Battery Management System BMS Knowledge and Function, An Introduction

1) What is BMS?

The full name of BMS is Battery Management System. It is a device that monitors the status of energy storage batteries. It is mainly used for intelligent management and maintenance of individual battery cells, preventing overcharging and overdischarging of batteries, extending battery life, and monitoring battery status. Generally, BMS is represented as a circuit board or a hardware box.
The BMS is one of the core subsystems of the battery energy storage system, responsible for monitoring the operating status of each battery in the battery energy storage unit and ensuring the safe and reliable operation of the energy storage unit. The BMS can monitor and collect the status parameters of the energy storage battery in real time (including but not limited to single cell voltage, battery pole temperature, battery loop current, battery pack terminal voltage, battery system insulation resistance, etc.), and perform necessary analysis and calculation on relevant status parameters to obtain more system status evaluation parameters. It can also achieve effective control of the energy storage battery itself according to specific protection control strategies to ensure the safe and reliable operation of the entire battery energy storage unit. At the same time, the BMS can interact with other external devices (PCS, EMS, fire protection system, etc.) through its own communication interface and analog/digital input interface to form a linkage control of various subsystems in the entire energy storage power station, ensuring safe, reliable, and efficient grid-connected operation of the power station.
2) Architecture
From the perspective of topology architecture, BMS is divided into two categories: centralized and distributed according to different project requirements.
Centralized BMS
Simply put, centralized BMS uses a single BMS hardware to collect all the cells, which is suitable for scenarios with few cells.
Centralized BMS has the advantages of low cost, compact structure, and high reliability, and is commonly used in scenarios with low capacity, low total pressure, and small battery system volume, such as power tools, robots (handling robots, assistive robots), IOT smart homes (sweeping robots, electric vacuum cleaners), electric forklifts, electric low-speed vehicles (electric bicycles, electric motorcycles, electric sightseeing cars, electric patrol cars, electric golf carts, etc.), and light hybrid vehicles.
The centralized BMS hardware can be divided into high-voltage and low-voltage areas. The high-voltage area is responsible for collecting single cell voltage, system total voltage, and monitoring insulation resistance. The low-voltage area includes power supply circuits, CPU circuits, CAN communication circuits, control circuits, and so on.
As the power battery system of passenger vehicles continues to develop towards high capacity, high total pressure, and large volume, distributed BMS architectures are mainly used in plug-in hybrid and pure electric vehicle models.
Distributed BMS
At present, there are various terms for distributed BMS in the industry, and different companies have different names. The power battery BMS mostly has a master-slave two-tier architecture:
The energy storage BMS is usually a three-tier architecture due to the large size of the battery pack, with a master control layer above the slave and main control layers.
Just like batteries form battery clusters, which in turn form stacks, the three-tier BMS also follows the same upward rule:
From the control: battery management unit (BMU), which collects information from individual batteries.
Monitor the voltage and temperature of the battery cell
Battery equalization in the package
Information upload
thermal management
Abnormal alarm
Master control: Battery cluster management unit: BCU (battery cluster unit, also known as high voltage management unit HVU, BCMU, etc.), responsible for collecting BMU information and gathering battery cluster information.
Battery cluster current acquisition, total voltage acquisition, leakage detection
Power-off protection when the battery status is abnormal
Under the management of BMS, capacity calibration and SOC calibration can be completed separately as the basis for subsequent charging and discharging management
The battery array management unit (BAU) is responsible for centralized management of the batteries in the entire energy storage battery stack. It connects to various battery cluster management units and exchanges information with other devices to provide feedback on the operating status of the battery array.
Charging and discharging management of battery array
BMS system self-checking and fault diagnosis alarm
Battery pack fault diagnosis alarm
Safety protection for various abnormalities and faults in the battery array
Communicate with other devices such as PCS and EMS
Data storage, transmission and processing
Battery management layer: responsible for collecting various information (voltage, temperature) of individual batteries, calculating and analyzing SOC and SOH of batteries, achieving active equalization of individual batteries, and uploading abnormal information of individual batteries to the battery pack unit layer BCMU. Through CAN external communication, it is interconnected through a daisy chain.
