How to Know End-of-Battery-Life

A critical concern among battery users is knowing “readiness” or how much energy a battery has at its disposal at any given moment. While installing a fuel gauge on a diesel engine is simple, estimating the energy reserve of a battery is more complex — we still struggle to read state-of-charge (SoC) with reasonable accuracy. Even if SoC were precise, this information alone has limited benefits without knowing the capacity, the storage capability of a battery. Battery readiness, or state-of-function (SoF), must also include internal resistance, or the “size of pipe” for energy delivery. Figure 1 illustrates the bond between capacity and internal resistance on hand of a fluid-filled container that is being eroded as part of aging; the tap symbolizing the energy delivery.

Relationship of CCA and capacity of a starter battery

Figure 1: Relationship of CCA and capacity of a starter battery

The liquid represents capacity, the leading health indicator;
the tap symbolizes energy delivery or CCA. While the
energy delivery remains strong, the capacity diminishes
with age.

Courtesy Cadex

Most batteries for critical missions feature a monitoring system, and stationary batteries were one of the first to receive supervision in the form of voltage check of individual cells. Some systems also include cell temperature and current measurement. Knowing the voltage drop of each cell at a given load provides cell resistance. Elevated resistance hints to cell failure caused by plate separation, corrosion and other malfunctions. Battery management systems (BMS) are also used in medical equipment, military devices, as well as the electric vehicle. 

Although BMS serves an important role in supervising of batteries, such systems often falls short of expectations and here is why. The BMS device is matched to a new battery and does not adjust well to aging. As the battery gets older, the accuracy goes down and in extreme cases the data becomes meaningless. Most BMS also lack bandwidth in that they only reveal anomalies once the battery performance has dropped to 70 percent. The all-important 70–100 percent operating range is difficult to gauge and the BMS gives the battery a good bill-of-health. This prevents end-of-life prediction in that the operator must wait for the battery to show signs of wear before making a judgment. These shortcomings are not an oversight by the manufacturers, and engineers are trying to overcome them. The problem boils down to technology, or the lack thereof. Over-expectation is common and the user is stunned when stranded with a dead battery. Let’s look how current systems work and examine new technologies.

The most simplistic method to determine end-of-battery-life is by applying a date stamp or observing cycle count. While this may work for military and medical instruments, such a routine is ill suited for commercial applications. A battery with less use has lower wear-and-tear than one in daily operation and to assure reliability of all batteries, the authorities may mandate that all batteries be replaced sooner. A system made to fit all sizes causes good batteries to be discarded too soon, leading to increased operational costs and environment concerns.

Laptops and other portable devices use coulomb counting for SoC readout. The theory goes back 250 years when Charles-Augustin de Coulomb first established the “Coulomb Rule.” Coulomb counting works on the principle of measuring in- and out-flowing current of a battery. If, for example, a battery is charged for one hour at one ampere, the same energy should be available on discharge, but this is not the case. Internal losses and inaccuracies in capturing current flow add to an unwanted tracking error that must be corrected with periodic calibrations.

Calibration occurs naturally when running the equipment down. A full discharge sets the discharge flag, and the subsequent recharge establishes the charge flag (Figure 2). These two markers allow the calculation of state-of-charge by estimating the distance between the flags.

Discharge and charge flags


Figure 2:
Discharge and charge flags

Calibration occurs by applying a full charge, discharge and charge. This can be done in the equipment or externally with a battery analyzer as
part of battery maintenance.

Courtesy Cadex

Coulomb counting should be self-calibrating, but in real life a battery does not always get a full discharge at a steady current. The discharge may be in form of a sharp pulse that is difficult to capture. The battery may then be partially recharged and be stored at high temperature, causing elevated self-discharge that cannot be tracked. To correct the tracking error, a “smart battery” in use should be calibrated once every three months or after 40 partial discharge cycles. This can be done by a deliberate discharge of the equipment or externally with a battery analyzer. Avoid too many intentional deep discharges as this stresses the battery.

