial- and megastructure-scale engineering), you will be told "the concrete is still curing!" by the guides.

This is apparently not technically precise; someone from the concrete industry please correct me if I'm wrong here. I gathered from a civil engineering page that Hoover Dam's concrete is considered cured, but it is still undergoing the hydration process as long as water is present in the concrete. [2] I think the Hoover Dam guides can be cut some slack for colloquially using the terminology "curing" in light of this technical distinction that most laypeople won't understand.

[1] http://boards.straightdope.com/sdmb/showthread.php?t=258217

[2] http://www.engr.psu.edu/ce/courses/ce584/concrete/library/co...

As you probably gathered from your references, concrete curing is a continuous process - every time a stray water molecule binds with some cement the concrete will get a little bit stronger. We use 28 days to call the concrete 'cured' because it's a reasonable time on construction schedules (why it's in an even number of weeks) and hits a nice fat part of the exponential strength curve. It's a bit like asking when a pile of plutonium is done decaying. What will happen is that at some point, the strength gains are so marginal that various degradation processes will cause the concrete to start losing overall strength.

But - the idea that the concrete is curing and will be finished in 50 years is fairly goofy. No one is doing high-temperature long-duration concrete curing research, but the idea that you can extrapolate the laws of heat transfer to figure out how long some bubble of superheated wet concrete in the middle of the dam will last isn't very reasonable.

Fun fact on the topic of curing, the Hoover Dam was built with chilled water coolant lines running throughout the bulk of the dam in order to dissipate heat through the curing process. Without cooling, the sheer mass of the dam and concrete's exothermic curing process would have basically caused the dam to self destruct from expansion/contraction.

http://www.usbr.gov/lc/hooverdam/history/essays/concrete.htm...

    With a small dam the water can sometimes be drained,
    allowing repairs to be done safely. But imagine trying to 
    drain the Hoover Dam to repair cracks at the bottom--it 
    can't be done

I don't understand why they think it's a problem to drain the dam. It's done regularly to check and repair them. (Every 30/40 years, at least in France).

Recent video of this in France : https://www.youtube.com/watch?v=JflFg5un5zg

It is this dam, this is not a small one : https://fr.wikipedia.org/wiki/Barrage_de_Sarrans

According to their respective Wikipedia articles, the lake behind the Hoover Dam has a volume 108x larger than the lake at Sarrans. [1] It's unclear whether the same techniques apply.

[1] http://www.wolframalpha.com/input/?i=32.22+km^3+%2F+296*10^6...

IMHO, the main parameter seems to be how many time is needed to fill/empty the dam and to apply a reasonable ratio to ensure the dam is working 95% of the time.

From Wikipedia on Hoover Dam: "Filling of Lake Mead began February 1, 1935, ... ... In the latter half of 1936, water levels in Lake Mead were high enough to permit power generation ..."

To empty Hoover Dam once every 60/80 years doesn't sound absurd to me.

Give us a few years, we'll drain it for you.

--Love, the Western US.

Based on my last visit to the Hoover dam, draining it would be little or no problem at all these days. Water levels are incredibly low.

Caveat that it would require trucking in water to Las Vegas it was suggested to take advantage of the drought to work on the sediment levels and do potential maintenance. Not something you can just decide on the spur of the moment though.

A sadder thing is that while the Marinas that are now on dry land have a chance to do some upgrades/maintenance, they have no cash since they are very day to day based on the cash inflow from slip fees. That is a tragedy I think, you could completely rebuild the piers and develop a much more durable infrastructure for future use I would expect.

I'm surprised this isn't an underwater robotics project.

By last count (few years old), the number of dams was estimated at ~84,000 (not 76,000). Close, but that numerical discrepancy should point out one thing: it's an estimate. USACE maintains an authoritative database of dams in USA (and also 32 in Puerto Rico and 1 in Guam) and the numbers are always a moving target. Believe at this point they're working with pattern matching and satellite imagery to fine-tune the database. USACE is currently more concerned with inspections near urban populations than cataloging.

The write-up is a bit vague and misses one major point: not all of these dams in the U.S. are concrete. Not even close. Here's a construction type breakdown:

  70,278 Earth
   9,031 (Unknown)
   1,446 RCC
   1,215 Gravity
     724 Concrete
     455 Other
     420 Rockfill
     201 Masonry
     136 Buttress
     103 Arch
      58 Stone
      51 Timber Crib
      16 Multi-Arch

Only 847 of those dams are tailings dams, used for industrial processing. More than 0, but less than 84,000. Of those, none are listed as concrete (730 Earth, 61 Unknown, 43 Other, 13 Rockfill).

> The write-up is a bit vague and misses one major point: not all of these dams in the U.S. are concrete

The write-up is in response to a question asking how long dams like the Hoover Dam, which is a large concrete dam, are designed to last. It wasn't suggesting that all dams are concrete.

