The Day Our Epoxy Batch Didn’t Set
It started with a routine Q3 audit—inspection of a 500-gallon batch of epoxy resin hardener meant for industrial concrete floor coatings. The spec sheet said ‘density: 1.15–1.17 g/cm³’. The lab came back with 1.13. Not dramatically off, but off. I flagged it. The production lead shrugged: “It’ll still cure, just a little slower.” I’ve heard that line before.
Background: The Concrete Floor Job
This batch was destined for a distribution center—20,000 sq ft of polished concrete needing a high‑abrasion epoxy coating. The client had specified Olin’s EPICURE™ curing agent (our standard for heavy‑duty floors). The job had a hard deadline: the warehouse was reopening in 10 days. Every day of delay cost the client $8,000 in lost throughput. No pressure.
I assumed the density deviation wouldn’t matter. After all, the chemistry still looked right on paper. Didn’t verify. Turned out the supplier had substituted a cheaper diluent to cut costs, which altered the stoichiometric ratio. The floor coating never reached full hardness—it remained tacky after 48 hours. Workers tracked footprints everywhere. The client called. The project manager was not happy.
“I still kick myself for not running a small test cure. If I’d mixed a 100-gram sample and waited 6 hours, we’d have caught the problem before 500 gallons hit the floor.”
The Fix: Rethinking Our Verification Protocol
We halted the job. The cost of stripping and recoating: $22,000. Plus a week of delay. The supplier agreed to cover materials, but the client’s trust? That takes longer to rebuild.
Looking back, I should have enforced a three‑point check on every incoming hardener batch: density, viscosity, and a mini‑cure test. At the time, the standard delivery window seemed safe (this was back in Q2 2024). It wasn’t.
How to Sand Epoxy Resin Edges (The Lesson From Our Rework)
During the rework, our crew had to sand down the partially cured epoxy edges to prep for recoat. That’s when I learned a trick I wish I’d known earlier:
- Timing matters. Sand too early (within 24 hours) and the surface is gummy—gums up the sandpaper. Sand too late (past 72 hours) and the epoxy is fully hardened, requiring heavy grit and elbow grease.
- Use a block, not just your hand. A rubber sanding block keeps the edge flat and prevents dipping. (Essentially, the same reason you use a level for concrete floors.)
- Lubricate with water. Wet‑sanding reduces heat buildup and clogging. Start with 120‑grit, progress to 220, then 400 if you want a matte finish.
The question isn’t whether you can sand epoxy edges—it’s when and how. And that depends entirely on the curing system. Our EPICURE system, for example, reaches sandable hardness in about 36 hours at 25°C (based on Olin’s technical data sheet 1320‑R2).
The Honest Limitation: When Olin’s Product Isn’t Right
I’m proud of our materials, but I’ll also tell you when they’re a bad fit. If you’re coating a residential garage that gets casual foot traffic and occasional tire drag, our heavy‑duty epoxy floor coating is overkill—both in cost and application complexity. A water‑based epoxy from a local supplier might serve you better. Why? Because our system requires precisely mixed ratios and temperature–humidity control. In a garage with fluctuating conditions, that’s a risk.
“No product is universal” is a boring statement. Saying “I recommend this for industrial floors over 10,000 sq ft with consistent climate control, but if your floor is under 2,000 sq ft and uninsulated, look elsewhere” actually helps people make good decisions. It also builds credibility.
What I’d Do Differently (Hindsight, 20/20)
If I could redo that decision, I’d invest in better upfront specifications. But given what I knew then—nothing about the supplier’s substitution trick—my choice was reasonable.
Here’s what we changed after this incident:
- In‑bound quality checklist now includes a 200‑gram cure test on every hardener batch. Rejects in Q1 2025 dropped from 4.2% to 0.7%.
- Supplier contract updated to forbid substitutions without written approval, with a penalty of 150% of batch value for non‑compliance.
- Training session created for installers on how to detect under‑cured epoxy (finger nail test, acetone wipe) before applying the topcoat.
Final Thought: The Total Cost of Assumptions
The rework cost $22,000. The original batch saved maybe $800 in raw material (the cheap diluent). Net loss: $21,200 plus a bruised relationship. That $21,200 could have bought a lot of test kits and quality control hours.
“Calculated the worst case: complete redo at $22,000. Best case: save $800. The expected value said go for it, but the downside felt catastrophic. Next time, I trust my gut.”
For anyone specifying epoxy coatings for concrete floors, my advice: verify, then verify again. And if you’re sanding edges, water is your friend. (Prices and spec references as of January 2025; always check current Olin technical bulletins.)
— A quality inspector who learned the hard way.