Fifty years ago, the Volkswagen Beetle had few battery problems. The only battery management was ensuring that the battery was being charged while driving. Onboard electronics for safety, convenience, comfort and pleasure have added to the demands of the battery in modern cars. For the accessories to function reliably, the battery state-of-charge must be known at all times. This is especially critical with start-stop technologies, a future requirement in European cars to improve fuel economy.

When the engine of a start-stop vehicle turns off at a stoplight, the battery continues to draw 25–50 amperes to feed the lights, ventilators, windshield wipers and other accessories. The battery must have enough charge to crank the engine when the traffic light changes; cranking requires a brief 350A. To reduce engine loading during acceleration, the BMS delays charging for about 10 seconds.

Modern cars are equipped with a battery sensor that measures voltage, current and temperature. Packaged in a small housing and embedded into the positive battery clamp, the electronic battery monitor (EBM) provides a SoC accuracy of about +/–15 percent on a new battery. As the battery ages, the EBM begins to drift and the accuracy drops to 20–30 percent. This can result in a false warning message and some garage mechanics disconnect the EBM on an aging battery to stop annoyances. Disabling the control system responsible for the start-stop function immobilizes engine stop and reduces the legal clean air requirement of the vehicle.

Voltage, current and temperature readings are insufficient to assess battery SoF; the all-important capacity is missing. Until capacity can be measured with confidence on-board of a vehicle, the EBM will not offer reliable battery information. Capacity is the leading health indicator that in most cases determines the end-of-battery-life. Imagine measuring the liquid in a container that is continuously shrinking in size. State-of-charge alone has limited benefit if the storage has shrunk from 100 to 20 percent and this change cannot be measured. Capacity fade may not affect engine cranking and the CCA can remain at a vigorous 70 percent to the end of battery life. Because of reduced energy storage, a low capacity battery charges quickly and has normal vital signs, but failure is imminent. A bi-annual capacity check as part of service can identify low capacity batteries. Battery testers that read capacity are becoming available at garages.

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 when in a start-stop configuration. The standard flooded lead acid is not robust enough for start-stop, and carmakers use a modified AGM (Absorbent Glass Mat) to attain longer life.

Automakers want to make sure that no driver gets stuck in traffic with a dead battery. To conserve energy when SoC is low, the BMS automatically turns unnecessary accessories off and the motor stays running at a stoplight. Even with this preventive measure, SoC can remain low when commuting in gridlock. Motor idling does not provide much charge and with essential accessories engaged, such as lights and windshield wipers, the net effect could be a small discharge.

Battery monitoring is also important in hybrid vehicles to optimize charge levels. The BMS prevents stressful overcharge above 80 percent and avoids deep discharges below 30 percent SoC. At low charge level, the internal combustion engine engages earlier and is left running for additional charge.

The driver of an electric vehicle (EV) expects similar accuracies on the energy reserve as is possible with a gasoline-powered car. Current technologies do not allow this and some EV drivers might get stuck with an empty battery when the fuel gauge still indicates reserve. Furthermore, the EV driver anticipates that a fully charged battery will travel the same distance, year after year. This is not possible and the range will decrease as the battery fades with age. Distances between charges will also be shorter than normal when driving in cold temperatures because of reduced battery performance.

Some lithium-ion batteries have a very flat discharge curve and the voltage method does not work well to provide SoC in the mid-range. An innovative new technology is being developed that measures battery SoC by magnetic susceptibility. Quantum magnetism (Q-Mag™) detects magnetic changes in the electrolyte and plates that correspond to state-of-charge. This provides accurate SoC detection in the critical 40-70 percent mid-section. More impotently, Q-Mag™ allows measuring SoC while the battery is being charged and is under load.

The lithium iron phosphate battery in Figure 3 shows a clear decrease in relative magnetic field units while discharging and an increase while charging, which relates to SoC. We see no rubber band effect that is typical with the voltage method in which the weight of discharge lowers the terminal voltage and the charge lifts it up. Q-Mag™ also permits improved full-charge detection; however, the system only works with cells in plastic, foil or aluminum enclosures. Ferrous metals inhibit the magnetic field.

