How Does a UPS Decide That a Battery Is “Weak”?

A UPS cannot directly look inside the battery and determine its exact remaining capacity. Instead, it uses measurable electrical behavior to estimate battery condition.

Depending on the UPS design, a self-test may briefly disconnect the charger and apply a controlled load to the battery. The UPS then observes parameters such as battery voltage before the test, voltage drop under load, recovery voltage after the test and sometimes the estimated internal resistance of the battery bank.

The UPS firmware compares these measurements with an expected battery model. If the observed response falls outside the programmed range, the UPS may classify the battery as weak.

This approach can work well when the UPS is used with the type of VRLA battery for which the algorithm was originally designed. After converting to LiFePO4, however, the electrical response may be different enough to confuse the diagnostic logic.

LiFePO4 and Lead-Acid Batteries Have Different Voltage Curves

Lead-acid battery voltage changes noticeably with state of charge. UPS manufacturers can therefore use battery voltage as one part of their runtime and health estimation.

LiFePO4 has a much flatter discharge curve through a large portion of its usable capacity. The battery may remain near a relatively stable voltage for much of the discharge, then fall more rapidly near the lower SOC region.

This means that a UPS algorithm expecting a typical lead-acid voltage response may not interpret lithium voltage correctly. The UPS may report an inaccurate battery condition even though the LiFePO4 battery itself is operating normally.

Internal Resistance Is Also Different

UPS battery tests often pay attention to how much voltage drops when a load is applied. This relationship is influenced by battery internal resistance. A new LiFePO4 battery can have much lower effective resistance than an old VRLA battery bank. That might sound like it should always make the UPS happier, but the diagnostic algorithm may expect a particular voltage-response pattern rather than simply “lower resistance is better.”

The situation becomes even more complicated when the lithium battery contains an electronic BMS, current sensors, MOSFETs or contactors that were not part of the original passive lead-acid bank. The UPS is no longer testing the same type of electrical system it was designed to evaluate.

The BMS Can Change the UPS Self-Test Result

A LiFePO4 battery BMS monitors current, cell voltage and temperature continuously. During a UPS self-test, the load may change suddenly.

If the test creates a high current step, the BMS may temporarily limit current or, in a more severe case, enter protection. From the UPS perspective, battery voltage suddenly changes or disappears, so it concludes that the battery is weak. From the battery perspective, the BMS may simply be responding to an electrical event that exceeds its configured limits. This is why UPS lithium conversion must consider maximum discharge current, not only battery capacity in Ah or kWh.

A Battery Can Have Enough Energy but Not Enough Power

This distinction is critical for UPS dealers. Suppose a LiFePO4 battery has 5kWh of usable energy. That sounds sufficient for the customer’s backup requirement. However, the UPS may demand 6kW or 8kW during a self-test or high-load condition.

If the battery BMS is rated for only 100A on a 51.2V system, the approximate DC power available is limited. Once inverter losses and battery voltage are considered, the UPS may demand more current than the battery can supply. The battery therefore has enough energy, but not enough instantaneous power. A self-test failure caused by current capability should not be mistaken for a capacity problem.

Example: 51.2V 100Ah Battery on a 6kW UPS

A 51.2V 100Ah battery stores approximately: 51.2V × 100Ah = 5.12kWh

At first glance, this seems suitable for a 6kW UPS because the battery has more than 5kWh of energy.

However, if the UPS draws 6kW from the battery, DC current can exceed: 6,000W ÷ 51.2V ≈ 117A

After considering inverter losses and lower battery voltage under load, actual current may be even higher. If the battery BMS allows only 100A continuous discharge, the UPS can trigger BMS protection even when SOC is 100%. The problem is not battery capacity. It is battery power capability.

A UPS Battery Test May Use a Large Percentage of Rated Load

Different UPS models use different self-test methods. Some perform only a brief light-load test, while others may place a more significant demand on the battery.

For this reason, two UPS models with similar output ratings may behave very differently with the same lithium battery. A battery dealer should not promise universal compatibility merely because nominal DC voltage matches. The UPS manufacturer, model, DC bus range, load level and battery-test behavior all matter.

“Replace Battery” Can Also Be a Maintenance Timer

Some UPS products generate a replacement warning based partly on battery age or service history rather than real-time battery performance. If the original VRLA batteries were installed several years ago, the UPS may have a stored battery-installation date or replacement timer.

