The Environmental Relevance of Capital Goods in Life Cycle Assessments of Products and Services
Goal and Scope. Many life cycle assessment case studies neglect the production of capital goods that are necessary to manufacture a good or to provide a service. In ISO standards 14040 and 14044 the capital goods are explicitly part of the product system. Thus, it is doubtful if capital goods can be excluded per se as has been done in quite a number of case studies and LCA databases. There is yet no clear idea about if and when capital goods play an important role in life cycle assessments. The present paper evaluates the contribution of capital goods in a large number and variety of product and service systems. A classification of product and service groups is proposed to give better guidance on when and where capital goods should be included or can be neglected. Methods. The life cycle inventory database ecoinvent data v1.2 forms the basis for the assessment of the environmental importance of capital goods. The importance is assessed on the basis of several hundreds of cradle-to-gate LCAs of heat and electricity supply systems, of materials extraction and production, of agricultural products, and of transport and waste management services. The importance within product (and service) groups is evaluated with statistical methods by comparing the LCA results including and excluding capital goods. The assessment is based on characterised cumulative LCI results using the CML baseline characterisation factors of the impact categories of global warming, acidification, eutrophication, human toxicity, fresh- water acquatic toxicity, terrestrial ecotoxicity, ionising radiation, and land competition, based on proxy indicators (fossil and nuclear) cumulative energy demand, and based on the end- point indicators Eco-indicator 99 (H,A) mineral resources, human health, eco system quality and totals.Results. The analysis confirms the fact that capital goods cannot be excluded per se. On one hand, toxicity related environmental impacts such as freshwater ecotoxicity or human toxicity are more sensitive towards an inclusion or exclusion of capital goods. On the other, certain products like photovoltaic and wind electricity are very much or even completely affected by capital goods contributions, no matter which indicator is chosen. Nuclear electricity, agricultural products and processes, and transport services often behave differently (showing a higher or lower share of capital goods contribution) than products from other sectors. Discussions. Some indicators analysed in this paper show a rather similar behaviour across all sectors analysed. This is particularly true for 'mineral resources', and − to a lesser extent − for 'Eco- indicator 99 total', 'acidification' and 'climate change'. On the other hand, 'land use' and 'freshwater ecotoxicity' show the most contrasting behaviour with shares of capital goods' impacts between less than 1% and more than 98%. Recommendations. Capital goods must be included in the assessment of climate change impacts of non-fossil electricity, agricultural products and processes, transport services and waste management services. They must be included in any sector regarding the assessment of toxic effects. Energy analyses (quantifying the non-renewable cumulative energy demand) of agricultural products and processes, of wooden products and of transport services should include capital goods as well. The mixing of datasets including and excluding capital goods is no problem as long as their share on total impacts is low and partial omissions do not lead to a significant imbalance in comparative assertions. Perspectives. If in doubt whether or not to include capital goods, it is recommended to check two things: (1) whether maintenance and depreciation costs of capital equipment form a substantial part of the product price (Heijungs et al. 1992a), and (2) whether actual environmental hot spots occur along the capital goods' supply chain.