Skip to Content
EnergiesEnergies
  • This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
  • Article
  • Open Access

24 September 2026

24 Pages

Wireless 22 kW Charging at Scale: An N-Box Architecture for Light-Duty EV Fleet Charging Infrastructure

,
,
,
,
and
1
Institute of Electrical Energy Conversion, University of Stuttgart, 70569 Stuttgart, Germany
2
EnBW Energie Baden-Württemberg AG, 76131 Karlsruhe, Germany
3
ifak Institut für Automation und Kommunikation e.V., 39106 Magdeburg, Germany
4
Physikalisch-Technische Bundesanstalt (PTB), 38116 Braunschweig, Germany

Abstract

Automated bidirectional charging of electric vehicle fleets offers massive energy system benefits that have barely been leveraged so far. Inductive charging, due to its contactless power transfer, represents a particularly robust solution for automation and avoids the mechanical wear and robotic-handling requirements inherent to conductive systems with high plug-in cycles. This article investigates scalable topologies for inductive charging infrastructure with per-point power levels up to 22 kW. For this purpose, an N-box design architecture is introduced to systematically map power electronics sub-components into functional boxes between the grid interface and charging pads. The proposed framework thus enables a comparative cost assessment of supply schemes, i.e., DC link, 50 Hz AC, and 85 kHz high-frequency power distribution. The results show that using an optimized N-box architecture with grouping of four to seven charging points reduces the total CAPEX costs of an infrastructure from two to 50 parking lots by up to 32% on average compared to scaling state-of-the-art designs from private inductive home charging. Architectural grouping is therefore identified to be more important than the choice of power electronic topology. Further conclusions are listed at the end of the article.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.