Biopowered - vegetable oil and biodiesel forum
Biodiesel => Chemistry and process => Topic started by: dgs on September 29, 2016, 11:38:02 PM
-
I should really make notes so that when I post some thoughts about the process the information seems to flow more logically - but I don't! It just comes from the grey cells (or whatever is left of them) So I will try and make
this as simple as possible.
I water wash quite aggressively with a compressor line at the bottom of the wash tank. After a wash which is usually a max of 20 mins per go I drain the water after settling. The dreaded emulsified mono layer is sometimes hardly noticeable and at other times it is considerable.
To create biodiesel the oil (triglyceride) molecule has to have the three fatty acid chains stripped from the glycerine molecule, which happens in sequence. When one acid chain is left, this monoglyceride molecule will be left in the biodiesel unless it is also converted, leaving the glycerine molecule.
If there is any unreacted oil left in the mix there will be a higher proportion of di and monoglycerides left as well.
According to tests done some years ago by gas chromatograph the methanol phase in our 10/90 tests can hold up to 2% oil even though it is clear, so when we conduct a 10/90 test we don't know if the methanol is free of oil or if it contains the 2% (in which case there will be a higher % of di and monoglycerides present)
A batch of bio that was processed some weeks ago washed really well with hardly any mono's showing. The batch I completed a few days ago was an absolute swine to wash with a considerable emulsified layer showing.
I decided to try some higher concentration variations of the 10/90 test on these 2x batches. Firstly the batch that washed well (low mono's) I first did a 20/80 which was clear, then a 30/70, then a 40/60, all the tests were totally clear with an immediate clear methanol phase.
Now to the hard to wash batch (high mono's) I only had to do the 20/80 test which showed 3.2 mls of dropout. Yes 3.2 mls I could hardly believe it. Let me say at this point that both batches showed a clear 10/90
If the poor batch really had the 2% of oil left in the methanol phase of the 10/90 then increasing the bio amount in relation to methanol would have resulted in some oil coming out of solution and because of the high amount of biodiesel dissolved in the methanol some of this would rather mix with the oil than stay in the methanol. Hence the massive dropout.
So, it seems that after a clear 10/90 if we do a 20/80 it will tell us if some oil still remains in the bio and (more importantly) how well we can expect it to wash.
We cannot use the 20/80 test to calculate the catalyst amount for a re-process and if the 20/80 is a fail and the 10/90 was a pass then how much catalyst do we add. Logically I would say 0.5gms/litre if using potassium. 0.3gms/litre for sodium and 1.0mls/litre for ASM.
Before anyone says 'why should we bother with all that extra complication' I do appreciate having really well converted bio can be said to be over doing it if using older vehicles, also for those who don't water wash the above is not as significant. However, for water washers, passing the 20/80 test will certainly make the washing process easier.
Quick after thought, having 2x dropout tubes for the above tests makes things a whole lot easier
-
If there is space in your measuring deice just add more bio to your 10/90 test to make the 20/80 ratio based on 90 ml of methanol. I think it works out at an extra 12.5ml of bio to add. I do my test in a jar so adding extra after the 10/90 is not a problem.
-
That is a really good idea Bill, so it is like two tests in one, thanks.
-
Me and me mate Gord have been doing this trick for years, we try to out do each other with who can get the most bio into a 10/90 before we get a dropout.
IIRC I got to 40/60ish once.
-
As I have now found Jim, bio and methanol are not miscible in all proportions, regardless of how well the bio is converted.
This is not immediatley obvious when the tube is shaken as the two liquids stay miscible for quite a while but if the bio concentration is more than (as you say) something like the 40% region the liquid will split into 2x phases but it takes time, something like 48 hours.
-
Excellent idea Dave. Well worth doing the 10/90 and 20/80 tests and taking time to get as much conversion as possible prior to washing. My current jelly batch (rescued following your advice) has washed so easily due to it being fully converted. I've had some batches that must have only just passed the 10/90 that have been somewhat of a nightmare to wash.
Will try your test on my next batch.
David
-
I've just tried this exact test so I will report back my findings. ;)
-
I've just tried this exact test so I will report back my findings. ;)
BB, with double the amount of bio in the mix the test is very temperature sensitive, make sure you do it at 20 degs if you can.
