BU-908: Battery Management System (BMS)

Learn about old and new methods and how innovations may change old habits.

A battery should be equipped with a state-of-charge (SoC) gauge that shows the remaining charge. This alone is not complete without revealing the capacity as the battery fades. The user may have been accustomed to a battery that delivered full capacity, but this condition is temporary and cannot be maintained. With use and time, the capacity drops and a state-of-health (SoH) indicator should be foremost in a battery management system (BMS). Knowing SoC and SoH provides state-of-function (SoF), the ultimate confidence of readiness, but technology to do this effectively is still in development.

Building a better BMS is a challenge when considering that we still don’t have a dependable method to read state-of-charge, the most basic measure of a battery. (See BU-903: How to Measure State-of-charge) Reading the remaining energy in a battery is more complex that dispensing liquid fuel. While a fuel tank has a fixed dimension delivering fuel that and can be measured with great accuracy, an electrochemical storage system is constantly reducing in size and the in- and out-flowing coulombs cannot always be reassessed accurately.

A BMS not only reveals SoC but it also provides protection during charging and loading under all environmental conditions and it must disconnect the battery if a failure occurs. Communication is commonly achieved with the System Management Bus, also known as SMBus (see BU601: How does a Smart Battery Work?). Vehicular BMS use the CAN Bus (Controller Area Network) or the LIN Bus (Local Interconnect Network) among other communications protocols. 

Stationary batteries were among the first to include monitoring systems, and the most basic form of supervision is voltage measurement of the individual cells. Some systems also include cell temperature and current measurement. Recording a slight difference in cell temperature hints to a problem, and measuring the voltage drop of each cell at a given load reveals cell resistance. Dry-out, corrosion, plate separation, and other malfunctions can thus be identified.

Although the BMS is effective in detecting anomalies; capacity fade, the most predictable prospect, is difficult to estimate because voltage and the internal resistance remain the same. Reading capacity fade from 100 percent to 80 and 60 percent would be most valuable, but most BMS cannot do this effectively and the battery is given a clean bill of health even if the capacity has dropped to replacement level. Most BMS only respond to anomalies that lay outside capacity estimation, such as voltage differences among cells caused by cell imbalances and changes in internal resistance.

Some industrial and medical device manufacturers use a date stamp to determine the end of battery life, others observe the cycle count. While counting cycles may be simplistic, no convention exists that defines a cycle and some systems simply call it a cycle when the battery is charged. Date-stamping has similar shortcomings in that it promotes premature replacement of batteries that are seldom used and for short durations while the heavy hitters could stay in service too long. To reduce the risk of failure, authorities mandate early replacement, and a two-year service life is common. Other manufacturers state a three year service life from date of manufacturing; prolonged storage will give the batteries a short life.

Batteries have improved and live longer than in the past, but they also cost more. Biomedical engineers say that a Li-ion in a medical device is good for up to five years and they are aware that most batteries are replaced too soon. A manager at DOE, USA, said that every year roughly one million usable lithium-ion batteries are sent in for recycling with most having a capacity of up to 80 percent. Meanwhile, military leaders say that their battery arsenal for combat is so poor that soldiers carry rocks instead of batteries.

Over-expectations with a BMS are common and the user is stunned when stranded without battery power. iPhone owners complained that their smartphones show 100 percent charge when the battery is only 90 percent charged. Engineers at SAE meeting in Detroit, USA, disclosed that the SoC readings on new EV batteries were off by as much as 15 percent. According to reported cases, EV drivers have run out of charge with 25 percent SoC readings left on the fuel gauge.

Let’s look at how a BMS works and examine up-and-coming technologies that could change the way batteries are monitored.

A BMS takes the imprint of the “chemical battery” during charging and loading activities and establishes the “digital battery” that communicates with the user. Figure 1 illustrates the battery components consisting of stored energy, the empty portion that can be refilled and the inactive part that is permanently lost. The rated capacity refers to the manufacturer’s specified capacity in Ah (ampere-hours) that is only valid when the battery is new; the available capacity designates the true energy storage capability derived by deducting the inactive part. State-of-charge (SoC) refers to the stored energy, which also includes the inactive part that cannot be claimed.

