Environmental Assessment of Food Losses and Reduction Potential in Food Value Chains
Open access
Autor(in)
Datum
2018-12-12Typ
- Doctoral Thesis
ETH Bibliographie
yes
Altmetrics
Abstract
In order to enable future generations to lead a decent life, human consumption of natural resources and impacts on the environment must urgently be reduced. A cornerstone among all human activities is food consumption, which is responsible for roughly one third of all environmental impacts of consumption. The food supply chain is inefficient, as present studies estimate roughly one third of the edible food to be wasted globally. There are numerable political commitments to dramatically reduce FW, notably the UN’s recently released Sustainable Development Goals (SDG) calling to halve per-capita retail and consumer food waste (FW) by 2030. In order to identify promising interventions for FW reduction and to involve the key stakeholders able to successfully implement such interventions, detailed quantitative information on the amount, origin, and environmental impact of FW is needed.
There has been an increasing body of literature related to FW in the past years. Nevertheless, due to data inconsistency and a narrow temporal, geographical, and food supply chain coverage in present literature, FW quantification is still associated with large uncertainties and based on many assumptions. Due to insufficient data about the composition and the treatment methods of FW, existing environmental assessments are rough estimations. Thus, the present state of knowledge is insufficient to under-stand the current situation and to quantify the future reduction potential.
The goal of this dissertation was to provide methods and data to identify FW hotspots in terms of amounts and environment and assess reduction measures. We therefore developed a new approach that can be applied to food systems of any region or country and that would provide a solid information base to support the identification, prioritization, and implementation of effective strategies for FW reduction.
To reach this goal, we defined three subgoals: 1) The creation of a simplified model of the food value chain in form of a mass flow analysis (MFA) in order to understand the system and to quantify FW by origin and type of food. 2) The extension of the model with life cycle assessment (LCA) in order to quantify environmental impacts of FW and to identify hotspots of environmental relevance. 3) For a selection of hotspots identified in subgoal 2, the assessment of case studies, in which measures for FW reduction are exemplarily implemented and their effect measured in terms of mass and environmental impacts.
The thesis starts with a bottom-up quantification of FW across the entire food system related to Swiss food consumption. We chose this life-cycle consumption based perspective, which includes domestic production and net imports, in order to capture the FW-related resource use and emissions induced by Swiss consumers. The result is an MFA of the entire food value chain including the stages ‘agricultural production’, ‘trade’, ‘processing’, ‘retail’, ‘food services’, and ‘households’ and encompassing relevant methods of FW treatment (‘animal feeding’, ‘anaerobic digestion’, ‘composting’, ‘incineration’, ‘disposal in the sewer’). We thereby differentiated 33 food categories as well as edible and inedible parts of food (avoidable and unavoidable FW). Since the unit “wet weight” of FW, which was used in the MFA, is not an appropriate indicator for the nutritional value of food, we converted the MFA into an energy flow analysis (EFA) based on the nutritional value of food and FW. The results identify wasted ‘fresh vegetables’ and ‘cereals’ to be the main quantitative hotspots in terms of mass and ‘cereals’ and ‘oils and fats’ in terms of nutritional energy. The stage of the food value chain contributing most to total FW amounts were ‘households’ (40% in terms of energy). However, these results do not necessarily reflect the environmental relevance of FW.
In the next step we therefore coupled the MFA with life cycle inventory data. We adopted and extended the system boundary of the MFA in order to take the entire life cycle of all inputs into account (agricultural production, transport, cooling, processing, cooking, and partly packaging). In addition, we modelled the environmental impacts of FW treatment. In order to consider useful outputs from FW treatment (e.g. energy and fertilizer from anaerobic digestion), we adopted the method of ‘system expansion’ and substituted heat from natural gas, electricity from the Swiss grid, nutrients by inorganic fertilizer, and organic matter by peat. Since the nutritional values of the products within some of the 33 modelled food categories varied considerably, we allocated environmental impacts to consumed and wasted food based on their nutritional value. This is important since allocation by mass would imply that, for instance, 1kg of whey can substitute 1kg of cheese, which is unrealistic. The life cycle impact assessment was carried out for the impact categories ‘climate change’, ‘biodiversity loss due to land and water use’, and the aggregated method ‘ReCiPe’. The results showed that the total climate change impacts of food consumption could be reduced by 25% if all edible FW was avoided. Furthermore ‘fresh vegetables’, ‘whey’, and ‘beef’ were identified as hotspots for climate change and ‘cocoa’, ‘beef’, and ‘wheat’ as hotspots for ‘global biodiversity loss’. The impact assessment confirmed the results of the MFA that ‘Households’ are key actors for FW, contributing 51% to the climate change impacts and 41% to biodiversity loss caused by total FW.
