Consistency and robustness of forecasting for emerging technologies: The case of Li-ion batteries for electric vehicles. (July 2017)

• Record Type: Journal Article
• Authors: Sakti, Apurba; Azevedo, Inês M.L.; Fuchs, Erica R.H.; Michalek, Jeremy J.; Gallagher, Kevin G.; Whitacre, Jay F.
• Is Part Of: Energy policy. Volume 106(2017)
• Page Start: 415
• Page End: 426
• Publication Date: 2017-07
• Subjects: Electric vehicle -- Lithium-ion battery -- Battery design -- Expert elicitation -- Technology forecasting; Energy policy -- Periodicals; Politique énergétique -- Périodiques; Electronic journals; 333.79
• DOI: 10.1016/j.enpol.2017.03.063
• Languages: English
• ISSNs: 0301-4215

Abstract

Abstract: There are a large number of accounts about rapidly declining costs of batteries with potentially transformative effects, but these accounts often are not based on detailed design and technical information. Using a method ideally suited for that purpose, we find that when experts are free to assume any battery pack design, a majority of the cost estimates are consistent with the ranges reported in the literature, although the range is notably large. However, we also find that 55% of relevant experts' component-level cost projections are inconsistent with their total pack-level projections, and 55% of relevant experts' elicited cost projections are inconsistent with the cost projections generated by putting their design- and process-level assumptions into our process-based cost model (PBCM). These results suggest a need for better understanding of the technical assumptions driving popular consensus regarding future costs. Approaches focusing on technological details first, followed by non-aggregated and systemic cost estimates while keeping the experts aware of any discrepancies, should they arise, may result in more accurate forecasts. Highlights: New technology cost projections often confuse underlying technical assumptions. We use a novel method combining expert elicitations and production cost models. 55% of experts' component-level cost projections contain internal inconsistencies. Technical assumptions driving popular consensus regarding future costs are key.