I have made 6 batches in my processor so far (capacity 175 litres), the last 4 using PaulG's conc aqueous NaOH method. I can report that it works brilliantly!!! NaOH seems to dissolve more readily in water than in methanol, and doesn't generate any fumes, it does get quite hot however. The aqueous NaOH seems to dissolve instantly in methanol without any agitation. I have not noticed any increase in soap production.

My question to PaulG or anyone else using this method is have you developed a safe way to mix say 5 to 10 Kg batches of conc aqueous NaOH? If it can be mixed and then stored safely it will reduce the labour required to prepare each batch of biodiesel.

Greg

Greg, I usually make up 2.5L of conc aqueous NaOH at a time and store it in a plastic winchester. The type that used to be used for conc hydrochloric acid. To be on the safe side I store the bottle in a plastic bucket.
To dissolve the NaOH I use a 5L plastic beaker and glass stirring rod. Heat generated helps solution, occasional stirring required. I prefer to do this in fume cupboard as the fumes given off aren't very friendly to your lungs.
Chemical supply companies sell 1,000L "Bulky Boxes" of 50% aqueous NaOH. This is slightly weaker than conc and could work well for an industrial installation.
As a matter of interest I made three different mixes of "methoxide" and reacted them with the same WVO. The first was made from sodium and methanol, the second from NaOH and methanol and the third from conc aqueous NaOH and methanol.
GC testing revealed that the organic composition of the BD from the three different methoxides was practically identical.

Regards Squarepeg.

Nice to see you got the results back how did everything work out ?

BC

PaulG

You wouldn't expect the catalyst type to have any effect on the quality of the product after washing, provided the amounts used were equivalent. Only the quantity would be expected to differ.

quote:

---

Originally posted by PaulG:
I usually make up 2.5L of conc aqueous NaOH at a time and store it in a plastic winchester.

---

OK, I'm convinced that using conc. aqueous NaOH instead of dry caustic has advantages when mixing with MethOH. I suppose the same process would apply to KOH? If I dissolved 1kg of KOH in approximately 800ml of water and then topped it up to 1L after it dissolved, that would give me a very convenient 1g/ml solution.

It would also simplify making up titrant as I could pipette 1ml(unless it would be too viscous to pipette ) of this solution into more distilled water and top it up to 1L. In theory then I would be titrating with the same solution as I would use for the reaction. Hopefully, this would minimize measuring errors.

Thoughts?
Patrick

TDI Patrick.
I guess conc aq KOH would work, haven't tried it.
Commercial dry KOH usually contains some water ie it is not pure.
Though the 1g/ml solution would be convenient I would only recommend using a concentrated solution. Anything less could cause too much soap production. Don't have figures for solubility of KOH at 20 deg C available at the moment. Will look the figure up and adjust for various purities.
Post a table in next few days.

Diluting the 1g/ml solution to make a titrating solution could work. As a Lab technician I am more in tune with molarities (why did they change back from normalities, that made a lot more sense to me) than percentages when considering the interaction of chemicals.
When titrating WVO I use a 4ml sample of oil and titrate with 0.1M NaOH, because that is what I happen to have plenty of.

Regards Squarepeg.

Neutral, the BDs from the three different catalysts were tested unwashed but had been heated to drive off residual methanol. An additional sample of the conc aqueous NaOH batch was washed and tested also.

1L of used ETA Salfry (mixture of Canola and Soy oils) titration 1.69ml. Sample of same GCed.

Methoxides
A/. 2.98g metalic sodium dissolved in 250ml methanol.
B/. 5.19g solid NaOH dissolved in 250ml methanol.
C/. 6.23ml Conc aqueous NaOH mixed into 250ml methanol.

Mixing. 1L oil heated to 55 deg C, 2L PET Bottle for reaction vessel. Methoxide added to oil in bottle, bottle shaken vigorously for about 30 seconds,and intermittently over one hour. Reaction product allowed to stand for another hour.

results after 1 hour settling,
Raw BD............Glycerine
A/. 1057ml............. 150ml (dark brown liquid).
B/. 1045ml..............170ml.. .. .. .. ..
C/. 1035ml..............185ml.. .. .. .. ..

Raw BDs filtered through Whatman No. 1 filterpaper.
SG.................Viscosity (100ml pipette).
A/. 0.8794 at 22.5 deg C 33.01 sec at 25 deg C
B/. 0.8801 at 22.5 deg C 33.34 sec at 25 deg C
C/. 0.8815 at 22.5 deg C 34.31 sec at 24 deg C

After seven days the raw BDs were still clearing and depositing glycerine in small ammounts.

Methanol removal, heated to 110-120 deg C.
A/. % wt loss...2.75%
B/..............2.37%
C/..............2.41%
Observation: All samples of raw BD lost their turbidity and became crystal clear once heated. This clarity was maintained when the samples cooled.

