Loading...
1 results
Search Results
Now showing 1 - 1 of 1
- Planning of power distribution systems with high penetration of renewable energy sources using stochastic optimizationPublication . Santos, Sérgio da Fonseca; Catalão, João Paulo da Silva; Cabrita, Carlos Manuel Pereira; Fitiwi, Desta ZahlayDriven by techno-economic and environmental factors, there is a global drive to integrate more distributed energy resources in power systems, particularly at the distribution level. These typically include smart-grid enabling technologies, such as distributed generation (DG), energy storage systems and demand-side management. Especially, the scale of DG sources (mainly renewables) integrated in many distribution networks is steadily increasing. This trend is more likely to continue in the years to come due to the advent of emerging solutions, which are expected to alleviate existing technical limitations and facilitate smooth integration of DGs. The favorable agreements of countries to limit greenhouse gas (GHG) emissions and mitigate climate change are also expected to accelerate the integration of renewable energy sources (RESs). However, the intermittent and volatile nature of most of these RESs (particularly, wind and solar) makes their integration in distribution networks a more challenging task. This is because such resources introduce significant operational variability and uncertainty to the system. Hence, the development of novel methodologies and innovative computational tools is crucial to realize an optimal and cost-efficient integration of such DGs, minimizing also their side effects. Novel methodologies and innovative computational tools are developed in this thesis that take into account the operational variability and uncertainty associated with the RES power generation, along with the integration of smart-grid enabling technologies. The developed methodologies and computational tools are tested in real-life power systems, as well as in standard test systems, demonstrating their computational proficiency when compared with the current state-of-the-art. Due to the inherent uncertainty and variability of RESs, stochastic programming is used in this thesis. Moreover, to ensure convergence and to use efficient off-the-shelf solvers, the problems addressed in this thesis are formulated using a mixed integer linear programming (MILP) approach.