Back to basics???

[Julian]

In an attempt to stop boredom at Tosser Towers ...

Been thinking about the way our processors work right from basics.

We circulate oil TYPICALLY using a Tam 105 and, for the sake of argument, a 100 litre processor.  We introduce Methoxide at the venturi where some mixing and presumably reaction takes place, then dump the oil into the processor, where due to the cross sectional area, the flow back to the pump suction is very low. This, assuming we are not blessed with the odd eductor or two.

What happens in the processor with the very low flow created?  My guess is that glycerine starts to drop out and fall, due to its density, to the bottom of the processor faster than the bio/oil mix.  At the bottom it meets a nice cone which concentrates it towards the pump suction.

In this scenario, we are trying to react Glycerine rich oil with the incoming Methoxide.  Once all the methoxide is introduced we are still circulating Glycerine rich oil through the point which gives the most effective mixing ... the pump.

Two stage or non titration processing will reduce the effect but not completely remove it.  Eductors too will improve matters, but limit the flexibility of the processor.

Not a new idea as it was being used before GL introduced us to the venturi, but what if we went back to adding the Methoxide a little way off the pump on the suction line?  You would require a valve and a vacuum gauge on the pump suction to ensure that the line was under negative pressure and then very slowly introduce the Methoxide over a protracted period via a needle valve to control the flow.  You would have to make sure the pump wasn't throttled to the point of cavitation as this could be potentially dangerous.  I recon this would give the most intimate mixing possible without incurring too much additional cost.

This arrangement may work but is still hampered but the Glycerine.  Again, not a new idea as I've seen it muted but dismissed by long standing members on the VOD, but what if there were some method of splitting the Glycerine from the oil flow before it returned to the processor?  I've got a couple of ideas, but can anyone come up with a clever bit of design?

[K.H]

I have always introduced the methanol before the pump as with not de mething i've never needed a venturi,if ive needed to introduce air i use the sight tube.
I shut the main valve slightly and open the meth valve,obviously the more i close the main valve the quicker the meth introduction,i have always gone with latest idea that it didn't matter how long,so as quick as possible,a bit like lovemaking in the Suffolk area so i hear,it should be easy enough to do quick and slow meth introductions and see the result.
I guess with the glyc using gravity in some way would be the obvious thing,something like an inline centrifuge or what about James "cyclone",how did that work? and what was it used on?
Another thought that doesn't get rid of the glyc but may aid mixing is that the meth has a tendency to float and the glyc wants to sink but as you say we take the thick from the bottom and put it in the top to immediately sink,would there be a benefit if just while mixing we took from the top and reintroduced it the bottom?

[Julian]

Didn't realise you had no venturi.  Have you ever tried seriously starving the pump of suction whilst introducing Methoxide?

When I say slow introduction of Methoxide I was reckoning on about half an hour and reckoning on catching the glycerine as part of the process.  The aim being to have, not a complete reaction at the pump discharge because I'm sure it will be oil rich, but to use as much of the Methanol / catalyst as possible in one pass ... not really explaining myself very well here but I think I know what I mean!

It's not really Jim's cyclone, nor Dyson (as he seem to claim) for that matter, the technology it's been around for a very long time.  I first came across it 30 years ago for separating water from a gas stream on oil tankers, and I'm sure it was around long before that.  But yes, that was one possibility.  Why not have the processor tank as the cyclone with the suction at the centre of the vortex?

Another is to replicate the effect in the processor, with a smaller tank before it gets to the processor.  Unless emptied regularly the tank would have to be sized to the batch, which, like eductors rather limits the flexibility of the system.

I already see this effect on my system, as my sample point Tees vertically down off the pump suction.  It always fills with glycerol early in the process, so glycerine will apparently drop out of a flow in a pipe.

Quite like the idea of sucking from the top.  Or sucking from the 20% level, leaving a dead volume at the bottom (still dictates batch size though). Didn't GLs latest design use a reverse flow, but possibly for a different reason ... need to check his site.

[K.H]

If i seriously starve the pump when introducing methanol i would of thought there is a fair chance of going into orbit,are you trying to kill me as well as James!
Anyway what about inline static mixers?,they would help with the first part of your idea.
Its a pity i haven't got my old set up if i angled the input i could get a vortex straight to the bottom,so good the pump would suck air.
Going off topic-With the reverse suction idea,what if it was piped both ways? run a 80% first stage as normal,allow to settle then use the reverse idea from just above the glys level with the remaining 20%,get a pass then back to normal to mix it all up,a sort of "intank two stage"
Back on track,the cyclone seems the better option of the two,i will have to have a think and see if i can think of anything else,think i might google "weir systems" and see how they work

[Tony]

Someone I know that makes bio (not on the forum but has done for years) always draws the Methoxide in before the pump.  He says if the oil is too hot and the Methoxide instantly starts boiling it makes quite a racket but no issues as no oxygen present so it still mixes and reacts quickly.

