johno,
I will be giving your method a shot as soon as I get my setup going again!
Thanks for the instructions!
Jon

Take 1cc of reagent (KOH or Acid or Chocolate syrup), add it to 1 liter of distilled water. That's a 0.1% solution. Use it to do a titration. The beauty is that the math of a titration of 1cc of oil, or biodiesel, or milk using this solution and a pH indicator (like phenolpthalein) works out so easily. If it takes 1 cc of solution to get the phenolpthalein to change color, then you need to add 1cc to each liter of your oil, or biodiesel, or milk.

For KOH or NaOH we are using weight (well, mass really), so we add 1 gram to a liter of water. The titration to make biodiesel then gives the weight (mass) of KOH to add for each liter of oil.

For my acid solution, I use 1 cc acid (that's volume, not weight) to a liter of water. The titration gives the VOLUME of acid to add per liter.

Let me know if this isn't clear. Once you've done it it really does seem simple, but it's complicated to read.

Cheers,
JohnO

Oop's, My bad!
I see I wasn't paying attention when I linked to the soap titration instructions!
I edited my previous post...

This method sounds much easier and more accurate than trying to watch the colour change in the sight tube!
I intend to use turmeric for the indicator as I am all out of phenolphthalein, do you see a problem with using turmeric?
Jon

Johno,

Thanks for this great idea. I am about to run my first batch today, April 13 2009 after a year of assembly and testing (yes I know this is far too long) but could you explain the following statement a little more in depth.

quote:

Blank some IPA (rubbing alcohol) with Phenolphthalein and the regular KOH solution in a beaker, in preparation for the acid titration.

Here is what I take it to mean:
I assume you mean put a few drops of Phenolphthalein in some rubbing alcohol, but what is the regular KOH solution? The .1% Titration solution? As I write this, it seems clearer to me that is what you mean, but I just want to make sure. So the final instruction for the ATE test (Acid Ester TEST - I just made that up): "In a beaker, put a few Milliliters of Rubbing Alcohol, add a few drops of Phenolphthalein, add a few Milliliters of titration solution and add 1 CC of ester - all of which will be now bright pink. To this you will drop in 1 ml at time a solution of .1% acid (HCL or Sulphuric) until the pink disappears. Record this amount of acid used." "Now for the AGE test (Acid Glycerin Test), repeat the same process except this time instead of adding ester to the "Rubbing Alcohol, Phenolphthalein, Titration Solution add 1 CC of Glycerin"

Thanks
Doug

bernyjb,

Careful with that Keg can. I think its aluminum and I think Methanol and aluminum are not too compatible.

Johno - in your instructions on titration there appears to be an error in that you twice refer to titrating esters and not at all to titrating glycerine.

Does anyone realise that when phenolpthalein turns clear the pH is not neutral but roughly 8.3? http://www.chemguide.co.uk/phy...eqia/indicators.html Likewise turmeric returns to yellow at pH 7.4.

I understand that a handy broad-range pH indicator can be made from red cabbage. But wasn't there a thread from someone who distilled bio and gly together without neutralizing and didn't experience the dreaded, mysterious and hard to find back reaction????

Paulus - thanks for catching that. I've edited it now.

A couple points to clarify:

When titrating with acid and phenolpthalein, the IPA/indicator (before adding the esters or glycerin) needs to be PINK to start with. That's why we do a blank - so we've got a known starting point. IPA tends to be acidic to begin with. To do a blank, add the indicator to the IPA and add just enough KOH solution (the usual solution used to titrate oil for making biodiesel) to get the IPA to just turn pink. Now it's sensitive to subtle changes in pH - just a little alkali will turn it bright purple, or just a little acid will turn it clear.

The goal is to add enough acid to the esters or glycerin to turn the indicator CLEAR (actually it'll be cloudy - we want it to turn "not-pink").

The point is to push the pH below 8.2 (or 8.3 according to some phenolpthalein data). Above 8.2 the transesterfication reaction can reverse. Below 8.2 it can't. We could push it waaaaay below 8.2 by adding a definite surplus of acid and not bothering to measure, but Alkaline fuel ages better, won't attack steel, and is pretty well understood. Acidic biodiesel is generally undesirable, since it ages poorly, fails 3/27, and forms "bad things" (I wish I knew more about the chemistry), and hasn't really been studied that I know of. If someone wants to intentionaly try some, let us know what kinds of lab results it gives.

