I have two 60 KW generators 208V running parallel to power some equipment using 118KW load. Will IEEE 1584 applies to arc flash hazard calculations?
It is my understanding that arc flash analysis is not required for voltages 208V generated by less than 125 KVA transformers. How about Generators with 150KVA load?

The passage which is more of a passing reference than an actual "rule" which states that for cases with a single transformer not more than 125 kVA in size with 208 V or less that it "need not be considered" presumably because the incident energy is less than 1.2 cal/cm2 although this was not stated in the standard. Internal data in the joint IEEE/NFPA working group as well as published data show that there are cases where incident energy can exceed 1.2 cal/cm2 below this limit so the revised standard is likely to have a lower limit. It is also not useful for conditions such as any circuit fed by a source other than a single 125 kVA or smaller transformer.

Second, IEEE 1584 contains 3 calculations. The first is a theoretical calculation based on maximum power transfer that can apply to anything but it gives a value that is higher than will ever occur in reality, often much higher. The second is a set of equations that work with a specific set of equipment (fuses and breakers) so it can only be applied with that type of equipment so it doesn't see much use. The third is the empirical model which is based on laboratory tests (apply a specific voltage, current, arc gap, etc., and measure the result) where the model is curve fit to the data. This model has only a single data point at 208 VAC (the rest of the tests failed to produce a stable arc), so the validity at 208 VAC is questionable at best. But if we simply accept that the model works to 208 VAC then it can work but we have to supply the inputs (system voltage, available fault currents, arc gap, etc.). This data comes from some other source of information and IEEE 1584 itself does not specify the source. Typically practitioners use an IEC method which essentially uses Ohm's Law because it uses fewer short cuts and simplifications compared to for instance the most commonly used ANSI method.

So to state whether or not IEEE 1584 "applies" to a specific set of equipment is not easy to answer. It can be applied to anything, especially with the theoretical model. However to say that the results are actually valid and meaningful is a much different judgement call. With anything below 250 VAC it is clear that arc flash does exist but numerical modeling does not appear to be valid. I've penned a paper on this forum on the subject of arcs below 250 VAC summarizing the available literature and I believe one of the software vendors (ETAP) has published something very similar.

Generators are particularly difficult because they have very low fault current capability. With standard 10X Inst trip breakers, they almost never make it into the INST trip region, and thus you have high Cal values because of the long time component. Programmable breakers can over come this because you can adjust the INST range. If being on generator is just a 'once in while' event, like a power outage, I calc the Cal values based upon normal grid fault levels, and include in my label, "Live Work Prohibited while supplied from generator".

hey
We always plan on running a normal utility case and a standby generator case when preparing arc flash hazard analysis because we want to know the short circuit contribution for coordination purposes and then the incident energy available in that mode of operation.

question:
Does anyone know what the flash hazard requirements are for large diesel generator powered mobile equipment like impact rock crusher/jaw crushers? There are electrical components that operate at 480 volts. Is a study required? Warning labels? Is this clearly defined in 70E or anywhere else?

Please advise.
Thanks

No blanket guidance but let's start with the task. For what you are describing obviously there is no need to do energized work as far as changing the wiring so we can throw all those out. Really the only time this becomes a hazard is when you are troubleshooting. So when you are doing that let's say you use a multimeter for testing voltage. These days they come with covers for the probe tips that leave only a tiny amount exposed. Often these get thrown away. But if you are using the probes with the covers, you don't have enough metal exposed to accidentally cause an arc flash. Also consider the panels. If they're hinged, there's no way for the panel to fall into energized equipment even accidentally but it can if it's just bolted on. So for the vast majority of tasks we lack a way to cause an arcing fault in the first place and that should be your first approach and what 70E covers in detail.

The second issue then is everything else and figuring out what PPE is necessary. In the case of most diesel powered aggregate plants though the whole goal here is portability. The engineering studies are designed for equipment that doesn't move. Mining is basically an engineering "nightmare" for the standard method (IEEE 1584.1). So unless you have an engineer on staff, this can get prohibitively expensive paying for an engineer to work out the details with every move, but there are some practical alternatives. The first one is that a recent research paper showed that we can calculate worst case arc flash based on the performance of a circuit breaker without specifically knowing the exact details of all the other equipment. So we could model your situation with say just the generator, circuit breaker, and maybe the crusher, without considering cables, starter buckets, switch sleds, etc. We can establish a reasonable maximum incident energy. This is the best approach because so far the results (via IEEE 1584) are all but guaranteed (no severe injuries or fatalities). It will probably end up at under ATPV 12 which means your electirician(s) wearing arc flash uniforms and maybe a face shield and gloves, nothing out of the ordinary except when they whine in the summer about having to wear a long sleeve shirt like a lineman. The other alternative is to use the table based approach in NFPA 70E. This approach is recommended when an engineering study can't be done for practical reasons but probably will end up with much higher PPE ratings for your situation, ATPV 40. It's designed for the mobile technicians who work in a different operation every day, many of which have not done arc flash studies.

Generator's Arc Flash Label

I like the idea of the mfg putting the arc flash label on but it wouldn't help much except for working on the live generator. The cable connections is about all it would apply to. <deleted link to commercial website> usually produce lower fault current than the utility system and make the whole plant have greater arc flash energies when they are running unless they are running WHEN the utility generation is running. Lots of scenarios that all require due diligence in a well designed arc flash study. We have run as many as 5 scenarios on normal plants and many more on T&D systems and power plants.

Maybe, I guess. I like running normal utility and <deleted link to outside website for advertising> cases for analysis.