Why going beyond standard LCI databases is important: lessons from a meta-analysis of potable water supply system LCAs

Purpose

Our aim is to assess the comparability and generic applicability of harmonized published lifecycle assessment (LCA) studies on water supply systems. In the absence of localized life cycle inventories for water systems, generic or country specific databases may be inadequate if applied elsewhere. The objectives of this paper are to calculate the potential magnitude of errors introduced by this practice and recommend ways to better account for sources of impact variability.

Methods

In this study, harmonization has been carried out rigorously, utilizing a systematic differentiation of the subsystems, functional units, and system boundaries referenced in over 100 candidate studies, resulting in a comparable subset of 34 LCA studies. Statistical techniques (cluster analysis and Welch’s analysis of variance) were used to isolate and validate the main sources of variation in impact scores and identify the sub-systems in which these are most pronounced. The significance of technology-specific contribution to the impacts was compared to the significance of electricity as a contributing factor to the global warming potential (GWP) by applying statistical correlation analysis.

Results and discussion

Our review revealed that most of the published LCAs analyzed water systems in well-developed countries. Large variation was found in the impacts of water supply systems (e.g., GWP between 0.16 and 3.4 kg CO2-eq/m³ of supplied water), with mean value of 0.84 kg CO2-eq/m3 and median of 0.57 kg CO2-eq/m3. The main contributor to GWP is water production and desalination in particular, making water production the most important differentiating factor. Cluster analysis also showed that production technology is the most important differentiating factor with respect to terrestrial acidification, ozone depletion, eutrophication, and abiotic depletion impacts of water production systems. There is a weak correlation between impact scores of electricity mixes and entire water supply systems.

Conclusions

An LCA of water-intensive products drawing from a standard life cycle inventory databases could be substantially inaccurate, especially in a region with desalination. More accurate results can be achieved by taking local water production technology into account. Meta-analysis is a useful tool to explore the sources of variance in the impacts of water systems. Applying harmonized results is a cost-effective way for obtaining more accurate LCA results as compared to applying generic databases only.