Convex Relaxations for Dual Market Energy Arbitrage

The spread of renewable energy generation has increased electricity price volatility, incentivizing battery energy storage systems (BESS) to perform energy arbitrage in multiple markets simultaneously.
However, when a BESS interfaces with the grid via a single point of common coupling, physical power flow must remain one-directional.
Optimizing trading schedules across both Day-Ahead (DA) and Intraday (ID) markets while ensuring a one-directional power flow induces a non-convex sign-consistency constraint.
While exact Mixed-Integer Quadratic Programming (MIQP) enforces this constraint using binary variables, it is computationally expensive, restrictive for rolling-horizon and near-real-time dispatch.
This thesis develops and evaluates a family of convex alternatives that encourage physical trade feasibility without binary indicators, formulating the dispatch problem as continuous Quadratic Programs (QPs), solvable in polynomial time.

M.Sc. student under the supervision of Prof. Yoash Levron.