Optimum storage scheduling in distribution grids to minimize the ne load curve slope

Abstract

The demand for electrical power is increasing rapidly due to developments in modern technology. Consequently, the availability and affordability of electricity have become essential for users. To address this, power distributors are now exploring more cost-effective and accessible alternative energy sources. In addition, we are investing more in renewable energy sources, which release less carbon dioxide (CO2), as a result of the problems caused by climate change. However, because they depend on the weather, these renewable energy sources aren't always accessible. In order to guarantee a steady supply of energy, energy storage will be required. A thorough understanding of the problems associated with energy production and consumption is essential for managing this energy storage effectively and preventing the generation of excess energy. Renewable energy includes all forms of energy that are abundant and infinite, at least from a human perspective. Geothermal, biomass, hydraulic, wind, and solar energy are the five primary forms of renewable energy. In California, the Independent System Operator Corporation (ISO), which is in charge of most of the state's electrical grid, has seen a significant rise in solar photovoltaic energy. The ISO grid operator in California has multiple challenges as it looks to the future with more renewable energy. The Duck Curve problem, which is depicted by a widely used figure in the energy communities, is the most significant of these challenges. This graph illustrates the difference between the overall load that a public utility serves and how this load appears after some of it has been serviced by wind and solar energy. This graphical representation looks like a sitting duck. During the day, the demand for electricity from the grid dropped as solar generation grew (the duck belly). The surplus photovoltaic energy is the reason for this. The network then undergoes a peak in demand when the sun sets and people start going home in the evening (the duck's neck). It seems that the grid demand decreases during the day and then rises once more in the evening. These load patterns put traditional generating units under operational stress, threaten system stability, and necessitate flexible, fast-ramping resources to keep the power balance. These rapid fluctuations in net load frequently require expensive and ineffective ramping of peaking units in conventional grid operation. Insufficient utilization of clean energy investments and a reduction in renewable energy can potentially result from the duck curve if it is not adequately mitigated. Thus, new strategies are required to increase the flexibility of distribution systems, reduce ramping needs, and smooth the net load curve. Current strategies, including tariff reforms, advanced ancillary services, improved inverter capabilities, and enhanced control systems, primarily aim to flatten the load curve and tackle minimum system load issues. However, the hourly scheduling of battery energy storage systems remains largely overlooked, despite the introduction of various proposed solutions. This thesis addresses the duck curve problem by proposing an optimization-based battery dispatching strategy using Battery Energy Storage Systems (BESS) to flatten the net load curve during the daytime. The idea is to optimally control when and how much of a specific BESS unit will charge or discharge as a function of predicted system load and PV output profile. By storing surplus solar energy during the day and releasing it during high demand in the evening, BESS can significantly flatten the duck curve. A modified IEEE 33-bus radial distribution grid with six PV generators and six BESS units is considered in this study. The goal of the optimization problem is to minimize the net load curve's sum squared slope over 24 hours. The ramping issue is specifically taken into account by the slope technique, which minimizes the change between two successive hourly net load values. To solve the optimization problem, two nature-inspired metaheuristic algorithms— Particle Swarm Optimization (PSO) and Grey Wolf Optimizer (GWO)—were implemented and compared. Each algorithm was used to determine the hourly dispatch power of each BESS unit and the initial State of Charge (SoC) of each unit, which is crucial for maintaining energy balance over the day. The optimization was subject to several constraints, including SoC boundaries, charge/discharge power limits, SoC continuity (initial SoC equals final SoC), and energy conservation across time steps. The findings of this thesis offer valuable insights for distribution grid operators, planners, and researchers seeking to deploy energy storage as a flexible resource in high-renewable environments. The proposed method supports the reliable integration of solar energy, reduces the reliance on fast-ramping thermal units, and contributes to the broader goal of decarbonizing the electric power sector.