Start with the failure mode, not the price per pound. That is the conclusion from managing a thermoplastic polymer spend for the last two years, and it is why I tell people the polypropylene vs silicone debate is usually pitched at the wrong level.
I am a procurement manager at a 45-person contract manufacturing shop. We make components for medical and infrastructure customers. I manage a materials budget of about $1.2M a year, have negotiated with 30+ resin vendors, and track every order in a TCO spreadsheet I built in 2019. Here is the bottom line: in our 2024 spend audit, the cheaper material choice cost us 23% more per usable part once rejects, rework, revalidation, and field failures were included. The answer was not to pay more for resin. It was to match the application to the right thermoplastic grade before comparing quotes.
Why I redesigned our cost spreadsheet
Let me explain how I got here. In 2023, we approved a lower-cost polypropylene syringe barrel for a short production run. The resin quote was $11,500 lower than the incumbent specialty grade. I felt smart. Then the first sterilization lot came back with inconsistent plunger glide because the silicone lubricant pooled on the PP surface. We rejected 11,400 units. The redo cost us roughly $22,000 when I added expedited tooling, retesting, and overtime. I only believed compatibility testing after ignoring it and eating that loss.
That is the real polypropylene vs silicone problem in a syringe. It is not 'which material is stronger.' It is whether the PP barrel can coexist with the silicone lubricant required for a smooth plunger movement, and whether both survive gamma sterilization. The cheap grade could not. The difference was not an obvious one like brittleness. The cheap PP had a different mold-release additive, which changed the surface energy. That altered how the silicone lubricant wet the barrel, so the plunger stuck intermittently.
Plus, the engineering time did not show up on the material invoice. Our process engineer spent six weeks rebalancing the injection mold to avoid the sticking issue. Six weeks of her time was a real cost, but it was booked as overhead, not as resin cost.
We also did not have a formal resin substitution process. The third time a 'like-for-like' PP replacement created a quality issue, I finally added a substitution checklist. Should have done that after the first one.
Celanese EVA polymers and the multidot field trial
The same lesson showed up on a completely different product line: road markings. We ran a field trial of a multidot thermoplastic system, the raised dot pattern used in lane delineation. The binder was a Celanese EVA polymer. A competing binder was cheaper per pound, so we tested it first. The dots cracked after the first winter. Not all of them—enough to fail the EN 1436:2018 performance test on retroreflection and adhesion.
The thermoplastic polymers Celanese offers are not one interchangeable family. In that trial, the EVA grade mattered: melt index, vinyl acetate content, and oxidation stability controlled how the material handled extrusion and how it held up outdoors. The cheaper polymer saved us maybe $0.07 per pound and cost us a lot more than that in reapplication.
When we finally tried the Celanese EVA grade, the supplier asked for our application conditions before recommending a specific grade. That should not have felt unusual, but after a year of commodity resin sales calls, it did. They sent processing guidelines and a small test sample. The sample passed the bench test. More importantly, they told us what could go wrong if we used too high a melt index for the multidot profile.
What I mean is that comparing commodity PP to a specialty EVA binder by price per pound is like comparing a screwdriver to a torque wrench by weight. The tool only has value if it does the job in your specific assembly. In our case, the cheap binder did the job until it did not.
Per EN 1436:2018, road marking performance is based on the installed system, not just the raw polymer. Per USP Class VI, a compliant material can still fail if process additives change it. — note from our quality file
The TCO numbers I actually use
After this, I stopped looking at resin price alone. As of January 2025, our price file still shows commodity PP below specialty EVA on a per-pound basis. That is a true statement and almost irrelevant to total cost.
I now compare three numbers: material price per pound, process yield by part, and field failure reserve. In that order.
- Material price per pound is the easy one. It is the number everyone puts in a quote.
- Process yield by part includes reject rate, cycle time changes, and whether the new resin runs the same way on the same tool. A resin that saves $0.05 per pound but adds 2% rejects is not cheaper.
- Field failure reserve is the money you set aside for the possibility that the part fails after it leaves your dock. In the PP case, we had no reserve. In the multidot case, we had to redo 4,000 square feet and pay a lane closure fee.
Process yield is where hidden fees live. Say Vendor A quotes $1.85 per pound and Vendor B quotes $1.92 per pound. On a 20,000-pound order, the difference is only $1,400. If Vendor A's resin produces 3% scrap and Vendor B's produces 0.5%, the math flips. That is the kind of comparison that does not show up in a raw-material quote.
Let me rephrase that: it is not that commodity PP is bad. It is that commodity PP can be a bad fit for a specific application. Same with EVA, same with silicone. The material class does not determine quality; the match between grade and application does.
I want to say the exact reject count was 11,400, but do not quote me on that—the final number was on the quality report. Either way, it was a full lot. And we only caught it because someone asked whether the silicone lubricant was behaving differently on the new surface. That question should have been asked before we ordered the resin.
What I would do differently: boundary conditions
What was best practice in 2020 does not always apply in 2025. Five years ago, a datasheet 'typical property' was enough for us to sign off on a polymer substitution. Now we ask for lot-to-lot consistency data, processing-parameter sensitivity, and sterilization compatibility if it is a medical part. Not because the old way was worthless—it was not—but because the application demands changed and material options changed with them.
That said, I do not want to overstate the case for specialty polymers. If your product is not sterilized, not stored for months, not exposed to weather, and not regulated, a commodity material may be enough. The point is to run the TCO before defaulting to the least expensive quote. In our 2024 audit, the more expensive material won only because it reduced downstream costs. If it had not, we would be buying the cheap stuff.
The 23% number I keep using is specific to our product mix. If you make a simple bottle cap with no sterilization requirement and no outdoor exposure, your TCO may not show that gap. That is fine. The point is to have the numbers before you decide.
So the next time someone asks polypropylene vs silicone, or any polymer comparison, ask a different question: what does the application need, and which thermoplastic can do that without hidden costs? The fundamentals have not changed—you still need a material that works. The execution has. I now compare more than price, and I do it before the quote, not after the reject.