LiFePO4 cells are out of balance when the spread between the highest and lowest cell voltage climbs past what the pack normally holds. A large spread does not automatically mean a defective cell. A small spread does not guarantee equal capacity. At rest, a healthy pack typically holds 20mV to 50mV. A spread of 50mV to 100mV needs monitoring, and one above 100mV that never closes needs a controlled top-balance cycle. Diagnose first, balance second. Fix the cause behind the difference, not the number.
The most common mistake is reading one voltage snapshot and calling the battery faulty. Voltage readings move with current, temperature, internal resistance, measurement accuracy, and the time since the last charge or discharge. A reliable diagnosis needs operating context, repeated measurements, and a look at how each cell behaves under load. This guide walks through the six mechanisms behind cell imbalance, an 8-step diagnostic sequence, real documented case studies, and the criteria for balancing, further testing, or cell replacement.
1. What Cell Voltage Spread Actually Is
LiFePO4 cell voltage spread is the difference between the highest and lowest measured LiFePO4 cell voltage in a pack at one moment:
Consider a 16S LiFePO4 battery, the common 51.2V home storage layout of 16 LiFePO4 cells:
| Parameter | Example reading |
|---|---|
| Highest cell voltage | 3.430V |
| Lowest cell voltage | 3.370V |
| Cell voltage spread | 0.060V (60mV) |
| Series configuration | 16S |
| Nominal pack voltage | 51.2V |
A 60mV spread in LiFePO4 cells is an observation, not a diagnosis. Whether it is acceptable depends on the manufacturer's limits, the operating conditions, and the behavior of the individual cells. A battery can show a measurable spread and work normally. Two LiFePO4 cells can also read the same voltage while holding different capacity or internal resistance. Published battery-pack research shows that capacity differences limit energy utilization in series packs, while resistance differences affect power performance. The spread is the starting point, never the verdict.
Why the spread matters comes down to the weakest-cell effect. Every LiFePO4 cell in a series string carries the same external current, but cells do not share identical capacity, state of charge, or resistance. During charging, the cell that reaches its upper limit first can force the BMS to cut or reduce charging while the others still have room. During discharge, the weakest cell can hit the lower limit first and shut the system down. One cell, not the sum of the pack, decides how much energy you can actually use.
2. Why a Spread Does Not Equal a Bad Cell
LiFePO4 cell chemistry makes voltage a poor ruler for state of charge through most of its range. The open-circuit voltage curve stays nearly flat between roughly 20% and 80% SOC, typically 3.20V to 3.35V per cell. Inside that window, the voltage slope is under about 0.5mV per 1% SOC. Two LiFePO4 cells with meaningfully different SOC can display nearly identical voltages, and two cells with equal voltage can hold different usable capacity.
Near the top and bottom of the usable range, the curve steepens. Small SOC differences produce larger voltage differences there, which is why imbalance becomes visible at the end of charging or discharging instead of in the middle of normal use. Research on the LiFePO4 voltage plateau and its hysteresis confirms that SOC estimation from voltage alone is unreliable in this chemistry.
Three diagnostic principles follow from this:
- Record the charging or discharging state of every LiFePO4 cell. A spread measured at 50A discharge is not comparable with one measured after rest.
- Compare measurements under similar conditions. Temperature, current, and recent history should match as closely as practical.
- Watch the trend, not a single number. A spread that grows repeatedly under the same conditions deserves attention; a transient difference that vanishes when current stops does not.
There is no scientifically defensible universal spread threshold that classifies every LiFePO4 pack as healthy or defective, and no single number fits every LiFePO4 cell. Limits depend on cell specification, pack design, BMS configuration, and measurement conditions. The honest question is not "is the voltage equal" but "does this LiFePO4 cell behave like its neighbors under the same conditions."
3. The Six Mechanisms Behind Cell Imbalance
Different causes produce similar-looking readings, and each needs a different fix. Confusing them is the most common diagnostic error.
3.1 Different states of charge
Cells can start with different SOC, or their SOC can diverge through cycling from initial charge conditions, self-discharge variation, or incomplete balancing. A BMS can correct part of this difference in LiFePO4 cells if its balancing function is active under the right conditions. Because the voltage curve is flat, equal voltage does not prove equal SOC, and balancing restores charge alignment rather than cell health. The LiFePO4 SOC troubleshooting guide explains how SOC estimates drift on this chemistry.
