Consumables Waste on Site: What the Numbers Usually Show

Every so often someone asks about consumables waste and brings a number with them. A percentage they heard, a target someone set. The number is usually the least useful part of the question.

Here's why we won't give you one: waste is whatever the site decided to count, and most sites count it in a way that produces a comfortable figure. Borrow a benchmark from a shop welding flat fillets at a bench and try applying it to pipe work in position — you've now got a target nobody can hit and a monthly argument about why.

So instead: four things worth counting, how to count them, and what each one is telling you to change.

1. Issued, returned, and the gap — by grade

No instruments needed. What left the store, minus what came back unopened, against the deposited weld metal recorded for the same period. Do it per grade and per diameter — the total hides the one line that's out of step.

Its weakness: it treats everything not deposited as waste. Wrong. Some is legitimate handling loss, some is rework, some is still sitting in a part-used carton on a bench. That's why the next three measurements exist — they split the gap into causes instead of arguing about the size of it.

2. Stub length — collect them, don't eyeball them

Every stick electrode leaves a stub, and how long that stub is isn't fixed by physics. It's set by the holder, the current, the position and mostly by habit. So collect stubs from a defined period and measure them. Estimating is how you end up with a number that matches your opinion.

Two things to watch when you read the result. Shorter isn't automatically better — chasing the last few millimetres cooks the holder, damages the coating at the grip, and buys arc instability with the savings. And the figure only means something against your own site and process, because a rutile electrode run cool and a cellulosic one run hot leave very different habits.

3. What's left on the spool

Same idea for wire. Weigh a sample of spent spools at changeover, or just write down why each one came off. Easy to do, almost never done.

Residue has two causes and they need opposite responses. Spool changed early because the wire at the end has rusted or kinked — that's storage and packaging, and the fix is a dry store and intact barrier bags, not a welding tweak. Spool changed early because the shift ended or the job changed — that isn't waste in any material sense, and counting it as waste is how a site spends a quarter chasing a number that was never going to move.

4. Rework consumables, booked separately

Rework eats consumables twice: once for the weld that got rejected, once for the repair. On most sites that cost is real and invisible, because the material went out to the job and nobody charges it back.

This is a booking change, not a measurement project. Give repairs their own cost code and you get a quantity per repair for free. The number that actually matters then isn't the material — it's the rate at which you're doing repairs. A shop with a low reject rate and a generous allowance is in a completely different place from the reverse, and in total consumption they look identical.

5. So what do you compare against?

Yourself, last month. Measure those four things for one representative month, and write down the process, the position mix and the grade mix next to them. Then compare every month after that to your own baseline.

Movement against your own recorded figure is a real signal. Movement against somebody else's percentage is usually just a change in how the number gets calculated.

6. Each number points somewhere different

That's the whole reason for keeping them apart. Big issued-to-returned gap with normal stubs and normal residue? Look at booking and store control. Long stubs? Holder practice and current settings. Residue caused by surface condition? Packaging and storage. Rework eating visible quantities? That's procedure and operator qualification, not the consumable.

No new equipment required. One month of consistent recording, and the discipline not to average the four together afterwards.

If you're setting this up and need the pack sizes, unit markings and batch references your store has to record in order to reconcile issued against deposited, drop us a line — info@steeltoolschina.com.