Distributed Energy Storage in Energy Internet

China Energy Storage Network News : Traditional energy networks are mainly based on fossil energy, using a unidirectional power flow tree topology, centralized control, vertical integration of energy supply and energy consumption (including the current intelligent regional energy network), and segmentation between different energy systems. With the rise of distributed energy and the rapid growth of energy demand on the user side, the traditional energy system's way of solving energy demand by continuously increasing fossil energy consumption is no longer sustainable, and it cannot effectively solve the access of large-scale distributed renewable energy and the energy supply needs of emerging industries, such as electric vehicles and cloud computing and other national strategic emerging industries. Therefore, how to fundamentally change the energy structure and improve the comprehensive utilization efficiency of energy has become a hot issue of common concern to the government and all sectors of industry, academia, research, and use.

In recent years, the idea of ​​energy internet has emerged as a powerful theoretical support for solving the above problems. In essence, the rise of energy internet is also the inevitable result of reconfiguring and optimizing various production factors in the energy field after information (data) has evolved into a new type of production material. From a technical point of view, energy internet, as a new type of energy system, is the physical infrastructure that supports the realization of "Internet + smart energy". At the same time, it is also a customized energy service industry ecosystem centered on user energy experience. Energy internet integrates resources horizontally and drives industries through the system architecture that tightly couples information systems with traditional energy systems (such as electricity, heat, gas, oil, nuclear, transportation and other energy networks that operate independently), and realizes the activation of heavy asset stocks in traditional energy industries through information light asset increments, supports the integrated and efficient use of multiple energy sources, and thus generates new value. As an energy internet that integrates the open genes of the edge of the Internet, its generation is first driven by the increasingly diverse energy demand on the user side. We can easily draw this conclusion from the popularity of new user-side bidirectional power sources such as distributed photovoltaics, electric vehicles, and home energy storage. In the era of energy internet, the market positioning of traditional users has changed from simple energy users (consumers) to dynamic energy users (prosumers). This role change will have a significant and far-reaching impact on the evolution and development of the energy system. From the user side, the energy internet must meet the C2B and C2C models of energy production and supply driven by diversified energy demand. Therefore, the energy internet is also regarded as a specific example of "Industry 4.0" in the energy field. Its main output is to provide flexible and diverse customized energy services for various user needs.

In the era of energy internet, there will be a large number of bidirectional energy nodes at the edge of the traditional energy network, which will pose a great challenge to the B2C model supported by the traditional energy network system with vertical integration, one-way flow and closed edge in terms of technology and business model. In particular, the emergence of large-scale bidirectional energy nodes at the edge of the network breaks the design and control boundary conditions of the traditional energy network system. The transformation of users from consumers to prosumers has brought more complex randomness, suddenness and uncertainty. Moreover, the random superposition of these suddenness and uncertainty will interfere with the operation stability and security of the core grid of the traditional energy network, triggering the opposition and contradiction between the distributed energy system and the traditional energy system at the network control level, thus affecting the development of the entire new energy industry.

From the perspective of system theory, the traditional energy system is a tightly coupled one-way analog system and a simple stress system without memory, while the energy Internet is a loosely coupled two-way digital system and an intelligent information-physical system. Therefore, the evolution from the traditional energy system to the energy Internet requires the addition of a large amount of distributed storage (memory) capabilities to the traditional energy system to support energy virtualization and digital processing and optimize the global efficiency and stability of the energy system. Therefore, distributed energy storage is one of the core technologies for the rapid development of the energy Internet. From the perspective of technical implementation methods, there are currently many ways of distributed energy storage, including battery energy storage, phase change energy storage, cold and heat storage energy storage, flywheel energy storage, pumped water energy storage, supercapacitor energy storage and compressed air energy storage. Among them, battery energy storage has incomparable advantages over other energy storage methods in terms of kilowatt-hour to megawatt-hour energy storage. Therefore, in this article, we will take the most widely used battery energy storage as an example to specifically explain the problems and countermeasures of the development of distributed energy storage systems.

Distributed battery energy storage system is the core equipment to achieve dynamic matching between power generation curve and power consumption curve. It has the functions of smoothing fluctuations, matching supply and demand, shaving peaks and filling valleys, improving power supply quality, enhancing disaster recovery capabilities, delaying power grid upgrades and transformations, etc. It is an indispensable device in distributed energy systems. In distributed battery energy storage, lithium-ion battery energy storage has the advantages of fast effect speed, high energy conversion efficiency, fast cost reduction and good scalability, and is listed as the first choice. Therefore, lithium battery energy storage is one of the energy storage methods with the most market application prospects. However, due to problems such as production process, safety, reliability, and ease of use, the capacity of lithium battery cells cannot meet the load's demand for energy storage capacity, so the battery grouping method is inevitable. In terms of battery grouping technology, small-capacity monomer grouping or networking has incomparable advantages over large-capacity monomers in terms of safety, reliability and manageability of energy storage systems. This view has been confirmed by many application practices. For example, Tesla Model S electric vehicles use more than 8,000 18650 (i.e. cylindrical batteries with a diameter of 18 mm and a height of 65 mm) small-capacity monomer batteries. However, the energy storage system composed of a large number of battery cells has problems in terms of efficiency, operation and maintenance costs, safety and reliability, which is currently recognized as a technical problem in the world. The scientific problem is the mismatch between the inevitable differences between battery cells and the rigid system architecture of fixed series and parallel connection adopted by the battery grouping technology, which causes the "short board effect" of the battery pack. In addition, the battery itself is an electrochemical reaction process, and its working process exhibits very strong nonlinear characteristics. Therefore, the battery state of charge SOC (state of charge) is difficult to measure and estimate accurately, which further increases the difficulty of battery system management and control. Since the traditional battery management system (BMS) is an information measurement system superimposed on a fixed connected battery pack, it cannot completely solve the problems of single cell balance, efficiency, maintainability, service life, cascade utilization, reliability and safety that exist in large-scale battery packs. Therefore, we need to rethink the technical route of distributed battery energy storage systems from the perspective of the tight coupling of energy flow and information flow in the energy Internet, introduce new ideas of distributed energy management and control based on energy virtualization and energy informatization, and realize "separation of batteries from management and control systems, and separation of battery systems from application systems". The battery energy is virtualized into a computable and measurable Internet resource, which strongly supports the development of distributed energy storage in the energy Internet.

In addition, the distributed energy storage system in the energy internet is a necessary device for the distributed energy to access the traditional energy system, so the function and performance definition of the distributed energy storage system should match the distributed energy access requirements, just as the diversity of information storage systems in the information field corresponds to the different requirements of information processing. As shown in the figure, similar to the information storage architecture from the top register to the bottom Internet storage, the unit storage cost from top to bottom is gradually decreasing, but the response speed and delay interruption performance of the storage system are constantly increasing. The energy internet energy storage system architecture has the same characteristics. The construction of the energy storage system in the energy internet needs to cope with the requirements of various distributed power sources to randomly access the traditional power grid in an disorderly manner. We need to introduce a software-defined energy management and control system to achieve the use of different types of energy storage systems in combination to achieve the purpose of reducing unit costs and optimizing system performance.