A Data-Driven Scenario Screening Method Considering Peak Shaving and Reserve Contribution of Power Systems with High Penetration Renewable Energy

Renewable generations are one of the most significant measures to achieve the goal of carbon emission peak and carbon neutrality. However, high penetration renewable energy results in massive scenarios and pressures heavily on the operation of power systems, due to its randomness and intermittence. This paper proposes a data-driven scenario screening (DDSS) method to figure out crucial scenarios which may lead to insufficiency of peaking shaving and reserve contribution of power systems and release the heavy burden of massive scenarios. First of all, massive long-term operation scenarios including loads, network topologies, and outputs of conventional units and renewable generations are provided from the control centers of power grids or generated by Monte Carlo method. Three types of scenario screening rules are then summarized into Types A, B, and C. Types A, B, and C are all considered leading to insufficiency of peaking shaving and reserve contribution. Types A and B present scenarios with demand peak and demand valley respectively, while Type C presents scenarios with rapid rise/decline in demand. Each type consists of four detailed screening rules. Predetermined number of crucial scenarios can be obtained in descending order by each rule. Furthermore, a flow is illustrated for subtle scenario screening. The number of crucial scenarios will be double, once potential operating risks such as wind/solar power limitation and load shedding are found. The scenario screening comes to an end if and only if no more potential operating risk exists. Finally, crucial scenarios are verified by calculating a series of evaluation indexes such as renewable energy penetration, unit ramp rate, and wind/solar power limitation. Case studies are implemented on a provincial power grid of the China Southern Power Grid to validate the efficiency of scenario screening method.