I'm always interested in trying something new, and having had several very low yielding batches last year, I thought that I would test the method posted by Mark Imisides.
Like dgs, I did a full batch over Christmas of 200 litres, and my overall impression was that this method did give me a higher yield AND significantly less soap. I say "overall impression" as my test was qualitative rather than quantitative; it was difficult to provide accurate data due to the starting oil titration being 8koh, necessitating me using an enhanced prewash that contained some added methanol/KOH as well as the residual chemicals in the glycerol.
I did encounter one problem, which was the production of Calcium Soap; these remained suspended in the bio even after washing, but came out after I passed the bio several times through oak shavings held in a bucket suspended over my processor. The bio produced was bright, light coloured and very transparent. I'm using it in my Ford Focus car now.
Moving on, I thought it would be interesting to myself and maybe the wider biodiesel community if I carried out a more controlled test. At this point I should say that I am in no way a scientist; my knowledge of science is restricted to my experience of teaching science to children at Primary School (ages 4 to 7).
The first stage of my testing is to prepare the methoxides. I have done this today, and will share my practice here. I welcome any criticisms or improvements anyone might like to make as I go on. Just remember I am not a scientist.
I prepared the methoxides in accordance with the method outlined by Mark Imisides. I started with 1500ml of virgin methanol and added 150 of KOH. Once dissolved, I added 400ml of this solution to three bottles. Having read Tilly's warning about mixing methoxide in plastic bottles (his patent pending Dr Pepper technique), I used three identical glass bottles with screw caps.
I didn't add any drying material to the first bottle. This is my control.
To the second bottle I added 60g of quicklime. This came from a sealed drum that dgs gave to me before Christmas.
To the third bottle I added 60g of ordinary Portland Cement. This was from a new bag.
I have shaken the bottles regularly to maximise surface contact between the powders/liquid.
The bottles will now have to sit a couple of days to allow the fines to drop and leave clear methoxide at the top.
I have purchased three new litres of vegetable oil from my local discount supermarket. I thought that using new oil with next to no titration would be a better test than using used oil. I have also bought three identical 2litre plastic water bottles in which to react each methoxide with the new oil.
I will post further as the test progresses over the next few days.
image.jpg (24 Kb, 25 downloads)
There was some interesting testing performed and results posted on another forum using this procedure about 5 years ago.
I look forward to your results.
These are the three methoxide bottles 24 hrs after I last shook them. The liquid above the layer of quicklime and cement is clearing but still not completely clear. I will leave it another 24 hrs to see if any more fines drop out of suspension.
image.jpeg (42 Kb, 28 downloads)
I have used mostly new corn oil in my experiments in making various types of biodiesel to minimize problems caused by higher levels of free fatty acids. I titrated the new corn oil. It did have some free fatty acids in it. My titration number was 0.6 grams sodium hydroxide per litre of oil. Even though it's new oil it still does have some free fatty acids in it.
can you be more specific what quality or qualities of this procedure you are testing and how you plan on testing them.
Also I noticed that you mixed 1500ml Methanol with KOH,you poured 400ml into each of three bottles which totals 1200ml. There is 300ml methanol/ KOH mix missing.
You are correct; I mixed a bit more methoxide than I needed rather than trying to mix an exact amount. I did this in case I lost a bit during the mixing due to my carelessness or evaporation of the hot liquid.
I measured the three lots of 400ml of methoxide using a laboratory glass measuring flask and carefully poured it into my methoxide bottles without spillage. I then disposed of the surplus methoxide. I do not intend using all this methoxide in my test, but using 400ml samples allows me to draw off the required amount of methoxide for my experiment without disturbing the sludge at the bottom of the bottles.
I would value any input you or anyone could make regarding how to proceed with the test from this point.
I propose to react three lots of 1000ml of new oil using the ratios outlined in Mr. Imisides's posted method.
What I am measuring/observing: The claim of this method is that it reduces the water that is produced when mixing KOH with methanol. The water is supposed to be reduced by adding quicklime (CaO) or ordinary cement to the methoxide. For my first experiment, I think that it is reasonable for me to measure and compare the amount of glycerol produced in a reaction using the CaO/cement methoxide and compare this to the amount of glycerol produced by the control sample of methoxide. My rationale for this is that anhydrous methoxide should produce less soap. The majority of soap usually migrates to the glycerol layer, so the anhydrous methoxides ought, in theory, result in a better yield (more biodiesel and less glycerol).
Once again, I am not a scientist and I would welcome any suggestions about how I can improve the test.
There are other claims for the procedure that catch my eye, such as:
1. It is foolproof (no titration required)
2. It proceeds at room temperature. (The reaction appears to go to completion in 10 minutes)
3 15% methoxide required in a single stage reaction without glycerine prewashes, enhancments, neutralizations, etc etc.
Ok Tilly, I hear what you are saying.
But as a first experiment I think that 1st stage yield is a reasonable start, as it is this aspect that I find most interesting. If it is true, then for minimal extra effort I will be able to achieve more efficient reactions and easier washing (due to less soap).
