L'économie circulaire est-elle bonne pour l'environnement ? Essais sur la mesure de la circularité

The Circular Economy (CE) concept is inspiring new governmental policies along with company strategies. This has led to the emergence of a plethora of indicators to quantify the “circularity” of individual companies or products. Preserving resources and reducing emissions are both necessary conditions to achieve a sustainable CE, and the corresponding assessment tools should be selected accordingly. Approaches behind these indicators build mainly on two implicit premises: (1) maximizing material circularity contributes to mitigating environmental impacts and (2) closing material loops at product level leads to improvements in material efficiency for the economy as a whole. While the first case gives rise to environmental trade-off situations, the second case disregards the scale effects. To the question immediately formulated in the title "Is circular economy good for the environment?", we argue that yes, provided that these two premises can be sufficiently well verified to be monitored. With the literature increasingly questioning these premises and the development of new indicators related to the popularization of the CE concept, we asked ourselves the following question: should we develop new tools dedicated to CE? The literature review reveals different types of tools either used to measure the degree of circularity of a system defined in time and space, or to measure the effects on the ecosphere. To date, no tool allows (1) to capture the whole range of CE strategies, (2) to monitor trade-off situations nor (3) to integrate scale effects. Without providing a new analytical tool, this thesis lays the foundations for a methodological approach that integrates these three criteria. In the first paper, we challenge the first premise that material circularity contributes to mitigating environmental impacts using the Life Cycle Assessment (LCA) and the Material Circularity Indicator (MCI) through two case studies from the tire industry. In the most comprehensive case study, we analyze three scenarios to process used tires in Brazil: baseline, retreading and regrooving. We propose an approach to discuss MCI and LCA results that identifies four pathways toward or away from the CE goal: coupling, decoupling, trade-off related to resource consumption or trade-off related to emissions. The case studies reveal that extending lifetime through retreading and introducing recycled material improve the MCI of a tire, but do not necessarily improve impacts on human health and ecosystems. In the second paper, we challenge the both premises, on the environmental trade-off and the scale effect, with a case study on closed-loop recycling of polyethylene terephthalate (PET) bottles in the USA. The MCI and the impact on climate change revealed by the LCA show that closing the material flows within the scope of a bottle lifecycle increases its circularity and decreases the environmental impacts. This, however, shift burdens to other sectors where recycled PET is applied and where using high quality PET is not required. However, expanding the assessment scope to the entire PET market, open-loop recycling by including other applications to introduce recycled PET, i.e. fiber textile, sheets and films, appears more beneficial and that increasing the post-consumer bottle reclamation rate is the major environmental improvement lever for the PET market. In the third paper we develop a versatile approach to assess the environmental performance of CE strategies, building on an empirical use of the IPAT (Impact = Population x Affluence x Technology). Our approach can distinguish activities by region, industry or process for instance, and the product of environmental efficiency factors, embodied in T, illustrates key drivers for CE implementation, e.g. GHG/materials and materials/product. CE is often linked to decoupling the economic activity from the degradation of the environment. Hence, we use our approach to quantify the contribution of a CE strategy to an overall decoupling according to the scenarios of technological efficiency factors. This type of approach allows, for example, to deduce the maximum allowable increase in the production of a material to ensure the effectiveness of circularity strategies. We illustrate this approach with data from the scenario analysis carried out by the Swedish consultancy Material Economics, then taken over by the Ellen Macarthur Foundation. The first paper shows trade-offs in the results between the MCI and the LCA resource indicator and thereby highlight relevant issues to the design of future assessment tools; while the MCI measures the degree of circularity, LCA indicators measure the environmental impact of a circular strategy. Results of the second paper show that improving the circularity of a single product does not necessarily improve products average because product circularity does not necessarily reduce virgin resource demand and environmental impacts at a broader market perspective. Hence, we conclude that a CE assessment scope should be systemic enough to enhance progress towards reducing environmental impacts and that the way product-level circularity assessment is currently performed is contradictory to the symbiotic purpose of CE. The approached developed in the third paper is versatile enough to meet the three criteria that current tools in the literature fail to integrate – CE strategies, monitoring trade-off situations and integrating scale effects – but it currently does not allow the integration of market effects in a pragmatic way. The use of general equilibrium models could fill this gap. To the question "should we develop new tools dedicated to CE?", this research work demonstrate that a better integration of existing tools would suffice. A possible research avenue could be based on, for instance, using the decomposition of the indicator proposed in the third article to identify the “economic structure” characteristic of a product, a component or a material by combining existing methodologies based on a matrix computational structure. The combined use of these tools would assess the contribution of a product to a CE, rather than assessing the degree of circularity of the product itself.