Biodiesel and SVO Forums Home
27/3 how soon?

This topic can be found at:
http://biodiesel.infopop.cc/eve/forums/a/tpc/f/9411061471/m/2927024973

September 29, 2014, 12:53 PM
fuelfarmer
27/3 how soon?
How soon can you do a 27/3, or any of the other multiples, after drawing a sample? Does glycerin cause an error if it is not settled?
September 29, 2014, 01:43 PM
fabricator
I just take a little jam jar and fill it up, it takes about 5-10 minutes for the glyc to settle, then I take my sample.
September 29, 2014, 02:28 PM
Jon Heron
I let it settle till there is a defined line between the layers, about 15 or 20 minutes.
The glycerine dissolves readily into the methanol but if it has not settled at all you have to wonder how close to 3ml of actual biodiesel you have compared to the glycerine thats in solution..
Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 29, 2014, 09:48 PM
fuelfarmer
When doing a no titration two stage process it would be nice to pull the sample and do the 27/3 right away so the drop up could settle while the glycerin settles in the processor. That would save a little time when starting the second stage. I might had to do some experimenting.
September 29, 2014, 10:50 PM
Jon Heron
Maybe one of those hand crank fuges would work to speed it up?
http://www.ebay.ca/itm/Lab-Han...4-10ML-/321272484092
Or you could likely roll your own pretty easy too...
Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 30, 2014, 08:30 AM
Dgs
fuelfarmer, yes thats correct.
It would be nice to do a dropout and have it settle while the batch is settling.
Trouble is with all the glycerol in there (usually @ 10%) only 90% of your sample is crude bio.

The other very important point is that if you take the sample too soon it will take ages to settle due to all the contaminants in the crude bio. When I say ages I mean several hours,which when doing a sample "midstream" is totally impractical.
The optimum time to leave it to settle is about 30 mins.

A sample taken after 30 mins will settle in about half the time compared with one taken after 10 mins.

It all depends why you are doing the dropout. If like myself (I do 10/90) you need the dropout to calculate the chemical for the 2nd reaction this time delay presents a problem.

I want to get an accurate result but not wait hours for it.
So i leave it to settle 30 mins,do the test then from how cloudy the top methanol phase is, estimate how much more dropout there is to fall.

Doing a test on finished bio is easy and you can get an accurate result in 15 mins to a couple of hours.
If you take a sample of "crude" too early it can take up to 12 hours for the methanol phase to be clear.

It's a wonderful test but needs to be "tweeked" to use it to advantage.

Look at the "3/27 and beyond "thread.
September 30, 2014, 04:10 PM
Jon Heron
Excuse my ignorance on this method, but why not just titrate?
What is the perceived advantage of this no tiration thing? It just seems like more work/time and fiddling with very little reward, but I am probably missing something along the way there... Smile
Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 30, 2014, 04:51 PM
john galt
quote:
Originally posted by Jon Heron:
... why not just titrate?


Because titrating is fiddly, takes longer, requires continually fresh test reagents, and can be inaccurate.
A simple 10/90 test is so much easier than titrating.
It's the difference between seeing the amount of conversion instead of taking a calculated guess at how much reagent is required.

Both methods work, it's largely a mater of personal preference. Some find one method better than the other.



September 30, 2014, 04:55 PM
Dgs
I titrated for years and found the method inaccurate.

If a second stage is required it is far more precise to calculate the amount of chemical required as a direct result of the dropout.

If a process has been taken to completion in one stage the water produced by FFA neutralisation has been present for all the reaction and therefore produced more soap as a result of the saponification reaction,making the single stage process less efficient.
September 30, 2014, 05:09 PM
Jon Heron
quote:
Originally posted by john galt:
quote:
Originally posted by Jon Heron:
... why not just titrate?


Because titrating is fiddly, takes longer, requires continually fresh test reagents, and can be inaccurate.
A simple 10/90 test is so much easier than titrating.
It's the difference between seeing the amount of conversion instead of taking a calculated guess at how much reagent is required.

Both methods work, it's largely a mater of personal preference. Some find one method better than the other.

Interesting, what is fiddly about titrating? I find it to be the opposite, titrating is simple consistent and accurate with no fiddling involved. Titrating has never let me down.
Your assertion that titrating takes longer is ridiculous if you need a second stage for the no titration technique, never mind the issues with settling described by FF and DGS above.
I guess it seems obvious its a matter of personal preference then with no real advantage?
Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 30, 2014, 05:12 PM
Jon Heron
DGS, interesting, has anybody done any tests to compare the efficiency of single stage vs 2 stage processing?
I would be interested in the results if so.
I have never had an issue with accuracy and titrate with turmeric.
I have only ever done single stage as it works great for me and takes much less of my precious time, therefore I have no direct experience with two stage...

Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 30, 2014, 05:23 PM
Tilly
Hi Jon,

I performed some fairly extensive testing using the latest "Titrationless" method that is so much the current rage and found that it suffers severe limitations.

The other problem is that the drop-out in the Warnqvest test does not tell you the amount of un-rected oil remaining. It overstates the amount of unconverted oil remaining so you end up using excess chemicals in the second stage.

I will post the test series I performed with the titrationless method later.






September 30, 2014, 05:41 PM
Jon Heron
Thanks Tilly. That would confirm my suspicions.
I just cant see in what way this method could be more efficient or faster, particularly if a second stage is required. But then again I have never found titrating to be fiddly, inaccurate nor time consuming...
FF, sorry to hijack your thread!
Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 30, 2014, 05:56 PM
Jon Heron
I am wondering how any of you folks complaining of inaccurate titrations are performing them?
I use 99% iso and turmeric for the titration with an automatic burret to measure the titration amount and a small 200ml beaker on a stir plate for the mixing and titration.
To mix the solution I measure 1L of RO water into an erlenmeyer flask and then use an accurate scale to measure in 1 gram of the caustic of choice. I refresh the solution every few months whether it needs it or not.
This has never let me down nor shown any signs of being inaccurate.
Here is a pic of my titrating rig;

Cheers,
Jon

This message has been edited. Last edited by: Jon Heron,


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 30, 2014, 07:13 PM
Dgs
Jon,are you saying that with all the single stage processes you have done that you have never had any dropout when conducting a warnqvist test. Or do we take it that you don't do the test.
September 30, 2014, 07:23 PM
john galt
quote:
I use 99% iso and turmeric for the titration with an automatic burret to measure the titration amount and a small 200ml beaker on a stir plate for the mixing and titration.
To mix the solution I measure 1L of RO water into an erlenmeyer flask and then use an accurate scale to measure in 1 gram of the caustic of choice. I refresh the solution every few months whether it needs it or not.

The 10/90 test I use is much simpler than that and only requires methanol, a 12mm syringe, and a simple tester that Imakebiodiesel posted a while back made from a bottle and a 3mm syringe. It's accurate enough to produce good quality biodiesel, and that's all I'm interested in.

Yes it's a personal choice and I've stated my reasons without personal comments like "Your assertion... is ridiculous" because that would be rude and it adds nothing useful to the discussion.



September 30, 2014, 07:53 PM
Jon Heron
quote:
Originally posted by Dgs:
Jon,are you saying that with all the single stage processes you have done that you have never had any dropout when conducting a warnqvist test. Or do we take it that you don't do the test.

Hi DGS, yes pretty much. One time I had a fail with my old processor and attempted a reprocess that ended up with a tank full of gel... I documented the mess I made HERE with pics and all, in the end I managed to save it.
In retrospect, I would have never even attempted to reprocess it as a bit of dropout will not harm anything.
With my current processor, knock on wood, I have never had a 3/27 fail, I preform it carefully and religiously and when it passes (after about 20 minutes of processing) I start the recovery (I do WBD exclusively).
I have no secret recipe, I dewater under vacuum, titrate carefully and accurately, mix chemicals carefully and accurately and process, voila 3/27 pass every time.
Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
September 30, 2014, 10:57 PM
fuelfarmer
Jon, the advantage is the no titration method is very forgiving and you get a boost in yield. I can't tell you how forgiving or the yield increase.

As for the chemical savings, here is what I think. I do not know this to be fact. When a titration is done you are trying to decide how much extra to throw at a known problem with the oil. The unknown, like water, can still bite you on the backside.
With the no titration method, I use the base amount of chemicals for 80% of the oil volume. When the glycerin is drained the bad stuff, or unknown stuff is removed from the remaining oil and fuel. Then the 10/90 in my case show me how much "clean" oil is remaining in the fuel. As stated the amount of the drop out may not be precise, but the results work for me. When I add the chemicals used in both stages it is less than what I would use in a single stage using a titration number. And because there is less soap the yield is improved a little.

There is also some benefit to removing most of the glycerin near the end of the reaction because glycerin can interfere with the reaction as it nears completion.

Stopping to drain the first stage does add a step, but the results make me happy. And can you put a price on happiness? Smile
October 01, 2014, 07:58 AM
Jon Heron
Thanks for the detailed explanation FF!
So your experiencing better yield and using less chemicals, that's a good thing!
My pal on here Dannodiesel has been talking about how much he likes this no titration process too, we just haven't had a chance to talk about what all it entailed.
It makes some more sense to me now though it would sure be nice to see some quantitative testing to determine if the extra steps are really worth it in the long run.
You being happy with it is certainly the important part for you. Smile
Cheers,
Jon


___________________________

Simple schematic for a pump and heater control with a high limit
Sensor for the biodiesel/glycerin layer
October 01, 2014, 08:16 AM
fuelfarmer
I can show the yield increase using sodium methylate instead of lye. 240 gallon batches.