How to calculate the total cost of owning a laser over five years ART-165

Define the TCO object

The most common mistake is to call the object a “laser” while counting only the main machine in one quotation and the entire cell in another. The boundary must be identical. For a practical decision, include everything required to execute the agreed production program:

• machine, source, head, control system and tables;
• chiller, compressor, air drying and treatment;
• extraction, filtration, ductwork and make-up air;
• gas infrastructure, manifold, tanks or generation;
• CAM software, licenses and workstations;
• internal sheet logistics, lifting equipment and storage areas;
• installation, commissioning, training and acceptance tests;
• spare parts, optics, nozzles and an initial consumables stock;
• electrical, building, network and safety changes.

If an automated storage system is purchased in year three, include it in the scenario with the cash-flow date. If it merely “can be added” but no decision has been made, create a separate scenario instead of hiding uncertainty in an average number.

A five-year horizon is 61 points in time, not one total

A useful model has at least monthly or quarterly periods for utilization and cash flow. The initial year includes the advance payment, delivery, site preparation, installation and ramp-up. Subsequent years differ in volume, tariffs, service events and productivity. Year five ends not at zero, but with an assessment of residual value, dismantling or continued use.

NIST defines life-cycle cost analysis as a technique for assessing the total cost of an investment over its life cycle. For a laser cell, this means every cost has an amount, date, driver, data source and uncertainty range.

Recommended register columns:

| Field | Example content | Why it is needed | |---|---|---| | cost_id | ENERGY-CHILLER-01 | keep track of an item across versions | | date/period | month 18 | discount correctly and plan cash | | category | energy, labor, service | compare the cost structure | | driver | cutting hours, sheets, shifts | build scenarios | | quantity | measured or forecast | separate volume from price | | unit_cost | dated tariff | update independently | | confidence | high/medium/low | identify weak points | | source | manual, quote, meter, ERP | verify provenance | | owner | finance, process engineer, service | assign responsibility |

Initial investment

CAPEX is not limited to the supplier's invoice. Include equipment and options, international and domestic logistics, transit insurance, customs and brokerage costs, unloading, rigging, foundations, electrical supply, ventilation, gases, networking, fire-safety measures, guarding, lighting, personnel preparation, initial process qualification and test material.

Separate assets that remain with the company regardless of the specific machine, such as an upgraded incoming power supply or building reconstruction. These costs are still a real cash requirement for the decision on one machine, but some may become shared infrastructure for the next investment. State allocation rules transparently.

ART-166 examines omissions from the initial budget in detail. They should already be consolidated in TCO.

Variable costs for the production program

Link variable costs to the operating mode, not simply to powered-on hours. Different materials and thicknesses change cycle time, gas, electricity, nozzles, optics, waste, changeover frequency and good-part yield.

Key drivers:

• machine-on, beam-on and productive hours;
• material type, thickness and mass;
• cut length, number of pierces and contour complexity;
• assist-gas type and actual consumption;
• number of loading operations, changeovers and trial parts;
• nesting yield, usable remnants and scrap price;
• urgent jobs, recutting and rejects;
• actual replacement frequency of consumable components.

Do not apply an advertised “consumption per hour” figure to the entire order portfolio. Build a representative basket of parts and validate it with CAM data, trial cuts and measurements after commissioning. ART-150 shows how to check the margins on real orders, and ART-139 explains how to value waste in monetary terms.

Energy and auxiliary systems

The electrical boundary includes the machine, source, chiller, compressor, dryer, extraction, filter, gas generator, automation, IT and HVAC. Rated installed power is needed for engineering design, but annual cost is determined by the profile: load during cutting, waiting, startup, cleaning cycles and standby.

The model should contain:

`Energy cost = Σ (measured kWh of subsystem in mode × hours in mode × tariff for period)`

This is an editorial formula. Energy and tariff are stored separately to test sensitivity. For peak demand, reactive energy or different tariff bands, add the components actually present in the company's contract. ART-152 covers the engineering power balance; TCO uses its results.

Labor: more than the operator at the console

Include programming, nesting, locating and delivering sheet, loading, unloading and sorting, quality control, marking, finishing, cleaning, consumables replacement, scheduled maintenance, dispatching, data management and training in the labor map. Separate productive labor from support labor.

