According to Jamie :
Chris Lewis wrote:

I don't believe that different H/N lengths make much difference to
GFCI or AFCI, except in extreme and unusual situations. Eg: if you
wrapped the neutral many times around an operating fluorescent
ballast I could see it tripping a GFCI. _Maybe_.

Yes, they have to be. [the same length]

That is why many people have problems with GFCI outlets false tripping.

CFCI's work on load and phase balancing with respect to the race ways.

Remember that by definition, a simple two pole device _cannot_ see
"phase differences". There's nothing to compare the phase _to_. All
it can see are instantaneous voltage between the two leads plus
current in the individual leads. You need three conductors (the
ground doesn't count here because a GFCI doesn't care about
the ground) to see "phase".

Yours was an industrial situation (likely very noisy EMI), and not
knowing exactly how these things were wired and what they fed, it's
difficult to tell what was going on. Eg: common ground points on what
was being fed with the equipment and a tiny bit of leakage on the other
leg. Completely isolating both current-carrying leads of the GFCI
(as you ultimately did) from the tool would eliminate much of that,
without having anything to do with individual conductor length.

It could have just been severe EMI on the line side of the GFCI
overwhelming the device or an imbalanced inductive effect.

I really don't think you'd see anything like this in a residential
situation.
--
Chris Lewis,

Age and Treachery will Triumph over Youth and Skill
It's not just anyone who gets a Starship Cruiser class named after them.