Mixing vs. processing

[dgs]

I've saved at least 35 mins at 1.5 KW's You do seem to have 'hit' on something here Julian, in that the process seems to be continuing at a certain rate without mixing.

I'm doing a small S2 at the moment and keep stopping the pump and letting it rest for a while. Of course these S2's I do after I sometimes get a pass after S1 (ideally I like 0.3 mls d/o after S1 ) are something of a 'black art' in that processing
the mono's to completion are a bit of a guess.

[Julian]

Could we be heading back to the bucket and stick method everyone used to poo poo!

[dgs]

Strange you should say that. I've recently taken to processing the bio for the c/heating in a cubie, after adding the methoxide I shake it 3 times in the first half hour, then just leave it.

I normally don't bother testing for conversion but a couple of weeks ago I did a 10/90 which showed only 0.4mls d/o,  I was amazed.

Recipe is 15lts oil (not my best stuff) 150gms K and 250mls methanol. This time of year when we only use the boiler for hot water this method is so much easier than a 'proper' process.

[Julian]

Ok, so started the glycerine wash and I thought I'd try the start, stop trick with the glycerine.  A 2min run of the pump seems to produce a homogeneous mix, so I'll try using 2mins as a basis for the whole batch.

I have a much smaller processor that you, only 80ltr batches, so with the same pump it shouldn't need such a long run to mix to a similar degree.

Just thinking about the oil glycerine mix, both are quite viscous (this is being done with no heat, current processor temperature is 22°C) and so should separate more slowly and hence need less mixing.  Only issue may be separating the two at such low temperatures … any experience of this?

[Julian]

Answer my own question here ... the sample I took during the first mix has separated quite clearly, so I don't think draining the glycerine will be a problem.

[dgs]

As I glyc prewash in the IBC it is difficult to judge the separation time by the glycerol level, but by the colour of the oil getting lighter I would say it takes several days.

When I've done a prewash in the processor at low temperature it takes 3 days to get most of the glyc out.

From the outside (pre wash) IBC the oil is pumped into a 400 litre tank near the processor. It can sit there 3 to 6 weeks before being processed. Even then there is a few litres of glyc that has dropped to the bottom.

[RichardP]

I tried this on my last batch a week ago and got similar results:

Glyc washed 170L batch, very little conversion afterwards. I used 1600ml ASM and 650g Koh in 20 liters meth.Temp at start was 50c.

After getting in the methoxide I mixed for 5mins, 10/90 showed 1.7ml dropout, good glyc separation in the sample jar. Did another 10/90 on the same sample after 40 mins and this was 1.5ml dropout
1 hour later mixed for 5mins and 10/90 showed 1.2ml dropout
1 hour later mixed for 5mins and 10/90 showed 1.0ml dropout
After another hour and a 5 mins mix there was 1.0ml dropout so no further conversion, probably down to the chemicals needing adjusting upwards.
Carried on with a normal 2nd stage.

Interesting seeing the conversion after such a short overall mixing time as I normally process stage 1 for 45-60 mins for a similar final dropout.

[neisel]

I'd be interested to see the conversion %s if you were to reduce the dosage of chems by 25%. I suspect there'd be not much difference.

[Julian]

So, bio production is rather slow here at the moment … I'm just finishing off the batch I started a month or so ago!

I used my normal methanol and catalyst volumes and mixed for 2 mins, five times on the hour and ran two stages (with only an 80 ltr processor combined with Frankinpump I figured that should be sufficient mixing).

The result was a perfectly clear 3/27, so it seems to work.  I wonder if more gentle mixing, something like a bubbler would work, certianly less energy than Frankinpump, although I have had issues with bubbling air into oil adding water from atmospheric humidity.

[dgs]

A bio friend from the next village tried this a few weeks ago but it didn't work for him. He does however use a 105 that he says is worn out. From memory after his 1st 5 min mix his dropout was 6.5mls. The next two hours showed something like 5mls then 4.5mls. He didn't go any further and mixed for 1 hour to get a 2.0 dropout.

Julian, don't you think the methanol loss would be a problem if using a bubbler. I don't think R/H would be an issue.

