The new energy industry is showing signs of entering a new technological cycle. On September 18, the 2025 International Digital Energy Exhibition opened in Shenzhen, featuring over 2,000 companies from more than 50 countries. The event showcased more than 300 cutting-edge technologies and created a 50,000 square meter innovative ecological platform, providing a “China solution” and a “Shenzhen model” for the global energy revolution through a comprehensive perspective of the “source-grid-load-storage” chain.
Firstly, Shenzhen is exploring digital energy, offering a replicable and scalable “Shenzhen solution” for promoting energy transformation in densely populated cities, with a potential market size reaching trillions. Secondly, the integration of AI has become a core focus. For example, Huawei’s “end-edge-cloud” intelligent energy storage system has improved operational efficiency in power plants by 50%. Additionally, CNOOC’s “Marine Energy” model has achieved predictive maintenance for equipment, reducing production downtime.
Moreover, the technological breakthrough in large-scale energy storage cells has opened up new pathways for long-duration storage. For instance, BYD has launched a 2710Ah blade battery designed for a 10MWh energy storage system, while Xinwangda’s 684Ah cell meets the spatial demands of energy storage cabinets. As a result, the recent recovery in lithium battery demand has led to a tight supply of battery cells, prompting leading companies to accelerate capacity planning and strategic alliances, signaling the start of a new industrial cycle.
The impetus for energy transformation in major global cities stems from the challenges posed by high-density buildings, which are often subjected to urban heat island effects, heat waves, and high energy consumption. Rapid urbanization requires consideration of multiple energy sources and storage mediums. The “2025 Shenzhen Digital Energy White Paper,” released at the exhibition, outlines the implementation path for Shenzhen to become a global leader in digital energy, focusing on six major directions: “source, grid, load, storage, data, and carbon.”
Shenzhen has specifically proposed to create the world’s first “Super Charging City 2.0,” establish a network for power charging and storage, and develop a virtual power plant management cloud platform. In terms of the Super Charging City 2.0 initiative, Shenzhen began construction in 2023, aiming to exceed the number of supercharging stations compared to traditional gas stations by December 2024. Currently, the city has built and put into operation 1,055 supercharging stations, with over 420,000 charging piles.
In the past year, Shenzhen has introduced six pioneering local standards nationwide, which not only specify that the rated power of supercharging equipment should not be less than 480 kilowatts but also establish standards for the site selection, layout, and power quality requirements of centralized public charging stations for electric vehicles. In terms of virtual power plants, Shenzhen’s adjustable capacity has reached 1.3 million kilowatts, accounting for over 5.4% of the city’s peak grid load, the highest ratio in the country, and successfully meeting predetermined targets ahead of schedule. This means that during peak electricity usage, Shenzhen’s virtual power plants can significantly alleviate demand, enabling the equivalent of 870,000 households to use air conditioning simultaneously.
In March of this year, the first specialized policy document for virtual power plants was officially released. Additionally, Shenzhen established the country’s first local standard committee for virtual power plants, continuously improving the standard system covering management and technology. The related industry has also rapidly expanded, with over 60 virtual power plant operators in Shenzhen. In the recently announced list of virtual power plant operators in the southern regional electricity market, 70% were cultivated in Shenzhen.
Technologies such as blockchain and artificial intelligence have made the operation of Shenzhen’s virtual power plants greener, ensured better data privacy, enhanced multi-level monitoring efficiency, and matured operational models. Since the beginning of 2023, Shenzhen’s virtual power plants have conducted load adjustments over 150 times, the highest frequency in the country, supporting a 20% reduction in overload rates during peak load periods, resulting in direct economic benefits exceeding 200 million yuan.
Shenzhen has applied the “smart scheduling” function of virtual power plants to demonstrate vehicle-grid interaction, helping participating vehicle owners save an average of 15% on charging costs and increasing utilization rates of charging piles during off-peak hours by 300%. “Next, we will continue to expand the scale of resource access, promote the construction of Shenzhen’s vehicle-grid interaction demonstration, facilitate the absorption of new energy, and enhance the market-oriented mechanism design for virtual power plants, raising the development level of the industry, and promoting the transformation of Shenzhen’s virtual power plant achievements,” said Cheng Renli, general manager of the Shenzhen Virtual Power Plant Management Center.
The exhibition featured a planned area of 50,000 square meters, with the AI empowerment zone becoming the focal point. Artificial intelligence is not just a “power-hungry monster,” but also a core tool for energy transformation and efficiency enhancement. Currently, the relationship between energy and AI has formed a mutually empowering and symbiotic coupling. Companies such as Huawei and Xinwangda showcased the deep application of AI algorithms in energy forecasting and operational optimization. The clean and low-carbon exhibition area presented innovations in carbon-neutral pathways, including CCUS technology and carbon asset management tools, while the international enterprise zone gathered global energy technology giants, creating a window for direct interaction with cutting-edge technologies.
The share of renewable energy sources like wind and solar in China’s energy structure is steadily increasing. However, their intermittent and fluctuating characteristics pose challenges to stable grid operation. The energy power industry’s unique nature imposes a high demand for reliability and stability. To bridge this gap, the value of AI comes into play, primarily in data processing, forecasting, and decision-making assistance. Currently, AI applications in the energy sector show the most potential in predictive maintenance and load regulation. For example, AI can utilize machine learning algorithms to compare normal data with real-time data, identify early anomalies, provide proactive fault warnings, and reduce downtime, thus extending the lifespan of equipment.
