when doing my first tests to get baseline readings on 10 new critical motors, i noticed the megger readings increased incrimentally from l1 to l2 to l3. doing a 30 second and 60 second check on each phase the reading on l2 at 30 seconds would be 2-3 G ohms higher than l1 was and the same for l3 compared to l2 at the same time (30 seconds). 60 second values increased in the same way.i was wondering what caused this.was it that the winding was partially polarised from the l1 test which allowed the l2 test to charge up faster? or am i completely off the track? any ideas?

quote:

this.was it that the winding was partially polarised from the l1 test which allowed the l2 test to charge up faster

Yes, I'll bet that is the explanation. The resistance reading is expected to increase gradually as voltage is applied. It may take up to 10 minutes to reach it's final value.

I suspect the windings you tested are tied together at some point. For example internal wye connection. So each time you apply the voltage, you are testing the whole winding (regardless of which lead you pick) and the resistance goes up. In this type of winding, there is not much point to performing the test on each termination (unless you can internally separate the winding).

Can you clarify the motor connections? How many leads are accessible (3, 6, 9?)

I would agree with electricpete that this is why you are seeing this increase. If you want to prove this to yourself do the same test three times on l1 and I would suspect you will see the same type of increase with each test.

My First question is where you took the Megger test from the motor or the mcc bucket. Depending on where you did the test you could be getting readings that include the T-Leads which could have got nicked in the process of pulling in the wires or there is a bad connection in the motor J Box. So the data that you are seeing could have more investigation that needs to be done. You also need to wait 4 times the amount that you put the voltage through the motor before you can retest a different phase. For example if you put 1000V through Phase 1 for 1 minute than you would need to wait 4 minutes to let the system drain before you did phase 2 and repeat for phase 3.

Regards

I agree with Epete,

It's a Y connected motor.
Do as JOA1 suggests and test phase 1 twice. See if you don't get the same rise.

D

I agree with RRS_Dave and Epete. Meggering all three motor leads is a waste of time.

I disagree with jkays about the part of having a possible nicked wire needing more investigation. A RISE in megger values will not indicate a nicked wire or any kind of leakage. And even if you had a bad reading, testing the other two phases won't tell you any more than you already know.

The only investigation that needs to be done is personal investigation for learning purposes if so desired.

Just my opinion.

I am not disagreeing with the fact that the motor could be all one circuit or the leads could be hooked up wrong 3,6,9,12 lead motor, but that if the test was done from the MCC bucket that one of the phases could have a nick in the wire and the other two phases could Carrie a higher insulation value to hide the nick in the one wire. This making the one phase that the tester is on show low and when the test is on the other two phases it hides the anomaly. This will hold true to a bad connection in the motor J box. If the test was done at the Motor J box then it would be very difficult to determine what the problem could be. We need more information on how the test was done and we also have to take in to consideration that moisture could be a factor. Just my opinion

Regards

I agree with all,
but I noticed Pennichuck’s reading with motor insulation resistance in 30 and 60 seconds.I heard about values 60 / 15 second ( koeficient of absorption- name in my country)and ratio should be between 1,2 - 2 for good motor .Does this 30/60 seconds reading mean same things?
Good hunting!!

This message has been edited. Last edited by: Panta,

thanks for all the responses. I thought all that was required was to test only one leg.since these motors are wye connected.so if you test one leg to ground you are testing all 3 legs at the same time.i suppose if you got a low reading you could unwire the motor leads and find which winding the fault is in, but on a motor like this we need to replace it to get it right back in service anyway.
i took my readings at the motor disconnect. i would have to disconnect the leads at the bucket as these are run off of vfd's.dont think the vfd's would like 1000 vdc.these are 9 lead 480 v motors. they are all wired correctly and functioning fine.I'm just getting baseline reading for the future. So my baselines will be more accurate just testing one phase? I thought this was the case but someone on the fluke forums told me I HAD TO TEST ALL 3 PHASES.. no explanation as to why but thats what he said.maybe he thought the motor was unwired.

Im taking my readings at 30 and 60 seconds just to be sure the 60 second is higher than the 30 second reading,from what i understand if the 60 second reading isnt higher than the 30 second the windings might be contaminated with dirt/ oil.
again i might be off on some of this, i just got hired here to be the head maintenance electrician and really have no experiance with maintenance. I was the electrical contractors forman running a large upgrade job here and they wanted me to come to work for them. right place right time kinda thing. Thankfully ive found some good info on the ol' internet. thanks for any input

Most meggers allow you to measure the residual voltage on the winding & it is normal procedure to check that this is zero (or minimal). If this has been checked, the variation cannot be due to residual charge.

Have you ensured that the guard is connected? If not, leakages could give you variations, especially if you are testing close to energized equipment.

Also, keep an eye on whether the IR value increases steadily or shows occasional fluctuations while charging. This would indicate weak layers in the insulation.

I believe that the minimum DAR value should be similar to the minimum PI value, viz 2.0 for Class B & F insulation (not just greater than 1.0).