Since it is unrealistic to avoid all FW, in a next step we analyzed the effect of measures for FW reduction in real case studies. We therefore selected the food service sector, since the rate of FW has been identified to be largest in households and food services and since the Swiss Federal Office for the Environment chose the food service sector as a starting point to develop its strategy to reduce FW. We analyzed 13 case studies, in which food services implemented measures for FW reduction and measured their FW before and after implementation. We then extrapolated the achieved reduction to the entire food service sector, by weighing the subsectors ‘restaurants’, ‘school and university canteens’, ‘hospitals and care centers’, ‘business canteens’, and ‘hotels’ proportionally to the number of meals consumed in each subsector. In order to increase the reliability of the status quo FW amounts in individual subsectors, we included additional publications from Germany, Austria, Finland, and the UK and thus based our results on 1’042 measurements of status quo FW amounts. Considering the FW composition in the status quo and the reduction scenario, we calculated the environmental benefits of potential future FW reduction. In addition to this base scenario, which assumes that all food services achieve the same reduction as our case studies in the corresponding subsector on average, we calculated an extended scenario, in which food services additionally buy 50% of their vegetables from non-marketable origin and thus prevent them from being wasted in the supply chain. The results show that in-house FW is reduced by 38% and related climate impacts by 41% in the base scenario. In the extended scenario an additional 32% of FW and 17% of climate impacts can be saved by using products which otherwise would have been wasted in the supply chain. Thus, the SDG of halving per-capita FW was not reached in the food service sector by the base scenario, but by the extended scenario it was. This shows the importance of considering all stages of the food value chain in order to develop effective reduction strategies. Additionally, we quantified FW per meal in the entire supply chain of a progressive restaurant specialized on FW minimization. With 26 g/meal FW over the entire food value chain, this restaurant only causes 10% of the 252 g/meal estimated for average food services, suggesting that FW reduction on the long-term is larger than the achievements in our case studies, if innovative approaches are implemented.
Another way of reducing FW is to improve supply chains logistically, e.g. due to improved cooling systems or packaging for enhanced food preservation. In addition to the environmental benefits from reducing FW, in such cases also the additional environmental impacts of the improved cooling or packaging system need to be considered. We therefore coupled the LCA with a quality evolution model based on the thermophysical cooling history of the product. With the new methodology we exemplarily analyzed different supply chain options for oranges imported from South Africa and Spain to Switzerland, differentiating 3 cold chains (‘forced-air precooling’, ‘cold storage’, ‘ambient loading’) and three types of packaging (‘standard box’, ‘supervent box’, ‘opentop box’). The results identify a trade-off between direct environmental impacts of the cold chain and indirect environmental impacts from potential FW reduction due to better quality, which can only be evaluated by coupling the product’s quality evolution empirically to the FW amounts. While this was not yet done in the current study, in some cases the optimal solution could be identified without further analyses, e.g. in the case of precooling with solar energy, which saves environmental impacts compared to diesel-driven cooling in the container and provides better quality of the products.
We conclude that the method applied in the thesis of coupling MFA and EFA with LCA turned out to be an appropriate methodology to calculate environmental impacts of FW. The methodology represents a solid basis for further development and extensions into a model to evaluate scenarios and to support stakeholders in the food industry and policymakers to develop successful strategies to reduce FW and related environmental effects. By combining LCA with quality evolution modelling (and its implications on FW), such a model could be used to logistically improve supply chains. Digitalization and monitoring of parameters influencing the products’ quality (such as temperature, quality, degrees brix, etc.) can give new insights about the products’ storage life and help to improve food management. A further breakdown of food categories and the integration of a dynamic transport and seasonality model, which calculate environmental impacts depending on the season and the origin of the food, would further improve the quality of the results. Mehr anzeigen
Persistenter Link
https://doi.org/10.3929/ethz-b-000347342Publikationsstatus
publishedExterne Links
Printexemplar via ETH-Bibliothek suchen
Verlag
ETH ZurichThema
food waste; food waste prevention; MFA; LCAOrganisationseinheit
03732 - Hellweg, Stefanie / Hellweg, Stefanie
ETH Bibliographie
yes
Altmetrics