This is a very common problem in industrial control panels for the same reason. These are panels such as used in assembly lines, machine tools, and so forth that hold multiple open face contactors mounted either on DIN rail or screwed to a back panel. Panel fuses or breakers, jf present, are mounted the same way. There are often relatively small instrument transformers stepping down to lighting voltages, which then run controls. If you are building these, I highly recommend compartmentalizing into 3 compartments, and sheet metal is cheap. Compartment 1 contains an incoming protective device. Compartment 2 contains the motor contactors, drives, and voltage conversion. Compartment 3 houses controls. Many companies (GM and Ford I know for sure) have tried to go to under 24 VDC only in the control compartment. The problem with this is that it is difficult to operate actuators for hydraulics and pneumatics (solenoids typically exceeding 1-2 A), and contactors usually end up with a very nonstandard voltage like 48 VAC. Or a layer of isolation relays gets added which reduces reliability and consequently increases exposures to the 'high voltage' compartment. I recommend instead using 24 V where practical and lighting voltages (120/240 in US) where needed in isolated terminal strips. Cut and strip everything to proper length, use fusing or circuit breakers in the noncontrol cabinet, and use hardware that is 'touch safe' which means screws, terminals, etc. are recessed. That is how most gear is built today. Even if you have an older panel, it is usually relatively easy to achieve touch safe status. Open frame transformers are usually the biggest problem and they can be insulated or caged relatively easily. Once this is done then look at how an arc flash can be created. Depending on the environment, opening the doors may or may not be prone to causing an arc flash such as if there is conductive dust or liquids around that could start something. Once the doors are open then visual inspection does not require crossing the restricted approach boundary. If the tools are insulated to prevent both phase to ground and phase to phase contacts (using insulated screwdriver, insulated meter probes with the tip covers on) then at that point even with 480 Volts, initiating an arc is very remote. I dealt with this exact problem in a foundry that had a lot of very old equipment and a lot of industrial control panels. The trick as was already mentioned is attacking the source of the problem (initiating an arcing fault). Throwing PPE at it is not really the recommended approach. Eliminating sources of arcing faults engineers the hazard out of the system rather than simply accepting it and doing what we can to protect personnel. Predicting arcing faults is NOT an exact science. IEEE 1584 shows that if proper PPE is worn, there is a 95% chance it will work. The other 5% is due to the variable nature of arcs. If we can engineer out the things that electricians can do to accidentally cause an arc, the remaining equipment failure related cases run at about 0.0001% to 0.00001% or lower (about 1 in a million), based on statistics on older designs, compared to human error rates that are around 0.1 to 40% depending on a large number of largely uncontrolled variables as verified by various military studies. Even with the PPE, that still puts arc flash injury likelihood in the absolute best of circumstances well above an engineered approach. H/RC 3 is only going to lower visibility, dexterity, comfort, ergonomics, heat stress, and otherwise contribute to increasing odds of an arc flash. It is better to take human error out of the equation as much as possible. And if the remaining arcing fault likelihood is tolerable, then adding PPE would no longer be necessary, and may even be counterproductive to increasing safety because it increases other risks such as dropped tools, stabbing injuries, heat stress, slips/trips, and misidentification due to poor visibility.
|