Glycerol Pre-treating versus the 10/90 dropout test.

[dgs]

A few years ago I was conversing with Countrpaul via PM about conversion after glycerol pre-treating. He told me he always achieved a conversion, whereas I never did.

The question was resolved when I realised that my syringe delivered 11 mis of sample into my dropout tube that overestimated the volume, reading a total of 13mls. So an easy fix of pouring the oil in until the level read exactly 10mls resolved the situation, after which I always had a small conversion.

More recently ( especially after reading some of 'Neutrals' old posts on infopop) another anomaly has arisen and I have (hopefully) solved it.

Ok, IMO  glycerol pre-treating using the previous batch glycerol and mixing it with the oil lowers titration but doesn't usually achieve any conversion even though a 10/90 test result may show a dropout of < 10mls. How can this be.

When I have 'split' glycerol with sulphuric acid the top FFA layer is usually about 40% of the total volume. I (like many others) always thought that this meant around 40% soap in the glycerol. This is not the case.

According to Neutral for every litre of soap formed during the reaction it takes with it 2 litres of biodiesel into the glycerol. (I recovered a jelly batch for someone 18 months ago and recovered 70 litres of bio from 100 litres of solidish jelly)  This means that the top layer in the Sulphuric acid treated glycerol was about 2.7 litres of FFA's (from the soap)  and 5,3 litres of biodiesel.

So, when we glycerol pre-treat, the biodiesel (or most of it) that is held by the soap in the glycerol migrates into the oil, leaving most of the soap in the glycerol. This of course explains why the oil has a reduced acid value, but is still acidic, so no conversion has taken place, yet still shows a reduced volume of dropout from the 10/90 test.

Just onto an associated  subject, the 10/90 test and temperature. I must say when Jan first came across the test I don't think he realised exactly how it works. Most think that the biodiesel is dissolved into the methanol, leaving the residual oil to drop to the tube bottom. The test doesn't work like that. Lets take a dropout of 2.4mls, that dropout is composed of 1.0 mis of oil and 1.4mls of biodiesel.

It just shows the amount of intermediates (Di's and mono's)left in the bio as adding chemicals for the whole 2.4 mis of dropout doesn't result in any over conversion.

 I am absolutely convinced that the 20degsC temperature for the test is too high. A better level is something like 10degs. A sample I recently tested for someone (no dropout at 20degs) had 0.2mls dropout when left in my bio room overnight. That dropout would be composed of about 50/50 oil and bio. The fact that there was any dropout there at all means the presence of mono's meaning a nightmare to water wash. Sure enough the poor guy had washed it 7 times without getting clear wash water.

Had the guy done his dropout at a lower temperature the dropout would have been revealed and given him chance to do another stage.

[Dickjotec]

Very interesting, thank you.

[countrypaul]

Dave,

I suspect most glycerol from the previous batch contains an amount of unreacted catalyst. One reason for glycerol washing reducing the acid value will be the neutralisation of these also leading to more soap. In the case of very low acid value oil, and a high residual amount of catalyst there could/should be some conversion if there is enough methanol present.

I agree that diluting the oil with bio (from the glycerol) will appear to reduce the acid value but not the total amount of acid present (more volume of oil+bio but with a lower acid value).

[dgs]

Agreed Paul. I think 'Neutral" many years ago developed the 5% prewash method, not to de-soap the bio but to separate the bio from the soap in the glycerol, thus increasing the yield.

I save my wet glycerol from my version of the 5% prewash method to add to my glycerol that I treat some oil with, the addition of the wet glycerol seems to encourage more bio to be released from the soap.

[Tony]

I've had little luck pre-treating batches with glycerol, even when I've added extra meth to it.  The batch seems to go better afterwards, but I've not seen evidence of a partial conversion from the 3/27 so always put that down to reducing fatty acids and taking out any water.

What the glycerol is astonishingly good at is treating the nasty wet whites and dregs left over.  If I put them all in a drum, mix in 20-25% by volume glycerol with a stick and leave for a month or two I always end up with a clean, pumpable liquid layer on top, and a layer of thick glycerol with chips in.

[dgs]

Tony, when you are doing a 10/90 to test any conversion, try just pouring the sample in the tube until you are exactly at 10mls. If your tube reads the same as most others I have tested then exactly 10mls reads around 12 on the tube, in which case you probably have conversion which doesn't show re the tubes incorrect calibrations.