The influence of system boundaries and baseline in climate impact assessment of forest products

Artikel i vetenskaplig tidskrift, 2019

Purpose
This article aims to explore how different assumptions about system boundaries and setting of baselines for forest growth affect the outcome of climate impact assessments of forest products using life cycle assessment (LCA), regarding the potential for climate impact mitigation from replacing non-forest benchmarks. This article attempts to explore how several assumptions interact and influence results for different products with different service life lengths.
Methods
Four products made from forest biomass were analysed and compared to non-forest benchmarks using dynamic LCA with time horizons between 0 and 300 years. The studied products have different service lives: butanol automotive fuel (0 years), viscose textile fibres (2 years), a cross-laminated timber building structure (50 years) and methanol used to produce short-lived (0 years) and long-lived (20 years) products. Five calculation setups were tested featuring different assumptions about how to account for the carbon uptake during forest growth or regrowth. These assumptions relate to the timing of the uptake (before or after harvest), the spatial system boundaries (national, landscape or single stand) and the land-use baseline (zero baseline or natural regeneration).
Results and discussion
The implications of using different assumptions depend on the type of product. The choice of time horizon for dynamic LCA and the timing of forest carbon uptake are important for all products, especially long-lived ones where end-of-life biogenic emissions take place in the relatively distant future. The choice of time horizon is less influential when using landscape- or national-level system boundaries than when using stand-level system boundaries and has greater influence on the results for long-lived products. Short-lived products perform worse than their benchmarks with short time horizons whatever spatial system boundaries are chosen, while long-lived products outperform their benchmarks with all methods tested. The approach and data used to model the forest carbon uptake can significantly influence the outcome of the assessment for all products.
Conclusions
The choices of spatial system boundaries, temporal system boundaries and land-use baseline have a large influence on the results, and this influence decreases for longer time horizons. Short-lived products are more sensitive to the choice of time horizon than long-lived products. Recommendations are given for LCA practitioners: to be aware of the influence of method choice when carrying out studies, to use case-specific data (for the forest growth) and to communicate clearly how results can be used.