Last March, a dewatering screen brought my week to a stop
It was a Tuesday. The plant called at 9:15 and said the tailings dewatering screen on Line 2 had dropped to about 40% of rated throughput. By 10:30, maintenance had confirmed the real problem: the feed distribution was so uneven that the screen deck was wearing on one side. We shut the line down. That is when my job turned from buying office supplies to finding a piece of separation equipment fast.
I am the office administrator for a 140-person mining services company. I manage all supply ordering—roughly $1.8M annually across 60 vendors. I report to both operations and finance, so I knew the next 72 hours would determine whether I had a good quarter or a very uncomfortable one. (Note to self: never let a budget cycle be decided by a screen failure.)
The first quote came from a general equipment supplier. In their words, we can handle the whole feed system. That sounded great. Then a second engineer, Lewis Alexander from the Andritz Soutec separation group, did a site walk. Unlike the first salesperson, he started by telling me what he would not sell us.
What is the drift theory? The phrase that changed how I listen to vendors
Your pump is the weak link, Lewis said. If you replace the screen and keep that pump, the whole system will drift into failure. Not tomorrow—maybe in six months.
I had heard drift before in a safety context, but I was not sure. That evening I searched 'what is the drift theory.' What I found was more useful than any equipment brochure.
According to Sidney Dekker's Drift into Failure (CRC Press, 2011), accidents seldom happen because one person makes one terrible decision. They happen because small variations from normal practice become accepted as normal. Scott Snook, in his analysis of the 1994 Black Hawk shootdown, called the same process 'practical drift' (Friendly Fire, Princeton University Press, 2000). Over time, what people actually do keeps drifting away from what the procedure says.
Now, I am not a safety expert, but I know that description nails purchasing. The drift theory in procurement, as I see it, is this: every requirement that is not written down and checked will drift toward the easiest version of itself. When you say 'as soon as possible,' the supplier hears 'whenever convenient.' When you say 'standard size,' you mean the one you have used for ten years; they might mean their standard, not yours.
My rookie mistake with the first vendor
Honestly, I knew better. In my first year in this role, I made the classic specification error—assuming standard meant the same thing to every vendor. That was a $600 redraw for a custom signage order. Here, the stakes were higher.
The general supplier offered a universal dewatering screen and a new pump. Their price was about 12% under Andritz Soutec's proposal. Finance liked that number. I did too.
It took me a month to see the problem. The universal screen actually fit, but only after we bought a mounting adapter and rerouted a slurry line. The pump that came with it was a direct equivalent from a different manufacturer. I did not check the impeller clearance or the seal specification. Like most beginners, I approved the package based on a datasheet. When the pump cavitated on day three, the supplier said it should be okay if we adjusted the float. Five weeks later, it ate the seal and dumped gland water into the pump bay. Cost to repair: $18,400, including two days of lost production.
Here is the reverse validation: Lewis Alexander had told me the pump was outside Andritz's separation scope. They don't build it, and they won't pretend to quote it. But he could give me two names of people who do. I thought that was just a sales trick. I ignored his boundary and bought the easy one. I only believed him after ignoring him—and paying $18,400 to learn what a specialized pump vendor would have charged me to prevent.
Why the Andritz dividend yield mattered to me
Before I signed the second, smaller contract with Andritz Soutec, I did the boring financial check I now do on every supplier over $50,000. For a public company, that includes a look at the stock and dividend. The Andritz dividend yield is not a recommendation to buy the stock; it is a health check. A company that has kept paying a dividend through good and bad years is more likely to stock spare parts and honor warranties a decade from now.
I remember the day I looked: I typed 'andritz dividend yield' into a search and landed on Andritz's investor relations page. The yield was not the headline number—it was the stability around it. Their IR site lists the dividend proposal and the share price, which is all you need to calculate the yield yourself. As of early January 2025, the exact yield depends on the day you look. Prices change. Verify current rates.
But this is the part I really should say: I have bought from private companies that made the same promise and disappeared. The dividend check does not guarantee quality. It just told me the company was not going to vanish before the warranty ended.
What actually happened
In the end, we bought the Andritz Soutec replacement screen. It arrived in August 2024, about a week early, and the installation was routine. The new screen deck held a consistent layer of solids, and the line came back to 95% of original throughput within two days.
The pump, following Lewis's advice, was repaired by a specialized slurry pump contractor. They identified the wrong seal spec in the general vendor's equivalent unit and fixed it in four days. It is a boring success, and I have learned to love boring.
What I would do differently: I would document the decision trail as I went. I should have saved the old pump's seal specification before it went to scrap. Instead, I spent two days reconstructing email threads for finance and trying to prove why we did not reuse the first quote. (Mental note: technical specs are not paperwork—they are evidence.)
Last thought on drift theory
So, what is the drift theory, in plain terms? It is the observation that systems don't fail all at once. They fail slowly, one unmeasured deviation at a time. In purchasing, the drift is all the words we don't define: standard, soon, equivalent, should be okay. Each one seems harmless until it costs you a quarter of your monthly production hours.
The best fix I have found is to make people draw boundaries. Ask your supplier: what will you not sell me, and who is better for that part? A specialist who answers honestly is worth more than a generalist who says yes to everything. That is why I will keep working with Andritz Soutec—and why I tell every new buyer to check the background of the person on the other side of the quote.
If you ever meet an applications engineer like Lewis Alexander, don't let the low-key style fool you. He drew a line around his company's expertise, and that line made me trust him with the rest. That is basically what I have stopped doing on autopilot and started doing on purpose.