I followed the Imisides method to the letter. I've described it as I've gone along. The 400 ml of methoxide was my starting quantity that I used to make the sample solutions dried with cement and quicklime.
I added 150ml of methoxide to each 1 litre sample of oil. This 150ml of methoxide contained 15g of koh.
Ps. From the above you will see that I have 250ml of each methoxide sample still unused. I'm thinking of repeating the experiment using USED oil. I have 220 litres in my processor that I can use. It's a mix of very grotty oil and reasonable oil. Is there any merit in my plan to test again using waste oil?
Interesting 3/27 results.
What did you deduce from these results?
Can you post the production method you used.
For instance, did you perform the reactions at room temperature for ten minutes which is what the instructions say is all that is required.
I think performing the procedure with used oil is a very good idea.
Interesting results David.
I am suprised that the dropout in the 10/90's is so high.
One small point is that the 150mls of methoxide added will not contain 15gms of KOH. It will be more like 14.3gms.
The original 40 gms of KOH added will occupy about 20mls of space, so you will end up with slightly less than the 10% KOH in the mix.
Yes, I did wonder about exactly how much koh was in the mix. I gathered it would be slightly less than 15g but wasn't sure how to calculate. But I think that the important thing is that I followed the Imisides recipe exactly both in terms of preparing the methoxides and the volume of methoxide added - 100/15 ratio by fluid volume.
I think I'm going to perform the 10/90 tests again just to make sure. More in my next post when I reply to Tilly.
Pity we never got to see the three completed batches in the initial photo that failed to load. Did the control batch (No CaO or cement drying) split into two layers of byproduct as well?
Does everyone agree that the 10/90 result tends to indicate residual di and triglycerides ie the reaction did not go to completion?
Hello Tilly and thank you for your comments.
This was the first time I've used your Pat Pen Dr Pepper technique, so I tried to follow your instructions for the tests. The reaction temperature was 30deg. This is more than room temperature but it is about the temperature I use when doing a full batch in my processor. After adding the methoxide I shook for 10 seconds and shook for another 10 seconds after 20, 40 and 60 mins. Already I can see that in terms of processing time and temperature I have varied from Imisides's instructions, however I think at the moment I am most interested in whether the quicklime and cement has a beneficial effect on the reaction. I would agree that my test involved a higher temperature and greater mixing time than the published method. I'm not so interested in Imisides's claims of short mixing times and lower temperature.
Tilly- I wanted to ask you about the 10/90 tests. I think I'm going to perform them again just to double check my results. I only have one dropout tube. I'm a little concerned that I ran one of my tests overnight and one in the daytime when the methanol would be at a different temperature. Is there anything I should be doing to make sure it's a fair test? Run it overnight over three days to reduce temperature variations?
Here is the picture again. I've managed to reduce it in size for posting. The samples are identical. The byproduct is split into two layers in all three samples (yes, including the control sample)
image.jpg (14 Kb, 19 downloads)
I think it is safe to say that if there is fallout in the 10/90 procedure then the biodiesel is probably not ASTM conversion.
That is about all that you can say.
It is necessary to have the biodiesel and the methanol in all three tests at the same temperature.
Of course the 3/27 test is only a Pass/fail test. I know some people pretend that it will tell you the percent conversion but for some reason they always fail to provide the proof this is so.
"The control sample had a total of 0.8ml of dropout. Of this, 0.6 ml was clear oil with a 0.2 ml lighter cloudy layer above."
So the parameters of the control were such that it failed to produce a good batch. The calcium treatment improved somewhat but still failed to produce a good batch.
Out of 10ml of biodiesel in each of the three tests there was a difference in results of 0.3ml drop out.
Is this tiny difference significant or would it fall within "experimental error"?
I'm going to run the 10/90 test again Paulus - just to confirm my result. This time I will photograph the dropout in the flask..
I know something about doing organic chemistry research. A part of it is conditions at the time. I have been reading some of this conversation. As I recall Imisides claimed that drying the potassium methoxide methanol solution caused the reaction to proceed at room temperature in a reasonable period of time. Now my room temperature is twenty three degrees centigrade. What is the temperature, the conditions you are doing your experimental reaction at? Thanks for your time.
The claim made by Imisides was a 10 minute reaction to completion at 8deg C.
The outside temperature here in Leeds is 4C. My house is between 17 to 22C inside.
I actually warmed the three samples in a bath of water, and the reaction temperature was around 30 to 31C
How is the snow in Leeds this morning?
A couple of things,
From testing I have done, your dropout of 0.5mls in the CaO sample equates to about 3.5% triglycerides remaining in the bio.
(I'm sure that will spark a response from someone)
Also, I'm wondering what titration your new oil had. When I have tested new oil before it does suprisingly have more FFA's present than one would expect. I think Wesley has also confirmed this.
Heavy snow over here in Leeds, which seems to have caught everyone out! Just got home from school in time before it started.
I honestly didn't titrate the new oil, as I assumed it would have a titration of 0KOH. My bad! What I can do is go and buy another litre from Lidl (they had a stack of it so I think it will all be the same) and titrate that.
I am redoing the 10/90 testing again, this time under standardised conditions (in my cellar!).
I have now completed the soap testing. Done that twice, and I will post my results in a few mins.
I am going to run the full experiment again using oil that has been drying in my processor. Its a mix of good, bad and downright ugly oil. I've been drying it in my processor for three days (heating and bubbling) and it is now clear. You will remember that Dr. Imisides's sole criteria for whether the oil was dry enough was that it was clear.
You will be relieved (coz I know you will ask!) to know that I have titrated the oil in my processor, and it comes out at 3.5 KOH.
Soap tests for the NEW OIL to follow.
Here are the results of the soap test for the bio produced by the control, quicklime dried and cement dried methoxides that were reacted with 1 litre of new oil
I carried out the soap test twice , and although there is slight variation, the overall result in terms of trend is the same
SOAP TEST RESULTS
-----------------------TEST 1 (ml HCl titrant/ppm)--------------------TEST 2 (ml HCl titrant/ppm)
CONTROL-----------------------1.1ml / 352ppm-------------------------------------1ml / 320ppm
QUICKLIME---------------------1.2ml / 384ppm-------------------------------------1ml / 320ppm
CEMENT------------------------2.2ml / 704ppm-------------------------------------2ml / 640ppm
The control and quicklime dried methoxides produced biodiesel containing almost exactly the same amount of soap as each other. The cement dried methoxide produced almost exactly double the amount of soap
I am happy that these results are accurate. I've been very careful in terms of measuring the sample size and the amount of titrant added.
You have done tests that show this???
Please describe the tests you performed that shows this to be the case.
If your testing shows there are about 3.5% triglycerides present, what percent mono-glycerides and di-glycerides did your testing show was also in the biodiesel?
|Powered by Social Strata||Page 1 2 3 4 5 6 7|