# Centrifuge Separation Engineering: Peeler vs. Pusher vs. Decanter vs. Disc Stack in API Manufacturing

# Scaling Centrifugal Acceleration, Sigma Theory ( $\Sigma$ ), Dewatering Kinetics, and Cake Compressibility

Solid-liquid separation is often the critical rate-limiting step in active pharmaceutical ingredient (API) manufacturing. While Agitated Nutsche Filter Dryers (ANFD) excel at contained solvent washing, centrifugal separators provide orders of magnitude higher dewatering driving forces, slashing batch filtration and deliquoring times from hours down to minutes.

Selecting the right centrifuge geometry—Horizontal Basket Peeler, Continuous Pusher, Scroll Decanter, or Disc Stack Clarifier—requires balancing particle size distribution (PSD), cake compressibility, washing efficiency, and containment requirements.

Pharma Centrifuge Taxonomy & Equipment Comparison

# 1. Governing Centrifugal Mechanics & G-Force

Centrifuges replace gravitational acceleration ( $g = 9.81\text{ m/s}^2$ ) with centrifugal field acceleration ( $a_c = \omega^2 \cdot r$ ). The dimensionless Centrifugal Separation Factor ( $G$ ) is:

G = \frac{\omega^2 \cdot r}{g} = \frac{\left(\frac{2\pi N}{60}\right)^2 \cdot r}{9.81} = 1.118 \cdot 10^{-3} \cdot r_{mm} \cdot N_{rpm}^2

Where:

$r_{mm}$ : Basket or bowl radius in millimeters.
$N_{rpm}$ : Rotational speed in revolutions per minute.

                      CENTRIFUGAL FIELD G-FORCE SPECTRUM
   Gravity (1G)      Pusher (400-900G)     Peeler (800-1500G)     Decanter (2000-4000G)    Disc Stack (5000-15000G)
  ◄──────┼───────────────────┼─────────────────────┼───────────────────────┼────────────────────────►
   Nutsche Filters     Large Granular       Pure Finished APIs      Slurries & Bio-Mass      Broth & Bio-Polishing
   & Decantation       Inorganics           (5 - 200 µm)            (High Solids 30%)        Sub-micron Clarification

# 2. Ambler's Sigma Theory ( $\Sigma$ ) for Scale-Up

Ambler's Sigma Concept ( $\Sigma$ ) represents the equivalent gravitational settling area of a centrifuge:

Q = 2 \cdot v_g \cdot \Sigma

Where:

$Q$ : Volumetric throughput ( $\text{m}^3/\text{s}$ ).
$v_g$ : Stokes' settling velocity under gravity ( $1\text{ g}$ ).
$\Sigma$ : Physical index of centrifuge capacity ( $\text{m}^2$ ).

# 2.1. Scale-Up Formulation

When transferring a centrifugation process from pilot laboratory to production plant:

\left( \frac{Q}{\Sigma} \right)_{pilot} = \left( \frac{Q}{\Sigma} \right)_{production}

For a tubular or solid-bowl decanter centrifuge:

\Sigma = \frac{2\pi \cdot \omega^2 \cdot L}{g} \cdot \left[ \frac{3}{4} r_2^2 + \frac{1}{4} r_1^2 \right]

# 3. Technology Comparison Matrix

Centrifuge Type	G-Force Range	Mean Particle Size	Cake Compressibility	Operational Mode	Primary Pharma Application
Horizontal Basket Peeler	$800 - 1,500\text{ G}$	$5 - 200\,\mu\text{m}$	Low to Moderate	Discontinuous Batch	Premium pure APIs, multi-solvent displacement washing, cGMP clean-in-place.
Pusher Centrifuge	$400 - 900\text{ G}$	$> 100\,\mu\text{m}$	Incompressible / Granular	Continuous Axial Stroke	Bulk actives, intermediate salts, high-volume production ( $1 - 15\text{ tons/h}$ ).
Scroll Decanter	$2,000 - 4,000\text{ G}$	$2 - 500\,\mu\text{m}$	Moderate to High	Continuous Sludge Discharge	Heavy slurries, fermentation cell mass separation, recovery from mother liquors.
Disc Stack Clarifier	$5,000 - 15,000\text{ G}$	$0.1 - 20\,\mu\text{m}$	Non-applicable (Clarification)	Continuous Liquid / Intermittent Shot	Polishing mother liquors, bio-pharmaceutical broths, vaccine cell debris removal.