If we leave the pH above 8.2 and suck out the methanol from both the glycerin and esters at the same time, some of the ester molecules will recombine with the glycerin to form mono-glycerides and other undesirable things. Those nasty things will clog filters as well as reduce the yield of biodiesel. GL's process runs the reaction to completion, then 'kills' it with acid so it can't reverse. That makes it safe to suck out the methanol for recovery without reversing the reaction and without polluting. This also produces a little more recovered methanol, saving everyone a little money, too.

I must emphasize that the method I've described might not work in general. The assumption that 25% is glycerin and 75% is ester might not work for other batches. My oil doesn't (normally) titrate very high, so I might just have been lucky. If a few other brave homebrewers can document their results, with a range of feedstocks and techniques, then we might be able to come up with a simple formula to compensate for the unknown problems we'll find. For example, the simplest change would be for an individual to use the known yields of esters/glycerin from their process. Let's say it was 30/70% instead of my 25/75%. This would be consistant with higher titrating oil, known to give lower yields. That would be nice, if it worked out. Does everyone understand how to change the calculation to reflect the different percentages?

Cheers,
JohnO

quote:

Careful with that Keg can. I think its aluminum and I think Methanol and aluminum are not too compatible.

Thanks Doug for your concern. The can is sheetmetal, mild steel I think.
Anyways, aluminum and methanol don't cause any problems. Aluminum is contraindicated for long term storage of methanol.

I read somewhere aluminum was a problem but it was with Methoxide, sorry my bad. Here is one link:

Material Comaptibility

Ok well, Yesterday was the first time I ever attempted making Bio Diesel as well as putting my Push/Pull Rig through its full paces. A year ago I tried to have a friend video me starting this process. Too bad they were not there not taping yesterday, because I would have put Tim Allen from the TV show Home Improvement to shame. Let me cut to the chase: I did not end up with Bio-Diesel and hard/gooey glycerin. What I ended up with was what you see in the pictures.






The Mason jar full of clearer brown liquid was simply scooped off the top of the "glycerin" buckets this morning - it is clearly some kind of bio diesel. The Pyrex measuring cup was filled from the bottom of my holding tank. As you can see it has some liquid "glycerin" at the bottom and the possibly biodeisel brown liquid on the top. I am full of questions, dejected, but with a glimmer of hope.

1. What do I have? Where is the Glycerin?

2. How can I fix I this batch or should I just start dump it out and start over?

3. What went so wrong?


Ok here is recap of what happened yesterday.

1. I started with 130 Liters of more or less dry used soybean oil. It was not as dry as Virgin Olive Oil, but no longer cracked/pop and had very little bubbles in the frying pan test.

2. The oil was now at about 145F and titrated to about 2.5 to 3, so I called it 2.75. Using KOH, I added (7+2.75)*130= 1268 grams of KOH to 26 Liters of Methanol.

3. I injected the mixture slowly, about 30 minutes through a Venturi, and looking at my sight tube, the mixture went from oil brown to a darker brown. After the Methoxide was added, I ran the process for another 90 Minutes with temps slowly going down to 133. During that time, I occasionally opened the Venturi to add air with the idea that it would aid in the mixing.

4. After 90 minutes, in a mason jar, I mixed a teaspoon of Turmeric into 50Ml of methanol and sucked that in through the Venturi. I found out then that Turmeric does not dissolve easily into methanol, and after the sucking, I was left with a lot of extra turmeric powder at the bottom of my jar. So I added some more methanol, stirred more vigorously and re-injected. This is where things got worse.

