ReDigital: Minimizing residual contamination risks in recycled plastics through digital process-accompanying analysis methods

The approach pursued in the project aims at continuous monitoring of recycling extrusion with highly sensitive real-time digital analysis and enables almost complete analysis and AI-supported optimization of recyclate quality and composition during the ongoing extrusion process.

For recycling companies, this means minimizing risks, costs, and time expenditure. Processors receive reliable recyclate qualities for defined applications and thus meet the legal requirements for recyclate quotas. The recyclate market is evolving.

The current state of the art in monitoring the quality and chemical composition of recycled plastics is that only a small amount of the recycled batch produced is taken in the form of granulate and analyzed on the basis of random samples. This is often done using manual, cost-intensive, time-consuming, and specialized offline laboratory methods, e.g., melt flow index (MFI) measurement, gas chromatography-mass spectrometry (GC-MS), Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy, etc.

Depending on the method, sample quantities vary from approx. 50 g to the µg range. Recycling companies (often SMEs) cannot afford their own testing laboratories and usually have random samples of their recyclates tested by an external laboratory for quality approval. As a result, batches are only released after days or weeks. In the meantime, costs for temporary storage are incurred.

The measurement results of these extremely small test sample quantities are used to approve several tens of tons of recycled material. This, in turn, carries the risk that the small sample quantities measured are not representative of the approved tonnages of recycled plastic, which consist of very heterogeneous input streams.

At the same time, the non-representative measurement results can lead to an entire batch not being allowed to be used for its intended purpose.

Due to these risks, many users categorically reject the use of recycled plastic. 

Investment at IKK

Procurement and installation of four different inline/online analysis modules on an existing semi-industrial recycling extrusion line to determine the chemical composition of the plastic melt and the gas phase simultaneously using different, complementary measurement principles. The data generated by the various modules is collected and, in combination with innovative processes from the fields of

• artificial intelligence (AI),
• machine learning (ML), further regression models, and
• the use of chemometrics, 

is used for the direct analysis and optimization of the ongoing extrusion process and the development of prediction models. 

Funding: € 1,261,530

Contact

Dr. Madina Shamsuyeva
shamsuyeva@ikk.uni-hannover.de
+49 511 762 18345