I'm keeping an open mind.
That is fine.
The fellow on the other forum did a number of experiments too that were quite interesting.
I await the results of your tests.
Do be aware that this procedure was not designed around the idea that it would produce high conversion biodiesel. However, if you use a 20% methanol/ lye mixture in the reaction that was mixed according to the instructions, that works out to 20g KOH per litre of WVO.
That on it's own should give pretty high conversion with most waste oils you are likely to get from a shop
I'm going to post in a minute regarding the stage I'm at with the testing, but thought I would answer the important point you raised above separately.
The reason why I am enthusiastic to test quantitatively whether the method works is not particularly because I am trying to achieve high conversion biodiesel over one stage. My interest is in whether this process improves the yield, over however many stages it takes to achieve full conversion.
At this point in the test, I am focussing upon the amount of glycerol that is produced after just a single stage. I think that is a pretty good starting point for us to discuss whether the claims of the anhydrous process are valid, the hypothesis being that using quicklime or cement to dry the catalyst results in a more efficient reaction and higher yield.
Once I have performed the initial reaction using the three catalysts, we can then consider what additional testing would be useful.
Hello again Tilly
Before I go on with the testing (I'm now ready to go) I would like to ask you to check your calculations. According to the method as originally written, Mr. Imisides says that after mixing the methoxide solution, he adds this at a ratio of 100/15 oil to methoxide. This means that for my test I will be mixing 1000ml of oil with 150ml of the methoxide. That is 15% methanol and 15g of KOH. I don't want to mess up the test - am I missing something?
Here's where I am with the test - all ready to go. I have carefully calibrated my reaction bottles using a glass lab measure and added exactly 1000ml of oil to each. I'm waiting for the oil to warm up a bit and also some clarification whether the recommended amount of methoxide is 15% or 20%.
Current oil temperature is 19.5C.
image.jpg (30 Kb, 10 downloads)
According to MI as the process produces no soap, I wouldn't think the 15gms KOH is important. As for the methanol amount, I will let someone else comment.
What I will say is that using these amounts the test will be biased towards the 2x dried methoxide tests in that the undried methoxide will have 3.2% water plus the water in the methanol and also 15gms of catalyst is a massive amount to add to 1 litre of new oil.
I think that maybe I should stick to the method as originally posted, that is a 100/15 oil/methoxide mix. And yes, and initial hit of 15g/l of KOH during just the first stage is a massive amount.
Hopefully I will have time to do the test tomorrow.
The instructions are "Now just add the methoxide solution to the WVO in your normal quantities at room temperature."
Because this is a single stage method as written, most people I know would normally use about 20% methanol in a single stage reaction.
He then goes on to say "I generally use a mixing ratio of 100:15."
He made it clear many times that high conversion biodiesel was unimportant and this is the amounts he uses.
He seemed to consider that if there was separation then he had successfully made biodiesel.
As a side note, the last biodiesel I made I used about 14% methanol and 10g KOH per litre.
If you use only 150ml methanol as you point out that is still 15g KOH in new oil which is quite substantial.
An interesting comparison would be to use 8g KOH mixed into 150ml methanol without the CaO drying with new Veg oil and see what the difference is.
I edited this post when I noticed that I had mistakenly written 15ml instead of 15%
I changed it to 150ml methanol for clarity.This message has been edited. Last edited by: Tilly,
Water is only the result of methoxide production.
One of the claims made in this procedure is that only a small amount of methoxide is actually produced when you mix KOH with methanol without using the drying stager.
This means that only a small amount of water is actually produced in mixing KOH with Methanol.
The chemist neutral also thought that mixing KOH with methanol resulted in very little actual methoxide production and mostly what you had was KOH dissolved in methanol which does not produce water.
Of course there is also the 10% water content in the KOH if your KOH is 90% purity.
My sandy brae tests on Potassium Methoxide did indicate that several % of water is formed when the methanol and KOH are mixed.
I very much doubt if all 10% of the missing purity of KOH is all due to water.
Yes, there is the 10% water contained in the KOH as I pointed out in my last post.
Everything I have read such as this MSDS says the remaining impurities is water.
From Cleartech MSDS
Potassium Hydroxide- 89.5-90.5%
What do you think the remaining 10% is?
"KOH forms a series of crystalline hydrates, namely the monohydrate KOH·H2O, the dihydrate KOH·2 H2O, and the tetrahydrate KOH·4 H2O" - wikipedia. And another contaminant is carbonate.
Dissolving K in an excess of MeOH produces ions K+ and MeO- in solution. KOH and MeOH produces K+ and MeO- and H+ and OH-. As we know, these latter two can participate in side reactions, whether or not you have produced "methoxide." It's all just a big soup of ions floating in an excess of MeOH. Equations not balanced, note.
I suppose it's worth nothing that pure "methoxide" is a white crystalline (ionic) powder. You can have pure table salt too which is a white crystalline powder, and clearly when you dissolve that in water it's no longer "pure" in the strict sense.
|Powered by Social Strata||Page 1 2 3 4 5 6 7|