BIODIESEL & SVO DISCUSSION FORUMS





Sponsors    Biodiesel and SVO Forums Home    Forums  Hop To Forum Categories  Biodiesel  Hop To Forums  Making Biodiesel    The Warnqvist Test and re-processing

Moderators: Shaun, The Trouts
Go
New
Find
Notify
Tools
Reply
  
The Warnqvist Test and re-processing
 Login/Join
 
Member
posted
The Warnqvist Test and re-processing

Over the last several years, a procedure has gained in popularity which uses the dropout amount from a warnqvest test to calculate the amount of chemicals required to reprocess or perform the second stage of a 2 stage reaction.
This method involves measuring the percent of dropout in the warnqvist test and then treating this as the percent of oil with a zero titration remaining from the initial reaction to be reprocessed.

An Example
You have a 40 litre batch of oil that you react and find that when you then perform a 10/90 version of the Warnqvest test you have 3ml of dropout.
3ml dropout is 30% of the 10ml warnqvist test sample, so you then calculate the un-reacted oil remaining to be reacted in the second stage as 30% of 40 litres.
30% of 40 litres is 12 litres.
Because all of the FFA's have been converted into soap in the first stage, you treat this remaining 12 litres of oil to be reacted as if it were new oil with a zero titration.

When reacting oil with a zero titration in a single stage, most people will use 5g NaOH/ 8g KOH mixed into 200ml methanol for every litre of oil being reacted.
In the above example, you have calculated that there is 12 litres of oil remaining to be reacted.
The KOH amount required in the reprocess is 8g X 12= 96g KOH
The methanol amount required in the reprocess is 200ml X 12= 2400ml Methanol

This method has been proven to be quite reliable in performing a second stage reaction/ reprocess that does produce biodiesel that subsequently passes the Warnqvist test.

Recent information posted to another thread on this forum shows there is NOT a 1 to 1 correlation between dropout and conversion in the Warnqvist test.

People have asked why the above procedure works so well at reprocessing if there is not a 1 to 1 correlation between dropout and unreacted oil.

Some preliminary information
Most people want to make very high conversion biodiesel that will pass the Warnqvest test while spending as little time and money as possible.
Testing has shown that the formula of 5g NaOH/ 8g KOH and 200ml methanol is about the least amount of chemicals you can use to react a litre of zero titration oil in a single stage that will probably pass the Warnqvest test assuming you have a properly designed reactor.
If you use more chemicals than 5g NaOH/ 8g KOH and 200ml methanol, this will give you even higher conversion and a greater probibility of passing the Warnqvest test.

For instance, a few months ago I made a litre of biodiesel from new oil using about 16g KOH and 250ml methanol.
This turned out to be extremely high conversion biodiesel, but it also cost more to produce. I would not use this formula for my daily driving biodiesel.

Why the above procedure works so well if there is not a 1 to 1 correlation

Recent testing has shown that the Warnqvist test actually has more dropout than would be the case if there were a 1 to 1 correlation, at least in the range of conversions we are normally testing.
In other words, the above procedure overstates the amount of oil remaining to be reacted and as a consequence, if you use the above procedure, you end up using more than 5g NaOH/8g KOH and 200ml methanol for every litre of oil remaining to be reacted.
This results in a very high likelihood that you will pass the warnqvist test after reprocessing using this procedure.

This message has been edited. Last edited by: Tilly,






 
Location: ลึก ประเทศอินเดีย | Registered: March 03, 2001Reply With QuoteReport This Post
Member
posted Hide Post
I express no opinion about the accuracy of the information posted; frankly, I have reached the point of no longer knowing or caring -- a point which corresponds to my inability to remember when I last (or even if I ever) had to run a re-process. Sure feels like one more way to beat the "no 1-to-1 correlation" horse -- which I thought was dead and buried already.

Cheers, John
 
Registered: June 17, 2007Reply With QuoteReport This Post
Member
posted Hide Post
Hi Dukegrad98,
quote:
Originally posted by dukegrad98:
I express no opinion about the accuracy of the information posted; frankly, I have reached the point of no longer knowing or caring --
Thank you for advising us of your lack of knowledge or interest in the subject of this thread.


quote:
a point which corresponds to my inability to remember when I last (or even if I ever) had to run a re-process.
Do you mean that you always pass warnqvest when you make biodiesel or do you do like me and not bother to check it.






 
Location: ลึก ประเทศอินเดีย | Registered: March 03, 2001Reply With QuoteReport This Post
Member
posted Hide Post
quote:
Originally posted by Tilly:
The Warnqvest Test and re-processing

Over the last several years, a procedure has gained in popularity which uses the dropout amount from a warnqvest test to calculate the amount of chemicals required to reprocess or perform the second stage of a 2 stage reaction.
This method involves measuring the percent of dropout in the warnqvest test and then treating this as the percent of oil with a zero titration remaining from the initial reaction to be reprocessed.

An Example
You have a 40 litre batch of oil that you react and find that when you then perform a 10/90 version of the Warnqvest test you have 3ml of dropout.
3ml dropout is 30% of the 10ml warnqvest test sample, so you then calculate the un-reacted oil remaining to be reacted in the second stage as 30% of 40 litres.
30% of 40 litres is 12 litres.
Because all of the FFA's have been converted into soap in the first stage, you treat this remaining 12 litres of oil to be reacted as if it were new oil with a zero titration.

