Chernobyl disaster and Matthew Walker’s “Why We Sleep”

Alexey Smirnov writes:

I was outraged with Walker’s claim that Chernobyl disaster was caused by the operators’ lack of sleep (do not remember exact page).

I happened to live in Ukraine during the period, and the subject was somewhat of a painful area of memory.

While the book up to that point seemed fresh and interesting, this particular claim had dropped trustworthiness of the author to a car sales level.

I do not assume you care all that much about details, and majority of people around the world got pretty much as much details as they care about from HBO miniseries, but just in case, for your reference.

(1) There was a reactor designer who did not correct design despite multiple prior accidents, possibly due to extra cost being stopped by Soviet authorities.

(2) Responsibility of (1) for disaster was brought up by a scientist familiar with the matter.

(3) More than 10 years prior to Chernobyl there was a serious accident similar in causes on a reactor of the same design.

(4) The scientist who investigated disaster, Prof. Legasov, knew it was faulty design.

(5) Top Soviet nuclear scientist actively suppressed descent in analysis, forcing (2) to retire (details in Russian version).

Lack of sleep caused Chernobyl … never let facts get in the way of good commerce …

I looked up Why We Sleep online and searched for Chernobyl and found this very short passage:

All he says is “There is speculation that . . .” with no reference. I guess it’s good that Walker has no references, given his well-documented problems with the literature in the past (see here, here, and here, for example).

But then I was curious: what was this speculation, exactly? Who was doing this speculating? So I did what anyone would do, I googled *Chernobyl Matthew Walker sleep* and just plain old *Chernobyl sleep*.

The first search led to this journalistic puff piece with the phrase, “Other researchers have found that severe sleep deprivation impairs people’s ability to follow preestablished procedures for making a “go” or “no-go” decision, something that researchers say contributed to the explosion of the space shuttle Challenger, the Chernobyl meltdown, and the Exxon Valdez disaster.” So not it’s gone from “speculation” to “researchers say.” But the linked paper describes a lab experiment on 26 people; nothing about Chernobyl. Another link went to this book review which flatly says, “Some of the worst disasters in recorded history have been caused by lack of sleep. Both the Exxon Valdez and Chernobyl were caused by sleepy operators.” But again with no references.

The second search led to this research article, “Catastrophes, Sleep, and Public Policy: Consensus Report,” which reports:

Perhaps most disturbing is the fact that the nuclear plant catastrophe at Chernobyl is officially acknowledged to have begun at 1:23 a.m. as the result of human error (39,40). The limited amount of information currently available about the human factors component of this accident, however, makes it difficult to draw any firm conclusions about the contribution of sleep-related errors in performance or judgment.

The two references are:

39. International Nuclear Safety Advisory Group. Summary report on the post accident review meeting on the Chernobyl accident (25–29 August, 1986; Vienna, Austria). International Atomic Energy Agency, November 1986. ISBN #92-0-123186-5

40. U.S. Nuclear Regulatory Commission. Report on the accident at the Chernobyl nuclear power station. January, 1987, NTIS NUREG 1250.

The reports are online. Here’s the report from the IAEA (you have to pay 24 euros for the actual report), and here’s the NRC report, which has a pdf link on the webpage but the link doesn’t seem to work. I was able to find this NRC report, “Implications of the Accident at Chernobyl for Safety Regulation of Commercial Nuclear in the United States,” but it didn’t mention sleep at all.

Going back to the Google search, we find this page which states, “Investigators concluded that fatigue – due to 13-hour shifts – was a leading contributor to the human error that led to the explosion. (A flawed reactor design was also to blame.)” That link didn’t work but I found it on the Internet Archive, and here’s the quote: “The power plant exploded after engineers had worked 13 hours or more. Investigators ruled that sleep deprivation was behind the accident that led to the explosion.” No documentation, though.

The Google search also revealed this NPR report, “Short On Sleep? You Could Be A Disaster Waiting To Happen,” which states, “Sleep-deprived operators failed to prevent the Chernobyl nuclear power plant meltdown and the Exxon Valdez oil spill”—but with no reference to the claim that the Chernobyl operators were sleep deprived. There’s just a link to the aforementioned experiment on 26 people. The search also turns up this Guardian article pointing to a report, “The Impact of Sleep Deprivation on Decision Making,” that “blames sleeplessness and long shifts for ‘human errors’ for nuclear disasters from Chernobyl to Three Mile Island . . .”