Rated Capacity Figure 1: Three parts of a battery

A battery consists of stored energy, the empty portion that can be recharged and the inactive portion that is permanently lost due to aging.

A BMS is set up to the rated capacity and measures the in-and-outflowing coulombs that relate to the available capacity. As the capacity drops, the coulomb count decreases. This discrepancy enables capacity estimation, and the most accurate readings are possible when counting the coulombs from a fully discharged battery during a complete charge, or by discharging a fully charged battery to the cut-off point. Such clean starts are seldom possible in real life and the capacity estimations get muddled.

A modern BMS uses the opportunity to calibrate the battery after receiving a full discharge and charge. During a rest period, the BMS may calculate SoC on hand of the stable open circuit voltage and begin counting the coulombs during charge and discharge from that vantage point. Some BMS also look at voltage recovery after removing a load to estimate SoC and/or SoH. 

Battery Sensing by Voltage-Current-Temperature

In simpler days, the Volkswagen Beetle had minimal battery problems. Its management system did nothing more than to charge the battery while driving. Since then, modern vehicles have been inundated with onboard electronics to enhance safety, convenience, comfort and pleasure, features no one knew we needed. These extra loads add demand on the battery, and for the accessories to function reliably the state-of-charge of the battery must be known at all times. This is especially critical with start-stop technologies, a requirement for European cars to improve fuel economy that is adopted worldwide.

When the engine of a start-stop car is off at a red light, the battery draws 25–50 amperes to feed the lights, ventilators, windshield wipers and other accessories. The battery must have enough charge to crank the engine, which requires an additional 350A for a brief moment. When the engine is running again and the car is accelerating to the posted speed limit, the battery only begins charging after a 10-second delay. This deferral allows channeling all energy to vehicle acceleration.

Realizing the importance of battery monitoring, luxury cars are fitted with a battery sensor that measure voltage, current and temperature. Packaged in a small housing that forms part of the positive clamp, the electronic battery monitor (EBM) provides vital information about the battery. Figure 2 illustrates a battery sensor that is based on voltage-current-temperature.

BMS Figure 2: Battery sensor for starter battery

The sensor reads voltage, current and temperature to estimate state-of-charge and detect anomalies; capacity assessment is not possible.

The EBM works well when the battery is new but most sensors do not adjust correctly to aging. The SoC accuracy of a new battery is about +/–10 percent, and as the battery ages the EBM begins drifting and the accuracy can drop to 20 percent and lower. The drift in accuracy is in part connected to capacity fade, a deficiency most BMS cannot estimate effectively. This is not an oversight by engineers; they fully understand the complexities and shortcomings.

A typical start-stop vehicle goes through about 2,000 micro cycles per year. Test data obtained from automakers and the Cadex laboratories indicate that the battery capacity drops to approximately 60 percent in two years while in in start-stop configuration. The standard lead acid is not robust enough to for start-stop use and carmakers revert to modified versions, including AGM and the Advanced Lead-carbon. (Also see BU-806a: How Heat and Loading affect Battery Life)

Automakers ensure that no driver gets stuck in traffic with a dead battery. To conserve energy, modern cars turn off unnecessary accessories when the battery is low on charge and the motor will stay on at a stoplight. Even with this measure, the state-of-charge can remain low if commuting in gridlock traffic because an idling motor does not provide much charge to the battery. There could even be a small discharge when lights, windshield wipers and electric heating elements are engaged.

Battery monitoring is also important on hybrid vehicles to optimize charge levels. Intelligent charge management prevents overcharge and avoids deep discharge. When the charge level is low, the internal combustion engine (ICE) engages earlier than normal and is left running longer for additional charge. On a fully charged battery the ICE turns off and the car moves on electric energy in slow traffic.

An EV driver expects similar accuracies in energy reserve as is possible with a fuel-powered vehicle but current technology does not allow this. To compensate, the EV battery is overrated and the fuel gauge is adjusted to preserve extra energy when the charge drops low to cover for inaccuracies.

