Carlos said:
Do you have more documents like this for generators and transformers?
thread237-364123
More for motors - EPRI 5936, volume 16, volume 17
generators - EPRI 5936, volume 1
transformers - EPRI 5936, volume 2
IEEE 43 mentions in an appendix there are other variants of the ratios using different times, but the limits aren't standardized.
There are 3 main components of currents measured in insulation resistance testing:
1 - capacitive
2 - polarization
3- resistive (includes surface leakage and volume leakage).
Capacitive decays the fastest, mostly gone within 30-60 seconds
polarization lasts longer, mostly gone in 5-10 minutes (for form wound)
resistive is generally constant.
The rough idea of the P.I. is that you're using polarization current as a benchmark for evaluating resistive current. In simplified terms you might think 1 minute current is mostly polarization and 10 minute is resisitve. So it is something close to (not exactly) ratio of resistive current to polarization which (to me) is vaguely reminiscent of ratio resistive to capacitive current we look at in power factor measurements. In both bases we use something like a ratio to a reference current (capacitive for pf or polarization for PI) to account for geometry factors when we evaluate resistive current.
The motivation to use D.A. (ratio 1 minute/30 sec) is as mentioned in the Biddle document to save time. But it also comes into play more often in conjunction with random wound machines. While form wound machines have almost exclusively dieelctric insulating materials between conductor and ground, random wound are going to have slot liner materials and air (and varnish for that matter). So random wound tend to not have as much polarization components. For some random wound motors it is hard to pass PI even when the insulation is dry and healthy.
As a result PI is not often recommended for random wound motors. D.A. is sometimes recommended in its place and it makes a little sense that in the case of D.A. we are using an earlier time for the reference which may include the capacitive current. Also saving s in time may be more relevant for a small random wound motor than a big form wound.
Here are some references directly or indirectly implying limitations of applying PI to random wound
* IEEE 62.2 states that PI values will often be around 1.00 for small motors
• EPRI NP7502 – Motor predictive maintenance – recommends PI for form wound, not for random wound.
• EPRI PM Basis for Medium voltage motors 106857 Volume 9 recommends PI. EPRI PM basis for low voltage motors 106857V8 does not.
• EPRI 1008964 (Repair Spec for low voltage motors) states PI is required for form wound stators only.
•
EPRI 5036, Volume 6, page 6-94 states "A PI of 2 or more is considered satisfactory for form-wound coil insulation. (The less uniform nature of random-wound insulation renders both IR and PI tests of little value; a PI of 1 is generally acceptable.)"
• Army Corps of Engineers UPS Guide - PI for >= 200HP
• NETA MTS-1: PI for motors > 200hp, DA for motors =<200hp
Note the last one - DA is substituted for PI for smaller motors which would likely be random wound.
=====================================
(2B)+(2B)' ?
