Every tonne of steel a plant ships has already paid for the tonnes that never made it. Scale burns off in the reheat furnace, crop ends are cut and returned as scrap, trimmed edges and cobbles leave the line, and rejected coils are downgraded or remelted. Each loss is small on its own, yet together they raise the cost of every good tonne. Modeling that effect turns yield from a percentage on a report into a cost per tonne with an owner and a plan. Plants can see how iFactory AI converts each yield loss into cost on live production data before setting next year's yield targets.
Yield Loss Cost Modeling for Steel Plant Production
Convert scale, crop, scrap and rejection losses into cost per tonne, then build a roadmap that shows where to recover it first.
Follow 100 Tonnes From Melt to Dispatch
Losses stack up along the route. The illustrative example below starts with 100 tonnes of crude steel and shows what is left after each stage.
Eight tonnes out of every hundred never reach the customer in this example. Each tonne lost late in the route carries more cost than one lost early, because more processing has already been paid for.
The Four Losses to Price
Each loss has its own cause, location and lever. Pricing them separately stops the team from averaging away the biggest one.
Metal oxidises and flakes off during heating. Longer soak times and hot furnaces raise it.
Head and tail ends are cut off to remove defects, and the cuts return as scrap.
Edge trim, rolling cobbles and size mismatches leave the line as scrap.
Surface, shape or property defects lead to downgrades, rework or remelting.
See Your Yield Losses Priced by Type
Book a 30-minute session and iFactory AI will show how your scale, crop, scrap and rejection losses would convert into cost per tonne and rank by value.
Turning a Yield Point Into Cost
The maths is simple. Cost per shipped tonne is the cost per input tonne divided by yield. The table below assumes 100 units of cost per input tonne and ignores scrap credit for clarity.
| Yield | Cost Index per Shipped Tonne | Yield Loss Adds |
|---|---|---|
| 90% | 111.1 | 11.1 |
| 92% | 108.7 | 8.7 |
| 94% | 106.4 | 6.4 |
| 96% | 104.2 | 4.2 |
Moving from 92 to 94 percent cuts about 2.3 units per shipped tonne in this example. Scale that across annual output and the value of a yield point becomes clear. Test the same arithmetic on your own volumes in a short guided session with the iFactory AI team.
Scrap Is Not a Total Loss
Most lost metal is recycled, so an honest model separates recoverable value from true loss.
A Reduction Roadmap in Three Phases
Recovery works best when effort matches payback. Start with actions that need attention, not capital.
Where iFactory AI Fits
iFactory AI reads production, quality and cost data together, so each tonne lost has a location, a cause and a price.
Loss-by-loss cost view
Scale, crop, scrap and rejection are shown separately, each converted into cost per shipped tonne.
Stage-aware valuation
Late-stage losses carry the processing already spent, so the ranking reflects real cost, not just weight.
Roadmap prioritisation
Opportunities are ranked by value and effort, helping leaders pick which phase to fund first.
Ask in plain language
Managers can ask why yield fell this week and receive the loss type and unit that drove it.
iFactory AI arrives pre-configured on an NVIDIA server that ships racked and ready with software pre-loaded. Scope covers cabling, network, ERP and MES integration, team training and 24x7 remote monitoring.
Frequently Asked Questions
How do you convert yield loss into cost per tonne?
Divide the cost per input tonne by the yield, then subtract the input cost to see what the loss adds to each shipped tonne. Deduct scrap recovery value for a fair net figure. Doing this by loss type shows which one costs most. You can watch the calculation run on your own data in a live session.
Which yield loss is usually the biggest?
It depends on the product and route. Crop loss often leads in long and flat rolled products, while scale depends heavily on furnace practice and rejection on quality control. The only reliable answer comes from pricing each loss on your own line. To find yours, request a loss ranking walkthrough with the iFactory AI team, or ask support how the data is mapped.
Does scrap recycling cancel out yield loss?
No. Recycled scrap recovers raw material value, but the energy, labour and capacity spent on those tonnes are gone, and scale leaves the loop entirely. A good model credits scrap at a fair value and still shows the remaining cost. A short product tour of the scrap credit logic shows how this is handled.
What data do we need to start?
You need input and output tonnes by stage, scrap and rejection records with reasons, and cost per tonne at each stage. Most plants already hold this in production and quality systems but rarely join it. See which of your systems can be connected first so results arrive within weeks.
How do we decide which reduction projects come first?
Rank opportunities by annual value and the effort needed, then sequence them from quick wins to capital projects. Priced losses make the business case direct, since each project has a value attached before funding is asked for. Schedule a walkthrough of roadmap prioritisation to see how the ranking is built.
Recover the Tonnes You Are Already Paying For
iFactory AI converts scale, crop, scrap and rejection losses into cost per tonne and ranks the reduction opportunities. Book a walkthrough to see it on your own plant.
