Batteries are not ‘fit-and-forget’ assets!

Author: Megger North America Technical Support Group (TSG)

Regular testing of storage batteries, particularly those used to provide emergency supplies, is essential. The batteries often sit unnoticed and unused for long periods, and they give little outward indication of deterioration or failure. Yet if they fail to perform as expected when called upon to do so, the result can be catastrophic.

The two most widely adopted approaches to assessing battery condition are impedance testing and discharge testing. Impedance testing is an on-line procedure that can be carried out frequently to identify individual weak cells before they fail. This test estimates the performance that can be expected from the battery in its current condition. It provides valuable information, but the results are always ‘best estimates’ rather than a definitive evaluation.

In contrast, the discharge test, which is also known as a load test or a capacity test, is an off-line test that measures the actual output of the whole battery string. It is the only test that can accurately measure the true capacity of a string, and for this reason, it is required by IEEE standards. A discharge test reveals what will actually happen if the battery is required to take the load.

Relevant standards are IEEE 450-2002 Recommended Practice for Maintenance, Testing, and Replacement of Vented Lead-Acid Batteries for Stationary Applications, and IEEE 1188-1996 Recommended Practice for Maintenance, Testing, and Replacement of Valve- Regulated Lead-Acid (VRLA) Batteries for Stationary Applications.

Because they are time consuming and they require the battery to be taken off-line, discharge tests are, in most applications, performed infrequently. Typically, it is recommended that this type of test be performed in any of the following conditions:

• When the battery is new, as part of the acceptance test.
• Within two years of the initial test, for warranty purposes.
• Subsequently, as a minimum, every 25 % of the battery’s expected service life or every 6 years, whichever is the shorter interval.
• Annually, when the battery has reached 85 % of expected service life, or if the capacity has dropped more than 10 % since the previous test, or is below 90 % of the manufacturer’s rating.
• If the impedance value of the battery has changed significantly.

Concerns are sometimes expressed that discharge testing reduces the life of a battery: in fact, it has been called destructive testing because weak cells may fail during the test. It is, however, better to discover these weak cells during a test than when the battery is required to supply its load!

In theory, the test does indeed slightly shorten the life of the battery. However, a typical battery will have a life of at least 1000 charge/discharge cycles, and discharge tests are likely to be performed only four or five times over the battery’s entire life. Since this is such a small percentage of the total available charge/discharge cycles, the impact on battery health and overall life, in practical terms, is negligible. In reality, it is far better to know the true capacity of the battery and to confirm that it will actually support the required load, than to worry about the minuscule effect that load testing may have on overall battery life.

Problem-free discharge testing

While discharge testing is the only true test of the capacity of a battery string, it undeniably requires a considerable amount of time and effort, hence it is important to make sure that it proceeds smoothly and without the need for re-runs. The following steps will help to ensure that this is achieved:

1. Make sure that the battery (or batteries) to be tested has been maintained in its fully charged condition (typically by float charging) for at least 72 hours before starting the test. This will ensure that the discharge test results accurately represent the battery’s capacity.

2. Carry out an impedance test and measure the resistance of the inter-cell connections before starting the discharge test. This will ensure that the electrical path in the battery string has been checked thoroughly before high current discharge commences.

Figure 1: Sample battery discharge test specifications sheet

Figure 2: Current capacity vs voltage for the TORKEL 900 Series

3. Decide on the type of discharge test to carry out. There are many different types of discharge test including constant current, constant power, constant resistance, and load profile. Constant current is the type of test performed most frequently.

4. Check the discharge test specifications for the battery under test. This will help with planning the test. The specifications will include the end cell voltage (which is typically 1.75 V or 1.8 V per cell for lead-acid batteries) and a table of discharge rates. Using the table, the test duration can be chosen based on the duty cycle of the battery and this will allow the corresponding test current to be determined. As an example, with the table shown in Figure 1, a test current of 19 A would be needed for a 5-hour discharge test on the selected battery model.

5. Arrange for a backup battery bank if needed. A backup battery bank can be used to supply the load while the battery string under test is off-line. The backup battery will also be needed after the test is completed to allow time for the string which has been tested to be recharged.

6. Make sure that the load bank can handle the required test current. With high test currents, a single load bank may not be sufficient. This issue can be addressed by using additional load banks connected in parallel, or by using a lower test current and increasing the duration of the test. For load banks in the TORKEL 900 series, comprehensive information about discharge capability is given in the data sheet (see Figure 2). As a further aid, the TORKELCalc software package can be used to determine the configuration needed to suit a particular discharge current.