Battery management layer: responsible for collecting various information from individual batteries uploaded by the BMU, collecting various information about the battery pack (pack voltage, pack temperature), battery pack charging and discharging currents, calculating and analyzing the SOC and SOH of the battery pack, and uploading all information to the battery cluster unit layer BAMS. Through CAN external communication, it is interconnected through a daisy chain.
Battery cluster management layer: responsible for collecting various battery information uploaded by BCMU and uploading all information to the energy storage monitoring EMS system through RJ45 interface; communicating with PCS to send relevant abnormal information of the battery to PCS (CAN or RS485 interface), and equipped with hardware dry nodes to communicate with PCS. In addition, it performs battery system BSE (Battery State Estimate) evaluation, electrical system status detection, contactor management, thermal management, operation management, charging management, diagnostic management, and performs internal and external communication network management. Communicates with subordinates through CAN.
3) What does BMS do?
The functions of BMS are numerous, but the core and what we are most concerned about are three aspects:
One is sensing (state management), which is the basic function of BMS. It measures voltage, resistance, temperature, and ultimately senses the state of the battery. We want to know what the state of the battery is, how much energy and capacity it has, how healthy it is, how much power it produces, and how safe it is. This is sensing.
The second is management (balance management). Some people say that BMS is the nanny of the battery. Then this nanny should manage it. What to manage? It is to make the battery as good as possible. The most basic is balance management and thermal management.
The third is protection (safety management). The nanny also has a job to do. If the battery has some status, it needs to be protected and an alarm needs to be raised.
Of course, there is also a communication management component that transfers data within or outside the system through certain protocols.
BMS has many other functions, such as operation control, insulation monitoring, thermal management, etc., which are not discussed here.
3.1 Perception – Measurement and Estimation
The basic function of BMS is to measure and estimate battery parameters, including basic parameters such as voltage, current, temperature, and state, as well as calculations of battery state data such as SOC and SOH. The field of power batteries also involves calculations of SOP (state of power) and SOE (state of energy), which are not discussed here. We will focus on the first two more widely used data.
Cell measurement
1) Basic information measurement: The most basic function of the battery management system is to measure the voltage, current, and temperature of the individual battery cells, which is the foundation for all top-level calculations and control logic in the battery management system.
2) Insulation resistance testing: Insulation testing is required for the entire battery system and high-voltage system within the battery management system.
3) High-voltage interlock detection (HVIL): used to confirm the integrity of the entire high-voltage system and initiate safety measures when the integrity of the high-voltage system loop is compromised.
SOC calculation
SOC refers to the State of Charge, which is the remaining capacity of the battery. Simply put, it is how much power is left in the battery.
SOC is the most important parameter in BMS, as everything else is based on it. Therefore, its accuracy and robustness (also known as error correction capability) are extremely important. Without accurate SOC, no amount of protection function can make BMS work properly, as the battery will often be in a protected state, making it impossible to extend the battery’s life.
At present, the mainstream SOC estimation methods include open-circuit voltage method, current integration method, Kalman filter method, and neural network method. The first two methods are commonly used. The latter two methods involve advanced knowledge such as integration models and artificial intelligence, which are not detailed here.
In practical applications, multiple algorithms are often used in combination, with different algorithms being adopted depending on the battery’s charging and discharging status.
open circuit voltage method
The principle of open-circuit voltage method is to use the relatively fixed functional relationship between open-circuit voltage and SOC under the condition of long-term static placement of the battery, and thus estimate SOC based on open-circuit voltage. The previously commonly used lead-acid battery electric bicycle uses this method to estimate SOC. Open-circuit voltage method is simple and convenient, but there are also many disadvantages:
1. The battery must be left standing for a long time, otherwise the open circuit voltage will be difficult to stabilize in a short period of time;
2. There is a voltage plateau in batteries, especially lithium iron phosphate batteries, where the terminal voltage and SOC curve are approximately linear during the SOC30%-80% range;
3. The battery is at different temperatures or different life stages, and although the open circuit voltage is the same, the actual SOC difference may be large;
As shown in the figure below, when we use this electric bicycle, if the current SOC is displayed as 100%, the voltage drops when accelerating, and the power may be displayed as 80%. When we stop accelerating, the voltage rises, and the power jumps back to 100%. So our electric scooter’s power display is not accurate. When we stop, it has power, but when we start up, it runs out of power. This may not be a problem with the battery, but may be due to the SoC algorithm of the BMS being too simple.
