Active control of smart grids
- Autores: Brenda Carolina Rojas Delgado
- Directores de la Tesis: Hortensia Amaris (dir. tes.), Mónica Alonso Martínez (codir. tes.)
- Lectura: En la Universidad Carlos III de Madrid ( España ) en 2019
- Idioma: español
- Tribunal Calificador de la Tesis: Irina Temiz (presid.), Carlos Álvarez Ortega (secret.), Gregorio Ignacio López López (voc.)
- Programa de doctorado: Programa de Doctorado en Ingeniería Eléctrica, Electrónica y Automática por la Universidad Carlos III de Madrid
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
According to the International Energy Agency, power generation from renewable energy sources (RES) has increased worldwide from 15% in 2004 to 25% in 2017. In regard to solar energy, the decline in the cost of photovoltaic (PV) elements has allowed continuous growth in the photovoltaic (PV) market, which has reached a 34% growth in power generation by the end of 2017. The greatest growth occurred in China followed by the United States and Japan. At the European level, the data published by Global Market Outlook For Solar Power for the 2018–2022 period indicate that Germany has experienced the greatest growth in installed photovoltaic (PV) power. In 2017, Germany experienced a 11% growth, while Italy’s market installed capacity increased by 5% and Spain’s increased by 1% compared with 2016 data.
In 2016 the World Energy council in 2016 states that marine generation capacity worldwide is 32 PWh/year, which points to the possibility that wave energy is capable of becoming more energy competitive when compared to other forms of renewable generation in the future. Nevertheless, in spite of its viability, costs are still high. The installation and operation of grid-connected wave energy plants still need to overcome important knowledge and technological barriers to meet the criteria in economy of scale. Apart from the problems derived from generation unit costs, wave energy sources also faces a variety of environmental, infrastructural, and socioeconomic obstacles, such as uncertainty of grid-connection infrastructures, obstacles related with the integration in the electricity market, power quality problems, supply chains growth, and environmental impact mitigation.
Moreover, growth expectations for some high-penetration scenarios have reached new targets of up to the 80–100% renewable energy sources at the global level. These growth expectations are very promising, but they also reveal the difficulties faced by operators of electricity networks in balancing demand and generation. Finding this energy-mix balance with a high penetration of fluctuating renewable energies that are stochastic, highly variable and scattered throughout the network means that the challenges of maintaining the safety and reliability of electrical systems will become more critical in coming years.
In most cases, renewable energy sources are connected to distribution networks that include both low-voltage (LV) urban networks (e.g., photovoltaic (PV) panels installed on house roofs) and medium-voltage (MV) distribution networks (e.g., solar or wind farms). In many countries, energy regulation policies have focused on injecting all of the energy generated by renewable sources into the power networks. These strategies can be considered acceptable as long as the penetration of renewable energy is maintained at low penetration levels because they do not generate enough significant technical problems in the electrical networks that could limit its expansion. However, as the level of renewable energy penetration increases the operation of electricity networks may be affected, technically, thus making it necessary to disconnect some renewable energy units (energy curtailment) from the network.
In some situations, installing sources of renewable origin onto distribution grids creates network congestions. This is mainly due to the fact that those energy sources are stochastic in nature and that distribution networks are not prepared enough to manage the bidirectional flow of this energy. In some EU-countries, grid codes are being defined so that the distributed generation (DG) can supply complimentary services to the grid at the connecting point when the distributed generation is connected to the distribution grid. Currently, there is a need of implementing tools that help distribution system operators (DSO) to manage the distribution grids. These grids are beneficial for distributed generation integration, electric vehicles connection, and the participation of the energy end-users, as the grids help to maintain both power efficiency (i.e., electric losses reduction) and the electricity supply quality.
