Biopowered - vegetable oil and biodiesel forum
Biodiesel => Chemistry and process => Topic started by: julianf on April 26, 2013, 09:53:30 AM
-
Feedstock -
Mix of real grim kebab fat (semi liquid) and some beef tallow (very hard)
Process -
Preheated in water heated pre-heat tank. No pumping, so the water / bits of meat fall out nicely.
Into reactor. Added 60 ltrs of virgin koh glyc. Pumped for longer than normal / drained. Remember being surprised at how clear the glyc looked, and thinking what a good job the new preheat drum was doing. Oil after the glyc was quite dark, but nice and clear.
300 ltr batch.
Normally i try and maximise the first stage, but, this time, i thought id do a smaller first stag, and larger 2nd. So -
40 ltrs of virgin meth
1600g of koh
(ie i was looking at reacting 200 ltrs of imaginary titration 1 oil - ie 20% of 200 / 200 x 8g)
At that point its all gone wrong.
I thought maybe my venturi had got blocked (again) so i took that apart, and its clear.
So i think its a flow issue. Either something is wrong with the pump, or it cant handle the oil that now seems to look like whites, even though the temp is still 64c.
Some of this tallow needs to be real hot to flow, but the oil was clear before i added the methoxide, so im confused. Ive got the batch climbing to 70c, to see what happens then.
I have more methoxide mixed, such as to make it a 1 stage / 300 ltr batch, but with the venturi not sucking, im going to have issues getting it in the reactor.
-
Sounds to my like an ineffective dry regime, adding the lye may have created a form of emulsion/soap.
WHY do people rely so heavily on glyc pre-wash.
IF or when I do do a glyc wash I always dry my oil first to gain the greatest advantage from the phase.
-
Preheated in water heated pre-heat tank. No pumping, so the water / bits of meat fall out nicely.
-
Oil has cleared again now, so im sure it was just a temperature issue.
Your venturi, however, still is not sucking, and i have no idea why.
-
If a venturi doesn't work it'll be down to the pump or pipework.
-
Not sure if this is the problem but I have found a first stage of at least 75% conversion is needed in my instances to get any glyc to drop at all. if your first stage was only aiming at 60 odd percent then maybe this could be causing issues.
-
Not sure if this is the problem but I have found a first stage of at least 75% conversion is needed in my instances to get any glyc to drop at all. if your first stage was only aiming at 60 odd percent then maybe this could be causing issues.
I have to agree with you there Steve, had a 70% first stage on my last batch and the glyc separation took quite a while, any less and I don't think it would have split.
I'm using the cruddy WVO I couldn't use in the colder weather.
-
Sounds to my like an ineffective dry regime, adding the lye may have created a form of emulsion/soap.
WHY do people rely so heavily on glyc pre-wash.
IF or when I do do a glyc wash I always dry my oil first to gain the greatest advantage from the phase.
I rely totally on glyc washing to dry my oil....but I have the advantage of knowing my oil, as I suspect many homebrewers do. Having another drying method in the process, would to me, be a complete waste of time and effort. Others are different, I know, but there are many like me who just use glyc washing.
Every now and again though a spanner gets lodged in the works...just like Julian was experincing when he opened the thread.
-
My last batch was heated and settled, then condensed off at 95°, then gly washed,
and I still had damp oil, it looked good but gave only a fraction of the convertion I was expecting,
no gly dropped either, I had to add more gly (another 20L) to get anything to drop.
This was mostly the crud I've stockpiled over winter.
So sometimes no matter what you do you can't remove all the water.
In the end I had to do three stages but ended up with good bio.
-
Ive now got pretty much 100% conversion, but the glyc wont drop.
I do wonder if its the silly amount of meth that ive put in (30% in total - multiple stages) holding the glyc in suspension.
I now have about 400ltrs of this *stuff* - i think it may be time to give up, and can it (and then feed it into later batches if it still does not settle)
I have considered bringing in a sample to demeth on the cooker (to see if that makes the glyc drop) but, to be honest, i dont much like that idea.
-
Ive now got pretty much 100% conversion, but the glyc wont drop.
I do wonder if its the silly amount of meth that ive put in (30% in total - multiple stages) holding the glyc in suspension.
