Provision of Power System Flexibility for Higher Penetration of Renewable Energy Sources

  • Baraa Mohandes

Student thesis: Doctoral Thesis


The ability of the electric power system to adjust its operating point by modulating its inputs and outputs is known as Power System Flexibility. A power system with a low flexibility characteristic is deemed insecure, and is unable to cope with very fast or unforeseen disruptions. In classical power systems, navigating the system safely through unpredictable events was a challenge limited to accommodating load variations, and rare equipment failures, and thus, conventional generation units were sufficient to provide the classical power system with the flexibility needed. In contrast, operators today are faced with larger uncertainty, caused by renewable energy sources (RES). The existing flexibility in classical power systems falls seriously short of accommodating the variability of RES. In fact, the lack of power system flexibility remains the main obstacle in the face of higher penetration of RES in the power grid. To achieve high penetration of RES in the power system, an operation framework that keeps a holistic view of the system is direly needed. This dissertation adopts this approach, utilizing all sources of flexibility to operate the system securely, and optimally. This thesis is comprised of four studies. In the first study, response time delay (q) is proposed as a new index of flexibility. This index is essential for an accurate and realistic description of the flexibility provided by distributed devices. In the second study, a smart DR contract is designed to empower DR in the residential sector as a potential source of flexibility. The proposed smart contract takes into consideration some concerns and characteristics of demand response participants which are not captured in the classical operation models of the power system. The two remaining parts study the optimum design and operation of a power system that features renewable energy sources, and comprises different sources of flexibility. In more detail, the third part optimizes the design of an islanded microgrid for an autonomous commercial building. The fourth study optimizes the design and operation of a utility-scale hybrid renewable energy plant. Demand response (DR) is explored in this thesis as a valuable and underutilized source of flexibility. DR differs from conventional central generation units in many ways. Because of that, it is widely accepted that deployment of DR requires changing many concepts and practices in power system operation. The components of power system flexibility were defined in 2010. However, the first study in this thesis reveals that the current taxonomy of power system flexibility indices is not sufficient to describe the flexibility of distributed energy devices. The time delay in the response of distributed devices is proposed as a fourth flexibility index, and a case study is run to highlight the effect of the new flexibility index. Later on, the second study in this thesis highlights the shortcomings of the classical unit-commitment problem formulation in addressing the characteristics and costs of DR units, and the problem formulation is updated accordingly. With extended UC formulation, a smart DR contract is designed to appeal specifically to residential customers. After presenting the contract terms, the financial aspects of the contract are optimized to reduce the generation cost, and also reduce RES spillage. In the third study, a commercial building is to be designed as an autonomous islanded microgrid. The building is powered by a rooftop PV plant, and features two types of loads with different potential to participate in load curtailment, and two types of Energy Storage (ES), with different characteristics. Three operation mechanisms for dispatching the microgrid components are proposed. The three mechanisms differ in their underlying logic and operation principles. Each mechanism regulates the RES unit's output, the charge/discharge processes of the two ES, and fulfills the requirements of two types of loads. The three mechanisms are tested under different weather scenarios and operation standards. The study highlights the role of thermal energy storage in facilitating demand response, and reducing load curtailment (i.e. minimizing customer discomfort). In the fourth study, a 'prevention is better than cure' philosophy is adopted in addressing the power system flexibility problem. Rather than calling for provision of more reserve and flexibility to the power system, the volatility and uncertainty of RES are addressed, partly, at the renewable energy plant side. This is achieved with a new design concept and a novel operation scheme for the renewable energy plant. The criteria of the hourly dispatch problem are suppressing extreme ramping events in the HREP output, and also adhering to the contracted output of the HREP in the hour-ahead market. The hourly dispatch problem employs a forecasting ensemble enhanced further by a machine-learning algorithm called the Multiplicative Weights Update method. The outcome of hourly operations at the end of each season are used, among other factors, to evaluate the quality of a candidate plant design, and the plant design is optimized accordingly.
Date of AwardMay 2020
Original languageAmerican English


  • Power System Flexibility
  • Renewable Energy
  • Energy Storage Systems
  • Demand Response
  • Microgrids

Cite this