Control techniques for power quality improvement in grid-connected dfig-based wind turbines

In the past 20 years wind energy has proved itself as a viable and increasingly economic means of generating electricity. However the qualitative jump in this technology has arrived together with new developments in the power processing field, and nowadaysalmost all wind turbines (WTs) are totally or partially controlled by means of power converters. At the beginning of the last decade, most installed WTs consisted of standard squirrel cage induction generators (SCIGs) that were directly connected to the grid. This kind of system, that still represent a 30% of the installed wind power (WP), operate at almost fixed speed. As a consequence the power injection to the grid can vary only a + 2 % with respect the nominal power. This feature played against the efficiency of WP technology, as this narrow range of operation didn't permit to recover the costs of wind farms in a short term. The solution for that problem arrived with variable speed WTs, that were able to generate power in a wider slip range. In this group the DFIG-based WTs have played an important role. This WTs are able to operate in the synchronous and the subsynchronous area, and normally their operation range permits to inject between 70% and 130% of the generator's nominal power to the grid. Despite the fact that other variable speed WTs are able to manage the full power of the generator, their cost is higher if compared with the DFIG proposal, hence this kind of systems have become very popular due to its good trade-off between costs and efficiency, and, nowadays, they constitute the 50% of the installed WTs in the world. At the present time several countries in Europe as Spain, Germany or Denmark cover more than 6% of their electricity demand by wind power. The already installed 80GW of wind power (WP) in Europe are expected to grow until 300 GW in 2030. If these future projections are finally achieved the installed wind farms in 2030 would be providing around the 23% of the electricity demanded by the grid. The rapid expansion of renewable energysources (RES), specially WP, has made necessary to redesign the existing grid code requirements (GCR). Now that power plants based on RES are becoming important players, the transmission system operators (TSOs) demand more reliability for technologies such as photovoltaic (PV) and WP. Therefore the standards, regarding operation and maintenance of grid connected RES power plants, are becoming more restrictive. The fault ride through (FRT) requirements for wind power plants have gained a great importance in the latest grid codes. A major penetration in the grid implies that WTs cannot get disconnected under fault conditions, as they used to do in the past for safety reasons. By means of the low voltage ridethrough (LVRT) graphs, each TSO describe how wind turbines (WTs) should behave when a transient voltage fault occurs. These curves determine the fault boundaries among the ones the system should remain connected to the network, in function of the fault depth and time duration, as well as their operationmode in each situation. Some standards, as the German and Spanish ones, have gone even further, specifying the amount of reactive power that WTs should deliver to the grid, during and after the fault. The most extended connection topology of DFIG WT's is based on the grid connection to the network by means of two power converters in back to back, one connected to the rotor of the machine and another linked to the network. The shaft of the generator gets the mechanical power from the wind by means of the secondary of a gearbox, that is in charge of reducing the input torque while boosting the speed of the system, as normally these generators do not have a high number of poles. In this topology the rotor side converter is responsible of controlling the amount of active and reactive power that is being injected to the grid, while the grid side converter is in charge of maintaining the DC bus voltage. This control concept can be considered as the 'classical control' of DFIG-WTs. The operation of DFIG-WTs is satisfactory under grid balanced conditions, however its performance is not so good when there is a voltage fault or unbalances near the point of common coupling (PPC). As the stator of the generator is directly connected to the grid any voltage change in the PCC voltage affects the behaviour of the machine, as the magnetic variables are closely linked to this parameter. Due to this coupling, any voltage drop in the PCC produces a high current peak that is produced by the machine in order to maintain the magnetic flux in the generator. These current peaks, that can be seven times higher than the nominal values, pass through the stator winding and are transmitted to the rotor. This forces the disconnection of the rotor side converter, as the high inrush currentswould destroy it. Obviously this operation mode prevents DFIG-WT's from supporting the grid during fault conditions, something that could be reasonable if this technology was not called to be an important electrical power source in the near future. Nowadays, fulfilling the new FRT requirements is an important challenge for wind power industry. In modern WTs the symbiosis between electrical machines and power electronics is permitting to achieve noticeable headways regarding that topic. The installation of Dynamic Voltage Restorers (DVRs) and Distribution Static Synchronous Compensators (D-STATCOMs) have contributed to improve the FRT response, as both devices permit boosting the grid voltage level if there is a voltage dip. However the connection of dedicated facilities to assist the generators in these conditions increases the costs of wind farms. For this reason the current trends in WP industry are focused on designing new control algorithms for the power converters of each WT that would improve the response of the system under faulty or unbalanced conditions. In the particular case of DFIG-WTs the rotor side and front end converters are being reviewed by many authors. Nevertheless there is not yet an established control strategy for controlling DFIG-WTs in these scenarios. This thesis is devoted to enhance the performance of DFIG-WTs in steady state conditions but on transient situations as well. This study will be specially focused on improving the FRT of these WTs, as well as establishing the operation limits of the different strategies that will be implemented. In this work not only the problematic of voltage sags will be treated, but other as harmonics and flicker will be reviewed as well. The solutions for those problems will consider new control algorithms for this kind for DFIG-WT's, as the classical one does not offer guarantees in generic situations. Techniques such as adaptive resonant controllers will be discussed and their performance will be tested and analyzed by means of simulations and experimental tests. The main objective will be to expand the functionalities of the existing power converter that are installed in a DFIG-WTs with back to back converters, inorder to achieve a major efficiency of the whole system. While proposing these new functionalities the operation limits of the power converters and the electrical machine will be continuously considered, in order to get the optimal trade-off between costs and performance of the system.