Cluster article
Staffing a Flour Mill: Headcount, Roles and the Training That Separates Good Mills from Mediocre Ones
How many people a flour mill actually needs, what each role does, the skills that take years to build, and the training structure that keeps a mill running when the head miller leaves.
Key takeaways
- Labour is 4–8% of flour mill operating cost, but the head miller decides 20–30% of your extraction rate.
- One head miller per 100–150 t/d is the norm; below 60 t/d the role is usually combined with production management.
- The single biggest staffing risk is a mill that depends on one person's undocumented knowledge.
Headcount by capacity
| Capacity | Production | Maintenance | Lab/QC | Admin/sales | Total |
|---|---|---|---|---|---|
| 20 t/d | 2–3 | 1 (shared) | 0.5 | 1–2 | 4–6 |
| 60 t/d | 5–7 | 1–2 | 1 | 2–3 | 9–13 |
| 100 t/d | 8–10 | 2–3 | 2 | 3–4 | 15–19 |
| 200 t/d | 15–18 | 3–4 | 3 | 4–6 | 25–31 |
| 500 t/d | 30–40 | 6–9 | 5–7 | 8–12 | 49–68 |
The rule of thumb: one employee per 10–15 t/d across the whole business, with larger plants at the efficient end.
The roles that matter
Head miller
The most consequential person in the plant, and the one most often hired last.
| Responsibility | Consequence of getting it wrong |
|---|---|
| Roll gap setting | 2–5 points of extraction |
| Flow sheet balance | Ash specification, throughput |
| Sifter screen selection | Granulation, sifting efficiency |
| Purifier setting | Semolina purity, ash |
| Adjustment to wheat change | Everything downstream |
| Quality troubleshooting | Customer rejections |
What a head miller actually does cannot be written in a manual: judging stock by squeezing it, listening to a sifter, knowing that this wheat needs 0.03 mm less gap than the last load. It is built over 5–10 years.
Two head millers is not a luxury at 200 t/d. It is risk management. Mills with one head miller and no deputy lose 3–9 months of performance when that person leaves.
Shift miller / operator
Runs the plant on shift. Monitors, adjusts within limits, responds to alarms, records.
| Skill | Training time |
|---|---|
| Start-up and shutdown sequence | 2–4 weeks |
| Normal operating parameters | 1–2 months |
| Alarm response | 3–6 months |
| Minor adjustment within limits | 6–12 months |
| Independent judgement | 2–3 years |
Maintenance technician
| Discipline | Per 100 t/d |
|---|---|
| Mechanical | 1.5 |
| Electrical | 0.7 |
| Instrumentation/automation | 0.3 |
Mechanical dominates because bearings, belts, screens and spouting are the bulk of the work. Instrumentation is increasingly needed as PLCs and sensors proliferate — a mill with 200 sensors and nobody who understands them has 200 liabilities.
Laboratory technician
| Test | Skill level |
|---|---|
| Moisture, ash | Basic, 2 weeks training |
| Protein (Kjeldahl/NIR) | Intermediate, 1 month |
| Falling number | Basic, 2 weeks |
| Farinograph | Intermediate, 1 month |
| Extensograph | Intermediate, 1 month |
| Alveograph | Advanced, 2 months |
| Baking test | Advanced, 3–6 months |
A mill producing to specification needs at least one person who can run a farinograph and interpret the curve. Without it, the mill is producing to a number it cannot verify.
Labour cost
| Plant | Staff | Annual labour cost | Cost per tonne |
|---|---|---|---|
| 60 t/d | 11 | USD 88,000 | USD 5.60 |
| 200 t/d | 28 | USD 280,000 | USD 5.40 |
| 500 t/d | 58 | USD 580,000 | USD 4.60 |
At USD 8,000 average annual cost per employee, typical of Chinese inland provinces; multiply for higher-cost markets.
Labour is 4–8% of operating cost in most markets — low enough that cutting headcount is rarely the right response to a cost problem, and high enough that productivity matters.
Training structure
The four levels
| Level | Content | Duration | Who |
|---|---|---|---|
| 1. Induction | Safety, plant layout, emergency procedures | 1 week | Everyone |
| 2. Role | Specific equipment and procedures | 1–3 months | Everyone |
| 3. Cross-training | Adjacent roles and the wider flow sheet | 6–12 months | Operators |
| 4. Judgement | Flow sheet behaviour, quality diagnosis | Years | Potential head millers |
Most mills do levels 1 and 2 and stop. Levels 3 and 4 are what produce people who can run the plant when something unusual happens.
What level 4 looks like
A head miller diagnosing a problem thinks like this:
- Ash is up 0.04% and extraction is down 1.5 points.
- Check whether wheat changed — if yes, adjust dampening.
- If wheat unchanged, check break release. Break 1 release down means flutes worn or gap drifted.
- Check sifter tailings. Rising tailings on 2M means screen blinded.
- Check purifier. Reduced semolina purity raises ash in the patent stream.
- Only then adjust.
An untrained operator’s version: close all the gaps and hope.
Documentation as training
| Document | Purpose |
|---|---|
| Flow sheet with all passages | The plant on one page |
| Operating parameters by passage | Normal ranges |
| Roll gap specification by passage | What to set and why |
| Sifter screen specification by passage | What mesh, why |
| Troubleshooting flowchart | Symptom → cause → action |
| Start-up and shutdown procedure | Written, signed |
| Emergency procedures | Fire, dust, injury, power loss |
A mill whose knowledge lives only in the head miller’s head does not have a training problem. It has a continuity problem, and it will surface the day that person is unavailable.
Shift patterns
| Pattern | Hours | Recovery | Notes |
|---|---|---|---|
| 8-hour, 3 shifts | 24 h coverage | Good | Standard |
| 12-hour, 2 shifts | 24 h coverage | Fatigue risk | Fewer handovers, more fatigue |
| 24/5 with weekend shutdown | 120 h/week | Full weekend | Maintenance window |
| 24/7 continuous | 168 h/week | Rotating | Needs more staff, higher cost |
Most 60–200 t/d mills run 24/5 with a weekend maintenance window. It aligns maintenance with low demand and avoids the cost and error rate of night-shift-only operations.
Handover discipline matters more than the pattern. A 10-minute structured handover — what changed, what is abnormal, what is pending — prevents more failures than any amount of training.
The staffing risks
| Risk | Mitigation |
|---|---|
| Single point of knowledge (head miller) | Hire a deputy, document everything |
| No instrumentation skills | Train one technician or contract it |
| No lab capability | Shared lab arrangement with a nearby mill or a service lab |
| Operator turnover above 15%/yr | Pay above local average, train visibly, create progression |
| No cross-training | Deliberate rotation, annual |
| Aging workforce | Apprentice programme, 3–5 year horizon |
The most common failure in new mills is not equipment. It is a plant with no one who can set it up, staffed by people who were trained by the installer for two weeks and then left alone.