During RP2, the project achieved fiber purity targets for tires and hoses, but productivity was insufficient. Separator productivity improved by April 2024. LAVARIS optimization for tires failed, while MTB optimization for hoses succeeded with a two-stage shredding process. However, separator performance for ELT and ELH fibers did not meet TRL5. Pilot productions and scale-up are ongoing. WP1 also saw IRIS and Synergies TLC defining specifications for a TRL7 prototype for real-time, multilayer PET textile identification, focusing on PU detection. IRIS sorted 140kg of complex clothing, yielding 32kg of high-PET garments, reaching TRL5 after manual dismantling.
WP2 focused on 4 feedstocks:
- Puffy jacket (100% PET, Inditex): Carbios processed 320kg, yielding 216kg of pellets. Expansion quality needs improvement, but PET content (93%) and crystallinity (17%) are acceptable.
- Garment (>80% PET, IRIS sorted): Two batches sorted by IRIS achieved >80% and >90% PET content. After hard point removal, IFTH shredded the materials. Crystallinity (19%) exceeded the 15% goal, possibly due to PU or process conditions.
- ELT : Thermal amorphization was chosen. The "Main Scenario" (diluting ELT with industrial waste fibers) improved processability, leading to 14% crystallinity and 91% depolymerization. This scenario was chosen for pre-industrial scaling. Pilot-scale thermal amorphization at IFTH yielded 200kg of pellets (14+/-2% crystallinity), now micronized for Carbios.
- ELH (Mandals fibers sorted by PPrime and Lavaris): Plasma amorphization was selected due to PU content. Lab tests at Michelin, Carbios, and IFTH achieved <15% crystallinity (except with nitrogen).
Process upscales were also investigated:
- Extrusion (ELT and clothes): Chosen for cost and productivity. Puffy jackets were amorphized using Carbios' extruder (200 kg/h) after compaction. ELT fibers required IFTH's pilot extruder (4-6 kg/h) and fine shredding. Michelin plans to develop a feeder for fluffy fibers and upscale extrusion.
- Plasma (ELH feedstock): Chosen due to PET degradation with extrusion. Michelin is building a pilot process for 100 kg of non-woven fibers (early 2026), followed by semi-industrial upscale. IFTH and Michelin are developing a protocol for determining the amorphous/crystalline ratio, addressing PA removal.
In WP3, 12 new ELT samples were evaluated, with a PET-enriched tyre waste feedstock (50% ELT, 50% PET fiber) validated for pilot-scale. ELH and ELT processes were separated in early 2025. 14 new ELH samples were evaluated, but PET-enriched hoses waste has not fully depolymerized at lab-scale. Pilot-scale depolymerization of lab-validated ELT feedstock (69% PET) occurred without MEG recovery. While kinetics were slow, yield was compliant. TA separation was complex. Carbios will produce 40 kg of recycled PTA by July 2025. Puffy jacket depolymerization kinetics were also slow (70h vs. 24h target), though yield was acceptable, and TA purification was challenging. Carbios produced 78 kg of recycled PTA. Lab trials for complex garments began in April 2025, with pilot recycling planned for April 2026.
Carbios produced anhydrous sodium sulfate (98.5% recovery) during the ELT pilot, but impurities were present. Future pilot loops will focus on feedstock selection, drying, and filtration. Wet filter cake was sent to MICHELIN for valorization assessment. Carbios produced 1.5t of rPET in june 2023 but it was insufficient for WP4. A new production in early 2025 yielded 1.8 tons of improved rPET pellets for WP4 and WP5 testing.
Moving to WP4, DITF converted 45kg of PTA from puffy jackets into 48.7kg of rPET and prepared 50.4kg of vPET for comparison. They optimized PET's diethylene glycol for ELT fibers. IFTH prototyped 14kg of HMLS multifilament yarn. Puffy jacket r-PET yarns were non-homogeneous, brittle, and had lower breaking strength and higher elongation. Kordsa's v-PET dipped cord studies produced a 200-meter sample for Michelin's tire tests, showing feasibility despite cord breaking strength issues. Amorphous r-PET chips are near specifications but need optimization.
Due to the unavailability of PET yarns and cords (WP4), WP5 focused on r-PET specifications, risk management, and technical prognoses. Raw material specifications and test plans for tyres, hoses, and garments were adapted. r-PET limitations pose technical risks, especially for tyre performance. Polyester textile specifications and test plans were based on pilot scale constraints, technical issues, and using v-PET as a reference. r-PET performance will be compared to v-PET through characterization and product tests. Producing v-PET under the same conditions as r-PET is crucial for technical prognoses.