The single sentence: "Indefinitely with regular maintenance" is a rather short article length, so Straight Dope decided to educate their readers that the large dams, which make great action-packed movies when they fail catastrophically, are the very least of our worries since earthen or other natural material based man-made dams, which generally have significantly less water capacity, are in fact at more risk of failure and potentially even more damage (like the cascading failures in SC this past year, or the Banqiao Dam incident).

> The write-up is in response to a question asking how long dams like the Hoover Dam, which is a large concrete dam, are designed to last. It wasn't suggesting that all dams are concrete.

But the author mixes concrete dams with tailings dams, of which none (according to USACE) are concrete. That's what I meant by vague: starts with a large large number talking about US dams with, talks about concrete, mentions danger of tailings and then talks about foreign dams. The data is chosen for a narrative and ends up implying a lot without clear references.

That number is nowhere near accurate as we learned during the South Carolina floods back in October. The state estimates there are between 10,000 and 20,000 unregulated dams in South Carolina alone: http://www.thestate.com/news/local/article41315256.html

Those are the dams that pose the greatest threat. You can't plan for something you don't know about and the threat posed by such an extensive system of "invisible" dams is immense. A small earthen dam on a farm may not pose much of a threat by itself but its failure can cause a cascade of failures further downstream if it breaks when those dams are already at their limit.

So, in South Carolina, The State reports that the state estimates that there are between 10,000 and 20,000 dams in the state which are not regulated by the state.

You'd think that a publication called 'The State' would be a lot more careful about throwing that term around, but within that article, it talks about "the state" meaning South Carolina as a geographical area, "the state" meaning the SC government as a regulatory authority, "the state" meaning government actors in general, and "the state" as a newspaper (though they do refer to themselves as "The State newspaper". Presumably they aren't claiming to be the official state newspaper, just a newspaper called 'The State').

Anecdotal story: This summer I was out at a Dam spillway installing some of my systems and got talking with the operator out there. He was commenting about the engineering of the facility and how impressive it was for the age it was built (50+ years ago) in comparison to the modern dam my government was currently building and all the lawsuits and issues its has even before opening. To him, at least, and I'd agree from his comments, business got too involved with engineering. They started running too many cost estimates and too many lowest-bidder contracts instead of building things for a century plus lifetime.

Shame to see that. I hope that our engineers continue to fight for proper surveys and designs rather than a race to the bottom.

For reference, ~100 people died in terrible conditions building the Hoover Dam. It was a tremendously unsafe worksite. The environmental impact of the project was barely considered.

I think all things considered we've made good progress.

On a relative scale the number of accidents building the Hoover dam was high but not spectacularly high when compared to the number of people dying in high-rise construction and road building.

The Hoover Dam hasn't even properly set yet!

Curing is an exponentially decreasing process, there will always be a little bit more concrete that still needs to cure, much like radioactivity half-life.

The Hoover Dam is quite an interesting piece of civil engineering, it broke ground in many ways other than the physical one.

Tons of interesting stuff here:

http://www.usbr.gov/lc/hooverdam/history/essays/concrete.htm...

Concrete, while on the surface very boring is actually a super interesting engineering material.

Another interesting tidbit: for the longest time the dam was actively cooled to whisk away the heat from the curing concrete (concrete curing is an exothermic process).

The threat posed by small dams and their short designed life expectancy was demonstrated in dramatic fashion earlier this year. The historic flooding that impacted South Carolina's midlands earlier this year was driven by a cascade of dam failures.

Everyone saw the pictures of Columbia, SC but that flooding was mostly due to "normal" causes. The city's drainage system was overwhelmed by rainfall that exceeded historic levels by several orders of magnitude (higher than even a 1,000 year storm) and the Broad river consequently overflowing its banks. However, that wasn't the only place impacted by the storm. South Carolina has an enormous number of small earthen dams, built on private land with the land owner responsible for maintenance.

Obviously, these dams were not built to handle such a storm but even if they were, the lack of oversight led to a lack of maintenance, worsening their chances of surviving the initial floods. As upstream dams broke, downstream dams were subjected to massive surges, causing a cascading series of failures, greatly increasing the amount of damage caused by the storm.

There are 10,000 to 20,000 unregulated small dams in the state according to the state. They pose an enormous risk as the flooding clearly demonstrated but SC isn't alone in sharing this burden.

Accounting for failure frequency due to natural disasters occurs at the design (initial or retrofit) stage, not the maintenance stage. Should every mom-and-pop dam adhere to the once-in-ten-thousand-year-rule (eg: nuclear powerplants) or once-every-hundred-years (eg: steel framed commercial construction)? Cost effectiveness is also a huge concern.

Note that maintenance doesn't increase the expected failure frequency, it maintains it at the design spec. And I'm willing to bet that even with proper maintenance those small dams in SC would not have survived; they were way outside their designed specs.