3.2 Different cell capacity
Nominally identical cells are not perfectly identical. Manufacturing variation and aging produce real capacity differences. Take two series LiFePO4 cells: Cell A holds 100Ah, Cell B holds 90Ah, and both start near the same SOC. Under a 45A discharge for one hour, each loses 45Ah. Cell A loses 45% of its capacity; Cell B loses 50%. Cell B reaches the low-SOC region first, its voltage falls sharply near the discharge limit, and it becomes the pack's limiting cell. The same effect appears during charging, where the smaller LiFePO4 cell reaches its upper limit first.
The diagnostic point is hard and important: balancing can correct a difference in charge state, but it cannot restore lost capacity. A cell with permanently reduced capacity stays the limiting cell even after its voltage is balanced. Cell grading explains why fresh cells vary, and our guide to grade A LiFePO4 cells covers the tolerance spread at the factory.
3.3 Different internal resistance
Cell voltage moves with current. A simplified electrical model describes the effect:
V(charge) ≈ V(OCV) + I × R
V(OCV) is the open-circuit voltage, I is the current, and R is the effective internal resistance. These are approximations; real cells also show dynamic polarization, hysteresis, and temperature dependence. A LiFePO4 cell with higher resistance drops more under discharge and rises more under charge. It can make a LiFePO4 cell look like the weakest in the pack under load, and then recover when the load stops. That is a resistance pattern, not an SOC imbalance.
A worked example shows the scale. Two LiFePO4 cells with a 1 milliohm difference in effective resistance, carrying 50A, produce a resistive voltage difference of 50mV before any other effect is counted:
Resistance mismatch matters most for power performance, and it becomes more visible at higher currents.
3.4 Temperature differences
LiFePO4 cell temperature changes resistance, reaction rate, voltage response, and available capacity. A colder cell behaves differently under the same current, and a persistently warmer cell deserves investigation if it runs away from its neighbors. Temperature differences are a reason to compare voltage with temperature data, not proof of a faulty cell.
Charging limits must come from the actual battery specification. Most LiFePO4 batteries restrict charging below 0°C unless the system is built for low-temperature charging. Charging a cold cell outside its limits can damage it. Never "fix" a voltage spread by disabling temperature protection, because the protection exists to prevent lithium plating and capacity loss. The temperature-aware charging guide covers this in detail.
3.5 Deep discharge and uneven usage
Excessive discharge is a documented trigger for serious imbalance in LiFePO4 cells, especially when one cell reaches its low-voltage limit before the pack voltage looks unusual. Small continuous loads also matter. Control electronics and standby equipment keep draining a battery after normal operation stops, and a pack left in that state drifts faster.
A field-service bulletin for a 48V LiFePO4 system documented LiFePO4 cell imbalance linked to over-discharge. Its prescribed inspection measured the cell-voltage difference with the battery idle instead of diagnosing from an active charging battery alone. A battery that recently survived a deep discharge should not be treated as a routine balancing case; check its history first.
3.6 Incorrect voltage readings
Not every apparent imbalance in LiFePO4 cells comes from inside the cells. Faulty or poorly connected cell-sense wiring, wrong BMS channel mapping, calibration drift, loose sensing connections, and intermittent contacts all produce fake spreads. A documented BMS troubleshooting reference checks the cell-tap wiring and pack-voltage sensing when the reported pack voltage disagrees with the sum of the individual cell voltages.
This distinction is critical: balancing cannot repair a wrong measurement. If the reported highest or lowest cell is incorrect, balancing on that reading can make the pack worse. Verify the measurement before touching the balance.
4. What Imbalance Actually Costs You
A small spread costs capacity; a large one costs the battery. The series string stops at the weakest LiFePO4 cell in both directions.
- Charging stops early. One LiFePO4 cell hits 3.65V and the BMS cuts charging, so the pack settles around 70-90% while the other cells still have room.
- Discharge stops early. One cell reaches the low-voltage cutoff and the BMS opens the output, stranding usable energy in the rest of the pack.
