you mean that no one has tried this on the fullsize batch???

I'm the first...oops I wish I had read the post closer. I might have used less KOH..

too late!! or time will tell. in about 1.5hrs

Stats
oil - titrated at 3 KOH(90% pure). 56gal(212L)
methanol - 8 gallons. stage 1
KOH 10g + titration * 1.15 = 13.5g/L
total of 2862g
process temp 130F

stage 2
1.5 gal of methanol and 275g of KOH

if found that I need to 10g + titration and 23%methanol to pass the 27/3 test. it looks like I should have use 8.9g + titration * 1.15 for stage 1.

-dkenny

at the moment stage 1 is still mixing

dkenny,

No, you are not the first to try it. One person did post, but it did not pass.

So far as I know, it has only worked for neutral and tilly and only in test batches.

Look forward to your results.

One thing you may want to try - take samples at various times in processing, particularly during stage 2 so you can see at what point it passes, or stops improving.

Too much KOH should only help conversion, and your methanol is almost 17%.

What type of mixing does your set up have?

How long will you run each stage?

Andrew

Results so far...

setup:
80gal waterheater with 1.5" Northerntool pump
methoxide input rate adjust to take about 30 minutes for 13 gallons.

stage 1 mixed for 1 hrs after methoxide was added
settled for about 45 minutes. drained off 8 gallons of glycerine.

stage 2 mixed for about 1 hrs after methoxide was added.
transfered to settling tank
after settling for about 30 minutes I tried the 27/3 test. failed, but not by much.

tomorrow after settling all night I'll retest.

before I needed to use the same amount of KOH and 22% methanol to get about same results.

so far..this would save me about $0.12/gallon based on methanol cost alone. not bad

-dkenny
EDIT 11/21/2006
I retested but it still fails. darn..
I got a total of 14gal of glycerine out of this batch of oil. this is much higher than normal. I think its fine food stuff in the oil. the yields from other batches using this oil are down as well.

Hello Andrew
I missed the post of the failed attempt. Can you give a link?

quote:

Originally posted by Andrew Morris:
dkenny,
No, you are not the first to try it. One person did post, but it did not pass.
Andrew

Tilly,

Here is the link:

he uses 16.7% methanol though. Nice detail in his post.

Andrew

dkenny

The 2-stage process with methanol test works well because you can test the product in the second stage and if not good enough add a bit more KOH/methanol until it is right.

Neutral,
thanks next batch I try that. What you recommend that I change? Does it matter that I couldn't get all the glycerine out before the 2nd stage?

thanks for the help

-dkenny

I don't think a little glycerine left from stage 1 will matter.

I am coming round to the view that, as we are not trying to push the reaction hard in the first stage we should use a bit less caustic. I really don't see why you couldn't drop back to the old 3.5 g NaOH plus titration. And perhaps use straight titration, not titration times 1.15. I suspect this will make a little less soap and give a cleaner separation for the first stage. Using KOH instead of NaOH also helps if using low methanol as the glycerine will be more fluid.

I just got through using neutrals base/base procedure and came up with some pretty promising results. I did a control by performing TLC's (Thin Layer Chromatography) on normal base transesterification technique that has become commonplace in the biodiesel community. I compared those plates with the plates I made after using neutrals base/base procedure.

FIY, my oil titrates at 6 using NaOH. I use a standard hotplate/magnetic stirbar. I only have a 1L Erlenmeyer Flask so I do 500 mL batches. My solvent system for my TLC plates are 90% Hexanes/ 10% Ethyl Acetate with a couple drops of acetic acid. I stain my plates with Iodine. Now for the good stuff.

The control was done using 22% Methanol, ((6g + 4g)/2) grams of NaOH, and 500ml of WVO which was reacted for 30 minutes at 56 degrees Celsius. The reaction was complete by TLC so I the separated the glycerine off the bottom. I then dry washed the biodiesel using GL's dry wash technique. Basically removing the methanol and water (shouldnt be any) which crashes out the fats and soaps which then are filtered off through a 50,10 micron bag filter. Yield was 470 mL or 94% yield

For the base/base procedure I used 500ml of WVO. For the first stage I used ((4.2g + ( 6g x 1.15) ) / 2 ) which equals to 5.6 grams of NaOH (remember I do 500mL batches so I have to divide by 2). I added the NaOH with 70 mL of methanol. I reacted the WVO and the methanol/hydroxide mixture for 30 minutes at 56 degrees celsius. I TLC'd the reaction mixture and separated. The reaction was not complete as you can see below in the attatchment (note scanner is broke so I used MSpaint to draw the TLC plates to the best of my ability.) I noticed a soap emulsion layer in the phase cut. I drained the glycerine and soap layer off and started the second stage. I used 0.4 grams of NaOH in 11.5 mL of methanol. I reacted at 56 degrees celsius for 30 minutes. I TLC'd the reaction and the reaction was complete or nearly almost complete (TLC's are judgemental by nature especially in reversible reactions like this where reactions are almost never 100% complete). I then cut the glycerine layer off and used the GL dry wash method described above. Yield was 390 mL or 78% yield.

I think this is a promising result and needs to be experimented further. I am dissapointed with the yield but I only ran one batch and could of been my error in not letting it cut long enough or whatever. I will run futher tests and post my findings.

