Although additive manufacturing (AM) is a promising advanced manufacturing technology, it consumes much more electricity energy than conventional manufacturing techniques during the production phase. This is a barrier to promoting the use of AM technology across manufacturing industries. Energy consumption (EC) of AM has been a subject of academic studies, especially the EC modelling of AM. However, the energy-saving methods to reduce the EC of AM are limited and insufficient. The energy benchmark has been recognised as an effective analytical methodology and management tool to improve energy efficiency. Nonetheless, research about energy benchmarks for the AM is lacking. Selective laser melting (SLM) is one of the most promising and widely used AM processes. As such, SLM is selected as a representative AM process for energy reduction. Because most additively manufactured parts have new designs without historical energy data, we develop an energy benchmark for the
SLM using the prediction model. Besides, we design a dynamic rating system to be consistent with the SLM’s energy-saving potentials, which can effectively evaluate the energy efficiency of each loading scheme of the parts. The case study shows that 18.99 % of EC of SLM for a turbine rotor is reduced by using the proposed energy benchmark methodology. A trade-off between energy efficiency and production tardiness is discussed.