5. I put 150ml of 95-98% Sulphuric acid (Utah bio) in the same Mason jar, and began injecting it through my Venturi. It was hard to add slowly. I have it set up to go through a very small piece of Tygon Tubing (1/16 ID) and a 1/4" ball valve, but basically the ball valve is just too big and the acid will always be added too fast. So I pinched the tube and tried to manually add the Sulphuric Acid slowly. It probably took 15 minutes to add the 150ml of acid. I have 2 sight tubes: a big 1" 1/4 glass one and a small 3/8 tube that I can also pinch to better see the color change. I never noticed a color change and with the oil/methoxide as brown as it was it, I don't think you ever will. So.... somehow in my Methane/Sulphuric/KOH high, it seemed like a good idea to add the last 40ml of Sulphuric to the mix! There was a little voice in my head that told me no, but hey, I was worried about storing the opened bottle with 40ml left... in the moment it made perfect sense, so I just tried to add it slowly as well.

6. Worried that the acid went in to fast, I ran the mixture for another 5-10 minutes just to make sure the acid mixed in well.

7. Now things get comical. Turned on the vacuum pump and low and behold, it would no longer pull a vacuum. It had operated fine early in the day but I think between dripping oil and getting too hot, I damaged the Venturi used in the Vacuum setup. So now I am in small panic. I was worried the Glycerin would harden if I turned everything else off... so.... with the processor pump running, I decided to re-rig the plumbers delight and turn the system back into an GL1 and instead of using the vacuum, just use more heat. I still was going to attempt to run the reverse flow.

8. With the quick use of many unused parts (well parts that have been in and out of my setup multiple times) I was able to connect the vacuum hose directly to the plumbers delight. I turned on the heat, and reversed the flow... and...well... quite anticlimactic. Nothing seemed to happen. I expected to at least hear gurgling up from the bottom of the tank. Nothing. So, I let the heat rise and waited. 140, 150, 160...not much. I was worried the pump was not strong enough to push the Bio back up through the bottom without the vacuum, but as the my electronic temp gauge was not skyrocketing, I knew I had some circulation around my heating elements (My heating elements are inline with the plumbing – not in my tank. So when I don't have good circulation, the temp begins to shoot up a degree every few seconds. With the temp only slowly rising, it indicates I have a good flow). At 178F or so, I played with the different valves to see if I could shake things up and one time left a valve open that made it so I was pulling from the middle of the tank (bio diesel) and the bottom at the same time. Low and behold, the tank gave a shudder and began to gurgle. I figured out my 'mistake,' shut that valve down and now the system definitely was pushing liquid up through the bottom of the tank. Also I started to see methanol trickle back into my collector. (It might have been happening at a lower temp, but in all the confusion might have missed it. It was a constant even small flow.) Over the next 90-120 minutes, I let the system get up to 203F and in the end had recollected about 6 liters of the 26 liters of methanol. At 10:30 at night, I was still getting a small trickle of methanol but it was dark, I was exhausted and just wanted the day to be over, I decided I had recovered most of the methonal. I shut everything down.

9. Placing a bucket under the processor, I expected to drain off goopy glycerin. What came out was a rush of dark liquid. Could this be glycerin? I realized I was draining Bio from the middle!! I forgot to shut off the valve from the middle of the tank! I shut that off and continued draining from the bottom. What came out was still liquid and dark...not thick at all... I drained off 3 five-gallon buckets worth of this liquid, looking for a color change to golden honey. Never saw it. So I pumped the remainder of the liquid to my holding tank. (Let me tell you, I got lucky with the those buckets as they held up to the 200 degree oil. They went a little soft, but next time I will never drain such hot oil into that kind of plastic again.)

10. Now its the next morning. Everything has cooled down. I collected those samples above. What are they? I have no hard/gooey glycerin. Why not? I put my hand in a plastic bag and touched the bottom of the all the 5 gallon buckets and no discernable glycerin. There certainly seems to be some liquid glycerin in the holding tank by looking at the Pyrex measuring cup, but its not thick, just darker liquid (I think I should count this as a blessing right now as I am surely gone to need to empty that stuff.)


I am going to take the day off. Clean up, calm down and see what comes of this. If any of you want to call and give advice over the phone I am up for that as well as I am now felling VERY in over my head. I am sure the light at the end of the tunnel is near, but darn this was disappointing.

