The environmental impact of additive manufacturing (AM) in consumer-electronics repair remains largely unexplored. AM allows on-demand production of spare parts, making it possible to repair products without available replacement parts (RPs). However, its per-part impact is higher than conventional injection molding (IM). When no digital part exists, redesigning it—such as through the 3DPfR framework—often requires multiple testing and printing iterations, further increasing its environmental impact. Literature focusses on either emphasizing qualitative approaches like AM-enabled sustainable design strategies or quantitative life cycle assessment studies that overlook these trade-offs and the broader system dynamics. Few take a systematic, quantitative approach to evaluating AM in the context of product repair. This study addresses this gap through a life cycle assessment case study of the Philips Senseo HD6569/00 coffee machine.
The following input variables are defined: lifetime extension, part mass and five responses to product failure (RtPF). The five RtPFs are compared in this study are: product replacement (Replace), repair using a single injection molded RP (IMRP), IMRP including n overproduced IM RPs (IMRP-n), on-demand RP production using a pre-existing AM-ready digital model (AMRP), and using the 3DPfR framework to design an RP for AM and requiring n printing iterations to achieve an acceptable part (AMRP-n).
LCA is employed as the primary method due to its ability to systematically and quantitatively assess environmental impacts across an entire product lifecycle. The study uses ReCiPe 2016 (Hierarchist) to evaluate environmental impact midpoint categories, with the functional unit defined as “providing one year of coffee machine use for a consumer in the Netherlands”. The study takes a cradle to grave approach, including a logistics scenario, but excludes the machine’s use- phase impacts. The next phase of the study involves performing a sensitivity analysis and contribution analysis as well as depicting tipping points between different repairs compared to product replacement.
The study defines one alternative as environmentally ‘favorable’ over another if it has a lower impact score in all impact categories compared to its alternative. Results identify electricity consumption during printing, 3D printer production, and RP material impact differences as key drivers of AM’s increased environmental impact over IM. As a result, AM-based repairs always lead to a higher environmental impact compared to IM-based repairs. This difference becomes more evident when additional AM print iterations are required, or the number of overproduced IM parts increases. When comparing repair to product replacement, IM-based repairs are far more likely to be environmentally favorable over product replacement compared to AM-based repairs. Therefore, the number of IM overproduced parts is not considered a sensitive variable, whereas the number of print iterations is. However, the exact tipping point is highly depending on the input variables.
IM-based repairs should be prioritized whenever available. If no IM RP is accessible, minimizing the number of AM print iterations is crucial, with single print-on-demand manufacturing being the most favorable option. When neither an IM RP nor an on-demand AM RP is available, a part can be redesigned using the 3DPfR framework. In this case, the environmental favorability of repair over replacement depends on the number of required print iterations, part mass, and expected lifetime extension, and careful assessment should be made. While the scenario evaluation tool developed in this study provides insights into these trade-offs, its findings are highly context dependent. If AM RPs designed via 3DPfR contribute to an open-source, on-demand manufacturing library, their broader sustainability benefits may justify their initial environmental impact.
The impact of overproducing IM RPs is less significant compared to the number of AM iterations. Specifically, for the same part mass and lifetime extension, producing 10 IM RPs or performing 2 to 6 AM print iterations can be environmentally favorable over product replacement, depending on the packaging type used for transporting the IM RPs.