Battery capacity testing: Implementing standards and field experience

Stationary lead-acid batteries play an increasing role in the industry today by providing power and back-up energy during emergencies. When an emergency situation occurs it is essential that the battery systems perform as designed, or the substation may be left unprotected, considering that the batteries provide power to the protection and control equipment. IEEE and NERC are active in the development of standards for battery testing.

Although many test methods can be performed to estimate the condition of the battery, the discharge test at specific time intervals and specific discharge rates are presented as the necessary tests, and the only method that can measure the real capacity of the battery. Nowadays the capacity tests are not performed because of the belief that this type of testing is too expensive, requires a lot of time, and it damages the battery. Most battery users today apply the internal resistance measurements in order to determine the battery state of health and save time.

This approach is not wrong. However the real capacity cannot be decided from this test. This article describes certain initial conditions, which should be met before the discharge tests, as well as how to save the time required by the discharge testing using the load bank for internal resistance measurements. The type of testing discussed in this article is the constant current discharge test along with the load bank and external load, as well as the discharge test using modified load profiles.

Fig. 1: Measured battery voltage as a function of time in five discharge tests on Powerfit S312/7S.

Test preparation

A test preparation is important to perform the test correctly. Before the test starts, it is necessary to prepare the equipment to comply with the safety standards. After that, it is important to visually inspect that cell connections are in a proper condition as well as that the charger has been disconnected. The load connected to the battery needs to be adequately backed up, or the current taken by the load needs to be included in the discharge rate. It is important to emphasise the test preparation and conditions before its start can affect the test results. Having that in mind, it is important to check the key parameters such as the cell voltage, float voltage, and electrolyte (cell) temperature before the test starts. Once the initial conditions are met, load unit and additional systems may be connected.

Myths and misunderstandings related to capacity testing

In battery testing, especially capacity tests, there are some facts that are misunderstood. Mistakes are made as a result of myths, misunderstandings and lack of training and experience.

Myth 1: Capacity testing ruins the battery and is expensive

This myth creates confusion among the test professionals. Excessive charging and discharging, with too many cycles will reduce the service life. However, when conducted in accordance with the applicable standards and manufacturer’s recommendations, the capacity testing will not reduce the battery service life and the consideration that this test is not needed is wrong.

The effect of the performance test on a battery was analysed on a Power-fit S312/7S battery. The discharge test was performed with higher discharge rate and reduced test time. The graph in the
Fig. 1 illustrates the test results collected from the battery with manufacturer declared five years of service. After being subjected to five discharge tests, the voltage values and the capacity levels at the last test were unchanged (even the voltage was bit higher in last test) compared to the results of the first test. This confirms that five more discharge cycles will not ruin the battery.

Besides the opinion that this kind of a test ruins the battery, the capacity test is not performed since it is considered as expensive, and the load bank is replaced by the hand held instruments for internal resistance measurements. The main reason is that the internal resistance would tell us the battery state of health. Although the internal resistance does correlate with a capacity, it is not a direct correlation. A capacity cannot be determined using this kind of a test. The internal resistance indicates there is a problem and this data needs to be confirmed with additional measurements, such as the capacity testing.

Fig. 2: Load profiles for internal resistance measurements using load banks.

For the team members that are involved in a battery testing, the capacity testing using load banks is recommended. In addition, internal resistance can be tested using a proper load bank which provides different load profiles settings. In order to measure the internal resistance using a load bank, it is necessary to establish a specific discharge characteristic and establish the two points. After 20 s discharge, at a defined current, the voltage is read and this provides the first point. Then, the discharge is interrupted and the battery is left in an open circuit state between two and five minutes. After the five seconds in the second discharge interval with a higher current level, the voltage value is recorded again.

The internal resistance is calculated using the following formula as per IEC 60896-21:

R_i=frac{}{U_1-{U_2{}}}{I_2-{I_1{}}}          [1]

Measurements based on this method were done on a Powerfit S312/7SR batteries while fully charged. The resistance that we calculated did not changed in five tests as well as the capacity as shown on the Fig. 3.

Additional tests were done on the Sonnenschein 2EPzV 200 battery. The battery was subjected to the capacity test in order to determine week cells in the string. After the discharge test the battery was recharged and subjected to the internal resistance test according to the above mentioned test method with load bank in order to see the differences in resistance between the cells that passed capacity test and the cells which failed the capacity test. Cell voltages during two discharge intervals were measured with a battery voltage supervisor system (cell by cell scanner).

The differences in internal resistance on cells with better capacity and on cells with bad capacity are show on the Fig. 4. The cell with lowest cell voltage at the end of the capacity test had the highest internal resistance.

