Is there test data out there showing breaker failure rates as exercising time increases? NEMA AB-4 recommends annual for mccb's as do all manufacturers and NETA MTS. However this is clearly not industry practice anywhere. NFPA 70B recommends 3-6 years for long cycles which also seems like a huge spread. So it seems like somewhere between 1 and 6 years js the rjght number for exercising (not necessarily testing) is what is recommended but I'm wondering if there is some experimental or theoretical data out there, especially these days with better bearings and lubricants.

I find exercise time spans vary wildly with frame size.
Most <100A MCCB seem to never need regular operation while those >1200 seem to have issues starting after 3 years and especially after 5.

Just as a general rule, where I've worked in the past and where I work now, I've worked with a 3 year maintenance cycle for all distribution breakers.

Our company has been tracking this for 30 years, we test thousands of breakers every year and have found the following:

• Over 30% of low and medium voltage power circuit breakers tested that have been in service for more than 24 months in industrial applications will not perform to specification when “as foundâ€� trip tests are performed.
• After exercise and operation, cleaning and proper lubrication this is reduced to less than 12 %.

Additionally A 2007 IEEE paper addressing the reliability and integrity of low voltage overcurrent protective devices reviewed various surveys. It was found that nearly one�third of all circuit breakers failed while in service and thus would not have been identified unless proper maintenance was performed. In addition, 16% of all circuit breakers failed or were damaged while opening.

NETA also conducted a global study on the topic (Attached)

I have seen the same data from NETA, IEEE, etc. This data is great IF it can be assumed that failure rates are essentially purely random. If they were then failure rates would linearly correlate to maintenance frequency...double the PM's, failure rates are cut in half. Thus we can simply pick an acceptable failure rate such as 10^-5, and then adjust the PM frequency directly based on estimated failure rates of the equipment. This approach was referenced in the above NETA PowerTest survey from an IEEE IAS paper back from the early 1980's.

However, Zog's comments from personal experience also confirm my suspicion that lubrication plays a huge role in reliability of the breaker itself. This is corroborated by NEMA AB-4, NFPA 70B, and NETA MTS, all of which mandate annual exercising followed by 2-5 year cycles for cleaning, relubrication, and testing, although the cycle rate is somewhat arbitrary and seems to suggest that it would be more dependent on the random failure rate of for instance trip units. This is clear evidence for the idea that exercising is a requirement for certain breaker designs.

JBD's comments though also enlighten what most people intuitively believe about breakers. Nearly all residential breakers are 250 A or smaller frames, and PM's (exercising and testing) are almost never done, and that the breakers themselves are of a sealed design where cleaning and regreasing is impossible except in a factory certified rebuild or with special "mining duty" breakers. The referenced PowerTest survey shows that although failures do occur, the rates are less than for other electrical components such as receptacles. Comparing this to the giant IEEE 493 data set suggests that MCCB failure rates with <400 A frames would be acceptably low and that exercising and testing would be required only in the most critical of applications, or again by considering failures as purely random.

The manufacturers of molded case circuit breakers all reference NEMA AB-4 which mandates annual exercising and 3-5 year testing. NETA MTS and NFPA 70B echo the requirement for annual testing and 3-5 year testing. I have only been able to uncover one manufacturer (Powell) which has suggested that extended maintenance frequency breakers with an 8 year cycle are available, although it is clear that they only recommend this for certain models.

I just was hoping that there was something more definitive out there. I'm getting push back understandably for the idea of implementing any kind of testing more frequently than every 3-5 years, never mind following manufacturer recommendations for which there is clear evidence that the requirement is unnecessary and not valid based on collective personal experience at least with one particular application (residential).

What percentage of installed breakers fall into this molded case <400A group?
When people are worried about regular switching of breakers, as switches, aren't most of them molded case?