Who this is for
Farmers integrating forward, cooperatives, millers and feed producers, oilseed crushers and packhouses in South Africa who need to know which tier they are really in, what it costs in 2026 and what the site and the supply chain must have before an order. It reads across our agro-processing price guides so you can find your row and go to the page that prices it.
The conditions that apply to every agro-processing line
| Condition | What it means for the line | What to have in hand |
|---|---|---|
| Crop supply and working capital | A mill or press runs on the grain or seed you can buy and store; the working capital in stock is the number first-time processors forget and it can rival the equipment | Supply contracts or own-farm tonnage, storage for one to eight weeks by tier, a working-capital line |
| Fortification and food premises | Maize meal and wheat flour are subject to mandatory fortification regulations; any food line needs a certificate of acceptability for the premises under R638 of 2018 | Fortificant dosing in the mill scope; premises layout reviewed before fit-out |
| Power | Mills, presses and pellet presses are motor-heavy: a 30 t/day maize mill is roughly 80–140 kW installed; three-phase supply and a backup that finishes the batch during an outage | OEM installed-load list; generator and UPS decision with the power guide, solar options in the solar guide |
| Import duty and VAT | Most milling, pressing and feed machinery under HS 84 enters at 0% duty-free; the exceptions to plan for are refrigeration equipment, some motors and large generating sets; import VAT is 15% on the customs value plus a 10% uplift plus duty, recoverable for a VAT-registered manufacturer | Eight-digit tariff codes per machine; run the landed cost estimator |
| By-products and effluent | Bran, oilcake and pellet fines are products with buyers; refining effluent and washdown water need a permitted route | By-product off-take priced into the model; effluent route agreed with the municipality |
The scenarios, one by one
Scenario 1: a first maize mill
Thirty tonnes a day, one to one-and-a-half shifts, at US$250–400k in mill and packaging equipment, for a business with up to about 700 tonnes of meal a month it can actually sell. Oversizing is the single most expensive mistake in this category. Maize mill and flour mill prices and sizing carries the three tiers; the regional maize mill case study shows a 60 tonne-per-day delivery.
Oversizing is expensive because a mill earns on what it sells, not on what it can grind. A 30 tonne-a-day mill running one comfortable shift against secured offtake beats a 60 tonne mill running a third of the time: the larger plant carries more capital, a bigger grain buffer and a payroll sized for shifts it is not working. Grow into the next tier on demonstrated sales, not on a projection.
Two things decide the business before the machine list does. Grain has a season and a price that moves with it, so working capital in stock is usually the binding constraint rather than equipment; meal has a shelf life and a delivery radius, so the sales plan has to empty the warehouse on a schedule. A mill sized to a market you have not yet won is a warehouse with a motor in it.
Scenario 2: a regional 60 tonne-per-day mill
Two to three shifts, a branded product, two to four weeks of grain buffer and a team of eight to fourteen, at US$700k–1.1M. It is a different business from the 30-tonne mill, not a bigger version of it, and the sizing guide sets out the off-take that justifies it.
At this tier the product is branded and sold into formal retail, which brings obligations the entry tier does not carry. Maize meal and wheat flour are subject to mandatory fortification, so a dosing system, a premix supply and an inspectable record belong in the plant from commissioning; printed packaging adds a supplier and a stock item. Neither is expensive alone, and both are awkward to retrofit into a running mill.
Running two or three shifts also changes maintenance from a Sunday habit into a schedule. Roll and sieve wear is proportional to hours, so a planned window and a wear-part stock become part of the operating model — and with scheduled interruptions on the grid, the restart discipline after an outage matters as much as the maintenance itself.
Scenario 3: hammer mill or roller mill
A straight hammer mill is the workhorse for animal feed and coarse community milling; a roller mill makes the graded meals and flours retail buyers pay for. They are different purchases with different by-product streams, and the choice follows the product, not the budget. The maize guide draws the boundary and prices both classes.
The choice is decided by what you sell, not by what you mill. A hammer mill produces a whole meal at low capital and high tolerance for mixed grain; a roller mill separates the fractions, which is what gives you super maize meal, special maize meal and a bran stream that has its own buyer. Retail buyers pay for the grade, and the grade needs rolls.
By-products are the hidden half of the comparison. A roller mill's bran and germ are saleable products with real demand from the feed sector, and that revenue changes the economics of the line; a hammer mill has nothing to separate and therefore nothing to sell. Count the by-product income before comparing the two capital numbers.
Scenario 4: a small feed pellet line
One to three tonnes an hour at US$60–150k landed and commissioned, for a farm or a local feed business that formulates its own rations. Hammer mill, mixer, pellet press, cooler and bagging make up the line; the formulation and the raw-material supply decide whether it pays. Animal feed pellet machine prices and feed pelletising line sizing cover the tiers.