The hourly rate should be a fully loaded management rate agreed with finance, not just take-home pay. At the same time, do not double-count overhead: if it is already included in the rate, do not add it again as a separate percentage.

Automation can reduce touch time while adding monitoring, maintenance and planning. Therefore, compare roles and hours in each scenario rather than assuming that an “unmanned shift” means zero labor.

Service, consumable parts and degradation

Break service down into scheduled contracts, inspections, filters, fluids, optics, nozzles, ceramics, emergency visits, parts delivery, diagnostics and major repairs. Each item needs an interval or probability, price, lead time and consequence of unavailability.

Do not treat a component's service life as guaranteed without conditions. Use three sources: the manufacturer's recommendation, experience with a similar configuration and your own log. Introduce low/base/high scenarios. If a component is scarce and takes a long time to arrive, its downtime impact may exceed its purchase cost.

Account separately for technological obsolescence: discontinued software support, the need to replace an industrial PC, a license or a network component. This does not mean replacement will necessarily happen; the model should contain a risk allowance or scenario.

Downtime cost

Downtime cannot automatically be valued as “revenue per hour.” Some orders are rescheduled, some are processed on another machine, and some genuinely lose margin or miss a deadline. Create a downtime event model:

`Impact = lost contribution margin + overtime + outsourcing premium + logistics + penalties actually applicable + restart scrap`

For each failure class, specify a frequency range, mean time to restore, spare availability, alternative routing capability and deadline sensitivity. Do not insert penalties that are absent from contracts. Model scheduled service, emergency repair and production stoppages caused by a lack of gas, material or an operator separately.

Working capital and ramp-up

A new cell needs stocks of sheet, gas, nozzles, optics, filters, spare parts and work in progress. Cash may be paid before the customer pays for the finished products. The model therefore needs a working-capital peak, not just P&L expenses.

The first few months rarely match steady state. Add training hours, trial sheets, lower productivity, additional inspection and gradual order-portfolio growth. Define ramp-up as a curve with checkpoints: SAT completed, representative parts accepted, shifts trained, CAM accuracy confirmed and planned yield achieved.

Financing, taxes and accounting

For a cash-flow comparison, add the advance payment, payment schedule, interest, fees, insurance, foreign-currency payments, collateral requirements and final buyout. IFRS 16 requires lessees to recognize a right-of-use asset and a liability for most leases; this matters for reporting but does not determine whether leasing is economically preferable.

Depreciation is not a direct cash outflow, but it affects accounting and may affect taxes. Do not mix accounting profit with TCO cash flow. The file should have separate layers: operational cash, financing cash, accounting view and tax view, reconciled by finance.

Residual value and the end of the horizon

At the end of year five, the machine may continue operating, be sold, upgraded or dismantled. Residual value must not be an optimistic “asking price from an advertisement.” A range, technical condition, hours, service history, support availability, dismantling costs, logistics and site reinstatement costs are needed.

In discounted TCO, salvage value is deducted when receipt is expected, and disposal cost is added. If no reliable estimate exists, show several scenarios, including zero residual value.

Nominal or real prices

There are two valid approaches. In a nominal model, each future tariff includes expected inflation, and the discount rate is consistent with nominal flows. In a real model, prices remain at constant purchasing power and the rate is real. Mixing nominal costs with a real discount rate distorts the result.

Model foreign-currency costs separately: reporting base currency, exchange rate for each payment, hedging or no hedging. Do not hide exchange-rate risk in a general “contingency” allowance.

Basic formula

Present value can be used to compare scenarios:

`TCO₅ = Initial cash + Σ(Cash flow_t / (1+r)^t) − Residual value₅ / (1+r)^5`

where `r` is the agreed rate per period, and `Cash flow_t` includes operating, service, downtime, working-capital and financing components according to the selected boundary. If the model is monthly, the rate and exponent `t` must be expressed on a monthly basis.

Do not compare only the lowest TCO if scenarios produce different numbers of good parts. Add `TCO per good part`, `TCO per productive hour` or present value for an agreed production program. The metric must account for quality and capacity, not reward a cheap but inadequate system.

