New Data Center Protects Against Solar Storm and Nuclear EMPs

dcblogs writes "In Boyers, Pa., a recently opened 2,000-sq.-ft. data center has been purpose-built to protect against an electromagnetic pulse (EMP), either generated by a solar storm or a nuclear event. The company that built the facility isn't disclosing exactly how the data center was constructed or what materials were used. But broadly, it did say that the structure has an inner skin and an outer skin that use a combination of thicknesses and metals to provide EMP protection. Betting against an EMP event is a gamble. In 1859, the so-called Carrington solar storm lit the night skies and disrupted the only telegraph communications. William Murtagh, program coordinator at U.S. Space Weather Prediction Center, said there is ongoing concern that the earth may see an solar storm that could impact electronics on the ground. "We're concerned that can happen," A 2012 solar storm, that missed the earth, "was very powerful, and some have suggested it would have been on par with a Carrington-level event." One researcher put the odds of a catastrophic solar storm by 2020 as one in eight.

Solar storm EMP vs nuclear weapons

[magarity]

OK, solar storm I can understand protecting against. But nuclear weapons EMP? Better to use the data center as a bunker in that case and never mind the data.

Re:

[olsmeister]

A nuclear produced EMP would likely be from a detonation high in the atmosphere that would do little or no damage on the ground. A bunker would not be needed for such a scenario.

Re:

[Obscene_CNN]

A nuclear EMP can be produced by any nuclear detonation. Although an enemy might choose a high detonation to try to knock out just electronics the vast majority of nuclear warheads are going to be used for their blast effect. Thus a bunker would be useful

Re:

[Obscene_CNN]

Secretly detonate a nuclear warhead... Lets ponder the stupidity of that statement for a moment.... I think someone will notice... Also detonating a nuke would be a full fledged war

Re:

[Obscene_CNN]

Ummm..... your satellite attack scenario has a few holes in it. Orbital mechanics is well understood. Also NORAD finds and tracks items as small as a bolt in space. Keep dreaming, you'll might come up with a plan that is good enough to get yourself on the watch list someday.

Re:

[MobSwatter]

EMP permeation is relative to the event that created it. No blanket can be created to cover the actual lock (materials created to deflect the charged particles created by an event not foreseen). Only a couple of methods of creating one have been documented, but many are there, same with working in any limited perception in the aspect of energy. Hey uncle sam, think gagging the scientist was all that smart now?

Re:

[thechemic]

Also detonating a nuke would be a full fledged war

Lets ponder the stupidity of that statement for a moment...

Several different nations have detonated over a combined 2000 nukes throughout history, and none of them resulted in a "full fledged war". Furthermore, only two nukes have been detonated during wartime, however it was a "full fledged war" that resulted in the detonation of the nukes. It was not the detonation of the nukes that resulted in the full fledged war. In fact, throughout history exactly ZERO detonated nukes have ever resulted in a full f

Re:Solar storm EMP vs nuclear weapons

[EvilSS]

While any nuclear detonation creates an EMP, it's really only high-altitude detonations that you need to worry about for EMP effects. Their interaction with the magnetosphere can amplify and spread their EMP effects over a very large area . By contrast the effective range for a normal ground or airburst nuclear EMP is less than its thermal and overpressure damage radii. When you are pincushioned by debris and on fire, you tend to re-evaluate your level of concern about your electronics.

Re:

[theshowmecanuck]

Because you would be dead.

Re:

[mbone]

EMP is not biologically dangerous, unless you are wearing something like a pacemaker. And, all you need for protection is a suitable Faraday cage and isolation from the grid, so the same shielding can protect against both EMPs and Solar Storms.

Re:

[CanHasDIY]

EMP is not biologically dangerous, unless you are wearing something like a pacemaker.

In which case, it's still not biologically dangerous. It's just dangerous to you because you have non-biological parts performing essential functions.

2000 square feet?

[mbone]

That's it? With a datacenter that small, I wonder they didn't put it deep underground (unless this is a typo).

Can I convert my basement into a data center and get it on slashdot too?

Re:

[__aaclcg7560]

Can I convert my basement into a data center and get it on slashdot too?

Putting your kids out on the streets wouldn't be good for society.

Re:

[decep]

This is basically a closet in the data center world. It is certainly not nothing, but there are probably still satellites in operations that could not discern a building of this size. :-)

Re:

[mbone]

Generally, in the data centers guarantees really mean that you get a payment (or a reduction in fees) if the guarantee is violated. (You might get 1 day's service fee off if you lose power for X minutes, for example). So, if it doesn't work, expect a reduction in the bill, as specified in the contract.

So, if you bet your business on something like this, you had better have a plan B in case of outages.

Poor comparison...

[Mister Transistor]

A "Carrington-level" event nowadays would most likely be much less disruptive, as back then all the early radio and spark gap stuff was well under 50 MHz, which is where almost all of the natural noise winds up in the spectrum. Ever notice, for example you can hear your shaver motor on an AM radio but not an FM one. This is not due to AM vs. FM, (well, it is a little) but mostly due to the fact that AM is about 1 MHz and FM is about 100 MHz, well above the "static line" around 50 MHz.

