UC Berkeley

Global resource consumption and anthropogenic carbon emissions are increasing at an unsustainable rate, causing noticeably adverse changes to our ecosystem and jeopardizing the ability for future generations to thrive. This realization has brought together designers and engineers to holistically incorporate all aspects of sustainability in the product's entire life cycle using principles such as green engineering and design for environment (DFE) and eco- design tools such as life cycle assessment (LCA). However, to properly assess and facilitate designs and technologies that are indeed more environmentally benign, changes are needed to shift from the conventional serial LCA to a more coupled and integrated sustainable life cycle design (iSLCD) approach that resonates the three pillars of sustainability. A unique concept of the Product Life Cycle Zodiac (PLCZ) is introduced that reveals the complete holistic product life cycle from Earth to landfill and enables the information flow of the different life cycle phases to be fed back or looped for product development and process planning. In addition, the precision of the iSLCD approach can be vastly improved by the leveraging of green manufacturing, such as the scales of green manufacturing (SGM), where changes at the manufacturing process level can propagate throughout all downstream stages.

A case study reflecting the influence of design and manufacturing using the iSLCD frame- work is considered. A potential proxy for large carbon emission reductions is the electrifi- cation of the automotive industry, which has promised to provide a renewable and cleaner alternative to the conventional internal combustion engine (ICE). Alternative energy vehi- cles such as the Polymer Electrolyte Membrane (PEM) fuel cell vehicle utilizes compressed hydrogen to offer zero emissions during the operational use phase. However, despite being commercially available for over a decade, current annual production volumes are more than several orders of magnitude lower than todays conventional ICEs. At current low production volumes the processes for PEM fuel cell manufacturing are burdened with large inefficiencies such as low throughput batch processing (as compared to continuous roll-to-roll processing), high equipment idle times, low material utilization and processing yields. These inefficiencies contribute to an increase of the specific energy consumption (SEC) and hence the environmental impact of the fuel cells to a point where the benefits of zero emissions may potentially be outweighed by the emissions during the manufacture of the fuel cell. Furthermore, the low production volumes and the use of exotic materials such as platinum catalysts impedes the adoption of the technology due to prohibitively high cost. Therefore, it is of interest to analyze in, parallel with the environmental impacts, the cost implications and where identify area of potential cost reductions.

The case study investigates the environmental and economical performance of PEM fuel cell manufacturing for automotive applications. The research is in part a collaborative effort with Daimler-Benz in attempts to assess and improve the current state-of-the-art manufac- turing practices by leveraging the SGM. Detailed unit processes are modeled in terms of energy consumption as a function of manufacturing inputs and are integrated into a facility scale HVAC energy consumption model. The life cycle phases included in the model follow the product life cycle zodiac (PLCZ) from raw material extraction to product distribution and the various end-of-life pathways. The economical aspect is investigated using a design for manufacturing and assembly (DFMA) technique in conjunction with the environmental anal- ysis. A thorough analysis of the results and the breakdown of the component contributions and sensitivity analysis of the model is conducted. The sensitivity analysis provides insights to not only the the fuel cell manufacturing, but also highlights the importance of integrating the SGM. Lastly, the influence of data uncertainty is incorporated using a stochastic Monte Carlo technique.