Conversion cost on a 30-inch, 28-gauge single jersey machine lands at about USD 0.17–0.18 per kg of grey fabric at 85% efficiency, and only USD 0.038 of that is machine depreciation. Move the machine price 20% and the per-kilo figure shifts by under 1%. Drop efficiency from 85% to 60% and it rises 42%. That ratio, not the sticker price, is what most machine decisions actually turn on.
The Five Lines That Make Up Conversion Cost
Conversion cost is everything spent turning purchased yarn into greige fabric, divided by kilos shipped. Yarn is excluded here because it is a pass-through the machine barely influences. The five lines below are indicative values for the base case: 30-inch, 28-gauge, 96 feeders, single jersey at 180 g/m², 25 rpm, three shifts.
| Cost line | USD per running hour | USD per kg at 22 kg/h | Basis and assumption |
|---|---|---|---|
| Machine depreciation | 0.84 | 0.038 | USD 45,000 landed, 8-year life, 6,000 h/yr, 10% residual |
| Power | 0.83 | 0.038 | 7.5 kW total draw, USD 0.11/kWh (band 0.08–0.15) |
| Needles and spares | 0.47 | 0.021 | USD 2,800/yr consumption at this utilisation |
| Labour | 1.10 | 0.050 | USD 1,100/month all-in, 1 operator per 2 machines, 500 h/machine-month |
| Facility and overhead | 0.60 | 0.027 | Floor space, compressed air, chiller share, administration |
| Total conversion | 3.84 | 0.175 | Indicative, excludes yarn and finishing |
Every figure is a range in real quotations, not a fixed cost. The 0.175 is typical for this specification in a coastal Chinese mill; a mill paying USD 0.08/kWh and one operator per four machines lands nearer 0.14. Depreciation assumptions are covered in more depth in the total cost of ownership model.
The Formula, and One Worked Example
Write it once and you can re-run it for any quote in ten minutes:
Cost per kg = (D + P + S + L + O) ÷ Q, where D = landed price × (1 − residual share) ÷ (life years × running hours per year), and Q is output in kg per running hour.
Base case: D = 45,000 × 0.90 ÷ (8 × 6,000) = 40,500 ÷ 48,000 = USD 0.84/h. Adding P 0.83, S 0.47, L 1.10 and O 0.60 gives USD 3.84/h. At 22 kg/h, cost is 3.84 ÷ 22 = USD 0.175/kg.
Add yarn to reach fabric cost: 1.04 kg of 30s combed cotton yarn per kg of fabric at USD 3.05/kg gives USD 3.17, so grey fabric lands at about USD 3.34/kg. Conversion is roughly 5% of that total — which is exactly why the machine line has to be modelled rather than argued. Landed price composition is broken down in the import cost calculator.
Why a 20% Dearer Machine Lands Within 1% per kg
Take two quotations for the same diameter and gauge. Machine A costs USD 38,000 FOB and runs 22 kg/h. Machine B costs USD 45,500 FOB — 19.7% more — but carries more feeders and holds 23.5 kg/h on the same fabric. Both are priced at a CNY 7.10–7.20 per USD band, the 2026 Q4 range.
| Line | Machine A | Machine B |
|---|---|---|
| Landed price (USD) | 45,000 | 53,500 |
| Depreciation (USD/h) | 0.84 | 1.00 |
| Output at 85% efficiency (kg/h) | 22.0 | 23.5 |
| Depreciation per kg (USD) | 0.0382 | 0.0427 |
| Total conversion per kg (USD) | 0.175 | 0.173 |
| Annual conversion spend at 6,000 h (USD) | 23,047 | 24,449 |
A 19.7% sticker gap becomes a 1% per-kilo gap, and an annual saving of about USD 1,400 against USD 8,500 of extra landed cost. Simple payback is 6.1 years on an 8-year asset — inside life, but thin. Machine B is not automatically the better buy; it is only better if you can sell the extra 9,000 kg/yr and hold the machine six years or more. Sticker ranges for this class of machine are in the 2026 price guide.