It would take a much stronger signal than back then to cause the same level of disruption. Not saying that can't happen, but modern radio communications are quite a bit more robust than they were back over 100 years ago.

Re:

[Bengie]

I would be more concerned about power lines or transformers melting.

Re:

[chihowa]

You are correct in that AM frequency is generally labeled in khz... in the US the range is 535-1605 kHz ... of course any one who isn't from the US could tell you that 1000 kHz is equal to 1 MHz....
which means the parents statement about AM being "around 1MHz" is a fairly accurate statement, more accurate would be 1MHz plus/minus ~600 kHz.

Ultimately weather something is measured in kilohertz, megahertz or gigahertz, is a matter of scale, is something oscillating at of thousands of time per second, or millions, or billions?

Anyone who isn't from the US probably needs to fix their intermittently failing shift key, though. SI actually assigns meaning to capitalization in units and there is a big difference between mHz and MHz.

Re:

[cyberchondriac]

in the US the range is 535-1605 khz ... of course any one who isn't from the US could tell you that 1000 khz is equal to 1 mhz.... w

Is that an anti-american dig? Just because we don't use metric for everything doesn't mean we're clueless about it. Most people I know would know that 1000k=1M, and millions of us work or study in IT or other scientific fields where kilo, mega, giga, micro, pico, tera, etc..prefixes are common knowledge, despite our continued use of the Imperial system. Anyone who's graduated from a decent high school should know that.

Re:

[rubycodez]

sure it can be, USA AM band is 0.540 to 1.710 MHz

Also, US FM band is 87,500 to 108,000 KHz.

Re:Poor comparison...

[jdschulteis]

A "Carrington-level" event nowadays would most likely be much less disruptive, as back then all the early radio and spark gap stuff was well under 50 MHz, which is where almost all of the natural noise winds up in the spectrum. Ever notice, for example you can hear your shaver motor on an AM radio but not an FM one. This is not due to AM vs. FM, (well, it is a little) but mostly due to the fact that AM is about 1 MHz and FM is about 100 MHz, well above the "static line" around 50 MHz.

It would take a much stronger signal than back then to cause the same level of disruption. Not saying that can't happen, but modern radio communications are quite a bit more robust than they were back over 100 years ago.

The concern is not so much about the disruption of radio communications, but the power grid [infowars.com]. Our society might not survive a massive, long-term (months or even years) blackout (a huge number of transformers might be destroyed all at once by the induced EMF).

Um.

[caffeinated_bunsen]

They do know that solar storms can do absolutely nothing to a data center but maybe cause power outages, right?

Re:

[killkillkill]

Sure they know, but the suckers they sell hosting to at a premium won't.

What about the pipes?

[__aaclcg7560]

Protecting the data center from EMP is one thing. If the pipes to Internet aren't protected against EMP, data entering and leaving the data center will get corrupted. Garbage in, garbage out.

Re:

[Bengie]

You can detect corruption and it would probably cause a completely loss of network connection, which is still better than complete loss of some data and hardware.

Re:

[disambiguated]

If only there were some way [wired.com] to simulate physical systems like that. Where will they find the computing power? Plus the software probably costs [petr-lorenz.com] a [femm.info] fortune [sourceforge.net].

Totally works but we can't tell you how

[GameboyRMH]

My data center is completely safe against tigers. It's due to the construction materials used but I can't go into any detail.

Re:

[caffeinated_bunsen]

Lisa, I want to buy your rock.

Clueless

[Animats]

This keeps coming up. The effects of an electromagnetic pulse and a solar storm are completely different. EMP is a big RF pulse with a risetime in the nanoseconds. This is a risk to input transistors connected to external wiring. Twisted pair, coax, and small mobile devices are relatively immune. Fiber optics are totally immune.

Solar storms induce DC voltages across long distances of conductive landscape. This is a risk only to transformers with grounded center taps connected to long transmission lines.

Re:

[rogueippacket]

Interesting slide deck from PJM - what's more interesting is the reference to a piece of damaged equipment for which the manufacturer quoted a two-year timeline for replacement (during normal business operations!), but a suitable spare had been found elsewhere and put in place within 6 months.
From what I have read on the topic, this is the largest concern - spare parts simply do not exist, and if enough small pieces are damaged at once, they may never be replaced in a reasonable amount of time. Entire comm

Re:

[nblender]

You can design your power supplies to withstand a nuclear event using a NED http://www.maxwell.com/product... [maxwell.com] and when the pulse is over, your stuff just boots and it's business as usual. At least that's my understanding.

Lightning is an EMP too

[fuzzywig]

A good rule of thumb is that if your equipment is protected from a direct lightning hit, then it'll do fine against any EMP that won't destroy any reason for the equipment. ie a really close nuke could produce enough of an EMP to fry it, but would also destroy so much infrastructure there would be no point in having the equipment there in the first place.