# 4. Cake Filtration & Centrifugal Deliquoring Kinetics

Filtration flow rate through a porous cake on a rotating perforated basket is governed by the centrifugal modification of Darcy's Law:

\frac{dV}{dt} = \frac{\pi \cdot L_{basket} \cdot \rho_L \cdot \omega^2 \cdot (r_b^2 - r_0^2)}{\mu_L \cdot \left[ \frac{\alpha \cdot M_{cake}}{2\pi L_{basket} \ln(r_b/r_c)} + R_m \right]}

Where:

$\alpha$ : Specific cake filtration resistance ( $\text{m/kg}$ ).
$r_b, r_c, r_0$ : Radius of the basket, inner cake surface, and liquid pool surface.
$R_m$ : Filter cloth medium resistance ( $\text{m}^{-1}$ ).

# 4.1. Centrifugal Dewatering & Capillary Desaturation

Residual cake moisture after centrifugal spinning depends on the Capillary Number ( $N_{cap}$ ):

N_{cap} = \frac{\rho_L \cdot a_c \cdot k_{perm}}{\sigma \cdot \cos\theta}

Where:

$k_{perm}$ : Cake permeability ( $\text{m}^2$ ) via the Kozeny-Carman relationship:

k_{perm} = \frac{\epsilon^3}{5 \cdot S_v^2 \cdot (1 - \epsilon)^2}

$\sigma$ : Liquid surface tension (N/m).
$\theta$ : Contact angle.

As $N_{cap} > 10^{-2}$ , capillary trapped moisture is forced through pore throats, achieving residual moisture contents under $5 - 12\text{ wt}\%$ (compared to $25 - 40\%$ on Nutsche filters).

# 5. Comprehensive Worked Industrial Case Study: 400 kg Batch API Isolation

# Problem Statement:

A pharmaceutical plant isolates an intermediate crystalline API from a mother liquor slurry:

Slurry batch volume: $V = 3,000\text{ L}$ ( $15\text{ wt}\%$ solids, $\rho_{slurry} = 1,080\text{ kg/m}^3$ ).
Dry cake mass: $M_{dry} = 450\text{ kg}$ .
Dry cake bulk density: $\rho_{cake} = 650\text{ kg/m}^3$ .
Cake volume: $V_{cake} = 450 / 650 = 0.692\text{ m}^3$ ( $692\text{ L}$ ).
Select and size a Horizontal cGMP Peeler Centrifuge.

# Step 1: Centrifuge Basket Selection

Standard basket sizes: $\Phi 1,000\text{ mm}$ , $\Phi 1,250\text{ mm}$ , $\Phi 1,600\text{ mm}$ .
Evaluate a $\Phi 1,250\text{ mm} \times 630\text{ mm}$ basket:
- Basket Radius $R_b = 0.625\text{ m}$ .
- Basket Length $L_b = 0.630\text{ m}$ .
- Total Basket Internal Volume:

V_{basket} = \pi \cdot R_b^2 \cdot L_b = \pi \cdot (0.625)^2 \cdot 0.630 = 0.773\text{ m}^3 = 773\text{ L}

Working cake thickness: $t_{cake} = 150\text{ mm}$ ( $r_c = 0.625 - 0.150 = 0.475\text{ m}$ ).
Active Cake Volume:

V_{active} = \pi \cdot (R_b^2 - r_c^2) \cdot L_b = \pi \cdot (0.625^2 - 0.475^2) \cdot 0.630 = \pi \cdot (0.3906 - 0.2256) \cdot 0.630 = 0.326\text{ m}^3 = 326\text{ L}

Since $V_{cake} = 692\text{ L}$ , the batch will be isolated in 2 sequential feeding cycles ( $346\text{ L}$ cake per spin cycle).

