Besides geomagnetic storms, obviously. This has to be man-made. At first I considered a HEMP, but who would've guessed nuclear weapons had so many failsafes to prevent premature detonations? So then the question is "how could I cause an accidental global blackout without using nukes or The Power of the Sun"? Alternatively, "how could I make enough nukes detonate prematurely without any elaborate sabotage"?

P.S. The "accidental" part is really important. It has to be an unexpected consequence of something, or at least appear to be one on the surface.

  • 6
    $\begingroup$ How global and how accidental? Something that could blackout antipodal island nations like Ireland and New Zealand would be pretty extreme. $\endgroup$ – Nick T Jul 7 '17 at 22:07
  • 3
    $\begingroup$ It takes a lot of power to do something like this, it's hard to imagine humans even being able to it if they tried short of nuking everything. For any individual to be able to do it accidentally is even harder to visualize. $\endgroup$ – A. C. A. C. Jul 7 '17 at 22:14
  • 1
    $\begingroup$ Your alternative question of detonating nuclear weapons prematurely would be better as a separate question. It is too different from the topic of a global blackout to be part of it. Good to see you're posting questions again. For a while there you seemed to be elsewhere. $\endgroup$ – a4android Jul 8 '17 at 2:22
  • $\begingroup$ My guess is an accidental global blackout could only be triggered by an accident involving alien hypertechnology. Perhaps, something left behind by visitors from a Kardashev Type 2 or 3 civilization? $\endgroup$ – a4android Jul 8 '17 at 2:25
  • $\begingroup$ I was just about to ask this, beat me too it! :) $\endgroup$ – Noah Cristino Jul 9 '17 at 1:07

A security researcher accidentally loses control of a highly infectious worm that affects IoT devices.

IoT devices are increasingly common, and they're connected to things that require large amounts of power like air conditioners. They also have really bad security, though because of awareness, hopefully this will improve. However, the general trend is towards more of them controlling more things, and security will always be an issue in some form.

Power grids need to be balanced in terms of supply and demand, also across regions, otherwise bad things can happen like the 2003 Northeast US blackout. If you were to synchronously increase load and shed load you may be able to destabilize grids, triggering cascading failures.

So, if your worm turned everything it had control of on every even minute, off every odd minute, that might cause many power grids to react extremely poorly, and potentially shut off as power plants disconnect. More advanced patterns that exploit the connections in a power grid might be able to cause more damage to what is already aging infrastructure, but the on-off thing might be what someone creates as a simple test.

You could also have this worm created by an AI, or an AI infecting all the things, and the power plants, but even if power plant computers are throughly "air-gapped", crazy load patterns could still trip them into disconnecting. Omniscient, omnipotent AIs are kinda cliche though.

This is not likely to affect anything on the smaller scales like the T'au microgrid which are more resilient because they're small and relatively modern. Certainly is going to do nothing to your diesel generator out in the woods.

  • 1
    $\begingroup$ Especially if the worm was either legitimately aiming at something else, wasn't intends to be malicious, accidentally left systems in a state that some other system interpreted as incorrect and shut them down as part of a safety protocol, or were an incorrect security update. All quite plausible. After all over time systems may converge and owners of houses and businesses might have systems linked to their utilities. Wouldn't be impossible for one issue to trigger so many fallovers globally that the rest fell over too $\endgroup$ – Stilez Jul 8 '17 at 6:11
  • $\begingroup$ A related point - what would be the minimum that would have to happen, to create a global blackout? $\endgroup$ – Stilez Jul 8 '17 at 6:13
  • $\begingroup$ The worm was trying to mine bitcoins on smart meters, but it was badly written? There are already technothrillers about that ... $\endgroup$ – o.m. Jul 8 '17 at 6:34
  • $\begingroup$ @o.m. The meter/device itself would only consume a few watts, versus the kilowatts of what it's connected to. If the load is constant, that wouldn't do much to destabilize the grid as well. $\endgroup$ – Nick T Jul 8 '17 at 11:02
  • $\begingroup$ @NickT, I was thinking of a worm which accidentally crashes the smart meters themselves, instead of using them for bitcion mining. That causes a chain effect to crash the power grid. $\endgroup$ – o.m. Jul 8 '17 at 14:14

Linked: biggest blackouts of all time. http://mentalfloss.com/article/57769/12-biggest-electrical-blackouts-history

Here is described a huge one from 2003 which affected a big piece of the northeastern US and Canada. It was an attention getter because so many people were affected, and because blackouts of this sort were thought to be relics of the 1970s. When no explanation was forthcoming people started imagining weird explanations.

