Remagnetising rotating permanent magnets of a PMG 5

[edison123]

A 3000 RPM generator was out of service for two years. Since the utility owner did not use keeper plates, the rotating permanent magnets of the PMG (whose output supplies the AVR) lost some of its magnetism. During recommissioning the generator, the PMG output voltage was only 1/3rd of its rated voltage and hence the AVR is not working.

Any ideas/suggestions on how to restore the magnetism of these rotating permanent magnets would be greatly appreciated.

A photo of the PMG rotor.

Muthu

http://www.edison.co.in

 

[ScottyUK]

Generally magnetising a magnet requires a larger and more powerful magnet. The easiest option would be a big neodymium (sp?) magnet stroked along the axis of the weakened magnet. If that fails then you're looking at a coil with a big DC source behind it - large battery, heavy cable, close the switch and let a fuse clear it. ;-)

 

[waross]

I would consult the manufacturer.
Failing that:
Try making a block of mild steel that will fit snugly between a pair of poles.
Wind as many Amp turns as you can on it.
Remember, at a given voltage, twice the number of turns equals twice the resistance and half the current.
Your Amp turns will be determined by the wire gauge and the applied voltage. The more turns that you can squeeze into the available space, the lower the actual current for a given number of Amp turns.
You may want to use a fuse as a control. Let the current build until the fuse blows.
When you are working on a pair of poles, the other poles will not have a complete magnetic circuit.
Hopefully the large air gaps associated with the other poles will render the other poles some what immune to the magnetizing of one pair at a time.
Disclaimer; I have not done this myself. Just brainstorming.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

I agree with Bill only use the iron bar and wrap it a few times with welding cable then strike an arc with a DC welder and practice your welding for a minute. Do something in the 150~200A range. Step and repeat.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

I agree with Keith, but I would not add the impedance of the arc. Do like a beginner and stick the rod to the workpiece.
You don't need a lot of time. The rise time will be quite short.
Did I mention that while the Amp turns doesn't much change with more turns, doubling the number of turns will increase the inductance and the rise time by a factor of four.
I remember a re-magnetizer that used AC. The coil was almost a short circuit. A fuse was used sized to blow on the fault current during the first 1/2 cycle.
With AC the polarity was random. That was not an issue with the bar magnets being re-magnetized.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[catserveng]

I have never done this on a generator PMG, but I have been involved with the process on a Woodward PMG for a hydro governor. Look at pg 14 here,

https://www.usbr.gov/power/data/fist/fist2_3/vol2-3.pdf

Hope that helps, MikeL.

 

[ScottyUK]

Whoa, that looks brutal! I thought our method using the turbine emergency battery, a big switch, 630A fuselink and a handful of turns of heavy cable was bad enough. Throwing the switch wasn't for the faint-hearted: the fault current was in the order of 6kA - 8kA.

 

[catserveng]

Yeah, first time I did it was way before any 70E training, but the guy I was working with had done it a number of times. I guess was a common problem in smaller seasonal hydro's, at least according to him.

I haven't done it for several years, and probably today in a lot of places it would be tough to do, but at the time it was a well documented procedure and results were consistent all the times I was involved.

I think EASA has a procedure they publish for their member companies,

https://www.easa.com/resource_library

I know a couple of generator repair shops I have worked with in the past had procedures for remagentizing PMG rotors, but it usually got done when we sent int he entire generator or PMG/exciter asembly.

MikeL.

 

[itsmoked]

Good Point Bill 'stick the electrode'.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[edison123]

Thank you all for your tips. Both the static DC excitation of the magnets and the USBR method of momentary flashing of exciter stator seem to be trial and error methods. Each time, the generator has to be run up to speed to check the PMG output. Also for the USBR method, the magnet unit need to be removed from the generator rotor shaft, which is out of question due to the shrink fit.

We tried supplying the AVR with the station's 110 V, 50 Hz (instead of the rated 110 V, 350 Hz) but the generator voltage shot up way beyond the rated. Any reason, why the AVR would not work at the rated voltage but at a lower frequency? At lower frequency, the output DC ripple content will be higher but why would that increase the generator stator volts?

Muthu

http://www.edison.co.in

 

[waross]

It could be that something in the AVR is saturating at the lower frequency.
I doubt that the effect will be linear.
How about applying 50/350 x 110 = 15.7 Volts?
That should be safe and from the generator output Volts you should be able to calculate the required voltage at 50 Hz.
I have replaced a PMG AVR with a self excited AVR while waiting for a replacement PMG AVR.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[edison123]

Bill

I know nothing about the innards of an AVR. Would an AVR be affected by V/Hz issue? Isn't it basically a rectifier with voltage sensing and control?

Muthu

http://www.edison.co.in

 

[pwrengrds]

This is a procedure we used to magnetize a PMG. I would recommend asking the manufacture for theirs.

To magnetise;
We use 7 off car/van batteries.
Charged to give between 90-100V.
Place 200A fuses in circuit, one in positive and one in negative of supply. These should protect PMG winding if 3 pole 1000A contactor fails to disconnect D.C. The positive is looped via two of the poles wired in series and the negative wired through the third. These ends are then wired to the PMG. Positive to terminal 1. This is not really important but I do it.

The contactor coil is wired to a control circuit and timer set to apply Battery voltage via fuses and contactor poles to the PMG for just over 1second. Also include Emergency stop switch.
This pulse might magnetise the PMG but could also simply remove the residual magnetism.
A second pulse is now required to ensure magnetization has taken. However it must not be applied immediately you must wait 30 minutes to ensure PMG winding is not over heated.

Due to the battery impedance and low impedance of the PMG winding we expect between 400A and 800A to be applied.

Once magnetized the PMG requires stabilization.
To do this we disconnect the battery and connect the cable ends together to short the PMG using the contactor and timed to stay shorted for 60 seconds.
With PMG shorted the maximum current will be 60A most often 20 - 40A.

 

[waross]

Is the PMG stationary or turning at rated speed during this procedure?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[pwrengrds]

Stationary for the DC portion, ran at rated speed and short out the PMG output for the stabilization.

There is a Brush procedure (attached) for lining up a rotor pole with a stator pole and putting the DC through the stator coils, using the same DC setup as above. Have to limit the time to less than 2 seconds to avoid overheating the stator.

Then run it up to speed and short out the terminals for 1 minute for stabilization.

David

 

[edison123]

Thanks David. Will give it a try when the machine is idle.

Muthu

http://www.edison.co.in

 

[pwrengrds]

Just to be clear, the stator mentioned above is the PMG stator (also refereed to as the PMG output).