My army of squirrels are primed and ready to find the acorns I've planted in the power grid - please pay 100 Bitcoins or else......
What's the biggest danger to the power grid? Hackers? Terrorists? Er, squirrels
For decades now people have been claiming that the power grid could be taken down by terrorists. However, simple statistical analysis shows that the biggest danger isn't online hackers, but squirrels – aka rats with good PR. Cris Thomas, a strategist at Tenable Network Security who goes by the moniker Space Rogue, has been …
-
Thursday 19th January 2017 08:15 GMT Anonymous Coward
Why destroy them?
Just program the smart meters in an area to flip load on-off in a pre-programmed pattern. Then do it again after the grid has been brought up (it may take days to do that by the way). And again. Viva la smart metering - the greatest act of calamitous self-harm one can do to itself with the current grid designs.
While it is a bit more difficult now, because the resistive loads of yesteryear from lights and heaters are mostly gone, it is still doable today.
I remember my dad doing the math for this once every few years (as the grid changed) in the 1980-1990es in one small (and nowdays NATO) country. I helped him with the software parts for some of the models. We got pretty good money too.
-
-
Thursday 19th January 2017 12:31 GMT phuzz
If you check out cybersquirrel1.com you can search by time, and find that in January 2016 alone, there was a dozen squirrel related power incidents.
Monkeys are amateurs by comparison.
Is it time for elReg to provide us with a cybersquirell threat icon?
-
Thursday 19th January 2017 15:16 GMT Anonymous Coward
Storks beat monkeys
In 2000, a stork got itself electrocuted and managed to make half of Portugal, including it's capital, go dark for a couple of hours (https://www.publico.pt/destaque/jornal/e-ao-principio-foi-a-cegonha-143727 and https://pt.wikipedia.org/wiki/Apag%C3%A3o_no_sul_de_Portugal_em_2000 - both in portuguese)
-
-
-
-
Thursday 19th January 2017 19:12 GMT GrapeBunch
Re: Treasonous Squirrels
In all likelihood, they are Eastern Gray Squirrels, Sciurus carolinensis. Red squirrels tend to be retiring. "East" being the direction of many threats, and "Gray" the colour of heartless bureaucracy, maybe this appellation could win the Internet, before destroying it. I rather like Red Squirrels. Hate Eastern Grays and imposition of smart meters with a passion. I would not be a good Buddhist. See, on the Report Card, the Buddhism tea-cha wrote "Lacks compassion. Could do better".
-
-
-
Friday 20th January 2017 10:16 GMT Voland's right hand
Re: No birds dropping bread?
Birds dropping bread is a minor issue.
A Vulture or another large bird of prey taking a dump - different story. It is actually part of the design brief for substations in North America and other places where these can be found to be bird sh*t proof.
Think of the wonderful present from the friendly wood pidgeon or seagull which has graced your windshield. Multiply by at least 10 (if not 100) and make it a bit more runny. Apply between wires carrying 600KV. Enjoy the fun.
By the way - the standard British substation designs _ARE_ _NOT_. As we reintroduce more and more birds of prey and their population recovers we are going to see some fun.
Me coat (the one with the big torch and the candles in the pocket).
-
-
Thursday 19th January 2017 10:06 GMT John Smith 19
a "democracy-ending event."
Some would say the new President will do that all by himself.
BTW Wasn't the US (military or CIA) reputed to be looking at this in Viet Nam for deniable low intensity destruction of infrastructure. I think they were looking at training rats to gnaw on power line insulation in preference to other things.
-
Wednesday 15th February 2017 18:48 GMT Charles 9
Re: a "democracy-ending event."
"BTW Wasn't the US (military or CIA) reputed to be looking at this in Viet Nam for deniable low intensity destruction of infrastructure."
Sounds like they dropped the idea? Impractical because most places there didn't have infrastructure as we would know it?
-
Thursday 19th January 2017 11:18 GMT Dr. Mouse
In 2015, a fox shorting out a substation in Utah caused an outage that shut down an oxygen machine and led to the death of a patient.
In an era where all but the least important servers are protected by UPS, shouldn't a life-critical machine have a backup power supply?! Surely it wouldn't be too difficult to include a few hours of emergency power into it....
-
-
Thursday 19th January 2017 13:00 GMT Anonymous Coward
re: when you REALLY need it?
"And how would you know it's working when you REALLY need it? "
Well, back in the day when IT was a service to the business, rather than an empire for the Emperor of IT, people used to actually test things.