When reacting oil with a zero titration in a single stage, most people will use 5g NaOH/ 8g KOH mixed into 200ml methanol for every litre of oil being reacted.
In the above example, you have calculated that there is 12 litres of oil remaining to be reacted.
The KOH amount required in the reprocess is 8g X 12= 96g KOH
The methanol amount required in the reprocess is 200ml X 12= 2400ml Methanol

This method has been proven to be quite reliable in performing a second stage reaction/ reprocess that does produce biodiesel that subsequently passes the Warnqvest test.

Recent information posted to another thread on this forum shows there is NOT a 1 to 1 correlation between dropout and conversion in the Warnqvest test.

People have asked why the above procedure works so well at reprocessing if there is not a 1 to 1 correlation between dropout and unreacted oil.

Some preliminary information
Most people want to make very high conversion biodiesel that will pass the Warnqvest test while spending as little time and money as possible.
Testing has shown that the formula of 5g NaOH/ 8g KOH and 200ml methanol is about the least amount of chemicals you can use to react a litre of zero titration oil in a single stage that will probably pass the Warnqvest test assuming you have a properly designed reactor.
If you use more chemicals than 5g NaOH/ 8g KOH and 200ml methanol, this will give you even higher conversion and a greater probibility of passing the Warnqvest test.

For instance, a few months ago I made a litre of biodiesel from new oil using about 16g KOH and 250ml methanol.
This turned out to be extremely high conversion biodiesel, but it also cost more to produce. I would not use this formula for my daily driving biodiesel.

Why the above procedure works so well if there is not a 1 to 1 correlation

Recent testing has shown that the Warnqvest test actually has more dropout than would be the case if there were a 1 to 1 correlation, at least in the range of conversions we are normally testing.
In other words, the above procedure overstates the amount of oil remaining to be reacted and as a consequence, if you use the above procedure, you end up using more than 5g NaOH/8g KOH and 200ml methanol for every litre of oil remaining to be reacted.
This results in a very high likelihood that you will pass the warnqvest test after reprocessing using this procedure.


Tilly - you might have the decency to at least spell Jan Warnqvist's name correctly!

You might also do the general scientific community the favour of not using woolly terms such as 1 to 1 correlation. This term has no formal meaning and is at best slang for what I think you mean perfect correlation! If not tell me what you actually do mean!

Why would anyone ever have thought there is perfect correlation between the test dropout and the biodiesl from which it came. The distribution of glycerides in the dropout could not possibly be perfectly correlated with the biodiesel sample that produced the dropout for the very simple, basic and obvious reason that triglycerides, diglycerides and monoglycerides do not have identical solubility in methanol which would be a first requirment for perfect correlation.

And would you like to elaborate a little more on precisely how and why the test overstates the amount of oil remaining to be processed? As I see it there are at least four factors at play


  • The dropout potentially contains material other than try/di/monoglycrides (hence overstating)
  • not all the glycerides are in the dropout as some are methanol soluble (hence understating)
  • cannot distinguish between glycerdies in the dropout therefore we effectively assume they are all triglycerdies which will require more conversion than the true multi-mix of glycerides that are actually there (hence overstating)
  • 8g KOH per litre not a minimum (hence effective overstating)


Your extensive testing (alluded to by Jon) might help you in any response. Just point me to where you have written up (or published) the output of your testing or better still, present it in this thread where it would have an immediate and effective impact and would be scientific good practice.

All of the above would help you in your pursuit of REAL SCIENCE.


2001 Saab 93 TiD B50
2006 Mercedes C200 CDI (no DPF) B50 (for now)
 
Location: Stroke County | Registered: July 01, 2014Reply With QuoteReport This Post
Member
posted Hide Post
quote:
Originally posted by Tilly:
Do you mean that you always pass warnqvest when you make biodiesel or do you do like me and not bother to check it.


Can't remember that I ever had a batch which didn't pass the 3/27 test on first process. Starting with dry oil and titrating carefully just isn't that hard, and always led to good results for me. I agree that a bit of underprocessed oil has probably never been the root cause of an engine problem, but I also admit that I was more careful with my quality testing when running high blends (up to B100) in expensive engines with common rail and direct injection. An old inline Benz will burn nearly anything, but I was a bit more careful what I put through my VW V10TDI.

With D2 currently at ~$2.30/gal (US) around here, and given the value of and demands upon my time, I've got a good 500-700 gallons of oil at my ranch which won't see processing anytime soon. One of these days I should pause to train one of the staff on how to do it...but again, time...

Cheers, John
 
Registered: June 17, 2007Reply With QuoteReport This Post
Member
posted Hide Post
 
Location: Wellington County, Ontario Canada | Registered: February 07, 2008Reply With QuoteReport This Post



Member
posted Hide Post
quote:
Starting with dry oil and titrating carefully just isn't that hard, and always led to good results for me. I agree that a bit of underprocessed oil has probably never been the root cause of an engine problem,

Ditto!
For the life of me I just cant understand why all the mucking about with 2 stage, NTM, etc just to save a few dollars in caustic. Get a good pump and well designed reactor and start making quality fuel in a single stage that passes a simple 3/27 test and then use it in your diesel.
I have over 100,000km running B100 in a common rail 2007 Duramax done in a single stage with conversion verified via 3/27 and with no water washing, it just works and really doesn't need fixing in my opinion...
Cheers,
Jon
 
Location: Wellington County, Ontario Canada | Registered: February 07, 2008Reply With QuoteReport This Post
  Powered by Social Strata  
 

Sponsors    Biodiesel and SVO Forums Home    Forums  Hop To Forum Categories  Biodiesel  Hop To Forums  Making Biodiesel    The Warnqvist Test and re-processing

© Maui Green Energy 2000 - 2014