  1. NORTHEAST UNITED STATES AND CANADA // AUGUST 14-15, 2003 It took months before the real cause of the Northeast Blackout of 2003 was finally determined. Initially, Canadian Defense Minister John McCallum blamed an outage at a nuclear power plant in Pennsylvania, which the state's Emergency Management Agency denied. What actually happened was a high-voltage power line in Northern Ohio brushed against overgrown trees, causing it to shut down. When the alarm system that would typically alert FirstEnergy Corporation failed, the incident was ignored. In the next 90 minutes, system operators tried to figure out what happened while three other lines switched off as a consequence of the first line's failure.

The official "tree branches" thing still smells a little bit to me like "weather balloons...the planet Venus".

Centralized power generation is efficient, but new technology comes with new hazards, and the possibility of old hazards manifesting in new ways. I could imagine a circumstance where a technological breakthrough (for example: fusion power and Tesla-type aetheric power transmission) results in the entire world getting their power from a single facility. This facility - enormously expensive but so efficient there is no need for duplication (and no funds or political will to do so) - becomes the proverbial one basket that all eggs are in. The disaster which shuts it down does not need to be some Siberian fireball; it should be some prosaic like tree branches or an anchor fouling some sort of stay, which echoes and ramifies thru the new system to result in a total shutdown.


You say you want to go with something accidental, but you can take it even further. What if some people do something they think is helping and it causes catastrophic failure.

Imagine this:

A private software firm develops a new program for power grid management. You may or may not be aware of this, but most power grids need constant maintenance to balance supply and demand of power. Too much provided, and the grid shorts. Too little supplied, and systems shut down. As a result, the power plants for each regions grid are constantly changing output. It is inefficient and relies nearly exclusively on human influence. Water flow through dams, burn rate for coal plants, and number of cells active in nuclear plants are all dependent on people, with very little computer influence. What if that changed? What if our software company found a way for a computer to observe supply and demand for power and learn when and why to adjust output, adapting to all forms of generation. This company would instantly become rich, as energy firms seek to use their hopefully patented tech. It starts small, taking root in cities first and then is applied to nationwide grids. It is amazingly effective, causing the energy industry to reduce failures and increase reliability.


It was a simple mistake. The software firm is working on an update to their system, one that energy firms are deadly waiting for. When it comes out, all of them want it, and they allow some of their newly computerized systems to accept the update and all goes smoothly. They wait a week, a month maybe. All seems well. So, they push the update to more systems it until all of their grids use it. After all, if it didn't fail within a month, it must be safe. And it was, for a time. But unknown to all, there is a glitch in the update, a miswritten definition in one of the program's library files. This particular file isn't called on by the main function very often, so the glitch remained unnoticed. Then, one day, catastrophic power failures everywhere. Coal plants stop working. Dams are frozen at their current flow rate. Nuclear plants completely shut down, their cells grow cold. The lights go out. The world is dark.

Basically this idea depends on a glitch in computerized grid maintenance causing a global shutdown. This idea, scarily enough, is plausible even today. I'm sure the energy industry would love a way to manage the grid without constant human attention, and a bad update is something everyone deals with eventually, from iPhones whose hardware can no longer handle the new iOS to government websites with broken links.

Hope this helps.

  • $\begingroup$ Take a look at Wikipedia's article on the Therac-25, a radio cancer treatment device that was lethal under a very obscure combination of factors. Software race conditions are as plausible and would very neatly account for nothing going wrong until it all goes wrong. The wiki describes exactly what had to happen for that to be the case. $\endgroup$ – Stilez Jul 9 '17 at 6:47
  • 1
    $\begingroup$ ... Specifically two errors in use: "The failure only occurred when a particular nonstandard sequence of keystrokes was entered: an "X" to (erroneously) select 25 MeV photon mode followed by "cursor up", "E" to (correctly) select 25 MeV Electron mode, then "Enter", all within eight seconds. This sequence of keystrokes was improbable, and so the problem did not occur often and went unnoticed for a long time." and "The system merely displayed the word "MALFUNCTION" followed by a number, so operators pressed the P key to override the warning and proceed anyway." Sadly an easy cause. $\endgroup$ – Stilez Jul 9 '17 at 7:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.