Our company has been discussing arc flash propagation for several weeks, do we use it, do we not, for panels, for MCC's..... The discussion finally made it to the point that this is not so much a technical issue (no real supporting data) but a legal issue. I am positive the organizations governing arc flash will not take an official stance without supporting evidence for or against. So the question arises, if there is a reason that I should need to defend my position do I lean towards the scientific evidence (or lack of) or do I lean towards the general consensus of my peers (propagation needs to be considered)?

Our lawyer pointed us to the Precautionary Principle. "The precautionary principle or precautionary approach states if an action or policy has a suspected risk of causing harm to the [url='http://en.wikipedia.org/wiki/General_public']public[/url] or to the [url='http://en.wikipedia.org/wiki/Natural_environment']environment[/url], in the absence off[url='http://en.wikipedia.org/wiki/Scientific_consensus']scientific consensus[/url] that the action or policy is harmful, the [url='http://en.wikipedia.org/wiki/Legal_burden_of_proof']burden of proof[/url] that it is not harmful falls on those taking an action." (http://en.wikipedia.org/wiki/Precautionary_principle)

Basically if a judge drove across town(or country) and rounded up engineers and electricians familiar with the subject would they say "propagation must be considered". Even if there is no proof we shoulder the burden of providing the safest arc flash study(action), failure to do so places the burden of proof on our shoulders that is the safest practice.

Thoughts?

I think you are seeing a slow (glacier speed) shift towards more common sense interpretations. When arc flash studies were in their infancy, most people would not use the switchgear main as the device that defines the arc flash clearing time. Also, most people would not use the 2 second rule, the list goes on. It was mostly "fear" of the legal backlash and not having any precedent for doing anything different.

Over time, people began to realize the propagation issue with switchgear may not be as significant when it comes to the main. People are now beginning to realize this may also apply to propagation from one section to another but I'm not sure anyone is factoring that part in yet.

Propagation depends on so many factors such as equipment design, short circuit current, duration etc. that people stick with the legally conservative interpretations until they have more data. I was involved with testing about 2 months ago with an MCC section as part of a forensic study. The arc flash was staged in the middle of the section and of course the arc ran away from the source towards the bottom. However, the plasma cloud was large enough that the arc re-struck towards the top of the section. Although there are lots of individual stories like this, there just isn't enough data to support good solid conclusions.

Unfortunately, the days of engineering judgment have given way to the potential for legal action and that drives many technical decisions.

[INDENT=1]Newton's 3.14159th law: "For every action, there can be an unequal and opposite law suit" [/INDENT]

(ok, I just made that one up )

Following Jims example and "lawyering" it, there is no litigious limit as to what MIGHT happen, so the conservative approach is to eliminate electrical power in the workplace. Who is to say that the protective device does not fail, propagation does not enter an adjacent building, meteors simultaneosly hit the building, and lawyers short everything out as they descend on an accident like a cloud of locusts.

The conservative approach is to stick with known consensus opinion, not conjecture and guessing. if we go down that road, the ambulance chasers win and only legislative action to limit tort law will bring us back from the brink of absurdity.

Adding fuel to the legal fire is the disfunctional way in which NFPA 70E and the other safety codes are written. These codes are clearly the work of disjointed, self-interested committees, with little or no over sight or editing authority. For example, article NFPA 2012 340.5 Hazardous Effects of Electricity on the Human Body; which is a good informative paragraph, but is only included as subset of "Power Electronic Equipment". Huh? do these hazards not apply in other contexts?

That phenomena is multiplied because there is the duplicity for many of these requirements within OSHA 29 CFR 1910 Subpart S and OSHA 29 CFR 1910.269 with two major overlapping safety chapters, any of which can apply to a site with medium voltage distribution. Clearly these documents have been heavily lobbied by groups whose interests have superseded the need to communicate clearly the safety requirements.

Finally, our State of Washington has transcribed many of these requirements into their own Washington Administrative Code. I used to be aware of what minor differences they included but at this point can not identify it. The WAC does not supersede OSHA standards, we remain liable for any intimacies that might slightly vary between the two.

If these codes were included as a work scope description to describe the terms for which work was to be completed, it is felt that a contractor could have a field day with change orders and extended deadlines; there is nothing succinct. A whole bunch of randomly organized suggestions that all become major liabilities if an accident occurs.

The present arrangement leaves one with a jaded feeling that the State Authorities desire this confusion so they can arbitrarily assess penalties anytime that an accident occurs. That might be the purpose.

It seems that you're considering two options regarding how to treat propagation/compartmentalization:

1) Rely of the judgement of your engineers and electricians, which is likely based on the personal experiences of this group of people, and reach a conclusion that is technically dubious, based on a small body of data, and legally perilous.

2) Follow the conservative consensus of the industry which considers metal-clad switchgear to be compartmentalized, but considers all other equipment to be a single compartment from a propagation perspective.

I would suggest a third alternative:

3) Conduct a series of tests to collect additional data for specific types of equipment of which you have a large installed base. The test data could be used to backstop an internal engienering decision to treat certain equipment in a certain way regarding propagation. This approach is commonly taken by some of the largest corporations to investigate the safety aspects of a particular situation in which the equipment manufacturer and the consensus standards are silent.

or you could do it this way:

3a) Make a donation to the [url='http://standards.ieee.org/about/arcflash/']IEEE/NFPA Arc Flash Research Project[/url] with the caveat that the funds be used to test the specific equipment you are interested in as described above. Then the testing and results would come from an independent third party, and the results could be shared with the P1584 working group and others actively working on better understanding arc flash phenomena.