Robust optimal operation of future power systems under uncertainties

• Main Author: Lahanda Purage Mohasha Isuru Sampath
• Other Authors: Gooi Hoay Beng
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering::Electric power
• Online Access: https://hdl.handle.net/10356/137301
• Institution: Nanyang Technological University

As a result of the government incentives based on the policies on the decarbonization of the energy sector, renewable energy sources (RESs) are increasingly integrated into the power systems. This will introduce significant real-time variability in power injections due to the stochastic nature of these sources and the demand. As such, optimal reserve allocation is essential for economical and reliable operation of future power systems. Further, as many of these RESs are remote generation facilities, transmission congestion can be an upcoming issue faced by the system operators. Therefore, multi-terminal DC systems may be an essential part of future power systems due to their inherent capability to integrate RESs, to support voltage management, and to reduce transmission congestion via flexible power flow control. Furthermore, the concept of the deregulated markets promotes the proliferation of microgrids (MGs) in future power systems. The MGs comprise various types of flexible sources, which make them more promising for locally managing uncertainties. As such, the aforementioned challenges and the structural changes necessitate the development of efficient and scalable unit commitment (UC), economic dispatch, and optimal power flow (OPF) tools incorporating accurate optimization models for the analysis, planning, and operation of future power systems. This thesis is broadly divided into three parts. The first part of the thesis proposes a new trust-region based sequential linear programming algorithm to solve the AC-OPF problem. Therein, the OPF problem is linearized using the Taylor series and a trust-region constraint is used to control the validity of the linear model. A feasibility restoration phase is introduced to locate a feasible point. The algorithm converges satisfying the first-order optimality conditions for the AC-OPF problem. The case studies illustrate the efficiency and computational performance of the proposed approach based on the standard IEEE and large-scale Polish systems. In the second part of the thesis, a two-stage robust optimization (RO) framework is proposed for the joint scheduling of power dispatch and reserves for hybrid AC/DC transmission grids (HTGs). The optimal day-ahead schedule, including the commitment, power dispatch, and reserve allocations is computed subject to the feasibility of nonlinear AC and DC network constraints (NCs) while respecting the nonanticipativity under a specified set of scenarios. The proposed decomposition approach renders network security evaluations into standard OPF problems. As such, the scalability issues observed with existing methods are resolved. Further, a second-order cone relaxation is applied to nonlinear AC and DC NCs to improve the computational tractability and the globally optimal solvability. In simulations, the proposed RO framework is illustrated using a recently proposed HTG architecture with a specific topology. The results demonstrate the effective utilization of the grid capacity to accommodate more demand and RESs. This HTG architecture induces exactness of the convex relaxation, supporting the validity of the optimal day-ahead schedule. The efficiency and computational performance of the proposed approach are compared to the existing literature and the robustness is verified using out-of-sample analyses. Recently, the concept of multi-microgrid (MMG) systems has come into prominence due to the economic benefits accrued through the sharing of resources between the constituent MGs. The final part of the thesis proposes an RO framework to determine the day-ahead schedule of an MMG system. Unlike the existing works in the literature, the proposed RO framework preserves the nonanticipativity in reserve scheduling. The proposed RO framework also includes a cooperative bidding-based trading scheme to facilitate the sharing of power and reserves between the constituent MGs of the MMG system while respecting privacy concerns. The results highlight the economic benefits obtained through the sharing of resources between the constituent MGs in an MMG system. Furthermore, the proposed nonanticipative RO framework outperformed the existing RO frameworks in the literature in terms of robustness for MMG systems.