Update on enhanced glycerol pre-wash

[dgs]

                                   
A friend of mine who is a member on here 'DavidShinn' emailed me a couple of weeks ago to say he had been experimenting with this method. He had achieved some excellent results by adding more methanol to the glycerol than normal to pre-treat the oil so I thought I would give it a try.

150 litres of lightly used rapeseed plus 50 litres of new rapeseed, so a 200 litre batch.

I added 30 litres of ex process glycerol to which was added 1.5gms KOH/litre (total 300gms) and 7 litres of methanol. Heated to 55degs (M67) and mixed for 1 hour.

Just done a 10/90 this morning which shows 5.5mls dropout (so approx 45% conversion)

For the next stage I will use 14 litres of methanol and 4gms KOH/litre. Will post with the outcome.

[Julian]

I've been revisiting glycerine pretreatment with the last few batches, having found it inconvenient in the past.

I'm using NaOH with very good, low titrating oil and it seems the lower soap levels leave the glycerin pumpable even at current temperatures.   I haven't done any tests after pretreatment, but I've been able to reduce methanol and catalyst quantities in subsequent stages to well below base amounts while still achieving good 3/27 passes. 

I've also taken a leaf form your book Dave, and started processing at ambient temperature (that batch I've just started was at 12°C).  A combination of Frankinpum, contorted pipework and a venturi sized for a Leo pump (so good at making emulsions), slowly raise the temperature of the batch and presumably promote good mixing.

Quite liking this production method so far!

[dgs]

That's good.

Yes, a drawback of pre-treating is the energy required to heat the mix. Then ideally leaving it to separate overnight where it gets cold again. This is why I use the m67 running on 80% converted bio made from free oil. I reckon this costs less than 10 pence to heat up 200 litres. Of course the Frankenpump will cost more running it for 90 mins+

[Julian]

The other thing I've tried is running Frankinpump against a partially closed valve in an attempt to aid mixing.

I've had it running at a pressure of 50 psi.  Difficult to evaluate though.

[dgs]

The Frankenpump on my processor is fitted with a pressure gauge on the delivery line to the eductor. With an unmodified 3/8 eductor the pressure was 71psi (very worrying, sounded like a little steam coffee maker)
Someone on VOD calculated the nozzle velocity at 28 metres/sec!

On Jim's recommendation I drilled the central jet out to 8.5mms so it's now calmed down to 32psi.