How to improve the reliability of energy storage battery management system

2019-03-26 14:58:09 0

Large-scale battery arrays can be used as backup and continuous power storage systems, and this usage is gaining more and more attention, as evidenced by Tesla Motors' recently introduced home and commercial Powerwall systems. The batteries in such systems are continuously charged by the grid or other sources of energy and then supplied with alternating current (AC) power to the user via a DC/AC inverter.

Using a battery as a backup power supply is not new. There are many battery backup power systems available, such as basic 120/240V AC and hundreds of watts of bench PC short-term backup power system, ship, hybrid or all-electric. The multi-kilowatt special-vehicle backup power system used in automobiles, the grid-level hundreds of kilowatts backup power system used in telecommunications systems and data centers (see Figure 1)...etc. While advances in battery chemistry and battery technology have drawn a lot of attention, there is another equally important part of a viable and battery-based backup system, the Battery Management System (BMS).

The battery's backup power supply is ideal for fixed and mobile use from kilowatts to hundreds of kW, providing reliable and efficient power for a wide range of applications.

There are many challenges in completing a battery management system for energy storage, and the solution is not simply "expanded" from a small, low-capacity battery pack management system. Instead, new, more complex strategies and key support components are needed.

The starting point of the battle is that the measurement values of many critical battery parameters are required to have high accuracy and credibility. In addition, the planning of the subsystems must be modular to be able to tailor the configuration to the specific needs of the application, taking into account possible expansion requirements, overall management issues, and necessary maintenance.

The working environment of larger storage arrays has also brought other major challenges. In the case of high inverter voltages/currents and thus current spikes, BMS must also supply accurate, common data in very noisy electrical environments and often in very high temperature environments. In addition, the BMS must provide extensive “fine” data for internal module and system temperature measurements, rather than a limited number of rough aggregated data, as these are critical for charging, monitoring, and discharging.

Because of the important role of these power systems, their operational reliability is inherently critical. To make this easy-to-represent goal a reality, the BMS must ensure data accuracy and integrity and continuous health assessment so that the BMS can continue to take the required action. Completing robust planning and reliable security is a multi-stage process where the BMS must anticipate possible problems for all subsystems, perform self-tests and provide fault detection, and then select the appropriate actions in the standby mode and mode of operation. The last requirement is that BMS must meet many strict regulatory standards because of high voltage, high current and high power.

System planning transforms concepts into real-world results

While supervising rechargeable batteries is conceptually simple, simply place the voltage and current measurement circuitry at the battery terminals, but the actual BMS is quite different and much more complicated.

The sturdy planning begins with a comprehensive oversight of the batteries, which puts some important requirements on the function of the circuit. Battery readings need to be millivolt and milliamp accuracy, and voltage and current measurements must be synchronized to account for power. The BMS must evaluate the effectiveness of each measurement as it requires maximum data integrity while the BMS must also identify incorrect or problematic readings. BMS cannot ignore unusual readings because such readings may indicate potential problems, but at the same time, BMS cannot act on the wrong data.