1/ Upstream collection // Foreground: Data collected upstream of rank 1 processing are specific and cover 100% of Small Household Electrical and Electronic Equipment tonnage managed by Eco-systèmes in 2014, i.e. 68% of Small Household Electrical and Electronic Equipment tonnage managed by the French take-back schemes. These data were provided by the Eco-systèmes logistics department and cover the methods of conditioning, transport distances, dimensions of Heavy goods vehicles (HGV), their load rate and their empty return rate. // Background: The upstream collection model was developed using data from ecoinvent V3.4 (allocation, cut-off). The range of European standards taken into account is representative of the French HGV fleet in 2014. ________________________________________________________________________________________________________________________________________________________ 2/ Rank 1 processing // Foreground: The rank 1 processing model is based on specific data collected via questionnaire from 13 operators, which represent 95% of the stream tonnage processed by operators under contract with Eco-systèmes in 2014, or 65% of the Small Household Electrical and Electronic Equipment tonnage processed by the French WEEE take-back schemes. The data collected concern the energy and material inputs consumed by operators during processing and specific emissions generated (notably dust). The operators covered represent the diversity of technology in France and Europe for rank 1 Small Household Electrical and Electronic Equipment processing. WEEE processing by rank 1 operators leads to the production not of fully sorted waste, but fractions which are a mixture of different materials to various degrees of complexity. An important issue in the project was to quantify the way in which each material analysed is distributed in the rank 1 output fractions; this analysis is required to understand the scope of final destinations reached by each material and the itinerary taken to reach them (Material Flow Analysis). 100 % of fractions produced by rank 1 operators were taken into account, even when they represented a very marginal proportion of the total tonnage. Each fraction was subject to a quantified description of its composition. Regarding plastic fractions, regulatory framework has evolved since 2014 and is still moving in 2018, implying evolutions in the operational management of plastics and especially for plastics containing brominated flame retardants. In this modelisation a conservative approach is adopted: sorted plastic fractions containing any kind of brominated flame retardants are considered ending in hazardous waste incineration units only. // Background: The rank 1 operator processing model was developed using data from ecoinvent V3.4 (allocation, cut-off) to model the energy and material inputs consumed. Rank 1 operators who process Small Household Electrical and Electronic Equipment are all located in France: electricity is modelled based on the French electricity mix (2015-2017 annual average). ________________________________________________________________________________________________________________________________________________________ 3/ Transport between rank 1 operators and rank 2 acceptors // Foreground: The transport model for fractions involving rank 1 and rank 2 operators is based on specific data collected via questionnaires completed by rank 1 operators. For each fraction produced, the 13 operators involved in rank 1 Small Household Electrical and Electronic Equipment processing provided the identity and location of rank 2 operators and split between them. The coverage rate is 100% of fractions produced; for Small Household Electrical and Electronic Equipment involved, this implies that almost 291 separate journeys were taken into account. // Background: The transport model for fractions involving rank 1 and rank 2 operators was developed using data taken from ecoinvent V3.4 (allocation, cut-off); the range of European standards taken into account is representative of the French HGV fleet in 2014. ________________________________________________________________________________________________________________________________________________________ 4/ Rank 2 and later operators (excluding final destinations) // Foreground: Rank 2 operators may be: i/ trade/bulking operators; ii/processing operators (separation and shredding/separation); iii/ final destination operators. i/ Data concerning rank 2 trade/bulking operators concern loss rates and power consumption by handling equipment: the loss rate is based on assumptions; handling equipment power consumption is considered to be identical to that for rank 1 operators. ii/ The data concerning rank 2 processing operators or later are based on lessons learned from the analysis of rank 1 operators, on the opinions of expert operators and/or Eco-systèmes processing experts using bibliographic data. iii/ See the last paragraph if rank 2 operators correspond to a final destination. // Background: The model for operations by rank 2 and following operators was developed using data from ecoinvent V3.4 (allocation, cut-off); rank 2 operators and later may be located in France, other European countries or in Asia: Electricity consumption is modelled taking into account the electricity mix specific to the country in which each of the rank 2 operators and later are based. Concerning the transport phases downstream of rank 2 operators, no specific visibility is possible about the transport distances. These steps are based on a market-specific approach which distinguishes for each material and each category of final destinations: the portion that is processed/recovered on a national level, the portion processed on a European level, the portion processed on major Asian export markets. These market scenarios were developed based on bibliographic data. _________________________________________________________________________________________________________________________________________________________ 5/ Final destinations // Foreground: For each material analysed, the range and breakdown of final destinations reached result from the MFA based on the composition of fractions produced by rank 1 processing, on the existence of separation processes and their efficiency during rank 2 operations and later. The evaluation of final destinations was done using two methods: i/ With benefits: all the environmental burdens of final destinations are taken into account; the benefits obtained by substitution effects between secondary and primary materials are also taken into account, with the substitution point in question being located at the output of the recovery operation (exit from steel mill, exit from aluminium refinery); ii/ Without benefits: all environmental burdens of final destinations are taken into account; no benefits generated by substitution effects are accounted for. In all cases, the modelling of the behaviour of a given material (e.g. steel) in a final destination it reaches in high proportions (e.g. steel mill) or in lower proportions (e.g. non-hazardous waste storage facility), even in marginal proportions (e.g. copper / precious metal refinery), was done to be as specific as possible to the material in question (e.g. in a steel mill, steel is considered to be largely preponderant in the steel melt, and marginal in steel slag; steel in copper/precious metal refineries is considered to be found in drosses). To the extent of usable knowledge the models of the environmental burdens and benefits take into account these specific behaviours. // Background: With a view to producing a model as specific as possible of the burdens and benefits associated to a given material in each of the final destinations it reaches, in-depth modelling work was undertaken using various data sources: i/ Databases: ecoinvent V3.4, Industry data from Plastics Europe (ecoprofiles published in 2015 for ABS, in 2014 for PP, in 2012 for PS and in 2011 for PC), Industry data from EAA, etc; ii/ Waste LCA tool: Wisard™; iii/ Technical and scientific bibliographies: BREFs, scientific publications, etc. This work was covered by internal documentation resulting in a 200-page report.