Got failover storage? Pull some cables and see what happens? Do it regularly, and check that the fail back (ie return to previous operating config) works too.
Got failover power ? See above.
Got a DR site/DR plan? See above.
Etc.
It's not rocket science, but as you and others have clearly shown, it seems to be a forgotten era for a great many people. Perhaps because it costs money (even though actually it should be a good investment) while producing little visible increase in the size of the Empire.
-
Thursday 19th January 2017 14:33 GMT Charles 9
Re: re: when you REALLY need it?
"Got failover storage? Pull some cables and see what happens? Do it regularly, and check that the fail back (ie return to previous operating config) works too."
But with Murphy's Law, the device will PASS the testing, then FAIL when the actual emergency hits because the only thing that can actually duplicate the full conditions of an actual emergency is an actual emergency.
-
Thursday 19th January 2017 14:56 GMT Anonymous Coward
Re: re: when you REALLY need it?
"But with Murphy's Law, the device will PASS the testing, then FAIL ..."
On numerous occasions it is the test which accidentally sets up the next failure. You do a manual override for the test - and forget to put it back into "auto" afterwards. You pull a cable - which gets damaged when you re-insert it. The back up generator doesn't get its fuel replenished and conks out in a real emergency.
If is a humbling lesson to learn that in real life:
There are a finite number of things you can think of that might go wrong ....and an infinite number that can go wrong.
-
Thursday 19th January 2017 14:57 GMT rhydian
Re: re: when you REALLY need it?
Many a company has found out that simply switching the power off to the server doesn't replicate a power cut properly. There's always one rack, router, switch or other minor bit of kit (NTE for fibre or EFM lines are usually good bets) that someone's plugged in to the unprotected mains "just for now" that doesn't stay up when the supply goes off.
The other classics are generators that only have enough diesel for a few minutes running (because some berk forgot to wire the fuel lift pumps to the UPS) or having the whole IT infrastructure hooked up the generator, but not the air handling plants...
-
-
Thursday 19th January 2017 23:43 GMT Anonymous Coward
Re: re: when you REALLY need it?
"X] Supply for computer room via UPS, [X} supply for computer room cooling via UPS, [ ] cooling for UPS itself ..."
IIRC a server room was in a low lying area with potential flooding. They discovered one day that the water pumps were only on direct mains power.
-
-
-
Friday 20th January 2017 10:51 GMT Tom Paine
Re: re: when you REALLY need it?
If you've ever heard grizzled on BCP / DR hands yarning late at night,.. I'm sure I remember hearing many such "but it worked when we tested it!" stories, either because the failure mode was different than that tested, or the surrounding environment was different, or just simple bad luck. ISTR working at a place where there was a genny soaktest for like six hours or something, powering a few small campus offices with a couple of server rooms. All fine. Unfortunately the following week a bit of backhoe fade took out the HT lines to the business park for real, and no-one had got round to topping up the diesel tanks after the test...
-
-
-
-
-
Thursday 19th January 2017 11:21 GMT Anonymous Blowhard
Die Hard 5 - This Time its Squirrels!
Looks like politicians are getting cyber-warfare "knowledge" from Die Hard 4.0, just like they got their "liquid explosives" information from Die hard 2.
I'll go and put on a clean vest, just in case...
-
-
-
-
Friday 20th January 2017 00:24 GMT Anonymous Coward
Re: Only 15?
Self censorship? Careless talk costs lives?
How much 'intelligence' do you need to read an Ordnance Survey map, know where the important transmission lines are (GW-scale power stations are hard to hide), and find some conveniently located Toolstations or similar (or just insecure building sites).
Taking out the UK grid for an extended period (blackstart, anyone?) would be trivially easy. I'm astounded it hasn't been done already. I can only imagine that people don't want to, for some reason.
-
Friday 20th January 2017 11:12 GMT Tom Paine
Re: Only 15?
The IRA would have worked it out in the end; their later bombing campaigns in Britian (early 90s) were nominally about economic damage (the Bishopsgate, Canary Wharf and Brent Cross truck bombs). Fortunately, -- well, Google "The Grugq IRA" and take it from there...
These days (I believe) the various bodies interested in the security of CNI have thought about this, and done things about it.
-
-
-
Friday 20th January 2017 10:30 GMT Voland's right hand
Re: Only 15?
How long does it take to remove a destroyed pylon and replace it exactly?
The issue is not the time to replace it. Same goes for the substations.