What Actually Moves the Number: Efficiency, Then Yield
Holding hourly cost at USD 3.84, output alone sets the per-kilo result. At 85% efficiency the machine yields 22.0 kg/h and costs 0.175/kg. At 75% it yields 19.4 kg/h and costs 0.198/kg. At 60% it yields 15.5 kg/h and costs 0.248/kg. A 25-point efficiency slide is a 42% cost increase — larger than any realistic negotiation on machine price.
The same arithmetic applies to yield loss. A 2% seconds rate at USD 3.34/kg of fabric removes USD 0.067 of value per kg sold, which is four times the depreciation line. Buyers who negotiate hard on the machine and accept a vague acceptance test are optimising the smallest number on the page. Spare-part consumption per 1,000 kg, not spare-part unit price, is what belongs in S; the spare parts consumption model covers how to measure it.
When This Conclusion Reverses
The model above assumes 6,000 running hours, a stable single-jersey mix and a machine held to end of life. Break any of those and the ranking changes:
- Utilisation below 5,000 h/yr. At 3,500 h/yr, depreciation becomes USD 1.45/h and total conversion rises to about 0.202/kg. Capital-heavy machines get worse fast when the floor is idle.
- Heavy fabrics and jacquard. Output drops to 8–14 kg/h on rib, interlock and structured jacquard, so the depreciation line triples as a share of cost. On that mix, price differences between machines matter far more than they do on plain jersey.
- Low power tariffs. Below USD 0.06/kWh the power line halves and the spread between machines narrows further; above USD 0.15/kWh a machine that draws 1 kW less earns its premium on electricity alone.
- Short holding periods. Sell at five years and residual value, not depreciation, dominates. Machines with a thin secondary market lose 65–75% of landed value in that window.
- Currency. A 5% CNY move against USD shifts landed cost by roughly USD 2,000 on this class of machine — more than the annual per-kg difference being debated. Fix the currency and validity window before comparing quotes, as the negotiation guide sets out.
If yarn is 92% of your fabric cost and your mix never changes, machine selection is close to noise and you should buy on service response instead. That is a legitimate outcome of this model, not a failure of it.
What to Demand in the Quotation So the Model Is Checkable
Ask for five numbers in writing and the model becomes auditable: guaranteed output in kg/h at a stated efficiency and fabric, measured power draw at that speed, needles and spares consumption per 1,000 kg, acceptance-test protocol with seconds-rate threshold, and a spare-parts price list with a validity date. Missing any one of them turns the comparison into a conversation about impressions.
Run all three scenarios — 60%, 75% and 85% efficiency — before signing. The machine that wins at 85% is often not the machine that wins at 60%, and your floor will spend real weeks at 60% during changeovers and yarn changes. AISUNNY quotes output against a named fabric and efficiency rather than a top speed, which is the only form of the number you can put into the formula above.
Frequently Asked Questions
Does cost per kg include yarn?
No. The 0.17–0.18 figure is conversion only: depreciation, power, needles and spares, labour, overhead. Yarn is added separately at roughly 1.03–1.08 kg per kg of fabric depending on gauge and structure.
Is USD 0.17 per kg conversion typical for single jersey?
It is indicative for a 30-inch, 28-gauge frame running 6,000 hours a year in a coastal Chinese mill. Well-utilised mills with cheaper power and wider operator spans reach 0.13–0.15; low-utilisation plants with older machines run 0.22–0.28.
Why does efficiency matter more than machine price?
Because hourly cost is nearly fixed while output is not. Hourly cost varies a few percent between machines; output varies 40% between a well-run and a badly-run floor. The divisor dominates the result.
What running hours should I use?
Your own. Three shifts at 90% uptime is about 6,500 h/yr; two shifts is about 4,300. Using a supplier’s 7,000 h assumption on a two-shift plant understates depreciation per kg by a third.
At what point does the dearer machine pay back?
When it produces at least 6% more usable kg/h, you run over 5,500 h/yr, you hold it six years or longer, and you can sell the extra output. Fail the volume condition and the premium becomes a donation.
Should I accept the supplier’s kg/h figure?
Only with the fabric, gauge, efficiency and measurement window attached. A kg/h number without those four parameters cannot enter the formula, and an acceptance test written without them cannot enforce anything.
