Mistake number one in payback math is taking the machine price as the investment. A complete project invests in at least six items:
| Item | Content | Typical share (industry ref.) |
|---|---|---|
| ① Core equipment | Process-section machines | 60%–75% |
| ② Auxiliaries | Elevators, cleaning, hoppers, controls | 8%–15% |
| ③ Logistics | Sea freight, inland transport, customs | 5%–12% |
| ④ Civil works | Building adaptation, machine foundations | 5%–15% (less with an existing building) |
| ⑤ Installation | Erection, commissioning, acceptance runs | 3%–8% |
| ⑥ Spares & training | First-year wear parts, operator training | 1%–3% |
Shares vary by project; the key move is demanding an itemized quotation that lands items ②–⑥ — a "machine price × 1.15" shortcut breaks on logistics and civil works.
AmGrainTech processing lines cover rice milling, flour milling and grain processing complete sets, exported to 30+ countries over 20+ years. Quotations are issued item by item (equipment / auxiliaries / logistics / installation), installation is by our own engineers, and every supply includes operator training and a first-year wear-parts list.
1. Revenue: Two Operating Models
Model A — toll milling: processing for farmers or traders for a fee. Revenue = annual throughput × fee per tonne. Simple, but fees are locally competitive and utilization swings with the harvest.
Model B — buy and sell: purchase paddy, mill and sell rice. Revenue = sales volume × (rice price − paddy cost converted), plus by-product income (bran, broken rice). Better margin, but inventory and price risk, and a higher management load.
Most real mills run a mix: toll milling through the harvest keeps utilization up; self-purchased processing in the off-season captures the spread. Model your actual mix — not the ideal one.
2. Operating Costs: Three Big Items
- Power: rice milling commonly runs 40–70 kWh per tonne of paddy (line-type dependent, industry reference). Tariff × unit consumption is the single largest operating item.
- Wear parts: rubber rolls, abrasive rolls, screens — rubber rolls are budgeted per tonne milled; they are a consumable, not a capital item.
- Labor: semi-automatic lines run 4–6 operators per shift (industry reference). Automation compresses labor but raises depreciation — balance the two.
3. Worked Example: A 2 t/h Rice Milling Line
Assumptions (illustrative figures, industry-reference frame):
- Total investment $160,000 (equipment $120k + auxiliaries, logistics, civil, installation, spares $40k)
- Annual throughput 4,000 t paddy (two shifts; concentrated harvest plus off-season stock)
- Mixed revenue: 60% toll milling at $18/t + 40% self-sold margin at $35/t (by-products included)
- Operating costs $52,000/year (power $22k + wear parts $12k + labor $15k + other $3k)
Calculation:
Toll revenue = 4,000 × 60% × $18 = $43,200
Self-sell margin = 4,000 × 40% × $35 = $56,000
Net annual cash flow = $43,200 + $56,000 − $52,000 = $47,200
Static payback = $160,000 ÷ $47,200 ≈ 3.4 years
Sanity check: well-run, high-utilization milling lines commonly pay back in 2.5–3.5 years (industry reference; trade-media models put quality engineering recovery at 24–36 months) — the example lands inside the plausible band.
4. Sensitivity: Which Assumptions Move the Answer
| Variable | Change | Payback moves |
|---|---|---|
| Utilization | 4,000 t → 3,000 t (−25%) | 3.4 → ~5.0 years |
| Tolling fee | $18 → $14/t | 3.4 → ~4.2 years |
| Power price | +30% | 3.4 → ~3.9 years |
| Head rice recovery | +2 percentage points (more self-sell upside) | 3.4 → ~2.9 years |
Two conclusions: utilization is the first variable, followed by unit fee/margin — which is why "capacity matched to real paddy supply" beats "as big as possible": utilization you cannot achieve sinks any machine price. Second, recovery improvement is systematically underweighted: recovery is not a fixed constant — it moves with paddy condition, operation and maintenance — but the +2–5 point improvement from renewed process sections (industry reference) compounds through self-sell margin year after year. Don't count only the electricity saved. For the recovery mechanics, see what drives head rice yield.
5. Three Errors That Distort the Math
- Missing working capital: the buy-and-sell model finances paddy in harvest season — it is not depreciation, but it dominates your real cash pressure. Plan it separately.
- Modeling at full utilization: the most common optimism bias. Use a ramp: 60% year one, 80% year two, 90% from year three.
- Ignoring price cycles: rice-paddy spreads swing between years. Use three-year averages and run pessimistic/neutral/optimistic cases.