Investigation of possible approaches for consequential electricity markets in ecoinvent

Version 3 of the ecoinvent database is the first background database to offer the possibility to link life cycle inventories (LCI) according to different linking algorithms. Using those algorithms, ecoinvent now provides both average and long-term, marginal background LCI data, (e.g. for electricity markets for different countries). Long-term, marginal data for electricity markets ideally represents the additional power plant capacity that needs to be installed to cover a change in demand for electricity in that market. Such LCI data is used in consequential LCA (cLCA) and is needed for an accurate assessment of the environmental impacts of a decision to alter the source or production method of a product system using electricity. The current linking of inventories into ‘consequential’ electricity markets is associated with shortcomings, e.g. due to the fact that market-specific conditions are not modelled, only already producing technologies are included, and electricity produced in combined heat-and power-plants is considered not taking part in future installations and results in unrealistic output of future build-margins. Because the majority of goods and services produced in every sector of the economy rely heavily on electricity, the modelling approach for the consequential electricity markets must be improved to allow the ecoinvent database to provide relevant results for cLCAs. In this thesis, a literature study was performed to investigate which approaches are currently being recommended for modelling electricity supply in cLCA. Currently, the consequential version of ecoinvent is defined for small-scale, long-term decisions. From the literature study, the best approach was determined to be the use of nation-wide official forecasts of the power sector covering the time horizon 2008-2035. To determine the reliability of that approach, it was applied to Brazil, China, India, Japan, Russia and Switzerland. The results of these applications demonstrate more diversified mixes for all markets compared to status quo, overcoming the current shortcomings. A continued evaluation of the approach using the IPCC 2007 GWP, 100a and the ReCiPe 2008 life cycle impact assessment (LCIA) methods shows changes in all midpoint indicators, e.g. a clear reduction in Global Warming Potential (GWP), particulate matter formation and acidification for the studied markets with large shares of fossil supply in the current consequential markets. In summation, it is recommended to continue using the forecasts from consistent data sources to improve the remaining consequential electricity markets in ecoinvent. Future work may include using several scenario and time-frame datasets for the ecoinvent user to choose between, as well as implementation of a dynamic optimization model to allow for large-scale decision modelling