Fig. 3: Internal resistance vs capacity on Power fit S312/7SR battery.

Myth 2: A battery needs to be equalised before a test

This statement depends on whether the battery system is new or it is in an existing system that is about to undergo a performance test. When a new battery is installed and the acceptance test is to be performed, a commissioning charge including an equalise charge should be performed. In that case the battery will be fully charged and ready for the test. In addition to that, a battery should be placed on float for at least three days. In every days practice batteries are tested from a float state.

Battery manufacturers publish their recommendations for a charging procedure in their installation, operation and maintenance manuals. Prolonged battery operation outside the manufacturers float voltage limits will reduce the life expectancy. For example, VRLA batteries do not need to be equalised before performance test unless it is recommended by the manufacturer.

Myth 3: Stop the test when the first cell reaches cell-end voltage

During the performance test it is not uncommon for a cell or a few of them to fail before the end of the test. Terminating a test as soon as the first cell fails may result in incomplete or wrong test results. The total string voltage should be used as a test terminating criteria.

Fig. 4: Differences between internal resistance on cells with higher capacity and failed cell.

However, a test person should immediately bypass a week cell except the test has run 90 to 95% of its course in which case the weak cell can remain in the string until the end of the test. During a test a cell voltage non-uniformity is present; some cells are above the end voltage value, some are below. Terminating the test when one cell reaches end voltage will not allow us to find other damaged cells in the string; which is the main reason for the test.

During the discharge test, one test team member should be prepared to bypass the cells. The bypass can be done by pausing the test, then bypassing the cell with appropriate connectors. All safety precautions should be followed during this procedure. The string voltage needs to be recalculated and the test continued with the new end-voltage value taking into account one cell less. The maximum allowed time to finish the bypassing process is six minutes or 10% of the total expected test time, and it should not be counted in the total test time.

Effect of temperature on discharge test

The temperature has a big influence on the cell performance. As the temperature of the cell increases, the internal resistance decreases and the chemical reaction increases what improves (increases) the capacity of the battery. However, keeping the battery warmer then recommended will cause a higher level of active material shedding, gassing, and grid corrosion that in turn would reduce the service life. This includes ambient and cell temperature. Battery manufacturers always publish recommended operating temperatures for their batteries. If the operating temperature of the battery system is above or below recommended, a correction factor needs to be applied after the test.

Test process

Once the initial conditions are meet, the load unit and other necessary test equipment needs to be connected and setup properly. The first step is to define the test current, end voltage, and the test duration. These parameters are found in the manufacturer’s published data and recommendations. The battery installation used for testing had 126 cells and the string voltage of 252 V. The manufacturer’s recommendations for the discharge test of this battery are shown in the Table 1.

Fig. 5: Total string voltage on the battery that needs to be replaced.

It is important to emphasise that current values are not adjusted for temperature. After completing the test and calculating the battery capacity it is necessary to perform a temperature compensation.

It is advantageous to use a powerful load bank or a load bank with an external current measurement feature, to measure and to include in the test the current that has been withdrawn from the battery by regular or any other additional connected load.

The terminal voltage was measured during the entire test as a function of time, and the battery should be discharged until it reaches the minimum terminal voltage, or once the total test time has been reached.

Coup de fouet phenomenon shown in the Fig. 6 represents a dip in the voltage drop present on a fully charged battery at the beginning of the test. This drop in the voltage and the maximal voltage after the recovery are dependent on the battery age, discharge rate, float voltage, and other operating conditions. Due to that partial discharge with controlled load discharge will cause the Coup the fouet. It can be considered as a useful tool in determining the battery state of health.

During the entire discharge test, it is important to monitor a behavior of the cells within the string by measuring the cell voltages at least three times during the test. One measurement should be taken at the beginning of the test and the others at specified intervals. The sample rate should be more frequent when the voltage starts to decrease faster. This approach makes it easier to spot a bad cell and predict failures well in advance.

Individual cell voltages can be measured manually by using a handheld voltmeters or using an automated monitoring system.

Note that in case of one cell voltage drops below the end voltage, the test should not be stopped and restarted once the failing cell was replaced, especially if it occurs early in the test. For example, if performing the 8-hour test, the week cell has to be bypassed within six minutes and the test continued. The general rule is a test interruption should be 10% of the test time if the test is interrupted for a longer time period and the string retested again, depending on the condition of remaining cells, some of them may fail and create more problems.

Fig. 6: String voltage as a function of time.