The aforementioned shows that grid operators do not possess the right tools that allow them to solve congestions in their grids, even though they do have distributed resources that allows them to solve these congestions. Moreover, they do not have the proper tools to coordinate with the transmission systems operators and they also cannot provide the assurance of offering distribution networks’ flexibility services in order to, eventually, solve congestion problems in higher voltage power networks. The aim of this doctoral thesis is to develop methods and procedures that can improve the voltage regulation of medium voltage (MV) and low voltage (LV) distribution networks that have a high penetration of distributed energy resources. The developed methods will allow distribution network operators to manage their distributed energy resources (distributed generation, storage systems, electric vehicles) in almost real-time; in this way, the distribution grid can regulate both voltages at low and medium levels, and it can offer a complimentary service of voltage control to the transmission network. Distribution power networks currently face a new paradigm dealing with the connection of new agents, such as distributed generation, storage units, flexible loads (e.g., electric vehicles) and prosumers. Along with these distributed energy generation and customers demand systems, distribution power networks have been enhanced with communication infrastructures for the past few years, which allows them to know their real-time power production and energy consumption, as well as to establish the required power flow bi-directionality over the grid. Because of this evolution, distribution grids have become active elements of the system (i.e., active distribution systems) which can manage the distributed controllable elements available in the system so that the whole network performance can be ameliorated. It has to be emphasized that the raise of distributed generation units connected to medium voltage and low voltage grids can cause voltage and network congestion problems. The traditional way of dealing with distributed generation sources known as “fit and forget” it is no longer valid, which makes it necessary to implement proper management of the controllable units connected at these grids. One of the main problems that literature highlights is the one that is related to over-voltages present in low voltage grids when the distributed generation surpasses the demand. This problem can propagate to the medium voltage grid, and under some occasions it can also extend to the transmission power network. Therefore, it is necessary to develop control schemes for power grids that can, coordinately, manage both medium voltage and low voltage grids, and can also, locally, manage distributed generation, distributed storage, and distributed flexible loads units. To date, proposed voltage control schemes have centered on grids at medium voltage levels. Traditionally, the voltage control exerted onto distribution grids have been carried out by means of regulating the on-load tap changers (OLTC), which are located at the power transformers of the primary substations. Alternatively, this can also be carried out by locally controlling the reactive power compensation devices and capacitor banks (CBs) connected at distribution networks. However, these techniques have demonstrated to be inadequate for abnormal performance situations or even when high penetration of renewable energy sources exists. With the above in mind, this doctoral thesis develops the idea of using voltage control techniques among all the controllable devices present in the grid in order to diminish power losses and to maintain the voltages of these within the limits established in the range of normal functioning. The undertaken techniques will also allow for both the energy transference and the information exchange among the grids to be at the different voltage levels. Framed around this global objective, the proposed specific objectives are as follows: • To solve voltage regulation and bidirectional power flows problems in commercially exploited distribution grids connected to transmission power networks.
• To improve the network observability in medium voltage and low voltage grids using the real time data and information measured and gathered by the advanced metering infrastructure deployed at distribution smart grids. • To develop optimization algorithms for the coordinated control of controllable devices present in the distribution grid (i.e., medium voltage and low voltage) at different temporary operative ranges and also for several situations of high penetration of renewable energy sources. • To define the energy and information exchange among the different medium voltage and low voltage levels so that the low voltage grid can offer a complimentary voltage service to the medium voltage grid.
• To develop control models and techniques for wave power plants that can be grid-connected at low voltage level.
To tackle the issues, this thesis proposes a coordinated voltage control architecture that can allow the low voltage grid to provide support to the medium voltage grid and can make the control in real-time of medium voltage and low voltage grids possible. In this case, the thesis aims to develop the following tasks: • Defining the coordinated control structure, such as the relationship among levels, border knots definition, and the interchange among levels (e.g., power and information) • Defining the exchanges of information and signal among the levels of the coordinated control • Designing the control algorithms and comparing them to classic local decentralized and distributed control techniques • Validating the control algorithms with real distribution grids (i.e., medium voltage and low voltage) that include wind farms, photovoltaic (PV) units, and wave power sources The research suggested in this thesis focuses on the flexibility services that is offered by the distributed energy resources (DERs) connected to distribution networks. Thus, it would be possible to control the distributed energy resources (i.e.: photovoltaic (PV) energy sources and off-shore wave power plants) that are connected to a low voltage network, as well as the distributed energy resources (i.e.: solar farms and wind power plants) connected to a medium voltage network. A number of important contributions were made in this thesis: Firstly, the thesis developed a wave power plant model that is connected to the low voltage grid by means of a submarine cable. This comprised electric generators, a flywheel energy storage system, submarine cables, a substation’s power transformer, and electronic converters, the latter of which are in charge of controlling and smoothening the power injected to the grid so that it would be flattened as much as possible. The modeling and control were validated with the experimental wave power plant located at the Lysekil Wave Energy Site developed by Uppsala University in Sweden. To do this, data gathered from the real emplacement was later utilized to implement the developed control algorithms.
