A bi-level approach to energy storage dispatch for congestion management in meshed power systems

Energy Storage Systems (ESS) are expected to become important tools congestion management in future power systems. While control of ESS for peak load reduction and firming of renewable energy in radial networks is well understood, congestion management in meshed power networks is more complicated as resolving a problem at one part of the network may negatively alter the power flows at other parts of the system, resulting in suboptimal market welfare. Merchandising surplus measures the value lost due to congestion in the electrical power systems. Optimal reduction in system congestion can hence be achieved by dispatching ESS to minimize the merchandizing surplus. However as this is derived from the shadow prices associated with the network constraints, the merchandising surplus is not known before the OPF is solved which prohibits use of a single level optimization. In this work a bi-level programming formulation is set up to directly model the impact of ESS dispatch on market price formation and the merchandising surplus. A framework is created for utility owned ESS operation that improves overall welfare of both generators and consumers in a pool-based electricity market by minimizing the shadow prices associated with congestion in the network. The effectiveness of the approach is evaluated by comparing it to an optimization problem that reduces system and nodal peak load