Minimum fluidization and terminal velocities, Geldart groups and regimes, distributor design, Davidson bubbles, Kunii–Levenspiel reactor model, TDH and elutriation from the ASTU CHE1816OE22 syllabus (2018-19 scheme).
What is inside
- Minimum fluidization and terminal velocity
- Geldart groups, regimes and distributor design
- Bubbles, clouds and the Kunii–Levenspiel model
- TDH, entrainment and elutriation
Syllabus: ASTU CHE1816OE22 Fluidization Engineering, semester 6, 2018-19 scheme (open elective 2). Official syllabus PDF.
Introduction and applications
Topics: Introduction to fluidized-bed systems; fundamentals of fluidization; industrial applications: physical operations; synthesis reactions; cracking and reforming of hydrocarbons; gasification, carbonization, gas–solid reactions, calcining and clinkering.
What fluidization is and why industry uses it
When gas (or liquid) flows upward through a bed of particles, the pressure drop rises with velocity as in any packed bed. At the minimum fluidization velocity the drag on the particles balances their weight less buoyancy; the bed loosens and the particles become suspended, so the bed behaves like a liquid: it finds its own level, objects float in it, and solids can be drained out like a liquid. Beyond this point extra gas usually passes as bubbles (aggregative fluidization), while liquid-fluidized beds tend to expand smoothly (particulate fluidization).
Fluidized beds offer vigorous solids mixing, so the bed is nearly isothermal even for strongly exothermic reactions; very high heat-transfer coefficients to walls and immersed tubes (200–600 W/m²·K for gas beds); the ability to add and remove solids continuously, and to circulate catalyst between reactor and regenerator; and the use of fine particles with little pore-diffusion resistance. Drawbacks: gas by-passing in bubbles lowers conversion compared with plug flow; back-mixing of gas and solids; attrition of particles and erosion of internals; entrainment of fines, needing cyclones; and scale-up is difficult.
Applications. Physical operations: drying of granular solids (fertilisers, pharmaceuticals, PVC), coating and granulation, heat exchange, adsorption, and mixing. Synthesis reactions: acrylonitrile by ammoxidation of propylene (SOHIO), maleic anhydride, phthalic anhydride, Fischer–Tropsch synthesis, polyethylene and polypropylene in gas-phase fluidized-bed reactors (UNIPOL), vinyl chloride by oxychlorination. Cracking and reforming: fluid catalytic cracking (riser reactor with fluidized regenerator), fluid coking and Flexicoking, catalytic reforming. Gasification and combustion of coal and biomass (bubbling and circulating fluidized-bed combustion boilers, CFBC, which allow low-grade coal and in-bed sulphur capture with limestone), carbonization, roasting of sulphide ores, calcining of limestone and alumina, and clinkering.
- Fluidized bed behaves like a liquid.
- Advantages: isothermal, high heat transfer, continuous solids handling.
- Drawbacks: by-passing, back-mixing, attrition, erosion, entrainment.
- Uses: FCC, UNIPOL PE, acrylonitrile, CFBC boilers, dryers, calcining.
Example: 500 kg of sand (ρ_s = 2600 kg/m³) is fluidized in a 1 m diameter column by air. What pressure drop must the gas supply across the bed?
- Area A = π × 1²/4 = 0.785 m².
- Δp = W(1 − ρ_g/ρ_s)g/A ≈ 500 × 9.81/0.785.
- = 6.24 kPa.
Answer: About 6.2 kPa, whatever the gas velocity above u_mf.
Example: Why can a fluidized-bed reactor run a highly exothermic reaction (e.g. ammoxidation of propylene, ΔH ≈ −515 kJ/mol) almost isothermally?
- Rapid solids circulation carries heat from hot spots throughout the bed.
- Particles have large heat capacity compared with gas, and the bed-to-tube coefficient is high.
- Cooling coils immersed in the bed (raising steam) remove the heat.
Answer: Solids mixing and high heat-transfer coefficients keep the bed within a few degrees.
Fluidization: at u_mf the drag on particles equals their buoyed weight and the bed behaves like a liquid. Particulate (liquid–solid) vs aggregative (gas–solid, bubbling) fluidization. Advantages: near-isothermal operation, high heat transfer, continuous solids flow and circulation, fine particles. Disadvantages: gas by-passing, back-mixing, attrition, erosion, entrainment, scale-up. Applications: drying, coating, granulation, adsorption; synthesis (acrylonitrile, maleic and phthalic anhydride, Fischer–Tropsch, polyolefins); FCC, fluid coking; coal and biomass gasification and CFBC boilers; roasting, calcining, clinkering.
- Pressure drop after u_mf equals bed weight per area.
- CFBC boilers capture SO₂ with limestone in the bed.
- UNIPOL makes polyethylene in a gas-phase fluidized bed.
- Cyclones return entrained fines.