OK, now maybe we’re getting somewhere! Here’s the research article, it’s by Yvonne Harrison and James Horne, published in 2000 in the Journal of Experimental Psychology: Applied, and here’s what it says about Chernobyl:

It is perhaps just a coincidence that some of the most renowned man-made disasters or near disasters concerning nuclear power plants, such as Chernobyl, Three Mile Island, Davis–Beese (Ohio), and Rancho Seco (Sacramento), all occurred in the early morning and involved human error in failing to contain otherwise controllable but unexpected and unusual mechanical or control room malfunctions. With all four, experienced control room managers misdiagnosed and failed to appreciate the extent of the fault and then embarked on courses of action that were inappropriate and continued to persevere in this way in spite of clear indications that their original assessment was wrong.

That’s it!

“It is perhaps just a coincidence . . .”

After all that googling, what a letdown.

What’s interesting is that, with pure speculation and zero evidence, other than the fact that the disaster occurred at night, this sleep/Chernobyl claim is so often taken as fact. Props to Walker for labeling this as speculation; anti-props to Walker for writing it as if it’s a real thing. Also anti-props for him changing “It is perhaps just a coincidence” in that earlier article to “without coincidence” in his book. It’s not even clear to me that “occurring, without coincidence” even means. What is clear is that Walker wants to attribute radiation that was “one hundred times more powerful than the atomic bombs dropped in World War II” to sleep deprivation. He’s on a mission, and lack of evidence isn’t gonna stop him.

40 thoughts on “Chernobyl disaster and Matthew Walker’s “Why We Sleep”

    • But time spent drinking can’t be time spent sleeping. Therefore, even if drunkenness is the cause, sleep deprivation is *really* the cause. Q.E.D.!

      But seriously: did Walker actually look into any of these cases, beyond superficial third-hand reports? It doesn’t look like it, which is awful.

  1. That’s one of the problems with talking about the causes of effects, right?
    It’s definitely believable that sleep deprivation impaired the operators’ decision making, but any system that could fail so catastrophically because one person is sleep deprived has major structural problems to begin with–the sorts of things Smirnov described.

    • Adam:

      Agreed. At the very least it should be possible to do a qualitative analysis, for example interviewing people and asking whether certain people were sleep deprived, etc. One of my problems with Walker’s claim is that it doesn’t seem to be supported by the investigators who looked into the Chernobyl event. Or, at least, it didn’t seem to make it into those reports.

    • Adam said, “That’s one of the problems with talking about the causes of effects, right?”

      To be more specific, “one of the problems with taking about the causes of effects” is that all too often we look for a single cause, when there often may be a combination of causes. That’s what seems to be the case in the Chernobyl incident: If the reactor had been better designed, the accident wouldn’t have happened; if the operator hadn’t been sleep-deprived, the accident probably wouldn’t have happened.

      • “all too often we look for a single cause”

        That’s right. When mature mechanical or electrical systems fail, they fail for a single root cause and (if you find the cause) any other damage can safely be assumed to have been cascading damage. But it doesn’t make sense to try to identify a single root cause of a process failure, and this was a process failure. Generally when processes fail, they fail at the weakest link. When designing fail-safe systems, the last, worst choice for mitigation measures is human intervention, which is only exploited if no other mitigation strategy is reasonable.

        After Chernobyl, reactor design finally progressed to fail-safe, but it was too late. In modern reactors, if something goes wrong, the fire goes out.

        • Well, sometimes you get something like a tsunami that makes a bunch of systems fail at once. Is that one cause, or several?

        • Did the systems really fail at once? The system flooded, a fault occurred, and the system stopped. That was the single failure.

          The simple point about the single failure is that it is surpassingly rare for two rare events to happen at exactly the same time. Think of your car quitting while you are driving. Could two independent failures really have occurred at exactly the same instant?

  2. “What’s interesting is that, with pure speculation and zero evidence… claim is so often taken as fact.”

    … that’s apparently the bottom line of this lonnng story, but people making unfounded assertions is extremely common in all sectors of human society, even among supposed experts in various areas.
    Perhaps there is a more focused bottom line concerning the merits of general skepticism.

  3. I worked in a factory as a kid. It was a non-union shop, so whenever there was an accident (there were alot), the boss would instruct the foreman to fill out the OSHA report to say, “Worker error.” Then all they had to tell OSHA is that they instructed the worker on better safety. Whenever I hear that an accident was caused by operator error, I immediately think someone in power doesn’t want to take responsibility and pay the costs of really fixing the system.

