Abstract: Under the background of carbon peaking and carbon neutrality, our country develop renewable energy power generation vigorously, such as photovoltaic and wind power and actively promotes the construction of new infrastructure such as Internet data center, 5G and industrial Internet of things. Renewable energy power generation has the disadvantages of intermittency and fluctuation, a large amount of access to the grid seriously affects the grid security and electricity quality. The research results of Scholars at home and abroad have show n that the virtual power plants or demand response technology can be used to balance the power supply side and demand side. IDC has the characteristics of high power load and large power consumption, equipped with energy storage battery, it can be an ideal controllable load of virtual power plant. On the other hand, the high power consumption of the data center leads to high operating costs, which can be significantly reduced by participating in peak valley arbitrage or marketoriented demand response. The value evaluation model of energy storage is applied to study the allocation of energy storage system in large-scale data to participate in peak valley TOU price and reduce operation cost, so as to minimize the construction cost and energy consumption cost of energy storage system. A large-scale IDC in Guangdong as an application example, the maximum load is 15 384 kW. The simulation calculation showed that operation cost of the IDC equipped with energy storage system was reduced by 9%,and payback period was 3.3 years, which has good economy.