Key formulas
- Δp = W(1 − ρ_g/ρ_s)g/A
- Δp/L = (1 − ε)(ρ_s − ρ_g)g
Practice questions
At minimum fluidization the pressure drop equals
- zero
- the bed weight per unit area (buoyed)
- twice the bed weight
- atmospheric pressure
Show the answer
Answer: B. the bed weight per unit area (buoyed) Force balance.
A major advantage of fluidized beds is
- plug flow of gas
- near-isothermal operation
- no attrition
- easy scale-up
Show the answer
Answer: B. near-isothermal operation Solids mixing.
A major disadvantage is
- poor heat transfer
- gas by-passing in bubbles
- no solids flow
- high pressure drop that grows with flow
Show the answer
Answer: B. gas by-passing in bubbles Bubbles.
UNIPOL is a fluidized-bed process for
- ammonia
- polyethylene
- sulphuric acid
- urea
Show the answer
Answer: B. polyethylene Gas-phase polymerisation.
In CFBC boilers, SO₂ is captured by adding
- ammonia
- limestone
- activated carbon
- water
Show the answer
Answer: B. limestone In-bed desulphurisation.
Theory questions
Discuss the advantages and disadvantages of fluidized beds as gas–solid contactors. (6 marks)
Show the model answer
Advantages: smooth liquid-like flow of solids allows continuous automatic operation and circulation between vessels; rapid mixing gives nearly uniform temperature, avoiding hot spots; high heat- and mass-transfer rates between gas and particles and to surfaces; small particles reduce pore-diffusion resistance; suitable for large-scale operation. Disadvantages: gas flow pattern hard to describe and bubbles by-pass solids, lowering conversion; rapid solids mixing gives non-uniform residence time of solids; attrition produces fines that must be recovered; erosion of pipes and vessels; friable solids and sticky solids are unsuitable; non-catalytic high-temperature operation can cause agglomeration and sintering; scale-up from small units is uncertain.
Describe three industrial applications of fluidized beds. (6 marks)
Show the model answer
Fluid catalytic cracking: VGO cracks over fine zeolite catalyst in a riser (fast fluidization); spent catalyst is stripped and regenerated in a bubbling/turbulent fluidized bed by burning coke; circulation transfers heat. Fluidized-bed combustion/gasification: coal, lignite or biomass burns in a bed of sand and limestone at 800–900 °C; low temperature limits NOx, limestone captures SO₂; circulating beds (CFBC) recycle solids through cyclones. Drying: wet granular solids (fertiliser, pharmaceuticals, food) are fluidized by hot air; intense mixing gives uniform, fast drying without overheating. Others: acrylonitrile synthesis, polyethylene (UNIPOL), roasting of zinc sulphide, calcining of alumina.
Gross behaviour of fluidized beds
Topics: Minimum and terminal velocities; types of fluidization; mapping of fluidization regimes; design of distributors; voidage in fluidized beds; transport disengaging height; variation of size distribution with height; power consumption.
Minimum fluidization and terminal velocities; regimes and Geldart groups
At minimum fluidization the bed pressure drop from the Ergun equation equals the buoyed bed weight per area. In dimensionless form: (1.75/(ε_mf³φ))Re_mf² + (150(1 − ε_mf)/(ε_mf³φ²))Re_mf = Ar, where Ar = d_p³ρ_g(ρ_s − ρ_g)g/μ² is the Archimedes number. Because ε_mf and φ are often unknown, Wen and Yu’s correlation is widely used: Re_mf = √(33.7² + 0.0408Ar) − 33.7. For small particles u_mf ∝ d_p²; for large ones u_mf ∝ √d_p. The terminal velocity u_t (Haider–Levenspiel correlation, or Stokes/Newton laws) sets the upper limit of operation; u_t/u_mf is about 90 for fine particles and about 10 for coarse ones.
As velocity increases a gas–solid bed passes through regimes: fixed bed; minimum fluidization; (for Geldart A powders) smooth expansion up to the minimum bubbling velocity; bubbling fluidization; slugging when bubbles approach the column diameter (tall, narrow beds); turbulent fluidization, where bubbles break up and the bed surface becomes diffuse; fast fluidization, with solids recirculated through cyclones (CFB, FCC risers); and pneumatic transport. Regime maps (Grace, Kunii–Levenspiel) plot dimensionless velocity u* against size d*. Geldart’s classification by d_p and ρ_s − ρ_g: group C (cohesive, < ~30 µm, hard to fluidize: flour), A (aeratable, 30–150 µm and low density: FCC catalyst; expands before bubbling, bubbles split, slow de-aeration), B (sand-like, 150–1000 µm: bubbles form at u_mf and grow), D (large and dense: grains, coffee beans; spouting).
