Who this is for
Plant owners, finance leads, and engineering managers comparing production-line options — new line, upgrade, or expansion — who want to make the decision on real lifetime economics rather than the sticker price. Also useful for anyone building a capex business case for a board or a lender.
Why the purchase price misleads
The purchase price is the most visible number and the easiest to compare, so it dominates decisions it shouldn't. But over a 10-year life, the equipment price is frequently 15–35% of total cost of ownership (TCO) for an automated line — and less for labour-intensive operations. The other 65–85% is what happens after the line is installed: energy, labour, materials, spares, maintenance, downtime, and the cost of the money used to buy it.
This is why "we went with the cheaper quote" so often turns into "this line has cost us a fortune." The capex saving was real; the lifetime cost was higher.
Decision rule: compare line options on 10-year total cost of ownership per unit of output — cost per bottle, per block, per tonne — not on purchase price. The line with the lowest cost-per-unit over its life is the cheapest line, even if it is the dearest quote.
The components of TCO
1. Acquisition cost
Not just the FOB equipment price — the full landed-and-commissioned cost: freight, insurance, duty, VAT (reclaimable for VAT vendors), clearance, inland transport, installation, commissioning, and initial spares. Typically 130–160% of FOB equipment price (see the import guide). This is the number most people mean by "the price," and it is only the start.
The gap between FOB and landed-and-commissioned is where most first budgets break, and it is not one surprise but a dozen small ones arriving in sequence. Freight and duty are the visible half; clearance delays, inland transport for an abnormal load, a slab that needs cutting, and a three-week wait for a transformer connection are the half nobody quotes. Budget the 130–160% band as a band, not as a point, and hold the top of it until the site survey is done.
The initial spares buffer belongs inside this number rather than after it. Bought with the line, wear parts arrive in the same container at the same freight rate; bought after the first failure, they arrive by air, at a premium, against a stopped line. It is the cheapest item in the acquisition cost and the one most often deferred.
2. Energy
Often the largest single operating cost over a line's life, especially for motor-heavy and thermal processes (injection moulding, milling, pelleting, drying). In high-tariff African grids, the difference between a fixed-pump-hydraulic and a servo-driven machine, or between an efficient and an inefficient motor set, compounds into a very large number over 10 years. Energy-efficient drives that cost more up front frequently win on TCO. See drive technology in injection moulding for a worked example of this trade-off.
Energy is the component where the specification decision and the operating cost are separated by a decade, which is why it is routinely under-weighted. A servo-driven injection machine or a variable-speed drive on a motor-heavy line costs more once and returns the difference every shift; on a South African tariff with a demand charge, the peak the line draws can matter as much as the units it consumes, because the demand component is set by the worst fifteen minutes of the month.
Sequencing is the free half of this. Starting heavy motors together sets a demand peak that is then paid for all month; a staged start-up sequence costs engineering time once and nothing thereafter. Ask for the installed-load list and the start-up sequence at quotation — both are ordinary deliverables, and neither is offered unless requested.
3. Labour
Operators, maintenance, supervision, and the labour intensity the line design imposes. A semi-automatic line with low capex can carry high labour cost; a fully automatic line reverses that. Over 10 years, labour cost (and its escalation) is a major TCO driver — and the reason automation pays where labour is expensive and doesn't where it's cheap.
Labour cost is not only a wage bill; it is a dependency. A line that needs a specific skill to change over or to recover from a stoppage is a line that runs at the speed of one person's availability. Automation buys out that dependency as much as it buys out the hours, and in a market where skilled setters are scarce and mobile, that is often the stronger argument.
Model the escalation, not the current rate. Over a ten-year horizon the difference between a semi-automatic and an automatic line is decided by where wages land in year eight, not year one — and by whether the automatic line's higher skill requirement is available locally or has to be imported with every fault.
4. Materials yield and scrap
The line's efficiency in converting raw material to saleable product. A line that scraps 7% versus 3% is burning material margin every shift for a decade. On high-material-cost products, yield differences swamp capex differences. This is often invisible until measured — see measuring OEE on an old line and the plastics OEE retrofit case study.
Yield is the component that most often decides the ranking, and the one no quotation contains. On a product where material is the dominant cost, moving scrap from 7% to 3% returns more over a decade than the entire difference in capital between two candidate lines — which means a line that cannot be measured cannot be compared on the thing that matters most.
