Pricing EV Charging Service with Demand Charge

Research Overview

Image Lightbox

Download Full Image

Graduate student and Resnick Fellow, Zachary Lee, is a member of a research team that has recently proposed a pricing scheme for large EV charging systems that provably captures the impact each charging session has on total cost for the system. The results of this study are summarized in the publication, "Pricing EV Charging Service with Demand Charge", and supported by simulations using real data collected from charging facilities at Caltech and JPL. The team's results suggest that costs can be allocated to individual charging sessions to precisely capture energy costs, demand charge, and congestion.

Scientific Achievement

We propose a pricing scheme for large EV charging systems which provably captures the impact each charging session has on total cost for the system.

Costs incurred by the proposed causal (online) scheduling algorithm and revenues from the proposed pricing scheme. All data is from real workloads taken from the Caltech and JPL charging networks. Note that revenue adequacy is not guaranteed since scheduling is sub-optimal, however losses are very small. Image Credit: Zachary J. Lee, John Z.F. Pang, Steven H. Low (2020) Pricing EV charging service with demand charge. Electric Power Systems Research, Volume 189. DOI:10.1016/j.epsr.2020.106694

Image Lightbox

Costs incurred by the proposed causal (online) scheduling algorithm and revenues from the proposed pricing scheme. All data is from real workloads taken from the Caltech and JPL charging networks. Note that revenue adequacy is not guaranteed since scheduling is sub-optimal, however losses are very small. Image Credit: Zachary J. Lee, John Z.F. Pang, Steven H. Low (2020) Pricing EV charging service with demand charge. Electric Power Systems Research, Volume 189. DOI:10.1016/j.epsr.2020.106694

Download Full Image

Significance and Impact

This work is a first step toward practical pricing schemes which encourage charging flexibility and reduce overall costs.

Flow of energy and revenues in a charging system. The Charging System Operator purchases energy from the Utility. The price paid for this energy is dependent on the amount of energy purchased, the time when the energy is used (time-of-use rates), and the peak power draw of the operator (demand charge). The Operator must decide how much energy to purchase at each time, which EVs to deliver that energy to, and how to allocate the costs to each EV. Our work uses model predictive control to solve the energy purchase/allocation problem and convex duality to solve the cost allocation problem. Image Credit: Z. Lee

Image Lightbox

Flow of energy and revenues in a charging system. The Charging System Operator purchases energy from the Utility. The price paid for this energy is dependent on the amount of energy purchased, the time when the energy is used (time-of-use rates), and the peak power draw of the operator (demand charge). The Operator must decide how much energy to purchase at each time, which EVs to deliver that energy to, and how to allocate the costs to each EV. Our work uses model predictive control to solve the energy purchase/allocation problem and convex duality to solve the cost allocation problem. Image Credit: Z. Lee

Download Full Image

Technical Details

• Prices are based on the dual of the minimum cost scheduling problem
• Revenue adequacy is guaranteed if charging is scheduled optimally
• Online model predictive control achieves within 10% of optimal costs.

Zachary J. Lee, John Z.F. Pang, Steven H. Low (2020) Pricing EV charging service with demand charge. Electric Power Systems Research, Volume 189. DOI:10.1016/j.epsr.2020.106694

Contact: Steven H. Low