Rectangular processor

[Tony]

I'm considering future space constraints and peering at two steel engine oil tanks I've got (about 500l each).

I've got 80lpm pumps and a Jim venturi and eductor for them.  I don't need to place the eductor near the surface as one pump will splash oil down through any floating methoxide layer, so the eductor can be positioned wherever in the liquid purely to turn it over.

Anyone got experience of using rectangular tanks as reactor vessels?

[Manfred]

Only seen two commercial bio producers. The first one was a 8000L cuboid. The main thing would be where to place the return to ensure no dead spots in the corners. You could play with a cubie and a hose pipe giving about 3 LPM flow  to see what happens in different places. If placed right you may even get better mixing with the corners creating extra turbulence.

 When I googled eductor last week I found they're used in plating to keep the metals from dropping to the bottom of the tanks. They're firing from the bottom upwards at cross angles to each other. Who says we have to input at the top.

[photoman290]

if you wanted to be really nerdy you could find a free CFD program and model it. i am sure someone has already done that. finding the data maybe a problem. autocad do a CFD rogram on a monthly rental basis. cant remember what it is called though. could find out if you are interested.

[Jamesrl]

Who says we have to input at the top.

We place our eductors just under the surface firing down for a specific purpose just as the platers place theirs for the best results.

[dgs]

Someone off infopop once told me not to have the eductor pointing down, they were wrong. If the pump is man enough, the processor contents will 'turn round' in a vertical direction.

Using weak pumps in the past with no eductor, I got used to seeing the methoxide layer sitting on top of the oil for a while until it got mixed in. Nowadays the methoxide goes in within 10 seconds through my 1" venturi and the actual mixing of methoxide/oil is instant, turning the whole mix opaque for about 3 mins until the reaction is complete enough for the excess methanol to be miscible.

See, us plastic boys can see this whereas youse metal boys can't.

Getting back to the square tank thing, if the mixing is adequate I can't see it making any difference.

[Julian]

I've been intending to make a square section processor and settling tanks for some time to try and maximise the space in my 6' x 6' shed.  I got as far as buying a couple of large steel tanks but lost interest when I saw the price of plasma cutters.

There are some really cheap ones on Ebay now circa £128 so I may summon up the enthusiasm to rekindle the project.

I was intending to make the cross section as near to square as possible exactly for the mixing reason.  I also considered cutting a pipe into quarters length ways and welding into the corners (void would need venting to atmosphere) all just thoughts at the moment!

[neisel]

My processor is a 150L rectangular box, it's the only one I've ever used.

What do you want to know?

[therecklessengineer]

Who says we have to input at the top.

My return is about 2/3 the way up the tank, and tangential. The liquid spins in the tank.

[Tony]

My processor is a 150L rectangular box, it's the only one I've ever used.

What do you want to know?

Just the assurance that it'll work without massively slowing the reaction - not that this really matters with a large enough batch size.

Do you have an eductor, and if so, where is it fitted?

[Chug]

Tony,

I have a squarish stainless steel tank with cone bottom I was thinking of making into a processor, but its sat there for ages so if you are interested I'll try and get a pic, looking or £50 for it.

[neisel]

My processor is a 150L rectangular box, it's the only one I've ever used.

What do you want to know?

Just the assurance that it'll work without massively slowing the reaction - not that this really matters with a large enough batch size.

Do you have an eductor, and if so, where is it fitted?

No eductor & no problem getting a 90% reaction (+ or -) every time. 1st stage mixes for an hour, maybe 75 mins max.

[Jamesrl]

Who says we have to input at the top.

My return is about 2/3 the way up the tank, and tangential. The liquid spins in the tank.

According to chemical engineers that the worst thing you can in a reactor.

[Tony]

Tony,

I have a squarish stainless steel tank with cone bottom I was thinking of making into a processor, but its sat there for ages so if you are interested I'll try and get a pic, looking or £50 for it.

I've seen your stainless tank in the field, it's a beauty.  I've got two steel tanks (at least I know I can weld mild steel) and although not as swanky I can stick em on blocks and they'll do just fine