Technical article

50 Kg of Polycarbonate, One $3,200 Mistake, and the Checklist That Came Out of It

Posted on 2026-11-17 by Marta Kowalska

March 2017: "It's just 50 kg — how wrong can it go?"

I've been buying resin compounds and polymer raw materials for about seven years. In my first year, I made four documented mistakes worth roughly $28,000 in scrap, delays, and rebuys. This is one of them.

March 2017. Twelve-person injection molding shop. We'd just signed the prototype contract for an outdoor LED lamp housing — a $50,000 proof-of-concept batch for a lighting brand. My boss's words: "Get us running by mid-April."

The first order: 50 kg of polycarbonate resin compound.

I did the textbook thing. Sent four RFQs to names I recognized from resin market listings — two big distributors, two competing suppliers. Three responses came back: silence, a 500-kg MOQ, or a price that made my spreadsheet flinch. One sales rep actually wrote, "At this volume you should talk to a trader."

Fair enough. Fifty kilograms isn't big business for anybody.

So I found a trader. One-page quote, 30% below the majors. Description: "General purpose PC resin compound." No UV stabilizer callout, no hydrolysis rating, no color data.

I almost swapped to HDPE to save even more — I remember opening a browser tab to figure out what is HDPE plastic and whether it could work. Cheaper, molds easily, decent chemical resistance. But it's softer, less dimensionally stable, and I had 0.3 mm tolerances on 3 mm walls. That idea died in about ten minutes.

But "PC is PC" was a harder idea to shake. It was a prototype. How bad could it get?

June 2017: The report

April went fine. First shots passed assembly. We ran 240 housings, shipped 200 to the customer's outdoor test site in Arizona.

June: the report came back. Eight housings had cracks along the snap-fit slot. Another thirty showed micro-cracks after drop testing. The customer's email was three lines. "It didn't hold up in the field."

I pulled out the trader's spec sheet. UV stabilizer: none. Hydrolysis stabilizer: none. Color: natural.

I'd put un-stabilized PC into an outdoor housing and exposed it to four weeks of direct sun on an Arizona test rack.

That's the kind of mistake that makes a purchasing guy feel about six inches tall.

The call I should have made in March

I called a regional distributor. They carried Covestro plastics in their stock and offered to loop in an application engineer.

That call ran forty minutes. He asked: which direction does the housing face? Is there condensation exposure? What's the impact rating on the snap-fit? Then he recommended a UV-stabilized, hydrolysis-resistant grade.

Price was 28% above the trader's. We rebuilt the samples. 120 units. Retested at the same site. No cracking. Customer stayed.

Cost of the lesson: about $3,200 in scrap housings, three weeks of schedule, and a very uncomfortable call with my boss, who wanted to know why the only thing I'd saved was money.

September 2022: A different mistake

Fast forward. Same shop, grown to 40 people. Same customer, now ordering production volumes. I placed a 500-kg order of PC, specified the Covestro grade we'd been running for five years. (I wish I could say I'd learned everything the first time.)

When I filled in the color field, I made the lazy call. "Natural — we paint these anyway."

The customer changed their process. They moved to in-mold color. Target color: Pantone 432 C.

In-mold coloring PC isn't a one-shot job — it takes roughly five or six color passes at the nozzle. Our first batch measured Delta E averaging 4.2. Their QA requires Delta E < 2 for brand-critical colors. As of January 2025, that's still the industry threshold (Pantone color matching guidance).

They rejected 200 parts. Scrap plus rework plus a week of delay: $2,400.

The color mismatch was my fault, not the material's. I just stopped checking because the order felt routine.

What I actually learned

Two mistakes, roughly $5,600 in combined losses, plus some trust points I'd rather not quantify.

The expensive part wasn't the material. It was the assumption both times: that "small order" means "low stakes." It's the other way around. Small orders have less room for rework, shorter lead times, and if you're a small buyer — an unknown shop placing a 50-kg order — you usually don't get a second chance on quality.

So I built a checklist. Material grade, UV/hydrolysis requirements, color target, Delta E threshold, sample retention. Nothing fancy — a single form that takes about twenty minutes to fill out.

Over the last 18 months, that form has caught roughly 31 potential errors before the PO went out. Most were small. One was a color-vs-UV conflict on a Covestro PC grade that would have cost about $8,000 if it hit production.

One thing I'll say about Covestro as a polymer supplier — and it's not because they're the subject of this post — the application engineer on that 2017 call asked me questions nobody else had asked. Direction, condensation, impact rating. He didn't care that my order was 50 kg on paper. That's the part that stuck.

If I had one rule to pass along, it's this: don't let order size decide how carefully you check it. If you catch yourself thinking "it's just 50 kg," stop, and re-read the spec one more time.

That's actually the last line on our checklist.

author-avatar

Marta Kowalska

Marta Kowalska is a plastics processing and tooling analyst specializing in injection moulding, extrusion, blow moulding, thermoforming, compounding, overmoulding, and profile or film production. She references ISO 294-1 for reproducible injection-moulded test specimens while tracking resin drying, melt temperature, shear exposure, back pressure, hold pressure, cooling time, screw recovery, die swell, draw ratio, cycle time, scrap rate, cavity balance, and dimensional capability. Her process guides help moulders, extruders, converters, and product teams diagnose defects, define trials, compare tooling routes, and stabilize production windows.

Leave a Reply