A scalable life cycle inventory of an electrical automotive traction machine-Part II: manufacturing processes

Purpose A scalable life cycle inventory (LCI) model of a permanent magnet electrical machine, containing both design and production data, has been established. The purpose is to contribute-with new and easy to use data for life cycle assessment (LCA) of electric vehicles by providing a scalable mass estimation and manufacturing inventory for a typical electrical automotive traction machine. The aim of this article (part II of two publications) is to present the manufacturing data with associated collection procedures, from material constituents to complete motor. Another objective is to explain the gate-to-gate system boundaries and the principles for linking the LCI model upstream, to database data, in order to create a full cradle-to-gate dataset. Methods Data for design and production of electrical machines has been compiled from books, scientific papers, benchmarking literature, expert interviews, various specifications, factory records, and a factory site visit. For the manufacturing part, new primary data was collected directly from industry, with a motor factory and a steel mill in Sweden as main contributors, and from technical literature. Other LCA publications were used, if presented in sufficient detail to be disaggregated and revised, to match the gaps of the model. The data represents the current level of technology and targets high-volume manufacturing to the largest extent possible. Also, flows crossing the system boundary have a recommended link to Ecoinvent data, or a request for an attentive selection of input data, depending on the user's object of study. A distinction was made between the regular and an extended system boundary, wherein the processing of some smaller subparts was accounted for through proposals of ready-made Ecoinvent activities for production efforts. Results and discussion An extensive new dataset representing electrical machine manufacturing is presented, and, together with the estimation of motor mass and configuration of article part I, it forms a comprehensive scalable LCI model of a typical automotive electric traction motor. New production data includes a complete motor factory, electrical steel production, and the fabrication of a neodymium-dysprosium-iron-boron (Nd(Dy)FeB) magnet. In addition, smaller, new datasets cover the composition of silicon steel, the making of electrolytic iron, enameling of copper wire, and die casting of aluminum. Conclusions Successful data generation required data building from multiple sources and access to expert support. Transparent, well-explained, and disaggregated data records were found to be crucial for LCA data validation and usefulness.