Battery-capacity testing is an essential part of battery maintenance. But how do you do it effectively?

How do you know your UPS will have "enough battery" to provide the electricity your operations demand when you lose your primary power source? In theory, you add up the nameplate ratings on your battery strings and come up with an answer. The reality is this answer will invariably be wrong. Because cells age, battery ratings are always greater than their real capacity. The best way to get an accurate answer is to perform capacity tests on your batteries. Let's begin by answering a basic question.

What is a battery's capacity? In a nutshell, battery capacity is a measure of the energy it can store. When you test for capacity, you're looking at the battery's ability to deliver a specified amount of current at a constant rate to a specified end voltage for a specified time. We determine capacity by filling the battery with energy and then seeing how long it takes to empty that energy.

An analogy for this is trying to determine how much water an odd-shaped tank will hold. You don't know how much water is already in the tank, nor what the inside of the tank looks like. So, you fill it up completely. Then, you open a tap on the bottom of the tank and measure the amount of water that comes out. This tells you the capacity of the tank. If you couldn't capture the water, you would put a flowmeter on that drain line and monitor the flow rate over time to calculate the capacity. This is a reasonably close analogy to how you measure battery capacity.

The main benefits of testing are:

- You determine where the battery is on its predictable life curve.

- You know when to replace a battery.

- You locate weak cells and faulty intercell connectors.

Testing frequency. How often should you do a capacity test? The Sidebars, on pages 94 and 95, show standards that seem fine when viewed separately, but create confusion when seen side-by-side. The reasons for the discrepancies go beyond the technical, but revisions are not far off. While these standards give you some idea of how often to test, they don't give you clear direction. However, case histories and field test data provide compelling support for the test-frequency plan shown in the Table, on page 94. Still, this plan is just a guide. The best way to determine when to do capacity testing is by monitoring the battery with internal resistance measurements. This reduces unnecessary testing without causing you to miss a test when you should make it.

The most effective testing frequency plan is one that is condition-based rather than calendar-based. Such a plan compensates for all kinds of variables that affect battery life and performance, while a calendar-based plan depends on assumptions that may not be true for your situation. Internal resistance measurements will show you there's a potential weakness in the battery; a load test will show you whether that weakness is severe enough to require action.

A capacity test reduces battery life. However, if you do it only occasionally, it will create an acceptable loss in exchange for knowing the condition of the battery. You do the test because you are more concerned with optimizing the reliability of the battery - knowing it will perform when you want it to - than optimizing its longevity. When a battery can't provide the backup time you need, then how long that battery will continue to provide any power is irrelevant to your operations. Let's see how to do the test.

What the test entails. To perform a load test, you subject the battery to a constant current load while measuring the string overall voltage and the individual cell voltages. You compare the time it takes for the voltage to decline to a predetermined level with the manufacturer's rated time. Then, you use the following equation to grade your capacity so the results are more meaningful.

Capacity = T subscript actual/T subscript rated x 100%

What size load should you use? Ideally, it will be as close to the real application as possible. Make sure you test UPS batteries with a constant power load. To preserve battery life as much as possible, use a higher discharge rate to give you a shorter test time. What end voltage you should use depends on the test equipment available and whether you test the battery online or off-line. Select the current level you will maintain from the manufacturer's table or load test curves.

You can run four kinds of tests:

- Full string test off-line using load banks: This applies to facilities such as power plants where you can't use an actual load.

- Full string test online using load banks: A typical application is substation testing.

- Single cell test online: This applies frequently to central office testing.

- Rundown test using actual load: You can supplement the actual load with a load bank, to control the desired discharge rate.

An ideal UPS test uses an AC load bank connected to the output of the inverter, while the batteries power the inverter. This test exercises the batteries and most of the UPS power components.

Each test requires four steps:

- Pre-test. Ensure hardware is calibrated and working. Connect all test leads and load modules. Program test parameters. Ensure spare parts and jumpers are available. Prepare plan of action in case of cell failure(s). Measure ambient or electrolyte temperature and compensate the load to be applied, using the applicable IEEE standard.

- Testing. Verify all voltage readings are correct, then disconnect charger and start testing. Within the first 10 sec to 20 sec, verify the test current is correct. Then, look for low cell voltages. If a single cell voltage is more than 30mV below all the others, it requires investigation. Identify the problem source (cell or intercell), and then decide if you must interrupt the test. The middle of the test is typically uneventful. Toward the end of the test, cell voltages are dropping fast, and you must watch them closely. Once cell voltage is below 1.75V, it will drop rapidly. At this point, stop the test before it can go into reversal. If a cell fails early in the test, you can momentarily halt the test and bypass that cell. You can halt the test only for 10% of the total test time, or 6 min - whichever is shorter. You normally halt a load test, manually or automatically, for one of the following reasons:

1) Overall voltage reached the end voltage,

2) At least one cell is failing, or

3) Test time exceeds the value equal to a 100% capacity battery.

- Post-test. Verify the load is properly disconnected. Reconnect the charger and verify charge current is correct and all cell voltages are recovering normally. Disconnect all load test cables. Disconnect all sense leads. Print out test report.

- Analyze data. Calculate battery capacity. If capacity exceeds 80%, the battery passed. Otherwise, it failed. Evaluate individual cells for replacement.

As with all tests, battery load testing gives you a snapshot of what you see now. It does not certify the battery will perform as expected for the next 12 months regardless of improper maintenance. However, if you maintain your batteries per industry standards and manufacturers' recommendations, load testing allows you to know when a cell needs replacement. That's important information whether you are settling a warranty issue or ensuring reliable backup power.