Regards,

Aditya

Pennichuck

When you record these readings at 30 and 60 seconds you are basically performing what is known as a DA test. Take the 30 second reading and divide into the 60 second reading and this will give you a ratio. A ratio of approx. 1.5 is normally considered a good reading. If you get something less I would recommend doing a full PI wich is a 10 min. test and you devide the 1 min reading into the 10 min reading looking for a ratio of 2 or more (record the readings every 30 sec.). However if in either case if you reach 5000 meg in the first 60 sec. don't be to concerned about the ratio. Keep in mind that these test are looking at the potential to ground due to contamination or breakdown in the insulation system. It is unlikely that a loose connection will show up, the best way to check for this is to take phase to phase resistance readings. If there is a loose connection you will see an imbalance in these readings. Testing from the T leads on 480 volt random wound motors less than 3% imbalance is considered acceptable, however in my experience anything over 1.5% warrants taking a look at connections. Its allways beter to address it earlier than later to avoid a failure. When doing this be sure your meter is sensitive enough to show readings 3 or 4 places right of the decimal. Most VOM's are not sensitive enough to see this.

i have been taking phase to phase readings aswell.IM using a fluke 289 vom for this.

My main question is is it nessecary to meg test all three phases or is one leg all that is required? i would think when you test one phase you are testing all three.assuming you first tested phase to phase resistance and all looked good.

quote:

My main question is is it nessecary to meg test all three phases or is one leg all that is required? i would think when you test one phase you are testing all three

When you test one phase, (in the common configuration where phases are connected together), you do in fact test all three phases to ground.
That is all we do for our insulation resistance test.

There is a small amount of additional information which might be gleaned if you actually separate the phases and then test:
T1 to ground with T2/T3 grounded
T2 to ground with T1/T3 grounded
T3 to ground with T1/T2 grounded

What it gains you is:
A1 - Higher sensitivity, since you can compare three phases, which should be similar.
A3 - if you saw a low reading to ground on the initial test (all three phases tied together), you might be able to determine whether it is a localized condition or a global condition.
B - It also adds a small ability to check the phase-to-phase insulation resistance. I don't think this is important in the slot section (even though some slots will carrry coils from two phases) because we have twice the groundwall insulation between phases as what we have to ground. But we also have phase-to-phase paths around the end turn blocking and tieing and the area of the scarf joints on connection end which can be vulnerable points. And there is some additional discrimination that can be made if you do multiple tests including a guarded configuration.

I'd say most people don't bother to separate the phases and just do the single which captures all three phases to ground. It may not be worth the effort to separate the phases IMO unless you see a problem on the initial test (that's up to you to decide). (of course the amount of effort required to separate the phases depends on the motor configuration... internally connected wye is pretty well impossible to separate unless you are in a motor shop, and even then it could be a tremendous amount of work to separate, reconnect, reinsulate, retest).

quote:

i have been taking phase to phase readings aswell.IM using a fluke 289 vom for this....assuming you first tested phase to phase resistance and all looked good.

I suspect what you are testing phase-to-phase is the winding resistance as opposed to insulation resistance. Those are two different things. The winding resistance primarily checks for loose connections, turn to turn shorts. The insulation resistance checks the groundwall insulation and perhaps phase-to-phase (but not the turn insulation). So having a good winding resistance test really sheds no light on whether you might want to separate the phases for an insulation resistance test.

This message has been edited. Last edited by: electricpete,

thanks for clarifing pete.. yes the phase to phase i was talking about was winding resistance. i was responding to JOA1 and saying if i test phase to phase resistance and it looks good, i see no reason to test all 3 phases for insulation resistance individually ie: taking a megger reading at l1 then at l2, l3 because the 3 phase are internally connected. thanks for everyones help with this basic stuff. Ive got quite a bit to learn.

does anyone know of a good source of info on what what i should be doing for regular maintenance (how often and what to be checking)on say water pump stations? im setting up a maintenance program for our main water plant and about 80 remote pump/well systems.

so far ive come up with:

1. test amp draws on all equipment
2. check all connections with IR thermometer.
3.torque connections in switch gear.
4.winding resistance checks on motors.
5. megger test motors.
6. general visual inspection (probably first step)
anything else i should be looking for?


some of these stations are critical if one goes down people have no water.

thanks in advance for any help

I would agree with electricpete, seperating the phase for testing insulation to ground values isn't going to gain you much. Im not familiar enough with the fluke meter your using to know if it is sensitive enough to pick up connection issues or turn to turn shorts.
So the answer to your main question is No it is not necessary to meg test all three phases individually.

quote:

yes the phase to phase i was talking about was winding resistance. i was responding to JOA1 and saying if i test phase to phase resistance and it looks good, i see no reason to test all 3 phases for insulation resistance individually ie: taking a megger reading at l1 then at l2, l3 because the 3 phase are internally connected.

Now I see what you're saying. Sorry that I misunderstood before.

quote:

does anyone know of a good source of info on what what i should be doing for regular maintenance (how often and what to be checking)on say water pump stations?

Some things that come to mind:
Check oil levels.
Relubricate greased bearings (depending on schedule... requires some more consideration of bearing types and speed).
Check vibration
Check that space heaters on standby / non-running motors are energized

Pennichuck, Sounds like your doing everything you can (with the equipment available to you) along with a couple of the thngs electricpete suggested in his last post. A tool you might want to look at to help you with your maintenance is ultrasound. Ultrasound is extremely versatile, relatively inexpensive and fairly easy to learn. It can help you with lubrication, leak detection, cavitation, bearing fault detection and electrical. This is the technology I recommend to anyone just trying to start a predictive program.