Project Description

Electrification of the transportation systems is able to significantly reduce the carbon emissions. It is projected that with grid decarbonization through integration of more renewables, emissions from electric vehicles can be reduced by 75%, from ~200 grams per mile of CO2 emission in 2021 to ~50 grams per mile in 2050. To reap the benefit of carbon reduction, policy makers, such as New York State, has signed into the law to ban the sale of passenger cars and light duty trucks with internal combustion engines by 2035. Together with the policy support, factors including improvements in battery technology and cost, and more charging infrastructures, have also driven the rapid surge of EV sales. Along with the rapid proliferation of the EV and its grid integration, EPRI projected that in New York alone, passenger EVs on New York roads will increase to 950,000 in 2025 and 2.3 million in 2030, respectively, and 50% of new light-duty vehicle sales by 2030 are expected to be EVs. The widespread adoption of EVs necessitates an extensive charging network and computational modeling and simulation of the energy systems are needed for the grid integration of EVs with high demand uncertainties.

In this project, the students will collect the data from real-world EV charging stations. Computational modeling techniques such as Gaussian Mixture Modeling, will be developed to analyze and predict the stochastic behaviors of EV vehicles such as the arriving time at a charging station, charging duration, and energy demand. These charging uncertainties result in congestion and voltage swing of power grids. The project will also leverage both HPC techniques and other metamodeling techniques to fast evaluate the grid impact from these resources with high demand uncertainties. This research provides the technical support to the electrification of transportation systems and decarbonization of the energy systems.