Methodology

A web-based survey was distributed to members of an e-bike sharing program in Madison, Wisconsin, through the University of Wisconsin-Madison Qualtrics survey center.  A total of 667 responses were received, and after data filtering, 450 were used as a final dataset. A mode choice model was developed based on the random utility maximization of mode preferences revealed through the survey. For the environmental impact assessment, a WTW life cycle assessment (WTW-LCA) was utilized based on the Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) model.

Findings

• The mode choice model developed from the user survey revealed that before the e-bike sharing program, the mode splits were: personal vehicle (42.5 percent), ride hailing (1.6 percent), bus (14.13 percent), e-bike (0 percent), conventional bike (12.76 percent), and walking (28.94 percent).
• After gaining access to the shared e-bikes with membership, the mode splits were: personal vehicle (35.91 percent), ride hailing (1.96 percent), bus (10.11 percent), E-bike (21.83 percent), conventional bicycle (8.63 percent) and walking (21.56 percent).
• The analysis of survey data also revealed that with the membership program, e-bike trips replaced trips previously made by other modes. Around 30 percent of the e-bike trips were previously made by personal vehicle, and 19 percent previously used the bus. However, 16 percent of e-bike trips were previously made by conventional bicycle, and walking was the previous mode for 33 percent of e-bike trips.
• The environmental impact assessment revealed that, as a result of modal shifts before and after the shared e-bike program, energy consumption decreased by 15.78 percent, and reductions in emissions of various pollutants ranged from 12.61 to 15.92 percent (see Table 1).

Table 1: Results for Environmental Impacts from Mode Shift Model. Source: University of Wisconsin-Madison