[Glycer-rides]

Interesting posts. Couple of ideas here:

1. The next person to try this "quick mix, then settle" regime, could try this:
After the settles, drain off X% of the glyc, before the next mix. As it might favour the glyc/bio equilibrium in a good way?
I've no idea what X might be...

2. Are we 'over-pumping' things? Remember the original GL plans used a CH pump!

They fell from favour, possibly due to the low flow rate not working too well with venturis? (can't remember if there was any other  reason).

With three way valves, a CH pump could be plumbed in, in parallel to our pump of choice, and take over the donkey work of keeping things ticking over.

[Julian]

A bio friend from the next village tried this a few weeks ago but it didn't work for him. He does however use a 105 that he says is worn out. From memory after his 1st 5 min mix his dropout was 6.5mls. The next two hours showed something like 5mls then 4.5mls. He didn't go any further and mixed for 1 hour to get a 2.0 dropout.

Julian, don't you think the methanol loss would be a problem if using a bubbler. I don't think R/H would be an issue.

Just a gut feeling, but I recon you want to turn the batch over two or three times to get reasonable mixing … that was the thinking behind running Frankinpum for only two minutes.  Perhaps with a TAM 105 and a higher volume processor he may want to try running it for longer.

Good point, I didn't think about methanol loss.  On the standard GL design you could run the condenser to collect it, but that's just adding complications I guess.

I tried drying either oil or bio (too many years ago to remember) in the processor.  I made a connection low down to which I fitted a sintered brass air silencer and blew air from a small compressor through the hot bio/oil.  As far as I could tell it had no drying effect at all and I put it down to atmospheric humidity (could well be wrong) … UK humidity apparently ranges between 70% and 90%. 

Interesting posts. Couple of ideas here:

1. The next person to try this "quick mix, then settle" regime, could try this:
After the settles, drain off X% of the glyc, before the next mix. As it might favour the glyc/bio equilibrium in a good way?
I've no idea what X might be...

2. Are we 'over-pumping' things? Remember the original GL plans used a CH pump!

They fell from favour, possibly due to the low flow rate not working too well with venturis? (can't remember if there was any other  reason).

With three way valves, a CH pump could be plumbed in, in parallel to our pump of choice, and take over the donkey work of keeping things ticking over.

Certainly worth a try, but I've always wondered where the bulk of the unused chemicals ends up.  If they're in the glycerine it may be counter productive.

Some time back I had thoughts on trying to collect glycerine during processing.  My sample point is a short vertical leg off the mian circulation pipwork.  During processing glycerine collects in the leg, so when I take a sample the first slug is always settled glicerine ... thoughts were to have a drip feed off the leg, but not knowing where the chemicals were I never bothered trying.

At least with what you suggest you can lob the glycerine back into the mix if you don't get a good enough conversion.

[dgs]

Interesting posts. Couple of ideas here:

1. The next person to try this "quick mix, then settle" regime, could try this:
After the settles, drain off X% of the glyc, before the next mix. As it might favour the glyc/bio equilibrium in a good way?
I've no idea what X might be...

IMB tried this some years ago and reported his results on infopop. Basically it didn't work. The very thing that hinders the reaction in the final stages ie glycerol is the medium that contains most of the Methanol and catalyst. It needs to stay in the mix until an equibrilium is reached.

Some years ago I did a process for some c/heating bio. It was mid winter and I processed at ambient which from memory was no more than 10degs C. Just as a test I heated it the next day and re-mixed with it's glycerol which I hadn't drained. To my suprise I got no further conversion (was about 10% unreacted)

[Julian]

Just to add to this … I've tried leaving this batch circa 24hrs between two minute mixes on the last knockings of the second stage and the 3/27 test is showing an improvement each time.

Going to try both stages this way on the next batch.  I'm starting to think that not continuously pounding the glycerine into the bio may be a distinct benefit.

[dgs]

I gave this method another go a few days ago (one hour between 5min mixes)

Dropout results were; 1.0mls  0.5mls  0.3mls  0.1mls I then left it and carried on with S2 the next day.

I've held a cubie back from the water washing stage to try to dry at ambient with dried air from the turbo dryer. I'll post the results on the other thread.