There was an uncanny similarity between the results obtained by GC analysis (as you expected).
Strangly enough the only significant difference was in the methoxide made with metalic sodium.
A peak of %area 1.63 appeared for CAS# 002455-71-2, 001179-05-1 and 002519-10-0 which did not show up on any of the other tests. There was another detection only on the A sample and another which was absent on the A sample but present on the B,C and D sample.
I had hoped that the GC result would show up the soap content of the unwashed BD. Will have to determine that from yields from washed samples.
Sulphated ash would be the way to go but I don't have any platinum crucibles.
Neutral, would I be right in assuming that GC results only show the organic composition of the test sample? Would you expect the sodium soaps to show up? There were only methyl esters listed in the results.
Any other tests you could suggest to determine the relative conversion rates of the methoxides.

Thanks, squarepeg.

Thanks PaulG (Squarepeg)

I have been using 1 litre polypropylene bottles (originally contained sterile saline solution for wound irrigation), shaking 600grams NaOH with 550mls of water. I do the mixing (at home) over a stainless steel tub of cold water, while wearing nitrile gloves, apron and face mask. Initially I carefully release the pressure as the temperature rises, and place the bottles in the water to cool between shakes. As the mixing proceeds the bottles cool and air needs to be let in to prevent collapsing the bottles. I use two of these in each 175 litre batch (7 grams per litre). My oil is used cottonseed with some tallow in it, that titrates between 1.5 and 2 grams per litre, and I use 37 litres of methanol (21%).

The polypropylene bottles are not showing any sign of distress yet, and compatability tables I have found indicate that they should be OK. I am not sure what kind of plastic the screw caps are made of, but they also seem to be OK so far. These bottles seem to be ideal for this purpose as they have a large neck to make adding NaOH easy, and shouldn't be too hard to get as hospitals throw thousands of them away.

I have not mixed and stored any NaOH solution in advance of using it yet, but intend to do that soon. I am intending to store the bottles of solution in a metal bucket with cat litter in the bottom.

Many thanks for coming up with what I believe is a great process improvement.

Regards
Greg

quote:

---

Originally posted by PaulG:
Though the 1g/ml solution would be convenient I would only recommend using a concentrated solution. Anything less could cause too much soap production.

---

Hmm... I was basing my assumptions(yeah, I know ) on your recipe found here of 100g of NaOH to 90ml H20. Extrapolating that out to 1kg and 900ml(used 800ml in my post above), I really didn't think the water necessary to make 1L total solution would be excessive. I did ignore the difference in solubity between NaOH and KOH tho'...

quote:

---

Don't have figures for solubility of KOH at 20 deg C available at the moment. Will look the figure up and adjust for various purities.
Post a table in next few days.

---

Please do!

Patrick

Greg
Thanks Greg happy to share the method with you.

Patrick TDI,
you weren't too far off with the 1 gram per ml, It may work but could be a bit weak.

Here is the data for conc aqueous KOH.

Commercial grade KOH is usually 83-86% pure.
Contaminants are usually water 10-13% and potassium carbonate 2-3%.
KOH is also available 88-90% pure.
One gram of pure KOH is soluble in 0.9 part H2O at 20 deg C.

To make a conc aqueous KOH (at 20 deg C) dissolve 1kg of KOH in the following volume of water.

Purity of KOH.........Vol of water to dissolve 1kg
...83%..........................617ml
...86%..........................639ml
...88%..........................697ml
...90%..........................720ml.

The above figures are theoretical calculations from published data. I have no practical experience with conc aqueous KOH solutions.

The aim is to make as concentrated a solution as possible that will remain in the liquid state.
Store in a sealed plastic containor as strong caustic solutions etch glass, glass stoppers in glass bottles will be cemented solid,CO2 from the air will turn the KOH to K2CO3
The potassium carbonate impurity in the solid KOH is insoluble in alcohol and will precipitate out as the KOH is mixed with the alcohol.

How much to use.
Measure the final volume of the conc solution in litres. Multiply the grams of KOH required by that factor.
(When using conc Aq NaOH the factor is 1.2)

Regards Squarepeg.

[This message was edited by PaulG on 16 July 2002 at 04:36 PM.]

You're probably aware already that mixing KOH with water is extremely exothermic (loads of heat). Always add the KOH slowly to the water and don't add any more until the last addition has dissolved. Dissolving KOH rapidly in water can make it boil explosively. Take care as it can really burn you if you get boiling KOH splashed on your face.

I had a friend who got covered in hot concentrated sodium hydroxide solution (similar thing). He was drenched in it and although his workmate washed him down with water straight away, he died within 5 days because of infection of the burns.

Be cautious and you'll be OK.

Good advice Hardeep, I overlooked mentioning that.
As you mentioned best to add the KOH to the water in portions with stirring to prevent boiling.
Avoid breathing the fumes of alkali mist. Wear gloves and have running water handy.
Sounds bad but no more dangerous than the conventional method of preparing methoxide.

regards Squarepeg.