The issue is that generating capacity will have to go offline and in emergency dump mode. If you take offline using the emergency procedure a large coal or nuclear power plant it will take days to reconnect it and you need to bring things up slowly bit by bit as the capacity comes online. You also risk damage if you go offline in emergency mode - after the grid collapses not _ALL_ of the capacity will be coming back without repairs.
If you hurry to bring it back online because you have the politicos breathing down your neck you are likely to overload the system, it will go into emergency dump mode again and you need to start from scratch. With some more damage to account for.
This is where wind, hydro and other renewables are quite handy by the way - most of them are on/off nearly instantaneously, while they cannot carry the grid on their own they can definitely help you balance the load while you are bringing the capacity online.
-
Friday 20th January 2017 11:08 GMT Tom Paine
Re: Only 15?
I've thought about this quite a bit. It wouldn't take long to replace even one of the 400 kV pylons -- a day or two perhaps? Now imagine that a large urban area that only has, say, 15 or 20 lines carrying power from the grid to the comedy beard-wearing craft beer swilling bread-whittling masses of the big city, and they're all knocked out simultaneously. How'd you think Paris*, say for instance, would look after 10 days without power?
* I have no clue whether NYC is vulnerable to this sort of attack. I'm just puzzled why more of those long range transmission lines aren't buried...
(Edit: the UK, I believe, has mitigations in place for attacks on the electricity grid. Hopefully we'll never get to find out if they work or not.)
-
Friday 20th January 2017 12:30 GMT Anonymous Coward
Re: Only 15?
"Now imagine that a large urban area that only has, say, 15 or 20 lines carrying power from the grid"
No need to imagine. Just observe relevant public domain information, and learn. Don't assume.
"the UK, I believe, has mitigations in place for attacks on the electricity grid."
Any references readily available? I'll understand if not.
Here's an example from 2008 of what happened when two unrelated incidents a few minutes apart cut off around 1.5GW of input to the UK grid (and the grid's response for hours afterward wasn't quite as expected, according to the Grid's own published analysis):
http://news.bbc.co.uk/1/hi/england/7423169.stm
"[...]Hundreds of thousands of homes and businesses across London, Cheshire, Merseyside and East Anglia lost power.
Blackouts were caused by Sizewell B nuclear plant in Suffolk and Longannet coal-fired station in Fife going off-line within minutes of each other.
[...]"
https://www.ofgem.gov.uk/sites/default/files/docs/2008/07/national_grid__system_events_of_27_may_for_dswg_16_july.pdf
Readers who enjoyed coverage of the 2008 incident may also enjoy this writeup of two separate incidents in 2003 affecting larege numbers of customers across a wide area, one incident in London, one in the West Midlands
http://www.rae.gr/old/cases/C13/london/OFGEM_PB_london_v1.pdf
-
Wednesday 15th February 2017 18:58 GMT Charles 9
Re: Only 15?
"* I have no clue whether NYC is vulnerable to this sort of attack. I'm just puzzled why more of those long range transmission lines aren't buried..."
Burying cables has tradeoffs, especially for longer distances, harder ground, and maintenance concerns. (Buried cable is MUCH more expensive, both in installation and maintenance, gets worse if you have to deal with harder ground, and doesn't lend itself well to upgrading). And IINM it gets more complicated when you're talking high-voltage transmission lines because now you have to take other things into consideration.
-
-
-
-
Thursday 19th January 2017 13:11 GMT Doctor_Wibble
Or a shovel through a fibre/wire bundle
OK I know it's not an animal* but in the 'more cyber damage than terrorists' stakes, I recall 15+/- years ago a chunk of London-near-Old-Street and more lost its interwebs when a mechanical digger vigorously encountered a bit of fibre (or possibly wire bundle) and that cut off a *lot* of people. Same thing nowadays would be serious devastation as it's all expanded somewhat and I doubt the multiple path (lack thereof) has been addressed unless someone got fired the first time.
.
* though it might be an alien in disguise, would that count?
-
Thursday 19th January 2017 15:01 GMT Anonymous Coward
Re: Or a shovel through a fibre/wire bundle
A few years ago a storm in the UK blew down one microwave tower out of the many possible ones in that affected region. Our Ops manager knew immediately that it was the single point of failure in a customer's otherwise resilient multiple diversity comms path.
The customer had been warned that the weakness would exist - just for a few months - until a new alternative was to be built.
