Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

We are in a fish processing unit where the temperature is constantly at -10°C (+13F).

If we use a machine in here, will it be at risk of damage from the cold? Is there any considerations we need to make?

share|improve this question
Is that 13°C or 13°F (or 13K :-))? –  RedGrittyBrick Nov 17 '12 at 11:04
@RedGrittyBrick I hope for the OP it's not -260.15°C :) –  amiregelz Nov 17 '12 at 11:10
Sometimes liquid nitrogen cooling is required. tomshardware.com/news/… –  Aki Nov 17 '12 at 13:47
-10 C is not especially cold, especially considering that the unit will warm 10 degrees or so once turned on. Your main hazard is when you're done in the cold and bring it into a warm area -- wrap it in plastic before moving and don't unwrap for an hour or two -- until it's had time to warm up to room temp. This is to prevent condensation inside the unit. (Taking it warm to cold, OTOH, do not wrap it first.) –  Daniel R Hicks Nov 17 '12 at 18:23
@RedGrittyBrick: You forgot 13°R. =P –  Mehrdad Nov 17 '12 at 20:03

5 Answers 5

If we use a machine in here, will it be at risk of damage from the cold? Is there any considerations we need to make?

Consider a hot summer's day, and you've just finished cutting the grass in your backyard. You go inside and bring out a cold beverage, sit down, and enjoy the landscape... That's when you notice that your drink is covered in condensation.

Consider a cold winter's day, you've just finished shoveling off your driveway. You go inside and bring out a hot beverage, sit down, and enjoy the landscape... That's when you notice, well, nothing - you won't have any condensation formed.

This is why in most cases, simply exposing a computer to the cold won't do any damage. Computers enjoy being cold rather than hot, and so long as you're not moving a cold computer into a hot room (like the first example, this could cause condensation inside the case and thus short-circuit some components), there shouldn't be a huge issue.

Do note that if you ever have to bring a freezing cold computer back indoors (i.e. you stored the computer outside at -10C, and bring it inside at 21C), you should make sure all of the condensation has been cleared before plugging it in and starting it up. If you're unsure, I would let the computer sit inside for a few days for any moisture to evaporate, and for the temperature/humidity in the case equalize.

Lastly, also ensure that if the computer is exposed to the weather elements (i.e. precipitation, high humidity, etc...), additional precautions/safeguards might need to be taken (waterproof case, moisture-proof case, proper ventilation).

share|improve this answer
As already noted in the comments above, if there's no existing condensation or frost on the computer before you bring it inside, a good way to prevent any from forming is to wrap the computer in something relatively airtight (e.g. a plastic bag) before bringing it in, and to let it warm up wrapped like that. That way, all the moisture from the air will condense on the outside of the bag, while the inside stays dry. –  Ilmari Karonen Nov 21 '12 at 11:33

As it happens, there's a bunch of experimental servers running quite close to my office — specifically, on the roof of the building I work in — that may provide some relevant real-world data.

They're part of an experiment by the Helsinki Institute for Information Technology on cooling servers with unfiltered outdoor air at ambient temperatures ranging from around -20°C to +20°C. Originally, the servers were simply placed in an off-the-shelf camping tent (to protect them from snow, rain and pigeon droppings) and left to run over the winter. Since then, they've built a small rooftop greenhouse to run the servers (and grow some plants) in.

So far, the experiment seems to be a success. As one of the researchers writes,

"The intake server air is not filtered in any way. The most frequently asked question is, whether the moisture condensing inside the servers harms them. However, the condensation works in reverse, since the servers are cooled by air, so they are always warmer than the air around them."

share|improve this answer

As far as component failure, you are in serious personal danger at -20, -30 degrees that the device will fail at. however, humidity and saturation will become a problem quickly as the components gain moisture.

if the device is is a constant very low temp environment, a sealed case and externally attached heat-sinks will do fine.

share|improve this answer
"Serious personal danger" is probably exaggerating it a bit, even at -30°C, unless you get locked in there overnight without adequate clothing to keep you warm. (If that does happen, you do run a serious risk of hypothermia, possibly leading to permanent injury or death. But some nice warm winter clothing should be enough to keep you safe, and maybe even relatively comfortable. I should know -- those are not unusual winter temperatures around here.) –  Ilmari Karonen Nov 18 '12 at 12:44
Mild 35–32 °C: shivering, vasoconstriction, liver failure (which would eventually be fatal) or hypo/hyper-glycemia (problems maintaining healthy blood sugar levels, both of which could eventually be fatal). link –  user174216 Jun 28 '13 at 12:45
The text you quoted is about reduced body temperatures (35-32 °C is 95-90 °F). The amount your body temperature drops in a cold environment depends entirely on how long you spend there and how well insulated you are (plus, to a lesser extent, physiological details like size and amount of body fat). With sufficiently insulating clothing, you can spend as long as you want in a -30 °C environment without your body temperature dropping significantly from normal. Just ask the Inuit. –  Ilmari Karonen Jul 1 '13 at 3:29

It may not work stable when temperature is too low because electrolytic capacitors did not work properly when temperature is too low. Especially when the power supply is cheapest possible (with small capacitors).

Additionally, low temperature is a problem for most aluminum capacitors: for most types, capacitance falls off rapidly below room temperature while dissipation factor can be ten times higher at −25 °C than at 25 °C.

Most limitations can be traced to the electrolyte. At high temperature, the water can be lost to evaporation, and the capacitor (especially the small sizes) may leak outright. At low temperatures, the conductance of the salts declines, raising the ESR, and the increase in the electrolyte's surface tension can cause reduced contact with the dielectric.

The conductance of electrolytes generally has a very high temperature coefficient, +2%/°C is typical, depending on size. The electrolyte, particularly if degraded, is implicated in various reliability issues as well.

Source: Wikipedia: Electrolytic capacitor -> Reliability and length of life

This is why only CPUs (and other semiconductor components) are cooled to low temperatures, not the whole computer.

The normal working range for most capacitors is -30°C to +125°C

Source: Electronics Tutorials: Electronics Tutorial about Capacitor Characteristics

share|improve this answer
Also HDDs cannot work under very low temperature as there are lubricants in there. They're also prune to damange in wet environments. You'd better get specially made industrial models. –  billc.cn Nov 17 '12 at 19:07
If the normal working range for "most capacitors" goes down to -30°C, I would think that -10°C shouldn't present too much of a problem... –  Michael Kjörling Nov 17 '12 at 19:07
What about the temperature increase from the computer running? –  cmorse Jan 3 '13 at 20:02

As long as you keep the machine dry, there shouldn't be any real problems from running a PC in the cold. Where you will run into problems is if condensation is allowed to form on the components.

I have installed computers into blast chillers and refrigeration factories without problems, but we did need to make sure the cases were put into enclosures that keep moisture out.

We actually purchased special electronics cabinets for that purpose, but its not expressly necessary as long as you can keep the moisture level under control.

share|improve this answer
There were companies cooling CPUs down to -40°C (for overclocking) and they sealed the underside of the CPU to prevent humidity to condensate there, which would have shorted the pins. So just to clarify, that condensated water does not need to turn to ice to create a problem. But Jareds answer is perfectly fine in my opinion. See this TomsHardware Article tomshardware.com/reviews/… –  TheUser1024 Nov 17 '12 at 10:55
edited. thanks for the correction. –  Jared Tritsch Nov 17 '12 at 11:08
@TheUser1024 - They only sealed the bottom of the socket because it was exposed to ambient (~20°C) air. You don't get condensation on a cold surface is the air it is in contact with is the same temperature. –  Fake Name Nov 18 '12 at 9:14

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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