Energy optimisation and management of off-grid hybrid power supply systems

• Main Author: Tazvinga, Henerica
• Other Authors: Xia, Xiaohua
• Language: en
• Published: 2015
• Subjects: UCTD
• Online Access: http://hdl.handle.net/2263/50799; Tazvinga, H 2015, Energy optimisation and management of off-grid hybrid power supply systems, PhD Thesis, University of Pretoria, Pretoria, viewed yymmdd <http://hdl.handle.net/2263/50799>

Renewable energy-based hybrid systems have become attractive energy supply options in many countries because of global environmental concerns, access to electricity, and depleting and rising cost of fossil fuel resources. These systems are increasingly becoming popular solutions for electrification of off-grid locations where the grid extensions are difficult and/or uneconomical. The main challenge, however, is the design of an optimal energy management system that satisfies load demand considering the variable nature of renewable energy sources and changes in demand. This thesis aims to achieve an overall hybrid power management strategy that is capable of coordinating the power flows among the different energy sources as the management and control of energy distribution is a major problem associated with hybrid systems. The work deals with the optimization of the operational cost of the photovoltaic–diesel–battery/ photovoltaic– wind–diesel–battery power supply system from an energy efficiency perspective, as one of the key characteristics of energy efficiency is the search for optimality. The optimization models proposed in this work take into account the non-linearity of the operation costs associated with the photovoltaic-diesel-battery/ photovoltaic–wind–diesel–battery hybrid systems and this necessitates the use of quadratic programming while considering the weekday, weekend and seasonal changes in demand, and variations in the renewable output. The first part of this work considers the daily energy consumption variations for winter and summer weekdays and weekends in order to compare the corresponding fuel costs and evaluate the operational efficiency of the hybrid system. The second part minimizes fuel and battery wear costs, and finds the optimal power flow, taking into account PV power availability, battery bank state of charge and load power demand. The last part incorporates wind energy, and the energy dispatch model satisfies the load demand, taking into account the intermittent nature of the solar and wind energy sources and variations in demand. Model predictive control is applied to the optimal energy management strategy of the hybrid system which is a more practical approach to the energy management problem. The results show that the hybrid model achieves considerable fuel savings in both winter and summer seasons for days considered when compared to the case where a diesel generator satisfies the load on its own. The model predictive control model is shown to be superior to the open loop model owing to its ability to predict future system behavior and compute appropriate corrective control actions required to meet variations in demand, radiation and wind speed. The results of this work are important for remote area electrification, designers, performance analyzers, control agents, and decision makers who are faced with multiple objectives to make appropriate trade-offs, compromises or choices. === Hernubare energie-gebaseerde hibriede stelsels het ’n aantreklike energietoevoer-opsie in verskeie lande geword, as gevolg van globale omgewingskwessies, elektrisiteitstoegang, verminderende hulpbronne en stygende brandstofpryse. Hierdie stelsels neem toe in gewildheid as oplossings vir die elektrifisering van buite-netwerkgebiede, asook vir areas waar netwerkverlenging moeilik of onekonomies is. Die grootste uitdaging is die ontwerp van ’n optimale energiebestuursisteem wat die aanvraag bevredig binne die veranderlike aard van hernubare energiebronne en aanvraag. Hierdie tesis beoog om ’n algehele hibriede drywingsbestuurstrategie te ontwikkel wat daartoe in staat is om drywingsvloei met verskillende energiebronne te koördineer. Die werk ondersoek fotovoltaïse-diesel-battery en fotovoltaïese-wind-diesel-battery-energievervoerstelsels vanuit ’n energiedoeltreffendheidsperspektief, omdat een van die kenmerkende eienskappe van energiedoeltreffendheid, optimaliteit is. Die optimeringsmodelle wat voorgestel word, neem die nie-lineêre operasionele koste van bogenoemde kombinasies in ag, en dit noodsaak die gebruik van kwadratiese programmering wat weeksdae, naweke, seisoenale veranderinge, en die veranderlikheid van energiebronne en aanvraag in ag neem. Die eerste gedeelte van die werk ondersoek die daaglikse energieverbruik-veranderinge vir winter- en somersweeksdae en naweke ten einde die ooreenstemmende brandstofkostes en operasionele doeltreffendheid van die hibriede stelsel te evalueer. Die tweede gedeelte minimeer die brandstof-en batteryslytasiekoste, en vind die optimale energievloei, in ag genome die fotovoltaïese drywingsbeskikbaarheid, batterybank-ladingstoestand, en drywingsaanvraag. Die laaste gedeelte voeg windenergie by en die energieversendingsmodel bevredig die drywingsaanvraag, in ag genome die wisselvallige aard van son- en windenergie en aanvraagsveranderlikheid. Modelgebaseerde beheer word toegepas om die optimale energiebestuurstrategie van die hibriede stelsel op ’n praktiese vlak op te los. Die resultate wys dat die hibriede model merkbare brandstofbesparings behaal in beide winter- en somerseisoene in vergelyking met ’n dieselkragopwekker wat alleen werk. Die modelgebaseerde beheermodel oortref die ope-lusmodel as gevolg van eersgenoemde se vermoë om korrektiewe beheeraksies te gebruik om veranderinge in die aanvraag, straling, en windspoed te akkommodeer. Die resultate van hierdie werk is belangrik vir die elektrifisering van afgeleë gebiede, asook vir ontwerpers, werkverrigtingsanaliste, beheeragente, en besluitnemers wat gepaste kompromieë moet aangaan om veelvoudige doele te bevredig. === Thesis (PhD)--University of Pretoria, 2015. === tm2015 === Electrical, Electronic and Computer Engineering === PhD === Unrestricted