Weird issue with insulation testing a cable with multiple vfd driven motors.

[SandyBFB]

Just thought id throw this out here to see if anyone has a less crackpot theory about what is going on here than me

So a bit of history and a description of the system; I'm providing electrical tech support and maintenance to a company that operates ROVs in the Pacific ocean. The system uses 5 vfds on the boat to drive the ROV under the water. All 5 thruster circuits run in a 20C tether between the boat and the rov and split into 5 separate cables at either end of the tether. The OEM machine came with a fancy all in one cable with each motor having a braided shield around the individual motor conductors. Due to the challenges of covid I've had to source out an American made alternative and I've done away with the braid because of vibration caused by a high-pressure water hose running parallel to our tether cause the cables to wear out from the friction of the braids in about 4 months. They have been working great for over a year now requiring rebuilds more or less just from operator "wear n tear". 400v system, 2.5hp motors, #16 wire, 100m tether, Flygt SR 4620 motors. Cable connects to thruster with a subsea connector resembling a pin/sleeve connector.

So anyways, I've had 2 cables recently fail in a weird way. When doing loop checks and insulation tests the cable passes perfectly. Plug it into a motor and fails at about 300kohms. It'll fail when plugged into any other Flyght 4620 motor. Cut the subsea end off and twist the power conductors together to simulate motor windings...pass on the insulation test, install new subsea connector and fail again.

So I replace the cable and everything is happy again. I leave the cable sitting at the shop for a couple months and it seems ok now but i haven't tried putting it back into service yet. Ill be running it for a few hours with a spare unit in a tank at the shop in a week or two. A week ago I go out to another boat, same issue with a different cable/core numbers inside the cable, troubleshoot it the systematically in the same way. I cant repair it at the location the boat is at so we make a plan for a week later. Operators run the machine and over the course of the weekend the cable takes out the thruster motor, motor is extremely hot when it comes out of the ocean. The motor ran for 2 straight days and i can only speculate that the short circuit current wasn't too high because the vfd is limited in the settings at 5.7A. I replaced the cable and thruster a couple days ago and the cable is on my bench to disassemble and troubleshoot. It appears to be a problem "with" the tether that "affects" the motor and not the tether.

Curious if anyone has seen something like this before and can shine some light on what could be going on or how to test for it? Best theory i can come up with is one of the cores is magnetized somehow which moves a some internal component in the motor? Or that some other quality of the wire insulation other than resistance has been degraded? I'm out of my depth a bit with this one.

 

[petersonra]

is it possible there is a filter of some sort at the motor terminals?

 

[SandyBFB]

Possible i suppose, I didn't physically see anything on the couple occasions I've had them open but i could investigate.
Here is the websites brochure for the family of motors.

4620 Compact Mixer | Xylem Canada

https://www.xylem.com/siteassets/brand/flygt/flygt-resources/flygt-resources/1103-flygt-compactmixers-mixing.pdf

https://www.xylem.com/siteassets/products/specifications/4610-4620.pdf

I did notice that i have 12ohms across the motor windings of these thrusters and I'm used to seeing 2ish ohms in my past experience up north.

 

[GoldDigger]

How are you testing the insulation resistance? If using a megger, what voltage are you applying?
If you can do the test with the motors in-circuit you must only be measuring impedance to "ground"?

The sequence of measurements you describe seems completely consistent with the 600k to ground being in the motor, either by design or as a result of previous damage or too high a test voltage.

In addition, not necessarily related to the current problem, the sharp pulse edges on the VFD output may be causing wave interference patterns on the wires that result in an applied voltage on the wire of double the pulse voltage. This effect might have been mitigated by the higher wire to ground capacitance with the shielded cable.

 

[SandyBFB]

How are you testing the insulation resistance? If using a megger, what voltage are you applying?
If you can do the test with the motors in-circuit you must only be measuring impedance to "ground"?

The sequence of measurements you describe seems completely consistent with the 600k to ground being in the motor, either by design or as a result of previous damage or too high a test voltage.

In addition, not necessarily related to the current problem, the sharp pulse edges on the VFD output may be causing wave interference patterns on the wires that result in an applied voltage on the wire of double the pulse voltage. This effect might have been mitigated by the higher wire to ground capacitance with the shielded cable.

Fluke megger, typically 500v setting. And yes its just a quick pin/sleeve style subsea connection so i do both meggering all the individual conductors to eachother and a final (or sometimes cursory) insulation test to ground while plugged into the motor. More to test if the grease in the connection is saturated with saltwater or other impurities. For the most part all my thrusters exceed the range of my megger to ground unless water has weasled its way through the seals. This is a relatively new issue I havent seen yet. (Happened once on 2 out of our 5 vessels)

Yeah its a wierd set up. The multiple braids in one cable just wont work with the pulsation from the pump, itd be a bit over 100k a year per machine including downtime replacing cables wearing out due to their lifespan nevermimd opperators pinching the cables. Not to sidetrack my thread but Ive been considering doing a price (and weight) comparison on running 5 separate cables with braided shields vs a single tether. But i dont know if the accountants will wanna shell out that kind of money on something new when these are working.

