Exploring plastic waste identification and sorting through NIR spectroscopy and automated colour sorting

Plastic waste recycling has become a growing concern throughout the world. In Canada, three million tonnes of plastic waste are produced every year, but only 9 % of it is actually recycled. The remaining plastics are either incinerated, placed in landfills, or mismanaged and pollute the environment. In Prince Edward Island (PEI), plastic waste is sorted manually. If a plastic item can't be identified, it can't be recycled. Colourful plastics will also often go unrecycled due to manufacturers finding them undesirable. In order to address these issues, automated sorting could be implemented instead. Two types of automation that see some use in the industry are near infrared (NIR) spectroscopy and colour sorting. In this thesis, the identification capabilities of NIR spectroscopy were tested on six common household plastics. A spectral library was created from the samples and the correlation percentages for the plastics was roughly 86 %. NIR could be used to correctly identify every plastic with the exception of high-density polyethylene (HDPE) and low-density polyethylene (LDPE), which both correlated to each other. Due to its presence in the recycling industry, NIR has some potential to be implemented. A fully automated colour sorting apparatus was used to explore the effectiveness of colour sorting. Three different colour inputs could be examined at once. Black, white, red, and dark gray samples were tested, and black samples could be successfully separated from the other colours. Other colour combinations were unable to be tested, so automated colour sorting will need to be further studied.