Setting Fluidization Velocity: From Minimum Fluidization to Operating Window
1. Three Velocities, One Window
Operating gas velocity is a working point between two boundaries.
The lower bound is minimum fluidization velocity, Umf: below it the bed stops moving. The upper is terminal velocity, Ut: above it particles leave with the gas. Practice puts the operating point at 2-5 times Umf, set by whether the process needs heat transfer, residence time or tolerance for entrainment.
All three can be calculated, none accurately. Formulas give magnitude; the real window comes from cold-flow testing.
2. Umf: A Starting Point, Not a Setpoint
1. How to Estimate
Below a particle Reynolds number of about 20, Umf equals particle diameter squared times density difference times gravity, divided by 1650 times gas viscosity. For a porous carbon scaffold with D50 around 10-30 micrometres and bulk density 0.4-0.6 g/cm3, nitrogen at ambient gives 0.5-3 cm/s.
Expect 30-50% error: real powders have wide size distributions and moisture, and fines agglomerate, so measured Umf falls below calculated.
2. Umf Drops Sharply When Hot
Umf is inversely proportional to gas viscosity. Nitrogen viscosity roughly doubles between 20C and 600C, so Umf falls to about half.
Set velocity from ambient data and, hot, your multiple of Umf doubles: 3x becomes 6x. Set it from hot properties and start cold, and the bed sits dead. Size on hot properties, and run a separate transition velocity at start-up.
3. The Ceiling Is Not Ut
Terminal velocity is the free-settling speed of a single particle, a few centimetres per second for 10 micrometre particles in nitrogen. It seems operating velocity must stay below it, yet industrial units routinely exceed it: cyclones and a return leg capture entrained particles and send them back, so the loop holds inventory through circulation ratio.
So the ceiling becomes two figures.
Transport disengaging height, TDH: above it entrainment stops rising. Higher velocity means a taller vessel.
Circulation ratio: entrained over feed mass, usually 5-20. Push higher and particles cycle repeatedly through the hot zone and cyclone; attrition accelerates and the size distribution broadens.
The floor is explicit too. Below 1.2-1.5 times Umf, channelling and dead zones appear and thermocouples stop representing the bed.
4. On Scale-Up, Distribution Is the Real Problem
Velocity is uniform in a small unit, not after scale-up. As diameter grows, bubbles coalesce; large ones rise up the centre and starve the wall, producing a fast core and slow annulus: some particles under-coated, others over-coated. No velocity adjustment fixes this; it is a distributor and internals problem.
Three responses: staged gas injection, internals such as baffles or tube bundles, and multi-stage fluidization with narrower sections in series.
To judge whether a supplier has scaled a unit, ask: what is the distributor's open-area ratio and hole layout, and on what basis; was a cold-flow test run; how many bed thermocouples and where. No answers means the parameter sheet is a small unit with bigger dimensions.
Cold-flow testing is not optional. It measures what hot calculations cannot: actual Umf, the velocity where bubbling starts, the entrainment curve, and whether distributor pressure drop reaches above 30% of total bed pressure drop. The rig need not match production size, but its velocity range must cover one step either side of the operating point.
FAQ
Q: How many times Umf should the operating velocity be? A: Two to five times normally, five to eight when heat transfer must be intensified or material is sticky. Below 1.2 times you approach a dead bed.
Q: Why is measured Umf far below calculated? A: Agglomeration: clusters fluidize as larger particles and lower the apparent Umf. Wide size distribution pushes it lower still.
Q: Can I set velocity from ambient cold-flow data alone? A: Not directly, but the test stays mandatory. Measure Umf and entrainment cold, then correct for hot viscosity.
Keywords: minimum fluidization velocity, operating gas velocity, fluidized bed, entrainment, dead bed