ISLANDING DETECTION AND VOLTAGE-FREQUENCY CONTROL OF MICROGRID

Abstract

Power demand has increased significantly over the years. The increasing power demand required the conventional power system to be modified to a modern power system involving various distributed generators (DG) closer to the load requirement. Mostly renewable energy based DGs are penetrated at the distribution level (low voltage level) with the main utility. The microgrid is an entity present at the low voltage (LV) network which has loads, DGs, storage units, and a defined structure for controllable functionalities that is well implemented by a command setup for monitoring and regulation of DGs and loads. When the microgrid is electrically disconnected from the remaining of power system, it is identified as Islanding. It can be intentional and unintentional. The microgrid operates in two different modes viz. grid-connected and islanded/off-grid. In the case of grid-connected operation, the leading utility, which has high inertia, predominantly controls the voltage and frequency (V-f) of a microgrid. However, isolated or islanded microgrids are more complex, requiring the more accurate operation of all associated systems and subsystems while controlling system volt age and frequency. The effect of generation and load deviations on an islanded microgrid’s performance is of bifold nature. On the one hand, it decreases overall system stability by incorporating more and more renewable energy sources, at the same time it lowers overall operation costs and encourages the use of green and clean energy. Furthermore, power elec tronics interfacing (PEI) is explored to connect the renewable energy sources (RESs) with the LV distribution network because the RESs have intermittent generation. The use of PEI with quick transient dynamic properties results in a low inertia microgrid system. Furthermore, the microgrid is subject to instability because of its intrinsic generation intermittency, making it vulnerable to any little deviation in power equilibrium, viz., the gap between supply and demand. With such an operational environment, maintaining V-f at the rated value for dif ferent buses within the microgrid under varied load conditions becomes challenging. There fore, from a control perspective, the islanded microgrid requires more robust approaches for system voltage control, mainly when the system’s dynamic generation and load model are uncertain. This thesis is focused on evolution, analysis and design of new passive islanding detec tion methods for microgrid, and voltage and frequency control of islanded microgrid. The research carried out in this work is divided into three proposed islanding detection techniques and controllers for islanded microgrid. All the proposed islanding detection techniques have been simulated in the RTDS/RSCAD environment, each one of which is briefly discussed below. The first proposed islanding detection method is based on event index value (EIV), which is derived from the superimposed components of sequence impedance with the considera tion of different operating configurations of the microgrid. Voltage and current signals are acquired at the terminal of distributed generators to compute the impedance. The technique successfully discriminates the islanding, and non-islanding events under different operating conditions, including zero power mismatch, and hence eliminates the non-detection zone and unwanted tripping due to the various types of non-islanding events. The detection was performed within a half-cycle for the different power mismatches. The technique has been illustrated on a 7-bus reconfigurable microgrid test system. A large number of non-islanding and islanding events have been simulated, considering the various operating conditions. The technique has also been verified on the modified IEEE34-bussystem. Theeffects of sampling frequency and noise have also been studied. At the end, the technique has been compared with the existing techniques reported in the literature, which proves the supremacy of the proposed technique. The second proposed islanding detection method is based on the superimposed angle of negative sequence impedance (SAONSI). The technique is designed with the presence of dif ferent types of RES in the system, and islanding has been determined considering each type of RESasthetargeted DG. The technique discriminates the islanding events from non-islanding events within a half-cycle even at the critical power mismatch and also restricts the undesir able tripping due to the several kinds of non-islanding events. The technique is also able to detect islanding under network reconfiguration. The technique has been demonstrated on a 7-bus microgrid test system, and also verified on a modified IEEE-34 bus system. Moreover, the comparison of the obtained results has been performed. To take forward, the third islanding detection technique is proposed that uses two new developed criteria; i) Discrimination factor of islanding (DFI) and ii) Superimposed angle of positive sequence impedance (∆φ1Z), for the discrimination between islanding and non islanding events. The technique utilizes the above two criteria in hybridization to complement each other. The efficacy of the proposed hybrid technique has been tested on a 7-bus micro grid test system for illustration and has also been verified on a modified IEEE 34 bus system. The proposed technique distinguishes the islanding from non-islanding events in the 9 ms even at the critical power imbalance; therefore, restricts the undesirable tripping due to the various kinds of non-islanding events. The islanding has been determined considering each type of DGs as targeted DG in various cases. A comparative assessment of the obtained re sults has been performed with the existing methods to show the dominance of the proposed technique. The effects of quality factor of load, sampling frequency and noise have also been studied. In the islanded microgrid, voltage and frequency control is a very challenging issue. Due to the generation uncertainties of solar PV, maintaining the nominal voltage and frequency for an islanded microgrid requires a robust control method. Therefore, for controlling the voltage and frequency, a robust V-f controller based on the decentralized fractional-order proportional-integral (FOPI) controller is proposed. The decentralized control relies on local parameter measurements and eliminates the need for expensive communication infrastruc ture. The performance of the designed controller has been evaluated under the uncertainties involved with source and loads. The performance of the designed controller scheme has been simulated under different operating conditions like variable solar irradiance, voltage set-point tracking with various amplitude modulation ratio (AMR), large-scale load pertur bations, faults etc. For this purpose, a solar PV and battery storage system-based islanded microgrid with an inverter supply to a three-phase network is considered. For maximum power extraction from the PV system, a Perturb & Observe (P&O) algorithm has been used. AFuzzy-PI controller has also been used to control the charging and discharging of the bat tery storage system. The amplitude modulation ratio is used as the reference signal, and a closed-loop dq control is applied. The FOPI controller provides improved response as com pared to the reported method.