I drive past Folsom dam (1952) every day and have watched the incremental maintenance as well as a few large scale projects progress first hand. One of the most amazing things to witness was the construction of the spillway they have been building around the outer perimeter. They essentially are creating a second dam around the main dam to release water from a lower point due to a scare many years ago. Months of sorting dirt and rock at some kind of large facility nearby and a good year of constant trucks moving the dirt down the road. Even the asphalt on the road began to sink from the constant pressure of the heavy loads.

A few of the old spillway gate failures over the years are outlined here:

https://en.m.wikipedia.org/wiki/Folsom_Dam

Currently, the Kariba dam is the one most likely to fail:

http://citizen.co.za/820414/kariba-dam-a-ticking-time-bomb/

How long a dam lasts and how long the reservoir behind it is usable are two different things. A well constructed dam may last 100+ years but if the water flowing into the lake carries heavy sediment loads, the lake will fill up with mud before then if not periodically drained and the muck removed. This is definitely an issue with Three Gorges as well as Lake Powell/Glen Canyon Dam on the Colorado River. In fact, one of the reasons the latter dam was constructed was to act as a catch basin so the downstream Lake Mead behind Hoover dam wouldn't fill so quickly.

Sediment loads aren't really a problem. The sediment drops out as soon as the water goes slack so it all accumulates at the far upstream section of the reservoir and raises the river level rather than filling in much capacity.

The Grand Canyon is filling in with Hoover Dam sediment while the reservoir remains unfilled. In fact, the Grand Canyon is now a mud lagoon starting above Separation Canyon and the water below Pearce Ferry has very little sediment. The upstream sediment slackwater continues to migrate upcanyon, not downcanyon.

On Glen Canyon, the sediment accumulated 40 meters deep over Hite but did not slip downcanyon. Instead, Cataract Canyon continues to fill in upstream. Moab will be covered in Glen Canyon's reservoir before any serious capacity is lost because sedimentation moves upriver, just as in any ocean delta.

The same effect has stopped the main flow in the San Juan and covered all the rapids up to Slickhorn and even above.

---

Above I said, "sediment loads aren't really a problem." I mean they aren't a problem for dam operators. They're bad for the environment, of course.

But that should be obvious. For all the selling of hydroelectric as green and sustainable, we should know that it's the most destructive and damaging of all electric generation technologies. Even a failed nuclear plant in full meltdown isn't as destructive and harmful as a properly operating hydroelectric dam. Hoover dam makes Tchernobyl look good.

Dredging can solve that problem, it's just a question of what is considered reasonable maintenance.

Worth a read:

https://en.wikipedia.org/wiki/Risks_to_the_Glen_Canyon_Dam

This gives a hint as to why huge dams cannot be drained for maintenance: there is so much water behind them that it would take too long to drain them via the spillways. In fact, the spillways would probably be irreparably damaged by experiencing full flow for that long.

"Engineers, about 30 percent of the more than 76,000 dams in the United States are older than 50 years--and by 2020, that number will increase to more than 80 percent. That's a lot of old dams, some of which hold back not just water but toxic sediments from early industrial operations."

See also the TVA's Kingston plant spill:

https://en.wikipedia.org/wiki/Kingston_Fossil_Plant_coal_fly...

It wasn't clear to me why having toxic sediment sitting behind the dam was that much better than it moving downstream?

Remediation is always more expensive and less effective when treating across a diffuse area (which happens when moving the toxic sediment downstream) than in a single contained area.

As a civilization we're gradually getting better over time at remediation of large-scale "waste" like this, but still primitive considering our state of the art capabilities; we still don't recognize at a systemic, industrial design level that waste is materiel that we don't understand an economic use for yet, or understand how to transform into materiel we want at a thermodynamically desirable cost.

Do you want it 100 km away or in your garden and living room?

100 km away is another family's garden and living room.

They're living at the bottom of Lake Mead?

(cue the Yorkshiremen sketch https://www.youtube.com/watch?v=Xe1a1wHxTyo )

at least with it piled up behind the dam you could theoretically say its contained. Of course you could always end up with it down stream and out of control like Brazil recently faced when a dam for a mining company broke.

With the right equipment you could just suck up the sediment and then isolate it further. Having watched more than one youtube video of river restoration where they blow the dam from below that amount of gunk flying down stream can never be good

The toxic sediments are buried. If the water were to flow, the sediment would get kicked up.

Some Roman concrete like Pantheon Dome is still good after 1900 years. The Dome was poured in 126 C.E.

I remember hearing somewhere, perhaps on a tour of the Hoover Dam that the concrete is very good for the time because although the ideal chemical properties of concrete were not well understood at the time, the local aggregates used were relatively pH neutral which has contributed to the dams longevity. I'm trying to find a source for this though and struggling. Has anyone else heard this about the dam?

Don't know about pH but very interesting read on history primarily from a civil engineering perspective - http://web.mst.edu/~rogersda/hoover_dam/Civil_Eng-Nov-2010-H...

Reply on news.ycombinator.com