Life cycle assessment of Australian sugarcane production with a focus on sugarcane growing

Abstract

Past life cycle assessments (LCA) of sugarcane (Saccharum officinarum) production have commonly been based on limited datasets, and variability has not been well described. In this work, Australian sugarcane production was assessed more comprehensively in order to generate a robust set of LCA results for use in subsequent assessments of sugarcane products and also to investigate: (1) variability due to regional differences, (2) factors influencing variability, and (3) significance of the impacts. An average scenario for Australian sugarcane production was modeled based on data for the state of Queensland (98% of Australian production). Life cycle impact assessment (LCIA) results were generated using Impact 2002+, modified to be more representative of Australian conditions, and with the inclusion of water use and land use indicators. A Monte Carlo uncertainty analysis, using minimum and maximum values for production data, was undertaken to evaluate variability. Different regional production practices were also modeled to identify factors that influence variability. Normalization aimed to show the significance of total Australian sugarcane production relative to total Australian impacts. Considerable variability was found in the LCIA results, with the key variables being yield, N use efficiency, the susceptibility of soils to N leakage, irrigation (water and energy intensity), and pre-harvest burning. N leakage was found to be an important issue that influences a range of impact categories. When normalized against total national impacts, water use and land use appear to be the most significant impacts (based on simple indicators of consumption), followed by eutrophication potential, acidification potential, and respiratory impacts, whereas non-renewable energy input and global warming are less significant. The results suggest that toxicity impacts are insignificant; however, this may not be supported by other observations that link pesticide loss from sugarcane to toxicity concerns in receiving waters and is a subject for further research. The potential for significant variability in the impacts from sugarcane growing suggests a need for LCAs of sugarcane systems to consider ranges for key variables. The key variables and significant impacts identified in this work can guide data collection priorities for future assessment of sugarcane and possibly other Australian cropping systems. To further develop LCA as a useful predictive tool for Australian agricultural systems, further development and testing of impact assessment models for eutrophication, toxicity, and land and water resource depletion appropriate for Australia and its subregions will be required.