- Protection events stress the weak cell. Repeated trips accelerate the drift that caused them.
- SOC reporting becomes unreliable. The pack-level estimate cannot reflect one cell that behaves differently.
The economics are simple. On a 15kWh bank, a 10% usable-capacity loss is 1.5kWh of storage you paid for but cannot use every day. Fixing the cause restores that capacity in every LiFePO4 cell without buying new cells. On a moveable pack, the same math applies every time the LiFePO4 cells are rolled to a new spot and cycled there.
5. Passive vs Active Balancing: Can the BMS Catch Up?
The BMS corrects a charge-state difference in LiFePO4 cells with one of two topologies. Whether it can catch up is a math problem. The time a balancer needs to close a SOC gap follows:
C(nominal) is the cell rated capacity, ΔSOC is the charge-state gap, and I(balance) is the balancing current.
| Balancer type | Typical balancing current | Energy efficiency | Time to close a 2% SOC gap on a 280Ah cell | Best fit |
|---|---|---|---|---|
| Passive resistor | 50 - 200mA | 0%, surplus becomes heat | 28 - 112 hours | Small packs under 100Ah |
| Active capacitor or inductor | 0.5 - 2.0A | 70 - 95%, energy shuttled | 2.8 - 11.2 hours | Large packs over 200Ah |
A 280Ah LiFePO4 cell with a 2% SOC mismatch carries 5.6Ah of error. A passive balancer at 50mA needs about 112 hours of continuous work to close it. Home solar batteries only give LiFePO4 cells 0.5 to 2 hours of constant-voltage absorption per day. A passive BMS on a large pack rarely catches up with daily drift. Documented field procedures confirm this reality. One 48V system uses a 50mA balancing current, switched on and off in alternating one-second intervals. It reports that balancing can take one to two days depending on pack condition. Patience is part of the procedure, not an optional extra.
Active balancing moves charge between LiFePO4 cells with four to ten times the current, closing the same gap in a single good solar day. Large home packs with big LiFePO4 cells, like our 30kWh moveable LiFePO4 battery and the 40kWh 48V pack, are exactly where this math matters most. The JM Blade battery with active balancer shows how active balancing is packaged in a production battery.
6. How to Measure Cell Voltages Safely
You need real per-cell data before deciding anything about LiFePO4 cells. Two safe routes exist for reading LiFePO4 cell voltage.
- BMS app or display. Most smart batteries show every LiFePO4 cell voltage in the app, sorted from lowest to highest. This is the easiest route. The app shows each cell as a numbered bar, so you can watch the spread change in real time. The solar battery monitoring app guide covers per-cell reads and what each field means.
- Multimeter at the BMS sense wires. If the battery has no app, measure between adjacent sense wires: B0-B1 is cell 1, B1-B2 is cell 2, and so on. Set the meter to DC voltage and hold the probes steady.
Safety comes first in both cases:
- Wear insulated gloves and eye protection.
- Work on a non-conductive surface, away from anything that could short a LiFePO4 cell.
- Never short LiFePO4 cell terminals with the probes.
- Never measure directly at LiFePO4 cell terminals on an assembled pack unless the manual allows it.
- Record every reading with the date and the battery state. A multimeter check is only as good as the contacts, so clean the sense wire tips first.
Do not open a sealed battery to reach the cells. High-energy packs can deliver dangerous fault currents, and sense wiring may not be safe without the manufacturer's isolation procedure. Use the normal monitoring interface for sealed units and refer internal electrical work to qualified personnel.
7. The 8-Step Diagnostic Sequence
This sequence separates a temporary voltage difference from a persistent SOC imbalance, a weak cell, and a measurement fault. Work through the steps in order.
Step 1: Check safety-related faults first. Review BMS alarms, LiFePO4 cell-voltage limits, and temperature readings. Do not start a balancing charge with a serious overvoltage or undervoltage alarm, a cell outside its permitted range, physical damage, abnormal heating, or other failure signs. Do not bypass the BMS, force-charge a single cell, or use a lead-acid equalization mode on a lithium pack.
Step 2: Record the complete operating context. Collect highest and lowest LiFePO4 cell voltage, individual cell voltages, battery current, cell temperatures, charge state, BMS alarms, balancing status, and recent usage history. Note the time of each measurement. An excursion during heavy load means something different from the same reading persisting after rest.