In addition, the thesis developed a coordinated voltage control algorithm for medium voltage and low voltage grids that were made of two optimization algorithms. In the medium voltage grid, the study developed an optimization algorithm that can determine the performance set points of the grid’s controllable units and can optimize the active/reactive power flow at each one of the border knots with the low voltage grid (medium voltage - low voltage substation). To do this, the algorithm disposes of the forecasted hourly demand planning of the clients connected to the medium voltage grid, as well as the prevision of the demand aggregated to the substation’s border nodes (i.e., medium voltage - low voltage). Likewise, the algorithm also disposes of both the day-ahead hourly forecast associated with the renewable generation connected to the medium voltage grid, and the renewable generation aggregated to the border nodes (i.e., medium voltage - low voltage). The optimization system of the medium voltage grid is non-linear nor-convex.
At the medium voltage level, it is defined as an objective function the minimization of the power curtailment of the renewable energy sources connected to the medium voltage grid, considering the necessary equality and inequality constraints to maintain grid stability in all its nodes. In the low voltage grid, the study used an optimization algorithm that aims to minimize the difference among the received active/reactive power set points at the node serving as a frontier between medium voltage and low voltage grids (i.e., medium voltage - low voltage substation). This is done with respect to the injected/absorbed power by a low voltage grid for each performance condition.
The controllable units belonging to the low voltage network are the photovoltaic (PV) installations, which offer the flexibility required to regulate their active and reactive power according to the performance curves corresponding to each unit. The reason for this is to provide a complimentary service of voltage control to the medium voltage grid. The optimization algorithm of the low voltage grid is non-linear nor-convex. Lastly, this study has provided a definition of the information and data to be exchanged between the low voltage and medium voltage grids so that low voltage photovoltaic (PV) units can offer a complimentary voltage service to the medium voltage grid. This aspect is one of the most critical that is currently unresolved.
The applicability of the proposed strategy is demonstrated in a real 140-node distribution network located in southern Spain where two different situations have been considered, namely a normal operation in the medium voltage and low voltage networks and abnormal situations in the medium voltage network. In each situation, three voltage control strategies are compared—local decentralized voltage control, distributed voltage control, and coordinated voltage control. In the comparison, the advantages and disadvantages of each control scheme have been highlighted.
The proposed coordinated voltage control (CVC) schemes have been compared with two conventional voltage control techniques: decentralized local control, where the interconnection requirements of renewable energy sources defined in the standard IEEE 1547-2018, “IEEE Standard for Interconnection and Interoperability of Distributed Energy Resources with Associated Electric Power Systems Interfaces”, have been applied, and, also, a distributed control composed of two independent controllers for low voltage and medium voltage grids respectively. In the distributed control, there is no information exchange between medium voltage and low voltage grids.
The analysis of the results shows that the proposed coordinated voltage control allows the operation of the complete distribution system (i.e., medium voltage network and low voltage grids) to be optimized in terms of power losses and curtailment associated with distributed generation units. Furthermore, the proposed coordinated voltage control is the only scheme capable of determining the optimal working set points for the controllable units’s under abnormal network operating conditions, such as the loss of a line that happens at the medium voltage grid.
In view of the results, the coordinated control algorithm has been applied to the grid under the consideration of different load and distributed generation situations. It can be concluded that the implementation of the proposed algorithm allows the power losses in the system to be minimized if a coordinated optimal assignation of the distributed generation units present in the whole system is undertaken. Due to this optimal assignation, the algorithm could reduce the required var capacity of medium voltage reactive compensation devices, since these have been employed at one-fifth of their rated capacity only for one out of the 96 hours that comprise the study period. With respect to the main transformers located at the secondary substations (i.e., medium voltage- low voltage), the application of the coordinated control allows the system to minimize the number of tap changes for the voltage regulation of these substations. The reason for this is because they employ two transformer ratios—one is for the hours in which the power is exported to the medium voltage grid, and the other one is for those hours in which power is imported instead, which increments their lifespan.
