surge capacitor

""Meggers" test low and medium voltage with a charge between 600 to 2,000 volts over the span of a minute. "Hi-pot" testers apply a much higher voltage in the range of 15,000 volts and above, to a maximum of 300 volts per mil of insulation. "Hi-pot" tests are performed over 15 minutes with readings taken every minute."

https://www.facebook.com/cablegi/posts/whats-the-difference-between-a-megger-hi-pot-test-high-voltage-testmegger-and-hi/1726906697335754/

Megger makes many different pieces of test equipment including megohmmeters. They may even make one designed to test capacitors.

A megohmmeter or Hi-Pot produces dangerous voltages. Charging a capacitor with either of these instruments should be done only with extreme caution. I don't recommend testing a capacitor with either of these devices. They will tell you very little useful information about the capacitor if the capacitor survives the test. If the capacitor is breaking down before the rated voltage these instruments can test for that breakdown voltage.

Using a dc megger, you could perform a dielectric absorption test, which, over time, measures the capacitance of insulation in rotating machinery. This should be able to confirm the capacitance of the unit, while also giving you some idea of leakage at voltage.

A simple capacitance meter is probably a better test, as you can do with a Fluke 324, and then hipot the capacitor at a suitable voltage, dc will be easiest.

A long time ago I used to test surge diverters with 3 separate tests.

First on 11KV diverters they were tested with a HV DC resistance tester to check they had not absorbed water and they had to have a resistance of at least 1 Gohm.

Next they were connected to HV ac at twice the nominal line voltage and had to withstand that for 1 minute.

Next they had to pass an impulse test where a back of capacitors were charged up in parallel at 10KVDC. They were then switched to series and generated 80KVDC through an RC network to Sharpen the wave front to emulate a lightning strike. The DUT had to pass 10 pulses in a row to be accepted as suitable for use.

We used to test them in batches of 1000 and always tested full stock batches. Every returned diverter was tested to the same standard to ensure they were fit for purpose. Batches could take a month of tests.

Which parameter(s) of the capacitor do you want to test?

The <...we...> needs to know what it wants to test before <...how...> is considered; <...how...> is secondary to what/where/who/when/why. Particularly why.

That information has been withheld from the forum.

Please describe the boundary between pass and fail. This needs to be established before any <...test...> is carried out.

Otherwise it isn't a <...test...>, merely a measurement.