Graham,
I just spent the last 5 days reading this thread with growing interest and am looking fordward to seeing the plans you mentioned a few weeks ago. I would like to incorporate your findings in my process but I know very little about electronics. I am looking forward to your next post.

I found some information that might be a "clue" to the inconsistent separation.

There is a technique called "electrophoresis" which is used to separate amino acids (like the bio researchers use). It is conceptually very similar to what you are doing here for fat / glycerol separation.

The interesting thing, is that it is very dependent on the exact charge on the molecules, which just happens to be very dependent on pH.

In the case of the amino acids, if the pH is less than 6, the separation goes in direction A, if the pH is 6, it does not happen at all, and if the pH is greater than 6, it goes in reverse.

It might be possible that your mixture needs a pH adjustment to work consistently.

I did not do an on-line search, but this was taken from my son's college chem text, "Organic Chemistry, A short Course, 11th edition, Harold Hart, Leslie Craine, David Hart, copywrite 2003.

I am sure there are other places to research this in more detail.

Thanks Harry,

I'll look closely at that, because this may explain why I'm having such difficulty dropping water from WVO, which is contaminated with FFA.

Clean unused oil drops water easily, but who has water in their unused oil?

Hi Graham,
Have you had a chance to do any experiments to test my water content theory as it relates to floculation of soap micro particles?

Regards
Tim

Hi Tim

I have dosed the soap laden biodiesel with various amounts of water, up to becoming hazy, but still get no effective electrostatic effect.

I am becoming less and less hopefull about electrostatic soap removal. Glycerol separation is wonderfully effective, soap separation is depressing.

I am simply using gravity now for my soap separation after methanol removal - it is remarkably good, if slow.

I would imagine a series of settling tanks in series might work well, feed in at around 50% level, take off at 100%.

I'm only using one 45 gallon settling tank at present - I never empty it, just keep adding soapy biodiesel to it after each batch and take off from the top to fuel the cars a few days after the batch is finished. It suits my usage pattern well, and has been plenty clean enough for the engines and fuel filters - I have recently not been using any filtration before the fuel goes into the cars. I have jam-jar samples of each final few litres of transfer, and none have shown any signs of dropping sediment, and all are well below 100ppm soap.

It's becoming a very lazy, simple process.

Although it's not a straight line as elegance would prefer... have you tried adding a small amount of methanol to the bio after the bulk of the soap has been removed? This should be less than originally there as there is much less soap now. In any case the method is easy enough that removing the meth twice is not such a big deal if it is recovered and reused each time.

Also upping the voltage might increase reliablity. Johno had good fast results with 7Kv.

and did you do anyting interesting with the pyrex tubing?

Hi Ant,

I haven't tried the methanol idea - will give that a try next go. All ideas go on my corkboard and stay there till tried.

Yes, the pyrex tube was really useful - I used it after the venturi to see the downstream flow pattern and mixture colour/texture when using the venturi to mix methoxide into the oil on the PRESSURE side of the pump.

That convinced me the venturi makes a bl***y good mixer, so I hard plumbed it and have made some wonderful conversion BD by injecting after the pump.

I could also see I needed around 6" straight pipe run after the venturi, to allow the flow to join up properly again, before any elbows.

It is now back in my inclined glycerol cell, with a pair of stainless steel rod electrodes.
One rod on the bottom of the tube, one 2/3 of the way up, not touching the sides. That feeds to a Tee at the bottom, with the lower branch going to the glycerol tank, upper branch back to the tank, via pump.

This is all on a side spur, taking a fraction of the total flow, as the cell gets completely overwhelmed by full pump flow rate. It also isolates the fragile cell from the full impact of the pump when it kicks in - the pyrex is there for experimenting only, not a long term installation.

I have a nice camera again, rather than my phone camera which I've been relying on the past 4 months, so this weekend will be spent photographing all the bits and bobs to do a proper write up.

Does this work for dewatering? I've got a 15kv transformer I'd like to try out on dirty, watery oil.