Doug
323-428-5115

IMG_1246.jpg (44 KB, 28 downloads) Jars

Cant access the top two pictures in your post. But the bottom picture of stuff in the pyrex jug looks pretty normal to me....bio on top with soap and glycerine settling. We all have some pretty evil days from time to time, lol. You'll get the hang of it.

One other idea. I am really hating the vacuum pump design. I mean its awkward, it leaks, it requires managing its temp and drawing another 1500 Watts for the heater plus the pump to run it. What about AC vacuum pump and manifold. I don't understand exactly how they work so if it is silly let me know, but doesn't that do the same thing? Would that be safer the refrigerator pump? Is the only drawback cost?


Doug

Doug: I know this may sound strange, but CONGRATULATIONS!

What you're going through now is similar to what happened to me with my first Dr. Pepper: I just couldn't believe it was so easy!!!
What you have in the Pyrex cup IS glycerin. The reason it's not thick is most likely because there's still a substantial amount of methanol in it. I've had batches in which the gly took several days to thicken up, even after being left under the sun.

Take a sample of the BD in the mason jar, and do a 3/27. Chances are you have good BD. Some pages back, Graham and others discussed using turmeric in the process. There were some people experimenting with it, but as far as I know, nobody had success using it. In any case, you don't need to fully dissolve the turmeric into the alcohol. When the alcohol turns yellow, the curcumin (which is what you're really after) is already dissolved. Just get rid of the sediment.

About the extra acid, I don't really know what its effect would be. For what Graham, and other more knowledgeable than me people said in the past, I'd assume you risk getting back your free fatty acids from the byproduct, so, if the 3/27 looks good, I'd say you're ok...

That sounds like my first batch with my push/pull.It was very hectic for a while.

130 ml. of sulfuric acid is what I add to my 340 liter batches.I then do a titration to check ph and that seems about right.You probably put in too much acid. I really like the idea of pulling a sample and doing an acid titration to determine the correct amount to add.
You already have to pull some oil for a 27/3 test and if it passes you can then use the same oil for acid titrating.

My acid is injected before the pump and sucks in very fast,less than a minute. I now pass 27/3 and the phlip test every time. I'm on my fourth batch.

When you crack the ball valve to take a sample of oil for a 27/3 test there could be unreacted oil in the hole in the ball of the valve. That oil will contaminate your sample and cause a false fail. What I do is pull about a quart out and set it a side. Then I pull my sample for my 27/3 test. The quart is used to flush out my acid injection tube after injecting the acid.I suck the entire quart into the processor thru the same tube I inject the acid with after injecting the acid.

I recovered alot of methanol from my first batch of 340 ltrs. and my glycerin was so thick I had to pressurize the tank to get it out. By the next morning it had hardened up so much I can stand on it.

I have not been able to recover as much methanol since then. My pump will start cavitating at around 12 inches of vacuum and stop pumping oil. I believe the seals are leaking air. I try to compensate by increasing the temperature but it hasn't worked as well as the first time.However, after seeing how hard the glycerin was the first time I'm now somewhat terrified about leaving it in the tank too long cause if it hardens I'll never get it out. I let it sit for no more than ten minutes before draining it.

Congratulations on discovering that the mid tank ball valve should be closed when draining gly.. It took me three batches to find that out. I was draining gly. into 5 gallon pails to settle out and the next morning I would have 3 gal. of gly. and 1 gal of bio in every bucket.

Dave

How is it looking now Doug? Did you try the 3/27 test on it?
As far as a mechanical vacuum pump goes I wouldn't bother if I was in your shoes, you already have your pump built now just work out the kinks. It will be less effort and a better setup once you get yours working right in my opinion.

quote:

My pump will start cavitating at around 12 inches of vacuum and stop pumping oil. I believe the seals are leaking air. I try to compensate by increasing the temperature but it hasn't worked as well as the first time.However, after seeing how hard the glycerin was the first time I'm now somewhat terrified about leaving it in the tank too long cause if it hardens I'll never get it out. I let it sit for no more than ten minutes before draining it.

