Megger versus hi-pot testing

[rasiga]

Hello everybody

I'm looking for some informations about the subject; I would appreciate any help (links, disscussions, etc.).

My particular case is diesel-electric locomotives; currently we are testing the power circuits as follow: 1) perform ohmmeter check first; 2) megger with 1000Vdc for 15 seconds (min.5M); 3) hi-pot with 2400Vac.

The effect on equipment insulation by means of high potential testing has always been a controversial subject; that's why I'm looking at the possibility to replace the hi-pot testing with a megger test at a higher voltage. Any comments/suggestions are welcomed.

regards, chris.

 

[busbar]

Search this site for a discussion or two.

Is your AC test go/no go? With DC, you can effectively "see it coming" by watching leakage current.

It’s a very old debate. If it does not directly affect the amount of money in your pocket, I wouldn't loose any sleep either way.

 

[electricpete]

Busbar is right that opinions vary.

By the way what voltage is your machine? Also what vintage? What type of machine (syncronous generator?) and if possible what type of insulation?

Relevant standards include IEEE43-2000 and IEEE-43.

Good writeup at

http://www.usbr.gov/power/data/fist/fist3~1/3~1_cont.htm

(also return to FIST listing link at bottom of page to access adobe document).

It is widely accepted in utility circles that dc testing is equivalent to ac testing if you increase the test voltage by a factor of 1.7 That factor is mentioned in quite a few standards.

I have direct experience that contradicts this 1.7 factor. A 13.8KV service-aged motor passed ac hi-pot at 18.5kv. Then it failed dc hi-pot at 24kvdc... far less than 1.7 above the ac voltage.

 

[rhatcher]

I am assuming that by 'power circuit' you are referring to the locomotive's generator and drive motor circuits. In my experience these are DC machines with an armature voltage of 500VDC. If this is the case for you I find the 2400VAC hipot test surprising.

I would recommend a meggar test at 1000VDC for this circuit. If you 'must' test at a higher voltage a meggar (or hipot) test at 2500VDC should be more than adequate to prove the insulation of the circuit.

 

[rasiga]

Busbar
the test is a go/no go. I could not see which site you refer to?.

Electricpete
As you can see from my initial posting we use a 2400Vac. (the tripping current is set-up at 1A).
Thanks for the site.

My point is that, while everybody agree that the megger testing is less destructive than hi-pot, why don't we try to do a megger test instead of a hi-pot (where is possible). We tried an experiment: we megger at a higher value: 5000Vdc instead of 1000Vdc and the results where almost 100% identical by using a 2400Vac hi-pot machine.

 

[electricpete]

megger is certainly less destructive than hi-pot.

But it is also less sensitive of a tests. Some faults will only show up as you increase the voltage towards hi-pot levels.

I see you have identified your test voltage levels but I still don't see what the operating voltage of the machine is. That provides the basis for selecting test voltage levels.

 

[rasiga]

Rhatcher
The power circuits are: main generator, companion alternator, traction motors, high voltage cables, power contactors and eventually dynamic brake circuits. the traction motors could either be dc or ac machines; the 2 generators are always ac machines.

Rhatcher&electricpete
the maximum voltage on any element of the power circuits is 600V.

 

[ccaserta]

Hi everyone, I'm jumping in a little late. I think you have to remember that insulation testing and checking the dielectric strenght are two different test. For most safety evaluations both are necessary. As I'm sure everyone knows the Megger is meant to verify the insulation between ground and primary circuits, thats all. The Hypot tester checks the dielectric strenght of an insulator between circuits of different potential. Thats two very different processes, therefore both are required. The leakage meter on a Hypot machine allows you to set a limit during the voltage curve on time. It is not meant to read leakage currents although it can give you an idea of where your at.
Most Hypot test are first tried with AC voltage, DC is acceptable if the AC does not pass. Christopher Caserta
ccaserta@enorthhampton.com
Ph:904-225-0360

 

[rhatcher]

rasiga, sorry for the delay in getting back to this thread. Before I go on I will say that although I have worked on a few locomotives that I am not an expert on them...

If your locomotives all have AC generators then they are of newer design (ie < 25 years old). This still raises some questions though. I would think that you would have a different test standard for each type of equipment. For example, one standard for the synchronous generator, one standard for the DC traction motors, and one standard for the induction traction motors. Are you using the same test for all of these motor types?

As well, the power circuit for an AC-DC (generator type-traction motor type) locomotive would also have to include a rectifier. An AC-AC power circuit would include a rectifier and an inverter (or inverters). Normally you would not want to subject this type of equipment to a hipot. I am assuming that you are disconnecting these from the circuit before the test. (?)

 

Los ensayos con Dc son más informativos que los de Ac.

Se pueden sacar buenos datos del envejecimiento y estado de los aislamientos, utilizando, por ejemplo, un Meguer de 2500Vdc, con la ayuda de un microamperímetro durante el proceso de &quot;carga&quot; que puede durar varios minutos. Acabada la inyección se pueden leer las corrientes de &quot;descarga&quot;. Con la forma de las curvas y sus integrales, se sabe mucho de la máquina. Debe repetirse periodicamente.

El ensayo con 2500Vac es más arriesgado, sirve para determinar si rompe el aislamiento, pero no tanto su estado. Si se lee la capacidad del devanado y se compara con la corriente de fuga medida con un microamperimetro se obtienen datos de interés.

 

[jbartos]

Suggestion: Visit

http://www.phenixtech.com/products.html

http://www.vector-instruments.co.uk/hivolt.htm

http://mitchellinstruments.com/01075.pdf

etc. for more info and prevailing test approaches.

 

[mchenney]

try

http://www.asresearch.com/

Dennis McHenney
mchenney@mindspring.com

&lt;dmchenney@merkle-korff.com&gt;

&lt;mchenney@mindspring.com&gt;