    As far as things happenning at night, there are alternative explanations. Better employees don’t want to work at night. The supervisors with the real power don’t work at night, and don’t want to be bothered. When the decision needs to be made to shut down the plant or that a patient needs to head back to the operating room asap, who wants to make the call to bossman that he has to get back to work. Delay the call, and accidents occur. That seems far more likely than sleep deprivation.

    • I think it’s in one of Don Norman’s books that there is a chapter on airplane crashes, in which the author discusses the shift in thinking, in the 1970s I believe, that led to a transition away from identifying “pilot error” as the ’cause’ of most airline crashes. There was an example from a small plane in which a plane crashed because the pilot moved the wrong lever, changing the propellor pitch instead of the throttle (or perhaps it was vice versa). A clear case of “pilot error”: the pilot moved the wrong lever, whaddyagonnado? But, as Norman pointed out, the levers were adjacent to each other, and looked and felt identical. And both levers change the sound of the propellor, not in exactly the same way, but still: you move the lever, you expect the engine sound to change, and it does. So, yeah, the pilot, in a plane being buffeted by turbulence, in the dark, moved the wrong lever. That’s pilot error but it’s also a design error: the answer to ‘whaddyagonnado’ is you’re gonna redesign the instrument panel so those levers are farther apart, and you’re gonna put a squared-off handle on the end of one lever and a circular one on the top of the other, and you’re going to make them different colors.

      Oh, I just remembered, there was a connection to nuclear power: Norman said that he once visited a nuclear power plant where the operators had put a Michelob handle on the top of one lever and Budweiswer on another, because these were more distinguishable than the original levers.

  4. From what I know, the definitive source on the causes of the Chernobyl disaster is version 7 of the INSAG report from 1993. (Your post has a reference to version 1, which was based on incomplete information provided by the Soviets.) I looked into the Russian version of it and didn’t find any mention of sleep there. It is believed, as Alexey Smirnov correctly claims, that the disaster was caused by the faulty design. But even when mistakes are ascribed to the operators, those mistakes are not the kind that someone would make out of inattentiveness, but rather from wrong beliefs about what operation regimes are safe.

    The people who were in control at the time of the explosion were the night shift. They were not overworked. The chief reactor operator Toptunov started his shift around midnight. Dyatlov, who was not directly controlling the reactor but was leading the experiment, was not part of the night shift. He worked from the previous morning, but had a 2-3 hour nap in the afternoon, as he recalls in his memoirs on the youtube.

    • Paul:

      Walker’s being fed a steady diet of dollars and adulation so no real motivation for him to wake up. And maybe he’s right that lack of sleep is a problem, so for him the ends (alerting people to the sleep issue) justify the means (misreporting the scientific literature). Kinda like how Sen. Hawley thinks it’s ok to support people who lie about election results and attempt to overthrow of the government because it supports his larger goal of stopping socialism or whatever. Or how Oliver Stone thought it was ok to spread entertaining fabrications about the JFK assassination because it supported his larger goal of . . . well, whatever his goal was. Etc. Utilitarianism is a hell of a drug.

  5. Don Basilio music teacher of the beautiful Rosina from Barber of Seville Rossini has an aria on the subject “La calunnia è un venticello”. The main idea is if you push the lie gently you have a better chance of planting it in people ears.
    I found listening to this (granted overture is better) most theraputic after reading responses on this page.
    😂

  6. I feel a lot of the criticisms that follow these disasters have this holier than thou attitude born out of hindsight.

    I don’t know about the nuclear industry but at least in the chemical industry every time I read an investigation report point to some “glaring mistake” I sheepishly remember the other sites and situations in which I have seen the same stuff done.

    I am not saying this is right just pointing out the hypocrisy of it all when we assign blame to particular actors.

  7. Operator’s error is a broad church. There are tons of operators whose tasks don’t require constant vigilance.

    Best to look at highly complex, time-sensitive such as flying an aircraft.

    I believe it has been researched quite a bit by a lot of very smart people and the effects of sleep deprivation are known:

    https://www.faa.gov/data_research/research/med_humanfacs/oamtechreports/1980s/media/am85-03.pdf

    https://www.faa.gov/data_research/research/media/NASA_Controller_Fatigue_Assessment_Report.pdf

  8. “The Chernobyl accident in 1986 was the result of a flawed reactor design that was operated with inadequately trained personnel.” is the World Nuclear Assosciations one liner, it’s fair enough, although one of the flaws of the design is that “adequate training” is a doctoral level competence in nuclear chemistry and engineering.

