← Back to all notes
R

Roel Huis in ‘t Veld

Director, Kriya Materials · Week of September 21, 2026

What a pilot plant admits

Covestro opened a recycling pilot in Leverkusen. The useful word in the press release is “pre-industrialization”. On 17 September Covestro opened a pilot plant in Leverkusen for chemical recycling of Vulkollan, the cast polyurethane in forklift wheels, railway buffers and jounce bumpers. Double-digit millions. More than ninety percent of the material recovered by mass, carbon footprint down by up to two thirds against virgin material. Thomas Braig, head of elastomers, called it “a key step in our pre-industrialization journey”. I like that word. It admits, in public, that this is not a business yet. A pilot plant says: we believe this enough to lose money on it for a while, and we do not know the number. Most press releases hide that under the word breakthrough. Covestro said it plainly and the story is more credible for it. Vulkollan is decades old. Nobody recycles it because the chemistry is new. They recycle it because forklift fleets and railway operators ask for it. A market decision dressed as a technology decision, which is the right way round. Our version is smaller. No pilot plant. A reactor that is not making product for a paying customer for a few weeks, a lab coater and one engineer. Same question as Covestro's, at a scale of thousands instead of tens of millions: which chemistry deserves those weeks. In twenty years at Kriya we got that wrong often enough to have three rules. One. A pilot needs a customer who asked for it in writing. A market report is not a customer, a conference panel is not a customer. An engineer who sent a specification with numbers and a date is. Every pilot we ran on a trend alone produced good data and no invoice. Our former CTO, did not live by this rule and when he left in 2013 we had years of "pilot materials" on stock which I depreciated to nil as nobody needed them. Two. A pilot has a number before it starts. What it costs in weeks and material, what it must return, by when. Covestro said double-digit millions. We say six weeks and a drum, and we write down the order that pays for it. If nobody can write that order down, the pilot waits, See rule number 1, I have enough stock without it. Three. A pilot has a stop date, and stopping is not the failure. In 2014 our largest business disappeared within two weeks due to legislation changes in the EU (nobody saw that coming) and took more than half the company with it. Afterwards we deliberately spread the work across many applications and wrote down everything we learned, including what went nowhere. That archive is now the most valuable thing in the building. When a customer arrives with a problem we already failed at once, we start from the report, not from zero. Where coatings come in: a coating supplier recycles nothing, we decide whether someone else can. A hundred nanometres of anti-reflective layer on PET can decide whether the film goes back into the loop or into the incinerator. Customers now ask what is in those hundred nanometres, and whether it contains fluorine. For a coating the metric is how little you put on and what it is made of, decided years before anyone opens a pilot plant. Covestro can afford a double-digit-million pilot. We can afford a disciplined small one. A small pilot with a customer, a number and a stop date beats a large one without them. Say pre-industrialization in your own budget meeting and watch the conversation get honest. Sources Coatings World, 17 Sept 2026 – Covestro opens pilot plant in Leverkusen — https://www.coatingsworld.com/breaking-news/covestro-opens-pilot-plant-in-leverkusen/