Design of a rotor-tied doubly fed induction generator

Olubamiwa, OreOluwa I. (2017-12)

Thesis (MEng)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: The severe e ects of climate change has seen a global push towards an increase in the use of renewable energy sources. Wind energy is a major renewable energy source and its usage in electricity generation is steadily rising. The doubly fed induction generator (DFIG) is the most commonly used generator in wind turbines due to its wide speed range of operation and easy power factor control implementation. DFIGs also use fractionally rated power converters which have lower converter losses compared to fully rated converters in induction and synchronous generators. Recently, a new DFIG topology (the rotor-tied DFIG) has been proposed. In this topology, the DFIG's rotor is connected to the grid and the stator to fractionally rated power converters. This topology has been shown to lead to higher efficiency as the higher frequency (grid frequency) is on the rotor core, which is typically smaller than the stator core, thus lower core losses. It was also suggested that designing DFIGs in this topology could lead to higher power densities. The topology requires no extra complexity in operations as a similar control system used with a conventional DFIG can be used. The steady state operation of the rotor-tied DFIG is first discussed and a new method of calculating slip for this topology is given. The proportions of the rotor and stator power to the input mechanical power to the generator at different slip values are illustrated. The transformer model equivalent circuit of rotor-tied DFIGs is also described. The purpose of this study is to design and optimize a low power rotor-tied DFIG. The design process is presented in a sequential manner from the calculation of the rotor size, to the rotor and stator winding parameters, then the slot and core dimensions. The obtained model is then evaluated with infinite element analysis (FEA) specific for rotor-tied DFIGs. The FEA is used to evaluate the power density, power factor and efficiency. The harmonic content in the model is also assessed. The design is then optimized to increase the power density and lower the harmonic content. A 5.5 kW rotor-tied DFIG is designed and its performance is evaluated using FEA. The optimization is executed using a response surface approximation of the FEA model with a genetic algorithm and this significantly reduces optimization time. The design of experiments for the response surface approximation is based on a combination of the latin hypercube sampling and composite sampling methods. Finally, a prototype is constructed and tested in a DFIG standalone mode. The tests are conducted in the sub-synchronous, synchronous, and super-synchronous regions of operation. The stator is excited with DC at synchronous speed, and slip frequency AC by the use of an AC drive at other speeds. The rotor is connected to a resistive load in all the tests. Tests results show that the prototype's efficiencies at synchronous and sub-synchronous speeds, for the rated stator current, are similar. The efficiencies at super-synchronous speeds are however lower with the same rated stator current due to the power rating of the AC drive used to excite the stator.

AFRIKAANSE OPSOMMING: Die ernstige gevolge van klimaatsverandering het tot 'n globale toename in die gebruik van hernubare energiebronne gelei. Windenergie vorm 'n belangrike deel van die energiebronne en die gebruik van wind vir kragopwekking styg aanhoudend. Die dubbel-gevoerde induksie generator (DFIG) is die mees algemeen gebruikte generator in wind turbines weens sy wye omvang van spoed werking en eenvoudige implementering van arbeidsfaktor beheer. DFIGs gebruik ook laer kapasiteits omsetters in vergelyking met gewone induksie en sinchrone generators. Onlangs is 'n nuwe DFIG topologie (rotorgebonde DFIG) voorgestel. In die topologie word die masjien se rotor aan die netwerk gekoppel en die stator aan lae kapasitiet omsetters. Dit is bewys dat die topologie tot 'n hoër masjien effektiwiteit lei, omdat die hoër frekwensie op die rotor is. Dit lei tot minder kernverliese, weens die feit dat die rotor kleiner is as die stator. Dit is ook voorgestel dat DFIGs wat so ontwerp word, tot hoër drywingsdigthede kan lei. Die bedryf van 'n rotorgebonde DFIG vereis ook nie ekstra kompleksitiet nie, aangesien konvensionele DFIG beheerstelsels toegepas kan word. Die bestendige toestand werking van die rotorgebonde DFIG word bespreek, waarna 'n nuwe metode vir die berekening van die glip vir die topologie gegee word. Die verhouding tussen die rotor en stator drywing en die megansiese intree drywing by verskillende glip waardes word geïllustreer. Die ekwivalente transformator model van die rotorgebonde DFIG word ook beskryf. Die doel van die navorsing is om 'n lae drywing rotorgebonde DFIG te ontwerp en te optimeer. Die ontwerp proses word op 'n opeenvolgende wyse aangebied, vanaf die berekening van die rotor grootte, die rotor- en statorwikkelingsparameters en uiteindelik die gleuf en kern dimensies. Die voorgestelde model word dan met behulp van eindige element analise (EEA) geëvalueer. Die eindige element analise word gebruik om die drywingsdigtheid, arbeidsfaktor en effektiwiteit te evalueer. Die harmoniese inhoud in die model word ook geassesseer. Die ontwerp word dan geoptimeer om die drywingsdigtheid te verhoog en die harmoniese inhoud te verlaag. 'n 5.5 kW rotorgebonde DFIG is ontwerp en die prestasie word geëvalueer met behulp van die EEA program. Die optimeering geskied met behulp van 'n generiese oppervlak reaksie benadering algoritme. Die metode verminder die optimeeringstyd aansienlik. Die ontwerp van die oppervlak reaksie benadering is 'n kombinasie van die latyn hiperkubus en saamgestelde monsternemingsmetodes. Ten slotte word 'n prototipe van die rotorgebonde DFIG gebou en getoets in 'n alleenstaande modus. Die toetse word uitgevoer vir die sub-sinchroon, sinchroon en supersinchroon bedryfstoestande. Die stator is opgewek met 'n GS-bron by sinchroon spoed en met glip frekwensie WS, by die ander bedryfstoestande. Die rotor is aan 'n weerstandslas gekoppel vir al die toetse. Die toetsresultate toon dat die prototipe se effektiwiteit by sub-sinchroon en sinchroon spoed vir ken stator stroom, is soortgelyk. Die effektiwiteit by super-sinchroon is egter laer as gevolg van die kendrywing van die WS omsetter wat gebruik is om die stator op te wek.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/102973
This item appears in the following collections: