Stability Improvement of MMC-Based Multiterminal HVDC Grids

Abstract

The 100% renewable European smart grid cannot be implemented without reliance on wind energy from the offshore wind farms in the North Sea. High-voltage direct current (HVDC) cables are generally used to transfer power from long-distance offshore wind farms to onshore grids at relatively low costs and energy loss. According to European expansion planning scenarios, the interconnection and expansion of the existing and newly built offshore point-to-point HVDC grids in the North Sea are inevitable. Implementation of the multiterminal HVDC grid configuration demands control measures to ensure grid stability since such grids were not initially designed to be multiterminal. Hence, dynamic interactions, either on the AC- or DC-side of the multiterminal grid, can be a source of detrimental oscillations and propagate to the entire network due to the droop control action of the HVDC converters, resulting in stability deterioration, faults, and subsequent blackouts with considerable damages and costs. This thesis aims to present an optimal analytical methodology to improve the stability of the multiterminal HVDC grids, focusing on offshore wind applications. The analysis is performed in MMC-based HVDC grids, the preferred offshore technology, to account for the converter’s control and dynamic challenges and opportunities originating from the arm capacitor energies. First, a centralized optimal linear feedback controller is introduced to ensure network stability under the worst-case perturbation scenario. The main aim of the centralized optimal controller is to minimize the oscillations caused by the poorly damped modes under the worst-case perturbation scenario while considering the constraints on the control inputs and state variables. The optimal controller can be found by running the optimization problem formulation once without analyzing the time-consuming dynamic and transient time-domain simulations for different scenarios. Additionally, there will be no need for repetitive tuning of the grid controllers under the interconnection and expansion procedure since the optimal controller alone can guarantee stability. The performance of the centralized optimal controller in minimizing DC voltage oscillations is investigated via small-signal eigenvalue stability analysis and time-domain simulations. DC voltage stability is a sign of power balance in multiterminal HVDC grids, and DC voltage oscillations can propagate to connected networks due to the droop control action and lead to voltage instability and faults, ultimately resulting in blackouts with significant costs. This thesis shows that the centralized optimal controller can ensure DC voltage stability under the worst-case perturbation scenario when implemented either in the presence or absence of the droop control gain, i.e., it can optimally readjust or substitute the droop gain in a multiterminal configuration to improve DC voltage stability. Second, the centralized optimal controller is replaced by the decentralized one to create the desired sparsity pattern such that there is no need for long-distance data communication between converter stations, resulting in a more reliable and stable solution. In the decentralized problem formulation, the constraints on the initial state variables’ perturbations and control inputs can be decoupled, which is physically more practical and sensible. This thesis investigates the applicability of the decentralized optimal controller in reducing oscillations caused by the poorly damped modes under the worst-case perturbation scenario and decoupled grid control inputs’ and state variables’ constraints in MMC-based multiterminal hybrid AC/DC grids. It is shown via the eigenvalue stability analysis and time-domain simulations that the proposed decentralized optimal controller can efficiently enhance the grid stability margins and reduce the oscillations, not only under the worst-case perturbation scenario but also under other critical fault conditions. Furthermore, the controller’s superior performance is validated in the absence or presence of the power system stabilizer (PSS) and MMC droop controller under small and large disturbances, and its robustness is assessed against the uncertainties in control parameters, grid parameters, and operating conditions under the disturbances on the AC- and DC-side of the grid. Finally, this thesis introduces a stability (oscillation) index that can be integrated into the transmission expansion planning (TEP) problems as a further decisionsupport criterion in the preliminary stages of expansion planning in multiterminal HVDC grids. This index can identify the HVDC link placement among several options with minimum oscillations on critical and poorly damped state variables under the worst-case perturbation scenario. For instance, the DC voltage oscillation index is applied as a potential decision-support criterion for placing a new HVDC link between two independent point-to-point MMC-based offshore HVDC grids while considering the wind intermittency effect on grid operating conditions.