How to Choose the Right Capacity for a Protein Mill (Split-Axis Air Classifier Mill for Dry Fractionation)

Capacity selection is not simply matching target finished protein output. For dry pulse protein fractionation, mill rated capacity refers to raw dehulled flour feed rate into the mill, not final protein concentrate yield. Many projects run into bottlenecks: over-sizing raises CAPEX and energy waste; under-sizing forces overfeeding, poor liberation, higher temperature, wider PSD and damaged protein functionality. This guide focuses on dry fractionation lines producing beverage-grade pea/fava bean protein.

1. Start from your target finished protein output (mass balance first)

Step 1: Define key mass balance parameters (yellow pea reference):

• Dehulled pea raw protein: ~22–25%
• Dry fraction fine fraction (protein concentrate) mass yield: 38–45% of inlet feed
• Protein concentrate purity: 58–66%
• Operating hours: typical plant = 20 h/day, 300 days/year

Formula to calculate required mill inlet feed rate:

Required feed rate (kg/h) = Target protein concentrate output (kg/h) ÷ Fine fraction mass yield

Example:
Target finished protein powder = 800 kg/h
Average fine yield = 40%
Minimum required mill feed = 800 ÷ 0.4 = 2,000 kg/h inlet flour

Important note: Middling recirculation adds internal circulation load. The mill must handle fresh feed + recirculated middlings. Internal circulation can increase total mill throughput by 30–70%. Always inform equipment supplier of expected recirculation ratio.

2. Critical rule: Capacity depends strongly on target fineness & product grade

The same mill model delivers different throughput based on required PSD:

• Standard grade protein (D90 20–30 μm): higher achievable feed rate
• Beverage-grade protein (D90 10–25 μm, narrow PSD, minimal oversized tail) → feed rate must be reduced by 20–40%
• If you push maximum nominal capacity while targeting fine narrow PSD:
  1. Particle residence time drops → incomplete liberation
  2. More undissociated composites → higher oversized fraction
  3. Classification cut-point blurs → lower protein purity
  4. Particle crowding increases collision heat → protein denaturation risk

Vendor nominal capacity is usually quoted for medium-fine powder. Always ask: What is the guaranteed capacity at my target D90 specification?

3. Material properties affect real achievable throughput

• Raw material: Yellow pea vs fava bean vs lentil differ in cotyledon hardness, starch granule size, oil content
• Moisture window: 8.0–10.5% optimal. Moisture >10.5% reduces mill capacity by 15–30% due to agglomeration
• Dehulling efficiency: residual hull fibre makes material tougher to grind, lowers effective throughput

4. Reserve factor (safety margin)

Never select a mill exactly matching calculated minimum feed rate.
Recommended reserve coefficient:

• Steady single-product production: 1.15–1.25
• Multi-product switching (pea / fava bean) or future capacity expansion: 1.30–1.40

Why?

• Seasonal variation in raw bean hardness and moisture
• Periodic wear of hammers/stators gradually reduces milling efficiency
• Scheduled downtime, cleaning cycles, temporary process upsets
• Potential future upgrade to stricter beverage-grade PSD requirements

5. Match mill capacity to upstream & downstream equipment (line balance)

The mill cannot operate in isolation. Identify your line bottleneck:

1. Upstream: cleaning, dehuller, pre-grinder, silo discharge feeding
2. Downstream: cyclones, baghouse, powder conveying, cooling, packaging
   If you select a large mill but downstream air handling and dust collection are undersized, airflow becomes restricted, separation efficiency collapses.

Air volume is equally critical as tonnage: larger mills require matched fan, ducting and filter capacity.

6. Equipment design impact: Split-axis vs single-axis when sizing

• Single-axis ACM: Grinding and classification speed coupled. At tight cut-point operation, you often sacrifice throughput to maintain PSD.
• Split-axis classifier mill: independent grinding rotor and classifier wheel tuning. More flexible to sustain stable capacity while maintaining beverage-grade fineness.
  When comparing capacity quotations, confirm whether test data comes from split-axis configuration.

7. Avoid these common capacity selection mistakes

Mistake 1: Confusing finished protein output with mill feed capacity

Buyers often say “I need 1 ton/h protein powder” and select a 1 t/h mill. This severely undersizes the system, because the mill inlet feed must be 2–2.5× higher plus middling recirculation load.

Mistake 2: Accept vendor maximum capacity without PSD constraint

Vendor brochures show peak throughput for coarse powder. If you require beverage-grade narrow PSD, peak rates are not sustainable. Request trial data at your target D90.

Mistake 3: Ignore thermal limit when pushing capacity

Overfeeding → particle overcrowding → poor heat removal → powder outlet temperature >65°C → protein denaturation, loss of emulsification/solubility. Functional quality cannot be recovered downstream.

Mistake 4: Underestimating middling recirculation load

High-purity operation requires higher middling recycle. The mill interior must process combined fresh feed + recycle. Many suppliers quote capacity on fresh feed only, excluding recycle.

8. Step-by-step sizing workflow checklist

1. Define target annual/daily/hourly protein concentrate output
2. Complete mass balance to calculate required fresh flour feed into the mill
3. Confirm target PSD specification (D50, D90, maximum coarse tail %)
4. Apply reduction factor for fine beverage-grade quality (−20~40%)
5. Add safety reserve factor (1.15–1.40)
6. Estimate middling recirculation ratio and communicate to machine supplier
7. Verify that matched air system (fan, cyclone, dust filter) supports the full throughput
8. Request vendor bench/pilot test data using your raw material at your target fineness
9. Confirm continuous running powder temperature remains below 60 °C at quoted operating feed rate

9. Simple practical example

Target: 1,000 kg/h beverage-grade pea protein concentrate
Fine fraction yield = 40%
Base fresh feed = 1,000 / 0.4 = 2,500 kg/h
PSD requirement demands 25% capacity derate → equivalent base capacity = 2,500 × 1.25 = 3,125 kg/h
Safety reserve 1.20 → 3,125 × 1.20 = 3,750 kg/h nominal mill feed capacity
Then confirm the model can run continuously at ~3,100 kg/h fresh feed plus recirculation without temperature rise and PSD drift.

The correct mill capacity selection starts with mass balance, accounts for PSD-induced throughput derating, includes middling recirculation and applies a safety reserve.
Prioritise sustainable capacity at your target quality specification, not just brochure peak throughput. For beverage-grade dry fractionated protein, avoid running the mill at maximum nominal load; moderate feed rates preserve particle liberation, narrow PSD and native protein functionality.