Don't order Andritz replacement parts by part number alone. You can get exactly what you asked for and still end up with the wrong piece. That's the most expensive lesson I've learned in nine years of handling orders for Andritz equipment.
I'm a process engineer, not a purchasing specialist. That's probably why I've made so many buying mistakes. In 2017, I placed my first big order for stock prep components and got the wrong flange facing. It cost $3,200 in redo work and two weeks of delay. Since then, I've personally documented 14 significant errors across Andritz SINDUS separators, Andritz pumps, and paper machine clothing from Andritz Fabrics and Rolls Latina. Those lessons ate roughly $62,000 of budget. Now I maintain the pre-order checklist our team uses. This article is basically that checklist, with the scars included.
Start with the conclusion
If you're buying Andritz parts for a pulp, paper, or separation process, the part number is a starting point, not an answer. The real question is whether the part matches your installed machine, your current process conditions, and your lead time. I've watched too many orders get rejected at installation because the frame size was right but the orientation was wrong, or because the fabric was designed for a different stock consistency.
From the outside, ordering a replacement part looks like a data entry task. You find the tag on the equipment, type the code, click buy. The reality is a matching problem between the original design, every modification you've made since, and the specific way you run the process. That gap is where the money disappears.
Why I started keeping a checklist
I don't want to pretend I'm a procurement expert. I'm the guy who has to explain to a maintenance manager that the new centrifugal screen won't fit because the inlet flange is rotated 90 degrees. That happened in September 2022. The unit came from Andritz SINDUS, and it was flawless except for one detail: I hadn't checked the existing pipe orientation against the new drawing. The machine was down for three days. The total cost wasn't just the rush freight; it was the lost production on top of it.
After the third rejection in Q1 2024, I stopped trusting my memory and started a simple pre-order checklist. We've caught 47 potential errors with it in the past 18 months. Some were small, like a missing bolt grade. A few were big enough to have killed a project. The checklist didn't make me smarter. It just made the things I already knew, but kept forgetting, impossible to accidentally skip.
Mistake #1: Trusting the data sheet over the machine tag
On older Andritz equipment, the nameplate can be missing, scratched, or simply wrong. A data sheet from 2019 might not reflect a retrofitted seal or a changed motor frame. Andritz's installation manual usually includes a table of flange dimensions and orientations, but you have to verify the actual machine before you order. If the serial number isn't visible, don't guess. Write down every dimension you can measure and send it to the service engineer. That one step would have saved me from the September 2022 failure.
Mistake #2: Ordering stock without confirming the process condition
I'm not only talking about inventory stock. I'm also talking about the paper stock or slurry in your process. A pump impeller with the same part number can behave differently if the temperature, pressure, or solids content has changed. The stock that ran at 4% solids last year isn't the same stock at 6% this year. In one case, we ordered a replacement impeller that was the exact same number as the one we had. It didn't perform because our feed pump was now pushing a denser slurry. The part was fine. The process had changed.
Mistake #3: Forgetting the interaction between consumables and stock prep
Forming fabrics and press felts are not one-size-fits-all. Andritz Fabrics and Rolls Latina has a product engineering group for paper machine clothing because the fabric interacts with the whole wet end. The fabric that worked on the same machine in another mill may be completely wrong for your consistency, chemistry, and press load. I've seen a mill blame the fabric for poor sheet quality when the real problem was a refiner plate pattern that was breaking the fibers too aggressively.
Never expected the cheapest "compatible" part to be the one that caused the longest shutdown. It turned out the imported equivalent was 2 mm shorter than standard on the bearing seat. That doesn't sound like much until the vibration readings start climbing. We replaced it with the correct Andritz part and the readings dropped back to normal. Genuine doesn't mean perfect, but at least the operational manual has a tolerance table for it.
Mistake #4: Ignoring which Andritz subsidiary made the original part
Andritz is a global name, but the manufacturing shop matters. An Andritz SINDUS part built in Brazil can have a different revision letter than the same part number from a European factory. The same is true for rolls and fabric components from Andritz Fabrics and Rolls Latina. The part number prefix often tells you which factory made it. If you assume one source is the same as another, you may get a compatible part that still needs a different housing, a different bolt pattern, or a different service procedure.
