Hi folks

This method should allow you to use less (possibly nil - edit 13 June 06) water in washing, and seems to make washing less prone to emulsions, as it allows you to scoop off much of the soap before the wash stage. Introduces no extra chemicals or materials. (I run a 5% prewash in the processor).

Must be done in a well ventilated area. [ The experimental method I used here seems to release too much methanol to the air. A better approach may be to distil the methanol out of the batch - GL ]

1. Empty your freshly reacted, raw BD into your wash tank, but DO NOT add water yet. Also 1/2 fill a Clear plastic bottle with some of this brew to compare with the end result.

3. Lower an air bubbler into the BD. I had my bubbler in the centre of the tank, like this...





4. Bubble air thru the BD for 24 hours. This does several useful things...

1. It helps evaporate and thus eliminate methanol.
2. The reduced methanol concentration reduces the amount of soap the BD can hold.
3. The liberated soap picks up air bubbles and floats to the surface as a thick scum.

[ Comment added 18 June 2006: Much of the soap also drops to the bottom of the tank. I'd suggest if you want to do a water wash after this process, do it in a separate tank, to avoid dissolving the settled soap in your wash water ]

So you end with with separated soap...




And now you can easily scoop off the soap with a sieve ...




Keep going until your soap scum recovery level reduces to negligible amounts.

Take another plastic drink bottle and 1/2 fill it with this soap-reduced BD.

Now do a gentle wash test on your 2 samples and you'll see the bubble processed BD has much better wash handling properties and of course less soap.

I bubbled at 50 degrees C in the beginning, straight out of the reactor tank, and let it cool down naturally to around 15 degrees C. A warmed tank would probably be better, maintaining a higher methanol evap. rate.

A recirculating pump thru a filter may further improve soap yield after the bulk has been removed manually. No idea what filter size would be best, but I guess it would need to be a bag filter to have enough capacity for the soap. At least bag filters are easy to wash out.

You can then wash your soap-reduced batch, with much reduced risk of emulsification and less water usage.

[ Comment added 18 June 2006: Much of the soap also drops to the bottom of the tank. I'd suggest if you want to do a water wash after this process, do it in a separate tank, to avoid dissolving the settled soap in your wash water ]

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

I am going to give this a go on my next batch, I will post results.

Graham,
Good work.
Great contribution!
Thanks.
Diff

Worthwhile discovery, Graham. Thank you for sharing!

Hmm.... I have been giving some thought to alternative pre-wash treatments (mostly because I really don't want to wash) and tried heating the post-reaction BD to 110C to ensure that most methanol is flashed off. (I know that is way to high but I wondered if it would dry it at the same time). This left the same type of soap scum that you got here but also a little in suspension too.

So, why wouldnt we just drop the post-reaction BD into the heated wash tank and heat to 70C for a couple of hours and get the same result before washing? That way, there is a potential for reclaiming the methanol too for those that wish to.

just as an afterthought - what is the maximum temp that you would wash the BD at?

Graham,
How much washing do you have to do after you have done the bubbling for 24 hours? Do you think that if you had multiple bubblers you could reduce this time even further?

We are working on a homebrew size processor for rural central american communities and were hoping that this could reduce the need for water quite a bit. Any thoughts on how we oculd increase efficiency?

Do you recover the biodiesel that is mixed into the soap?

Have you tried this with biodiesel made from high FFA oil using NaOH?

Hi Tripi,

The faster you liberate the methanol, the sooner you can harvest the soap. This is a question of energy input, tank lagging, and available time.

You'd need to take care of the methanol fumes, which at above 65C or so will be released quickly, so a closed low pressure system with an adequate capacity condenser might be a good idea.

Curt,
I estimate I halved the amount of soap in the BD after 24 hours, using an unheated, unlagged tank, going from 50 to 15C. The loss of methanol is slow at 15C, so to halve the remaining soap without more heat input would probably take a week? I guess the rate of release drops off rapidly, perhaps exponentially? Not sure here, this is a guess.

Multiple bubblers seem a good idea, as does a well lagged tank. Tripitaka's idea of condensing the methanol may be worth looking into too.

More experimenting needed to find a temperature and air throughput rate for best efficiency.

Tilly,

Yes, the scooped-off soap holds a fair bit of BD, which separates OK with a cheesecloth.

My oil titrated at +1g and I use NaOH. I haven't tried any other oil.

Just out of curiosity, what is a lagged tank?

Tank lagging is what us Limeys call insulation.

Its just a way of keeping heat inside the vessel that it is supposed to be in.

word

This is good as we are working on lagging our tank right now. We were thinking of just buying some fiberglass insulation and just duct taping it-maybe including a layer of tin foil and bubble wrap as well.

I just searched the forums for a better part of half an hour and couldn't find any threads on insulation. Do ya'all have any advice or want to shoot me in the right direction?

Graham,

This is the technique that I used for my last two batches of fuel. Eric of Burlington Biodiesel suggested to me to do a major air bubble wash immediately after the glycerin settling period. The idea is to get the excess methanol out of the BD.

