Every recycling method we’ve covered so far, mechanical, chemical, biological, depends on one thing happening first. The right material has to reach the right process. Before a bioplastic tray can go through decontamination, or a batch of PLA can undergo chemical breakdown, someone or something has to find it. They need to correctly identify it and pull it out of a waste stream made up overwhelmingly of everything else.
That step is sorting. It rarely gets much attention. MoeBIOS’s own waste characterization work shows exactly why it’s the hardest part of the whole chain.
What we actually found when we looked
MoeBIOS ran real waste characterizations as part of Work Package 1. The team sampled waste at the Montemarta Cónica waste treatment centre in Seville, covering both the general residual waste stream and the dedicated packaging waste stream. Workers broke each sample down manually, material by material.
The results were stark. Bioplastics made up just 0.17% to 0.25% of the total residual waste stream by weight, across two separate characterizations. They made up 0.38% of the dedicated packaging waste stream. Conventional plastics (PET, HDPE, LDPE, PP, PS, PVC combined) made up roughly 15-44% of the same streams by comparison. Bioplastics weren’t just a minority material. They were closer to a rounding error.
The textile side told an even starker story. Alia analyzed a 500 kg sample of separately-collected textile waste with NIR technology. The sample contained effectively no bioplastic content at all.
This matches what European Bioplastics’ own industry-wide market data shows. Bio-based plastics represent roughly 0.5% of the 431 million tonnes of plastic produced globally each year. The European Environment Agency’s Circularity Metrics Lab confirms this same figure independently. MoeBIOS’s local findings and the continent-wide statistics tell the same story from two different angles.
Why that tiny fraction is such a big sorting problem
Picture a sorting plant built to process tonnes of ordinary plastic waste per hour. It has to somehow detect and separate a material that might account for less than half a percent of what passes through it. That’s a fundamentally different challenge than sorting an established, high-volume category like PET.
Labeling makes it harder still. Manufacturers mark many products “bio,” “compostable,” or “biodegradable.” Others fall under the generic “category 7 other plastics” label, which also covers unrelated multi-layer packaging. MoeBIOS’s own waste analysis found that this labeling confusion pushes most bioplastics into the organic waste stream rather than the packaging stream. That makes them harder to isolate and more heavily contaminated once separated.
How sorting actually works today, and what has to change
At Montemarta Cónica, plastic sorting currently runs through a cascade of optical sorters. LEITAT and PICVISA are the technology partners developing and refining this equipment within MoeBIOS. One machine separates plastics from non-plastics. Each subsequent sorter then pulls out a specific plastic type: PET first, then HDPE, then a mixed PP/PS/PVC fraction. Whatever doesn’t match a known material profile at that final step gets rejected. Until now, that reject stream is exactly where bioplastics have ended up, unrecognized and lost.
MoeBIOS’s proposed fix works in two parts. First, update the very first sorter’s database so it recognizes bioplastics as part of the plastics fraction from the start, rather than routing them toward rejection immediately. Second, add an entirely new, fourth optical sorter at the end of the existing cascade. This sorter would specifically catch bioplastics out of what would otherwise be discarded.
Textiles: a different sorting problem entirely
Textile sorting doesn’t rely on the same optical sorter cascade. In the territory managed by ALIA, textile waste currently goes to companies for manual selection. A new Textile Hub, using Fibersort technology, will soon bring automated NIR-based sorting online. It’s set to process 20,000 tonnes per year of post-consumer textile waste and 13,000 tonnes per year of pre-consumer waste.
MoeBIOS’s own sample found essentially no bioplastic content in the current textile stream. That changes the sorting challenge here. It’s less about extracting a small fraction from a large one today, and more about building the detection capability now, ahead of where the bioplastics market is headed.
Agriculture: when the material can’t be sorted the usual way at all
Agricultural bioplastics present the strangest sorting challenge of the three value chains. A lot of what’s actually out there, like mulch film, is deliberately designed to biodegrade directly in the soil. That’s the opposite of what mechanical or chemical recycling needs: a stable, separable material. Similar compostability makes other products, like clips and twines, difficult to separate from the crops they become tangled with.
Because of this, and because so little bioplastic material currently exists in real agricultural waste streams, MoeBIOS took a different approach entirely. The project sourced new bioplastic products directly from manufacturers and farmers, then artificially simulated realistic contamination and weathering, rather than relying on existing waste streams alone.
Why MoeBIOS had to simulate the problem to study it
Here’s the practical consequence of everything above. Because real post-consumer bioplastic waste is still so scarce, MoeBIOS couldn’t simply collect it and study it directly. Instead, the project sourced bioplastic packaging, textiles, and agricultural products directly from producers. The team then deliberately contaminated them, mixing them with real waste for 48 to 72 hours, the realistic time between a bin being emptied and reaching a sorting plant, to simulate what genuine post-consumer bioplastic waste will look like as the market grows.
The actual data made this necessary. You can’t optimize a sorting system, or test a decontamination process, on a material that barely shows up in the waste stream yet.
Where this fits in MoeBIOS
This work is documented in full in D1.3, “Specifications of Waste Streams”, MoeBIOS’s public deliverable under Work Package 1. A follow-up deliverable, D1.4, will revisit these same characterizations later in the project to see how the picture has changed as bioplastics production continues to grow.