# Step 2: Rotational Speed & G-Force

Operating speed: $N = 1,200\text{ rpm}$ .
Separation Factor ( $G$ ):

G = 1.118 \cdot 10^{-3} \cdot (625\text{ mm}) \cdot (1,200)^2 = 1.118 \cdot 10^{-3} \cdot 625 \cdot 1,440,000 = \mathbf{1,006\text{ G}}

# Step 3: Cycle Time Breakdown per Charge (Cycle A & B)

Slurry Feeding (Charging at 600 rpm): $1,500\text{ L}$ slurry charged in $4.0\text{ minutes}$ .
Primary Spin (Acceleration to 1,200 rpm & initial dewatering): $3.0\text{ minutes}$ .
Displacement Cake Washing: 2 cake volumes of chilled Ethanol ( $650\text{ L}$ wash liquid) at 800 rpm: $5.0\text{ minutes}$ .
Final High-Speed Deliquoring (1,200 rpm, 1,006 G): $8.0\text{ minutes}$ (achieves residual moisture $< 7.5\%$ ).
Deceleration & Peeler Knife Discharge (at 120 rpm): $2.5\text{ minutes}$ .
Residual Heel Nitrogen Blowback: $1.0\text{ minute}$ .

Total Cycle Time per Charge: $23.5\text{ minutes}$ .
Total Batch Processing Time ( $2 \times 23.5\text{ min} + 10\text{ min}$ pre-inerting): $\approx 57\text{ minutes}$ (vs. 8 to 12 hours on an equivalent ANFD).

# 6. Process Safety & Inertization Control (NFPA 69)

Because centrifuges run at high rotational speeds in the presence of flammable organic solvents (Acetone, Toluene, IPA), electrostatic spark ignition is an ever-present catastrophe risk.

flowchart TD
    A["Centrifuge Start Command"] --> B["N2 Pre-Purge (3 Chamber Volumes)"]
    B --> C["Continuous O2 Sensor Check"]
    C --> D{"O2 < 4.0% v/v?"}
    D -- No --> B
    D -- Yes --> E["Release Safety Interlock & Accelerate"]
    E --> F["Continuous Dynamic N2 Blanket (+25 mbar)"]
    F --> G{"O2 > 5.0% or Vibration > 4 mm/s?"}
    G -- Yes --> H["EMERGENCY TRIP: Nitrogen Flood + Dynamic Braking"]
    G -- No --> I["Normal Cycle Progression"]

    style A fill:#e0f2fe,stroke:#0284c7
    style D fill:#fef3c7,stroke:#d97706
    style E fill:#dcfce7,stroke:#16a34a
    style H fill:#fee2e2,stroke:#dc2626

# 7. Operational Troubleshooting & Plant Failure Modes

Problem	Root Cause	Underlying Mechanism	Corrective Engineering Action
Severe Vibration Trip ( $> 4.0\text{ mm/s}$ )	Non-uniform cake thickness along basket length	Feed slurry pipe distribution nozzle is misaligned or choked	1. Replace single-point feed pipe with multi-nozzle spray header. 2. Lower feeding speed from 800 rpm to 450 rpm to allow slurry to level hydraulically before dewatering.
Cake Glazing / Blinding	Fine particle compaction into filter cloth	High initial G-force forces sub- $5\,\mu\text{m}$ fines into cloth pores	1. Implement two-stage ramped acceleration: feed at low G ( $150\text{ G}$ ), then ramp to full G ( $1,000\text{ G}$ ) only after cake bed forms. 2. Switch from polypropylene needle felt to monofilament PTFE satin weave.
Peeler Knife Heel Contamination	Incomplete cake discharge leaving $5 - 10\text{ mm}$ residual cake heel	Scraper blade kept $6\text{ mm}$ away from cloth to prevent ripping cloth	Install high-pressure Nitrogen Blowback manifold behind the cloth: blow back $6\text{ bar } N_2$ pulses during peeling to dislodge the remaining heel completely.
Mother Liquor Cloudiness	Fines breakthrough through cloth seams	Mechanical cloth fastening ring leakage	Inspect and replace O-ring clamping cords; verify filter cloth micron rating against laser diffraction PSD $D_{10}$ .