    What turns out to be sufficient to blow up this particular reactor is to turn the “throttle” down for a while, then to turn it back up, and then to turn it off. (It is fun to learn why this is.)

    Although this fact was known by the designers of the reactor and the designers of the protocol (to test a saftey subsystem) which involved throttling down and then a shut down, it likely wasn’t known to the operators, who for happenstance didn’t follow the protocol exactly. It takes no effort to imagine how they failed to intuit that the “Off” button became an “Explode” button.

    It is insane to imagine that an better sleep protocol would have made this situation anything close to safe. Maybe they didn’t eat enough vegetables either.

    (https://www.world-nuclear.org/information-library/safety-and-security/safety-of-plants/chernobyl-accident.aspx is a good starting source.)

  9. Adam Higginbotham’s book Midnight in Chernobyl more or less blames Dyatlov for precipitating the disaster. It contains the following (my emphasis):

    Beneath the sickly fluorescent strip lights of Control Room Number Four, a rancid haze of cigarette smoke hung in the air. The midnight shift had only just arrived, but the mood was growing tense. The turbine generator test scheduled to finish that afternoon had not yet begun. The station’s deputy chief engineer for operations, Anatoly Dyatlov, was now entering his second day without sleep. He was exhausted and unhappy.

    Higginbotham, Adam. Midnight in Chernobyl: The Untold Story of the World’s Greatest Nuclear Disaster (p. 75). Simon & Schuster. Kindle Edition.

    He footnotes that statement thus:
    Now entering his second day without sleep: Dyatlov’s defender, cross-examining Fomin in court, in trial proceedings reproduced in Karpan, states that Dyatlov managed the operations of Unit Four alone for two days. Fomin responds that Dyatlov had gone home for a “break” on the afternoon of April 25, but remained available by telephone. Chernobyl to Fukushima, 148.

    Exhausted and unhappy: Boris Stolyarchuk, author interview, Kiev, July 2015

    • It seems to me any design in which it’s possible for the operator to cause a meltdown without going far out of their way to override about 20 different manual overrides is broken to begin with.

      A nuclear plant should be able to have the entire crew go play volleyball for 12 hours and the worst that happens is some people don’t get power.

  10. Tangential Story: When I was a grad student I did a project with a bunch of chemists. They had gotten some data from a third party vendor who reported “non-detects” whenever the measured level of a toxin was below 3 SEMs of their measurement equipment (i.e. they right-censored the data. Apparently this is not uncommon.) I took it as my most important job to convince them to get the full dataset to analyze.

    I read in Helsel (2006) that improperly handling non-detects was a principal cause of the Challenger shuttle disaster, and that really got their attention. They ended up getting the full dataset.

    (I thought that was kinda funny cuz a lot more than 7 deaths could be caused by toxic vehicle emissions, which is what we were working on.)

    Anyway sometimes dramatic stories can get people’s attention. That’s not to excuse shoddy research, but still.

  11. This got me thinking: could the destruction of Pompeii by the eruption of Mt. Vesuvius in 79 AD have been caused by sleep deprivation? Alas, a bit of googling revealed the eruption most likely occurred at noon. Not exactly a time when people are asleep. But then this caught my eye: The dust “poured across the land” like a flood, one witness wrote, and shrouded the city in “a darkness…like the black of closed and unlighted rooms.” Could the citizens of Pompeii have been so sleep deprived that the unexpected onset of darkness caused them to fall into a deep slumber, thus ensuring their destruction? Could they have been snoozin’ while the lava was oozin’? Some are certainly speculating it.

  12. I’m a little baffled by your concerns Andrew. It’s just a form of the classic NPR “some say”, right? Dig it:

    Walker:
    “there is speculation that the operators were sleep deprived…”

    NPR:
    “Several investigations have attributed the disaster to bla bla. But **SOME SAY** the operators may have been deprived of (quality pre school, a living wage, sleep)…?

    If it’s good for an eminent Pillar of Democracy and Leader in Reliable Journalism like NPR, it’s gotta be good enough for science, no? I think NPR has a series of seminars on the Some Say, because Some Say Something in almost every report they produce.

    Anyway, it doesn’t matter who said it. Or when they said it. Or if they said it. Because whoever they are, it’s so certain that it might have been said that they could have said it even if they didn’t.

    Now I know what you’re thinking: this is a mere rhetorical device. But no, that’s not true, because someone really might have said it, and when you think about it with over 7 billion people it’s *almost certain* that someone has said almost everything there is to say. Humans are no typing monkeys. If it might be true they might have said it.

