Magnetic Separation? Don't Learn These Lessons the Hard Way (A 3-Step Checklist)

Who This Checklist Is For (And Why You Need It)

So you're looking at magnetic separators. Maybe you've narrowed it down to a system from Steinert, or you're comparing specs between a few vendors. You've got the brochures, you've got the performance claims—and you're about to run a test to make your final call.

That test is where things go sideways. I've been doing this for about 8 years now—handling equipment evaluation orders and process optimization for recycling and mining operations. In my first year (2017), I approved a test protocol that looked fine on paper. It wasn't. We wasted $4,200 on a pilot run that told us absolutely nothing useful. Since then, I've personally made (and documented) 11 significant testing mistakes, totaling roughly $27,000 in wasted budget and lost time. Now I maintain our team's checklist to prevent others from repeating my errors.

This checklist is for anyone about to run a material test on a magnetic separator—whether it's a Steinert, an Eriez, or any other brand. It gives you 3 concrete steps to follow to avoid the traps that look obvious in hindsight, but are shockingly easy to miss when you're in the middle of it.

Step 1: Define Your "Worst" Material (Not Just Your "Standard" Material)

This is the one I messed up spectacularly on my first test. I took a representative sample of what we normally process. It performed beautifully. 99% recovery, clean concentrate. I was thrilled.

Then we got a feed with high clay content because of a rainy season at the mine site. The separator choked. Recovery dropped to below 80%. The vendor claimed it was a "feed issue." They weren't wrong—but they also weren't wrong that their equipment couldn't handle it. The result? A 3-week production delay and $8,000 in reprocessing costs.

The fix: Don't just test your ideal material. Identify your worst-case scenario feed and test that too. Specifically:

• Particle size distribution: Test both the fines-heavy and the coarse-heavy ends of your range.
• Moisture content: If your material can get wet, test it wet. Magnetic separation and moisture often don't mix well.
• Mineralogy changes: If your feedstock varies seasonally or by source, test the most challenging variation. Pro tip: Ask your operations team what "bad day" looks like. That's your test material.

If I remember correctly, our test plan that first time only covered about 30% of the variability we actually see. That's a recipe for expensive surprises. Make sure your test material represents at least 80% of what you'll throw at the machine.

"From the outside, it looks like you just need to run a representative sample through the machine. The reality is that 'representative' in a lab is rarely the same as 'representative' in production. You have to deliberately stress-test the system."

Step 2: Measure the Right Throughput Variable (Hint: It's Not Just Tons Per Hour)

Everyone asks for throughput in tons per hour (TPH). It's a simple number. It's easy to compare. It's also—by itself—largely meaningless for magnetic separation performance.

I once ordered a test on a Steinert induction sorting system for a scrap metal application. The vendor quoted 5 TPH. The test ran at 5 TPH. Results were acceptable. But when we ramped up in production, we could never sustain 5 TPH without sacrificing recovery. Why? Because the feed in the test was perfectly distributed across the belt width. In reality, our conveyors didn't spread material that evenly.

That $200 savings on a cheaper feeder belt turned into a $1,500 problem when we had to retrofit a vibratory feeder. The real metric? It wasn't TPH; it was specific feed rate per unit width (e.g., kg/m/hour). A machine might handle 5 TPH at 1 meter width, but only 3.5 TPH at 0.7 meters of effective separation width.

The fix: When you run your test, measure these overlooked variables:

• Feed distribution width: What percentage of the belt width is actually carrying material? This needs to match your real-world setup.
• Layer thickness: For magnetic drums and eddy current separators, a thicker feed layer reduces separation efficiency. Measure it in mm or number of particle layers.
• Velocity at the separation point: Not just belt speed. The actual material velocity when it hits the magnetic field. These can differ due to slippage or feeding methods.

Let me put it this way: It's tempting to think you can just compare throughput numbers from a test sheet. But identical TPH figures from different feed configurations can result in wildly different separation outcomes. If you don't measure how that material reaches the separation zone, you are flying blind.

Step 3: Check the "Obvious" Mechanical Constraints (The Ones Everyone Forgets)

This one came back to bite me in September 2022. We had a test with a Steinert KSS unit. Separation performance was excellent. Recovery was high. The concentrate was clean. Everyone was happy.

Then someone asked: "How does this physically fit in our building?"

We had a 3.2-meter ceiling height. The KSS unit with its feed hopper and product chutes needed about 3.9 meters. It didn't fit. Not even close. The mistake affected a $3,200 test fee plus the cost of a site visit. We'd spent 6 weeks evaluating a machine we couldn't install.

I know, I know. It sounds stupid. But when you are in the middle of a technical evaluation, it's so easy to get caught up in the separation metrics and forget the basics. And it's not just height.

Before you approve a test, verify these 5 physical constraints:

1. Height clearance: Total height from floor to the top of the feed system. Include airlocks, dust collection, and chutes.
2. Floor loading: What's the static and dynamic weight per square foot? Industrial buildings vary wildly. A Steinert ISS system can weigh over 12 tons. Not every floor can handle that.
3. Access for maintenance: Do you need a mezzanine to reach the magnetic head? Is there room to pull out the rotor? We once had to cut a hole in a wall to extract a drum magnet. Don't ask.
4. Electrical supply: The test unit might run on 480V/3-phase, but your facility might only have 208V available. This seems obvious, but it's a common reason tests get delayed.
5. Dust and ventilation: A separator that generates fugitive dust needs a collection system. If you don't have one, the test results might look good but your compliance inspection won't.

Skipped verifying the height clearance because I assumed "it's basically the same as the old machine." It wasn't. That oversight cost a 2-week production delay and a $1,200 crane rental. A single phone call to ask for total height would have prevented it.

What to Watch Out For

Here are the gotchas that tripped us up, that didn't quite fit into the 3 steps above:

• Beware the "homogeneous" test feed: Lab-prepared samples are often perfectly mixed. Real-world feed is not. If you test with a sample that's been sieved and blended, the results will look better than what you'll get in production. Run a "dump truck" test if you can—literally grab a scoop from an incoming load.
• Test duration matters: A 30-minute test can look great. A 4-hour test reveals how performance changes as material builds up, as magnets heat up, or as screens start blinding. I want to say we saw a 12% performance drop between the first hour and the fourth hour on one test. But don't quote me on that exact number—it was a while ago.
• Don't assume the test center mirrors your control system: At the Steinert Practice Center, everything runs with their engineers. They compensate for feed inconsistencies automatically. Your plant might not have that level of automation. Ask about the control philosophy, not just the results.

"It's tempting to think 'I can just trust the test lab results.' But no test center can replicate your specific feed variability, your specific operator skill, and your specific maintenance schedule. The test gives you potential. Your facility determines reality."

最后一点：Don't Make These Mistakes Yourself

I've been doing this long enough to know that the expensive lessons don't come from the complex technical failures. They come from the simple things we skip because they feel like common sense—until they aren't.

So I'll leave you with this: If you're evaluating a Steinert system, or any magnetic separator, print this checklist. Go through it before you approve the test plan. It will probably save you more than the $27,000 it cost me to learn it.

Roughly speaking, about 60% of the tests we've reviewed at our facility with this checklist caught something that would have been a problem. That's worth taking 20 minutes to verify.

Note: Equipment specifications and pricing change. This checklist reflects my personal experience through early 2025. Always verify current dimensions, power requirements, and test protocols with the vendor.