Bed voidage rises from ε_mf (0.4–0.5) as velocity increases; in a bubbling bed (1 − ε_f) = (1 − δ)(1 − ε_mf), with δ the bubble fraction. The distributor (perforated, porous, nozzle/bubble-cap plates) must give even gas flow and support the bed; its pressure drop is designed at about 20–40 % of the bed pressure drop (minimum ~3.5 kPa), from which the orifice velocity u_or = C_d√(2Δp_d/ρ_g) and the number of orifices N_or = u₀/((π/4)d_or²u_or) follow. Above the bed lies the freeboard; ejected particles fall back up to the transport disengaging height (TDH), above which the entrained solids flux is constant; the size distribution in the bed coarsens over time as fines elutriate. Power consumption is the blower power to supply gas at the distributor + bed + downstream pressure drops.
- Ar = d_p³ρ_g(ρ_s − ρ_g)g/μ².
- Wen–Yu: Re_mf = √(33.7² + 0.0408Ar) − 33.7.
- Regimes: fixed → bubbling → slugging → turbulent → fast → pneumatic.
- Geldart C, A, B, D.
- Distributor Δp ≈ 0.2–0.4 bed Δp.
Example: Find u_mf for 300 µm sand (ρ_s = 2600 kg/m³) in air (ρ = 1.2 kg/m³, μ = 1.8 × 10⁻⁵ Pa·s) by Wen–Yu.
- Ar = (3 × 10⁻⁴)³ × 1.2 × 2599 × 9.81/(1.8 × 10⁻⁵)² = 2549.
- Re_mf = √(1135.7 + 104.0) − 33.7 = 35.21 − 33.7 = 1.51.
- u_mf = 1.51 × 1.8 × 10⁻⁵/(1.2 × 3 × 10⁻⁴) = 0.0755 m/s.
Answer: About 7.5 cm/s (Geldart group B).
Example: For the same particles, the terminal velocity is about 2.1 m/s. What is the operating window and what happens at u₀ = 0.3 m/s?
- u_t/u_mf = 2.12/0.0755 ≈ 28.
- u₀/u_mf = 4: a bubbling bed; the 300 µm particles stay in, but any fines with u_t < 0.3 m/s will be entrained.
Answer: Window ≈ 28 × u_mf; at 0.3 m/s the bed bubbles vigorously.
Example: Design a perforated plate for u₀ = 0.3 m/s, bed Δp 6.2 kPa, air 1.2 kg/m³, 2 mm holes, C_d = 0.6.
- Δp_d = max(0.3 × 6.2, 3.5) = 3.5 kPa.
- u_or = 0.6√(2 × 3500/1.2) = 45.8 m/s.
- N_or = 0.3/((π/4)(0.002)² × 45.8) = 2084 per m² ⇒ 1637 holes in a 1 m bed, triangular pitch ≈ 23.5 mm.
Answer: About 1640 holes of 2 mm at 23.5 mm pitch.
u_mf from Ergun at incipient fluidization (needs φ, ε_mf) or Wen–Yu correlation in terms of Ar; u_t from Haider–Levenspiel; operating window u_t/u_mf. Regimes: fixed, minimum fluidization, bubbling, slugging, turbulent, fast, pneumatic transport; regime maps. Geldart groups C, A, B, D. Bed voidage and expansion. Distributors: types, pressure drop 20–40 % of bed Δp, orifice velocity and number. TDH and freeboard; size distribution change by elutriation; blower power.
- Fine particles: u_mf ∝ d_p²; coarse: ∝ √d_p.
- Group A beds expand before bubbling.
- Slugging in tall, narrow beds.
- Distributor too weak → channelling and dead zones.
Key formulas
- Re_mf = √(33.7² + 0.0408Ar) − 33.7
- u* = [18/d*² + (2.335 − 1.744φ)/d*^0.5]⁻¹
- u_or = C_d√(2Δp_d/ρ_g)
- N_or = u₀/((π/4)d_or²u_or)
- (1 − ε_f) = (1 − δ)(1 − ε_mf)
Practice questions
FCC catalyst belongs to Geldart group
- A
- B
- C
- D
Show the answer
Answer: A. A Aeratable.
Sand of about 300 µm belongs to Geldart group
- A
- B
- C
- D
Show the answer
Answer: B. B Sand-like.
The Wen–Yu correlation gives
- terminal velocity
- minimum fluidization velocity
- bubble size
- TDH
Show the answer
Answer: B. minimum fluidization velocity Re_mf from Ar.
Slugging occurs when bubbles grow to about
- the particle size
- the column diameter
- 1 mm
- the TDH
Show the answer
Answer: B. the column diameter Wall-limited.
A distributor’s pressure drop is usually designed at about
- 1 % of bed Δp
- 20–40 % of bed Δp
- 200 % of bed Δp
- zero
Show the answer
Answer: B. 20–40 % of bed Δp For even distribution.