That makes measurement a purchase decision rather than a later improvement. A line specified with output, reject and downtime counters from day one produces the evidence to defend or challenge its own business case; a line without them leaves yield as an opinion for ten years.
5. Spares and consumables
Wear parts, dies, optics, filters, and the spares buffer. A line with cheap, hard-to-source, or fast-wearing parts costs more to keep running. Budget the recurring spares spend, not just the initial buffer. See spare-parts strategy.
The recurring spend, not the first buffer, is the TCO item. A die, an optic or a filter that is consumed on a schedule is a running cost with a currency exposure attached, and a part that is available locally is cheaper than an identical part that is not — not because it costs less, but because it never costs a stopped line.
This is why component selection at specification time is a ten-year decision. Choosing globally branded bearings, seals and contactors over house-brand equivalents adds little to the order and removes a recurring procurement problem. The same logic, by category and cost band, is in our spare parts strategy.
6. Maintenance
Planned maintenance labour and materials, plus the cost of the maintenance capability the line requires. A line on an obscure control platform that nobody local can service carries a hidden premium for a decade.
Maintenance cost follows serviceability, and serviceability is designed in or it is not there. Access for the routine jobs, standard fasteners, documented lubrication points and a drawing pack that matches what was built decide whether a planned service takes four hours or a day. None of that appears on a specification sheet, and all of it is visible in a factory acceptance test if you look for it.
The obscure-platform premium is the clearest example of a decision made once and paid for annually. A control system nobody local can service means every change, every fault and every spare is priced by one party with no competition — and it converts an ordinary maintenance budget into a negotiation.
7. Downtime
The cost of lost production when the line is down — often the most underestimated TCO component. A line that achieves 75% OEE versus 60% produces 25% more from the same shift, the same labour, and the same overhead. Reliability is an economic decision, not just an engineering one. Downtime cost includes both lost margin and the response/repair cost.
Downtime is underestimated because it is priced as lost hours rather than as lost margin. An hour lost at full rate on a line running to a delivery commitment costs the contribution on the product that did not get made, plus the restart scrap, plus the overtime to recover the schedule — and on a seasonal product, an hour lost in peak week cannot be recovered at all.
The corollary is that reliability spending competes with capacity spending, and usually wins. Moving from 60% to 75% OEE releases 25% more output from the same shift, labour and overhead, with no new machine to finance and no additional floor. Buyers who size the next line before measuring the current one routinely finance capacity they already own.
8. Finance
The cost of the capital used to buy the line — interest, opportunity cost, and the cash-flow profile. A China-import 30/50/20 payment structure has a different finance profile from a milestone-paid local build. For owners financing from operating cash, this matters as much as the headline price. See the import vs local framework on cash-flow.
Finance changes the ranking of otherwise similar options, because the payment profile is not the same shape. A China-import structure front-loads cash against milestones that precede any production; a local build spreads it against progress you can see. For an owner funding from operating cash, the question is not only how much but when — and the gap between the last payment and the first saleable output is the part to model.
Working capital belongs in the same calculation and is regularly left out. A milling or feed business ties up more in grain than in machinery; a plastics shop ties it up in resin. A line financed to the last rand with nothing left to buy raw material is a common and entirely avoidable failure.
9. End-of-life and upgrade path
What the line is worth (or costs to dispose of / replace) at the end of the horizon, and whether it can be upgraded rather than replaced. A line that boxes you into a forced full replacement at year 8 has a worse TCO than one with a viable upgrade path. See when to upgrade your PLC.
Upgradability is the difference between a line that is improved and a line that is replaced, and it is decided by things that cost nothing at order time: a mainstream control platform, documented I/O, mechanical access for an extra station, and space left in the panel and on the floor. A line with none of those is a forced full replacement at the moment the market moves.
Residual value follows the same logic. Equipment on a common platform with a complete documentation pack has a second-hand market; equipment on a closed platform, undocumented, is worth its steel. That is a ten-year consequence of a decision made in the specification, which is why it belongs in the TCO rather than in a footnote.