I now have about 400ltrs of this *stuff* - i think it may be time to give up, and can it (and then feed it into later batches if it still does not settle)
I have considered bringing in a sample to demeth on the cooker (to see if that makes the glyc drop) but, to be honest, i dont much like that idea.
I once had a conversation with Graham laming and excess meth is a reason for glyc to stay in suspension, perhaps if you were to demeth the batch a bit then it may drop.
-
Ive now got pretty much 100% conversion, but the glyc wont drop.
I do wonder if its the silly amount of meth that ive put in (30% in total - multiple stages) holding the glyc in suspension.
I now have about 400ltrs of this *stuff* - i think it may be time to give up, and can it (and then feed it into later batches if it still does not settle)
I have considered bringing in a sample to demeth on the cooker (to see if that makes the glyc drop) but, to be honest, i dont much like that idea.
I once had a conversation with Graham laming and excess meth is a reason for glyc to stay in suspension, perhaps if you were to demeth the batch a bit then it may drop.
Or add some more gly, this should swing the balance in your favor.
-
Ive split the batch and left about 150 ltrs in the reactor, which i am now WBD'ing.
Ill take a sample in a while, and see if the glyc falls. If it does, ill settle and drain, and then continue with normal demeth.
Thank you to everyone for the comments so far. Will update as i progress.
-
I promised an update...
Bad to worse.
I split the batch and started WBD on half of it. I came back to carnage.
My conclusion was that the batch thickened as demeth progressed, which lowered the flow, caused the inline heater to cavitate, and the 'pulsing' caused the pipework to 'blow' off the venturi inlet.
We have sackfulls of sawdust / bio mix now.
-
What a mess! :'(
-
I know it's a bit late now to be offering advice and you probably wouldn't have took it anyway but if something has gone wrong and it needs a fix don't go into it with your blinkers on. Do a jug sample first to get your "fix" and then scale it up to your batch size. Other wise, as you now have, you end up with a bigger mess.
It's really simple....for some.
-
Would some acid in there have made any difference? Would that have split it?
Still trying to learn... :-\
-
Who knows.
I think i may feed the remainder of the batch in with fresh wvo on the glyc wash stage. As mentioned, ive got full conversion now, but the part that was demethed turned to jelly.
The feedstock of this lot was collected two summers ago, and had more meat in it than would fit in a tesco carrier bag - i really want to get my titration kit back (or make another) as i think it would be worth using on batches like this, to adjust the first stage of the 2-stage.
This batch has all the symptoms of a wet batch, however, i spent more time prep'ing it than normal (both with long cycle heat and settle, and with glyc wash) so im really not convinced it was wet. I suspect just bad chemical adjustment, probably underdosing on catalyst on the first stage, and then overdosing on meth thereafter.
My usual method is to dose with 15% of the whole batch methanol, and then catalyst to a supposed titration of 2 [koh] (again for the whole batch), and then finish up with 20%+base for the remaining on s2 (which is generally under 10%) - this probably wasnt the time to change the method!
In all the time ive been brewing ive had a total of two jelly batches (this being one of them) and one failed water washing attempt, so, hey, some you win, some you loose (just need to make sure you win most of them!)
-
Whatever is wrong with this batch Julian you made a big mistake by not doing a jug sample to find a fix.
You claim a 100% conversion yet there is something fundamentally wrong with the way you conducted the 27/3. ie you have no glyc drop out so your testing that also..
Is having no glyc dropping out constistant with a "jelly" batch? I ask becuase I've never had this issue. Not to say that I won't at some point in the future.
You have an ideal opportunity to really find out what's going on here because you have half the batch to play with.
How much methanol did you recover before it went tits up?
Your brew shed must have been on the point of blowing up with all of that hot, methanol rich bio all over the floor.
-
Would some acid in there have made any difference? Would that have split it?
Still trying to learn... :-\
Different things work for different problems. The key is doing some samples in a jug to find out what does work and what doesn't. I had a slight emulsion yesterday after adding the water wash at the end of process. When I tried droping the glyc before pumping into the wash tank i could see there was something not quite right. I mixed it all up again and put some in a jug. I added a bit of stage one glyc to it and mixed it with a thermometer. It returned back to what I would normally expect. So I added 8lts of glyc into the mix and hey ho my problem solved. Easy peasy squeezy lemons
-
When I produced my jelly batches the gly did not drop, but I only got to the first stage when it thickened up.
