NFPA 70E 2018, the NEC continue to emphasize the condition of equipment and the importance of good maintenance practices. Does anyone else get the feel that quality preventive maintenance could almost be viewed as mandatory practice or at least one that meets "best practice" standards?

It would seem like periodic IR inspections would go a long way in fulfilling/ exceeding maintenance requirements. Any thoughts on benefits or compliance afforded with this practice?

Thanks!

Yes. See NFPA 70B, or maybe NETA MTS.

There are three ways to check for high resistance connections: infrared scans, torque testing (not sure this is really valid though...lots of evidence that it's not), or low ohm bridge testing, although again the latter might not be all that valid either as far as predictive power goes. So that leaves IR scans as the ideal. The other two are simply options to use if it's impossible to do IR scans.

It's worse than that. Carefully reading the informational notes basically states that the arc flash engineering study is total garbage and a waste of time, and that some of the "baked in" assumptions about equipment reliability go out the window if you don't properly maintain the equipment. There are no provisions for equipment that is not properly maintained so at that point you are on your own.

I think what's more instructive at this point is getting to the point of qualifying what the bare minimum necessary maintenance is from a safety perspective because there is a lot of stuff in NFPA 70B for instance (and NETA MTS is even worse) that has little or no bearing on equipment safety.

How do you define "no bearing on equipment safety"?

Does it have any direct bearing on potential injury due to shock, arc flash, fire, etc.?

In PSM plants one of the means of classifying equipment failures is to look at every failure mode of every piece of equipment. They get classified into 3 categories: critical, potentially critical, and noncritical equipment. Critical equipment is anything that would for instance in this case directly impact personnel safety. Potentially critical equipment is when a failure can occur but by the failure by itself would not be obvious unless combined with one or more additional failures which then lead to a direct impact to personnel safety. Noncritical is everything else. An example of critical equipment would be for instance a cover missing on an electrical panel. An example of a hidden but potentially critical failure is a breaker with a failed trip unit. You wouldn't know the breaker is bad until something happens where it should trip but won't because of the hidden failure. Under normal circumstances no one would even know the breaker isn't functioning any more until it is called into service.

This generalized approach works for ALL equipment and is well documented as a maintenance strategy for classifying equipment in some versions of RCM (not RCM or RCM2 per se).

Gotcha, I follow what your saying. I'm just hesitant to want a statement like "that has little or no bearing on equipment safety" to go unqualified. Most of what those standards provide is for the maintenance of equipment that does affect safety, maybe not in a critical path sort of way, but something as simple as cleaning the interior surfaces of a switchgear is actually a big deal, we've dealt with many pieces of equipment that arc'd because the surfaces were not clean and compromised the integrity of the insulation between the bus and the siding.

Not to say there are not things that have a very minimal affect on safety, but it's not always cut and dry.

Here are some examples of what I mean:

1. NETA MTS and I've seen it sho up in some other standards specifies to unwire and do insulation resistance testing and then presumably rewiring control wire insulatio. First off, what's the safety impact if any? This is where we get into those "potentially critical" situations. You have to have several failures...the insulation has to fail, AND the overcurrent protection has to fail, AND the wiring has to remain intact through all of this and not cause other operational problems such as causing a starter coil to drop out from low voltage, AND the grounding and bonding has to be so shoddy that elevated voltages are present on the equipment cabinet above 50 V, AND someone has to touch the equipment in just the right way to receive a sufficient electrical shock to die. How many levels deep are the odds here? Let's contrast that with the risk that someone performing this task fails to properly terminate the wiring once it is reattached again, or of the screw or other hardware stripping out creating an equipment failure all in its own, or someone taking short cuts and accidentally meggering something that can't take the 500 V test voltage specified in the standard...you get the idea. As a general practice, this one is just plain bad practice. I could envision scenarios with some sort of critical wiring which is frequently subjected to damage where basically this is the only practical way we have to test for faults but not as a general practice.
2. Testing 120 V/15 A breakers. Two issues here. First off, how many of these are used in residential panels throughout the world? Second, how many failures do we see (hidden or not)? Third, even if one fails, what's the consequence? Generally it means the main trips instead of the small one. While I agree for the most part with practices to exercise and visually inspect them as per NEMA AB-4 standards, actual testing as general routine maintenance for a circuit where a failed breaker really just means loss of overload protection but again we're into that multiple failure case here doesn't result in most cases in a serious problem, never mind the frequency of occurrence. We can then extend this same view point to other cases. For instance is there any logic in testing overload relays routinely? Does anyone actually do it?

I can make a case for testing molded case breakers if:
1. It supplies critical equipment (that's the hidden failure case) and the concern here is a breaker nuisance tripping, not really failure to trip, which would cause loss of the operation of the critical equipment. This makes it a potentially critical piece of equipment (treat same as critical equipment).
2. 800 A or larger frame MCCB's. IEEE Standard 493 data shows a 300% increase in failure rates. So it deviates from the expected failure rates for breakers in general.
3. It shows up as the OCPD on an arc flash study and controls the opening time in the event of an arc flash. This makes it critical equipment from a life/safety point of view. As a secondary effort we may want to look at tripping of the next upstream device (basically eliminate the breaker under question) and see what happens...the practical impact may be the same (PPE level doesn't change) in which case we can say that protection is sufficient (doubly protected) so with frequency of failure so low, no need to test. Or we may just eat it or find that we need to test in general and live with the result. This means that quite often MCC main breakers or breakers feeding MCC's (particularly with main lug only) would need to be tested with some frequency to ensure correct operation.
4. It's the main breaker in a high resistance system with ground fault protection.

Your cleaning example as a PM (as opposed to PdM) is a good one. This is an example of a variable-length PM frequency though. It depends on where it's at. Say you have a customer with a coal loadout facility (like I do). That equipment can get pretty filthy in a year's time. This is as opposed to a lot of power plants with great ventilation and filtering where after 10 years plus it still passes the "white glove" test.

Same thing can be said but is on more of a fixed schedule for lubricating open frame breakers. Most of them use Mobil 28 where the oils in the grease tend to evaporate, requiring periodic relubrication and most manufacturers specify every 3 years. Newer ones are starting to use Molykote 3451 which has at least an 8-10 year operating life according to third party (Dobles) testing.

Either way, the key thing is if it is critical equipment or potentially critical or not. If it's not, then we run to failure. If it is, then if there is a cost effective and proven predictive or preventative maintenance actiivity, we do that. For example do you change your oil in your car/truck? If you don't, it quickly fails. That's preventative maintenance at it's finest. Second do you check air pressure in your tires routinely? If you don't it can be come a serious safety problem and/or lead to lowered efficiency. So it's kind of a middle of the road sort of thing. Now, do you check the coolant pressure in your car/truck air conditioning system? Almost nobody does. That's an example of a noncritical system. The result of a failure, even if you live in the South, is really more of an inconvenience. You can always roll down the windows. But a lot of car/truck owner's manuals specifically say to check this on some particular maintenance frequency. If the coolant level gets low all you do is top it off and/or eventually start looking at replacing lines, seals, and pumps if it gets to the point where the leak is an issue. But basically the normal practice is to run to failure on air conditioning.

So I'm not advocating for not doing maintenance at all...far from it. What I'm advocating for is that the mainteannce standards are just that and are focussed on reliability while a standard that applies to safety would have a slightly different and much more restricted scope of focussing on the portion of the reliability standard that addresses safety. A great example is the difference between a "failure to trip" and a "trips too soon" failure with a circuit breaker. From a reliability point of view, both are bad. From a safety point of view, tripping too soon is harmless but failure to trip is basically the ultimate hidden failure issue.