Theory questions
Derive the equation for minimum fluidization velocity and discuss the Wen–Yu simplification. (6 marks)
Show the model answer
At incipient fluidization the frictional pressure drop through the bed equals the apparent weight per unit area: Δp/L_mf = (1 − ε_mf)(ρ_s − ρ_g)g. The Ergun equation gives Δp/L = 150μu(1 − ε)²/(ε³φ²d_p²) + 1.75ρ_gu²(1 − ε)/(ε³φd_p). Equating and multiplying by ρ_gd_p³/(μ²(1 − ε_mf)) gives (1.75/(ε_mf³φ))Re_mf² + (150(1 − ε_mf)/(ε_mf³φ²))Re_mf = Ar. Wen and Yu found 1/(φε_mf³) ≈ 14 and (1 − ε_mf)/(φ²ε_mf³) ≈ 11 for many systems, giving Re_mf = √(33.7² + 0.0408Ar) − 33.7, which needs no φ or ε_mf. Limits: Re_mf < 20: u_mf = d_p²(ρ_s − ρ_g)g/(1650μ); Re_mf > 1000: u_mf² = d_p(ρ_s − ρ_g)g/(24.5ρ_g).
Explain Geldart’s classification of powders. (5 marks)
Show the model answer
Geldart (1973) classified powders by mean size and density difference by their fluidization behaviour with air at ambient conditions. Group C: very fine (< ~30 µm), cohesive (interparticle forces exceed drag): channels and lifts as a plug, hard to fluidize (flour, cement). Group A: small (≈ 30–150 µm) and/or light: bed expands homogeneously between u_mf and u_mb before bubbling, bubbles split and coalesce frequently with a maximum stable size, gross circulation, slow de-aeration (FCC catalyst). Group B: ≈ 150–1000 µm, sand-like: bubbles form as soon as u > u_mf and grow by coalescence; bed collapses quickly when gas is cut. Group D: large and/or dense (> ~1 mm): bubbles rise slower than interstitial gas, beds tend to spout (grains, coffee beans, ores).
Analysis of bubble and emulsion phases
Topics: Davidson’s model; frequency measurements; bubbles in ordinary bubbling beds; flow model in bubbling beds; emulsion phase: experimental findings; bubbling-bed model for the emulsion phase.
Davidson’s model, bubble growth and the emulsion phase
Davidson’s model treats a bubble as a solids-free sphere rising through an emulsion that behaves as an incompressible fluid (the particles) through which gas percolates by Darcy’s law, with the emulsion at minimum fluidization. A single bubble rises at u_br = 0.711√(gd_b), like a gas bubble in a liquid. The key comparison is u_br with the interstitial gas velocity in the emulsion, u_f = u_mf/ε_mf. If u_br > u_f (fast bubble, common for Geldart A and B), gas leaving the top of the bubble is swept round and returns to its bottom, forming a cloud of recirculating gas around the bubble; in three dimensions R_c³/R_b³ = (u_br + 2u_f)/(u_br − u_f). If u_br < u_f (slow bubble, Geldart D), gas uses the bubble as a short cut and there is no cloud. Behind each bubble is a wake of solids (wake fraction α ≈ 0.2–0.4 of bubble volume) that is carried up and gives the strong solids circulation.
In a freely bubbling bed, bubbles form at the distributor, grow by coalescence as they rise, and are limited by the column diameter (slugging) or by splitting (group A). Correlations such as Mori and Wen give the bubble diameter at height h: d_b = d_bm − (d_bm − d_b0)e^(−0.3h/D_t), with d_bm = 0.652[A_t(u₀ − u_mf)]^0.4 and d_b0 depending on the distributor (cgs units). Bubble frequency and size are measured with capacitance or optical probes, pressure fluctuations and X-ray photography. The rise velocity in a bed is u_b = u₀ − u_mf + u_br (Davidson–Harrison).
The emulsion phase stays close to minimum fluidization conditions (ε_e ≈ ε_mf, gas velocity ≈ u_mf/ε_mf relative to solids). Experiments show solids moving up in bubble wakes and down in the emulsion, so the emulsion gas may even flow downward at high u₀/u_mf. The two-phase theory (Toomey–Johnstone) says all gas in excess of minimum fluidization passes through the bed as bubbles, giving the bubble fraction δ ≈ (u₀ − u_mf)/u_b and the expanded height L_f = L_mf/(1 − δ).
- u_br = 0.711√(gd_b).
- Fast bubble (u_br > u_mf/ε_mf): cloud forms.
- 3D cloud: R_c³/R_b³ = (u_br + 2u_f)/(u_br − u_f).
- Two-phase theory: excess gas goes as bubbles.
- Mori–Wen bubble growth; u_b = u₀ − u_mf + u_br.
Example: A 10 cm bubble rises in a bed with u_mf = 0.0755 m/s and ε_mf = 0.45. Find u_br, decide if it is fast, and find the cloud size.
