Biodiesel and SVO Forums Home
3/27 test & unconverted oil, question?

This topic can be found at:
http://biodiesel.infopop.cc/eve/forums/a/tpc/f/9411061471/m/3807023973

August 28, 2014, 08:16 AM
Rick G
3/27 test & unconverted oil, question?
I make a few hundred gallons of BD per year with the two stage water wash method using waste cooking oil. The oil looks terrible but usually titrates less than 5 ml/L (NaOH). I suspect but have never actually determined the contaminants are charred food particles. This is surprising given the titration but I believe the restaurant cleans the oil mid-week with magnesium sulfate to reduce soap.
I have two questions.
First, at the end of the two stage test I either have only a trace of unconverted oil or about 3% according to the 3/27 done in a conical graduated centrifuge tube. If I have 3% unconverted and try to reduce it further with a third stage and prolonged processing most of the time it will remain at 3% and I'll produce much more soap with reduced yield (vol finished BD/vol of initial oil).
Has anyone else experienced this outcome? One other observation, I haven't determined this carefully but when I have warmed the centrifuge tube to approximately 100F (40C) the oil remains undissolved. I've never tried virgin cooking oil but the question arises at what point will cooking oil dissolve in methanol?
My second question is about yield, defined as volume of BD per Vol of initial oil. Using my source of waste cooking oil I very consistently obtain 87-89% yield. Is this typical for other producers?
How does this yield compare to virgin oil?
August 28, 2014, 10:17 AM
fabricator
By stages are you talking about acid/base or base/base? Either way have you tried bumping up your base titration number? Unconverted oil is simply oil that has not reacted with the naoh, this could be because your oil is slightly wet as would be the case if it contained solids as these will hold water.
Your solution would be to either dry the oil somehow or add to your base titration number.
August 28, 2014, 02:42 PM
Dgs
Hi Rick G,
Yes, I have experianced the small amount of dropout that seems reluctant to react that you describe.
From the conversion you have I would think that you have some water in there.
You don't mention glycerol pre-washing(glycerol from the previous batch is mixed with your oil at @ 50degsC for 30mins,left to cool and separated)
This will help reduce the water in the oil and reduce the titration.
Your process will be far more efficient after doing this.
Your yield will increase and you may not get the tiny amount of dropout.
August 29, 2014, 10:08 AM
Rick G
Fabricator, by two stage I mean two Base treatments. Since my oil titrates consistently between 2.5-5 ml of 1 mg/ml NaOH per ml of oil I feel that an acid treatment isn't warranted. My concern is that once I have made two Base treatments there is sometimes this trace (</=0.1 ml per 3.0 ml sample)that remains. As I have indicated subsequent treatment with additional Base treatment using appropriate amounts of caustic soda (NaOH) & methanol even with prolonged agitation (overnight) and sometimes additional heating up to 130F ( 54C) this residual trace remains.
Dgs, I find it difficult to believe that after two good base treatments with glycerol dropout there can be much water remaining in the batch.
Just now I did a simple experiment where I placed one drop (approximately 0.1 ml of clean waste cooking oil (free of sediment) into my conical tube and added 27 ml of methanol. At 70F (21C) some of the oil remains undissolved (about 1/2 by eye) whereas at 100F (38C)almost if not all has dissolved. (Using waste cooking oil there is usually a small amount of non-methanol soluble sediment probably due to food material contamination ie. carbohydrates, proteins etc.)
I can't remember exactly whether I have ever performed this experiment with a unconverted sample but I believe so. If I remember correctly it didn't dissolve at the higher temperature. This leads me to believe that it isn't a typical fatty acid oil ie. glycerol ester.
As a further experiment I should add 2.9 ml of 100% converted biodiesel in order to determine whether biodiesel itself affects the solubility of cooking oil but I don't happen to have a sample of 100% biodiesel at this time. (Given that the 3/27 test is designed to take advantage of the solubility difference between cooking oil and biodiesel fatty acid esters the 9 to 1 ratio of methanol to biodiesel is right at the borderline of biodiesel solubility. Raise the temperature a little and more unconverted cooking oil dissolves, lower the temperature and some biodiesel remains undissolved.
Now that I think about it there is some possibility of oil-like contamination from spices etc. Spices are used specifically for the flavors & aromas from the oils they contain. If you cooked large quantities of spicy food in the deep fryer vat over the course of time they might accumulate enough to be detectable.
August 29, 2014, 11:22 AM
Dgs
Rick G,
Because your yield is low, I was refering to water in the oil,not in the bio.
This 3% or so dropout.-I have a pupil that sometimes has this,if he re-processes with the correct amount of chemical,the dropout is sometimes still there.He will then re-process with the same amount of chemical again and after this 2nd re-process the dropout is gone.
Re dropout temperature,I have found that 22degsC is the max the test should be done at. Any higher and you risk some of the oil dissolving into the methanol phase.
However on the low side,when I have a clear pass,I have never had a situation where the bio comes out of solution,down to @ 10degsC.

The 3/27 and beyond thread is interesting to read.

The only way you will ever absolutley know what your dropout is,is to have it analysed by GC.
August 29, 2014, 02:31 PM
Biotom
my experience for what it is worth, each point of titration reduces yield by about 2%. So your yield seems to closely fit my experience.


