Intelligent Control and Operation of Distribution System

Research output: Book/ReportPh.D. thesisResearch

Abstract

Increased environmental concern and favorable government policies in the recent years have resulted in rapid growth of Renewable Energy Sources (RESs), such as Solar Photovoltaics (PVs) and Wind Turbine Generators. Despite creating several benefits, the high integration of the RESs in an electrical network imposes potential problems. In particular, limited dispatchability of the RESs on top of intermittent generation is creating key scientific challenges to the research community. On the other hand, increased trends of electrifying heating, transportation and gas sectors have resulted in new electrical loads, such as Electric Vehicles (EVs), Heat Pumps (HPs), and Electric Water Heaters (EWH), in the distribution network. In addition to the increase in power demand, those loads create high, rapid and random fluctuations in the demand, thereby impacting the power balancing scenario negatively. Conventional solutions to address the aforementioned issues include building excess generation or grid scale storage to compensate the power imbalances resulting from intermittent generations and fluctuating demands. Even though those approaches perform technically well, they are capital intensive to be implemented with the existing technology. One of the potential alternatives is to intelligently control the electrical loads to make them follow the intermittent generation. This not only enables the end consumers to get reliable and cheap electricity but also enables the utility to prevent huge investment in counterpart. Therefore, the theoretical foundation of this research work is based on a paradigm shift in ‘generation following demand’ to ‘demand following generation’ scenario. The primary aim of this research work is to develop intelligent control architecture, control strategies, and an adaptive protection methodology to ensure efficient control and operation of the future distribution networks. The major scientific challenge is thus to develop control models and strategies to coordinate responses from widely distributed controllable loads and local generations. Detailed models of key Smart Grid (SG) elements particularly distribution network, controllable loads, namely EV, HP, and EWH, and local generation, namely PV, are developed. The outcome of the projects is demonstrated mostly by simulations and partly by laboratory setups. The research outcome will not only serve as a reference for ongoing research in this direction but also benefit distribution system operators in the planning and development of the distribution network. The major contributions of this work are described in the following four stages: In the first stage, an intelligent Demand Response (DR) control architecture is developed for coordinating the key SG actors, namely consumers, network operators, aggregators, and electricity market entities. A key intent of the architecture is to facilitate market participation of residential consumers and prosumers. A Hierarchical Control Architecture (HCA) having primary, secondary, and tertiary control loops is developed to establish coordinated control of widely distributed loads and generations. Moreover, a heterogeneous communication network is integrated to the HCA for enabling control and communication within each control loop of the HCA. In particular, each control loop of the HCA is designed with specific control latency and time coordinated with the other loops to establish coordination among each other. The outcome of a power and communication co-simulation demonstrated that the proposed architecture effectively integrate responses from widely distributed loads and generations. Detailed models of the controllable loads, local generation, and distribution network are done in the second stage. Transportation loads, namely EV, and heating loads, namely HP and EWH, are modeled as controllable loads, whereas the solar PV is modeled as local generation. Control strategies are developed to realize various DR techniques, namely autonomous, voltage controlled, incentive based, and price based. The performance of the developed models and control strategies are demonstrated in a low voltage distribution network by utilizing the intelligent architecture developed in the first stage. Even though the developed models and control strategies are tested in a particular low voltage network, they are generic to apply to any network. The outcome from this part of the study enables to exploit demand flexibility from the consumers/prosumers for technically supporting the grid as well as for enabling to trade the flexibility in electricity markets. In the third stage, a scaled down SG testbed is developed and implemented for practical demonstration of developed DR models and control strategies. A LV test case network is scaled down to a 1 kVA testbed and is built in a laboratory. An optical fiber and TCP/IP based communication is integrated on the testbed to enable data communication and control. Moreover, a centralized and decentralized control approach is implemented for optimized EV charging coordination in the testbed. As practical demonstration of the SG is significantly lagging, the developed scaled down testbed based approach provides not only a novel approach for practical demonstration of the SG control but also a new research perspective in the SG field. More importantly, the proposed method provides an economic and riskless approach compared to the existing pilot project based SG developments. In the final stage, an adaptive overcurrent protection is developed for the future distribution system having high share of RESs and Active Network Management (ANM) activities. In the future grid, the protection is affected not only by bidirectional power flow due to RESs but also due to the ANM activities such as demand response, network reconfigurations etc. Therefore, unlike the conventional protection methodology where the protective settings are made static, an approach to adapt relay settings based on dynamic network topologies and power infeed is developed. In particular, the developed methodology integrates the protection and ANM such that the protection is ensured for every change in network operation mode and topologies. A two-stage protection strategy, whereby an offline proactive stage combined with online adaptive stage is designed to realize the intended protection approach. The proactive stage determines settings of the relays for every mode of operation using offline short circuit analysis and dispatches the settings to the respective relays. The adaptive stage in turn identifies the operating status of distributed energy resources and operating modes (grid-connected, islanded, network reconfiguration etc.) in real time to adapt/activate proper settings. The performance of the developed method is demonstrated in a medium voltage distribution network using real time digital simulator. The simulation results demonstrated that the proposed adaptive protection approach reliably discriminate the faults and establishes protection coordination in dynamic network topologies. Overall, the research outcomes provide an innovative approach to exploit demand flexibility from controllable loads and establish the integrated ANM and protection to ensure effective control and protection of the future distribution grids. This not only enables the end consumers to get reliable and cheap electricity but also enables the utility to prevent huge investment in counterpart. Moreover, distribution system operators can implement the findings of the projects in their operational stages to avoid grid bottlenecks and in the planning stages to avoid or delay the grid reinforcements.
Close