- u_br = 0.711√(9.81 × 0.1) = 0.704 m/s; u_f = 0.0755/0.45 = 0.168 m/s.
- u_br > u_f: fast bubble with a cloud.
- R_c³/R_b³ = (0.704 + 0.336)/(0.704 − 0.168) = 1.940 ⇒ R_c/R_b = 1.25.
Answer: A fast bubble; its cloud extends to 1.25 bubble radii (≈ 12.5 cm diameter).
Example: In a 1 m bed with u₀ = 0.3 m/s, u_mf = 0.0755 m/s and 5000 holes/m², estimate the bubble size at 0.5 m and the bed expansion (L_mf = 0.5 m).
- A_t(u₀ − u_mf) = 7854 cm² × 22.45 cm/s = 1.763 × 10⁵ cm³/s.
- d_bm = 0.652(1.763 × 10⁵)^0.4 = 81.8 cm; d_b0 = 0.347(1.763 × 10⁵/3927)^0.4 = 1.59 cm.
- d_b(50 cm) = 81.8 − 80.2e^(−0.15) = 12.8 cm; u_b = 22.45 + 0.711√(981 × 12.8) = 102 cm/s.
- δ = 22.45/102 = 0.22 ⇒ L_f = 0.5/0.78 = 0.64 m.
Answer: Bubbles of about 13 cm; the bed expands to about 0.64 m.
Davidson’s model: bubble rise velocity u_br = 0.711√(gd_b); fast bubbles (u_br > u_mf/ε_mf) carry a cloud of recirculating gas (R_c³/R_b³ = (u_br + 2u_f)/(u_br − u_f)); slow bubbles short-circuit gas. Wake solids drive circulation. Bubble growth by coalescence (Mori–Wen), measurement by probes and X-rays, u_b = u₀ − u_mf + u_br. Emulsion stays near minimum fluidization; two-phase theory gives δ and bed expansion.
- Larger bubbles rise faster and carry more by-passing gas.
- Clouds thin out as bubbles get faster.
- Group D beds have slow bubbles and no clouds.
- Bubble wakes lift solids; emulsion carries them down.
Key formulas
- u_br = 0.711√(gd_b)
- R_c³/R_b³ = (u_br + 2u_mf/ε_mf)/(u_br − u_mf/ε_mf)
- d_b = d_bm − (d_bm − d_b0)e^(−0.3h/D_t)
- δ = (u₀ − u_mf)/u_b
- L_f = L_mf/(1 − δ)
Practice questions
The rise velocity of a single bubble is
- 0.711√(gd_b)
- u_mf
- g d_b
- u₀
Show the answer
Answer: A. 0.711√(gd_b) Davidson.
A cloud forms around a bubble when
- u_br < u_mf/ε_mf
- u_br > u_mf/ε_mf
- u₀ < u_mf
- the bed is fixed
Show the answer
Answer: B. u_br > u_mf/ε_mf Fast bubble.
The two-phase theory says gas in excess of u_mf passes as
- emulsion gas
- bubbles
- solids
- nothing
Show the answer
Answer: B. bubbles Toomey–Johnstone.
Bubbles grow mainly by
- evaporation
- coalescence
- compression
- reaction
Show the answer
Answer: B. coalescence Merging.
The wake behind a bubble carries
- gas only
- solids upward
- liquid
- nothing
Show the answer
Answer: B. solids upward Solids circulation.
Theory questions
Explain Davidson’s model of a rising bubble and the formation of a cloud. (6 marks)
Show the model answer
Assumptions: bubble is solids-free and spherical (circular in 2D); particles move as an inviscid incompressible fluid around the bubble with bulk density ρ_s(1 − ε_mf); gas flows through the emulsion by Darcy’s law (velocity ∝ pressure gradient); far from the bubble the pressure gradient is that of minimum fluidization. Solving gives the bubble rise velocity u_br = 0.711√(gd_b) and the gas streamlines. If u_br > u_f = u_mf/ε_mf, gas leaving the top of the bubble cannot escape upward fast enough and is carried round by the moving particles back into the bottom of the bubble, forming a cloud of thickness given (3D) by R_c³/R_b³ = (u_br + 2u_f)/(u_br − u_f). As u_br/u_f → ∞ the cloud becomes very thin; if u_br < u_f, emulsion gas uses the bubble as a short cut and no cloud forms.
What is the two-phase theory of fluidization and how is it used to estimate bed expansion? (4 marks)
Show the model answer
Toomey and Johnstone proposed that a bubbling bed consists of an emulsion phase at minimum fluidization conditions (voidage ε_mf, gas flow u_mf) and a bubble phase that carries all the gas in excess of u_mf. Then the bubble volume fraction δ satisfies δu_b = u₀ − u_mf, so δ = (u₀ − u_mf)/u_b with u_b = u₀ − u_mf + 0.711√(gd_b). The bed height then follows from conservation of solids: L_f(1 − δ) = L_mf, i.e. L_f = L_mf/(1 − δ). In practice somewhat less gas flows as bubbles (some extra throughflow), especially for group A.
