How to actually read an ISRI scrap code (without the cheat sheet)

The codes are a language, not a list

If you have been around a scrap yard for more than a month you have seen the colour-coded ISRI cheat sheet on the wall. Bare Bright. Birch. Honey. Mill. Talk. Zorba. Zurik. They look like a list of words. They are actually a language - a precise vocabulary that lets a mill in Houston and a yard in Lagos agree on what is in a container without anyone seeing the load.

That precision is where the money lives.

What a code actually encodes

Every ISRI spec tells you four things in one short string:

• The base metal (copper, aluminium, steel, etc.)
• The purity / form (bare, insulated, painted, mixed)
• The size / preparation (shred, briquette, baled, loose)
• The dirt budget - how much non-target material is tolerated

The "spec" your mill sells against is shorthand for all four. #1 Copper Bare Bright is not the same as #2 Copper Birch because Birch carries paint, solder and other contaminants - and the price gap reflects exactly that.

The non-ferrous codes you really need cold

You do not need to memorise the whole book. You need these by heart:

• Bare Bright (110) - bright, uncoated, clean #1 copper wire. Highest-paid copper.
• Berry (111) - clean uncoated copper, slightly tarnished. One step below Bare Bright.
• Birch / Cliff (122) - #2 copper with paint, solder, or insulation residue.
• Talk (130) - light copper. Sheet, gutter, downspout. Higher dirt budget.
• Mill Berry-style aluminium - clean wrought aluminium scrap.
• Tense (240) / Tale (245) - old cast and aluminium auto wheels.
• Zorba - non-ferrous fraction from automotive shredders. Mixed metals; sold to separators.
• Zurik - high-grade stainless and copper fraction recovered after Zorba processing.

Why mills care about the dirt budget

A mill is not buying metal. A mill is buying a predictable melt yield. When a 40,000-lb container of "#1 Copper Bare Bright" arrives, the mill operator wants 39,920 lb of copper to come out the other side. If the actual yield drops to 95%, the spec was wrong and the price was wrong. They will downgrade the load or claim back.

This is why your grading discipline at the scale is the single largest lever on margin. Every load you ship that meets spec increases your trust score with that mill; every claim erodes it.

How to grade fast without grading wrong

The pros do this in three passes:

• Wallet pass. Walk the load. Eyes only. Get the dominant code in your head before you touch anything.
• Hands pass. Pick up two or three pieces from different parts of the load. Are they consistent with the dominant code? If yes, move on. If not, downgrade.
• Magnet test where appropriate - especially for shred and Zorba-type fractions where ferrous contamination flips a spec.

The trap is going straight to "let me grade this individual piece" - you can grade a thousand individual pieces correctly and still call the load wrong if you missed that the dominant code is a step down.

Where software helps (and where it cannot)

Scale-ticket software can lock the grade dropdown to your current price list, force a photo of the load at intake, capture an operator note, and pin a grading rule to each material code so the same supplier gets the same call every time. That removes "did Mike grade this differently than Sarah?" variance.

What software cannot do is replace standing at the load and looking. Grade is judgement. The system makes the judgement consistent, fast, and audit-traceable - but somebody still has to look.

Where to read more

ISRI publishes the Scrap Specifications Circular - keep a copy in the scale house. The codes do not change often, but the dirt budget for things like Zorba moves as auto shredders get better separators. Read the circular every January and you will be ahead of 90% of operators.

The yard that grades sloppy buys high and sells low. Eventually it sells out.