-
Absolutely no fall out 2 1/2 hours later albeit the temp taken with my laser is 14.2deg.
Not too sure what the effect of temp +-20 deg will make.
However, I'm sure one of the more experienced amongst us will inform me.
-
The test is harder to pass the colder it gets (10/90) but doing a 20/80 at lower temps could mean any dropout you were seeing would be mostly bio with some methanol, whereas the top phase would be mostly methanol with some bio.
Leave it all night to get colder, but if it has been 2 1/2 hours it is looking good. you have proved it is a very good pass, probably >99% conversion. Min ASTM 14214 is 96.5% min.
-
This is my reply to Element from a thread presently on VOD ;Water washing discovery.
----------------------------------------------------------------------------------------------------------
Hi Element,
I totally understand what you are saying. When I have used acid in the recent past I have done a similar thing, reducing the acid per wash.
If most of the emulsified layer has been removed then the mono's will be greatly reduced, but not removed completely.
Even if a hard wash reveals no mono's there will still be some present. The nature of the reaction is such that they can't be totally avoided.
The EN spec for mono's is 7000ppm (0.7%) the other specs for di's and tri's are tighter than this. As the mono's are the last stage before the bio (FAME) they are the most difficult intermediate to totally get to react.
A friend ex the infopop forum who has become a biopowered member posted a thread which had several graphs of the reaction, (mathmatical models) showing the unreacted intermediates.
From his models if a 10/90 is clear but 'on the cusp' and the methanol phase is holding the 2% of tri's (which is possible even with a clear test) then the mono's are about 2.2% (22000ppm) Three times more than the max spec.
I think the important point to mention here is that having mono's this high would make the bio 'over spec' on bonded glycerine, which is why it is important to remove them.
So as you can see there is no wonder we get the considerable emulsified layer when doing a 'hard' wash, even after a totally clear 10/90
Increasing the bio ratio in the 10/90 seems to work but makes the test very temperature sensitive. A clear 20/80 usually gives a better wash with hardly any mono's showing (but not always) I will keep working on it.
-
Checked 20/80 test again today and still no fallout! ;)
-
Checked 20/80 test again today and still no fallout! ;)
Great news. My last attempt didn't work too well, maybe my batch just wasn't converted enough. Your result has given me new hope.
-
I am presently processing a small experimental batch to test the 20/80 test further. The testing will also include a different 10/90.
To Explain, today I added 12 mls of new rapeseed oil and 240mls of virgin methanol into a 250ml separating tube. I gave it a really good shake for a few mins and left it to settle. Tomorrow I will run off the oil and the methanol will be used for one of the 10/90 tests.
The idea is that as the methanol is (presumably) now saturated with oil, using it for the 10/90 test should give a more accurate drop out.
I will do 3x dropout tests on the batch tomorrow after S2 (I have 3x dropout tubes)
One will be a 'normal' 10/90, One will be a 10/90 with the tri saturated methanol and the third one a 20/80
i will report when the tests are completed.
-
Sounds interesting, Dave. Looking forward to the results.
We did have Jan Warnqvist sign up on here a little while back so I guess you could contact him via pm (which I assume will result in an email notification) if advise is needed.
-
Thanks Julian.
I think Drum was in contact with him before he posted the results of his models. It is a possibility.
-
I tried this 20/80 test today.
From the current batch I had a 3/27 test that was clear with no drop out and added 6ml of bio. After a good shake the result was cloudy but cleared within a couple of hours to reveal about a third of one graduation drop out, As my graduations are 0.05ml I guess that's 0.01666ml.
So I'm guessing that's a reasonable result.
EDIT ... I should say shed ambient is 25o C at present.
-
After S2 with todays test batch I did a conventional 10/90 which showed no dropout. A 20/80 showed 0.8mls dropout and a 10/90 with the triglyceride saturated (tri sat) methanol showed 0.4mls.
I reprocessed on the basis of the 0.4mls after which a 20/80 and a 10/90 with the tri sat methanol were both clear.