The EV driver also anticipates the same driving range as the car ages. This is not possible and the drivable distance gets shorter with each passing year, but some BMS make allowance. A new battery may only charge to about 80 percent and discharge to 30 percent. As the capacity fades, the bandwidth gradually increases, providing similar driving ranges as a new battery would. The distances traveled will be noticeable shorter when driving in cold temperatures because of reduced battery performance.

Capacity Estimation in BMS

The EBM has limitations in that it cannot estimate capacity effectively. Adding electrochemical impedance spectroscopy (EIS) with complex modeling corrects this shortcoming. (See BU-904: How to Measure Capacity) Figure 3 shows a BMS with common sensing points to which capacity estimation has been added. Such an addition will convert a simple battery sensor to the state-of-function (SoF) level to provide high confidence in readiness. Developments are in progress to reduce the circuit into a small and economic model for inclusion in the battery circuit.


BMS Figure 3: Spectro-BMS™ adds Reserve Capacity as key element to estimate battery state-of-health.

Knowing SoF improves battery validation, but some device manufacturers might be hesitant to reveal a capacity reading to the consumer that is less than 100%, especially during the warranty period. To conceal unwanted information, the data could be code-accessible for service personnel only.

Consumer concerns put aside, SoF signifies a momentous improvement to BMS in terms of battery reliability as it tracks capacity fade and calculates the true runtime on the available energy. Capacity-based BMS also predicts eventual replacement, an issue that cannot be fully satisfied with current BMS technologies. Future BMS will combine the information of the “digital battery” with that of the “chemical battery” to provide reliable SoF data through advanced learn algorithm.

Last updated 2015-06-01

*** Please Read Regarding Comments ***

Comments are intended for "commenting," an open discussion amongst site visitors. Battery University monitors the comments and understands the importance of expressing perspectives and opinions in a shared forum. However, all communication must be done with the use of appropriate language and the avoidance of spam and discrimination.

If you have a question, require further information, have a suggestion or would like to report an error, use the "contact us" form or email us at: answers@cadex.com. While we make all efforts to answer your questions accurately, we cannot guarantee results. Neither can we take responsibility for any damages or injuries that may result as a consequence of the information provided. Please accept our advice as a free public support rather than an engineering or professional service.

Or Jump To A Different Article

Basics You Should Know
The Battery and You
Batteries as Power Source


On January 8, 2012 at 10:34pm
Alfred Mahon wrote:

Learned alot about my Prius 12 Volt Auxillary battery,  that Toyota does not know or wants
to conceed lack of knowledgr Ihard to believe).  “Just buy a NEW battery whenever you
think you need one or come in and we Toyota) will ghage and check it for you )for a good
dolllar fee of cource>  What a guarnteed make buy/work system!!!! e I can locate a
CADEX—“Q-MAG Monitor.  and stop my ansiety evry morning ,—-” will it start today?.

On November 7, 2012 at 2:42am
Jeremy Gidney wrote:

I am looking for a reliable method to determine a car battery’s state of charge.  We perform emissions tests to determine catalyst performance.  Variations in the battery state of charge will change the CO2 and NOx emissions.  Would your device enable me to datalog useful information with respect to the battery condition?  How can I obtain one or more?

On February 4, 2014 at 8:09pm
Rex Roberts wrote:

Where can I find a Q-Max?

On June 18, 2014 at 5:58am
Josh Phillips wrote:

This information is very inaccurate as of 2014.  There are monitoring systems out there that can give you much more detailed information than what this article leads you to believe.  For instance, on an aging string of batteries you can still predict their life cycle by using a MidTronics meter for instance to see the current admittance, and comparing that against the batteries original admittance from the date of manufacture.  You can track this change over time by using a good battery monitoring system like the system available from Phoenix Broadband Technologies.  You can see the % of change from birth and setup alarms to email certain individuals when it falls below a certain %.

On September 22, 2014 at 10:21am
James Buniak wrote:

” This is a demanding task when considering that there is still no dependable method to read state-of-charge”

On lead acid non vented batteries state of charge can be read by reading the specific gravity. Just a side note.