Run the numbers — how much can smart welding save you each year?

① Price → per-Nm³: cylinder = price-per-cylinder ÷ volume (L × MPa ÷ 100, e.g. 40L×15MPa ≈ 6 Nm³, ideal-gas approximation, excl. temperature & compressibility); bulk = price-per-ton ÷ 1000 × gas density (Ar 1.78 / mix 1.82 / CO₂ 1.98 kg/Nm³). Bulk mixed-gas €/t = Ar €/t × Ar-share + CO₂ €/t × CO₂-share (default 80:20 → €1300/t, see Advanced). Nm³ = standard conditions (0°C, 101.325 kPa).

② Effective (ideal) gas = units × empirical flow × 60 × arc-h/day × workdays ÷ 1000; real gas = effective ÷ (1 − waste rate) (overshoot/spillover/drift/leakage consume gas on top of effective use); × unit price = annual gas spend.

③ Annual saving = gas spend × waste rate × recovery rate (default 30% × 90% ≈ 27% saved, conservative); groups summed, no double-counting.

④ Done by hardware (precision sensing + closed-loop control) — no AI, no change to operators.

📌 ≥10% saving or you don't pay; CO₂/mixed gas usually 20%+ (expected, not promised).

① Value = stations × annual labor cost × utilization uplift % (default 10%, up to 30%); annual cost = hourly × hours/shift × workdays (latter two in Advanced).

② Platform tasking + arc-on utilization tracking + PDCA loop lets the same headcount weld more (+10% capacity) — worth ~10% of the labor capacity.

③ ⭐ Equipment-level arc data only, no worker-level info (GDPR by-design) — raising OEE without monitoring workers.

📌 Strictly separate from ③ Downtime: this = more output in normal run; downtime = fewer fault stoppages. Never double-counted.

① Value = stations × annual hours (hours/shift × workdays, EU 2000h / CN 2400h) × downtime % (default 5%) × downtime cut % (default 30%) × hourly rate; when a machine stops the welder is paid but idle — valued conservatively at labor cost (hourly wages are the real loss in EU/US).

② Multi-sensor monitoring (current/voltage…) + anomaly time-slicing + alerts & remote diagnostics slash the time spent on the real bottleneck — finding the fault.

📌 Strictly separate from ② Productivity & ⑤ Rework: this = fewer fault/long-stop events; ② = more output in normal run; ⑤ = re-welding. Never double-counted.

① Value = stations × QA-compile h/yr·station (default 50h ≈ 10min/workday·station) × paperwork cut % (default 90%) × doc-staff hourly rate (default €15/h ≈ 60% of welder — a QA clerk is not a scarce role, kept conservatively low).

② It's not the welder's paper that's saved (filling records is mandatory in certified welding, unchanged) but the downstream QA "collect → re-key into spreadsheets → cross-check WPS → assemble dossier" hours — born-digital logging + auto compliance + one-click export/lifelong traceability remove ~90% of that re-keying.

📌 Only paperwork labor is priced; the legal value of traceability is a qualitative benefit, conservatively not monetized. Independent of other dimensions.

① Value = annual arc hours × weld rework rate (default 3%) × extra-time × (default 1.5×) × rework-rate cut % (default 20%) × welder hourly rate. Arc hours = Σ groups(units × arc-h/day × workdays), reusing the gas groups above.

② A bad weld = original 1 + grind-out 0.5 + reweld 1 = 2.5× total, i.e. 1.5× more time than a good one; the original is sunk, so only the extra 1.5× (grind+reweld) is counted.

③ Real-time compliance checks + tri-color alerts catch defects as they happen, cutting rework rate ~20% (conservative). Labor only — material, scrap, inspection & schedule losses are excluded, the most conservative basis.

📌 Separate from ② Productivity: ② = more new output via utilization; this = fewer welder hours re-welding. Never double-counted.