Flow pattern of gas in fluidized beds
Topics: Flow pattern of gas through fluidized beds; experimental findings; the bubbling-bed model for gas interchange.
Gas interchange and the Kunii–Levenspiel bubbling-bed model
Tracer experiments show that gas in a bubbling bed does not move in plug flow: most of it rises rapidly in bubbles, while gas in the emulsion moves slowly or even downwards with the circulating solids. Gas is exchanged between bubbles and emulsion by throughflow and diffusion. This explains why fluidized-bed reactors often give much lower conversion than packed beds with the same amount of catalyst, especially with large bubbles and fast reactions.
The Kunii–Levenspiel bubbling-bed model describes the bed with three regions — bubble, cloud–wake and emulsion — all quantities expressed per unit volume of bubbles. Gas transfers from bubble to cloud with coefficient K_bc = 4.5u_mf/d_b + 5.85D^½g^¼/d_b^(5/4) (throughflow plus diffusion) and from cloud to emulsion with K_ce = 6.77(Dε_mfu_br/d_b³)^½. The solids in each region per bubble volume are γ_b (≈ 0.001–0.01), γ_c = (1 − ε_mf)[3(u_mf/ε_mf)/(u_br − u_mf/ε_mf) + α] and γ_e = (1 − ε_mf)(1 − δ)/δ − γ_c − γ_b. For a first-order reaction (rate constant k per volume of solids), reaction in each region and transfer between them combine like resistances: K_f = γ_bk + 1/[1/K_bc + 1/(γ_ck + 1/(1/K_ce + 1/(γ_ek)))], and the outlet concentration is C_out/C_in = exp(−K_fL_f/u_b).
The model shows how design choices matter: smaller bubbles (fine particles, good distributors, internals, Geldart A powders) improve interchange and conversion; for fast reactions the conversion is limited by K_bc and K_ce, not by the catalyst; for slow reactions the bed behaves almost like a well-contacted reactor. It is the basis for comparing fluidized and fixed beds and for scale-up.
- Gas by-passes in bubbles; emulsion gas moves slowly.
- K_bc = 4.5u_mf/d_b + 5.85D^½g^¼/d_b^(5/4).
- K_ce = 6.77(Dε_mfu_br/d_b³)^½.
- K_f combines reaction and interchange; C_out/C_in = exp(−K_fL_f/u_b).
- Small bubbles → better conversion.
Example: With d_b = 0.1 m, u₀ = 0.3 m/s, u_mf = 0.0755 m/s, ε_mf = 0.45, D = 2 × 10⁻⁵ m²/s, α = 0.33, γ_b = 0.005, k = 1 s⁻¹, L_m = 1 m (ε_m = 0.4), find the conversion.
- u_br = 0.704, u_b = 0.929 m/s; δ = 0.2245/(0.929 − 0.0755) = 0.263.
- γ_c = 0.55(3 × 0.168/0.536 + 0.33) = 0.698; γ_e = 0.55 × 0.737/0.263 − 0.698 − 0.005 = 0.838.
- K_bc = 3.40 + 0.82 = 4.22 s⁻¹; K_ce = 0.539 s⁻¹.
- K_f = 0.005 + 1/[1/4.22 + 1/(0.698 + 1/(1/0.539 + 1/0.838))] ≈ 0.83 s⁻¹.
- L_f = 1 × 0.6/(0.737 × 0.55) = 1.48 m; X = 1 − exp(−K_f × 1.48/0.929) ≈ 0.73.
Answer: About 73 %, compared with 86.5 % for plug flow through the same catalyst.
Gas flow in bubbling beds: fast by-passing in bubbles, slow (even downward) emulsion gas, interchange by throughflow and diffusion. Kunii–Levenspiel bubbling-bed model: bubble, cloud–wake and emulsion regions per bubble volume; interchange coefficients K_bc and K_ce; solids fractions γ_b, γ_c, γ_e; overall first-order rate constant K_f in series–parallel form; C_out/C_in = exp(−K_fL_f/u_b). Smaller bubbles and slower reactions bring conversion closer to plug flow.
- Conversion in bubbling beds is usually below plug flow.
- K_bc dominated by throughflow for large particles.
- Internals (baffles, tubes) limit bubble size.
- Fast reactions are interchange-limited.