Details

Increased environmental concern and favorable government policies in the recent years have resulted in rapid growth of Renewable Energy Sources (RESs), such as Solar Photovoltaics (PVs) and Wind Turbine Generators. Despite creating several benefits, the high integration of the RESs in an electrical network imposes potential problems. In particular, limited dispatchability of the RESs on top of intermittent generation is creating key scientific challenges to the research community. On the other hand, increased trends of electrifying heating, transportation and gas sectors have resulted in new electrical loads, such as Electric Vehicles (EVs), Heat Pumps (HPs), and Electric Water Heaters (EWH), in the distribution network. In addition to the increase in power demand, those loads create high, rapid and random fluctuations in the demand, thereby impacting the power balancing scenario negatively. Conventional solutions to address the aforementioned issues include building excess generation or grid scale storage to compensate the power imbalances resulting from intermittent generations and fluctuating demands. Even though those approaches perform technically well, they are capital intensive to be implemented with the existing technology. One of the potential alternatives is to intelligently control the electrical loads to make them follow the intermittent generation. This not only enables the end consumers to get reliable and cheap electricity but also enables the utility to prevent huge investment in counterpart. Therefore, the theoretical foundation of this research work is based on a paradigm shift in ‘generation following demand’ to ‘demand following generation’ scenario. The primary aim of this research work is to develop intelligent control architecture, control strategies, and an adaptive protection methodology to ensure efficient control and operation of the future distribution networks. The major scientific challenge is thus to develop control models and strategies to coordinate responses from widely distributed controllable loads and local generations. Detailed models of key Smart Grid (SG) elements particularly distribution network, controllable loads, namely EV, HP, and EWH, and local generation, namely PV, are developed. The outcome of the projects is demonstrated mostly by simulations and partly by laboratory setups. The research outcome will not only serve as a reference for ongoing research in this direction but also benefit distribution system operators in the planning and development of the distribution network. The major contributions of this work are described in the following four stages: In the first stage, an intelligent Demand Response (DR) control architecture is developed for coordinating the key SG actors, namely consumers, network operators, aggregators, and electricity market entities. A key intent of the architecture is to facilitate market participation of residential consumers and prosumers. A Hierarchical Control Architecture (HCA) having primary, secondary, and tertiary control loops is developed to establish coordinated control of widely distributed loads and generations. Moreover, a heterogeneous communication network is integrated to the HCA for enabling control and communication within each control loop of the HCA. In particular, each control loop of the HCA is designed with specific control latency and time coordinated with the other loops to establish coordination among each other. The outcome of a power and communication co-simulation demonstrated that the proposed architecture effectively integrate responses from widely distributed loads and generations. Detailed models of the controllable loads, local generation, and distribution network are done in the second stage. Transportation loads, namely EV, and heating loads, namely HP and EWH, are modeled as controllable loads, whereas the solar PV is modeled as local generation. Control strategies are developed to realize various DR techniques, namely autonomous, voltage controlled, incentive based, and price based. The performance of the developed models and control strategies are demonstrated in a low voltage distribution network by utilizing the intelligent architecture developed in the first stage. Even though the developed models and control strategies are tested in a particular low voltage network, they are generic to apply to any network. The outcome from this part of the study enables to exploit demand flexibility from the consumers/prosumers for technically supporting the grid as well as for enabling to trade the flexibility in electricity markets. In the third stage, a scaled down SG testbed is developed and implemented for practical demonstration of developed DR models and control strategies. A LV test case network is scaled down to a 1 kVA testbed and is built in a laboratory. An optical fiber and TCP/IP based communication is integrated on the testbed to enable data communication and control. Moreover, a centralized and decentralized control approach is implemented for optimized EV charging coordination in the testbed. As practical demonstration of the SG is significantly lagging, the developed scaled down testbed based approach provides not only a novel approach for practical demonstration of the SG control but also a new research perspective in the SG field. More importantly, the proposed method provides an economic and riskless approach compared to the existing pilot project based SG developments. In the final stage, an adaptive overcurrent protection is developed for the future distribution system having high share of RESs and Active Network Management (ANM) activities. In the future grid, the protection is affected not only by bidirectional power flow due to RESs but also due to the ANM activities such as demand response, network reconfigurations etc. Therefore, unlike the conventional protection methodology where the protective settings are made static, an approach to adapt relay settings based on dynamic network topologies and power infeed is developed. In particular, the developed methodology integrates the protection and ANM such that the protection is ensured for every change in network operation mode and topologies. A two-stage protection strategy, whereby an offline proactive stage combined with online adaptive stage is designed to realize the intended protection approach. The proactive stage determines settings of the relays for every mode of operation using offline short circuit analysis and dispatches the settings to the respective relays. The adaptive stage in turn identifies the operating status of distributed energy resources and operating modes (grid-connected, islanded, network reconfiguration etc.) in real time to adapt/activate proper settings. The performance of the developed method is demonstrated in a medium voltage distribution network using real time digital simulator. The simulation results demonstrated that the proposed adaptive protection approach reliably discriminate the faults and establishes protection coordination in dynamic network topologies. Overall, the research outcomes provide an innovative approach to exploit demand flexibility from controllable loads and establish the integrated ANM and protection to ensure effective control and protection of the future distribution grids. This not only enables the end consumers to get reliable and cheap electricity but also enables the utility to prevent huge investment in counterpart. Moreover, distribution system operators can implement the findings of the projects in their operational stages to avoid grid bottlenecks and in the planning stages to avoid or delay the grid reinforcements.
Original languageEnglish
PublisherDepartment of Energy Technology, Aalborg University
Number of pages141
StatePublished - Oct 2015
Publication categoryResearch

Press/Media items

Download statistics

No data available
ID: 220285247