An-Shi integral method
The Anshicontinuous integration method directly calculates the SOC value in real time through the definition of SOC.
Given the initial SOC value, as long as the battery current can be measured (where the discharge current is positive), the change in battery capacity can be accurately calculated through current integration, resulting in the remaining SOC.
This method has relatively reliable estimation results in a short period of time, but due to measurement errors of the current sensor and gradual degradation of the battery capacity, long-term current integration will introduce certain deviations. Therefore, it is generally used in conjunction with open-circuit voltage method to estimate the initial value for SOC estimation with low accuracy requirements, and can also be used in conjunction with Kalman filtering method for short-term SOC prediction.
SOC (State Of Charge) belongs to the core control algorithm of BMS, representing the current remaining capacity status. It is mainly achieved through the ampere-hour integration method and EKF (Extended Kalman Filter) algorithm, combined with correction strategies (such as open-circuit voltage correction, full-charge correction, charging end correction, capacity correction under different temperatures and SOH, etc.). The ampere-hour integration method is relatively reliable under the condition of ensuring current acquisition accuracy, but it is not robust. Due to the accumulation of errors, it must be combined with correction strategies. The EKF method is robust but the algorithm is relatively complex and difficult to implement. Domestic mainstream manufacturers can achieve an accuracy of less than 6% at room temperature, but estimating at high and low temperatures and battery attenuation is difficult.
SOC correction
Due to current fluctuations, the estimated SOC may be inaccurate, and various correction strategies need to be incorporated into the estimation process.
SOH calculation
SOH refers to the State of Health, which indicates the current health status of the battery (or the degree of battery degradation). It is typically represented as a value between 0 and 100%, with values below 80% generally considered to indicate that the battery is no longer usable. It can be represented by changes in battery capacity or internal resistance. When using capacity, the actual capacity of the current battery is estimated based on data from the battery’s operating process, and the ratio of this to the rated capacity is the SOH. An accurate SOH will improve the estimation accuracy of other modules when the battery is deteriorating.
There are two different definitions of SOH in the industry:
SOH definition based on capacity fade
During the use of lithium-ion batteries, the active material inside the battery gradually decreases, the internal resistance increases, and the capacity decays. Therefore, SOH can be estimated by the battery capacity. The health status of the battery is expressed as the ratio of the current capacity to the initial capacity, and its SOH is defined as:
SOH=(C_standard-C_fade)/C_standard ×100%
Where: C_fade is the lost capacity of the battery; C_standard is the nominal capacity.
IEEE standard 1188-1996 stipulates that when the capacity of the power battery drops to 80%, the battery should be replaced. Therefore, we usually consider that the battery SOH is not available when it is below 80%.
SOH definition based on power attenuation (Power Fade)
The aging of almost all types of batteries will lead to an increase in battery internal resistance. The higher the internal resistance of the battery, the lower the available power. Therefore, the SOH can be estimated using power attenuation.
3.2 Management – Balanced Technology
Each battery has its own “personality”
To talk about balance, we have to start with batteries. Even batteries produced in the same batch by the same manufacturer have their own life cycles and “personalities” – the capacity of each battery cannot be exactly the same. There are two reasons for this inconsistency:
One is the inconsistency of cell production
One is the inconsistency of electrochemical reactions.
production inconsistency
Production inconsistencies are easy to understand. For example, during the production process, diaphragm inconsistencies and cathode and anode material inconsistencies can result in overall battery capacity inconsistencies. A standard 50AH battery may become 49AH or 51AH.
electrochemical inconsistency
The inconsistency of electrochemistry is that in the process of battery charging and discharging, even if the production and processing of the two cells are identical, the thermal environment can never be consistent in the process of electrochemical reaction. For example, when making battery modules, the temperature of the surrounding ring must be lower than that of the middle. This results in long-term inconsistency between charging and discharging amounts, which in turn leads to inconsistent battery cell capacity; When the charging and discharging currents of the SEI film on the battery cell are inconsistent for a long time, the aging of the SEI film will also be inconsistent.
