The battle for large-capacity batteries is in full swing, with energy storage integrated with AI technology accelerating.

Securities Times reporter Liu Canbang

Recently, Chen Haisheng, chairman of the China ZGC (000931) Energy Storage Industry Technology Alliance, released a forecast: by 2030, the cumulative installed scale of domestic new-type energy storage will reach 371.2GW—450.7GW, and the compound annual growth rate from 2026—2030 will be 20.7%—25.5%.

Data show that over the past five years, domestic new-type energy storage installations have grown by more than 40 times. At the ongoing 14th International Energy Storage Summit and Exhibition, more than 800 enterprises across the full energy storage industry chain have come to exhibit, highlighting just how hot the energy storage sector is.

Since the second half of last year, as policy guidance shifted from mandatory capacity allocation to market-based utilization, some links in the energy storage industry chain had at one point experienced supply tightness. At this energy storage expo, large-capacity cells have become the main battlefield where manufacturers “show their muscles.” In addition to technological iteration, AIDC (artificial intelligence data centers) energy storage as a key application scenario has also become the focus of competition among manufacturers. In addition, AI prediction models have been deeply embedded into the entire energy storage trading process to adapt to the new situation brought by power marketization reforms.

Accelerating iteration of large-capacity cells

At present, the mainstream energy storage cell model is 314Ah. At this energy storage expo, the reporter saw that besides models like 587Ah and 588Ah, multiple manufacturers also displayed 600Ah-class, 700Ah-class, and even higher-capacity cells.

The reporter learned from relevant personnel at CATL (300750) that since the 587Ah cell was introduced in 2025, its mass-production progress has been closely watched. Currently, its cumulative shipments have already exceeded 5GWh.

In application fields, the grid-side benchmark project Erlhan Tu 400MW/2400MWh independent energy storage power station located in Baotou, Inner Mongolia, is one of the largest grid-side independent energy storage projects in China. CATL introduced that this project applied 587Ah cells in the first large-scale batch. Compared with the traditional 314Ah方案, the number of components is reduced by about 40%, the system energy density has been significantly improved, and construction and operation/maintenance costs have effectively been lowered.

At the Ruiru Lanjon (Ruipu Lanjun) exhibition booth, the reporter saw that the company’s core products are 588Ah large-capacity energy storage cells. The company also showcased its latest Powtrix system solution, further strengthening its integrated energy storage capability from cells to systems. A staff member at the company’s booth said that each 588Ah super-large-capacity cell can achieve more than 10k ultra-long cycles, meeting the high-intensity requirements of long-duration energy storage.

The context for industry competition over large-capacity cells is that the standard for the third-generation energy storage cell has not yet been finalized. The “Zhijiu” series of long-cycle energy storage cells shown by CALB (Cen Innovation?) covers three capacity versions: 392Ah, 588Ah, and 661Ah.

“Judging from our perspective, the next three years will be the era of 500Ah-class big cells in the energy storage industry,” Wang Bing, product director of Penghui Energy (300438), told reporters. The company’s Fengpeng big cell 587Ah has entered the production line commissioning stage, with mass production and delivery expected to begin in June. “The Fengpeng big cell 587Ah is mainly for the next-generation large storage market. We’ve already received some orders from domestic customers. For the overseas market, it is expected to be put into the market in 2027.”

Currently, not only cell suppliers are rolling out large-capacity technology, but equipment manufacturers such as energy storage inverters are also actively pushing product adaptation. At the booth of Kehua Digital Energy, the company focused on showcasing products including a 6.25/6.9MW centralized energy storage inverter/step-up integrated unit and a 6.25MW modular energy storage inverter/step-up integrated unit. Their power has been significantly improved compared with the previous generation.

Feng Xuetao, product manager at Kehua Digital Energy, told reporters: “As battery capacity gradually increases, our products continue to iterate toward higher power and higher energy density. Currently, 5MW inverters will continue to be promoted, but in the future demand will gradually decrease; the 6.25MW/6.9MW products are expected to become the new market mainstream.”

In addition, as large-capacity and long-duration energy storage technologies accelerate deployment, the energy storage industry is also placing higher requirements on testing: it must meet higher voltage level and accuracy standards, and also balance efficiency and reliability in large-scale production. A person in charge of Ruilan shares told reporters that based on the actual needs of energy storage R&D and manufacturing, the company has built an end-to-end testing system covering every level, including energy storage cells, modules/PACK, battery clusters, energy storage systems, and MW-class containerized units.

AIDC energy storage is set to surge

If large-capacity cells are a popular path for energy storage technology iteration, then from the perspective of scenario applications, AIDC energy storage is one of the core directions attracting market attention.

At the expo, Trina Storage, together with Kehua Digital Energy, released an AIDC full-domain integrated solution. The solution is built on the “source-and-storage grid-forming as the foundation, and efficient load delivery as the core,” constructing an end-to-end, full-chain system from energy production, storage, and distribution to consumption. A company executive said the solution is not just a simple combination of product offerings, but a reconfiguration of energy flows from “grid to chip,” and a reshaping of energy flows from “grid to chip,” enabling high energy-consuming data centers to have an operational foundation that is efficient, stable, and low-carbon, providing sustainable energy support for AI infrastructure.