-
Thursday 19th January 2017 16:52 GMT An0n C0w4rd
Re: Or a shovel through a fibre/wire bundle
My favourite RFO* from a telco was that they found shotgun pellets embedded in the fibre. Someone had been taking pot shots at some fauna and had taken out the fibre instead. I guess it must have been hung from telephone poles, but can't remember to be honest
Yes, this was in the land where the 2nd amendment is used to justify way too much
* RFO = Reason For Outage
-
-
Friday 20th January 2017 15:20 GMT Anonymous Coward
Re: Or a shovel through a fibre/wire bundle
That one is missing one more. Car fires.
Quite a lot of telcos especially in France (including Level3 primary fiber routes into and out of France) have fiber running along the motorways in relatively shallow ducts. So when a car ends up on fire next to the road it also kills the fiber. Happens in other countries too, just less often.
-
-
Thursday 19th January 2017 17:00 GMT Sir Runcible Spoon
Re: Or a shovel through a fibre/wire bundle
"(or possibly wire bundle) and that cut off a *lot* of people"
If you are referring to an incident that happened in the late 90's, then it was a fibre bearer - 150 bundles iirc. Oh, and yes, the people who had purchased divergent connectivity were none to pleased to discover it wasn't.
It was odd seeing all the alerts come in. 5 red lights....15 red lights......screen full of red lights OMG WHAT HAPPENED TO LONDON!!! ;)
-
-
Thursday 19th January 2017 15:13 GMT Stevie
Bah!
"For decades now people have been claiming that the power grid could be taken down by terrorists. "
a) Decades? Really?
2) A local substation outage or a line down does not remotely constitute "taking down the grid".
$) Decades of shortsighted management, replacing decades of engineering knowledge specific to the business of power generation and distribution with software written by programmers with a total disconnect with the actual business, and refusal to invest in new infrastructure that doesn't replace knowledge base with IT do more to threaten the grid than any othr factors.
-
Thursday 19th January 2017 21:35 GMT Anonymous Coward
Re: Bah!
"2) A local substation outage or a line down does not remotely constitute "taking down the grid"."
Yes it does, in fact. The problem is that failures cascade, and NO ONE wants to build backup capacity that never gets used and costs money the customers aren't willing to pay. BOTH Northeast Blackouts started with a fault in a single substation that caused another down the line to overload, causing the next one to overload, and so on until a sizeable chunk of the US (including notably New York City) is without power.
-
Wednesday 15th February 2017 17:04 GMT Stevie
Re: Bah!
No, the 2004 northeast blackout started because the people running a power company in Ohio didn't know how to do that, having had all their actual power grid expertise fired and replaced by ... IT.
That is why they took down their only reactive power generation facility for maintenance at the height of reactive load season, ignored field reports of shorts and fires on the now overloaded lines (because the - stalled - computer instrumentation said "nothing wrong here") and when they finally got a clue started running their playbook (not a euphemism, they had an actual book) from the wrong page because no-one considered that the downed reactive power facility had already put them in a "class one" failover.
The cascade blackout was entirely unnecessary and could have been avoided had anyone in the Ohio control room understood the business they were in charge of and taken some simple, widely known, industry standard mitigation steps at some point in the hours of warning signs they had.
2004 pretty much informed my original post, and the official report on what, when and why is clear and interesting reading if you care to look it up.
Substations fail all the time. Cascades do not happen when they do. Yes the aging grid is vulnerable. It is not, however, that fragile.
Apologies for late response. Vacation, illness, work etc.
-
Wednesday 15th February 2017 18:40 GMT Charles 9
Re: Bah!
"The cascade blackout was entirely unnecessary and could have been avoided had anyone in the Ohio control room understood the business they were in charge of and taken some simple, widely known, industry standard mitigation steps at some point in the hours of warning signs they had."
Thing is, it spread BEYOND Ohio, meaning it wasn't JUST Ohio that was in trouble since the mitigations THERE failed, too. Plus there's the earlier 1960's blackout, which we KNOW started with ONE substation and cascaded.
-
-
-
-
Thursday 19th January 2017 17:00 GMT An0n C0w4rd
Soldiers unaware of the Faraday cage
"In the same year, three Sri Lankan soldiers were electrocuted after a squirrel caused a fire that broke power lines – causing them to fall on the soldiers' vehicle."
From what I understand, the soldiers who died got out of the vehicle, and were therefore electrocuted. The other soldiers in the same vehicle who stayed inside survived.