*SEI film: “solid electrolyte interface” (solid electrolyte interface). During the first charge discharge process of liquid lithium ion battery, the electrode material reacts with the electrolyte on the solid-liquid phase interface to form a passivation layer covering the surface of the electrode material. SEI film is an electronic insulator but an excellent conductor of lithium ions, which not only protects the electrode but also does not affect battery function. The aging of SEI film has a significant impact on battery health.
Therefore, non-uniformity (or discreteness) of battery packs is an inevitable manifestation of battery operation.
Why balance is needed
The batteries are different, so why not try to make them the same? Because inconsistency will affect the performance of the battery pack.
The battery pack in series follows the short-barrel effect: in the battery pack system in series, the capacity of the entire battery pack system is determined by the smallest single unit.
Suppose we have a battery pack consisting of three batteries: 微信截图_20240115195929e know that overcharging and overdischarging can seriously damage batteries. Therefore, when battery B is fully charged during charging or when the SoC of battery B is very low during discharging, it is necessary to stop charging and discharging to protect battery B. As a result, the power of batteries A and C cannot be fully utilized.
This leads to:
The actual usable capacity of the battery pack has decreased: Battery A and C, which could have used the available capacity, are now unable to do so in order to accommodate Battery B. It is like two people on three legs tied together, with the taller person unable to take large steps.
Reduced battery life: A smaller stride length requires more steps and makes the legs more tired. With a reduced capacity, the number of charge and discharge cycles increases, resulting in greater battery degradation. For example, a single cell can achieve 4000 cycles at 100% DoD, but in actual use it cannot reach 100% and the number of cycles will certainly not reach 4000.
*DoD, Depth of discharge, represents the percentage of battery discharge capacity to the rated capacity of the battery.
The inconsistency of batteries leads to a decrease in the performance of the battery pack. When the size of the battery module is large, multiple strings of batteries are connected in series, and a large single voltage difference will cause the capacity of the entire box to decrease. The more batteries connected in series, the more capacity they lose. However, in our applications, especially in energy storage system applications, there are two important requirements:
The first is long-life battery, which can greatly reduce the operation and maintenance costs. The energy storage system has high requirements for the life of the battery pack. Most of the domestic ones are designed for 15 years. If we assume 300 cycles per year, 15 years is 4500 cycles, which is still very high. We need to maximize the life of each battery so that the total life of the entire battery pack can reach the design life as much as possible, and reduce the impact of battery dispersion on the life of the battery pack.
The second deep cycle, especially in the application scenario of peak shaving, releasing one more kWh of electricity will bring one more point of revenue. That is to say, we will do 80%DoD or 90%DoD. When the deep cycle is used in the energy storage system, the dispersion of the battery during the tail discharge will be manifested. Therefore, in order to ensure the full release of the capacity of each single cell under the condition of deep charging and deep discharging, it is necessary to require the energy storage BMS to have strong equalization management capabilities and suppress the occurrence of consistency among battery cells.
These two requirements are exactly contrary to battery inconsistency. To achieve more efficient battery pack applications, we must have more effective balancing technology to reduce the impact of battery inconsistency.
equilibrium technology
Battery equalization technology is a way to make batteries with different capacities the same. There are two common equalization methods: energy dissipation unidirectional equalization (passive equalization) and energy transfer bidirectional equalization (active equalization).
(1) Passive balance
The passive equalization principle is to parallel a switchable discharge resistor on each string of batteries. The BMS controls the discharge resistor to discharge the higher voltage cells, dissipating the electrical energy as heat. For example, when battery B is almost fully charged, the switch is opened to allow the resistor on battery B to dissipate excess electrical energy as heat. Then charging continues until batteries A and C are also fully charged.
This method can only discharge high-voltage cells, and cannot recharge low-capacity cells. Due to the power limitation of the discharge resistance, the equalization current is generally small (less than 1A).
The advantages of passive equalization are low cost and simple circuit design; the disadvantages are that it is based on the lowest remaining battery capacity for equalization, which cannot increase the capacity of batteries with low remaining capacity, and that 100% of the equalized power is wasted in the form of heat.
(2) Active balance
Through algorithms, multiple strings of batteries transfer the energy of high-voltage cells to low-voltage cells using energy storage components, discharging the higher-voltage batteries and using the energy released to charge the lower-voltage cells. The energy is mainly transferred rather than dissipated.