Where the money actually goes — an illustrative split
The split varies enormously by industry, automation level, and energy cost. The table below is illustrative for a moderately automated line over 10 years — directional, not a quote.
| TCO component | Illustrative share (10 yr) | Biggest levers |
|---|---|---|
| Acquisition (landed & commissioned) | 15–35% | Sourcing split, scope discipline |
| Energy | 15–30% | Drive technology, motor efficiency, process design |
| Labour | 15–30% | Automation level vs local labour cost |
| Materials yield / scrap | 10–25% | Line quality, process control, OEE |
| Maintenance & spares | 8–15% | Serviceability, parts availability, PM discipline |
| Downtime (lost margin) | 5–20% | Reliability, OEE, spares strategy |
| Finance | 5–12% | Payment structure, cost of capital |
The point is not the exact percentages — it's that acquisition is one slice among many, and the operating slices are larger and more controllable than buyers assume.
Failure mode: choosing the cheapest quote, then discovering it draws more power, scraps more material, breaks more often, and is harder to service than the option that cost 15% more up front. The capex saving is paid back to you as higher operating cost, with interest, every year for a decade.
How to actually compare options on TCO
- Define the output unit — bottles, blocks, tonnes, parts per year.
- Estimate realistic annual output for each option at its real OEE, not nameplate.
- Build the 10-year cost stack for each option: acquisition + (energy + labour + materials/scrap + spares + maintenance + downtime) × 10 years + finance.
- Divide by total 10-year output to get cost per unit.
- Compare cost per unit, not capex. The lowest cost-per-unit option is the cheapest line.
- Stress-test the assumptions — what if energy tariffs rise, OEE is lower, or the line runs fewer shifts than hoped?
You don't need a perfect model. Even a rough 10-year stack reorders the options dramatically compared to a capex-only comparison, and usually surfaces the real decision: which option has the lowest, most robust cost per unit across plausible futures.
The African-specific TCO factors
- Energy tariffs and stability — high and rising tariffs make energy efficiency a larger TCO lever than in cheap-power markets; instability adds downtime cost.
- Service distance — remote plants carry higher spares-buffer and downtime cost; serviceability is worth paying for. See the maize milling case study.
- Skills availability — a line on a locally-supportable control platform has a lower maintenance TCO than one needing scarce or offshore expertise.
- Currency and finance — import payment structures and exchange exposure affect the finance component.
- Local content / B-BBEE — can affect competitiveness and therefore the revenue side of the business case. See B-BBEE and local content.
TCO and the sourcing decision
TCO thinking is exactly why the hybrid sourcing model so often wins: it buys specialist performance and reliability where they drive yield and uptime (imported), while keeping serviceable, low-energy, locally-maintainable scope close to home. The cheapest-capex "import everything" or "build everything local" options frequently lose on 10-year TCO. See Buy from China or fabricate locally?.
What CISH does in this part of the process
For line projects, our feasibility stage builds a TCO view alongside the capex options — so the decision is made on lifetime cost per unit, not sticker price. We surface the energy, serviceability, and reliability trade-offs explicitly. See Turnkey Production Lines and Line Upgrade & Digitalisation.
Frequently asked questions
What share of total cost is the purchase price?
For a moderately automated line over 10 years, the landed-and-commissioned acquisition cost is often 15–35% of TCO. The rest is energy, labour, materials, spares, maintenance, downtime, and finance — the parts that capex-only comparisons ignore.
Is the cheapest quote ever the right choice?
Only when it also wins on cost per unit over the line's life. Often it doesn't — a cheaper line that draws more energy, scraps more, or breaks more often costs more over a decade. Compare on 10-year cost per unit.
How do I estimate downtime cost?
Lost margin per hour of downtime (output rate × contribution margin) plus the response and repair cost. Even a rough figure shows why a few OEE points are worth real capex — and why reliability is an economic decision.
Does TCO favour automation?
Where labour is expensive and the grid is stable, usually yes. Where labour is cheap or power is unstable, a less automated line can win on TCO. The answer is specific to your cost environment, which is exactly why you model it.
Do I need a detailed model to use TCO?
No. Even a rough 10-year cost stack divided by 10-year output reorders the options versus a capex-only view and surfaces the real decision. Precision helps, but the directional answer usually doesn't depend on it.
Won't buying through CISH cost more than buying direct from China?
On the quote, buying direct is usually cheaper; for the project as a whole, not necessarily. The quote assumes you already have six capabilities — specification, supplier screening, negotiation, shipping and duties, commissioning, maintenance — and can carry the risk of the project failing. What those cost you, in time and in tuition, is the real comparison: Is buying through CISH more expensive?