Could such a simple change in methoxide introduction make such a big difference?

[mikethebike3]

Over the past few years I’ve made about 30 x 80 litre batches of Bio in my double-domed hot water tank derivative GL processor.  Most of these have been OK (eventually), and all of have been used with no problems.  More recently however, I’ve had a fair bit of trouble in getting my glycerol to separate fully (more methoxide & processing would eventually sort it), and dark bio foaming out in the closing stages of demething.

A week or so ago I started the first batch for a number of months (family issues…) which seemed to go surprisingly well.  A 15l glycerol wash returned only about 5 litres, but after the first stage of processing I got about 20 litres of well-separated glycerol out which was a great relief.  The 10/90, gave 7% dropout, meth mixed appropriately for stage 2 and processed again for a couple of hours (my normal), but meanwhile the stage 2 10/90 gave me a super-clear pass, no dropout, crystal clear liquid immediately the methanol was added.



I have never managed this before!  The nil dropout yes (with enough stages), but the mix has NEVER been remotely clear.

I lead a pretty simple life, so this has been quite a revelation!  Furthermore, I’m now demething at 70° and keeping a wary eye on the mix (through a 28mm compression penetration, so not ideal), but after the methanol boiled off (lots and lots of tiny but clear bubbles on the surface of the bio below) there’s now nothing visible on the surface.  No signs of foam in this batch!

It’s the same oil, the same KOH (4 years old but pretty dry), same methanol (including ~30% self-recovered), same processing temperature 55°, same everything.  The only difference is that instead of using the venturi to suck up the methoxide into the tank, this time I used the port next to the TAM105 pump inlet normally used to fill the processor with WVO or the glycerol wash.  I took about 5 minutes for the ~13 litres of meth to be introduced.  Given that this then forces the oil and the meth directly into the main mixing feature, i.e. the pump impellor (no eductors etc here), could this alone make such a huge difference?

Of course this is only one batch, but I’ll be doing another shortly, in exactly the same way.

Any comments gratefully received!

Mike

[Dickjotec]

Mixing is the key to fast production so, in your case, it could be that this is the way to go in future.

[dgs]

Hi Mike,

A quick and clear S2 dropout usually means that other contaminants are minimal. A long settling time after S1 often causes this, so the glycerol left in the mix is lower.

I don't apply any heat at all for S1 and leave it to settle overnight. I then use heat for S2 which I leave for 1 hour to mix at 50 degs. I can take a sample for a 10/90 15 mins after adding S2 methoxide which usually shows a clear methanol phase within 10 mins.

Using heat on S1 can cause self inflicted problems with a rush to get S2 underway so the mix doesn't loose too much heat. Doing it cold and leaving overnight greatly improves the process.

So my answer to your thread title is no. Always add your methoxide as quickly as possible for better reactions. Mine goes into the processor in 10 seconds.

[Head Womble]

If you want to see a really quick reaction try the AAF method (arse about face), you add the methoxide to the tank and then add the hot oil while the pump is running.

To give you an idea of the amount of mixing this gives in the first few minuets, picture adding a small amount of orange squash to a glass of water, then adding the squash first and topping up with water.

[mikethebike3]

Thanks for the feedback, it’s appreciated.  I realise that there are many different methodologies, but in my case, apart from the way the meth was added to the oil, pretty much everything I did and used in this latest batch was near identical to that which I’ve done and used in previous batches, yet the results were dramatically different.  Poor glycerol separation and (probably consequent) foaming of bio during demething have been a real pain in the past, then suddenly it all changes!

In the classical GL processor, I just picture the meth joining the stream of oil inside the venturi and entering the vessel, but (eductors etc. notwithstanding) afaiu most of the actual processing does not take place here.  With widely differing specific gravities and viscosities, isn’t it likely that they (the unprocessed oil and meth) will move through the processor tank in very different paths, arriving at the pump input in variable and often far from optimal ratios, whereas if the streams meet and immediately hit the impellor they are more likely to combine (and therefore process) more efficiently than in and just after the venturi?

Of course there are many on the forum who will understand these matters far better than I.  I’m just curious to explain how the results from this batch were so different!

Mike

[Tony]

Yes that is interesting, perhaps the pump mashing the two together avoids the methanol pooling on top of the oil when less violently introduced using a venturi.

I suppose a number of people are paranoid about pump mixing a flammable alcohol, though I do remember collecting meth from one chemical supplier from an IBC and he just used a standard TAM105 to dispense it.  Nothing ATEX rated. I'm not sure I'm that brave... for the sake of waiting a bit I think I'll stick with the venturi.

[Bio-boy]

I use a 12mm venturi to draw my methoxide into the processor. I have a rather large centrifugal pump with a flow capacity of 350L/min and 1.5" pipework. The agitation of chemicals to oil is rather aggressive but it does the trick exceptionally well. I have now amended my washing process adopting at least 2 pump agitation washes which has made a dramatic difference to drying times. So in answer to your questions I feel the pump plays a massive part in the reaction phase. The more aggressive the better.