Scenarios and sensitivity

The base case is not “the most likely number,” but an agreed set of assumptions. Build at least:

• low-load: slow growth of the order portfolio;
• base-load: planned product mix and a two-shift schedule;
• high-load: a higher beam-on share and faster consumption of resources;
• disruption: a prolonged failure of a critical component;
• energy/gas shock: a tariff change;
• automation stage: an additional module in a specified year.

Then change one driver at a time and build a tornado chart. Decisions are usually affected most not by small nozzles but by utilization, yield, labor, gas, downtime, financing and residual value. However, this must be demonstrated with your data.

Model quality gate

Before a decision, an independent reviewer checks:

1. the same system boundary for all alternatives; 2. the same production program and quality criterion; 3. no double-counting; 4. separation of quantity from unit price; 5. dates and sources for every significant input; 6. consistent nominal/real assumptions; 7. financing and accounting in different layers; 8. downtime without overstated lost revenue; 9. ramp-up and working capital; 10. transparent residual/disposal logic; 11. sensitivity to key drivers; 12. a change log and version owner.

How to turn an estimate into plan-versus-actual control

Pre-purchase TCO inevitably contains assumptions. It should not be archived after the contract is signed. Create a baseline as of the decision date, then replace assumptions with actuals at checkpoints: after contracting, delivery, SAT, the first month, quarter and year.

Keep three values for each driver: decision estimate, latest forecast and actual. Explain variances by their cause rather than rewriting the past. For example, energy cost can differ because of the tariff, a different number of hours or higher kWh per productive hour. These call for three different management actions.

A useful variance bridge:

`ΔCost = volume effect + rate effect + efficiency effect + mix effect + unplanned event`

This breakdown helps avoid blaming equipment for a weak order portfolio or, conversely, hiding low process efficiency behind volume growth. Adapt the formula to the data structure; it is not an accounting standard.

Update sensitivity after each review. A driver that seemed secondary may become critical after a material-mix or gas-supply change. At the same time, do not change the decision baseline: it is needed to assess the quality of the original business case.

How to avoid false precision

Do not enter many decimal places when an input has a wide range. For uncertain quantities, show low/base/high, the source date and the condition under which the value will become known. It is useful to distinguish:

• committed — contracts and costs already incurred;
• measured — actual meters and ERP;
• quoted — valid quotations with a date;
• engineered — calculations with input data;
• benchmark — an analogue with no guarantee of transferability;
• assumption — a management hypothesis.

The decision maker must see which part of TCO is confirmed and which depends on future utilization. This matters more than one attractive total.

Reconciling TCO with job costing

Annual TCO and individual job costing should share drivers but should not be one file. Check that total machine hours, material, gas, energy, labor and service allocation from the portfolio reconcile with the annual scenario. If job estimates show 4,000 productive hours but TCO shows only 2,500, one model uses a different boundary.

Do not allocate all fixed costs over an optimistic volume: unit cost will rise at lower utilization. Show the contribution view for short-term decisions separately from the full-cost view for long-term payback. ART-150 covers actual margins; TCO provides the control total for the entire asset.

Quarterly reconciliation should explain the difference between total cell costs and the amount allocated to jobs. Unallocated time, training, repairs and idle time must remain visible rather than disappear into “other costs.”

Typical mistakes

• comparing sticker price instead of the full system;
• using installed power as constant consumption;
• counting all hours as beam-on;
• excluding auxiliary systems and HVAC;
• adopting gas consumption without representative parts;
• treating training as a one-off formal day;
• multiplying all downtime by revenue;
• double-counting overhead in the labor rate;
• ignoring working capital and ramp-up;
• mixing real and nominal cash flows;
• accepting optimistic resale value without evidence;
• hiding weak inputs behind one contingency percentage.

Conclusion

A five-year TCO model is useful when it represents a specific production program and shows where every number comes from. Its strength lies not in a complex formula, but in consistent alternative boundaries, time structure, plan/actual updates and scenarios. The machine price remains important, but infrastructure, utilization, gas, labor, downtime, financing and residual value often change the decision.

Safe boundaries

This material does not specify a discount rate, depreciation period, tax regime, exchange rate, component life, gas consumption, energy price or residual value. All values require a dated source and approval by the responsible function.

Need service advice?