According to the Zhitong Finance APP, China Merchants Securities has released a research report stating that with the rapid development of the AI industry, the United States is making significant investments in the data center sector. Configuring energy storage systems can reduce the grid connection capacity of data centers and smooth out power fluctuations, effectively promoting the rapid grid connection of projects. Additionally, energy storage configuration helps achieve owners' climate goals and save on electricity costs. Currently, grid connection issues caused by power supply have become a bottleneck, and there are many explorations in the industry regarding power supply for data centers. Rapidly deployable energy storage systems can lower the grid connection capacity of data centers and smooth out power fluctuations, reducing the impact on the power grid and greatly facilitating the rapid grid connection of data center projects.
<h2>The main points of China Merchants Securities are as follows:</h2>
With the rapid development of the AI industry, the United States is making significant investments in the data center sector. This will significantly drive up electricity consumption across the United States, with data centers becoming major electricity consumers. The characteristics of high load density and significant demand fluctuations make the grid connection issue a major bottleneck for industry development. In Texas, the queue time for grid connection can reach up to 11 years, and grid connection capability has become the top priority consideration for data center site selection in the United States.
Configuring energy storage systems can reduce the grid connection capacity of data centers and smooth out power fluctuations, effectively promoting the rapid grid connection of projects. Additionally, energy storage configuration helps achieve owners' climate goals and save on electricity costs. Therefore, the application of energy storage in data centers is expected to see significant growth. It is estimated that by 2030, the demand for energy storage driven by data centers in the United States will be between 122-245 GWh.
Data centers are becoming major electricity consumers. Over the past decade, the annual electricity consumption in the United States has remained stable at around 3900 TWh, with an estimated total electricity consumption of 3937 TWh in 2024, and data center electricity consumption reaching 142 TWh, accounting for 3.6%. With the explosive growth of data center scale driven by AI, it is expected that by 2030, data center electricity consumption will reach 323 TWh, accounting for over 8%.
Grid connection issues caused by power supply have become a bottleneck. Data centers exhibit characteristics of high load density and significant power demand fluctuations. Against the backdrop of the rapid development of the AI industry, these characteristics are being further amplified. The resulting power supply issues are becoming significant challenges for industry development, with many projects queued due to grid connection issues. As of the end of 2024, the operational scale of data centers in the United States is 34.7 GW, while the scale of queued grid connection projects has reached 100 GW. In major data center development areas, the grid connection queue time is 3 years in the Chicago area, 7 years in Virginia, and as high as 11 years in Texas. Grid connection capability (queue time, capacity) has become the top priority consideration for data center site selection in the United States.
Emphasizing the application of energy storage in AIDC. In response to the bottleneck of difficult grid connections for data centers, the industry is exploring various power supply solutions for data centers. Rapidly deployable energy storage systems can lower the grid connection capacity of data centers and smooth out power fluctuations, reducing the impact on the power grid and greatly facilitating the rapid grid connection of data center projects. Considering that major data center operators in the United States currently have relevant climate goals, the use of energy storage systems is also beneficial for achieving these goals
In addition, electricity expenses account for more than half of the operating costs of data centers, and energy storage systems capable of peak-valley arbitrage can effectively save data center operating expenses. According to estimates by China Merchants Securities, by 2030, data centers are expected to drive U.S. energy storage demand by 122-245 GWh (based on 4h and 8h storage).
Recommended attention to: Sungrow Power Supply (300274.SZ), Canadian Solar (688472.SH), TRINA SOLAR (688599.SH), Kehua Data (002335.SZ), Haibo Sichuang (688411.SH), WDI (603312.SH), Shenghong Co., Ltd. (300693.SZ), Sineng (300827.SZ), etc.
Risk warning: The penetration rate of energy storage in data centers may not meet expectations, changes in industry policies, increasing trade barriers, and certain subjectivity in calculations, etc

300274

603312

300827

688599

China Merchants Securities released a research report stating that with the rapid development of the AI industry, investments in data centers in the United States have significantly increased. Configuring energy storage systems can reduce the grid capacity of data centers, smooth out power fluctuations, promote rapid grid connection of projects, and help achieve climate goals and save electricity costs. It is expected that by 2030, the demand for energy storage from data centers in the United States will reach 122-245 GWh, and electricity consumption by data centers will increase from 142 TWh to 323 TWh, accounting for more than 8%

China Merchants Securities: AI-driven investment expansion in the data center sector, emphasizing the application of energy storage in AIDC