Electric vehicle charging can cause inefficient energy consumption, resulting in overheated wires, transformer stress, and possible power outages, according to the recent results of a research project conducted by the University of Michigan Transportation Research Institute (UMTRI) in Ann Arbor and Utilidata, an AI-powered technology company in Rhode Island.
The project studied the relationship between EV driving and charging behaviors to understand how those behaviors impact the electric grid.
UMTRI installed Utilidata’s distributed AI platform, Karman, via electric meter adapters at multiple EV charging stations on the University of Michigan campus to measure grid performance.
Vehicle monitoring devices also were installed on the EVs of drivers who charge their vehicles at one or more of the selected charging stations to uncover driving and charging behavior.
Data was collected for 16 weeks and found that while there is uncertainty around charging behavior and impacts to power quality, there is opportunity for EVs to add value to the grid.
Findings from this initial study include:
- EV charging can cause unexpected rapid swings in current draw which can lead to inefficient energy consumption, resulting in overheated wires, transformer stress, and possible outages.
- Areas with high volumes of unmanaged EV charging are more likely to experience power outages.
- Utilities need the ability to calculate high resolution measurements with local computation at individual locations to understand which locations can accommodate EV charging.
- EV charging can cause equipment degradation and possible failure as it lowers power quality by introducing current harmonics generated from the conversion of AC to DC power.
- With granular real-time data, utilities can manage EV charging in the presence of faults to mitigate the fault impact and facilitate faster recovery.
“It’s clear utilities need tools at the grid edge to fully understand the impact of EV charging in real time,” says Josh Brumberger, CEO of Utilidata. “Fortunately, the Karman platform identifies these issues and allows utilities to leverage AI models right at the source of charging to predict, model, and manage EV charging to prevent outages, lower grid costs, and add value to the grid.
“Working with UMTRI allowed us to combine our expertise in AI and the electric grid with UMTRI’s expertise in innovation and transportation to identify opportunities and challenges on the grid as EV charging demand increases.”
Utilidata’s distributed AI platform, Karman, is being deployed with utilities to provide real-time visibility of the grid and leverage AI to predict and model grid conditions. This level of visibility and analysis should improve grid reliability and manage distributed energy resources like solar and EVs.
Built on a custom module, based on the NVIDIA Jetson Orin edge AI platform, Karman is designed to be embedded into any grid edge device. Aclara, a division of Hubbell Inc., is the first meter company to embed Karman in a residential smart meter.
In 2022, Utilidata established an advisory board to bring together executives from utility companies, auto manufacturers, and technology providers to explore ways to leverage grid edge data to solve industry challenges like EV charging demand.
“As EV ownership increases, energy providers need real-time data to manage EV charging demand and provide effective customer programs,” says Ryan Jackson, executive director of corporate strategy at Consumers Energy and a member of Utilidata’s advisory board. “This computing platform can help us deeply analyze the impacts of a variety of EVs and EV chargers. These insights can be incredibly invaluable as we see more EVs connect to the grid.”
To view the full results of the project, visit here.