Capacity calculation methods

The battery capacity is usually expressed in Ah. However, the capacity at the end of the test is always expressed in a percentage of the manufacturers published capacity. IEEE standards define the two methods for capacity determination. Once one test method has been chosen, for trending purposes it should be used during the battery’s entire service life.

Time adjusted method

In the time adjusted method, the test current is kept constant as defined in the manufacturer’s published table as a function of the selected test time duration. The battery capacity is calculated after the completion of the test by using the published performance data at 20°C. This method is recommended for the test whose duration is longer than one hour.

The capacity is calculated as the ratio of an actual test discharge time to a published discharge time.

C=frac{T_a{}x100}{T_mx{K_t{}}{}}          [2]


C = Capacity at a specific temperature recommended by the manufacturer.

Ta = Actual time duration of the test until reaching the specified string end voltage

Tm =     Manufacturer’s rated time to reach the string end voltage

Kt = Temperature correction factor as per the IEEE standard.

This method is recommended for performance or acceptance tests whose duration is one hour or less. There are two options of this method based on the discharge rate for testing: adjusting the manufacturer’s recommended rating for the end of life condition or using the full recommended rate.

The adjusted manufacturer’s rate method is preferred since testing an aged battery at a high rate may result in a short runtime. In this method, a time duration is kept constant and the rate is adjusted based on the derating factor in order to simulate the end of life condition (80% of the rated capacity).

Following this, from the Table 1, for a 15 min test and derating factor of 0,8, the rate needs to be adjusted from 764 to 611 A. The test needs to be conducted with 611 A for 15 minutes until the end voltage is reached. The capacity is calculated using the following formula:

C=frac{left (X_a{}x K_c{} right )x100}{X_t}          [3]


C =  Capacity at the specified temperature

Xa =  Actual rate used for the test

Kc =  Temperature correction factor as per the IEEE standard

Xt =  Manufacturer’s provided rating for time to reach the specified end voltage

Table 1: Manufacturer published ratings at 20°C.

Replacement criteria

The recommended practice is to replace the battery if its capacity is below 80% of the manufacturer rating. Following the test, it is necessary to review the battery sizing to conclude if the remaining capacity is sufficient for the battery to perform intended function. A capacity of 80% indicates the battery (cell or string) rate of deterioration is increasing even there is still a capability to support the load.

Additional characteristics such as abnormality of the cell temperature and the cell voltage are often determinants for complete battery or cell replacements. The cell voltage is a good indicator for further investigation and cell replacement. In case we use replacement cells, it is important to ensure they have electrical characteristic compatible with the existing cells and they should be tested before installation. In case that only one or a few cells need to be replaced, the battery manufacturer should be consulted.


One of the main goals in the battery testing area, besides learning the real condition of the battery, is to reduce the test time. Several techniques and test systems are used today in order to achieve this goal. If we take a look at the battery manufacturer’s publications, there are two parameters provided in these documents that a user should be referred to. These parameters are the capacity and the internal resistance. The battery capacity is the only proven way to judge the condition of the battery during the entire service life and it does not depend on the measurement equipment that is being used. The internal resistance measurement technique is not standardised and the techniques are proprietary. This article offers the three general conclusions.

  • When conducted according to the applicable IEEE recommended practice and following the manufacturer’s maintenance instructions, the periodic performance testing will not ruin the battery and cannot be considered as excessive or unnecessary. Using load banks with extended features enables users to increase a loading capability, reduces a test time, and in addition, creates less wear and tear of the battery.
  • A load bank that enables creating different load profiles can be used for internal resistance measurements, in order to observe significant changes in a battery condition. Depending on the degree of a change, a capacity test interval can be planed.
  • According to regulations set forth by NERC and IEEE, the battery capacity testing is recommended and should be performed properly. It is important to emphasise that all generation as well as transmission companies must be compliant with these regulations to avoid penalties.

Considering all the above, it is highly recommended that users involved in battery testing have a load bank available.


[1]  IEEE 1188-2005: “EEE Recommended Practice for Maintenance, Testing and Replacement of Valve-Regulated Lead-Acid (VRLA) Batteries for Stationary Applications”, 2006.
[2]  IEEE 450-2010: “IEEE Recommended Practice for Maintenance, Testing and Replacement of Vented Lead-Acid (VLA) Batteries for Stationary Applications, 2011.
[3]  Tressler Rick, “Myths Misunderstandings and Mistakes, Battcon Stationary Battery Conference
[4]  IEC 60896-21: “International Standard for Stationary Lead Acid Batteries, 2004.

Contact Jacques Franken, Action Instruments, Tel 011 403-2247,

The post Battery capacity testing: Implementing standards and field experience appeared first on EE Publishers.

Source: EE plublishers

More news