The problem is there are several variations on a theme. There's no right way of doing things, it really boils down to personal preference and what works for you with your feedstock.
I can tell you the way I prefer to do it, but someone else will inevitably say their way is better. You'll also find you start taking short cuts after you've been at it a while ... then you'll have a bad batch and go back to basics to get it right. Next thing you know you're slipping back into cutting corners again!
My way (without too many short cuts or bad habits) is ...
Filter WVO through landscaping fabric filter socks (
http://www.biopowered.co.uk/wiki/Tips_and_wrinkles_1#Homemade_sock_filters) and settle for as long as possible. I believe food particles can retain moisture, so I aim for the oil to be as clean as possible.
Perform HPT and dry oil using condenser if required.
Titrate oil and mix methoxide at something less than the recommended ratio usually about 4.0 - 4.5g per litre (NaOH).
Bring oil to 67°C ish and add 80% of Methoxide. I turn the heater off while adding the Methoxide but run the rest of the reaction with the heater on and the PID set to 63-65°C.
React for about an hour, turn heater off and settle Glycerol for about an hour.
Drain Glycerol, pumps and heater on and add the balance of the Methoxide. Do a 3/27 after about half an hour. If it's OK (I never look for a "perfect" pass, clearish with no dropout is acceptable to me) I carry on to demeth, if not continue processing until a satisfactory pass is obtained. You can take a chance at this point as the initial stage of demeth is no different to processing, so you could accept a small amount of drop out and hope that the first part of the demeth finishes the reaction.
Once 3/27 is ok, add back the first stage Glycerol and start demething. I don't demeth at elevated temperatures and no longer bother controlling the still head temperature, I just set the PID to around 80°C and let the system run until the Methanol has slowed to two or three drops per minute.
Heater off, add 7% water and run for about 15 minutes.
Stop the pumps and allow the Glycerol to settle for an hour or two, or until the temperature has dropped to around 60°C.
Drain the glycerol (add any winter additives and mix for 5 minutes or so) and pump out to a settling barrel.
Bubble dry over night and allow to settle for as long as possible, but at least a week.
If the hoses serving it didn't sweat as much as they do, I'd pump through my dry wash tower and into a second settling barrel and again settle for another week or so.
Then it's pumped out through a 5µ filter and into the car.
If you check out the VOD forum, Hyfly currently has a thread running advocating no titration, instead relying on 3/27 tests to progressively add the second stage methoxide in varying concentrations, then do a couple of full water washes in the processor. Several guys have tried this with some success, my efforts sadly resulted in an emulsion that took a week to break, but have, I must admit, resulted in some of the clearest bio that's left my processor. I'll give the idea another couple of tries as it certainly seems to have it's merits, but if I need a batch in a hurry I'll go back to what works for me.
Hope that helps, but as was said previously ... ask the question of someone else and you'll get a different answer!
