Smart grid is a communication network to collect and analyze data from distinctive components of the electricity grid to forecast power demand and supply that leads to better power management [1]. The smart grid contains four fundamental subsystems: power generation, transmission, distribution, and utilization.
Internet of Things in Smart Grid
IoT services can support the smart grid functionalitecties by incorporating communication technologies, sensors, actuators and smart meters. Smart grids are built upon two types of communication flow. The first flow occurs from sensors and appliances to smart meters, while the second flow is between smart meters and energy suppliers’ data centers. For the first flow, power line communications or wireless technologies such as ZigBee, Bluetooth, and ultra-wideband are used. The second flow is accomplished through cellular networks such as 3G and 4G or the Internet. Employed IoT applications enhance real-time and bidirectional data reading among the electricity corporation, customers, and power equipment in smart grid.
Fig. 1 shows smart grid architecture presenting power subsystems, power flow and information flow. Smart grid is formed by three categories of networks which are a wide area network (WAN), a neighborhood area network (NAN) and a home area network (HAN). The power flows through the subsystems and the information flows through the networks.
HAN is deployed within residential units or industrial plants. Smart devices, home appliances, electric vehicles, and renewable energy sources such as solar panels can connect via a HAN. Electrical appliances send the power consumption data to smart meters; therefore, consumers’ power demands are managed in this network. Bluetooth, family standards of IEEE 802.11, power line communication, or ZigBee technologies are utilized for home area networks. Furthermore, NAN is the second layer of the smart grid, consisting of smart meters located on different HANs.
NAN facilitates communication between distribution substations and field electrical devices for the distribution layer. NAN enables collecting the metering information from multiple HANs and transmits this information to WAN through data concentrators using the communication technologies such as family standards of IEEE 802.11, 3G and 4G wireless cellular networks, and optical networks. WAN is the third layer of a smart grid, and it acts as a bridge for the network gateways or data concentrators. WAN enables communication among power transmission, generation systems, renewable energy sources, and control centers. Long-range and high-bandwidth communication technologies such as fiber optic and wireless cellular networks, e.g., WiMAX, LTE, and LTE advanced, can be used in WAN.
To accomplish the communication and information flow, a number of IoT architectures have been proposed for smart grid systems. Layered IoT architectures are widely used (e.g., 3-layer or 4-layer) in the literature. Lloret et al. proposed a 3-layers IoT architecture when classifying all components in smart grids. Layer 1 includes smart meters, network devices, and communication protocols to enable smart metering via the Internet.
Layer 2 is formed by the devices receiving data on the utility side, while layer 3 is designed for intelligent systems to deliver information to the decision systems and billing systems. In particular works, 3-layer IoT architecture is extended to 4-layers architecture to represent the smart grids better. Wang et al. defined a 4-layers IoT architecture for the smart grid system consisting of the terminal layer, field network layer, remote communication layer, and master station system layer.
Remote terminal units and smart meters in the terminal layer gather information from IoT devices and transmit this information to the field network layer. The field network layer can use wired or wireless communication based on IoT device type to transfer the collected data to a remote communication network layer. Wired (e.g., optical) and wireless (e.g., 2G,3G, LTE) networks support remote communication network layer. This layer acts as middleware between IoT devices and the master station system layer. The master station system layer aims at controlling and managing smart grid functions, including generation, transmission, distribution, utilization.
Since the layered architectures are generic for many IoT applications, the specific aspects of smart grid, such as the networks (i.e., HAN, NAN, WAN) and functionalities (i.e., power generation, transmission, distribution, consumption) are covered by the alternative architectures for smart grids in the literature. For example, cloud-based smart grid architectures utilize cloud computing as a key design component for data storage and computation in smart grids. Pan et al. used a cloud-based smart grid architecture in their proposal of an energy-efficient IoT testbed that can be employed in a wide range of buildings and help energy savings on different levels.
Having used information and communication technologies such as mobile smartphones with location service, distributed control, and cloud computing, the authors involve the individuals and organizations in the energy-saving process to build energy-saving policies and use them in decision making. Mohanty et al. proposed web-enabled smart grid architecture where both non-renewable and renewable energy sources are connected to digital energy meters. The household energy consumption data collected by digital energy meters is firstly sent to IoT gateways. Then the collected data is periodically transmitted to the servers where web services are running. Thus, the end-users can utilize the data for monitoring or further analysis.
