Thickener Design for Light Particles

[TiCl4]

I am assisting a sister plant with debottle-necking production. They produce microsphere particles with a density of ~0.9 s.g. The particles also have a average particle size of 10-50 um. They currently use tall, skinny tanks (aspect ratio is ~1.5-1.8) after the spheres are created to thicken the tank (increasing solids from, say, 20-30% to 40-50%). They let the particles rise to the top over the course of 9-12 hours, then draw the relatively low-solids bottoms water until solids are in spec.

They only have a few of the thickening tanks, but the combination of throughput and different product grades means they have exceeded the capacity of the tanks - they often package these tanks into IBCs to allow for further thickening because they need the tank for an upcoming production run.

Thickeners are typically designed for settling solids, but the same principle for design should apply, correct? In this case, gravity will work against the buoyant force, but the particle should still have a rising rate. If I'm thinking about this properly, a wide and shallow thickener design should produce much higher throughput/lower thickening time than a tall/skinny thickener due to the lower travel distance you need for separation, correct?

Any tips on thickener design? Jar tests can give me rising rates for different products, but what other effects do I need to consider? Brownian motion? Other differences from general thickener design parameters?

 

[SuperSalad]

My experience is more in getting solids to remain in suspension, but it sounds like the wider tank should help. I'm not sure of other effects but one of the first thoughts I had is that you'd have less "head" of thickener per unit area pushing back against the buoyancy of the particles below it as they collect. That plus the shorter distance to rise like you mentioned make it sound plausible to me.

Is a centrifuge not applicable in this situation?

If it is, you may be able to continue using the existing tanks if you were able to install a centrifuge prior to the tanks in the process and possibly skip much of that early separation time. It sound like you are producing several grades from a single base stock, so you can set the centrifuge output to be near your minimum grade and that would then become your "base" if that is the case which should reduce your separation time for all the other grades and make your "low" grade ready right away.

Andrew H.

http://www.mototribology.com

 

[TiCl4]

The particles are polymers that are stabilized with surfactant. I'm not sure how stable they would be in a centrifuge - we try to avoid high-shear environments to avoid agglomeration.

The other issues is that there is no base stock - each grade is individually produced and thickened. I need to design a thickener that can handle a batch well within the given cycle time of the reactor.

 

[EdStainless]

A low speed centrifuge may be worth exploring. Even if you just took a small amount of the liquid out it would help.
What happens if you added slight aeration to assist the particles in rising?
You might try looking in mining references. Ores are crushed and then flotation is used to separate by density. I know that they have designs for flow-through tanks where product cascades from one tank into the next with each giving a little separation.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[TiCl4]

Any recommendations on centrifuge manufacturers? I haven't dealt with industrial centrifuges before.

 

[3DDave]

What volume of material is being introduced to the tanks?

 

[TiCl4]

Tiny - the process is a batch one that takes 10 hours to produce approximately 2,000 gallons. So a 10-20 gpm capacity would be more than sufficient.

 

[EdStainless]

You might search for centrifuges that are used in food products, many of those are low shear devices just meant to aid de-watering.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[jari001]

GEA and Alfa Laval are common names in pharma and food. I have a generally good impression about these companies, though I have only ever interacted with them on the pharma/biotech side of their operations.