Over the w/end I tried a few experiments, and one of them was a real surprise to me as I've read on this and other forums many times about reversing reaction.
It has been said that if you try to recover methanol from bio with glyc still present then you will end up reversing the reaction and unmaking your bio back into oil again.
But some commercial biodiesel companies allegedly do it using high temp and vacuum.
Some homebrewers have been neutralising the NaOH/KOH with acid to stop this reversing happening and are then successfully recovering methanol from the whole batch.
I thought I would try recovering the methanol from a small 30L batch without removing the glyc or neutralising NaOH/KOH with acid.
Processed using my GL style processor with venturi and condensor.
After the reaction was run and a 3/27 with no fallout obtained I just opened up the valve for the venturi and started running cold water through my condensor.
After almost 4 hours I was getting almost nothing coming out of the condensor and the cubee had a nice amount of methanol in, and was about what I would usually expect to recover when I demeth the bio and glyc seperately.
I performed another 3/27 test with no fallout and then turned everything off and left it for the glyc to settle out, after a couple of hours it was starting to thicken and I drained glyc off, and then bubbled the remaining bio for 2-3 hrs to remove any residual methanol, next I performed another 3/27 and still no fallout, I also tried a shake em up water test for soap and got very little.
I then successfully repeated the test on another 30L batch to make sure it wasn't a fluke
So it seems the reaction didn't reverse and I saved time and energy over my normal method.
I've run this past Neutral and he says I may well be on to a good method, reducing complexity, and he is interested to hear how it turns out for others.
I've got one or two guys on the goat forum who are also going to try demething a whole small batch, and is there anyone here brave(read crazy) enough to try a demething a whole small test batch?This message has been edited. Last edited by: Chug,
Hmmm sounds interestings. So did you use KOH or NaOH?
I use NaOH, I'm just waiting for a mate to report back as he was demething his whole batch a couple of hours back and should be finished by now. I'll report back when I hear how he got on.
just heard back from Jim (jamesrl venturi dude on the forum) who has also tried a whole batch demeth and it worked for him, he tests the recovered methanol and it's looking like 99% purity so thats good, and he recovered 36.36% of the methanol used, starting out with 22% of the wvo batch.
I did another successful whole batch demeth yesterday.
I demethed this whole batch for 2 and 3/4 hours and then stopped after I worked out that the stoichiometric quantity being around 12.5% (which comes out around 18.5 litres) should leave 13.5 litres excess which should be recoverable, once I had recovered 13.5 litres I stopped.
Last time I kept on going until the methanol stopped coming but it was probably starting to reverse (Jim also kept going until the meth stopped coming and then had slight fallout)
A shake em up test with water on the first batches is now showing a tiny trace of dreaded white stuff (DWS) on the interface between bio and water. so I think I had recovered more than the excess over the stoichiometric quantity in the first two batches.
But this batch shows no DWS on the water/bio interface.
I'm also gonna borrow a camera and take a video of the glyc settling out as it's amazing, you can actually see it happening before your eyes. I've never seen glyc drop that fast before and it's something to behold, and I thought I had virtually no soap last time but this time it's confirmed, almost zero soap, so it must be settling out superfast in the glyc too.
All in all I'm pleased with this experiment and it may be me but the resulting bio seems to be less viscous than my usual. I'm definitely going to try this again.
Nice going on this line of testing.
I question coming to the conclusion too firmly based on 3/27 alone, since it mainly tests for TG. As I understand, in the event of a reverse reaction, TG levels would be the last to rise of the 3 MG first, then DG, then TG, no?
Any way you can check for rise in MG and or DG?
Do you have access to safetest, CG or pHlip tests?
03 Dodge 2500 B100 homebrew
79 Rabbit B100 homebrew
I agree with you Andrew,
the only other test as suggested by Graham Laming was the shake up test of de-methed, de-soaped bio with distilled water, where the DWS shows as scum on the interface of bio and water, my first batches had a tiny amount after 24 hrs, this last batch has none after 24hrs.
You know yourself from experience when you get a good batch, or if it's a bit iffy, to me this one looks good.
I'm liking what I find so far and will continue with more trials, but even if you do single stage or 2 stage reaction if you have a GL style processor it's still worth doing a small whole batch demeth one time just to see glyc and soap settling out so quick.
Hi Chug, followed your link from the GL Ecoprocessor thread.
