Clio MKII intermittent ignition failure

[ClioSportAccount]

SYMPTOMS
Ignition turns once and then dead. There's sometimes a kind of buzzing but it was already doing that.

VIDEO:
Sound is there but quiet:
https://www.dropbox.com/s/y6p16h9x9k1t3wq/VID_20180115_170637.mp4?dl=0

RELEVANT INFORMATION(?):
Top left of the windscreen interior is damp. No faults reported by OBD.

TRIED SO FAR:
This is only the second time the problem has happened, last time fixed itself overnight so I haven't had opportunity to do much other than test the battery which came in at 12.4 and is less than 2 years old. I also cleaned the key and put electrical contact cleaner down the ignition.

 

[ClioSportAccount]

Oh, important detail perhaps! The first time it happened, I was sitting in the car with the blowers on max to demist the windscreen when the power suddenly dropped - the blowers audibly went quiet - but didn't cut out. I turned the key to see what was going on and the car wouldn't start.

 

[ClioSportAccount]

No wonder it wouldn't start, down to 10v after a whole bunch of fault-finding cranking.

The car starts absolutely fine with a running car attached to the battery, so I guess either the battery is borked or an earth is dodgy and the jump cables are a working substitute? I don't think the battery is failing to charge as there was at least two weeks between failures and plenty of driving.

I'll be damned if I can find the earths from above to check them and it's currently stranded in a place I can't jack it.

 

[ClioSportAccount]

What's with this forum and editing limits? Now I gotta spam the place up.

Battery jumped high enough for the car to start as normal first time, now reads around 12.6v and then 14.3v while revving at 2k.

So I guess I'm looking at corroding or fraying, somewhere. I've already cleaned the popularly faulty starter motor wire, maybe that was the problem and the battery was already too dead to prove it but the wire looked fine.

I feel like I'm going to be waiting to see if it happens again.

 

[ClioSportAccount]

Oh, Lord save me.

Battery was stone dead 24 hours later undriven. Charged it up again, looked for parasitic drain, measured 0.005a. Plugged my ELM327 OBDII back in and drain shot up to *0.075a*. About the same as a interior lightbulb?

I don't know if it's constantly scanning for receivers or what, but a quick Google says that's absurdly high. Torque Pro was complaining it might be a fake. So in the absence of another explanation, that's that.