NOTE: HAVING TROUBLE UPLOADING THE PICTURE. SORRY...DOES ANYONE HAVE ANY SUGGESTIONS

I suspect your yield will be higher if you cut back the NaOH in stage 1, as suggested above. I believe it will reduce the wasteful conversion to soap as the caustic cannot do its proper work if the methanol is in short supply, so does things we don't want. You should also get a cleaner separation so will not risk losing oil in the glycerine. It is also possible that it will turn out that using KOH works better with low methanol.

Just a quick up-date
Latest teletype message from mickey.

Mickey made a 50 litre batch yestereday from WVO titrating 2.5.
The oil was about 60% tallow and 40% palm oil.
The batch failed the Warnqvest Test.
Mickey then reprocessed using another 17ml methanol and 1.5g KOH/ litre WVO.
This time it passed Clear and bright immediatly. NO precipitation.
This was a total of 180ml methanol per litre WVO

As there is some thought that Tallow may give false failures with the Warnqvest test, Mickey says he will try a few test batches using Veggie oil and if these work out he will try another 50 litre batch.

Tilly,

Thanks for your report.

Yes, the degree of saturation of oil may have an effect.

Nice job on the 18%. How and how long did you mix your 13 gallon batch?

Andrew

Hello Andrew

Mickey's reactor is a 75 litre Black HDPE drum with a widemouth screw on top.
Mixing is accomplished with a drill and plaster mixer. The "Nose" of the drill fits snuggly through a hole in the top.
An initial 1-2 minute hard mix followed by further 1 minute Hard mixes every 10- 15 minutes.
Heating is solar. At this time of year the oil is between 50- 55deg C by Noon. But it is not summer either.
Very simple. Very effective.

Mickey said The First stage was mixed for about 2 hours.
45 minutes for the glycerine to settle and then drained.
Second Stage 2 hours mix.
Allowed to sit for 45 minutes.
Failed warnqvest.
Allowed to sit a further 45 minutes and still failed warnqvist.
Drained glycerine (Not Much)
Mixed 3rd stage for an hour and a half.
Waited over night.

Easily passed Warnqvest the next morning on first test.

Tilly,

Thanks for the detailed report.

Andrew

quote:

Originally posted by Andrew Morris:

Yes, the degree of saturation of oil may have an effect.

Andrew, You could be right. As far as I know, no one has looked seriously at how hydrogenated oils fare in the Warnqvest Test.

Tilly, good to see Mickey investigating the Warnqvest test again. (I hope I'm reading your post right) There have been a number of false failures or odd results that have been reported that haven't been fully investigated. We know more about the test now than we did when you initially introduced the test (thanks to all the people that have tried it and reported their results), but there is still a lot more testing and investigation left to go.

Tilly,
Can your last two posts be edited together like below to include the mixing details:

-----
Heating is solar. At this time of year the oil is between 50- 55deg C by Noon.

Mickey made a 50 litre batch yesterday from WVO titrating 2.5.
The oil was about 60% tallow and 40% palm oil.
Mickey said The First stage was mixed for about 2 hours.
45 minutes for the glycerine to settle and then drained.

The batch failed the Warnqvest Test.

Second Stage 2 hours mix.
Allowed to sit for 45 minutes.
Failed warnqvest.
Allowed to sit a further 45 minutes and still failed warnqvist.
Drained glycerine (Not Much)

Mickey then reprocessed using another 17ml methanol and 1.5g KOH/ litre WVO.
Mixed 3rd stage for an hour and a half.
Waited over night.

Easily passed Warnqvest the next morning on first test.
This time it passed Clear and bright immediatly. NO precipitation.
This was a total of 180ml methanol per litre WVO
-----
George

Hello George

The Warnqvest test was only performed after the second stage and third stage.
It failed after the second stage but passed after the third stage.

steved,

When the test was introduced, it came with a disclaimer about not working with BD from fish oil. Some people get odd results depending on temp. TG is not totally unsoluable in methanol. So, I suspect the type of oil is one factor, along with temp. that will effect the percent of TG which disolves in methanol and thus skews this test. How much? Don't know yet. After Thanks GIving I have a simple test to do, but wish I could GC some stuff.

Andrew

Further updates

Mickey did two 1 litre test batches with Palm WVO titrating 2.5
One batch was performed as per the original formula and the other batch was performed as per Neutrals new Idea which includes an increase in the methanol in the second stage.
Stage 1- 140ml methanol 4.9g KOH + (Titration X 1.2)
Stage 2- 30ml methanol 2.1g KOH

Both batches passed the Warnqvest test, however the second batch was cleared and bright immediatly while the first batch took a few seconds to clear.

Mickey decided to do a 50 litre batch using the new formula and see how it turns out.

More info as it comes to hand.

quote:

Originally posted by steved:
So what is it ? The warnqvist test does not work the same on different kinds of oil or the 2 stage doesn't work the same on different kinds of oil ? Or both ?

RickDaTech, you have a GC, right ? What's going on ?

steved, It's a little of both. With different oils you would have to adjust your recipe a little to make the two stage work, but you'd expect that even with the single stage method.

From time to time I have access to a GC. I may get one next year. We're moving to a new location before the end of the year and all my testing and experiments have been put on hold while we're preparing for the move.