Biodiesel and SVO Forums Home
Getting that last few % of water.

This topic can be found at:

May 27, 2014, 04:58 AM
Getting that last few % of water.
Ive come back to methanol recovery after a break of a couple of years. Im making a biodegradable chainsaw lubricant from glycerol and the first stage of the process is to remove the methanol.
I put together a small 100litre still from a 200 litre steel drum. A 3000w 240v stainless steel immersion heater is installed horizontally near the bottom of the drum, controlled by a PID controller and a 40a SSR.

A reflux column, 2" wide and 13" long, filled with stainless steel wool, is mounted onto the 3/4" outlet. The K type temperature probe is fitted into the top of the reflux column.
The still is filled and emptied via the 2" outlet and the 2" cap has a pressure release valve attached to it.

The condenser is a 4 ft long copper plumbers delight type.( not shown in the picture.)

The still works very well. I put 100 litres of KOH glycerol into it and it takes about 2 hours to heat up, I then run it for 12 hours with the reflux temp set at 67.3°C. I get between 11 and 16 litres of methanol at a purity of 95%.

My priority is to demeth the glycerol as completely as possible but obviously I want to be able to use the methanol for my biodiesel processing. To that end I have been trying a couple of ways of removing that last 5% of water. Im looking at 2 methods. Molecular sieve and quicklime.

Ive had some molecular sieve in my workshop for a few years now and I decided to try it out. I added 250 gms of MS to 1 litre of my recovered methanol and stirred for 5 minutes. I then strained the methanol off and tested for purity using my carbide manometer. result 98.5%. not bad but I think the MS may have been lying around for too long and would have worked better if i had regenerated it first. I am going to try that next time.

MS is expensive and has to be regenerated at very high temperatures so I want to try another method, Quicklime. I can get quicklime calcium oxide cheaply enough that I can use it once and then dispose of it. Calcium oxide reacts with water to produce Calcium hydroxide ( lime) but does not react with methanol. My problem is that I dont know how to use it so I enlisted the help of Wesley B, a member here on the forum and a great source of chemical knowledge. The following posts are our communications so far.
May 27, 2014, 05:08 AM
My first post to Wesley,
I have a question. How much water would be required to react 1gm of Calcium Oxide into Calcium hydroxide?
Im experimenting with drying my recovered methanol with quicklime.

His first reply

One gram of Calcium Oxide reacts with about 0.249934517 grams of water to form Calcium Hydroxide. That's stoichiometry. But there may be some kind of equilibrium between the three materials. In drying agents (dessicants) some are fast, some are slow. Sodium metal is a fast, strong dessicant. Sodium sulfate is a slower drying agent (water absorber), and not so strong as sodium metal. A book said to allow 8 or more hours at elevated temperature to allow calcium oxide to react with water in alcohol. Then distill off the alcohol, leaving the water with the calcium hydroxide, in the still pot. That's one way to make fairly anhydrous alcohols. Then the calcium hydroxide may be heated to drive off the water and regenerate the calcium oxide.

His next reply

Calcium Oxide reacts with carbon dioxide in the air to give calcium carbonate. So when using CaO try to avoid air contacting it, maybe dry your methanol in a sealed container. Wikipedia gives that heating calcium hydroxide to 512 degrees celcius produces calcium oxide and water, but Calcium carbonate (CaCO3) undergoes thermal decomposition when heated above 825 degrees centigrade. Calcium Carbonate decomposes to calcium oxide and carbon dioxide. Calcium oxide swells when it is used to dry alcohols, as the calcium hydroxide forms. A book titled "A Manual of Inorganic Chemistry" by Charles W. Eliot & Frank H. Storer (Massachusetts Institute of Technology) published 1872 gives "Of the anhydrous oxide of calcium...It unites with water very energetically." and "Hydrate of calcium...At red heat it gives off water, and is reconverted to quicklime." Quicklime is calcium oxide.

his next reply,
I looked at the book again. Drying alcohol with calcium oxide was done by refluxing the alcohol with CaO for 6 hours, then letting the reaction mixture sit overnight. The next day the anhydrous alcohol was distilled off. The CaO becomes Ca(OH)2 and increases in volume (it swells). I believe it produces about 99.5% or more anhydrous alcohol, if it is done correctly.
May 27, 2014, 07:06 AM
imakebiodiesel; Calcium hydroxide is a white finely divided solid powder. If you use calcium oxide to get the last bit of water out of your recovered methanol, how will you empty your still pot of the calcium hydroxide? I expect it might be difficult to clean up the equipment.
May 27, 2014, 01:42 PM
Could you increase the length of your fractionating column? Could you try broken glass or marbles or rashig rings for packing?

