EPRI Examines Feasibility of Collecting EV Driving and Charging Data the Tesla WayPosted to Electric Power Research Institute (EPRI)
- Dec 21, 2020 10:08 pm GMTDec 21, 2020 6:57 pm GMT
- 447 views
As the number of electric vehicles (EVs) on America’s roads and highways increases over the next decade, utilities need to prepare for new charging loads. “To handle the impacts and determine where they might need extra grid capacity, utilities need data on when, where, and how much EVs charge.” said Jamie Dunckley, an EPRI expert on EV charging and driving patterns. “The problem is, there isn’t enough data today.”
Between 2016 and 2018, EPRI and Salt River Project tracked the charging and driving patterns of 100 EVs using data loggers installed under the hood. While the study provided helpful data about average annual electricity use (between 2,700 and 3,300 kilowatt-hours) and where charging took place (80% was done at home), conducting a project like this on a large scale would be cost-prohibitive.
“For data collection to be cost-effective and timely, we need to find ways to do it without a physical device placed on the vehicle,” said Dunckley. “This can reduce costs and simplify logistics because a device wouldn’t have to be mailed to customers and installed on each vehicle.”
All vehicle manufacturers track the driving and charging behavior of their vehicles remotely, though Tesla is the only manufacturer that makes this data available to drivers. Tesla does this using an application programming interface (API), which provides password access to vehicle data stored in the cloud (with driver permission). Tesla drivers can in turn allow others to access the information through password access.
EPRI conducted a study to determine if using a publicly available API was a feasible method for collecting large amounts of data. Working with two data aggregation companies and 10 utilities, EPRI collected data on the charging and driving behavior of 37 Tesla drivers (all utility employees) over 60 days. The drivers granted permission for the study to proceed. Data included:
- Each charging session’s duration, location, maximum charge level, and average power use (in kilowatts)
- Where and how long vehicles were parked when not being charged
- Where vehicles were driven and charge levels at the start and end of each trip
- Vehicle load shapes
Drivers interacted with the project through either a smartphone app or a web dashboard (depending on which data aggregation company the driver was assigned to). All drivers received emails with project data.
EPRI developed three legal documents to protect the drivers’ data while complying with rigorous privacy standards:
- A customer enrollment policy describing the data that would be collected and how it would be used and protected
- A letter to participating utilities establishing strict limits on how collected data could be used. One example: utilities interested in having a third-party company analyze the data needed to inform EPRI and ask customers to sign a waiver
- A document outlining requirements for third-party data collection and analysis, including data protection measures and assurance that all data would be deleted within 60 days of the project’s completion
These documents can potentially serve as a model for future EV data collection initiatives.
While API-based data collection removes the logistics and expense of data collection with physical devices, the study revealed some challenges that need to be addressed. First, researchers found that the energy consumption associated with data collection (which occurred every two to seven minutes) can vary depending on whether the EV is parked, driving, or charging. When a car is parked, for example, additional energy is needed to turn on or “wake up” the battery in order to transmit data. Frequent data collection when the vehicle is not plugged in for long periods could drain the battery. Future research could explore solutions, such as less frequent data collection or collecting more frequently during charging and less at times that could deplete the battery. Second, a few drivers opted not to participate in the study because of concerns about granting access to their driving data. While many drivers were willing to participate, the experience pointed to the importance of good communication with drivers about password and data protection.
The project demonstrated that relative to data loggers, the Tesla API offers a more secure, cost-effective way to collect data. “With good program design and communication, remote data collection provides a greater opportunity to scale compared to physical devices because of simpler logistics, lower costs, and elimination of installation activities,” said Dunckley.
EPRI has been investigating data collection methods similar to those examined in this project. Since 2012, EPRI has worked with 9 EV manufacturers and 15 utilities to develop and test the Open Vehicle-Grid Integration Platform. A novel approach to gather EV data without data loggers, the platform uses existing vehicle on-board telematics systems—which include cellular, global positioning systems, and other technologies—to collect and transmit data about a vehicle’s location, speed, maintenance requirements, and charging habits. The idea behind the platform is that all EV manufacturers would provide data to a web dashboard, and utilities and other interested parties could download the data for insights on regional EV driving and charging patterns.
This article appeared in the winter issue of EPRI's Efficient Electrification newsletter. To sign up to receive future issues, visit: https://www.electrificationcommunity.com/contact-form.