Load Uncertainty-Aware Economic Dispatch

Uncertainty in energy demand poses significant obstacles to the efficient and secure operation of power systems. The purpose of this work is to provide a detailed analysis of uncertainty quantification in economic dispatch, with a specific focus on different scenario development methodologies. This research utilises various uncertainty quantification techniques such as bounded intervals, probability distributions, ellipsoidal sets, hybrid sets, and non-Gaussian distributions to produce scenarios that accurately reflect a wide range of load fluctuations. The optimal power output of generating units is determined for each scenario by solving the economic dispatch problem, which takes into account cost characteristics and demand constraints. The scheduling result P
1
-P
5
in (MW) and P
cost
in ($/h) of bounded intervals [465.14, 184.93, 278.50, 139.67, 188.28, 1.89e+07] and hybrid sets [448.65, 172.56, 265.58, 126.63, 173.80, 1.67e+07] have higher scheduled power and scheduling expenses relative to the base scenario [437.32, 164.35, 256.85, 117.81, 163.87, 1.54e+07] for period 2. Ellipsoidal sets schedule [425.61, 155.77, 247.79, 108.82, 153.80, 1.40e+07], probability distributions schedule [432.01, 160.33, 252.65, 113.68, 159.26, 1.48e+07] and non-Gaussian distributions schedule [433.95, 161.83, 254.19, 115.23, 160.98, 1.50e+07] are optimal for achieving efficient results in scheduled power and scheduling costs for period 2. The present study contributes to the domain of uncertainty quantification in power systems by facilitating the identification of suitable techniques for economic dispatch in real-world scenarios.