Very interesting to hear what you've achieved. I would like to do some testing of this as well but I'm afraid I didn't do too well in Chemistry, and I'm struggling with the stoichiometry. Essentially what I need to know is how much methanol I should be looking to recover, but I'm also interested in how this is calulcated if you don't mind explaining.
I start with 80 litres of WVO and add 17.6 litres of methoxide (22%).
Any help will be appreciated.
the stoichiometric quantity depends on the feedstock oil and on the amounts of the various different fatty acids in the oil, the amount varies from approx 11.5% for rapeseed oil (canola) to 16.5% for coconut oil. But the average is taken as 12.5%
Most common oils that are used for making biodiesel only need 12 - 13% methanol (or 12.5% average) and the excess up to 22% should be available to be recovered.
If you use 22% @ 17.6 litres of methoxide.
Divide 17.6 by 22 to give you 1% or 0.8 litre
then multiply 0.8 by 12.5 = 10
subtract 10 from 17.6 and that leaves 7.6 litres which should be recoverable.
Thanks for that Chug - seems pretty simple when it's explained like that. I use low titrating canola so I'll work on 11.5%.
I'm going to give it a go in a week or so when I have time. I thought of a possible test for MG and DG's - how about reprocessing a small batch of completed BD. Wouldn't need much NaOH, maybe 1g/litre. If the conversion is already complete this shouldn't produce any glyc I would imagine, so if glyc did fall out this could indicate partial reversal.
Let me know your thoughts as to whether this would be valid. It would require care to not make soap.
Yes reprocess test with 1g NaOH per litre should produce no glyc if all the oil is converted to methyl esters. Soap shouldn't be a problem as all the free fatty acids (FFA's) will have been neutralised.
And I would stick with 12.5% even for rapeseed oil (canola) as 11.5% is the value given for virgin oil, used oil will have more FFA's.
The calculation for recovered methanol is going to depend on the methanol used, the molecular weight of the oil, and the density of the oil.
Hi chug, you mention that after you demeth and drain the glyc, the BD got no soap?
that means you don't need any wash, or dry wash at all? That's very impressive....
can you confirm that you don't need wash after the drain? And how long did you settle?
Yes just demeth the whole batch then settle the glyc for 30 - 45 mins (much longer and it can start to solidify in pipes) then pump bio to settling tank and bubble with aquarium pump overnight to remove any residual methanol, then settle for 24 hrs then filter and use, there is a little soap/glyc in the bottom of settling tank.
sorry to dig up an old thread... but are you using a 5% pre-wash at all?
i've been trying this with a 5% pre-wash with limited success.
I do no no prewash or post washing, to keep the recovered methanol as pure as possible (it usually tests at 97-98%)
just process as per your normal routine until 3/27 test pass is reached, then go straight into demething, try to maintain 65°C at the still head and recover methanol until the stoichiometric amount is almost reached then stop,(if you need help with working this out I made a guide here http://www.biofuel-uk.net/recoverable%20methanol.doc )
allow 30 - 40 mins for glyc and some soap to settle out then drain it off, pump bio to settling tank and bubble with air to remove any residual methanol then allow remaining soap to settle for at least 24 hrs then decant/siphon/tap off the and filter and use or store.
Has anyone one your side of the pond (or my side) done any GC testing on before/after samples to determine if any reverse reaction has occured? Not to beat a dead horse I just haven't kept up with the the process as well as I should have.
No Gc testing as far as I'm aware and we don't even have Phlip tests over here yet, but it would be nice if someone with access to GC equipment could check it out, as although I've been doing this since March other folks are only recently starting to try this out, my gut feeling from experience is that as long as you don't recover methanol past stoich amount then reversal shouldn't be a real problem, time will tell.
I am in a good position to compare your method with my regular process. My reactor is a 75-liter GL-Eco. I use 22% methanol and KOH processed with base/base.
I am staring a batch later to day. I try to time it so the 1st base can settle over night. It has been in the high 90’s and hot KOH glycerin is to hard to separate. It will be a few days by the time I do the reaction and let it settle fore a while, but I will post the results.
Have you/anyone done a soap test on your bio after it has settled? I am finding that even though I am using the GL system and the soap is dropping out it still tests from 500 to 700 ppm of soap. You may be having better results with KaOH. KOH soap is hard to get rid of, but my oil is too bad for KaOH.
see reply on the main forum in the other WBD thread
|Powered by Social Strata|