My processing steps went like this:

1. Process as normal using NaOH.
2. Separate for 12 hours.
3. Drain glycerin.
4. Move to wash tank and air bubble with two blocks for 18 hours. (This produced scum on the surface that looked just like your photo.) I skimmed it off.
5. Applied 200g of Magnesol and blended for 30 minutes.
6. Filtered Magnesol out of BD using a Spinner II model 25 centrifugal oil filter. I cleaned the Spinner II every 10 minutes and since the flow rate through housing is about 3.5 liters per minute it would be quite full after 10 minutes.
7. After about 40 minutes the entire batch was transferred to the centrifuge drum and I then recycled the fuel back through the Spinner II cleaning about every 30-40 minutes as required until no deposits were left. The Spinner II filters down to .5 micron.
8. Transferred fuel to storage drum for settling. Perfectly clear horizontally and vertically.
9. Specific gravity: .875
10. Now about 2 weeks later if I draw a sample from the bottom of the tank I have some white wispy stuff floating around suspended in the BD. If I draw a sample from the middle or above it is still very clear. I think I am getting some residual settling out of gycerin and FFAs that did not completely react. I may have had some methanol left in the mixture.

It will be interesting to see how these batches look after more settling time in storage. Anyway, I think this completely waterless method of removing impurities is going to work. It needs some fine tuning but I plan to keep experimenting with it.

Thanks for your post.

Don

Graham,
Your idea sounds like a good one. Tonight I put a 27L batch of 2 titrated BD in the wash tank with a sigle wooden bubbler and a titanium heatwand set at 30* C. I will post my results as they occur.
Fruitcakesa

Is there some way this could be done in the processor .,.this way the fumes would vent outside..How small do the bubbles need to be.
i am thinking of low pressure compressed air
delivered through the site tube..

hey fruitcakesa where in vt are you

So... just to clarify, the methanol is being driven to the top by the bubbles and being evaporated, right? Does the scum form because the soaps/residual glycerin had been carried by the methanol?

I'm wondering whether doing a prewash will interfere with this bubbling method. It appears during the prewash that the reacted bd picks up a bit of water and has a slight OJ appearance. Hopefully that won't cause any kind of problem (emulsion) if you bubble like this before adding any additional wash water.

I'd be interested in trying this method to see how much it cuts down on water washing. My oil titrates at +1.5 or so with NaOH and I have to wash 5-6 times on average (35 gallon batches using 30% cold wash water).

Hi OB1,
I am down in Cavendish, Southern VT Ludlow /Springfield area. Skimmed off a thin layer of scum from the pre bubbled BD today and now have added 15L H2O and am now bubbling that.

You guys are not far from me, I am in the Keene,NH area.

quote:

Originally posted by Don G.:
Graham,

This is the technique that I used for my last two batches of fuel. Eric of Burlington Biodiesel suggested to me to do a major air bubble wash immediately after the glycerin settling period. The idea is to get the excess methanol out of the BD.

My processing steps went like this:

1. Process as normal using NaOH.
2. Separate for 12 hours.
3. Drain glycerin.
4. Move to wash tank and air bubble with two blocks for 18 hours. (This produced scum on the surface that looked just like your photo.) I skimmed it off.
5. Applied 200g of Magnesol and blended for 30 minutes.
6. Filtered Magnesol out of BD using a Spinner II model 25 centrifugal oil filter. I cleaned the Spinner II every 10 minutes and since the flow rate through housing is about 3.5 liters per minute it would be quite full after 10 minutes.
7. After about 40 minutes the entire batch was transferred to the centrifuge drum and I then recycled the fuel back through the Spinner II cleaning about every 30-40 minutes as required until no deposits were left. The Spinner II filters down to .5 micron.
8. Transferred fuel to storage drum for settling. Perfectly clear horizontally and vertically.
9. Specific gravity: .875
10. Now about 2 weeks later if I draw a sample from the bottom of the tank I have some white wispy stuff floating around suspended in the BD. If I draw a sample from the middle or above it is still very clear. I think I am getting some residual settling out of gycerin and FFAs that did not completely react. I may have had some methanol left in the mixture.

It will be interesting to see how these batches look after more settling time in storage. Anyway, I think this completely waterless method of removing impurities is going to work. It needs some fine tuning but I plan to keep experimenting with it.

Thanks for your post.

Don

Don, What was your soap concentration (soap test) before and after the bubble wash. 200g of Magnesol seems awefully small amount to be using. When I titrated my 30 gallons, My starting soap was 609 PPM and ended up less than 87 PPM and that was using over 1Lb of Magnesol.
I real curious as to how you get away with such a small amount.

Respectively,

Ian

Graham,
When you said that you take the freshly reacted biodiesel out and put it straight into your wash tank, how long of a settling time had you given it?

How long should we let the mixture settle for before transferring it? One one hand heat conservation is nice, on the other I would have to have a bunch of unsettled glycerol in the wash tank that didn't need to be there.

Hi Curt,

I had drained off the glycerol before I transferred to the washtank. I do a 5% water prewash in the reactor. So, what I put into the washtank is BD ready to wash - only I don't put water in there to sart with.

-------
Caddytd: You made a good point about the pre-wash - I think you're right that the water introduced by the prewash could impair soap separation.

I'll do a new batch next weekend and will skip the prewash, so there will be no water at all in the BD or wash tank.

I'm also going to try KOH instead of NaOH, because I want to have a good long settling period (24 hours) in the processor to drain off as much glyc as possible. If I use NaOH I know for sure I'll block the whole thing up if I wait that long.

However, I've never used KOH before, so I don't know if the soap will coagulate like it does with NaOH - just have to find out.
-------

[ edit 20 June 06 - Distillation of the methanol would help to limit its release into the environment and would make it available for use in the next batch - GL]

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