It is highly challenging to measure the efficiency of electric vehicle charging stations (EVCSs) because factors affecting operational characteristics of EVCSs are time-varying in practice. For the efficiency measurement, environmental factors around the EVCSs can be considered because such factors affect charging behaviors of electric vehicle drivers, resulting in variations of accessibility and attractiveness for the EVCSs. Considering dynamics of the factors, this paper examines the technical efficiency of 622 electric vehicle charging stations in Seoul using data envelopment analysis (DEA). The DEA is formulated as a multi-period output-oriented constant return to scale model. Five inputs including floating population, number of nearby EVCSs, average distance of nearby EVCSs, traffic volume and traffic congestion are considered and the charging frequency of EVCSs is used as the output. The result of efficiency measurement shows that not many EVCSs has most of charging demand at certain periods of time, while the others are facing with anemic charging demand. Tobit regression analyses show that the traffic congestion negatively affects the efficiency of EVCSs, while the traffic volume and the number of nearby EVCSs are positive factors improving the efficiency around EVCSs. We draw some notable characteristics of efficient EVCSs by comparing means of the inputs related to the groups classified by K-means clustering algorithm. This analysis presents that efficient EVCSs can be generally characterized with the high number of nearby EVCSs and low level of the traffic congestion.

1. 서 론
2. 문헌연구
3. 연구방법
    3.1 투입요인 및 산출요인
    3.2 분석방법
4. 효율성 분석결과
5. 결 론
References

• Dong-Hoon Son(홍콩과학기술대학교 토목환경공학과) | 손동훈
• Yeong-Su Gang(인하대학교 아태물류학부) | 강영수
• Hwa-Joong Kim(인하대학교 아태물류학부) | 김화중 Corresponding Author