In this way, during charging, battery B, which reaches 100% voltage first, discharges to A and C, and the three batteries are fully charged together. During discharge, when the remaining charge of battery B is too low, A and C “charge” B, so that cell B does not reach the SOC threshold for stopping discharge so quickly.
Main features of active balancing technology
(1) Balance the high and low voltage to improve the efficiency of the battery pack: During charging and discharging and at rest, the high-voltage batteries can be discharged and the low-voltage batteries can be charged;
(2) Low-loss energy transfer: energy is mainly transferred rather than simply lost, improving the efficiency of power utilization;
(3) Large equilibrium current: Generally, the equilibrium current is between 1 and 10A, and the equilibrium is faster;
Active equalization requires the configuration of corresponding circuits and energy storage devices, which leads to large volume and increased cost. These two conditions together determine that active equalization is not easy to be promoted and applied.
In addition, the active equalization charging and discharging process implicitly increases the cycle life of the battery. For cells that require charging and discharging to achieve balance, the additional workload may cause them to exceed the aging of ordinary cells, resulting in a greater performance gap with other cells.
Some experts believe that the two expressions above should correspond to dissipative equilibrium and non-dissipative equilibrium. Whether it is active or passive should depend on the event that triggers the equilibrium process. If the system reaches a state where it has to be passive, it is passive. If it is set by humans, setting the equilibrium program when it is not necessary to be balanced is called active equilibrium.
For example, when the discharge is at the end, the lowest voltage cell has reached the discharge cut-off voltage, while other cells still have power. At this time, in order to discharge as much electricity as possible, the system transfers the electricity of high-energy cells to low-energy cells, allowing the discharge process to continue until all the power is discharged. This is a passive equalization process. If the system predicts that there will be an imbalance at the end of discharge when there is still 40% of power left, it will start an active equalization process.
Active equalization is divided into centralized and decentralized methods. The centralized equalization method obtains energy from the entire battery pack, and then uses an energy conversion device to supplement energy to the batteries with less energy. Decentralized equalization involves an energy storage link between adjacent batteries, which can be an inductor or a capacitor, allowing energy to flow between adjacent batteries.
In the current balance control strategy, there are those who take the cell voltage as the control target parameter, and there are also those who propose using SOC as the balance control target parameter. Taking the cell voltage as an example.
First, set a pair of threshold values for initiating and ending equalization: for example, in a set of batteries, when the difference between the extreme voltage of a single cell and the average voltage of the set reaches 50mV, equalization is initiated, and when it reaches 5mV, equalization is ended.
The BMS collects the voltage of each cell according to a fixed acquisition cycle, calculates the average value, and then calculates the difference between each cell voltage and the average value;
If the maximum difference reaches 50mV, the BMS needs to start the equalization process;
Continue step 2 during the equalization process until the difference values are all less than 5mV, and then end the equalization.
It should be noted that not all BMSs require this step, and subsequent strategies may vary depending on the balance method.
The balance technology is also related to the type of battery. It is generally believed that LFP is more suitable for active balance, while ternary batteries are suitable for passive balance.
The stage of intense competition in BMS is mostly supported by cost and reliability. Currently, the experimental verification of active balancing has not yet been achieved. The level of functional safety is expected to move towards ASIL-C and ASIL-D, but the cost is quite high. Therefore, the current large companies are cautious about active balancing research. Some large factories even want to cancel the balancing module and have all the balancing performed externally, similar to the maintenance of fuel vehicles. Every time the vehicle travels a certain distance, it will go to the 4S store for external balancing. This will reduce the cost of the entire vehicle BMS and also benefit the corresponding 4S store. It is a win-win situation for all parties. Therefore, personally, I understand that this may become a trend!
3.3 Protection – fault diagnosis and alarm
The BMS monitoring is matched with the hardware of the electrical system, and it is divided into different failure levels (minor failure, serious failure, fatal failure) according to the different performance conditions of the battery. Different handling measures are taken in different failure levels: warning, power limitation or direct high-voltage cut-off. Failures include data acquisition and rationality failures, electrical failures (sensors and actuators), communication failures, and battery status failures.
A common example is when a battery overheats, the BMS determines that the battery is overheating based on the collected battery temperature, then controls the circuit of this battery to disconnect, performs overheating protection, and sends an alert to management systems such as EMS.