Future also simultaneously showcased its first sodium-ion energy storage dedicated cells and an end-to-end energy solution for the AIDC edge-to-edge. It is reported that the company’s first sodium-ion energy storage dedicated cell officially rolled off the production line in March 2026. Its capacity is over 180Ah, with a cycle life of more than 20k times. In the future, it will match differentiated scenarios such as AIDC backup power and extreme high/low temperature environments, complementing lithium batteries in synergy.

In addition, Future Energy (also known as Vison?) first released an end-to-end energy solution from “chip to grid,” covering the grid-side, station-side, load-side, and control-side, directly targeting the power bottleneck of data centers in the era of AI computing power. The company is working with leading AI enterprises to build what is expected to be the world’s largest zero-carbon AIDC park in Ulanqab (Ulan Tu), and the project is expected to be delivered this year.

“AI large models drive a sharp increase in global computing power; large computing power must be supported by electricity. The company’s AIDC business growth last year exceeded 600%, and we have collaborated with major domestic internet companies,” said a person from a manufacturer focusing on the AIDC energy storage market.

Song Wan, Minister of Energy Storage R&D at Goldwind Technology, told reporters that in AIDC scenarios, the reliability requirements for the power supply system are extremely high. Once accidents such as a power outage occur, the impact on data service providers and possible property losses would be significant. Therefore, reliability is the core threshold for energy storage to enter AIDC.

Song Wan Guang believed that large-scale application of AIDC energy storage can consider two paths: first, further improve the reliability of energy storage inverters (PCS)/DC-DC, strengthen product performance, and further reduce failure rates; second, implement redundant thermal backup designs for AIDC energy storage, configuring additional energy storage equipment as backup so that seamless handover can be achieved when a failure occurs, thereby lowering the risk of power interruption in AIDC.

“Based on a comprehensive view of information such as the frequency of customer inquiries, order delivery schedules, and market-collected information, we judge that AIDC energy storage will experience a surge by the end of this year,” Wang Bing said.

AI deeply embedded in energy storage trading

In the past two years, the energy storage industry has benefited from policy tailwinds. Last year’s Document No. 136 canceled mandatory energy storage procurement, and promoted all electricity of renewable energy entering the market. This opened a market-based profit model for independent energy storage, and broadened new paths for diversified returns and asset appreciation for high-quality development of renewable energy. This year’s Document No. 114, for the first time, included grid-side independent new-type energy storage into the capacity pricing system, fully establishing a “capacity minimum protection + market upside” new mechanism for revenue.

As Yang Bo, General Manager of Hemi? said, currently the energy storage industry is accelerating its shift from policy-driven momentum to market-oriented value creation, and from scale expansion to high-quality development. This trend is also the direction of Hemi’s strategic layout. With regard to a new-type power system, competition in energy storage is no longer limited to single equipment, but has been upgraded to a comprehensive contest involving full-scenario solutions, value across the full life cycle, and coordination capabilities across the full industry chain.

Industry insiders believe that in the future, energy storage will be deployed more on the user side. As the power market shifts from fixed tariffs to time-varying tariffs, the profitability model of peak-valley arbitrage for user-side energy storage will be constrained, and tariff fluctuations will further test energy storage charging/discharging strategies and trading models.

To deal with this situation, the enablement of AI digital technology appears especially critical for energy storage trading models. At this energy storage expo, the reporter saw that multiple manufacturers are focusing on improving returns and creating value.

For example, Guoneng New Vision showcased its newly released Kuangming AI intelligent system—Kuangming meteorological forecasting model 4.0 and Kuangming power trading model 1.0. Company representatives told the reporter that the two models deeply collaborate to thoroughly move power trading from an “experience-based game” to a new paradigm of “data-driven, intelligent decision-making.” This transforms the high economic utilization of renewable energy assets from probabilistic events into computable, controllable, and actionable deterministic outcomes.

“The company uses an AI-driven smart energy (600869) solution, providing support for safe and stable operation of the power grid through its self-developed WE platform and a virtual power plant management platform, and also opens a new path for market-based returns,” said people from Goodwe.

“In the future, energy storage trading will need to forecast spot electricity prices every day. More AI-based power trading agents will be introduced, improving the accuracy of electricity price predictions. AI investment on power trading platforms will be the key battleground for manufacturers,” said an exhibitor.

(Editor: Zhang Yang HN080)

     【Disclaimer】This article only represents the author’s personal views and is not related to Hexun. The Hexun website maintains a neutral stance toward the statements and judgment of opinions contained in the article, and provides no express or implied guarantee regarding the accuracy, reliability, or completeness of the content. Please read for reference only, and the reader should bear full responsibility on their own. Email: news_center@staff.hexun.com

Report