Cluster article
Understanding Flour Extraction: Definitions, Measurement and Benchmarks
A technical note on flour extraction rate — how it is measured, why mill extraction differs from commercial extraction, and benchmark figures by wheat type and product.
Key takeaways
- Mill extraction is measured on a clean wheat basis; commercial extraction on as-received wheat. The two differ by 1.5–4%.
- Laboratory extraction (Bühler test mill) is the benchmark — production mills typically run 1–3 points below it.
- Benchmark: soft wheat 72–75%, hard wheat 76–79%, durum semolina 62–68%.
Three definitions that get confused
1. Mill extraction = flour ÷ clean wheat into mill × 100
2. Commercial extraction = flour ÷ wheat purchased × 100
3. Laboratory extraction = flour ÷ wheat in a Bühler or similar test mill × 100
| Definition | Hard wheat example |
|---|---|
| Laboratory | 80% |
| Mill | 77% |
| Commercial | 74.5% |
The 2.5-point gap between mill and commercial is cleaning loss plus moisture shrinkage. The 3-point gap between laboratory and mill is scale, flow sheet differences and operator reality.
When a supplier quotes an extraction figure, ask which definition. The difference between 78% and 74% commercial is USD 165,000 a year at 60 t/d.
Benchmarks
| Wheat / product | Laboratory | Production mill |
|---|---|---|
| Soft wheat, patent | 76–78% | 72–75% |
| Medium wheat, two grades | 79–81% | 75–78% |
| Hard wheat, two grades | 80–82% | 76–79% |
| Hard wheat, high extraction | 83–85% | 79–82% |
| Durum, semolina | 65–70% | 62–68% |
| Whole wheat / Atta | 97–99% | 85–95% |
Measurement
The only reliable production measurement is a mass balance over a defined period:
- Weigh wheat entering the mill (clean wheat, after cleaning)
- Weigh flour leaving (all grades, before packing loss)
- Weigh bran leaving
- Reconcile — the three should sum to the input within 1%
Run it over a full 24 hours, not a shift, to average out surge and buffer effects.
Common measurement errors:
| Error | Effect |
|---|---|
| Using wheat purchased instead of clean wheat | Understates extraction by 2–4 points |
| Ignoring flour in the dust collector | Understates by 0.2–0.5 points |
| Weighing bags instead of using a flow meter | ±1–2 point noise |
| Short measurement window | Random error dominates |
Improving extraction: ranked
| Action | Gain | Cost |
|---|---|---|
| Add bran finisher | +1.5–3.0 points | USD 6,000–15,000 |
| Re-flute rolls on schedule | +2–3 points | USD 40–120 per roll |
| Re-balance purifier air | +2–4 points | Technician time |
| Add a break passage | +1–2 points | USD 30,000–45,000 |
| Add tailing passage | +0.5–1.5 points | USD 15,000–30,000 |
| Fix sifter screen leaks | +0.5–2 points | Negligible |
| Improve tempering control | +0.5–1.5 points | USD 12,000–25,000 |
The first three are cheap or free and are the reason well-maintained older mills outperform neglected newer ones.