Figure 3: Voltage sense leads connection

Figure 4: Discharge test setup with BVMs for cell voltage measurement

Figure 5: Screenshot from the TORKEL GUI showing the test limits

7. Make the test connections safely while the battery to be tested is still connected to the charger. The connections need to be properly made to ensure that the high current flowing during the test does not lead to excessive heating. The battery terminal voltage can be measured accurately by using separate voltage sense leads, as shown by the dotted connections in Figure 3. This arrangement eliminates the effect of voltage drop in the current leads between the test set and the battery under test.

8. Monitor individual cell voltages. Bad cells in a string can discharge much faster than good ones. To allow the discharge test to continue, bad cells may need to be bypassed to avoid effects like polarity reversal. It is therefore important to monitor the voltage of each individual cell in the battery string while the discharge test is being performed. This can be done with battery voltage monitor (BVM) accessories, as shown in Figure 4. The correct voltage probes should be used to ensure that the connections to individual cells can be made easily.

9. Program the test parameters in the discharge test set. These include the test method, capacity calculation method, test temperature, test current, test duration, nominal capacity (test current x test duration), warning limits, and stop limits. A warning limit could be set for the individual cell voltage (for example at 1.75 V per cell). In addition, a stop limit could be set for the battery voltage (for example, 1.75 V per cell x 24 cells = 42 V). Examples of these settings on a TORKEL test set are shown in Figure 5. An additional warning limit could be set at a voltage slightly higher than the end battery voltage, so that the person performing the test is alerted when it is almost complete.

10. Be aware that some cells will reach the end voltage earlier than the others. The discharge test discharges all cells, and inevitably some will discharge sooner than others. The test should not be stopped when one cell reaches the end cell voltage, rather it should carry on until the average cell voltage is equal to the end cell voltage. For example, if the end cell voltage for the battery under test is 1.75 V and the battery has 60 cells, the test should continue until the battery voltage is 60 x 1.75 V = 105 V. At this point, it is perfectly possible that some cells will be at 1.8 V while others are at 1.6 V.

11.Be ready to bypass bad cells. Some cells in the battery string will discharge faster than others. IEEE test procedures for lead acid batteries (VLA and VRLA) state that the discharge test may be paused once for bypassing cells that are nearing polarity reversal. The maximum permitted duration of this “downtime period” is 10 % of the test duration or 6 minutes, whichever is shorter. After the bypass, the end battery voltage needs to be adjusted based on the remaining number of cells in the string. It is also necessary to evaluate the need for bypassing cells. If there are only a few bad cells in a string, the test can continue, but if, for example, half the cells in the string discharge prematurely, the test should be stopped and the battery replaced.

12. Record the float voltage for each cell. Having BVMs connected makes this much easier. For a lead-acid battery, the float voltage will typically be around 2.2 V.

13. Prepare to start the test. Turn off the charger, disconnect the load, and if necessary, transfer it to the backup battery bank (testing a battery with the load connected is possible, however, by using an accessory CT to measure the external current flow).

14. Start the test.

15. Monitor the discharge data to ensure the test is progressing smoothly. Real-time data capture makes it possible to view live test values and evaluate progress in relation to the programmed test limits.

16. At the end of the test, save the discharge data and note the percentage capacity. For a discharge test that runs for one hour or longer, the following formula can be used to calculate the percentage capacity:

The calculated time to reach the end voltage should be available in the battery manufacturer's data (see Figure 1). The manufacturer may also provide the temperature correction factor but, if not, the values provided in IEEE 450 can be used.

17. Reconnect the battery to the charger. Note that the charging current will initially be high as the battery has been heavily discharged during the test. The charger will need to be in good condition to supply the current required.

A sound investment

Batteries are costly assets that play an important and often critical role in modern power systems. It is therefore essential to ensure that they are maintained in good order and that their performance is regularly and accurately assessed. The key to achieving this is to implement a program of testing that includes both routine impedance testing and, at less frequent intervals, carefully planned discharge testing. Modern test instruments, such as those in Megger's battery test portfolio, deliver dependable results and make both types. of testing easier to perform. For all battery users, such testing is a sound investment that will yield an excellent return, not least by helping to eliminate the risk of batteries failing to perform when they are needed most.

Watch our webinar on 'Introduction to battery Testing Bite, BGFT and Torkel'