Magnetic field measurements of a lithium iron phosphate during charge and discharge


Figure 3: Magnetic field measurements of a lithium iron phosphate during charge and discharge

Relative magnetic field units provide accurate state-of-charge of lithium- and lead-based batteries.

Courtesy of Cadex (2011)

Q-Mag™ also works with lead acid. This opens the door to monitor starter batteries in vehicles. Figure 4 illustrates the Q-Mag™ sensor installed in close proximity to the negative plate. Knowing the precise state-of-charge at any given moment optimizes charge methods and identifies battery deficiencies, including the end-of-battery-life with on-board capacity estimations.

Q-Mag™ sensor installed on the side of a starter battery

Figure 4: Q-Mag™ sensor installed on the side of a starter battery  

The sensor measures the SoC of a battery by magnetic susceptibility. When discharging a lead acid battery, the negative plate changes from lead to lead sulfate. Lead sulfate has a different magnetic susceptibility than lead, which a magnetic sensor can measure.

Courtesy of Cadex (2009)

Q-Mag™ is also a candidate to monitor stationary batteries. The sensing mechanism does not need to touch the electrical poles for voltage measurements and this poses an advantage for high-voltage batteries. Furthermore, Q-Mag™ can assist EVs by providing SoF accuracies not possible with conventional BMS. Q-Mag™ may one day assist in the consumer market to test batteries by magnetism. It is conceivable that one day an iPhone or iPad can be placed on a test mat, similar to a charging mat, and read battery SoC and performance.


As a medical test at the doctor’s office or a weather forecast broadcasted through the media requires multiple data to derive a result, so also can no single instrument fully assess a battery. Several methods are needed. Voltage as medium to estimate SoC works best if the battery type is known and the pack has rested for at least four hours. A charge or discharge falsifies the voltage and the battery needs several hours to neutralize. Temperature also alters the voltage; it is lower when hot and higher when cold. Coulomb counting offers better SoC accuracy but requires periodic calibration. The coulomb method is not immune to battery aging and the accuracy decreases with time. Measuring the internal resistance gives vital battery information but it presents only a snapshot and cannot predict end-of-battery-life. The resistance of most batteries stays low while the capacity fades as part of aging. Measuring battery state-of-charge with magnetic susceptibility is promising but the technology is only in research laboratories. In time, new technologies will evolve that promise clearer insight into the mystery of a battery. This moment cannot come quickly enough.


On July 5, 2011 at 9:50am
Rob Van wrote:

The soon to be available LEAD/CARBON battery by AXION purportedly addresses the needs of EV applications including quick charging in only 20% the time , 70% the weight and much greater use of capacity with deep discharge and 2 or 3 times the life of traditional lead/acid batteries. Won’t this revolutionize the battery industry and therefore dramatically affect the performance, desireability, and practicality of EVs and dozens of other applications???

On July 5, 2011 at 4:08pm
shivakorn krieng ati-butr wrote:

I   feel the items useful both   for end-users &  for   service   technicians.
I need   more& deeper informations for vocational & academic   training   in school
since   actually   very   little people   know   really what   the battery is ,how   the
battery work   and   could not   verify   exactly   the   useful   life   of   the battery.
And since battery   becomes   energy storage   source for   many   electric   power
sources   i.e,  wind power, solar and water stream &  so on, most   vehicles   must
turn to be electric   motivated   as fossil energy   crisis also pollutions awareness,
not even hybrid , there ‘ll becomes   purely   electric[battery] driven   ones.

On July 6, 2011 at 6:41pm
Roger D. Rognas wrote:

As a former (retired) developer of remote telcom battery monitoring devices I found it most informative. Thank You.

On July 28, 2011 at 7:07am
Hans Welschen wrote:

Most lead acid batteries come in packs of e.g. 6 cells. Charging - discharging control is done with only the total pack information. In practise one cell, the worst one, determines overall performance but cannot be distinguished from outside. Do we need build-in Q-mag sensors per cell and how can we influence individual cell charge like with Lithium?