-
Friday 20th January 2017 06:49 GMT Charles 9
Re: Soldiers unaware of the Faraday cage
Sounds better. I would've thought the vehicle wouldn't readily conduct the power from a fallen wire. See, I've been told a vehicle is actually one of the better places to hide out in an electrical storm; the air-filled rubber tires put a layer of insulation between the vehicle and the ground, reducing the risk of a direct discharge.
-
-
Thursday 19th January 2017 17:12 GMT Runty Dog
Nut jobs
Two loons with a bit of readily available explosives nearly got away with dropping the communication grid in 1961. As it was, it took Ma Bell & the US government over six days to re-establish AT&T Long Lines' transcontinental relay network. http://www.beatriceco.com/bti/porticus/bell/longlines-expdam.html
In 1988 on Mother's Day the Bell system suffered a knee to the groin from a single point of failure, dropping cross country & local communications into the dumpster: http://articles.chicagotribune.com/1989-03-11/news/8903250918_1_state-fire-marshal-alarm-electrical-power
A disgruntled employee with an axe & a pack of matches dropped the US air traffic control system into the bin: http://www.chicagotribune.com/suburbs/aurora-beacon-news/news/ct-nvs-faa-report-reaction-st-1002-20150930-story.html
Our infrastructure is not as robust as we would like, but in each case the people directly affected overcame the problems. And not because some pin head, pencil pushing, government administrator came up with a study identifying probable threats. As the squirrel study shows, the real problems come from diverse random events, and are not predictable. It's usually some nut job (squirrels included) acting out that causes all the havoc.
-
Thursday 19th January 2017 21:40 GMT Charles 9
Re: Nut jobs
"Our infrastructure is not as robust as we would like, but in each case the people directly affected overcame the problems. And not because some pin head, pencil pushing, government administrator came up with a study identifying probable threats. As the squirrel study shows, the real problems come from diverse random events, and are not predictable. It's usually some nut job (squirrels included) acting out that causes all the havoc."
Which is why the reports notes what could happen if someone got SERIOUS about sabotaging the US. If one nutjob can bork a chunk of the US for a day or two, imagine a group with 9/11 levels of resources? Or worse, a State using it as the prelude to a Decapitation attack?
-
Friday 20th January 2017 13:38 GMT Tom Paine
Re: Nut jobs
Thanks for the links, hadn't heard those before and I love a good incident story.
As the squirrel study shows, the real problems come from diverse random events, and are not predictable.
Well, but. Without getting too Foundation about it, you can't predict that squirrel #2897467 will break cable #230487 at 17:42 on 15th May 2017, but I imagine the incidence of squirrel attacks stays reasonably constant (or constant with respect to the size of the squirrel population, anyway); and if it's happening dozens of times a year, it's clearly 100% predictable that "squirrels will have a go at infrastructure *like this*, even if it's not this particular box or cable". Evidently hardening fibre or copper cable against incisor attack is too expensive to be worth it. Not many squirrels live underground, though...
-
Friday 20th January 2017 16:59 GMT Charles 9
Re: Nut jobs
Yes, because squirrels are rodents like rats. And if you'll recall, rats are notorious for being able to chew through well-nigh anything, including cinder block and metal. And squirrels have an edge over rats in their great leaping ability and propensity to reach. That's why they're a handful around seed feeders. IOW, if they want in, they'll find a way in in spite of God, Man, or the Devil.
-
-
-
Thursday 19th January 2017 22:16 GMT veti
In proportion
Does no-one remember ANYfuckingthing from more than a few weeks ago any more?
Eight deaths? In 35 years? Boo hoo. In 2003, one software bug in the grid system - in America - killed 12 people in 2 days. Reference.
Software engineers - and, by extension, hackers - are way, way more dangerous than any number of rodents.
-
-
Friday 20th January 2017 12:28 GMT J P
Re: The real danger isn't terrestrial
Apologies to those who've seen me post this link before - Lloyd's of London have done some modelling on the likely impact of a Carrington Event - you can find the pdf to download here: https://www.lloyds.com/~/media/lloyds/reports/emerging%20risk%20reports/solar%20storm%20risk%20to%20the%20north%20american%20electric%20grid.pdf
-
Friday 20th January 2017 12:39 GMT Anonymous Coward
Re: The real danger isn't terrestrial
"modelling on the likely impact of a Carrington Event"
Interesting, thank you, not seen that document before although I'm aware of the original Carringto event and the much more recent Quebec one.
I've not read it properly yet but for those who might be interested, It should probably be pointed out that the scientific analysis was done *on behalf of* (not "by") Lloyds. Small but important distinction.
-
-
Biting the hand that feeds IT
