Effective Risk-limiting Carbon Emission Aware Economic Dispatch: An Algorithmic Perspective

Increasing public concern over climate change calls for high-level penetration of renewable energy sources into the future power grid, which makes the operation of the power grid fragile. One way to enhance the reliability of power grid operation is to equip each renewable generator with uncertainty management facilities such as conventional fast-responding generation units or storage systems. We identify a unified risk-limiting model for these diverse facilities. Specifically, in this paper, we consider two kinds of such facilities. The first one is the storage system, which has been traditionally utilized to enhance system reliability and reduce carbon emissions. We then propose the carbon allowance reserve (CAR), which, in a carbon emission aware economic dispatch, achieves the same goal as storage system does. The key to CAR is that it adopts conventional fast-responding generation units to conduct uncertainty management. We characterize the value of CAR by comparing the two kinds of facilities in the unified risk-limiting model. However, this is challenging because in a multi-period setting, solving the unified model alone is often intractable. Thus, we design an effective algorithm under mild assumptions on the renewable generation distributions. Next, we theoretically examine the robustness of the proposed algorithm, which highlights the practicability of the proposed algorithm. Numerical simulations further verify its effectiveness and provide comprehensive comparisons between the two kinds of uncertainty management facilities.