Last September I stood next to a scrap bin in our molding area and watched a machinist feed the last of 250 rejected clear enclosures into a granulator. Next to him sat the pallet of resin we had bought at a price that Finance called “a win.” From the outside, it looked exactly like the material our client had specified: same clear pellets, same product category, same nominal melt flow. The supplier’s website even used the words “medical grade.” It was not the same thing. Learning that difference the hard way is why I now buy Covestro medical grade polycarbonate without second-guessing the purchase order.
Background: I am the office administrator for a 35-person contract manufacturing company. We build two kinds of products: custom machined equipment for other manufacturers and low-volume plastic housings for medical device startups. I manage the purchasing for all of it—roughly $900,000 a year across 40-plus vendors. I report to operations on the shop side and to finance on the budget side. When I took over purchasing in early 2024, finance gave me a mandate: cut material spend by 8 percent without “sacrificing quality.” That phrase is easy to say and hard to define, and 2024 gave me three very different lessons in what it actually means.
The foam board that taught me to read a purchase request twice
The first lesson came in March, and it was not about a medical device. Our shop supervisor, Dave, put a request in my queue for foam board. We cut a lot of Baltic birch plywood into panels with a track saw, and the crew lays a full sheet of rigid foam across two sawhorses below the plywood. The saw blade cuts through the panel and stops in the foam, which keeps the back of the plywood from tearing and protects the blade. Dave wrote exactly what he wanted in the comment field: “Best foam board for cutting plywood we have used is the 2-inch dense polyurethane. Do not substitute the beadboard stuff.”
The beadboard stuff was $47 a sheet. The dense polyurethane board he wanted was $130. From the end of an aisle, both look like rigid foam board, and I decided Dave was being picky. I ordered the cheap one.
It did not work. A full 4×8 sheet of birch plywood on a $47 sheet of EPS across two sawhorses flexes in the middle. The panel dips when the saw passes over, the kerf closes, and the blade binds. We scrapped a $92 sheet of plywood and lost half a day of Dave’s time while he recut the part. Dave communicated the rest of the lesson with his eyebrows when he put the next request on my desk.
That little incident is worth telling because it made me notice how many polyurethane uses are hiding in a normal fabrication shop: cart wheels, machine bumpers, gaskets, and the dense foam board under our panels are all polyurethane, but they are different grades with different specs. The price of any of them only means something after you know what the job requires. Most buyers focus on the ticket price and miss that part entirely.
When a “cheap fix” is actually the expensive option
The second lesson arrived in May. One of our production fixtures uses a sheet of food-grade polyethylene as a liner. The liner had deep gouges from a rough winter of use. Replacing it meant $480 for the material plus a morning of labor. Our maintenance lead found a kit for polyethylene plastic repair and asked me to order it. The kits use a hot air welder and polyethylene filler rod to melt the damage back together. The name-brand kit was about $210. I found one for $95 from a seller I had never heard of.
The repair looked fine for about a week. Then the patch delaminated under cutting fluid, and filler fragments ended up in a batch of parts. We tore the fixture apart, scraped off the failed repair, ordered the replacement liner anyway, and ran the job again. The $95 kit ended up costing close to $1,100 once labor and downtime were counted, and the fixture was down an extra day. I am not saying polyethylene plastic repair never works. With the right equipment and a sound substrate, it can. That liner was worn and load-bearing, and no amount of welding rod was going to make it young again.
The polycarbonate order that made me rewrite my vendor checklist
The big lesson came in July. We signed a program with a medical device startup to mold 250 clear plastic enclosures for a point-of-care diagnostic instrument. The client’s spec named Covestro medical grade polycarbonate as the reference material. It allowed equivalent materials, but only if they could meet the same requirements: manufacturer traceability, a regulatory file, test data, and notice of any changes to the manufacturing process. I read that paragraph as “clear polycarbonate, medical-ish.” Finance read it as “opportunity.”
A broker quoted us a resin he described as “the same thing, you will never tell the difference,” at about 12 percent less than the authorized Covestro distributor. On a 2,000-pound order, the savings came to roughly $2,300. Finance liked the number. I liked the number. We cut the purchase order and moved on.
For two weeks everything looked good. The resin molded cleanly. The first articles passed dimensional inspection. Then the client’s quality engineer asked for the documentation package, and the broker’s answer began with “no problem” and ended with a spreadsheet that listed one lot number across three different date codes.
I am not a regulatory specialist, so I cannot walk you through the details of ISO 10993 or what belongs in a full medical grade file. What I can tell you from a purchasing perspective is simpler: if your supplier cannot prove what they sold you, you did not buy medical grade polycarbonate. You bought a plastic lottery ticket.
The molded enclosures were put on hold and then rejected when the client’s lab ran its sterilization validation and the samples crazed and clouded. Maybe the resin was a different grade. Maybe it was mislabeled. Maybe it was fine and only the paperwork was missing. That was the entire problem: we had no way to know.
We reordered from an authorized distributor. Covestro medical grade polycarbonate shipped next-day freight with a clear lot number, a manufacturer’s certificate, and written answers to questions. The molder requalified the tool. The client’s lab reran its test samples. We paid overtime to protect our production slot. By the end, the $2,300 paper saving had become about $5,500 of expedite fees, revalidation charges, and overtime—not counting the two-week schedule slip or the phone call where I told a client we were late.
What I would tell another buyer
I still buy generic materials for plenty of things. I still compare quotes and push suppliers on price. But I added one rule to our purchasing checklist in late 2024: the unit price is only a meaningful number when nothing about the application is specified, documented, or regulated. The moment a customer names a brand, references a standard, or asks for proof, the quote is not the decision. Total cost is the decision.
That is why Covestro resins are now on our internal “do not substitute without engineering sign-off” list. Not because the brand is perfect and not because generic polycarbonate has no place outside regulated products. It is because we nearly lost a good client by treating a specified material like a commodity SKU.
The same logic applies at $47, at $95, and at $2,300. The $130 foam board is cheaper than the $47 foam board plus a ruined $92 sheet of plywood and half a day of a skilled fabricator’s time. The $480 replacement liner is cheaper than the $95 repair kit plus a scrap batch and a day of downtime. The full freight and revalidation cost of the right medical grade resin is cheaper than the discount on the wrong one.
Bottom line: when a broker tells you they have “the same thing for less,” ask one question—what can you prove? If the answer is a data sheet and a hand wave, it is not a discount. It is a risk that you will explain to your finance director and your customer after the granulator stops running.
Ask a material question