I have a fleet (100+) of 100 kVA and above uninterruptible power supplies across the UK. They are designed to be super reliable. I work as the "reliability engineer" so local engineers deliver and oversee the maintenance, I am supposed to do most of my job from a desk with occasional site visits.

A few UPSes have exploded recently with big arc flashes. The OEM maintains these UPSes, and when they came in to investigate, the OEM said they couldn't find the root cause of the arc flashes but they observed the interior of the UPS cabinets were very dusty and dirty.

I am aware that dirt and dust can cause thermal issues, although I do not think there should be a problem with arcing because the boards are conformal coated.

I would like to develop a key to advise the local engineers on if levels of dust and dirt are a problem. In theory they know about Foreign Material Exclusion and should know any amount of it is bad, but I have been challenged back to clearly define when dust goes from "just one of those things" into something that needs to be cleaned or becomes a risk to the reliability and safety of the machine.

I haven't had a lot of luck finding a clear way to define levels of dust and dirt in terms of maintenance of power electronics - does anybody here know where I could find that information?

  • \$\begingroup\$ Is dust and dirt deposited on your equipment and you want to quantify how much is there? It's not airborne? \$\endgroup\$
    – jonathanjo
    Commented Oct 18, 2022 at 7:06
  • 2
    \$\begingroup\$ Doesn't the UPS supplier have a commitment to finding the root (or at least, assignable) cause and then corrective actions?? This is industry standard stuff. If they just push this off back onto you by suggesting "dirt" at your site, that's not being very pro-active. They should take this far more seriously, I think. Or find another supplier that does meet the ISO 9000 series of quality controls. \$\endgroup\$
    – jonk
    Commented Oct 18, 2022 at 7:18
  • 2
    \$\begingroup\$ They should not be throwing mud at the wall to see what sticks with you. They should analyze their own product, find the cause, and then tell you exactly what it is. If it is your problem, you at least know for sure because they tracked the problem down to the root cause. On the other hand, they may find that they have a design flaw. Which they should fix. They should tell you what caused the problem. \$\endgroup\$
    – jonk
    Commented Oct 18, 2022 at 7:25
  • \$\begingroup\$ do you want to measure dust on surfaces of the equipment? there isnt a standard for that because while the effects of dust are known it will change depending on the equipment. I also highly suggest that you also measure temperature and humidity as these also have a big effect on high voltage arcs \$\endgroup\$
    – Juan
    Commented Oct 18, 2022 at 7:26
  • \$\begingroup\$ I was imagining some kind of cotton-pad wiping the surface and compared against reference darkness. Good refs here: Estimation of dust concentration by a novel machine vision system. And I wondered if your equipment has an IP5X rating? Potential ways of continuous and airborne measurement: link \$\endgroup\$
    – jonathanjo
    Commented Oct 18, 2022 at 7:32

2 Answers 2


The pollution degree should be able to provide an answer, or at least a start.

What's the pollution degree the devices are rated for, and how would you rate the pollution degree that is to be expected when looking at the actual environment the devices are operated in?

In theory this is standard stuff (pun intended), cf. IEC 60947-1 and IEC 60664-1.

Of course, we all know that reality is sometimes more messy, but the numbers given in the standard may be a start when discussion issues with your customers or techs.

The number used to rate the pollution degree is based on the amount of conductive or hygroscopic dust, ionized gas or salt and on the relative humidity and its frequency of occurrence, resulting in hygroscopic absorption or condensation of moisture leading to reduction in dielectric strength and/or surface resistivity.

Standard IEC 60947-1 distinguishes four pollution degrees.

  1. No pollution or only dry, non-conductive pollution occurs.
  2. Normally, only non-conductive pollution occurs. Occasionally, however, a temporary conductivity caused by condensation may be expected.
  3. Conductive pollution occurs, or dry, non-conductive pollution occurs which becomes conductive due to condensation.
  4. The pollution generates persistent conductivity caused, for instance, by conductive dust or by rain or snow.

Based on what degree a device is designed for, isolation spacing will have to be laid out in different ways, and it tells you what type of fan or air filter, if any, you may use when designing a device.


Here's a starting point:

  • National Research Council 1982. Classification of Dusts Relative to Electrical Equipment in Class II Hazardous Locations. Washington, DC: The National Academies Press. https://doi.org/10.17226/10952

Down as guest from http://nap.nationalacademies.org/10952


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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