Yep that is the problem with mechanical seals under vacuum. I have been doing allot of research about pumps for my new 340L reactor. When using vacuum, in my opinion, you need a mag drive pump or a wet rotor pump like the Wilo pump Graham uses on his setup. The only seal to leak on a wet rotor style pump is the gasket on the pump body, that means there is no mechanical seal to dry out under vacuum and fail...
If you can, try valving off your pump and draw a deep vacuum on your reactor and see if you gain any less methanol than before when you were circulating. I have done this a few times before, demething just the biodiesel without the glycerin and it seemed to work just as fast as when circulating it.
Jon

Johno:

I tried your method of doig an acid titration today and had some problems and observations.
I reacted 335 liters, passed the 3/27 and then pulled out 760ml. of oil to settle. After settling I had 150ml. of glycerin. about 20%. So far so good. I then pulled off 1ml.of bio.from the top and acid titrated it three times. It took between one to two drops to neutralize it, turning it from pink to yellowish oil color.
I then pulled out some glycerin and it took 168 drops to neutralize it, did it three times so its a good number.It appears that most of the lye is in the glycerin.

I know from past experiance that it takes about 130 to 150 ml. of sulfuric acid to neutralize the entire batch.

I believe that 1 ml. is about 25 drops. If that is true then by my calculations I would have to use 718ml. to neutralize the entire batch.

I must have done something wrong. Do you have any suggestions?

Dave

Hummerbk: thanks, that's exactly the kind of feedback I was hoping for - by my calculation you made a math mistake, not that it solves the problem with my method:

Batch size = 335 liters
Glycerin = 20%, titration = 6.72 ml (wow!), therefore the acid to neutralize the glycerin will be:
335 x 6.72 x 0.20 = 450cc acid

Esters = 80%, titration = 0.04 (one drop), therefore the acid to neutralize the esters will be:
335 x 0.04 x 0.80 = 67 ml.

Therefore, total acid to add = 450 + 67 = 571 cc, which is WAAAAAAY too much, especially if you've already got some experience that says it should be around 130 - 150 ml.

OK, the formula needs to be tweaked. I'm suspicious of the glycerin acid numbers you got - 6.72cc of Sulfuric Acid is an enormous amount, which makes me wonder why you still got that value three times! Any suggestions? I assume you did a blank, and your titration acid solution is good (is it possible the water you used to make it was alkaline, weakening it?)

For you to arrive at a 150 ml quantity would require a correction factor of 0.18 to multiply your glycerin acid addition by, which blows my theory out of the water.
Cheers,
JohnO

Johno,

I actually did 4 titrations on the glycerin. I got 85 drops,169,167 and 185 drops. I threw out the first titration since it was nowhere near the last three.
I'm not giving up on this method since I believe it has great potential. I just need to understand it better and learn how to tweak it for my processor.

The water for the acid soultion was distilled water. I did not check it ph.

For my next batch I'll make up all new solutions in case I screwed something up.

Just curious as to wether your glycerin had much higher numbers also.

Dave

Oops, sorry, I wasn't clearer in my earlier posting, and forgot to comment in the last posting.

Yes, the glycerin will be grossly more alkaline than the esters. It might even be possible to skip titrating the esters, if they always take 1 drop (or 0.1cc in my tests) to neutralize. That leaves the glycerin to neutralize.

The fundamental question is - can we measure some characteristic of the glycerin that will tell us how much acid to add to bring the total pH below 8.2?

Hummerbk - do you have a powerful pump on your processor? 335 liters is marginally too big for the common "clear water pump". Also, I was surprised that you only assumed 20% glycerin - all my experience indicates I'll get a little more glycerin, and a little less ester. Then again, maybe it's possible to use a smaller, fixed value for glycerin in your case. That'd be pretty useful, especially if we can figure a simple way for each homebrewer to calculate their own fixed value (assuming they will differ)
Cheers,
JohnO

Johno,

My pump is a two horse 220 volt model that supposedly pumps at 110 gallons/minute.
I may have been off on the glycerin amount. I assumed the sample I drew off was representative of the entire batch. I pulled 760 ml. out and 150ml. of gly. settled out. I believe I get closer to 25% in most batches, never actually measured it.

I'm thinking that concentrating on neutralizing the gly. is the way to go since the esters seem to be insignificant.

Dave