    If that doesn’t justify it nothing does.

    • I might add that some say as well that
      I don’t know half of you half as well as I should like; and I like less than half of you half as well as you deserve.

  13. In response to Jim’s comments, Harry Frankfurt’s “On Bullshit” (https://en.wikipedia.org/wiki/On_Bullshit) seems highly pertinent. As wikipedia summarizes: Frankfurt determines that bullshit is speech intended to persuade without regard for truth. The liar cares about the truth and attempts to hide it; the bullshitter doesn’t care if what they say is true or false, but rather only cares whether their listener is persuaded.

    Also: has anyone else been so much subliminally influenced by this blog that they misread the title as ‘Chernobyl and Matthew Walker’s disaster “Why We Sleep” ‘ ?

  14. Seems that Walker is getting a bit of a pass for saying “There is speculation…” But he closes by saying the accident was “entirely preventable”, implying that more sleep was certain to prevent the disaster and turning unattributed speculation into fact.

  15. The classical art of rhetoric combined with modern psychology tricks are good for media opinion pieces and hostage negotiations. There is a fascinating description of the history of America’s pioneers in political manipulations in Jill Lapore These Truths ~ https://en.wikipedia.org/wiki/Whitaker_and_Baxter . I think it is a well known fact that a number of commerially “successful” scientists were also well … businessman. Edisson, Tupolev are a few I read about. Was Mr. Walker sloppy accidentally or on purpose, and was it a part of the agenda and concious manipulation – I have no idea. It is a pity that as I understand more than a few “slips” ended up in the book. Quite a few ideas do make sense, and everybody knows that your body gonna kick you in the butt when you are sleep deprived for more than a few days being it a seasonal flu due to lowered immune response or human / professional interaction degradation. I am definetely not equipped to discuss long term effects of sleep deprivation, and what seemed like a professional overview was fascinating – now what to believe is not clear which is pretty sad since subject is important.

    • Alexey:

      I doubt Walker made these mistakes on purpose. OK, I guess he probably cheated on purpose with that graph. But, other than that, my guess is that he sincerely believes that lack of sleep causes major problems—and maybe he’s right about that!—and so he didn’t really care about the details. Also, I guess they guy’s kinda lazy. I mean, I don’t know anything about sleep science but in about 15 minutes of googling I was able to find and read all the above sources and find out what was known about Chernobyl. It’s sad to think that Walker wasn’t willing to put in 15 minutes of work to check that paragraph in his book.

  16. A short note on the IAEA report and using it as a referece:

    An update of the IAEA report (“The Chernobyl Accident: Updating of INSAG-1”) is freely available online from their website, https://www.iaea.org/publications/3786/the-chernobyl-accident-updating-of-insag-1. It doesn’t mention sleep either.

    However, by the way the citations are ordered, I would assume that they should anyhow back up only that the accident was related to “human error”, not necessarily that it was caused by a lack of sleep.

  17. In related news, I’ve often seen claims that Chernobyl caused the deaths of “50,000 people”. Here, quote from the article, it’s “thousands”.

    Recently I looked up some reports. Here’s an example:

    https://www.unscear.org/unscear/en/chernobyl.html

    “.. The accident caused the deaths, within a few weeks, of 30 workers and radiation injuries to over a hundred others..

    .. Among the residents of Belarus, the Russian Federation and Ukraine, there had been up to the year 2005 more than 6,000 cases of thyroid cancer reported in children and adolescents who were exposed at the time of the accident, and more cases can be expected during the next decades. Notwithstanding the influence of enhanced screening regimes, many of those cancers were most likely caused by radiation exposures shortly after the accident. Apart from this increase, there is no evidence of a major public health impact attributable to radiation exposure two decades after the accident. There is no scientific evidence of increases in overall cancer incidence or mortality rates or in rates of non-malignant disorders that could be related to radiation exposure. The incidence of leukaemia in the general population, one of the main concerns owing to the shorter time expected between exposure and its occurrence compared with solid cancers, does not appear to be elevated. Although those most highly exposed individuals are at an increased risk of radiation-associated effects, the great majority of the population is not likely to experience serious health consequences as a result of radiation from the Chernobyl accident. Many other health problems have been noted in the populations that are not related to radiation exposure..”

    I’d be interested to hear of any scientific reports that deaths from the Chernobyl reactor are in “the thousands”.

Leave a Reply to Matt Skaggs Cancel reply

Your email address will not be published. Required fields are marked *