33

I got a field report of a product which is sealed in an IP66 aluminum die cast body. it was observed that water/moisture had condensed inside the enclosure over and over and creating some white marks inside (see attached). The product was installed in a US region about a year ago.

Corner of aluminum box sowing white discoloration and wiring

It was reported the top cover was screwed on properly, and there is no way water can get in. However one question in my mind is: we install this product on a street light pole, and the top cover is screwed on in the field only after the power cable gets in. So what if the atmosphere is already full of moisture during installation? That moisture will be trapped in the enclosure and sealed after covering the top. And due to the weather changing over and over that moisture gets condensed over and over and creates the problem. How can I get out of this? One way I think would be to use some kind of silica bag to absorb moisture. Any advice?

psmears
  • 678
  • 4
  • 6
Transformer
  • 667
  • 1
  • 7
  • 18
  • I don't think there's much more to say that you don't already know. You're confident that the enclosure didn't leak, so a desiccant (silica gel) and/or filling the enclosure with a dry inert gas are pretty much your only options. – brhans Aug 01 '18 at 17:25
  • 11
    Backfill with dry nitrogen gas just before sealing down the unit. But my guess is that the unit isn't as sealed as you imagine, with wires entering and exiting the box, and folks "man-handling" during installation or re-work. – jonk Aug 01 '18 at 17:25
  • 11
    Note on "trapped humid air when box was closed" At 25C, there is 22g/m^3 of water, or 22mg if this is a 1L box. Thats a 3mm ball of water. More than that and it got in afterwards – Henry Crun Aug 01 '18 at 22:16
  • A customer wanted to put some of our electronics that we supply in NEMA 4 boxes underwater and found that Magic Gel works great. http://www.raytech.it/product/low-voltage/magic-gel?lang=en – Warren Aug 01 '18 at 22:15
  • By "before one year" do you mean "less than a year ago", or "more than a year ago"? – psmears Aug 02 '18 at 15:53
  • approx 1 year completed this week – Transformer Aug 02 '18 at 16:24
  • OK - I've submitted an edit that makes it clearer :) – psmears Aug 02 '18 at 20:46
  • Were waterproof cables used? What about the IP rating of the glands, are those IP66 too? I have a tendency to stick IP68 glands (mushroom caps) on boxes which I know will be out in the weather, there's a definite difference between them. Were the glands properly closed too? Keyword in all this is 'creepage' of water. – Mast Aug 03 '18 at 07:23
  • Liquid tight, non-metallic strain relief cable gland with metric threads, black, thread type and size: M-16X1.5, UL recognized. Thread: Metric Thread type: M16x1.5 Construction material: Polyamide, CR Color: Black Clamping range: 0.177-0.392 Protection: 70 PSI IP rating: IP68, 5 Bar – Transformer Aug 03 '18 at 08:32
  • 1
    Also see [my answer at IoT Stackexchange](https://iot.stackexchange.com/a/1114/1728). – koalo Aug 03 '18 at 16:05

9 Answers9

53

Here's what happened to a product I designed that was IP69K (Protection against powerful high temperature water jets). It was installed in China and failed after about a couple of years. When looked into there was about 1.5 inches of water slopping about. Seemingly a mystery.

However, it turned out to be a progressive build up of humidity. Every night the product was turned-off and internally it started to cool over night. During that period the internal pressure lowered (because the internal air cooled) and it sucked-in high-humidity air from the surroundings. That settled out on top of the microscopic layer of water that was ingested the previous night. Gradually it built up until it reached a switch mode power supply then bang.

The way we got out of this was selling the option of having a drain plug fitted and each order was processed with attention to that humidity problem.

Silica gel will work but only has a limited lifespan before it needs to be replaced. A positive pressure inside will help - if it's sealed as good as you suggest you can buy slow leak gas cartridges that keep the pressure positive - they are used in the petrochem industry and are akin to flameproofing by using pressure to stop volatile gases entering and potentially being ignited.

Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • 9
    I love the idea of a slow leak gas cartridge to maintain a slight positive pressure -- if it is possible to include an indicator when the cartridge is empty so that it can be replaced. – jonk Aug 01 '18 at 17:31
  • I asked the vent manufacture like Gore, Wiska and porex for vent solution on this, waiting for their reply , anyone used this thing before how the field report of these vent, will they really solve the problem i have seen in my device. – Transformer Aug 01 '18 at 17:53
  • Also I want to recreate this issue on temperature chamber , can it is possible in short frame of time within a week. – Transformer Aug 01 '18 at 17:55
  • 1
    @Transformer cycle it hot and hard and find a way of examining the moisture build up without taking the test rig apart. Maybe get a window added to the box that is adequately sealed. – Andy aka Aug 01 '18 at 18:52
  • 7
    +1 Great story from the trenches. @Transformer Our environmental chamber has a control humidity option using wet/dry bulb. Maybe you can simulate that by heating an open tray of water inside the chamber if the local isn't high enough (heating the air will reduce the RH). – Spehro Pefhany Aug 01 '18 at 19:02
  • 14
    The definition of waterproof is that the hole in the bottom is bigger than the hole in the top. – Henry Crun Aug 01 '18 at 21:01
  • 2
    @Transformer When you have a heat source inside your unit (i.e it runs warmer that the outside environment, then porous vents are able to work better, as the vapor pressure inside will exceed that outside, so water will diffuse out. Daytime Sun can also create this condition. Note: goretex vents must be where liquid water won't cover them – Henry Crun Aug 02 '18 at 03:51
  • 2
    For testing it may require long hot cold cycles, as you are expecting moisture laden air to seep slowly past a seal. Each litre of air exchanged (which might take, say 100 cycles), only brings in a 3mm drop of water. 10 cycles over a couple of days isn't going to do it. – Henry Crun Aug 02 '18 at 03:59
  • 6
    @Andyaka Are you sure the "slow leak gas bottle" isn't something you send the apprentice for, along with striped paint and a long weight? – Henry Crun Aug 02 '18 at 09:20
  • @HenryCrun haha. It's a genuine part fitted to boxes using Ex 'p' protection: http://atex-enclosure.com/solutions/pressurisation-and-purge-systems-ex-p/ – Andy aka Aug 02 '18 at 09:25
  • Put it in a taller box, with the kit at the top. 1.5 inches every couple years so a foot of clear space should take 10 years to fill. What's the life expectancy of the item? – Criggie Aug 02 '18 at 20:01
  • 2
    @Criggie :- Its about 10Years – Transformer Aug 03 '18 at 08:36
22

Andy exactly describes the mechanism I have also seen . Because of this I favor inverted bucket case designs which do not rely on seals, wherever possible. "The definition of waterproof is the hole in the bottom is bigger than the hole on top" Note on "trapped humid air when box was closed" At 25C, there is 22g/m^3 of water, or 22mg if this is a 1L box. Thats a ~3mm ball of water.

I found scotch-brite pads to be a good breather vent and bug stopper for fixed equipment.

I have also just cut a disc of Gortex raincoat fabric and siliconed it into the case. (on top of metal mesh if you have gnawing bugs). I did some transpiration experiments years ago testing the 6 different types of breathable fabrics that I could get. (silicon fabric onto top of cup of water, and leave in hot water cupboard). Genuine Goretex was more than 2x better than the clones at the time.

There are specialist screw in and stick on breathers nowadays. (A lot smaller than the 1.5" Gortex patches I used though)

Mounting the case at 10 degrees to horizontal results in the water pooling in the corner. It is good to plan for where the water will go. A small drain hole there will see the overpressure force liquid water out the drain hole during the day. Put a some type of wick in drain holes so water drops do drain down.

A sunshade and/or white colour reduces the internal temperature, and therefore the overpressure that forces air out during the day (and thus sucks moisture in at night. As an aside, there is IR reflective paint that results in dark colors in the sun being 10degC cooler.

In a recent datalogger design used in the splash zone on boats, I put a hygrometer chip in, just to know if water had leaked in.

Henry Crun
  • 5,273
  • 11
  • 12
20

I once found water in a box like that installed INSIDE a house. I was baffled about where it could come from. Turns out the water wicked itself through the wire going into the box, coming from another (similar) box on the roof that was full of water. So basically the electrical wire was operating as a hose. I can't see very well on the picture you're providing, but if applicable, maybe investigate the possibility that water came in through the wiring? Even if the wires only come in from underneath, if the wire is going up, the pressure could force the water through the wiring. Hope this helps.

