EXBATT
    EXBATT
    All insightsInsight

    What a battery test report tells you

    5 min readLast reviewed: September 2026
    Short answer

    A capacity discharge test is the only measurement that proves how much energy a battery string still delivers, expressed as a percentage of rated capacity. An impedance test does not prove capacity; it finds individual weak cells quickly and without taking the system offline. A useful report states delivered capacity, per cell readings against a baseline, temperature, test conditions, and a clear replacement recommendation with a date.

    Written for the person who signs off the maintenance budget rather than the technician holding the instrument. What testing proves, what it costs in downtime, and how to read the report you are given.

    The two tests do different jobs

    A load bank capacity discharge test discharges the string at a defined rate to a defined end voltage and measures what it actually delivered. This is the only method that answers the question everyone cares about: will the autonomy be there when the grid fails. Internal impedance or conductance testing measures each cell against a baseline and identifies outliers in an hour or two, while the system stays available. Impedance trending tells you which cells are drifting. Capacity testing tells you whether the string still does its job.

    Capacity test versus impedance test
    AspectCapacity dischargeImpedance or conductance
    Proves autonomyYesNo
    Finds single weak cellsIndirectlyYes, quickly
    System availability during testReduced or on bypassMaintained
    Time on site4 to 12 hours plus recharge1 to 3 hours
    Typical intervalCommissioning, then 2 to 5 yearsEvery 6 to 12 months

    What the downtime really is

    Capacity testing does not require the protected load to stop. On a UPS with a maintenance bypass, or a DC system with a second string or a temporary supply, the load stays up while the string under test is discharged into a load bank. What the site gives up for the duration is redundancy, not power. The window to plan for is the test itself plus recharge, so the string is back at full autonomy before the risk period ends.

    How to read the report

    Look for five things. Delivered capacity as a percentage of rated capacity, at a stated discharge rate and temperature. Per cell voltage and impedance readings with the baseline alongside. The cells flagged as outliers, by position. Ambient temperature during the test, because it changes the result. A recommendation with a date: continue, retest earlier, replace specific cells, or replace the string. A report that gives readings without a conclusion has moved the decision back to you without giving you the basis for it.

    The 80 percent rule and what follows it

    A string is normally considered end of life when delivered capacity falls below 80 percent of rated capacity, because degradation accelerates after that point and the autonomy you designed for is no longer there. Below 80 percent, plan replacement rather than another test cycle. Between 80 and 90 percent, shorten the interval and watch the trend. Replacing individual cells is reasonable early in life, and a false economy late in life, because the new cells are then working against old ones.

    Frequently asked

    Do you test systems supplied by other companies?
    Yes. Testing is a standalone service and a large part of it is performed on third party installations, including data centers, substations and telecom operators.
    How often is testing actually necessary?
    Impedance trending every 6 to 12 months, capacity testing at commissioning, then every 2 to 5 years depending on criticality, and annually once the string passes 80 percent of expected service life.
    Is a voltage reading enough?
    No. A failing cell often shows a normal float voltage and collapses under load. Voltage alone tells you the charger is working, not that the battery will deliver.
    Can testing be tied to a maintenance contract?
    Yes, and that is the usual arrangement: scheduled visits, trending between visits, capacity testing at the agreed interval, and replacement planned before autonomy is lost.