# Applicable Engineering Standards & Codes Used

DIN 24400: Centrifuges: Safety requirements for construction and installation.
NFPA 69: Standard on Explosion Prevention Systems (Limiting Oxygen Concentration control).
ASME BPE: Bioprocessing Equipment cGMP design for pharmaceutical centrifuges.
ISO 20381: Industrial centrifuges - Technical specifications.
API 670: Machinery Protection Systems (Vibration and Bearing Temperature Monitoring).

Centrifuge Separation Engineering: Peeler vs. Pusher vs. Decanter vs. Disc Stack in API Manufacturing

Peer-Reviewed & PE Verified

# Centrifuge Separation Engineering: Peeler vs. Pusher vs. Decanter vs. Disc Stack in API Manufacturing

# Scaling Centrifugal Acceleration, Sigma Theory ( $\Sigma$ ), Dewatering Kinetics, and Cake Compressibility

# 1. Governing Centrifugal Mechanics & G-Force

# 2. Ambler's Sigma Theory ( $\Sigma$ ) for Scale-Up

# 2.1. Scale-Up Formulation

# 3. Technology Comparison Matrix

# 4. Cake Filtration & Centrifugal Deliquoring Kinetics

# 4.1. Centrifugal Dewatering & Capillary Desaturation

# 5. Comprehensive Worked Industrial Case Study: 400 kg Batch API Isolation

# Problem Statement:

# Step 1: Centrifuge Basket Selection

# Step 2: Rotational Speed & G-Force

# Step 3: Cycle Time Breakdown per Charge (Cycle A & B)

# 6. Process Safety & Inertization Control (NFPA 69)

# 7. Operational Troubleshooting & Plant Failure Modes

# Applicable Engineering Standards & Codes Used

Discussion

Peer-Reviewed & PE Verified

# Centrifuge Separation Engineering: Peeler vs. Pusher vs. Decanter vs. Disc Stack in API Manufacturing

# Scaling Centrifugal Acceleration, Sigma Theory (Σ\SigmaΣ), Dewatering Kinetics, and Cake Compressibility

# 1. Governing Centrifugal Mechanics & G-Force

# 2. Ambler's Sigma Theory (Σ\SigmaΣ) for Scale-Up

# 2.1. Scale-Up Formulation

# 3. Technology Comparison Matrix

# 4. Cake Filtration & Centrifugal Deliquoring Kinetics

# 4.1. Centrifugal Dewatering & Capillary Desaturation

# 5. Comprehensive Worked Industrial Case Study: 400 kg Batch API Isolation

# Problem Statement:

# Step 1: Centrifuge Basket Selection

# Step 2: Rotational Speed & G-Force

# Step 3: Cycle Time Breakdown per Charge (Cycle A & B)

# 6. Process Safety & Inertization Control (NFPA 69)

# 7. Operational Troubleshooting & Plant Failure Modes

# Applicable Engineering Standards & Codes Used

Discussion

# Scaling Centrifugal Acceleration, Sigma Theory ( $\Sigma$ ), Dewatering Kinetics, and Cake Compressibility

# 2. Ambler's Sigma Theory ( $\Sigma$ ) for Scale-Up