On August 15, 2011 at 9:27am
Philip Canard wrote:

As a recently retired biomedical engineering technician, I had much experience in the past with battery packs in portable medical devices. The most reliable method for guaranteed battery performance was to buy premium battery packs of recent manufacture and replace them on a regular basis based on history of past performance. Each device and its battery type and method of use determined frequency of replacement. I saw major differences between battery packs manufactured in USA/Europe/Japan and those from China. Low cost battery packs from distributors who sold on price instead of quality ended up costing more in the end. I can never remember a Hewlett-Packard supplied battery pack ever failing in a heart defibrillator as long as I observed their testing procedure or replaced it at 2 years, whichever came first. Their requirement was 90% of new rated capacity. My time requirement was to take the average time I saw major problems occur, and cut that time in half. All defibrillators had a new battery installed every 2 years even if well above 90% capacity, since 4 years was the expected time to failure of that battery chemistry.

On February 1, 2012 at 2:53am
John Fetter wrote:

Rob Van’s comments, (July 5, 2011), appear to be based on “spec. picking”. The batteries he describes are based on lead-acid technology but are functionally capacitors, with very large capacities.

And like capacitors, they can be charged quickly. And like capacitors they can be cycled many, many times. And like capacitors they can work effectively at a partial state of charge. But try using them as lead-acid BATTERIES and the self-same 130 year-old lead-acid shortcomings will once again prevail.

They are good at what they do but they are not good for everything.

On May 15, 2012 at 2:31pm
ron davison wrote:

Yes it will make a big impact.
Remember batteries suck so any significant improvement will have a big impact.
Not just a capacitor but also increases conductivity and reduce graidant build up all produce extra unwanted losses.
not sure if it will supplant lithum types or more anvanced types. But for cost sensitive applications (Can you say 10,000 dollar battery pack price tag?) and for making smaller lighter lead acid batteries for all standard application…slam dunk.

On June 13, 2013 at 9:35am
Kulchaya Tanong wrote:

I want to discharge all battery before i use for my experiment leaching. I have no idea how to do it. Do i need the machine for discharging or I can do it by something else?

On June 13, 2013 at 11:59pm
Hans Welschen wrote:


You can use any load, like a lamp or resistor to discharge the battery. Keep the initial current between C5 and C10. Most important: stop discharging at 1.75V per cel and recharge immediately after reaching that point. For this detection and switching to charge mode you might need some automation. Special equipment does excist but the investment is only worth if you use it frequently. Discharging the complete battery is not a guarantee all cells are discharged equally.

On June 14, 2013 at 11:55am
Kulchaya Tanong wrote:

Hans Welschen,

For me, i will get all financial support for the experiment. I think I will buy muiltimter..or something else to discharge battery autometically. Thank for your response smile.
In fact we did some battery cutting without discharging, it was quiet dangerous because all lithium when it reacts with an oxygen or humidity, hygroden gas will be produce…we need really speacial equipment and good planification.

On June 16, 2013 at 4:01am
Hans Welschen wrote:

Dear Kulchaya

Sorry for misunderstanding you original question. My answer was based on lead acid battaries and discharging for capacity measurements.

As you said, opening Li battaries is quite dangereous and should not be done without knowledge of the material and all protective measures required.

On September 25, 2013 at 9:41am
Jose Urena wrote:

If I’m a Distributor of batteries in my country, how the manufacturer should calculate my warranty claim percentage? They are only calculating the warranty percentage of sales and not considering the units in operations or units on field of previous years. Please help.

Thanks + Best Regards

On October 7, 2013 at 2:14am
Kabindra Kaphle wrote:

I want to repair inverter lead acid battery for using computer and lights. how to measure add sulfuric acid and distilled water?

On October 7, 2013 at 3:50am
Hans Welschen wrote:

Dear Kabindra,

If there was no spillage of acid from the battery you only have to top it up with destilled water. This, asuming you have a battery with removable caps. When it is a sealed or maintenance free version, topping up is not possible. Using a battery with low fluid level could have distroyed the battery. If the battery was discharged for a longer period, it can suffer from sulfation which is hard to remove if any. I wish you luck the battery is still in good health. If so, after topping up and fully recharging the acid density shouls read between 1.24 to 1.28 depending on the type.