Probabilistic Analysis for Optimal Power System Operation Using Flexible Smart Solutions

Abstract

Today’s power systems are rapidly changing. The low carbon technologies (e.g. wind and solar generation, electric vehicles and heat pumps) are increasingly being connected to electrical grids allowing zero fuel cost and less polluting network operation; on the other hand, these same techonologies cause greater intermittency and lower levels of system reliability. Furthermore, uncertain events such as adverse weather conditions that can cause network component failures lead to greater stress on the power system, as well as tighter security margins and greater operating costs. At present, many power utilities are seeing power system management as a challenge. To this end, smart energy solutions are being tested and applied as these can help mitigate operational and planning issues, while integrating the highest possible level of low carbon technologies. This thesis investigates how smart energy methodologies can help improve power system operation. Demand response, dynamic thermal ratings of overhead lines and FACTS devices are all considered as smart energy solutions that require further investigation. The modelling of these concepts is investigated and state-of-the-art methods are incorporated into the system reliability analysis. Assessment of power system operation is implemented using both probabilistic and deterministic criteria. Several contributions are presented in this thesis related to the field of reliability analysis for optimal power system operation. The first contribution of this research is a probabilistic framework for optimal demand response scheduling, which determines optimum ranking lists for both load reduction and load recovery based on reliability and economic risk metrics. The model also quantifies improvements in network performance, as well as customer profits received from participating in the demand response program for day ahead scheduling. The second contribution is the deployment of real time thermal ratings of overhead lines, which is applied in chronological analyses within both deterministic and probabilistic frameworks. The simulation results show that network-operating costs are lower under a probabilistic analysis than under a deterministic one. The third contribution is a probabilistic methodology to find the optimal deployment of wind energy sources, while minimizing wind curtailment to meet contractual obligations. The model gives the maximised hourly deployable wind capacities, minimised wind spillages, as well as reliability and operational cost indicators. The fourth contribution is a methodology for the optimal ranking of different FACTS devices based on their contribution to reducing both load and wind curtailments. Here, an additional investigation has been done, which determines the impact of FACTS and RTTRs on maximising the utilization of wind resources. Further contributions include improvements of simulation time for probabilistic analysis, implementation of a load-forecasting model for demand response loads, as well as the development of weather forecasting models for real time thermal ratings and wind generation output modelling.

Date of Award	31 Dec 2017
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Jovica Milanovic (Supervisor) & Victor Levi (Supervisor)