Zhou Jianjun, Vice President of Huawei and President of Huawei Digital Energy Global Marketing Service Division, stated at the International Digital Energy Exhibition that Huawei has developed the industry’s first “end-edge-cloud” collaborative intelligent energy storage system, enabling smart management throughout the entire lifecycle of power plants, achieving reduced personnel involvement in operations, and maximizing revenue from electricity trading. By deeply integrating AI into the entire process from “planning, construction, maintenance, to operation,” the system has reduced implementation and design errors by 40%, improved operational efficiency by 50%, and increased operational revenue by over 10%.
Industry giants are betting on the fusion of AI and energy technologies to reshape the operating logic of the power industry. On one hand, general large models enhance energy office efficiency, but more importantly, temporal large models accurately monitor the “pulse” of the grid. Combined with power market reforms, AI technology could redefine the value of “one kilowatt-hour” of energy, integrating real-time services such as power response and frequency modulation, while algorithms penetrate heterogeneous data to find optimal solutions in real-time, unlocking the potential of every kilowatt-hour.
The application of AI in traditional energy has also manifested in various ways. The Chief Engineer of CNOOC’s Shenzhen branch mentioned in an interview that since the launch of CNOOC’s “Marine Energy” artificial intelligence model last October, the company has achieved intelligent applications in multiple business scenarios. In equipment management and maintenance, CNOOC has developed an intelligent operation and maintenance model for pump equipment, utilizing large models for status assessment, anomaly detection, and automatic fault diagnosis, thereby achieving predictive maintenance and reducing equipment downtime, thus minimizing production interruptions.
However, as the energy sector is a typical traditional industrial domain, it faces significant challenges such as large volumes of diverse, dynamically changing data, often leading to “data silos.” The lack of unified data standards results in low efficiency in data integration and sharing. Additionally, data quality issues affect the accuracy of advanced intelligent applications such as status assessments and trend forecasts. Consequently, most AI models within the traditional energy sector remain at the experimental stage, especially those involving core production processes. Long integration cycles between production systems and AI platforms, along with the absence of model iteration and update mechanisms, lead to rapid model degradation. Furthermore, the high costs associated with training and deploying large models also pose significant barriers to the widespread promotion of intelligent applications.
The showcase of digital energy super products and technologies at the opening ceremony became a “highlight stage” for global energy companies to display their innovative capabilities. Several leading companies sequentially unveiled their latest technological achievements and solutions, covering key areas such as intelligent networking, new energy storage, solar thermal innovation, and supercharging networks. Zhou Jianjun shared the full-scenario networking technology and global practices, solving the global challenge of grid stability with high proportions of renewable energy through self-developed foundational hardware, dual architecture, and intelligent algorithms. This technology has been successfully validated in global projects, including the world’s largest 100% renewable microgrid in Saudi Arabia’s Red Sea and the Yalong River solar-hydropower complementary station in Sichuan, China, ensuring safe grid operation.
During the 2025 International Digital Energy Exhibition in Shenzhen, BYD released a gigawatt-level networking energy storage solution, encompassing the entire supply chain from battery cells to the grid. This includes self-developed integrated PCS machines (GCFlux) and energy management systems (GCMaster), showcasing a fully self-researched stack from cell to grid. On September 19, BYD’s Director of the Energy Science Research Institute, Cao Hu, unveiled the new generation energy storage product “Haohan,” which is equipped with the world’s largest 2710Ah blade battery, with a 10MWh equivalent capacity in a 20-foot container.
Likewise, Xinwangda presented two new energy storage cells, 684Ah and 588Ah. The 684Ah energy storage cell utilizes lamination technology, while the 588Ah cell employs winding technology. Despite differing technical paths, both cells demonstrate core advantages of long lifespan, high efficiency, and excellent economic performance. Liang Rui, Vice President and Chief Sustainability Officer of Xinwangda, stated in an interview that the current development of battery cells differs from previous designs based on electrochemical principles and product design convenience. Xinwangda designs products according to customer application scenarios, such as the 684Ah cell developed in collaboration with Sungrow to maximize suitability for outdoor energy storage cabinet space requirements. The entire energy storage cabinet is designed and configured to include battery cabinets, battery clusters, and fire safety monitoring thermal management systems, significantly enhancing system efficiency while improving overall energy density. Thus, in the future, we aim to leverage this new cell to capture a larger market share in the energy storage industry.
The capacity of battery cells has evolved from 280Ah and 314Ah to now seeing the emergence of 500+Ah and even 600+Ah cells. The primary driving force behind the large-scale development of battery cells arises from the urgent market demand for reducing the overall cost of energy storage systems. High-capacity battery cells can effectively reduce the amount of raw materials and manufacturing costs by decreasing the number of cells needed and simplifying wiring and structural components. Furthermore, they enhance the energy density of containers, enabling more energy storage in the same space, thus lowering land, transportation, and ancillary facilities costs, which directly improves project economics. Additionally, the increasing share of renewable energy generation is driving the demand for long-duration energy storage, which typically requires storage solutions lasting four hours or more, further fueling the need for high-capacity battery cells.
It is noteworthy that there has been a significant recovery in the lithium battery market recently, with a new lithium cycle expected to commence. As of June 30, 2025, China’s new energy storage capacity has surged from 73.76 million kilowatts at the end of 2024 to an impressive 94.91 million kilowatts. In the coming year and a half, the installed capacity is expected to double, marking the most solid performance support for the lithium battery industry cycle.
Original article by NenPower, If reposted, please credit the source: https://nenpower.com/blog/new-energy-enters-a-technological-era-innovations-in-cities-ai-and-energy-storage-at-the-2025-international-digital-energy-expo/
