Forecasting the composition of emerging waste streams with sensitivity analysis: A case study for photovoltaic (PV) panels in Flanders

Forecasting emerging waste streams is challenging because of high uncertainty on the product lifetime and material composition, which should be taken into account to forecast the evolution in amount and composition of such waste streams. In addition, it is impossible to collect complete product lifetime data for emerging waste streams. For this reason, a method based on the theory of the bathtub curve, used in reliability engineering for taking account of large data uncertainties, is presented here and used to forecast the materials that will appear in emerging waste streams. The presented methodology is applied in a case study to forecast the material composition of waste from silicon based photovoltaic (PV) panels in the region of Flanders in Belgium. For this case study, the lifetime distribution of PV panels is determined based on the analysis of the amounts of waste collected by the European collection program PV CYCLE, as well as expert consultation. The results of the presented forecasts indicate not only that a wide variety of materials has been used in different concentrations in PV panels, but also that a high uncertainty remains on when these waste streams will have to be recycled. The results also demonstrate that up to 22,000 ton per year or up to 3.4 kg per capita of silicon based PV panels will have to be recycled in Flanders in the near feature, which means that up to 0.2% of the annual environmental impact of Flanders could be mitigated by targeting all materials present in EoL PV panels for closed loop recycling. (C) 2017 Elsevier B.V. All rights reserved.