Optimatization of hybrid renewable energy systems on isolated microgrids : a smart grid approach

Tese de doutoramento, Sistemas Sustentáveis de Energia, Universidade de Lisboa, Faculdade de Ciências, 2016

The energy systems of small isolated communities face great challenges related to their autonomy and resilience, when looking for a sustainable energy future. Hybrid renewable energy systems, composed from different technologies, partially or totally renewable, potentiates a growing security of supply for these isolated micro-communities. Moreover, with a smart grid approach, the possibility to reschedule part of the electricity load is seen as a promising opportunity to delay further investments on the grid’s power capacity, enabling a better grid management, through peak load control, but also to promote a more efficient use of endogenous resources, maximizing renewable penetration. To identify the micro-communities main energy challenges, a literature review was taken, reporting the design and implementation of isolated hybrid renewable energy systems. Since electricity and heat energy vectors can be, in part, assured by endogenous resources, a methodology to optimize demand response on isolated hybrid renewable energy systems was developed, using the electric backup of solar thermal systems for domestic hot water supply as flexible loads. This approach is intended to increase energy efficiency of the energy system, reducing grid operation costs and associated CO2 emissions. A model of the electric impact of the implementation of solar thermal systems and heat pumps for domestic hot water supply was developed and tested for the Corvo Island case study, a small and isolated microgrid, located in the mid-Atlantic with around 400 inhabitants and a diesel power plant. An impact of 60% on peak load and 7% on annual electricity demand was found. In order to tackle this significant impact in the grid, a model for optimizing the economic dispatch of the island was developed, testing multiple demand response approaches to the backup loads, from heuristics to genetic algorithms, having this last one performed best to control the peak load and minimize the operation costs. Nonetheless, there was the need to compare and validate the demand response optimization strategies of this developed model with other available modeling tools, which in the end presented similar results. As the pillar of this thesis is the optimization of hybrid renewable energy systems, the influence of the uncertainties associated to renewables forecast had to be studied, in particular its impact on the demand response scheduling. Wind uncertainties demonstrated to have a greater impact on the grid than the solar ones. Finally, the methodology developed incrementally along the thesis and validated in Corvo Island, was tested on different scales and types of isolated systems. It demonstrated to be especially suitable for small systems with less than 20 MW power installed and over 25% renewable generation, with mostly residential load profiles.

Fundação para a Ciência e a Tecnologia (FCT, SFRH/BD/62446/2009)