The part nobody checks until it's late
Lead times. In my experience, the Andritz lead time for OEM replacement parts is usually six to twelve weeks, depending on the product line and the factory's schedule. If you order stock, meaning something sitting in a warehouse, it can be faster. But custom components from Andritz SINDUS, or a special fabric weave for a paper machine, aren't made until you place the order.
The kicker is that a shorter lead time isn't always better. I ordered a rebuilt item once because it was available in three days. It worked, but we had to reset our calibration schedule and it failed during a routine pressure test. The original component with a six-week lead time arrived a month later and ran for the next four years. Sometimes waiting for the right part is the efficient choice.
This is where I sound like a convert. I used to think efficiency meant shorter lead time and lower price. Now I think efficiency means fewer surprises. A predictable 15-week lead time is easier to handle than an unpredictable 6-week one.
The surprising role of digital monitoring
A few years ago, I would have laughed at the idea of a software tool on our Andritz equipment. Then our team turned on Andritz Peregrine, the remote condition monitoring service. I was skeptical. The surprise wasn't a fake alert or another dashboard with too many buttons. It was a real warning about a pressure screen vibration 36 hours before we could see or hear anything on the floor.
The alert came on a Sunday. We shut down the spare pump, swapped in a bearing block, and avoided what would have been a six-hour unplanned stop on Monday morning. That's the kind of efficiency I can defend: it doesn't eliminate the physical work. It just makes the work cheaper because you do it before the failure, not after.
Peregrine also changed how we buy stock. Instead of stocking every possible part, we could monitor condition trends and order the right consumables when the data said so. The automated process eliminated the data entry errors we used to have. Our stock checks went from a five-day turnaround to two days. That doesn't sound dramatic, but in a mill environment it's the difference between planned maintenance and a crash shutdown.
The "how to make hair look" question nobody asks until it's too late
One of my customers once called me asking, in an awkward translation, "how to make hair look" — he meant the fiber hair on the paper surface. He was focused on the forming fabric and the headbox. We spent two weeks changing stock conditions, retention aid dosages, and drainage profiles. Almost nothing helped.
In the end, the problem was the press felt. It had been installed too long and its surface had compacted. The "hair" he was seeing wasn't a forming issue at all. It was a press section symptom. That's a lesson I've learned to expect: the visible problem often points at a different part of the process than where it appears.
So when someone asks me about Andritz Fabrics and Rolls Latina, I tell them to think about the whole sheet-forming system, not just the fabric. The fabric, the roll covers, the press felts, and the stock preparation all work together. Change one variable and the "hair" comes back somewhere else. If you're in that situation, start with a simple question: What changed last? Not what looks abnormal right now. The answer is usually the cause.
When to ignore this advice
This checklist is for existing Andritz equipment and standard replacement orders. It does not apply to greenfield projects or major upgrades. If you're designing a new line and buying a complete Andritz system, the project engineers will handle the interface details. You won't need to worry about flange orientation because the whole system is drawn together.
It also doesn't apply as strongly to true commodity parts—filters, seals, or standard fasteners. For those, the risk is lower, and local supply is often fine. And if you're keeping a decommissioned machine alive only for a short planned shutdown, don't spend extra on OEM parts. Spend just enough to get to the finish line.
One honest caveat: digital monitoring like Peregrine is not a magic wand. It works best when the machine is already well understood and you have reliable baseline data. If your process changes every week or your operators skip the daily readings, the software will still give you useful signals, but it won't save you from your own chaos.
Bottom line
Don't buy Andritz parts the way you'd buy a shirt online. Check the serial number, check the machine tag, verify the process conditions, and confirm the lead time before you commit. The most expensive sentence in industrial procurement is "I assumed." The cheapest one is "Let's check." I do not want you to repeat my mistakes; that's the only reason I put the numbers in writing.