Stochastic Scheduling of Distributed Microgrids with High Penetration Levels of Solar PV and Run of River Hydro Generation

• Main Author: Anan Ayoub, Mustafa Dweekat
• Format: Thesis
• Language: English
• Published: Universiti Malaysia Sarawak (UNIMAS) 2021
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://ir.unimas.my/id/eprint/34954/1/Anan%20Ayoub%20Mustafa%20Dweekat%20ft.pdf; http://ir.unimas.my/id/eprint/34954/
• Institution: Universiti Malaysia Sarawak

The continuous consumption of fossil fuels in electrical power generation due to increasing load demand contributed in causing dire environmental impacts. This has directed to developing sustainable hybrid microgrids consisting of multiple distributed generations (DGs). DGs include renewable energy resources (RES) such as Run of River (RoR) small-hydro and solar Photovoltaic (PV) generations along with conventional generation. Integrated large shares of RES in hybrid microgrids help in minimizing emissions and operation cost of the power system to achieve an environment-friendly ecosystem. RoR small-hydro and solar PV generation vary depending on variable solar irradiance and water discharge respectively. Solar irradiance and water discharge vary during the day depending on the weather, such as sunny hours, clouds, rainfall and others. Intermittency and uncertainty nature of these energy sources in a hybrid microgrid makes energy dispatch a challenging problem. Therefore, this thesis presents an optimal energy generation scheduling to minimize the total power system operation cost. A mathematical model of RoR small-hydro and solar PV is proposed in this thesis to formulate a stochastic scheduling model. The model is solved using the chance constrained method and fitted into a Mixed Integer Linear Programming (MILP) approach. Solar irradiance and water discharge are modeled as stochastic random variables and simulated using the Autoregressive Moving Average (ARMA) method. The uncertainty of solar irradiance and water discharge are modeled using Probability Density Functions (PDFs). Chance constrained stochastic optimization method is employed to guarantee that power balance probability is equal to or greater than a predefined confidence level. Finally, the applicability of the proposed hybrid microgrid generation scheduling model is demonstrated through numerical simulations over 24-hour and 168-hour dispatch. The model is tested on four standard testbed systems IEEE 6-bus test system, IEEE 14-bus modified test system, IEEE 24-bus Reliability Test System (RTS) and 69-bus Microgrid Test System. Results show that increasing the penetration level of solar PV and RoR small-hydro generations in the microgrid will minimize total operation cost. Also, modeling solar PV and RoR small-hydro as stochastic RES is reflecting the intermittency nature of solar irradiance and water discharge when compared with capacity credit modeling. Using multiple RES in a hybrid microgrid makes the microgrid more dispatchable and increases power system reliability with lower operation cost. Finally, the right choice of the PDF that best fit the solar irradiance and water discharge data is affecting the optimal solution of the generation scheduling problem.