1. Mainstream technology route and characteristics of energy storage products
In terms of technology roadmap, it is mainly divided into five technical schools: centralized, string (distributed), high-voltage cascade, intelligent string, and decentralized.
Centralized: battery cluster → DC cable → DC combiner box → DC cable → centralized converter → AC cable → step-up transformer
Multiple battery clusters are directly connected in parallel on the DC side bus, and the DC current is converted into AC through an energy storage converter. This method is currently a widely used technical route, with the advantage of simple control and the disadvantage of generating circulating current when the voltage between battery clusters is inconsistent. In China, centralized energy storage currently has the highest proportion, simple structure, low investment cost, and convenient installation and operation in the future.
String type (distributed type): battery cluster → DC/DC → DC cable → single inverter → AC cable → AC combiner box → AC cable → step-up transformer
High voltage cascade: battery → H-bridge (DC/AC power unit) → H-bridge cascade → three-phase star connection
The system consists of multiple energy storage units, each consisting of an H-bridge and an independent small battery stack. Each phase is connected in series with multiple energy storage units to a certain voltage and directly connected to the AC power grid. The advantages are that there is no need for a step-up transformer, reducing system losses, minimizing footprint, no need for parallel connection between battery clusters, and eliminating inter cluster circulating current problems. The disadvantage is that only 5MW and above have economic viability, and can only output voltage levels such as 6kV and 10kV, lacking flexibility in industrial and commercial applications.
Intelligent string type: battery cluster → DC/DC (may not be available) → DC cable → multiple inverters → AC cable → AC combiner box → AC cable → step-up transformer
Similar to string type (distributed type), the difference is that DC current is converted into AC through multiple smaller capacity converters instead of using a larger capacity converter for conversion. The advantage is that a single inverter failure will not affect the entire energy storage system.
Distributed: battery cluster → DC cable → inverter → AC cable → AC combiner box → AC cable → step-up transformer
Each battery cluster is individually connected in series with an energy storage inverter, and multiple energy storage inverters are connected in parallel on the AC bus side, not on the DC side. The advantage of this method is that it can solve the circulation problem between battery clusters, and each cluster can be managed separately or fault isolated. The disadvantage is that due to the large number of inverters, the stability and reliability requirements of the system are high.
2. Characteristics of intelligent string energy storage technology
String type: Firstly, an energy optimizer is used to refine the energy management of the energy storage system to the Pack level, minimizing the impact of Pack series mismatch and improving the available capacity of the entire energy storage system; Secondly, through the battery cluster controller, the battery capacity is balanced during the charging and discharging process, and the parallel mismatch between batteries is minimized to achieve single cluster energy management; Finally, a distributed intelligent temperature control architecture is adopted, with each battery cabinet corresponding to a separate cascaded air conditioner. Each group of batteries independently and uniformly dissipate heat, reducing the temperature rise difference between reading rooms and improving the temperature balance of the energy storage system.
Intelligence: Firstly, advanced technologies such as AI and cloud BMS are applied to internal short circuit detection scenarios, which can accurately locate derivative internal short circuits, calculate internal short circuit resistance accurately, identify sudden internal short circuits in real time, timely warn of battery fire hazards, and provide 24-hour advance warning of thermal runaway faults to ensure the safety of energy storage systems; Secondly, AI technology can also be used to build relevant prediction models, estimate battery SoX parameters, and predict battery health in advance to reduce initial battery overfitting; Finally, multiple models such as battery life, battery behavior, and environmental prediction are applied to link intelligent temperature control strategies to find the optimal balance between battery degradation and temperature control energy consumption, ensuring real-time optimization of LCOS.
Modularization: Adopting a full system modular design. Firstly, the battery system should be modularized, allowing for the separate removal of faulty modules without affecting the normal operation of other modules. When replacing modules, there is no need to manually adjust the SOC on site; Secondly, the PCS will be modularized in design. PCS is a key core component in the energy storage system and has a significant impact on the availability of the power station. In the energy storage sub array, when a single PCS fails, other PCS can continue to work, and when multiple PCS fail, the system can still maintain operation.
3. Development trend of intelligent string energy storage technology
At present, the domestic energy storage market is mainly driven by new energy distribution and storage, and there are also some independent energy storage projects constantly being invested and constructed. Due to the imperfect electricity market mechanism, energy storage projects are facing profitability difficulties, which has led investment operators to pay extra attention to initial investment. The industry generally agrees that batteries need to be finely managed to solve usability issues, but the specific implementation methods are different.