Analysis of Energy Consumption at Slow Charging Infrastructure for Electric Vehicles

Straka, Milan; Carvalho, Rui; Van Der Poel, Gijs; Buzna, L'ubos · 2021 · DOAJ

DOI: 10.1109/ACCESS.2021.3071180

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the factors influencing electricity consumption at slow charging infrastructure for electric vehicles (EVs), addressing the need for data-driven insights to support infrastructure deployment and power grid planning. Motivated by the rapid growth of EV adoption and the resulting increase in electricity demand, the authors aim to identify which environmental, economic, and spatial characteristics correlate with energy usage patterns. The research focuses on slow chargers, which are projected to account for over 60% of global EV electricity consumption, and seeks to move beyond intuitive feature selection by systematically analyzing a large set of location-based variables. The methodology employs a data-centric approach using the EVnetNL dataset, provided by ElaadNL, covering 1,386 charging pools in the Netherlands during 2015. The dataset includes over 369,000 charging transactions, detailing connection times, idle times, and energy consumed. To characterize the spatial context, the authors compiled nine geospatial datasets, including population cores, neighborhood statistics, land use, traffic flows, and points of interest from OpenStreetMap. They extracted 195 candidate features representing the vicinity of each charging pool within a 350-meter radius. Statistical methods were applied to handle missing values, address multicollinearity, and select the most influential features correlated with annual energy consumption. The analysis reveals that energy consumption follows a transformed beta distribution, with the number of charging transactions identified as the primary driving factor. Among the spatial features, economic prosperity indicators were the most significant predictors. Charging infrastructure located near residents and businesses with high incomes, or close to expensive newly built housing, showed a positive correlation with energy consumption. Conversely, a high concentration of residents receiving social assistance had the largest adverse impact on usage. The study also differentiated findings based on rollout strategies: for strategically deployed infrastructure, business types, resident working sectors, and nearby public venues were linked to higher consumption, whereas for demand-based deployments, population age structure characteristics were more influential. The significance of this work lies in its methodological contribution to feature selection for EV charging models. By identifying specific, interpretable location features—particularly those related to socio-economic status—the paper provides a framework for collecting relevant data to improve prediction accuracy. These insights help planners understand the spatial heterogeneity of EV charging demand, enabling more efficient infrastructure placement and better integration of EV loads into power systems. The study highlights that while temporal forecasting is well-established, spatial analysis using robust statistical variable selection offers critical value for long-term infrastructure planning.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• demographic disparities