I manage purchasing for a 130-person mining services company. Every year I place roughly 60–80 orders for wear parts, rebuild kits, and service components across about 30 vendors. I report to both operations and finance, so I hear about every problem from two directions. I don't design the plant. I just make sure the right thing gets to the right site at the right time—and that finance doesn't reject the invoice afterward.
When I search 'FLSmidth company' I get the usual corporate overview: global engineering group, long product list, service centers around the world. That's the surface. The part that doesn't show up on the overview page is the revision history of each product line.
So when an operator asks for a FLSmidth cone crusher mantle or a FLSmidth Knelson concentrator bowl, I have to translate that into a purchase order. Usually it's easy. Sometimes it isn't. The hard cases aren't the ones where the part number is missing. The hard cases are the ones where the part number is printed, stamped, even confirmed by the vendor—and the part still doesn't fit.
A lot of buyers assume the problem is vendors: the aftermarket supplier sold me the wrong thing. That's sometimes true. But the deeper issue is that mining equipment isn't a static, standardized product. It evolves in ways that a model number alone doesn't capture.
I'm not saying industrial parts should be as simple as buying Simparica for dogs. But the comparison is useful. In consumer products, the label means the same thing on every order. In mining equipment, the label can mean a family of products with differences you can't see in a photo. A part number might cover several revisions, and the revision that's on your site depends on when the equipment was built and how it's been upgraded since.
One example: a plant I supported ordered a spare bowl for a FLSmidth Knelson concentrator. The bowl fit. Then the team realized the feed pipe had been upgraded in a later revision, and the new bowl didn't align with the old feed pipe. The vendor had quoted the correct model number for the concentrator. It simply wasn't the same revision that was on site.
Honestly, I'm not sure why the catalog didn't expose that revision. My best guess is that legacy data from acquired products doesn't always get cleaned up. FLSmidth has been serving the cement and minerals industries since 1882, according to its own website (flsmidth.com). That long history is great for process know-how. It's less great when you're trying to map an old part number to a new catalog.
Crusher mantles and Knelson bowls get attention. The failures that make me look bad to my VP are usually the small components with no good search path.
I once ordered a replacement grease pump for an apron feeder. The OEM bill of material said Lincoln, which should have been simple. But the part number we found through a third-party lookup was for a lower-pressure version. The stamping on the old pump was too worn to read. We caught it only because the supplier asked for a photo of the data plate and noticed a slight difference in the outlet port size. That photo took ten minutes. The wrong pump would have meant a weekend of rework.
You don't need me to solve the old 'hawk vs handsaw' question from Hamlet. In purchasing, the point is simpler: if you can't tell the difference between two revisions, you're going to buy the wrong one.
The apron feeder stopped. The wrong grease pump didn't start it. The mine operator called me while I was on the way to the airport. We had to expedite the correct pump from another state. Shipping cost $900. The pump itself was $1,400. The plant lost about six hours of feed because we had to do the job twice. Those six hours cost more than all of it, but the number on the invoice was only $2,300.
I had two hours to approve that expedited quote before the shutdown window closed. Normally I'd have pulled the original BOM and sent photos. There was no time. I approved it. In hindsight, I should have pushed back. But with operations waiting, I made the call with incomplete information. That was the mistake (note to self: read the whole BOM before approving).
I'm not saying every wrong part causes a catastrophe. Sometimes you can return it and feel foolish. But the 12-point checklist I created after my third mistake has saved us an estimated $8,000 in potential rework. 5 minutes of verification beats 5 days of correction.
The fix isn't exciting. I'm not asking for a full forensic audit of every purchase order. I'm asking for twenty minutes of boring checking that keeps the plant from becoming a testing lab.
This sounds like extra work because it is. It's also the cheapest insurance I've found. The vendor who couldn't provide proper invoicing cost us $2,400 in rejected expenses. The 12-point checklist I created after that mistake has saved us far more than it costs.
My experience is based on about 400 MRO orders for a mid-sized mining services company. If you're working with a brand-new mine or a major capital project, your process might need more detail. But I've seen enough examples to know: the part number is a start, not a conclusion. When in doubt, call the OEM. Their job is to know the revision differences that a third-party lookup won't show you. That call is the cheapest insurance you'll buy all year.
Discuss This Topic
If this article connects to an active wear issue at your plant, use the inquiry form to continue the conversation with our advisory team.
Tell Us What Is Wearing Fastest
Share plant stage, maintenance window, and the result you want to improve so our team can respond with a practical next step.