Development and validation of energy demand uncertainty model for electric city buses. (August 2018)

• Record Type: Journal Article
• Authors: Vepsäläinen, Jari; Kivekäs, Klaus; Otto, Kevin; Lajunen, Antti; Tammi, Kari
• Is Part Of: Transportation research. Volume 63(2018)
• Page Start: 347
• Page End: 361
• Publication Date: 2018-08
• Subjects: Electric vehicle -- Energy demand -- Uncertainty -- Modelling -- Simulation; Transportation -- Research -- Periodicals; Transportation -- Environmental aspects -- Periodicals; 354.76
• DOI: 10.1016/j.trd.2018.06.004
• Languages: English
• ISSNs: 1361-9209

Abstract

Highlights: Development of computationally efficient dynamic simulation model for electric bus. Validation with real-world tests resulted in a 0.1% difference in energy demand. The most crucial unknown factors were related to operating route characteristics. Abstract: The prediction of electric city bus energy demand is crucial in order to estimate operating costs and to size components such as the battery and charging systems. Unfortunately, there are unpredictable dynamic factors that can cause variation in the energy demand, particularly concerning driver choices and traffic levels. The impact of these factors on energy demand has been difficult to study since fast computing sufficiently accurate dynamic simulation models have been missing, properly quantified in terms of relevant inputs which contribute to energy demand. The objective is to develop and validate a novel electric city bus model for computing the energy demand, to study the nature and impact of various input factors. The developed equation-based model predicted real-world electric city bus energy consumption within 0.1% error. The most crucial unmeasurable input factors were the driven bus route, the number of stops, the elevation profile, the traffic level and the driving style. This understanding can be used to specify routes and stops for a given electric bus battery capacity. Worst-case scenarios are also necessary for electric bus sizing analysis. The best- and worst-case levels of the crucial factors were identified and with them synthetic best- and worst-case speed profiles were generated to demonstrate their effect to the energy demand. While the measured nominal consumption was 0.70 kWh/km, the computed range of variation was between 0.19 kWh/km and 1.34 kWh/km. For design sizing purposes, an electric city bus can have a broad range of possible energy consumption rates due to mission condition variations.