One of these batches is now sitting in Julians garden, so he'll be able to give more info on how he saves it.
-
So far, i have tried -
1)Heating in a pan, to make certain there is no methanol remaining.
Result - cooled to the same thickness, or, possibly, even thicker
2)Warming, and then mixing,50:50, with nice, road-reddy, bio.
Result - 2x as much jelly.
The result that i was wanting with this second test was that, if the soaps were diluted enough that jelly did not form, then they would have a chance to settle out. And then, if need be, the top (settled) layer could then be used to 'cut' more jelly.
I still think there is hope in this plan. Either by cutting with a higher percentage, or, i also thought -
Would the soaps settle from jelly if the batch was kept molten for long enough? Or would the convection currents stop the soaps settling and / or precipitating out?
-
1)Heating in a pan, to make certain there is no methanol remaining.
How hot did you get it? Did any bubbles form? Smell?
I'm trying to correlate against my "heat it until it sorts itself out" approach!
-
Im sure i wrote it somewhere.... (but evidently not on this thread)
I put it in a pan on the stove. Heated it, using the IR thermometer to keep an eye on what was going on.
At about 55 ish, i got a few bubbles forming on the surface, but, really, not that many. They dissipated fairly quickly, and i took the mix up to about 95c, and then stopped.
I think that there was a very small amount of meth remaining, however, not that much, and getting rid of it (unfortunately) made little difference - if anything the mix may have been slightly thicker since.
-
I thought 27/3 worked regardless of glyc as it is miscible with methanol
-
Right, an update -
The 100ml of jelly that i 'cut' with 100ml of good bio resulted in 200ml of jelly...
...but i still had hopes for it!
I put it back in a pan, water bath style, and heated it up again, and let it cool slowly (as the water retained the heat with it)
Result -
about 50ml of white soapy jelly / haze in the bottom, and 150ml of good bio on top. So, taking into account that id added 100ml of good bio, thats 50% recovery.
I then did the same again, and, sure enough, less layer at the bottom.
Ive did it one more time today, and i now have about 20ml in the bottom of the beaker. So that's 200ml, of which 100ml is the good that i added, and 80ml is the recovered.
Considering i did not take any glyc out of this batch at all, id be happy with 80% product (its certainly better than nothing!)
The 20ml now looks more like glyc than the soapy haze too. Ie its dark in colour.
I do not know the consistency of any of this, as im not shaking the beaker up at all. But the top is liquid, and, unless i have perfectly clear jelly lower down, id say its liquid to the glyc layer.
I will tip it out tomorrow to be sure.
From this test, im starting to think that, possibly, jelly batches can be cured by keeping the mix above the jelly melting point, and waiting for the soaps to settle whilst its hot. I cant do this in my processor, as the heater is inline, so i think im going to try chucking a radiator in a clip-top, and, hopefully, recovering the rest of the batch like that.
-
Julian, I've forgotten the start of this threadand being a lazy SOB I'll ask to save me reading it all ... At what stage in the production process did the soap/jelly form?
-
Sounds like a good time to get the water heating coil installed in your processor - then you can heat up the jelly with no circulation needed (except the water through the coil of course)
-
Julian, I've forgotten the start of this threadand being a lazy SOB I'll ask to save me reading it all ... At what stage in the production process did the soap/jelly form?
Demeth.
But the batch was having issues prior - glyc would not drop. Then on demeth, it thickened up, and started splattering out the suction port of the venturi, as my inline heater cavitated due to lack of cooling flow.
-
Julian, I've forgotten the start of this threadand being a lazy SOB I'll ask to save me reading it all ... At what stage in the production process did the soap/jelly form?
Demeth.
But the batch was having issues prior - glyc would not drop. Then on demeth, it thickened up, and started splattering out the suction port of the venturi, as my inline heater cavitated due to lack of cooling flow.
OK, that's a little different to the two batches I made and the one Womble has dumped in my back garden. Those went to jelly as soon as the meth/ASM was added. The fix for those appears to be a good mix with glycerine while heating to about 65°C, settle, drain glycerine and then process as normal, recon I got about 90% yield from the two batches I've done so far, yet to try Marks stuff in the processor.