How are you organizing the refluxing? ie does your set up introduce condensate back into the top of the tower to trickle back down through uprising vapor?
May 27, 2014, 02:49 PM
Jon Heron
I think you are heading in the wrong direction.
Considering we are dealing with methanol and not ethanol, which has an azeotrope with water whereas methanol does not, none of that stuff should be necessary, better control of your head or column temperature can produce 100% pure methanol with no need for zeolite, quicklime etc...
I am getting 98% pure methanol with WBD on 1425L batches with no reflux column, on my old 100L reactor I got 100% recovered methanol consistently. I would say if your only getting 95% from your current setup you have a big problem somewhere?
I should add that I have only used vacuum on my setups and have never tried just an atmospheric still, the mechanism is the same though.


Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
May 27, 2014, 05:23 PM
Jon Heron
Here is a nice chart for water and ethanol which should get you in the ballpark:
Here is another handy calculator for those using vacuum:
Good luck!


Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
May 27, 2014, 08:24 PM
The book "A Text-Book of Practical Organic Chemistry" by Arthur I. Vogel third edition, 1970 printed by Longman Group Limited, gives on page 169, instructions for producing absolute methanol. You don't just use simple distillation to do it. "Absolute methyl alcohol. The synthetic methanol now available is suitable for most purposes without purification...Frequently, however, the acetone may be as high as 0.1 per cent, and the water content 0.5-1 per cent.
Water may be removed, if necessary, by treatment with quicklime as detailed under ethyl alcohol, but the method is wasteful and is not generally employed unless water content is comparatively high, say above 3-4 per cent. An excellent method of particular value in recovering the alcohol, consists in fractioning through a highly efficient column."
Just using simple distillation in a common still might not get the purest product.
May 27, 2014, 09:30 PM
That's an interesting comment thank you Wesley. I know that methyl alcohol does not produce an azeotrope, and I wondered whether a refluxer would do the job, but I looked up fractionating column and it seems these are widely used to do this task. That was the reason for my suggestion that the 13" column mentioned above might not be up to the job.
May 28, 2014, 06:16 AM
I should explain that my priority is to remove as much methanol as possible from the glycerol because it causes problems later when converting the glycerol into a lubricant.
The purity of the methanol is a secondary but not unimportant consideration.

Ideally I would distill both the methanol and water off in a pot still and then redistill the methanol/water mixture in a very efficient reflux still but that would use too much energy and also I dont want to remove too much water from the glycerol as I would just have to add it back later.
I felt that the 13" reflux was a good compromise and indeed it does fulfill my first priority.

One of the reasons I am interested in the quicklime process is that the end product calcium hydroxide ( lime) has a commercial value. I live and work in a 19th century stone railway station complex and I regularly buy traditional lime for building maintenance work, its very expensive.
When quick lime is slaked in water to produce lime it gets very hot and is then left to complete the reaction for a month or more. Ihe lime settles into a solid putty in the bottom and the excess water can be decanted off. I had thought of trying something similar with the 95%methanol.
Fill a plastic barrel with 95% methanol,
add the appropriate amount of quicklime,
Mix thoroughly with a mixer.
close the lid and leave to settle in the barrel for 1 month or until the methanol is completely clear.
Siphon off the methanol
Dry the lime and use it.
May 28, 2014, 06:47 AM
Paulus, My picture doesnt really show much about the still because of the insulation jacket so I have posted a sketch here of it. The column is packed with stainless steel potscrubs. I realize that if I made the column longer or reduced the temperature to closer to the boiling point of methanol I would get higher purity but wouldnt it take longer to remove the methanol completely?

John Heron, perhaps Im misunderstanding your setup but isnt that sloping pipe at the top of your still acting as a reflux?

This message has been edited. Last edited by: imakebiodiesel,
May 28, 2014, 03:11 PM
Jon Heron
Thanks for that Wesley. That is some serious purification!
I can confirm that 100% purity methanol, as measured by a cheap alcohol tralles hydrometer, can be consistently achieved using vacuum and a temperature controlled pipe with a pot temp not exceeding 175F under 29Hg of vacuum. Whether or not that is called "simple distillation" I dont know? I test the accuracy of my hydrometer against the production methanol I use for brewing which I am certain is no where near as accurate as things are measured in a laboratory. In any event it works well with my set up.
John Heron, perhaps Im misunderstanding your setup but isnt that sloping pipe at the top of your still acting as a reflux?

