Comparing Steinert Magnetic Separators? Stop Looking at the Machine Price First

Compare the proof first. The purchase price is the last number I look at when evaluating a magnetic separator. If you are comparing a Steinert system with lower-priced alternatives and your spreadsheet only shows the quote, you are already missing the part where money leaks away. In my last five projects where we accepted the lowest bid, three ended up costing more than the mid-priced bid once installation, commissioning, and reprocessing time were included.

I work as a quality and compliance manager for companies that buy separation systems for scrap, minerals, and waste processing. I review acceptance criteria before any machine reaches production. In 2024, I rejected 11 of 57 first-delivery attempts. Most of those were not bad hardware. They were evidence gaps.

That is why a simple comparison of machine prices makes me uneasy. Price is the easiest number to compare and the least predictive number to trust.

The expensive separator that looked cheap

When I first started specifying magnetic separators, I compared machines the way I compare office paper: same grade, lowest cost, buy it. The wake-up call came in Q1 2023. The winning bid was 18% under the closest alternative. Factory inspection was fine. On site, the support frame did not match our conveyor height, the feed spreader had to be rebuilt, and the sorting threshold kept shifting. Seven weeks later we had it stable enough for trials, but not at the design feed rate.

After extra steelwork, process engineering time, and lost production during the rebuild, the 18% saving was gone. I do not remember the exact total. I remember writing 'TCO from now on' on the review sheet.

What is drift, in practical terms?

Drift is the slow change in a sensor's output or a magnet's performance when conditions change: heat, vibration, material buildup, component age. A sensor sorter is calibrated to make a cut at a certain material grade. Drift moves that cut without changing the number on the screen. The machine still fires accept and reject pulses, but some particles are sorted against the wrong decision.

Magnetic separators drift too. Rare-earth magnetic strength can change with operating temperature, and conveyor speed and belt tracking affect where material lands. If nobody tests the full operating window, drift stays hidden until recovery statistics become worse—and then it is expensive to trace back.

How I calculate total cost of ownership now

Before comparing quotes, I make each supplier price the same acceptance contract:

1. Machine, transport, commissioning: Include the work needed to fix interface problems on site, not just the shipment.
2. Performance test with my material: Not a reference sample that looks similar to ours. The real feed.
3. Performance after 3,000 operating hours: Including recovery, product grade, and drift behavior over temperature.
4. Service, spares, and downtime: Expected lead times for wear parts and the cost of a stopped line.
5. Reprocessing risk: If the separator makes a wrong decision, where does that material go and what does it cost to fix?

The quote comparison starts after that list, not before it.

What I learned from the Steinert practice center

I used to treat machine brands as legal disclaimers: necessary, but not very informative. A bad installation can ruin good equipment. But I have also seen good suppliers use test facilities to remove uncertainty. When we needed to verify a sensor sorting option, we ran a 10-tonne sample through a Steinert practice center before signing. That test told us more than the sales quote did: actual yield, middling behavior, and a baseline for drift sensitivity. It also made the site acceptance test easier because we knew what the machine could do with our feed instead of guessing.

I share this because total cost thinking depends on getting proof before purchase. If a supplier cannot or will not test your material, add a verification risk cost to that quote. It becomes less attractive very quickly.

When I ignore this approach

TCO-first assumes the process is continuous, product specifications matter, and downtime is expensive. If you need a temporary magnetic head pulley for a three-month cleanup job, the cheapest working unit may be the right answer. If you are removing nails from shredded wood and the next step accepts small variation, I would not demand a two-week acceptance test.

This approach worked for us because we run integrated processing lines and quality failures reach a customer. Your mileage will vary. That is fine, as long as the variation is a conscious choice and not an accident in the spreadsheet.