PaulG

That was a pretty interesting post comparing sodium metal, NaOH and aq NaOH. The difference between the methods appears to be pretty small but it is in the expected direction.

I cannot explain why there should have been an extra peak in the Na metal sample. GC does give erratic results sometimes due probably to the release of contaminants from injector packing material or from use of dirty needles. The only way to be sure a peak is valid is to repeat the injection. Contamination may occur in other ways as, for example, the Na metal may have been stored under an oil, some of which was carried to the reaction vessel.

>"I had hoped that the GC result would show up the soap content of the unwashed BD. Will have to determine that from yields from washed samples.
Sulphated ash would be the way to go but I don't have any platinum crucibles."

Sulphated ash will certainly give you the soap content but it is not a nice process. Many steps are involved and vast amounts of sulphuric acid vapour are generated.

Washed yields could allow calculation of the soap content, as you say, but the amounts may be hard to measure accurately enough. Being differences they would be small amounts which could be upset by variable losses in decanting and even by small errors in measuring the original oil.

It would be better to combine the washings, acidify them to about pH 4, or more acidic still, extract with hexane, wash with water, and then rotovap. That way you are measuring just the FFA, not the bulk of the BD, and the result will be much more accurate.

Another good way to determine the soap content is to titrate the BD with acid. It is simple, clean, accurate and causes no fumes. Titrate using bromocresol green. It goes from blue to yellow between about pH 4 to 5. Very easy to see. It gives the sodium content which is exactly what the sulphated ash method measures.

Why any organisation should put a difficult, expensive test in the ASTM when simple easy tests are available, is beyond me. Perhaps it is a legacy of the petroleum industry as a method for discovering how much sand was pulled up with the oil from the oil well and carried into the product. We should try to persuade the Australian authorities to be more rational when they produce their standard. Similarly it is irrational to include both a flash point and a methanol content test, as one determines the other.

>"Neutral, would I be right in assuming that GC results only show the organic composition of the test sample? Would you expect the sodium soaps to show up? There were only methyl esters listed in the results."

Whether of not you can expect to see soaps depends on the method of preparation. If you start by acidifying the sample all soaps will be turned into free fatty acids. These can best be seen in a gc if silylated. And of course if you acidify then wash a larger sample the acids can be measured using the standard FFA titration.

>"Any other tests you could suggest to determine the relative conversion rates of the methoxides."

As we all know there are several ways of estimating the degree of conversion, density and viscosity being the most accessible. These give good measures when comparing samples of modest conversion, but when looking at samples close to the ASTM standard the level of glycerides is so low that the effect on density and viscosity is small and hard to measure accurately enough. In this case the only method that I am aware of is gc. A method must be used which shows glycerides. The gc method used in your case was clearly not designed to measure the degree of conversion as no glycerides were reported, only methyl esters. The glycerides would have been there of course. They are still there, stuck in the injector or column!

.

[This message was edited by neutral on 17 July 2002 at 01:10 AM.]

This site IS the very cutting edge of technology in this science. Well there are other sites but this site on ocassion recommends DR pepper, no harm there. What we do here is not written in books or papers as a rule. If we do stumble on a patent or 2, (DR pepper inclusive) then no real hurt considering the staggerring odds of doing so.

[This message was edited by isoprop on 27 July 2002 at 10:09 PM.]

When using the concentrated solution, how much water do you end up adding per gramme of NaOH?

Can the stuff be bought as concentrated solution? Saving the mixing hassle.

Hi Dave

The proportion is 0.9 ml water to 1 gram NaOH. It becomes about 1.2 ml of concentrated NaOH solution.

regards
Greg

I'll have to try it.

Neutral
I think you have it with the suggestion that the sodium metal methoxide could have been contaminated with oil.
The sodium was stored under kerosene. Whilst I removed the surface skin of the sodium so as to use only shiny metal some of the kerosene must have carried over. These GCs sure don't miss much.

regards Squarepeg.

What if you didn't have a scale but you had an endless supply of 50% aqueous sodium hydroxide.

Could you start with a volume of this liquid and begin mixing dry NaOH into the solution and using a pH meter determine when the solution is at the proper strength? or is it not that simple?

If this is the case, could you not mix your catalyst (as aqueous) and never have a scale at all by starting with a given volume of water and start mixing catalyst until the pH is correct?

Would it work well enough to start with a set volume of water and begin mixing catalyst until it begins to thicken? Would that work well enough?

Is it possible to evaporate the water off the 50% solution instead of adding dry? I dont know if it would hurt to heat it but if that endless supply is free it may be worth it.
When I mix mine I never measure accuratly I just keep adding NaOH until it precipitates out. I found that if the product isnt perfect ( and it rarely is) then the amount you need varies by the amount of carbonized crud anyway. this is insoluble and will fall to the bottom (or sometimes float on top) and do no harm. the excess NaOH will do the same as you exceed saturation.
Just pour or pump what you need off the top and it works well
John

[This message was edited by JohnH on 20 July 2002 at 08:25 PM.]