Just shows, if I had carried on with the batch on the findings of the conventional 10/90 not only would I have had some tri's left in the mix but also a considerable amount of mono's (more than enough for several portions of Christmas pudding when emulsified)
Hopefully the batch is now converted to a point where the mono's won't show up during washing, we will see.
To make the tri sat methanol up I think the best way is the pop bottle method. just pour a few mls of new oil into the clean & dry bottle, top up with virgin methanol, shake and leave overnight for the methanol to clear.
When most of the methanol has been used, just top up with methanol and repeat. Keep checking that there is still a few mls of oil in the bottom.
-
I tried this 20/80 test today.
From the current batch I had a 3/27 test that was clear with no drop out and added 6ml of bio. After a good shake the result was cloudy but cleared within a couple of hours to reveal about a third of one graduation drop out, As my graduations are 0.05ml I guess that's 0.01666ml.
So I'm guessing that's a reasonable result.
EDIT ... I should say shed ambient is 25o C at present.
Not bad Julian,
Would be interesting to see what effect a cold night in your shed has on the test. Also a test with tri sat methanol would be interesting.
-
its looking like you've found a way to test the 100% pass Without going for the overdose method. It will be interesting to see how well it washes, also if the test is consistent over several batches.
Some have suggested that it could be unprocessed monos that are falling out of my good pass bio at the minute, even though my oil is clear before being processed. I will be brewing at the weekend so I will try the saturated methanol test and see what happens. I haven't washed for the past 14 months but I want to try and limit the waxing I get
-
Thanks Manfred. Yes let us know how you get on. If you indeed do have a lot of mono's do they not show up at the interface of a 50/50
-
I tried this 20/80 test today.
From the current batch I had a 3/27 test that was clear with no drop out and added 6ml of bio. After a good shake the result was cloudy but cleared within a couple of hours to reveal about a third of one graduation drop out, As my graduations are 0.05ml I guess that's 0.01666ml.
So I'm guessing that's a reasonable result.
EDIT ... I should say shed ambient is 25o C at present.
Not bad Julian,
Would be interesting to see what effect a cold night in your shed has on the test. Also a test with tri sat methanol would be interesting.
I'll have a look in the morning and post. Next batch I'll try a normal 3/27 and your tri sat methanol idea.
Just playing the devils advocate here (I have a habit of querying things without fully understanding them) ... if the 1/4 test is a more accurate indication of conversion why would Jan have suggested suggest a 3/27 test in the first place?
-
Thanks Manfred. Yes let us know how you get on. If you indeed do have a lot of mono's do they not show up at the interface of a 50/50
Sorry Dave it's been a while since I've done a 50/50. If the bio looks suspect after a dry wash I do a soap test on it. But now you say it I will do one on my last batch.
-
I would have to speak to him to answer that question. It is early days but I would say from initial testing the 10/90 tri sat method is an advantage in that chemicals for the next stage can be calculated from the dropout amount. If the 20/80 has dropout it tells us there is incomplete conversion but to add more chemicals from the result would be guesswork.
Any 1 to 9 ratio won't be as accurate when mixed with 'normal' methanol as the 2% of sample size (triglycerides) are absorbed into the methanol. The test only works well in an underconverted sample as long as it shows the correct amount of tri's as dropout, then some of the bio mixes with these tri's to give extra dropout which just happens to (roughly) equate to the di's and mono's left in the mix. (methanol phase)
-
This morning shed ambient was just under 20°C and the the test I did last night still showed the same drop out but the methanol was slightly cloudy.
Just wondering how accurate my maths was ... I think adding 6ml to a 3/27 test may have been too much. Can someone check my sums! Got a feeling I should have added 3.75ml. If so I'll redo the test this evening.
-
Yes a 20/80 on 27ml of meth is 6.75ml of bio total. So 3.75 would be he correct amount
-
Am I missing something here?
The 20/80 scaled down to a total of 30mls (3/27) is actually a 6/24.
-
Just got back from a very long lunch (circa 8hrs!) with a group of friends, a couple of whom are very smart and all of whom are way smarter than me and this problem was discussed with differing views for about an hour.
I don't think anyone came up with an answer of which they were 100% certain.
-
Am I missing something here?
The 20/80 scaled down to a total of 30mls (3/27) is actually a 6/24.
