Prevention Beats Cure: A Quality Inspector’s View of Steinert Separation Lines
Prevention isn't just cheaper than cure. In separation systems, it's the only strategy that fits into your production schedule.
I've spent the last four years at the point where theory meets failure: reviewing magnetic separators and sensor sorters before they're commissioned. My job is to say 'yes' or 'no' to a piece of equipment before it starts feeding a plant. In 2024, I rejected 11% of first-time commissioning checklists because of missing wiring diagrams, wrong splitter settings, and magnet gaps that were 2 mm off the drawing. None of those problems would have been expensive if they'd been caught early. All of them were expensive by the time they reached the site. And if you ask me where the value leaks in most projects, it's the 5-minute check nobody wanted to do.
How many systems have I reviewed? Around 200, maybe 180—I'd have to check the database. But the pattern is the same across all of them: when someone skips a 5-minute check, they usually spend 5 days trying to fix the result.
That's why I keep coming back to a simple rule: prevention beats cure.
The First Check Is the One That Pays
Most people think quality control happens at the end. It doesn't. It happens when someone writes the specification. I've seen a $180,000 order held up because one drawing said 'magnet gap: ±2 mm' and another said '±5 mm.' The vendor said both were within industry standard. Maybe they were. But our end customer was buying a Steinert KSS to produce a metal fraction clean enough to sell at a premium. 'Industry standard' wasn't the target. The contract was. We rejected the batch, the vendor reworked the assembly, and now every contract includes the exact tolerance. No amount of downstream sorting can fix a tolerance that was accepted too early.
The same logic applies to what some operators call the “eddie outlet”—the discharge chute on an eddy-current separator. On paper it's just a chute. In practice it's the border between sellable non-ferrous and contaminated waste. If the eddy current outlet is set too high, or the splitter knife is one notch off, you don't get a slightly worse product. You get a product that fails the customer's purity test. In my first commissioning job, I signed off on a separator without checking that splitter knife. It cost us a $14,000 redo and a very red face. Now it's the first item on the checklist. The only way to fix that after commissioning is to shut down, adjust, and run the test again. That's not a fix. That's a redo.
It's Not About Magnet Strength. It's About Feed Quality.
Here's where I get contrarian. It's tempting to think a stronger magnet or a smarter sensor solves everything. But when someone asks me “Steinert vs Hercules?”—and I do get that question—they usually expect me to talk about magnet strength, rotor design, or sensor resolution. I don't. I ask to see the feed analysis. Because every separator, no matter who makes it, is at the mercy of what you put into it.
I once spent an afternoon at a recycling site in Groves, Texas, watching a brand new separator reject clean aluminum along with the scrap. The operators were ready to send the machine back. The real problem was the feed: it had been sitting outside after a wet spell, and the fines were sticking to the belt. A 10-minute inspection showed that the screen deck ahead of the machine was running at about 80% capacity. The separator wasn't failing. It was being fed someone else's problem.
That's why prevention over cure has to start upstream. You can't inspect your way out of a feed problem. But you can test the feed, train the operators, and verify the setup so the separator can do what it was designed to do.
Documentation Is Part of the Machine
The least glamorous part of my job is paperwork. It's also the part that saves the most money. I've rejected more deliveries for missing wiring diagrams and maintenance schedules than for mechanical damage. People think that's bureaucratic. I think it's realistic. When a separator stops at 2 a.m., the technician on site doesn't need a stronger magnet. They need a diagram that shows where the limit switch should be and a manual that explains how to set the eddie outlet. Documentation is part of the machine. If it's missing, the machine isn't complete.
What About the 'Good Enough' Argument?
To be fair, there's a limit to checking. No amount of verification can fix a system that is fundamentally wrong for the material. If you're trying to separate stainless steel from aluminum, a drum magnet won't do it, no matter how carefully you inspect it. The check has to happen before the purchase order, not after delivery. That's why we run tests on customer samples before we quote. I'd rather tell a customer 'this won't work for your stream' in week one than have them discover it in week twenty.
I also get why people resist the upfront time. It feels like paying for insurance when nothing is wrong. But the numbers I've seen are lopsided. In our Q1 2024 audit, every hour spent verifying specifications before ordering saved roughly three hours of corrective work after installation. That's not a law of physics. It's just a pattern across 180-something projects. It's enough to make me feed the checklist before the order form.
At least, that's been my experience with metal recyclers and mineral processors. If you're sorting construction waste or glass, some details will differ—but the habit of checking won't.
My Bottom Line
So here's my stubborn position: I'd rather spend five days checking a Steinert line and start with confidence than spend five weeks correcting a line that was 'good enough' to buy, but not good enough to run. The equipment matters. The steel, the sensors, the rotor, the magnet—all of it matters. But what separates a good investment from a bad one is often not the machine. It's the 5-minute check that happened before the machine was ever loaded on a truck.
Prevention beats cure because by the time you need a cure, the production schedule has already moved on. And that's exactly where I come in.