Key formulas
- K_bc = 4.5u_mf/d_b + 5.85D^½g^¼/d_b^(5/4)
- K_ce = 6.77(Dε_mfu_br/d_b³)^½
- γ_c = (1 − ε_mf)[3(u_mf/ε_mf)/(u_br − u_mf/ε_mf) + α]
- K_f = γ_bk + 1/[1/K_bc + 1/(γ_ck + 1/(1/K_ce + 1/(γ_ek)))]
- C_out/C_in = exp(−K_fL_f/u_b)
Practice questions
In the K–L model, quantities are expressed per unit volume of
- bed
- bubbles
- solids
- gas
Show the answer
Answer: B. bubbles Bubble basis.
K_bc represents interchange between
- emulsion and wall
- bubble and cloud
- cloud and emulsion
- gas and distributor
Show the answer
Answer: B. bubble and cloud First step.
Smaller bubbles generally
- lower conversion
- raise conversion
- have no effect
- stop fluidization
Show the answer
Answer: B. raise conversion Better interchange.
For a very fast reaction in a bubbling bed, conversion is limited by
- catalyst amount
- gas interchange
- temperature
- pressure drop
Show the answer
Answer: B. gas interchange Mass transfer.
Tracer studies show emulsion gas may move
- only upward fast
- slowly or even downward
- radially only
- not at all
Show the answer
Answer: B. slowly or even downward Solids circulation.
Theory questions
Describe the Kunii–Levenspiel bubbling-bed model and derive the expression for conversion of a first-order reaction. (6 marks)
Show the model answer
The bed consists of bubbles of uniform size d_b rising at u_b, each surrounded by a cloud–wake region, in an emulsion at minimum fluidization. Assumptions: gas in the emulsion is stagnant relative to the bubble flow; all gas flows as bubbles; solids are distributed as γ_b, γ_c, γ_e per bubble volume. Balances on a bubble as it rises: −dC_b/dt = γ_bkC_b + K_bc(C_b − C_c); K_bc(C_b − C_c) = γ_ckC_c + K_ce(C_c − C_e); K_ce(C_c − C_e) = γ_ekC_e. Eliminating C_c and C_e: −dC_b/dt = K_fC_b with K_f = γ_bk + 1/[1/K_bc + 1/(γ_ck + 1/(1/K_ce + 1/(γ_ek)))]. Integrating over the bubble residence time t = L_f/u_b: C_out/C_in = exp(−K_fL_f/u_b), X = 1 − exp(−K_fL_f/u_b).
Why do fluidized-bed reactors often give lower conversion than fixed beds, and how can this be improved? (4 marks)
Show the model answer
In a bubbling bed most gas passes through as bubbles that contain few solids; it reacts only after transfer to the cloud and emulsion, which is slow for large bubbles. Emulsion gas is back-mixed. Both lower conversion compared with plug flow through a fixed bed. Improvements: use finer (Geldart A) particles giving small bubbles; good distributors with many small holes; internals (baffles, tube banks) to break bubbles; operate in turbulent or fast-fluidization regimes with better contact; staged beds; use deeper beds or more catalyst for slow reactions.
Entrainment and elutriation
Topics: Entrainment at or above the TDH; model for entrainment; application of the entrainment model to elutriation.
Freeboard behaviour, TDH and elutriation
When bubbles burst at the bed surface they eject particles into the freeboard, mostly from the bubble wake and nose. Coarse particles fall back; finer ones, whose terminal velocity is less than the gas velocity, can be carried out. Close to the surface the solids flux and concentration are high and fall roughly exponentially with height, F(h) = F_∞ + (F₀ − F_∞)e^(−ah). Above the transport disengaging height (TDH) only particles that the gas can carry remain, and the flux is constant, F_∞. Gas outlets and cyclone inlets are therefore placed above the TDH; correlations relate TDH to the bubble size at the surface (e.g. Horio: TDH ≈ 4.47√d_b,surface) or to the gas velocity (Zenz and Weil charts).
Entrainment is the carry-over of solids from the bed; elutriation is the selective removal of fines from a mixture of sizes. Above the TDH, the rate at which size fraction i leaves is proportional to its mass fraction x_i in the bed: R_i = −d(Wx_i)/dt = AK*_ix_i, with an elutriation constant K*_i (kg/m²·s) that depends on u₀, u_t,i and the system. Correlations such as Colakyan and Levenspiel’s K*_i = 0.011ρ_s(1 − u_t,i/u₀)² or Zenz–Weil give K*. For a batch bed with no feed, each fine fraction decays exponentially: W_i(t) = W_i0 exp(−AK*_it/W). For continuous operation with feed, balances on each size give the steady size distribution of bed, overflow and carry-over, and the load on the cyclones.
- TDH: above it entrained flux is constant.
- Horio: TDH ≈ 4.47√d_b,surface (m).
- R_i = AK*_ix_i; K*_i = 0.011ρ_s(1 − u_t,i/u₀)².
- Batch: W_i = W_i0 e^(−AK*_it/W).
- Cyclones return entrained fines to the bed.
Example: Bubbles at the bed surface are 0.15 m. Estimate the TDH.
- TDH ≈ 4.47√0.15 = 4.47 × 0.387.