Pellet quality is a conditioning question, not a die question. Steam conditioning before the die gelatinises the starch and gives a pellet that survives handling; under-conditioned pellets crumble into fines, and fines are feed the animal does not eat and the customer does not pay for. That is why the step from this tier to a commercial plant is mostly conditioning and dosing, not press size.
For a farm or a local feed business the case rests on formulating your own rations: buying maize, bran, soya and micro-ingredients and dosing them accurately is where the margin against bought feed comes from. That makes weighing accuracy and a micro-ingredient system earn their place long before extra tonnes an hour do.
Scenario 5: a commercial feed plant
Three to ten tonnes an hour at US$250–900k, with industrial plants at ten tonnes an hour and above from about US$1.5M and priced per project. Batching accuracy, dust control and bagging throughput are the design items that separate a plant from a large machine.
A commercial plant is a materials-handling business with a pellet press in it. Batching accuracy decides whether the specification on the bag is the specification in the bag, dust control decides whether the plant passes an inspection and whether an explosive dust risk is managed, and bagging throughput decides whether the plant can dispatch what it made. None of those three is the press.
Power is the constraint that surprises buyers. Feed plants are motor-heavy and run long hours; a three-phase supply that holds voltage, and a backup strategy that lets a batch finish rather than merely keeping the lights on, are part of the plant specification. An interruption mid-batch is scrap, not a pause.
Scenario 6: cooking oil pressing at 20 tonnes a day
A 20 tonne-per-day seed line, pressing and filtering to crude oil, is US$150–400k in equipment; the entry screw press under US$10k produces crude oil only and is a machine purchase the guide marks as a boundary rather than a tier. Cooking oil machine prices sets out the four tiers and what the seed yield means for the business case.
The entry screw press is the clearest example in this guide of a machine that is not a line. It produces crude oil and cake; what it does not produce is a filtered, settled, packaged product with a shelf life, and the gap between the two is filtration, storage and packaging. Buyers who start from the press usually end up buying the rest of the line afterwards, at a worse price and out of sequence.
At 20 tonnes a day the business sells crude oil and oilcake to other processors, which is a business-to-business operation with few packaging obligations and a short customer list. That is a legitimate place to start, and it is a different company from the one that sells a bottled product on a supermarket shelf.
Scenario 7: refining at 50 tonnes a day and above
Adding a refining section takes the line to US$400–900k at 50 tonnes a day and US$700k–1.5M and above at 100 tonnes a day. Refining brings effluent, steam and chemical handling into the project, and the oilcake becomes a feed product with its own buyers.
Refining is the step change, and it is a plant rather than an addition. Neutralising, bleaching and deodorising bring steam, chemical handling and an effluent stream into a project that previously had none of them, and the effluent needs a permit and a route. Budget the utilities and the permit alongside the refining section, not after it.
What you gain is the consumer product and the margin that goes with it — and a second revenue line in oilcake, which has a reliable buyer in the feed sector. The seed catchment has to support the tier: 50 and 100 tonne-a-day plants need thousands of tonnes a season, contracted, and that is usually the real constraint rather than the equipment.
Scenario 8: a packhouse or processor cold room
Chiller rooms at US$15–38k installed and blast freezers from US$60k are sized on throughput and dwell time. For fresh produce the cold room is often the first investment and the one that makes the packhouse saleable to retail buyers. Cold room and blast freezer sizing covers the calculation.
A packhouse is sized on two numbers buyers routinely merge: how much arrives at once, and how fast it has to come down in temperature. A chiller room holds produce already cooled; pulling field heat out of a fresh delivery against the clock is a different duty, and sizing on room volume alone produces a cold room that is full, cold and slowly spoiling what was put in warm.
Cold chain is also what makes the rest of the operation saleable. Retail and export buyers audit temperature records, so the room, the monitoring and the backup are part of the commercial case rather than an overhead. With scheduled interruptions the backup has to carry the compressors — a cold room that loses its compressors for two hours a day is not a cold room.
Scenario 9: recovering an old mill, or a cooperative and development-finance project
An existing mill with a dead controller and a worn roll stand is often a control retrofit at US$2–8k per machine and a roll regrind rather than a new plant. Cooperative and development-finance projects add a feasibility study a lender can read and, where the funder scores local content, a scope split between imported and locally fabricated elements. See new, used or recover, when to upgrade your PLC and local content scoring.
A control retrofit plus a roll regrind is the highest-return move available to an existing mill, and it is routinely skipped because it is unglamorous. The frame, the stands and the drives are usually sound; what has expired is the electronics and the roll surface. At US$2–8k per machine for the control and a regrind for the rolls, the plant comes back with the repeatability the new mill was being bought for.
Cooperative and development-finance projects add a requirement the equipment does not: a feasibility a lender can read, with offtake, working capital and a maintenance plan stated separately from the capital request. Funders decline on the business case far more often than on the machine list, which is why we scope the offtake before the plant.
