Hello folks,

After searching through old threads, I decided it would be a good idea to dedicate a discussion to the composition of crude glycerin. What's in it? The following is an educated guess at what is in ours, pending a more official result from a 3rd party lab that I may be getting in the future:

33% glycerin
25% methanol
18% biodiesel
14% potassium salts
9% FFAs
1% water

I pieced together those estimates based on a lot of experimentation, but it is hardly what I would call authoritative. The figure for K salts was arrived at entirely by process of elimination, after coming up with the other figures. I used a Karl Fischer on the crude glycerin soup to determine water. I based the other numbers loosely on acidulation results I've had in the past. It's certainly possible my "guesstimates" are way off. I'm curious if anyone else has tried to figure these things out with their own glycerin, or if anyone has thoughts on what these numbers "should" be.

BTW, the feedstock was run-of-the-mill UFO from restaurants, around 3-5% FFA. The glycerin came from a 1000-gallon batch, from which we got around 300 gallons of crude by-product. I think we got about 900 gallons of biodiesel, for a 90% yield.

posting to watch easier .
UFO , does that mean unidentified fryer oil ?

...or, Used Fryer Oil

...unidentifiable fried objects

quote:

Originally posted by john galt:
...or, Used Fryer Oil

Correct.

quote:

Originally posted by kumar:

33% glycerin
25% methanol
18% biodiesel
14% potassium salts
9% FFAs
1% water

Of course, potassium salts (pre-acidulation) are soap. I can't tell you what it should be because I rarely do soap tests for PPM on crude glycerine (I acidulate based on a direct titration with dilute acid, but don't bother calculating PPM since that's not the information I"m looking for- so I basically do the soap titration but with a HCL or H2SO4 that's been diluted so that the result of the titration is 'how many liters of acid do I need to use' rather than 'how many PPM of soap is in this stuff).

quote:

I pieced together those estimates based on a lot of experimentation, but it is hardly what I would call authoritative.

The figure for K salts was arrived at entirely by process of elimination, after coming up with the other figures.

if you're acidifying the glycerine's soap and catalyst, and trying to measure salts produced
Do the conversion from soap (pre-acidulation) to potassium content if you want to know how much potassium salts will be left after acidulation (which isn't what you're looking for in this thread of course since you're talking about 'crude' glycerine). There's a way to calculate it based on a soap test and based on how much/what kind of acid it took to acidulate the crude mixture.

measuring other salts produced after acidulation of soap:
If you're looking at salts after acidulating the soap with acid, you'll get a varied volume of stuff based on whether the salt is soluble in fats and glycerine or whether it precipitates. Sodium or potassium sulfate, which I believe is what's generated whne you use sulfuric acid for this (caution- it'll boil and spit if yo'ure using concentrated sulfuric acid, very dangerous), can be really high-volume looking- you'll get 1/3 of a container of cottage cheese sometimes. I'm not sure if it's because the potassium or sodium sulfate absorbs glycerine, biodiesel, or water (probably water) from the mixture and becomes 'fluffier'. Ive been meaning to press some of it through a filter and figure out what the heck it is that makes it look so high-volume.

Anyway, the offtopic point I'm making is that you should calculate potassium content if you're not doing it via ICP, because you can't just weigh or measure salts due to their partial solubility in other things.


starting amount of soap to expect:
AMount of soap in the glycerine varies wildly based on FFA of course, but also based on whether you do a 5% water prewash to the biodiesel or not. I think my numbers doubled when I do a prewash (can't recall starting FFA content now, this was months ago).

quote:

I used a Karl Fischer on the crude glycerin soup to determine water.

I think the KF coulometric method doesn't give accurate results for high water content samples
I've done this testing a couple of times and boy does the KF coulometer not like it- I can't get repeatable results at all over a certain percentage water, and it messes up the instrument I believe.

My impression is that you don't want to use the KF coulometric instrument for this, but can instead use another method called karl fischer volumetric.


I'll see Stephen Corlett next month (we finally got my GC going after 3 years or so!) and ask his opinion, I got the info on KF Volumetric from Rachel Burton. There's a KF Volumetric instrument at the NCDA lab that we NC residents have some access to, I've been meaning to send samples to them when I get back to some research that really requires me to know how much water is in my glycerine.
will the Sandy Brae tester measure water in glycerine
Is it possible to use the Sandy Brae meter to test for water in glycerine? Ill ask Graydon to ask that of the manufacturers, I"m trying to figure out how much the $200 sandy brae tester can replace a $5000 (???) KF in the research I'm trying to do this winter.

quote:

I based the other numbers loosely on acidulation results I've had in the past. It's certainly possible my "guesstimates" are way off.

I've done a TON of acidulation and methanol recovery over the years and with the exceptions above, the numbers sound about right to me.

methanol content:
I'd say to expect about 20% methanol, but it again depends on a lot of things about your process, whether you had water in the oil to begin with that makes the crude glycerine byproduct larger by volume, whether you made a lot of soap and therefore have a lot more byproduct due to the trapped biodiesel/soap increasing that layer, whether you're doing 5% water prewash or not, how much methanol you used on the front end, and the degree of biodiesel conversion you're attaining in the first place.


You should probably talk in terms of soap rather than potassium salts.

suggestion for how to test the biodiesel content
do an acidulation, then test the resulting FFA. Compare the percent FFA to the volume of recovered oils (which are FFA and biodiesle mixed together at that point, plus any 'soluble' salts such as those created by using HCL for the acidulation)

quote:

I'm curious if anyone else has tried to figure these things out with their own glycerin, or if anyone has thoughts on what these numbers "should" be.

BTW, the feedstock was run-of-the-mill UFO from restaurants, around 3-5% FFA. The glycerin came from a 1000-gallon batch, from which we got around 300 gallons of crude by-product. I think we got about 900 gallons of biodiesel, for a 90% yield.

.

This message has been edited. Last edited by: Shaun,