I don't think it matters how the Methoxide gets in so long as it is well mixed.  Really the venturi is just a convenient way of dosing slowly - the venturi is great for methanol recovery because it recirculates the same headspace gasses which reduces oxidisation.  In theory.  Though I still get dark bio if I overdo the heating/condensing with the venturi so other chemical processes may be at work.

IIRC the old JTF method included blowing the Methoxide into the processor vessel using an aquarium pump as well as introduction before the pump.

Many ways to skin the same cat.

[hifly]

i have added meth before the pump for the last couple of years, i have also cavitated the pump (tam105) loads of times and i'm not in orbit yet.

i have no venturi but still demeth via a bubble ring built into the processor and air pump. static mixers won't work as the meth is floating on top of the oil, what you want is good mixing in the top 6" of the liquid in the processor thats why eductors work.

sorry for the late reply.

[Julian]

Interesting.

Do add the Methoxide quickly in one big slug?

[kamaangir]

Surely if as I have read, most of the meth is in the glyc phase, you do not want to remove the glyc mid reaction (ie before all the catalyst is used up).

When speaking to Carrington, he was of the opinion that paddles are the way to go. They don't even have to rotate that fast. If I find something suitable with paddles already built in I might have a play and see if I can do some comparisons.

[Mubs]

I had an idea about taking the glycerol out of play.

What if you use a stand pipe to take the stuff away from the bottom of the cone?
If the standpipe is positioned several inches above the bottom of the cone, that gives several inches for the more dense substances to gather. These will effectively be taken out of the system as they won't be circulated.

[Rotary-Motion]

I had an idea about taking the glycerol out of play.

What if you use a stand pipe to take the stuff away from the bottom of the cone?
If the standpipe is positioned several inches above the bottom of the cone, that gives several inches for the more dense substances to gather. These will effectively be taken out of the system as they won't be circulated.

hey mubs good to see you chapperenee

answered ya internal email too 

[Tony]

I like the idea, I always thought of trying a 25l sealed container hanging off the bottom valve as a "glyc sump" which would collect the glyc as the process ran.  But your idea is significantly less likely to leak!

I did try single staging a couple of batches and draining off the glyc then just pump mixing the rest (same thing?) but that didn't seem to help completion, so perhaps keeping back 20% of the Methoxide is required for a second stage. And maybe the second stage works because if first stage glyc is drained/isolated, a it'll take with it the water from stage 1 Methoxide - which it won't do if it's still present but in some sump.

[Glycer-rides]

I had an idea about taking the glycerol out of play.

What if you use a stand pipe to take the stuff away from the bottom of the cone?
If the standpipe is positioned several inches above the bottom of the cone, that gives several inches for the more dense substances to gather. These will effectively be taken out of the system as they won't be circulated.

My heated and pumped wvo settle /glyc wash drum is set up in that way. The pick up is a swept bend in a tank connector on the side of the drum, so it's pick up height from the domed bottom can be varied. The idea was to get the optimum point where crud wouldn't jam the CH impellor but there would still be enough glyc from the previous 170L batch to wash effectively.

In theory I could brew bio in this tank. it could be worth trying a single stage like Tony did but draining off the denser 20L of glyc. However, I think that as this is in a dead zone, not being pumped, it would not be contributing to the reaction's equilibrium, so would not push it the way we want.
No harm in giving it a go, though. ATM, both the settle tank and the reactor are full of glyc washed wvo, so it would have to be next year!

[Mubs]

Tell us more about your swept bend buddy.

I take it you're able to get it to rotate? how does that work.

Cheers

[Glycer-rides]

Tell us more about your swept bend buddy.

I take it you're able to get it to rotate? how does that work.

Cheers

Really? OK...it's 22 mm copper pipe bent through 90 degrees to the 100 mm minimum radius achievable with pipe benders.
Yes, it can be rotated if the tank connector is loosened off and the pipe is 'poked with a stick', then the connector is nipped up again.

[Tony]

Presumably straight horizontal bit through tank connector (filed out to remove internal ridge) then 90 bend in tank and you just swivel (ahem!) appropriately.