After installing LiFePO4, the technician may need to reset the battery replacement date or maintenance counter according to the UPS manufacturer’s procedure. If this is not done, the UPS can continue showing “Replace Battery” even though the old batteries have physically been removed. This is one of the simplest causes and should be checked before advanced troubleshooting.

Battery Replacement Date and Battery Health Are Different

A maintenance reminder does not prove that the UPS has measured a battery fault. Dealers should first determine whether the message is a scheduled replacement notification or the result of a failed electrical self-test.

These two alarms require completely different responses. The UPS event log or service software can often provide more detail than the front-panel message.

Battery Voltage May Be Outside the UPS’s Expected Range

A lithium battery can be electrically functional but operate at a voltage that the UPS does not expect. For example, the UPS may have been calibrated around a specific number of lead-acid cells and may expect a certain float voltage, battery-low voltage and open-circuit voltage.

If the LiFePO4 replacement operates slightly differently, the UPS may interpret the voltage as abnormal. This is another reason that “48V UPS” and “51.2V LiFePO4 battery” should not automatically be considered compatible. The complete DC operating range needs to be checked.

Charging State Can Affect the Self-Test

Suppose the new LiFePO4 battery has just been installed at 40% SOC. The UPS charges it for one hour and then automatically performs a battery test. The battery may not yet contain enough energy for the expected test duration.

The customer sees a “Battery Weak” warning and assumes that the new battery has failed. For commissioning, allow the battery to reach an appropriate SOC before performing a meaningful runtime or self-test assessment. The required condition should follow the battery and UPS manufacturer’s instructions.

SOC Display Can Be Misleading After Retrofit

A legacy UPS may estimate SOC from voltage. The lithium battery’s BMS may estimate SOC using coulomb counting.

The UPS might therefore show: 100% Battery while the BMS shows: 65% SOC or vice versa.

If a self-test begins based on the UPS’s incorrect SOC estimate, the result can be confusing. When troubleshooting a lithium retrofit, record both UPS battery status and actual BMS SOC where possible.

Why a Full-Capacity Battery Can Still Trigger Low Voltage

Under high UPS load, battery terminal voltage can decrease because of internal resistance, cable resistance, breakers, connectors and busbars. If the UPS has a fixed low-battery voltage threshold, this temporary voltage sag can make the UPS conclude that the battery is nearly empty.

Once the load disappears, battery voltage recovers. The customer then measures the battery and says: “The voltage is normal, so why did the UPS say battery low?” The measurement was taken after the critical load condition had already disappeared. Voltage needs to be measured while the UPS is actually on battery under load.

Cable Resistance Becomes Very Important at UPS Current

UPS applications can demand much higher DC current than many solar applications. For example, 5kW at approximately 50V means around 100A. At 10kW, current can approach 200A. Even a small resistance in cables or lugs creates significant voltage drop at these currents.

A battery can therefore be perfectly healthy while the UPS sees insufficient voltage at its own DC terminals. Check cable cross-sectional area, total cable length, lug quality, torque, fuses and breakers before replacing the battery.

One Loose Connection Can Trigger the Battery Test Alarm

A slightly loose terminal may operate normally under light load. During a high-current UPS test, its resistance causes a large voltage drop and local heating. The UPS sees DC voltage collapse and fails the battery test.

After the test ends, voltage returns to normal and the loose connection becomes difficult to detect without a thermal or voltage-drop inspection. This is why battery-test failures should be investigated under realistic load. A visual inspection alone may not identify the fault.

Parallel Batteries Can Create Additional Confusion

Suppose four LiFePO4 modules are connected in parallel. The UPS expects a large battery bank. If one module is switched off, in protection or poorly connected, the remaining batteries still provide normal bus voltage. At light load, everything appears fine.

During the UPS self-test, however, the remaining modules carry much more current each. One or more batteries may then reach their BMS current limit. The system fails the test even though the total displayed voltage before the test looked normal. Dealers should confirm how many modules are actually online.

BMS Logs Can Reveal What the UPS Cannot

When a self-test fails, BMS event data can be extremely valuable.

Look for events such as:

• Discharge overcurrent
• Cell undervoltage
• Pack undervoltage
• High temperature
• Low temperature
• Contactor/MOSFET protection

If the BMS logs a clear protection event at the exact time of the UPS test, the problem is not simply an inaccurate UPS maintenance timer. The battery or system exceeded an operating limit. The next step is to determine why.

