Managing green Energy feed-in uncertainty for cost-efficient and reliable power system operation via AC CHance-constrained security constrained Optimal power flow (ECHO)

点击翻译按钮获取中文摘要

The ECHO project focuses on the security-constrained (or contingency-constrained) optimal power flow (SCOPF), aiming at improving the decision-making for day-ahead planning of power system operation, i.e., to ensure a cost-efficient and reliable power system scheduling for every hour of the next day. In addition, power systems are facing increasingly uncertain operation conditions, due to growing amounts of fluctuating renewable generation. The deterministic version of SCOPF, which fits only the most likely forecasted scenario, is not anymore suitable since it may lead to either sub-optimal or unreliable/risky operating conditions. To overcome this limitation, the ECHO project will develop a new SCOPF approach which manages the uncertainty via chance constrained optimization that enforces the set of constraints to satisfy a user-defined probability level. To this end, we resort to the inner-outer approximation method, which is a new method recently developed by the applicants in the area of chance constrained optimization. The very few existing works on this topic were based on the questionable AC grid model approximation as a linear direct current (DC) model (whose optimal control actions could be infeasible for real operation of the network as reactive power and bus voltages variations are ignored) and were applied to small systems. Unlike these previous works, the ECHO project is leveraged for the first time to the accurate fully nonlinear AC grid model for a medium size power systems (e.g. at country level). The project will comprehensively explore the aspects and implications of adopting chance constraints for security management. The ECHO project will demonstrate that a comprehensive SCOPF approach to security management of power system operation under uncertainty is indeed possible and this will lead to a better trade-off between optimality and reliability in the presence of a large penetration of renewable generation. We expect that this approach would be of interest to academia, utilities, and software developers, and it could support policy making of the network operators during the energy transition phase.