I'll do a 10/90 on that batch and report back after that. The skin doesn't seem to be getting any thicker although it has stood for about three week.
It won't be methanol.
Paul, you say bio and methanol are miscible in all proportions but when I've done expts on 10/90 ratio's I have found there is separation when the bio gets up to 30 to 40%. It's very hard to see if the bio colour is very light but its there.
I did say usually - because there is always the possibility of something strange happening.
I did have one batch where I got three layers, glycerol at the bottom, bio/veg oil in the middle and a third layer above that I am still unsure what it was. I had processed in the usual way but after stage 1 dropped everything into barrels and had left it far a couple of weeks. It was a clear liquid that decomposed when heated rather than boiling away. It left a carbon residue when heated strongly as it decomposed. For various reasons, I left it in the barrels with some additional catalyst and when I checked it between 1 and 3 months later the third layer had gone and the bio was a pass. It might have been soap, but as I used sodium why was it still liquid. The layer had been quite thick and obvious (1-2"), and immiscible with oil which would suggest not soap. Never worked out what it was.
I have on several occasions added more bio to a 27/3 test going as far as 27/15 but never seen a separate methanol layer - but probabably never looked for it that closely either.
In the same way that water in methanol causes the Warnquist test to give a high precipitation, could something else in the bio (or soluble in methanol and not bio) cause a separate layer of "methanol" to form.
David described it as a skin, so I would tend towards Nigel's view of it being soap.