Hi IMB, I suppose it is, I dont know a lot about it to be honest. My understanding was that a reflux column is not a temperature controlled column, rather the temperature was controlled by the heat in the pot? I also thought a reflux column was something filled with a media where the vapour circulates around for several moments? I could very well be wrong with both those understandings though...
What I do is control the temperature of the pipe separately from the pot. My pot temperature is not permitted to go above 175F (due to WBD) and my goal is to keep the pipe at 150F for methanol extraction (175F for dewatering).
Upon reflection I am thinking that when I increase my condenser size adding some media to the pipe may give more surface area for the water to condense on... As long as I can heat the media well it should not effect dewatering either... Hmmm... something to think about...


Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
May 28, 2014, 05:40 PM
Jon is the one with the practical experience in this, I would listen to what he has to say. Does 100% purity imply there is no methanol left in the distilland? I think Jon's suggestion of a heated riser to get the vapor up and over the hump is what makes his designs so successful.
May 28, 2014, 05:43 PM
john galt
I'm using a hydrophilic polymer [Quik 'n' Dri™]to get the last bit of moisture out of my fuel mix. These water absorbing polymers are commonly used in diapers or for retaining water in house plants. A quick search for "Water Absorbing Crystals" will find many sources.

I don't know how it would work for drying methanol, but it's easy to try.

May 28, 2014, 06:28 PM
Sometime close is good enough.

May 29, 2014, 04:02 AM
Does 100% purity imply there is no methanol left in the distilland?

I distilled a batch yesterday and this time I replaced the collection vessel when it was about half full after 6 hours. The first half was 98% pure and the second half was only 92%. So the answer to your question is no 100% purity would suggest that there is a lot of methanol left on the glycerol.
Im thinking now of a modification like this. This still is both a pot still and a reflux still. I have taken Jon's advice and added a band heater and pid to the extended tower ( Paulus's suggestion) so that I can control the temp independent of the pot.

With this still I can extract the first half of the methanol at very high purity >98%. I can then switch over to pot still mode and remove all of the rest of the methanol quickly but at low purity.
I would not need to do any purification to the first portion and after 10 or 15 batches I would have enough of the lower quality second portions to fill the still and re distill it. What do you think?
May 31, 2014, 08:11 AM
Further up the page John Galt suggested trying super absorbent polymers to remove water from methanol, so I thought I would try it. The jar on the right is my 95% methanol and the other one is virgin methanol. The saps havent swelled in either sample and the water content in my sample hasnt changed.

June 04, 2014, 12:21 PM
imakebiodiesel; I tried to determine what mesh size of molecular sieve to use to dry methanol, ethanol, isopropanol, 1-propanol etcetera (alcohols) can you tell me what mesh size is best, and the composition of the molecular sieve beads? I think mesh size 8-12 is available here.
June 05, 2014, 03:00 AM
imakebiodiesel; I didn't do my homework relative to the question I asked about the molecular sieve. I found an answer on line. The correct molecular sieve size to use for drying alcohols is 3 Angstroms, or type 3A.
June 05, 2014, 03:08 AM
Yes 3A is the type I use. The actual size of the pellet doesnt really matter. Smaller beads have a greater surface area and therefore work a little faster but in practice it makes no difference.
June 11, 2014, 12:29 AM
I'm not sure if it would ever be necessary or cost effective for the production of synthetic fuel to make"genuine absolute methyl alcohol". Again from "A Text-Book of Practical Organic Chemistry" By Arthur I. Vogel, 3rd edition, published 1970, by Longman Group Limited, on page 169 is given a proceedure to further dry commercial anhydrous methanol which maybe 1% water. If you dry your 95% methanol using calcium oxide, you should get 99% or dryer methanol. To get that last little bit of water out you can prepare it " by treatment with magnesium activated by iodine; the dehydration is due essentially to the magnesium methylate formed "...5 grams of magnesium and 0.5 grams iodine put in a 2 litre glass flask, fitted with a glass reflux condenser: add 50-75 ml of methyl alcohol, through the condenser and the mixture is warmed until the iodine disappears. Hydrogen gas is evolved (flammable). Next add 900 ml commercial anhydrous methanol (99%+ anhydrous). The mixture is boiled under reflux for 30 minutes. The product is distilled with exclusion of moisture. The first 25 millilitres of distillate is discarded. That is suppose to produce almost 100% anhydrous methyl alcohol.