Improving calculations of energy return on investment

Abstract

Maintaining energy supply is a critical challenge as we strive to transition away from fossil fuels. Energy return on investment (EROI) is a tool widely used by energy analysts to help understand the efficiency with which we extract, deliver and use energy. Initial research in this area focused on the EROI of extracting energy from nature, using direct energy costs where available and deriving indirect energy costs from economic data to infer relatively comprehensive energy cost assessments^1,2. More recent studies have increasingly expanded the boundaries of the denominator by including additional energy required to refine and deliver energy to its final point of use^3,4. Such studies, sometimes called harmonization studies, attempt to ensure consistent comparisons across different energy sources^5,6, and conclude that the EROI of renewables surpasses that of fossil fuels. We find this conclusion surprising, as it is opposite to earlier studies. While we agree on the importance of accounting for all costs associated with energy technologies and applaud the efforts of such studies to “compare apples with apples”⁶, we believe that there are at least five ways in which these assessments could be improved.

First, the most common approach to measuring EROI for renewable technologies is life cycle assessment (LCA). While this approach is usually regarded as accurate within its defined boundary, it is subject to two important types of truncation error⁷. The first is sideways truncation, where many small but collectively significant processes — such as service activities — are excluded because they are individually minor and too numerous to measure. Established LCA cut-off rules often lead to their exclusion, yet they can account for about half of the total energy costs, as demonstrated by more comprehensive environmentally extended input–output analyses (EEIOA) or energy intensities of financial activity^7,8. This truncation could halve the EROI of technologies like solar photovoltaics. The second is downstream truncation, where system-level processes that lie beyond the electrical busbar or inverter — such as storage, firming, and transmission — are typically omitted. These system-level processes are critical for understanding energy transition but are difficult to capture within the scope of an LCA-based EROI study. To address these limitations, studies must expand their boundaries of analysis.