Hi Graham

Would you PLEASE post the plans for the Glyc. separation. I see no conflict between using the glyc. separation, while experiments on soap / water separation continue..
Thanks

works very well for dewatering clean oil but has had variable results with waste oil. Possibly due to the ffa content as indicated by harry and graham a near the top of this page.

Try it and see.

Do limit the current flow on your transformer if you dont want to die. If you dont know what that is or how to do it then read untill you do. The info is in this thread. Do not use your transformer without current limitation.

Would fine filtering be good enough to clean up the oil before using electrostatic separation? My idea is just to drop two electrodes into a plastic 55 gallon drum of warm, filtered, and wet oil- then push 15kv across them and hope for dewatering. Is this realistic?

It may be realistic depending on the pH / FFA issue which has just been raised and is as yet unexplored.

Without wanting to give offence I think you sould put some time aside to read through the whole thread. I know it is long but it will just get even longer if we repeat information that is already there again and again, every time someone asks for a quick answer without wanting to sit down and read.

I know it is kind of a PITA, but it seems like perhaps the 2 stage glyc / soap clean out might be a real advantage. I am thinking that glyc which is removed this way would be better (cleaner) and perhaps more useful than if it has soap in it.

What? Soap is pretty much always produced to some extent or another. A proportion of what is produced is always dissolved in the glycerine. What are you talking about and how does it relate to this thread?

Hi Ant - sorry for not being more clear about the point I was try to make.

A few posts up, GrahamLaming posted:

quote:

Originally posted by GrahamLaming:

I am becoming less and less hopefull about electrostatic soap removal. Glycerol separation is wonderfully effective, soap separation is depressing.

I am simply using gravity now for my soap separation after methanol removal - it is remarkably good, if slow.

I would imagine a series of settling tanks in series might work well, feed in at around 50% level, take off at 100%.
.

The point I failed to articulate correctly is that as BD comes into larger scale production, there is going to be a lot of waste. At least one of these waste products, glyc, actually has a value which is enhanced by its purity.

Graham's electrostatic method appears to have the potential of enhancing the purity of the collected glyc because it essentially makes the waste separation a two stage event:
- Stage 1 - electrostatic glyc collection
- Stage 2 - gravity collection of "other stuff"

I hope that helps make my post a bit more on topic.

Ahh I see. I'm afriad that id doesn't work that way in my understanding of what happens.

What graham refers to as the glycerol is actually the same mixed byproduct that normally settles out. This is in fact glycerol with soa, methanol and caustic dissolved in it.

The soap that falls out later is the soap that normally stays in the bio, held there by the methanol. This is only a small part of the soap produced. The rest of the soap is still in the glycerine even when it is dropped electrostatically.

At least that is my understanding.

Hi, I found another "hint" for the process in a chemistry book. There is a technique used for chemical separation / chromotography that relies on using an electrostatic field, similar to what you are doing.

Of significant interest to me, were two points
- Pretty much every chemical has a constant associated with it, similar to the voltage constants of inorganic 1/2 reactions
- These "constants" are very temperature dependent, and as the temperature rises much above 25C, the associated separation ability changes dramatically. Just for fun, they do not change linearly or necessarily similarly.

BTW, the published constants are all for 25, as you would expect for thermodynamics.

In other words, the ability to separate using this technique is going to have substantial pH and temperature sensitivity.

I also read that the soluability of FFAs in water can be quite substantial. If it is like many other materials, this might also be a temperature dependent function.

I suppose one might be able to use this to "rinse out" some FFAs. I have not tried that one yet.

Its funny, I must have really slept through those organic chem classes, or was even more lost than I thought.

Hmmm heard of washing FFA out of used oil with methanol but never with water. Are you sure of your source and your interpretation? if you could just water wash the oil to reduce FFA I would have expected it to be a widespread practice by now.

I building another pressure reactor and I plan to try build in an electrostaic seperator. to generate the high voltage I fould a beast of a transformer out of a microwave


Those I'm not sure of specs on the transfromer but basse and what I see it could be over a 100:1