You are correct. But as Julian already had 27mls of methanol he needed 25% of that making 6.75mls.
-
I'll have a go at this with a clear pass sample tomorrow.
-
I see Manfred.
Chug, please let us know the outcome.
As both the 20/80 and the 10/90 with the tri sat methanol are both more saturated with tri's,di's, mono's, fame, they are a lot more temperature sensitive. I have noticed a small amount of dropout today in both my tests, the overnight temp was about 10 degs here last night and the tubes were left in my bio room away from the house in an outbuilding.
Yesterday evening both tests were clear of dropout when I locked up and I checked the temperature which was 15degs. As they were clear I am not worried by this, I think the big test is to see how the batch washes, which I will do on Sunday.
-
Sums never were my strong point, but I suspect the 3.75ml extra for the 3/27 test is correct as Manfred concurs.
The explanation I can just get my head around, but still a little unsure why he total volume is taken as the ratio, is as follows ...
I'm aiming for 20/80 which is (20ml bio in a total of 100ml) ... 100 / 20 = 5
If I add 3.75ml to a 3/27 test (3ml bio in a total of 30ml) I get 6.75ml in a total of 33.75ml. 33.75 / 6.75 = 5
Does that explanation make sense to anyone else? If so it should improve the results of the test I did quite dramatically.
-
After reading twice I understand it. More simple is a ratio of 4:1. So 27 meth /4 bio = 6.75. Minus the added 3 bio = 3.75 bio to be added.
Dave has altered the test from a 9:1 to a 4:1 so all we do is add 25% bio to any quantity of meth. The old 5/45 is now a 10/40.
When you look at it from this angle he is almost halving the methanol in the test. Jan's original 25/225 was to test a pass using titration to euro standard. Now we use it to minimise the chem quantity and say that it is not giving us the whole story, something that Jan never intended, as we would be using more chem with titration so probably getting the good pass without messing about.
Please don't think I'm knocking Dave's testing or results as I will try it out myself it's just my thinking why Jan went 9:1
Maybe there's still a place for titration afterall.
Thinking more on this after my last 100L batch used over 800g of NaOH.so I did a titration. It came out at 3. To get a good second stage pass I use a base of 6 so with titration I would have used 900g of NaOH. That's only 100 more and maybe a better conversion with it.
-
I read my last post again this morning and I can't work out what the hell the 2520 figure was meant to be so I've taken it out of that post!
=================================
OK, so I redid the test this morning. Shed ambient was 15°C. 3/27 was clear. 6.75/27 was cloudy but no drop out.
I brought the test indoors which, conveniently was 20°C and it cleared within a couple of hours with no dropout.
-
I washed the batch today, and the result--------------------------------------------TOTAL FAILURE.
It produced the same amount of emulsified mono's as the previous batch. So, just some thoughts and observations.
Based on my 200 litre batches the emulsified layer is about 4 litres. When split the actual volume of mono's is about 700mls. I use a small amount of acid to split it and have noticed a considerable increase in colour of the mono's when split (as opposed to letting it split naturally)
I am therefore wondering if the increase in colour when using acid in the whole batch is not an increase in the colour of the bio, but of the mono's when they are left in the bio.
Going back a few years I went through a period where I had no (or virtually no) mono's in the batch. At the time I was doing what I called the Maurice Mynah Method. M M replied with his method to one of my posts (can't remember if it was VOD or infopop) anyway, he basically glyc washes then uses 7 gms K/litre and processes for 4 hours at 60degs. He usually gets full conversion with this method but the process time is long.
I never managed the 7 gms, but close. However there was never any real emulsified layer in the bio when washing. Does this mean the key to reducing the mono's is more process time with temperature related?
-
I thought I had the monopoly on failures round here!
All the same it's not really a failure ... just an avenue of experimentation that returned a negative result eliminating the necessity to probe further.
Keep up the good work, it's only by chaps playing in their sheds (or poly-tunnels) that we gain any knowledge.
-
I know that failures can sometimes tell us as much as successes. I thought I had this one cracked. Even the test that was clear with the tri sat methanol is now showing about 0.2 mls dropout after a few days. It is maybe that the batch just wasn't converted enough but with a delayed action test it's a bit too late. I will keep trying.