" I don't know what I don't know until I know"
1994 GMC 6.5 Tubo 2005 Dodge ram 3500, 3 VW's 2000, 2002, 2005.
August 29, 2014, 03:19 PM
dukegrad98
How has the thread made it this far without a multi-puppeted diatribe about how the 3/27 test is meaningless and you can't tell anything from the fallout and and and and and...?!?!? Roll Eyes

Cheers, John
August 29, 2014, 06:07 PM
Dgs
I don't know,but isn't it nice. What a pain in the _rse that man is.
August 29, 2014, 07:04 PM
john galt
The Ignore List is very effective.
click on a user name, add it to your list, and make their messages disappear.



August 30, 2014, 07:33 AM
dannodiesel
Wow Tilly!! Very good!! Thank you...
Welcome back...
August 31, 2014, 10:01 AM
Rick G
Dgs, I have reprocessed a second time (a total of four base treatments) and still had the dropout only less but I can't tell you the temperatures now.
I mentioned above that my waste cooking oil comes from two sources. One, a pizza shop which in my opinion overheats their oil but does treat it midweek to remove some soap. The other, a bar &grill (mostly bar) which doesn't over heat the oil (much more clear, less charred material). Usually I mix oil from the two sources about equal parts. In a batch yesterday I used oil predominately from the bar&grill. As I expected it contained more soap but no dropout. The 3/27 following the second base treatment showed no (not even a trace of) dropout. Since the cardboard cover on the "cubbie" for the bar&grill were removed before I picked up I can't compare fresh oil sources but I can certainly ask at the next pickup.
I wonder if the residual that I see in the pizza shop oil is some kind of polymer generated by excess heating at the pizza shop? There is an extensive chemistry using fatty acid esters as paint polymers where the fatty acids react at the unsaturated double bonds to form such polymers (paint drying is just such a process where oxygen radicals initiate the polymerization). We normally think of those bonds as pretty resilient to degradation but concentrated lye has long been used as a paint stripper. I'll need to think about this a bit, I'm not sure I want those kinds of polymer compounds in my biodiesel available to form deposits in my engines.
For what it's worth.
August 31, 2014, 10:13 AM
Rick G
Kieran C. I share your concern about the proper interpretation of the 3/27 test. We know that not only does the fatty acid composition of the source oil vary (sometimes substantially) but the solubility of fatty acid esters as well as the mon, di & tri-glyercides in methanol probably differ as well.
Do you know if anyone has published a comprehensive study of the solubilities of the various components expected from incompletely converted biodiesel?
I would be very interested in seeing such a study. If I had access to reasonable quantities of the individual pure components (methyl esters & mon, di & tri-glycerides) it would make an interesting study.
That would be a great student paper or even a Master's degree project.
August 31, 2014, 11:41 AM
Rick G
Biotom,
I had forgotten the 2% factor, thanks for reminding me.
But this then begs the question. If you could remove the soap before the conversion how much fuel could you save? The losses are due to two principal types. First the soap itself and second the fuel entrained in the glycerol byproduct and wash water. If the lost is primarily due to the soap itself then little is to be gained. However if the loss is due to the entrained fuel maybe it will amount ot something.
We can estimate the amount of soap directly from the titration with NaOH(or KOH). If the titration solution is 1mg NaOH/ml. then one ml of titration solution corresponds to 0.025 millimoles NaOH/ml. Since the titration is one to one for NaOH to fatty acid and an rough approximate molecular wt of fatty acids in cooking oil is 280 grams/mole this means that one ml titration is 0.025*280 or 7 mg soap per ml of oil or 7 grams per kg of oil or roughly 6.3 grams per liter. And again assuming that the density of free fatty acid soaps is approximately the same as fatty acids waste cooking oil has about 0.6% (by vol) per liter oil. This is only a 1/3 of the lost so maybe 2/3 of the loss is fuel (1.4%).
If the oil titrates 5 ml NaOH solution this amounts to 7 liters of fuel per 100 liters of oil (1.4 x5 = 7). Now the question becomes is the fuel savings worth the cost of the magnesium sulfate and the pretreatment costs? It doesn't take into account the value of the improved efficiency or reduced water use.
For what it's worth.
September 01, 2014, 04:29 AM
WesleyB
Rick G. "The restaurant cleans the oil mid-week with magnesium sulfate to reduce soap." I used to work in a fish and chips restaurant as a fry cook. The senior cook treated the cooking oil with magnesium silicate not magnesium sulfate. A brand name Magnesol may have been what we used. It adsorbs the more polar materials in the oil, then the powder is filtered out. The last time I counted there were about 22 drops of water in a millilitre, not one drop in a tenth of a millilitre. I enjoy reading your posts. It seems you have a chemistry education. That's good.
September 01, 2014, 09:33 AM
Rick G
Wesley, Thanks for the tip about Magnesol. The point of using one drop of cooking oil was to simulate what I saw while testing my batch. There would be about one drop of unknown material remaining in the bottom of the centrifuge tube after testing a 3 ml sample. The question is, "what is that one remaining drop?" The reason why I get about 0.1 ml of oil from one drop is that the 5 ml pipet I use to sample the oil has been modified by grinding off the tip. This makes a bigger hole that doesn't plug easily with food particles and drains better. In order to maintain accuracy I use the volume graduation marks between the 3 and 4 ml carefully allowing the oil film to drain down the side of the pipet each time. You are correct a pipet designed to dispense low viscosity fluids does deliver many more droplets per ml.
September 05, 2014, 08:06 AM
dannodiesel
Funny!! Funny!, lol

Shouldn't the 10/90 conversion test be the new minimum standard now?
September 05, 2014, 10:14 AM
Dgs
Hi Jethro,
I pointed all your comments out to him at the time,but he was insistant that he was correct.
Looks as though we have someone else to add to the "sockpuppet" tribe.