Answer: About 1.7 m: the freeboard outlet should be at least this far above the bed.
Example: A 60 µm fraction (u_t ≈ 0.16 m/s, ρ_s = 1500 kg/m³) makes up 15 % of a 1000 kg bed in a 1 m column at u₀ = 0.5 m/s. Find its elutriation constant, initial loss rate and the fraction left after 30 min with no make-up.
- K* = 0.011 × 1500 × (1 − 0.16/0.5)² = 16.5 × 0.462 = 7.6 kg/m²·s.
- R = 0.785 × 7.6 × 0.15 = 0.90 kg/s initially.
- W_i/W_i0 = exp(−0.785 × 7.6 × 1800/1000) = exp(−10.7) ≈ 2 × 10⁻⁵.
Answer: This fraction is essentially lost in half an hour without cyclone recycle.
Bursting bubbles eject solids; solids flux decays with height to a constant value above the TDH; outlets above TDH (Horio, Zenz–Weil correlations). Entrainment vs elutriation; rate R_i = AK*_ix_i with elutriation constants (Colakyan–Levenspiel, Zenz–Weil); batch decay W_i = W_i0e^(−AK*_it/W); continuous size balances; cyclone recovery.
- Particles with u_t > u₀ are not elutriated above TDH.
- Fines are lost first, coarsening the bed.
- Cyclone diplegs return solids to the bed.
- Attrition continually produces new fines.
Key formulas
- TDH ≈ 4.47√d_b,surface
- F(h) = F_∞ + (F₀ − F_∞)e^(−ah)
- R_i = AK*_ix_i
- K*_i = 0.011ρ_s(1 − u_t,i/u₀)²
- W_i(t) = W_i0 exp(−AK*_it/W)
Practice questions
Above the TDH the entrained flux
- keeps falling
- is constant
- is zero
- rises
Show the answer
Answer: B. is constant Definition.
Elutriation is
- bubbling
- selective removal of fines
- attrition
- channelling
Show the answer
Answer: B. selective removal of fines Size-selective.
Particles are elutriated when their terminal velocity is
- greater than u₀
- less than u₀
- equal to u_mf
- zero
Show the answer
Answer: B. less than u₀ Gas can carry them.
The elutriation rate of a size cut is proportional to
- its fraction in the bed
- bed height
- distributor holes
- temperature only
Show the answer
Answer: A. its fraction in the bed R_i = AK*_ix_i.
The gas outlet of a fluidized bed should be placed
- below the TDH
- above the TDH
- at the distributor
- inside the bed
Show the answer
Answer: B. above the TDH Minimise carry-over.
Theory questions
Explain entrainment, TDH and their importance in design. (5 marks)
Show the model answer
Bubbles bursting at the surface throw particles (from their wakes and roofs) into the freeboard; large ones fall back, fine ones may be carried away. The entrained solids flux decreases with height above the bed until, at the transport disengaging height, only particles that can be transported by the gas remain and the flux becomes constant. Design: place the gas exit or primary cyclone inlet above the TDH to minimise the load on cyclones; size cyclones for F_∞; freeboard height sets vessel cost; TDH increases with gas velocity and bubble size (correlations by Zenz–Weil, Horio, Fournol). Entrainment also affects bed inventory, catalyst loss and emissions.
Derive the expression for batch elutriation of a size fraction and explain how elutriation constants are found. (5 marks)
Show the model answer
Above TDH, the rate of loss of fraction i is proportional to its concentration in the bed: −d(Wx_i)/dt = AK*_ix_i. If the total bed mass W changes little (fines are a small part), W dx_i/dt = −AK*_ix_i ⇒ x_i = x_i0 exp(−AK*_it/W). K*_i is obtained experimentally from the slope of ln x_i vs t in batch runs, or from correlations: Zenz–Weil (K*/(ρ_gu₀) vs u₀²/(gd_pρ_s²)), Colakyan–Levenspiel K*_i = 0.011ρ_s(1 − u_t,i/u₀)², Wen–Chen. For continuous systems, steady-state balances on each size give bed and entrainment size distributions.
Notes, examples and questions were written for this studio; the prescribed books are listed in the studio for reference.
Frequently asked questions
What is the Fluidization Engineering Studio?
Beds that flow like liquids. Minimum fluidization and terminal velocities, Geldart groups and regimes, distributor design, Davidson bubbles, Kunii–Levenspiel reactor model, TDH and elutriation from the ASTU CHE1816OE22 syllabus (2018-19 scheme).
What can I do in the Fluidization Engineering Studio?
Minimum fluidization and terminal velocity; Geldart groups, regimes and distributor design; Bubbles, clouds and the Kunii–Levenspiel model; TDH, entrainment and elutriation.
Is the Fluidization Engineering Studio free?
Yes. It is free and runs in your browser with no sign-up.