Microgrids
Smart grid is proposed as a solution for common electricity grid issues, including unpredictable power outages, consumer frauds, high greenhouse gas and carbon emission, and inflexible electricity prices. For instance, the peak hour energy demand needs to be overestimated to eliminate the power outage risks. Therefore, generating more electricity increases the utility costs. However, it has become essential to maintain efficiency and reliability in smart grids simultaneously.
The grid efficiency corresponds to the efficient use of power consumption and power system utilization. It can be improved with various mechanisms, such as renewable energy sources and energy demand management. Besides, the grid reliability is to guarantee the service to customers at all times. The reliability can be achieved by distributed energy generation, fault detection, and predicting peak demand.
Distributed energy generation enables renewable energy sources to go into service. Energy suppliers can place in new locations, especially energy sources near consumption points. The concept of microgrid has been introduced to manage the distributed power generation, and power supply operations with increased observability and control. Microgrids are interconnected local smart grids with renewable energy sources and local battery storage, as shown in Fig. 2. These microgrids are capable of self-generating energy by using wind turbines and solar panels.
They can be connected to the central system or isolated when potential risks such as failure and intrusion occur. Microgrids can improve the reliability of local distributions since these grids manage the generated and consumed power themselves. They have their power supply and control systems. Moreover, microgrids exchange information with other elements such as energy suppliers and service providers in the smart grid.
The benefits of micro-grid architecture are as follow:
1. Electricity distribution is shifted from centralized facilities towards decentralized micro-grids.
2. Micro-grids facilitate the penetration of low-carbon, renewable energy generation.
3. Micro-grids enable efficient and reliable operation, as compared with traditional energy management mechanisms.
4. Advanced real-time energy monitoring systems can be applied.
5. Service reliability can be achieved through micro-grids that can interact with each other in the presence of outages or critical load demand.
Energy demand management is enhanced with the integration of communication technologies and information management systems into smart grids. The implementation of energy demand management tools in microgrids ensures increased reliability and management of the impacts of renewable energy sources.
Studies related to energy demand management in microgrids gain attention. The referred studies mainly address the topics such as optimizing the demand response operation and energy demand forecasting in microgrids. Wang et al. proposed a genetic algorithm to reduce the operation costs of microgrids, considering the distributed generation, environmental factors, and demand response.
Yaprakdal et al. proposed an operational scheduling mechanism in microgrids using two models to forecast daily photovoltaic power generation and electrical demand. Moradzadeh et al. applied a hybrid model of machine learning methods for short-term load forecasting, using the data collected in a rural microgrid in Africa. López et al. developed a NAN model and evaluated M2M communications on short-term and medium to long-term scenarios for demand response.
Challenges for Smart Grids and Microgrids in Information Processing
Limited-capacity IoT devices: Since the resources of IoT devices, e.g., batteries, memory, and bandwidth, are limited, it is not possible to send all information to the destination. Thus, for efficient resource usage and data acquisition, data fusion techniques should be utilized to collect and aggregate data.
Network delay: It is required to optimize the smart grid network by employing the ideal number of gateway and IoT devices. Unless the number of connections between gateways is reduced, congestion occurs and causes delay and packet loss. Since IoT devices resend the data, the increased re-congestions degrades the grid performance.
Distributed energy generation: The usage of distributed generation sources in smart grids makes the control of the grid more challenging. Since renewable distributed generation sources, e.g., wind, solar, have a random nature, predictions of energy demand are dependent on the power generated by renewables. It makes the demand prediction more complicated.
Processing of big data: The IoT devices such as sensors and smart meters in a smart grid produce big data, and hence these devices consume too much energy and cause a bottleneck. Smart grid design should allow efficient storage and processing of a vast amount of collected data.
Security and privacy: Secure transmission and storage of data are essential for billing purposes and grid control operations. Effective measures and security standardization efforts are required to prevent cyberattacks and to avoid information leakage in smart grids.