The map on one table
| Scenario | Realistic tier | 2026 band (USD) | Site must have | Prices it |
|---|---|---|---|---|
| First maize mill | 30 t/day roller mill + packaging | 250–400k equipment | 1–2 weeks grain storage, ~80–140 kW, fortification dosing | Maize mill prices |
| Regional mill | 60 t/day, 2–3 shifts | 700k–1.1M | 2–4 weeks grain, team of 8–14, bran off-take | Maize mill prices |
| Hammer or roller | Machine class by product | Within the milling tiers | Product spec: feed and coarse meal vs graded meal | Maize mill prices |
| Small feed pellet line | 1–3 t/h | 60–150k landed and commissioned | Raw-material supply, formulation, bagging | Feed pellet prices |
| Commercial feed plant | 3–10 t/h | 250–900k; industrial from ~1.5M | Batching accuracy, dust control, intake | Feed line sizing |
| Oil pressing | 20 t/day seed | 150–400k; entry press under 10k is a machine, not a line | Seed supply, oilcake off-take | Cooking oil prices |
| Oil refining | 50 t/day with refining | 400–900k; 100 t/day 700k–1.5M+ | Steam, effluent route, chemical handling | Cooking oil prices |
| Packhouse cold room | Chiller room; blast freezer if frozen | 15–38k installed; blast from 60k | Compressor power and backup | Cold room sizing |
| Recover an old mill / cooperative or DFI project | Retrofit first; feasibility a lender can read | 2–8k per machine retrofit | Drawings and PLC access; funder scoring rules | New, used or recover |
| Fruit and vegetable drying | 2–5 t/day fresh intake heat-pump plant | 150k–350k landed and commissioned | Prep area, drying chambers, power backup | Dehydration plant cost |
| Sorghum and millet meal | Small mill set; complete plant per configuration | 10k–50k machine set | Dehulling, grading regulations, packaging | Sorghum milling cost |
| Pet food or floating fish feed | 1–2 t/h extrusion line | 75k–500k equipment | Steam boiler, Act 36 registration per product | Extrusion plant prices |
| Groundnut and nut processing | 300–500 kg/h groundnut line | 120k–300k landed and commissioned | Mycotoxin control, drying, export inspection for nuts | Nut processing cost |
| Small packhouse | 2–5 t/h wash-wax-size line | 10k–75k equipment; packhouse per configuration | PUC/PHC registration, cold room | Packhouse line prices |
| Organic fertiliser from poultry litter | 3–5 t/h granulation plant | 200k–500k landed and commissioned | Composting area, Act 36 fertiliser registration | Fertiliser plant cost |
Rows added October 2026 price new lines in their own guides; comparing across industries? See which production line to start in South Africa.
Bands are indicative USD, 2026, on the basis stated in each row; they are the same bands our price guides publish, so the row and the guide always agree. Machinery is priced in US dollars and the rand cost moves with the exchange rate.
What CISH does with your row
The feasibility call starts from the crop you can buy and the product you can sell, confirms the tier, prices the whole project including storage and power, and hands you a written scope a lender or a cooperative board can read. Delivery is the same for every row: turnkey lines commissioned to an agreed output on your grain or seed, operators trained during run-up, documentation handed over.
Frequently asked questions
What size maize mill should a first-time miller buy in South Africa?
The mill that matches the meal you can sell per month through channels you can name, which for most first-time millers is 30 tonnes a day at US$250–400k in equipment. A 60-tonne mill bought for growth runs two days a week and carries fixed costs for years. The sizing guide gives the off-take thresholds for each tier.
How much does a small animal feed plant cost?
A 1–3 tonne-per-hour pellet line is US$60–150k landed and commissioned; 3–10 tonnes an hour is US$250–900k. Formulation and raw-material supply decide whether the plant pays, and the feed pellet price guide sets out what each tier includes.
Is a cheap oil press a real business?
A screw press under US$10k produces crude oil only and suits own-use or a very local market. A 20 tonne-per-day line at US$150–400k presses and filters at a scale that supports a brand; refining at 50 tonnes a day and above is a further step with its own effluent and steam requirements.
Do I have to fortify maize meal?
Yes. Maize meal and wheat flour are covered by South Africa's mandatory fortification regulations, so a fortificant dosing point belongs in the mill scope and the premix in the operating cost. Ask for it in the quote rather than discovering it at the first inspection.
How much grain do I need to store?
One to two weeks of intake for a 30-tonne mill, two to four weeks for a 60-tonne mill, and four to eight weeks or more at industrial scale. The working capital in that stock is often the largest number after the equipment, and it belongs in the feasibility model from day one.
Can a cooperative fund a mill with development finance?
Development finance institutions fund agro-processing regularly, and what they need is a feasibility study with an off-take model, a site assessment and a scope a lender can read. We produce that document as part of the feasibility stage; we do not provide finance.