Climate impact and energy efficiency of woody bioenergy systems from a landscape perspective

Highlights: • Harvesting forest residues from a landscape decreases forest carbon stocks. • Willow grown on fallow land has a lower climate impact than forest residues. • Forest branches and tops has a higher energy efficiency than stumps and willow. • All studied biomass fractions have a lower climate impact than natural gas and coal. • A landscape perspective is important to capture both temporal and spatial variations. -- Abstract: The climate impact of bioenergy is debated, especially due to potential land use change effects and biogenic carbon fluxes. This study assessed the climate impact and energy efficiency of conventional long-rotation forest residues (branches, tops and stumps) and short-rotation forestry (willow) from a landscape perspective. A time-dependent life cycle assessment method, which considers the timing of biogenic carbon fluxes and the impact on global temperature over time, was combined with GIS mapping to assess the impact for a specific Swedish region (Uppsala County), i.e. a 'real' landscape. The results showed that harvesting forest residues decreased the forest carbon stocks over the landscape, while growing willow on previous fallow land increased the total carbon stocks. On average, energy ratios of 49 MJ MJ⁻¹ for branches and tops, and 30 MJ MJ⁻¹ for stumps and willow was found. Harvesting forest residues from the studied landscape resulted in climate impacts of around 0.8∙10⁻¹⁵ K MJ⁻¹ heat for branches and tops, and 1.3∙10⁻¹⁵ K MJ⁻¹ heat for stumps. Willow energy gave the lowest climate impact of about −0.6∙10⁻¹⁵ K MJ⁻¹ heat. The landscape analysis showed that spatial variations in the region had an effect on energy efficiency and climate impact, but that this effect was relatively small. A more important factor was the time frame chosen for the analysis, especially for long-rotation forest systems. Methodological choices such as spatial scale (stand or landscape perspective), allocation method and functional unit also influenced the results.