MartinF
  • 1,230
  • 7
  • 13
13

What you want is to prevent condensation from the thermal difference between the case and internal air. This depends on the relative humidity of the internal air to not exceed the dew point.

IP67 specifies prevent vapor pressure leaks submersed to 2~3m under water tested for >=1hr as per supplier specs.

IP66 only protects against external water spray and IP69 only protects against higher pressure water spray.

But neither can guarantee prevention of condensation if the RH of contained air is high.

Possible remedies

  • power wire entry voids the IP66 rating so liquid rubber sealant is needed
  • Epoxy Conformal spray to case interior
  • PCB with ENIG pads or equivalent gold alloy.
  • Conformally coated PCBA with mask for LEDs.
  • Desiccant.
    -Teflon gas plug
  • internal heater (Hygrotherm)
  • thermal insulation to case interior

The last suggestion requires measuring or calculating the need to conduct heat outside and prevent a temperature gradient on the inside inducing condensation above Dew Point.

The Dew Point drops with rising air pressure so allowing an H2 pressure leak with a Teflon plug still blocks Vapour Pressure of H2O.

Each of these solutions may cause new thermal problems if you need heat conduction, so this is a factor to consider.

  • The neoprene gasket seal must be a smooth interface on the case to have a constant high pressure seal which may be a problem on aluminum castings.

  • Testing should include rapid temperature cycling and monitoring internal condensation with condensation ink dots ( like those used in Apple products which change colour) and perhaps internal pressure.

  • Avoiding Solar heating also helps if possible for reducing thermal shock.

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182
5

I've observed, learned from (the hard way), and later protected from this phenomenon on a number of outdoor products. You can also see a similar effect if the enclosure becomes very hot (e.g. direct daytime sun) and then experiences a cold rain shower. The rapid temperature change cancreate a negative pressure capable of sucking water in right through NEMA 4X (hosedown) rated seals.

If you want to maintain your ingress protection rating while allowing for the exchange of air, you can use the product we ended up selecting, a breathable-membrane sealed vent from Gore: https://www.gore.com/products/categories/venting?view=protective-vents-for-outdoor-electronics

EnemyBagJones
  • 151
  • 2
3

As long as ambient air can reach the area changes in weather will cause condensation when the air is humid and the metal is cold. The same problem can happen in houses that don't have good vapor barrier protection in the walls- protecting the inside of the perimeter walls from the humid air the can come from living inside a house where bathing, cooking, breathing occurs. When the vapor barrier is inadequate to the point where humid air can penetrate the cold outside/perimeter wall if there is not something to fill the wall space- insulation, the humid air will condense on the wall and can cause ice dams, moisture collection and ultimately rot.
In looking at your problem from this perspective, air sealing the enclosure and filling it with insulation is my answer.

Sam Williams
  • 31
  • 1
2

You are already in the right direction. That happened also to us in a completely sealed product we designed.

Just to add a graphic description of what the other people are describing, check this video: https://www.bopla.de/en/service/technical-information/pressure-compensation-elements.html

Marcos
  • 144
  • 5
  • :- Thanks , i saw this video, one think i do not understand after test why i saw water in one unit which does not have vent, what physics behind this, is it pressure build up which sucks the water. – Transformer Aug 02 '18 at 16:30
1

Search for something "to avoid accumulation of water" like from Rittal:

Product – System accessories – Base – To avoid accumulation of water:

  • Pressure relief stoppers ...

  • Condensate discharge for reliable discharge from the inside ...

Community
  • 1
alex
  • 21
  • 1
1

Heating is the solution! It is really difficult to seal enclosures and prevent to getting water from outside when outdoor. One solution is to active heating when temperature going down and humidity up. It is handy to measure inside enclosure (humidity better than temperature) to save energy and prevent overheating - turn on heating only if humidity arise ~70% (or temperature steep down). It will helps by those principles:

  1. avoid to condensate by pushing up condensation point.
  2. helps to ventilate thru if enclosures fairly open (avoid insect by net)
  3. prevent to suck in outdoor cold and possibly humid air in relative tight enclosures

Heating definitely dry out outdoor enclosures, avoid to overheat in full sun. In some condition is enough native heating from running electronics with enough ventilation, but you need to mechanically protect from jet water.

winny
  • 13,064
  • 6
  • 46
  • 63
Ales
  • 11
  • 1