Life cycle assessment of metal laser powder bed fusion : A deep dive into the significance of system boundary expansion and improvement potential

Abstract: Metal additive manufacturing (MAM) is a manufacturing technology experiencing a rapid expansion rate. Metal laser powder bed fusion (ML-PBF) is among the most popular techniques in this field. The environmental implications of it are often discussed in literature and compared to conventional manufacturing. However, the system in its entirety, from a cradle-to-gate perspective, has not seen intense scrutiny so far. A Life Cycle Assessment (LCA) often serves as the evaluation method when investigating environmental impacts; however, this method has been proven to be complex and time-consuming. Efforts are made to reduce this burden by, among others, developing streamlined LCA tools for MAM. This thesis presents three different life cycle assessments, each with different system boundaries, methodologies, and data qualities. In all of them, Global Warming Potential (GWP) and CO2 emissions are focused on. The aim of this thesis is to investigate how large the environmental impact of ML-PBF is when considering the whole system, and to compare this to a streamlined assessment, per kilogram of printed AlSi10Mg based on an average production scenario. The database ecoinvent v.3 and the characterization method ReCiPe 2016 midpoint (H) are used for the analysis with wider system boundaries in combination with specific data. Whereas a third-party streamlined LCA tool is used for the LCA with narrower system boundaries, using the specific energy content of the material. Previous research in the field of ML-PBF often neglects the impact of inert gas and attributes a large portion of the impact to processing electricity. Moreover, post-processing and machine impacts are usually not included in the system boundaries but have been advocated by many to be worth investigating. The results in this thesis show that in contrast to previous research, argon gas accounted for the biggest GWP and where process electricity accounted for less than half of argon. A system boundary expansion was also found to lead to an increase of nearly 230 % of CO2 eq emissions, making it significant to the analysis. Many minuscule factors such as machining, various losses, idle time, machine impact and compressed air contributed to this contrast. Combining this with an improvement and generalizability analysis showed that the global warming potential associated with ML-PBF can be lowered by more than 75 % through either altering the electricity mix or optimizing process parameters, both at the company and upstream. Additionally, it was discovered that the LCA calculation method, and deviations in data quality, contributed to a higher difference in the environmental impact than expanding the system boundaries.