3.4 Communication
The normal operation of BMS cannot be separated from its communication function. Whether it is controlling the battery during battery management, transmitting battery status to the outside world, or receiving control instructions, stable communication is required.
In the power battery system, one end of the BMS is connected to the battery, and the other end is connected to the control and electronic systems of the entire vehicle. The overall environment uses CAN protocol, but there is a distinction between using internal CAN between internal components of the battery pack and using vehicle CAN between the battery pack and the entire vehicle.
In contrast, energy storage BMS and internal communication basically use CAN protocol, but its external communication (external mainly refers to the energy storage power station dispatching system PCS) often uses Internet protocol formats TCP/IP protocol and modbus protocol.
4) Energy storage BMS
Energy storage BMS manufacturers generally evolved from power battery BMS, so many designs and terms have historical origins
For example, the power battery is generally divided into BMU (Battery Monitor Unit) and BCU (Battery Control Unit), with the former collecting data and the latter controlling it.
Because the battery cell is an electrochemical process, multiple battery cells form a battery. Due to the characteristics of each battery cell, no matter how precise the manufacturing process is, there will be errors and inconsistencies in each battery cell over time and depending on the environment. Therefore, the battery management system is to evaluate the current state of the battery through limited parameters, which is a bit like a traditional Chinese medicine doctor diagnosing a patient by observing symptoms rather than Western medicine requiring physical and chemical analysis. The physical and chemical analysis of the human body is similar to the electrochemical characteristics of the battery, which can be measured by large-scale experimental instruments. However, it is difficult for embedded systems to evaluate some indicators of electrochemistry. Therefore, BMS is like an old Chinese medicine doctor.
4.1 Three-layer architecture of energy storage BMS
Due to the large number of battery cells in energy storage systems, in order to save costs, BMS is generally implemented in layers, with two or three layers. Currently, the mainstream is three layers: master control/master control/slave control.
4.2 Detailed description of energy storage BMS
5) Current situation and future trend
There are several types of manufacturers that produce BMS:
The first category is the end-user with the most dominant power in the power battery BMS – car factories. In fact, the strongest BMS manufacturing strength abroad is also the car factories, such as General Motors, Tesla, etc. At home, there are BYD, Huating Power, etc.
The second category is battery factories, including cell manufacturers and pack manufacturers, such as Samsung, Ningde Times, Xinwangda, Desay Battery, Topband Co., Ltd., Beijing Purrad, etc.;
The third type of BMS manufacturers are those with many years of experience in power electronics technology, and have R&D teams with university or related enterprise backgrounds, such as Eternal Electronics, Hangzhou Gaote Electronics, Xie Neng Technology, and Kegong Electronics.
Unlike the BMS of power batteries, which is mainly dominated by terminal vehicle manufacturers, it seems that the end users of energy storage batteries have no need or specific actions to participate in the research and development and manufacturing of BMS. It is also unlikely that they will spend a lot of money and energy to develop large-scale battery management systems. Therefore, it can be considered that the energy storage battery BMS industry lacks an important player with absolute advantages, leaving a huge space for development and imagination for battery manufacturers and vendors focusing on energy storage BMS. If the energy storage market is established, it will give battery manufacturers and professional BMS manufacturers a lot of room for development and less competitive resistance.
Currently, there are relatively few professional BMS manufacturers focused on the development of energy storage BMS, mainly due to the fact that the energy storage market is still in its infancy and there are still many doubts about the future development of energy storage in the market. Therefore, most manufacturers have not developed BMS related to energy storage. In the actual business environment, there are also manufacturers who purchase electric vehicle battery BMS for use as BMS for energy storage batteries. It is believed that in the future, professional electric vehicle BMS manufacturers are also likely to become an important part of the BMS suppliers used in large-scale energy storage projects.
At this stage, there is a lack of uniform standards for BMS provided by various energy storage system suppliers. Different manufacturers have different designs and definitions for BMS, and depending on the different batteries they are compatible with, the SOX algorithm, equalization technology, and communication data content uploaded may also vary. In the practical application of BMS, such differences will increase application costs and be detrimental to industrial development. Therefore, the standardization and modularization of BMS will also be an important development direction in the future.

Post time: Jan-15-2024