Development of Innovative Robust Stability Enhancement Algorithms for Distribution Systems Containing Distributed Generators

Low voltage (LV) residential grids were traditionally designed without consideration of renewable energy impacts. In other words, voltage drop was one of the main considerations in network planning with respect to the projected peak demand. For such design planning, the fixed tap at the secondary side of the distribution transformer is usually set between 1.02-1.05 pu in order to compensate high voltage drop during peak demand periods. However, as the network is growing and penetration of Distribution Generation (DG) units such as photovoltaics (PV) is increasing, the operation of networks is changing and voltage profiles in the network have been adversely affected. Voltage rise in high PV penetrated locations during midday and voltage drop during peak demand periods in the evening are becoming the main power quality issues for residential networks. As a current solution for overvoltage problem, if bus voltage of any PV inverter passes a pre-set limit, it will trip which causes loss of renewable energy. Additionally, non-linear loads inject harmonic currents into the grid leading to bus voltage quality degradation. These problems are addressed in this research work. As a practical and economical solution, this thesis proposes voltage regulation and quality improvement algorithms for LV network operation. As part of these techniques, a Dynamic Voltage Restorer (DVR) as a series voltage compensator is designed and modelled for regulation and quality improvement in both high generation and high demand periods as well as for improvement of harmonic distortion in bus voltages. In the proposed method, a continuous and online control strategy adjusts the DVR output voltage to minimise bus voltage deviations in the entire network while reducing harmonic distortions. This maximises PV power injection by eliminating overvoltage trips and also provides voltage compensation for the peak demand voltage drop. In addition, it reduces the increased network demand caused by voltage rise at constant impedance loads which increases branch currents and consequently power losses. One of the advantages of this method is that no communication link is required as the proposed algorithm uses the locally measured data of compensator in order to calculate a compensation voltage by using a provided equivalent line impedance (ELI) value for the downstream network of DVR. Avoiding communication links is iiDevelopment of Innovative Robust Stability Enhancement Algorithms For Distribution Systems Containing Distributed Generators important as they are costly and prone to interruption, and introduce complexity into the network. As the tap for LV distribution transformer is normally fixed, the proposed algorithm can provide the same operation as On-Load Tap-Changer (OLTC) for LV grids but with the benefit of continuous and online regulation as well as harmonic distortion reduction. In order to achieve the best DVR performance, an optimisation algorithm is also proposed using the hybrid of heuristic-based Particle Swarm Optimisation (PSO) and Genetic Algorithm (GA). This algorithm uses clustered load level of buses during a year to reduce the optimisation computations for finding the best location and rating of DVR as well as the best tap setting for distribution transformer to minimise the investment cost while maximising voltage quality at all buses in the entire year. A sample LV network with PV generation is selected in this thesis to study the effectiveness of the proposed approach. Furthermore, converter control and synchronisation are critical for using the DVR in the network, and therefore, proper algorithms for these issues are also proposed in this thesis. In this respect, a control algorithm is designed for tracking of harmonic voltage references with minimum error and high disturbance rejection using Internal Model Principle (IMP) and feedforward. In addition, an algorithm is proposed to preserve the stability and performance of the control system by compensating control loop delay and adaptation of control parameters to plant parameter variations. Furthermore, the synchronisation required by DVR controller is achieved by developing algorithms based on the Synchronous Reference Frame Phase-Locked Loop (SRF-PLL) and Frequency-Locked Loop (FLL). Some algorithms are proposed for improvement of these methods for having better harmonic rejection and response time. Furthermore, these methods are extended by proposed improved algorithms to measure selective harmonic voltage angles and amplitudes, apart from those of fundamental voltage. These measured values are used by the compensation and DVR controller for the generation of compensating reference voltage for DVR. The proposed algorithms in this thesis are simulated using PSCAD/EMTDC and MATLAB platforms. Several simulation cases are considered, and the results are presented to verify the effectiveness and performance of the proposed algorithms as compared with the proposed methods in literature. For some of the proposed algorithms, Hardware-in-the-Loop (HIL) tests have also been performed in order to prove their feasibility. Development of Innovative Robust Stability Enhancement Algorithms For Distribution Systems Containing Distributed Generators iii