 

[synchro]

... 400v system, 2.5hp motors, #16 wire, 100m tether, Flygt SR 4620 motors. Cable connects to thruster with a subsea connector resembling a pin/sleeve connector.

So anyways, I've had 2 cables recently fail in a weird way. When doing loop checks and insulation tests the cable passes perfectly. Plug it into a motor and fails at about 300kohms. It'll fail when plugged into any other Flyght 4620 motor. Cut the subsea end off and twist the power conductors together to simulate motor windings...pass on the insulation test, install new subsea connector and fail again.

So I replace the cable and everything is happy again.

For want of a better explanation, perhaps the following is happening:
The wires in the cable have capacitance between themselves, and the motor windings have inductance (as well as some stray capacitance). When the megger applies a step of 500V DC between one of the power conductors and the ground wire it might be causing the voltage on the other power conductors to "ring" because of the LC circuit that is formed. Perhaps that could cause a momentary overvoltage that initiates a sustained discharge through a defect in the insulation somewhere and produce the reading you see on the megger. It's just a guess, but I don't see a simpler explanation if replacing the cable but not the motor fixes the problem.

I suggest connecting all of the power wires together on the input side of the cable, and then use the megger to test between that common connection and the ground wire. That would constrain the voltages on the power wires and perhaps limit the amount of ringing and overvoltage that might be happening.

 

[SandyBFB]

For want of a better explanation, perhaps the following is happening:
The wires in the cable have capacitance between themselves, and the motor windings have inductance (as well as some stray capacitance). When the megger applies a step of 500V DC between one of the power conductors and the ground wire it might be causing the voltage on the other power conductors to "ring" because of the LC circuit that is formed. Perhaps that could cause a momentary overvoltage that initiates a sustained discharge through a defect in the insulation somewhere and produce the reading you see on the megger. It's just a guess, but I don't see a simpler explanation if replacing the cable but not the motor fixes the problem.

I suggest connecting all of the power wires together on the input side of the cable, and then use the megger to test between that common connection and the ground wire. That would constrain the voltages on the power wires and perhaps limit the amount of ringing and overvoltage that might be happening.

I do a pretty thorough isulation test at the shop, every conductor to every other conductor in the 20 pack (Very tedious) and that hasnt shown me anything out of ordinary in the tether (without motors hooked up) on either of the two cables.

I will definately do a bit of bedtime reading on RLC circuits because i have a feeling the answer might be related, its just a bit over my head. I havent had much to do with this kind of setup in my career before now. Mostly oil/gas background where we'd run a separate tek cable (or even vfd cable) for each motor. Its also challenging to troubleshoot with this tether because its not fixed and secured in a building, its a flexible cord so itll sometime fault when pulled on or bent a certain way. Although with this particular issue it doesnt seem to matter how the cable is positioned, which leads me to think the issue isnt a damaged section but possibly an odd degradation that doesnt ahow up on the megger of the insulation on particular conductors.

 

[synchro]

So anyways, I've had 2 cables recently fail in a weird way. When doing loop checks and insulation tests the cable passes perfectly. Plug it into a motor and fails at about 300kohms. It'll fail when plugged into any other Flyght 4620 motor. Cut the subsea end off and twist the power conductors together to simulate motor windings...pass on the insulation test, install new subsea connector and fail again.

So I replace the cable and everything is happy again.

A week ago I go out to another boat, same issue with a different cable/core numbers inside the cable, troubleshoot it the systematically in the same way. I cant repair it at the location the boat is at so we make a plan for a week later. Operators run the machine and over the course of the weekend the cable takes out the thruster motor, motor is extremely hot when it comes out of the ocean. The motor ran for 2 straight days and i can only speculate that the short circuit current wasn't too high because the vfd is limited in the settings at 5.7A. I replaced the cable and thruster a couple days ago and the cable is on my bench to disassemble and troubleshoot. It appears to be a problem "with" the tether that "affects" the motor and not the tether.

Fluke megger, typically 500v setting. And yes its just a quick pin/sleeve style subsea connection so i do both meggering all the individual conductors to eachother and a final (or sometimes cursory) insulation test to ground while plugged into the motor. More to test if the grease in the connection is saturated with saltwater or other impurities. For the most part all my thrusters exceed the range of my megger to ground unless water has weasled its way through the seals.

For the problem cable that's on your bench for troubleshooting, if it passes the megger test at 500V then I suggest turning it up to 1000V to see if any particular wire-to-wire combinations start to fail.
I'd only do this after you've completed your other measurements on the cable, and of course leave the motor disconnected so that you won't be over-stressing its windings. And I'm not suggesting you put 1000V on a new cable.