What If the BMS Shows No Alarm?

If the UPS reports “Battery Weak” but the BMS shows normal voltage, normal current, adequate SOC and no protection event, the alarm may come from the UPS’s own battery-health algorithm. In that case, check the UPS service manual for lithium compatibility, battery-type settings, battery-capacity configuration and self-test behavior.

Some UPS platforms support lithium batteries only with dedicated firmware or battery communication. Others may operate electrically with lithium but continue to provide inaccurate battery-health reporting.

The UPS Battery Capacity Setting Can Matter

Some UPS systems allow technicians to configure the installed battery capacity or number of battery strings. If the original system used a small VRLA bank and the new lithium battery provides much more energy, the UPS may still calculate runtime and battery health using the old configuration.

Likewise, if the configured battery-bank size is larger than the actual lithium installation, the UPS may expect more runtime than the battery can deliver and declare the test unsuccessful. After a retrofit, review every battery-related configuration parameter rather than changing only the physical batteries.

Automatic Runtime Calibration Can Help—When Supported

Some UPS equipment includes a runtime-calibration function. This intentionally discharges the battery under controlled conditions so the UPS can improve its runtime estimate. Whether this function should be used with a particular LiFePO4 battery depends on UPS compatibility and battery supplier guidance.

Do not run repeated full discharges merely to force the UPS to learn the battery. LiFePO4 does not need unnecessary deep cycling for routine maintenance. Use only manufacturer-approved calibration procedures.

Do Not Ignore a Real Battery Alarm

It is equally important not to dismiss every UPS warning as a false lithium alarm. If the battery actually has a weak cell, poor connection or BMS fault, the UPS self-test may be doing exactly what it was designed to do.

A useful troubleshooting sequence is to check UPS event logs, BMS alarms, battery voltage under load, DC current, terminal voltage drop and actual delivered runtime. A diagnosis based only on chemistry differences can miss a genuine problem.

A Dealer-Friendly Diagnostic Procedure

When a customer reports “Replace Battery” after a LiFePO4 upgrade, collect the following information before approving a replacement:

Item	Why It Matters
UPS brand and model	Determines test logic and compatibility
Original battery bank	Shows what the UPS was designed for
LiFePO4 model/quantity	Determines voltage, energy and current
Actual UPS load	Determines battery current
UPS alarm/event log	Separates timer from failed test
BMS SOC	Checks real battery state
BMS alarms	Identifies protection events
Battery voltage under load	Reveals sag
Cable size/length	Identifies DC losses

This small checklist can eliminate many unnecessary warranty replacements.

For UPS Distributors: Test the Combination Before Selling It Widely

If a distributor plans to promote one LiFePO4 model as a replacement for a popular UPS range, test the combination before ordering large quantities. Verify normal charging, full-load discharge, automatic battery test, low-battery shutdown and grid recovery.

Also check whether the UPS shows correct battery status after installation. A technically compatible battery that continuously generates nuisance alarms can still create a poor customer experience. Compatibility should include user-facing behavior, not only whether the UPS can run.

For Battery Importers: Build a UPS Compatibility Database

Battery distributors can create significant competitive value by recording which UPS models have already been tested.

For each model, document:

• Battery configuration
• Charger voltage
• UPS battery settings
• Maximum tested load
• Self-test result
• Runtime behavior
• Known limitations

This turns technical support into a sales tool. A dealer is more likely to buy from a battery supplier who can say, “We have already tested this UPS model,” than from one who simply says, “48V should work.”

Conclusion

A “Replace Battery” or “Battery Weak” warning after a LiFePO4 retrofit does not automatically mean the lithium battery is defective.

The alarm can result from UPS self-test algorithms, old maintenance timers, incorrect battery configuration, lead-acid-based voltage assumptions, high load current, cable voltage drop or BMS protection.

The correct response is to identify what the UPS actually detected before replacing the battery. For UPS dealers and battery distributors, testing lithium compatibility at the level of charging, discharge, self-test and runtime reporting can dramatically reduce unnecessary after-sales claims.

HIZN supplies LiFePO4 batteries for UPS backup, telecom and stationary energy storage applications with multiple capacity and BMS configurations.

If your customer wants to replace a UPS lead-acid bank with LiFePO4, send us the UPS brand/model, original battery configuration, load power and required backup time. We can help evaluate the battery voltage, capacity and current requirements before the retrofit.