Thanks everybody for your replies. More details on my set-up:
I extract the oil I get, direct from the fryers I clean as a business - usually at 175-185 deg C. It's filtered to 5 micron and is exclusively either rapeseed or GM Soya (no palm). I sometimes heat/settle it again, in my processor, to as hot as I can get it (70 deg+) for an hour or two and I don't get any water fall-out.
I titrate with either pheono or turmeric. The oil's been well used so it always titrates on the high side 4.0 to 6.5 using a potassium base are common. (I had some this week that titrated at nearly 10 - haven't dared try it yet.)
Batch processing sizes are 80 to 100 litres. Methanol is not virgin but is sold as 99% pure. The potassium hydroxide flake I buy is rated at 90%. I do a two stage process as described elsewhere on this site.
After the 2nd stage I sample 500ml of what's in the processor, allow the glyc to settle for 10-20mins, and do a 27/3 or 45/5 whilst the brew is still hot and roughly measure the amount of unreacted oil (often as much as 1ml of oil unreacted).
Erring on the high side I use the titration calculator on this site to calculate how much extra meth and potash to add for a third stage to try at getting a nil drop-out. I push it so far I sometimes end up with jelly but the reaction won't seem to go all the way.
I'm reading up on how to pre-treat the oil with an acid catalyst to esterificate it first before going on with a much less aggresive trans-esterification.
Three questions: Any advice on pre-treatment with sulphuric acid to esterificate oils with a high FFA? Could my process be making soap which in turn produces water which will stop the reaction going to 100% conversion? Has anybody else used the "extreme biodiesel" calculator - http://extremebiodiesel.com/titration-calculator.aspx?
Thank you