Step 3: Measure the spread after the battery rests. Isolate the charge source and loads as the manual directs. Let the pack rest for the prescribed period, then record every LiFePO4 cell voltage. Calculate ΔV and ask three questions: Does the spread shrink after current stops? Does the same cell stay highest or lowest? Does the spread repeat under comparable conditions? A spread that collapses at rest is current-related. One that persists points to SOC imbalance, capacity mismatch, self-discharge, or a sensing fault. Resting voltage alone cannot separate them, so do not treat a single number as pass or fail.
Step 4: Compare cell behavior under controlled current. During discharge, look for a LiFePO4 cell that falls faster than its neighbors at comparable current and temperature. During charging, look for a cell that rises disproportionately and approaches the upper limit early. After current stops, observe whether it returns toward the others. Large excursions in both directions suggest elevated resistance. A LiFePO4 cell that always reaches a limit early, even after balancing, suggests reduced capacity. Temperature, connection faults, and measurement errors can mimic both, so combine voltage, current, temperature, and history.
Step 5: Verify the measurement system. If readings are physically inconsistent or jump without a change in conditions, suspect the BMS. Review cell-sense and communication faults. Compare reported pack voltage with the sum of cell voltages where the system allows it. Check whether the same LiFePO4 cell keeps showing an implausible value. Confirm a measurement fault and repair it before balancing, or the BMS responds to incorrect information.
Step 6: Confirm that balancing can actually operate. Balancing is not continuous. The BMS may require a balancing-start voltage, a configured LiFePO4 cell spread threshold, a permitted charging state, an acceptable temperature, and no blocking fault. A battery that looks unbalanced may simply have a balancing algorithm that is not active under current conditions. Check the settings and status against the battery documentation before touching anything.
Step 7: Run a controlled balancing charge only if the cells are healthy. If the evidence points to an SOC imbalance with no weak cell, unsafe voltage, temperature problem, or sensing fault, use the approved procedure. Passive balancing dissipates energy from high cells as heat and is slow on large packs. Active balancing transfers energy between cells and is faster, but its effect depends on circuit design and current. Neither restores the capacity of a degraded LiFePO4 cell. Never raise the charging voltage to force balancing; the weakest cell may cross its safe limit before the others reach the balancing window.
Step 8: Verify the result after balancing. Repeat the cell measurements under comparable conditions. Check whether the resting spread narrowed, whether the same cell still diverges during charging, whether it still reaches the discharge limit early, and whether SOC reporting improved. A stable narrow spread supports a corrected SOC imbalance. A LiFePO4 cell that keeps diverging points to capacity loss, resistance growth, self-discharge, or a sensing fault, and more balancing will not fix it.
8. Case Studies: What the Evidence Shows
The three cases below separate documented field observations from simplified engineering calculations. A useful diagnostic example is not a universal specification.
48V system develops imbalance after over-discharge. A November 2023 service bulletin for a 48V LiFePO4 system links LiFePO4 cell imbalance to over-discharge. It instructs technicians to compare the highest and lowest cell voltages with the battery idle. It treats a differential above 300mV as significant for that system. It also describes balancing during charging when its specific voltage and spread conditions are met. The balancing current is low and switches on and off, and completion can take one to two days. The lesson: overall pack voltage alone cannot characterize cell condition, and the 300mV criterion belongs to that application, not to every battery.
BMS reports 3.35V on one cell and 3.00V on another. A BMS troubleshooting reference uses resting readings of about 3.35V and 3.00V as an example of substantially different LiFePO4 cell voltages. The difference is 350mV:
That is a reason for serious investigation, not a diagnosis. The LiFePO4 cells may hold unequal SOC, one may have reduced capacity, or the sensing may be wrong. Balancing can align charge states but cannot rebuild missing active material. The lesson: a large resting spread narrows the investigation but does not name the fault.
Spread grows only under a high load. Two LiFePO4 cells with a 1 milliohm difference in effective resistance at 50A produce a resistive difference of about 50mV before other effects. If the spread grows under load and narrows at rest, resistance, polarization, or connection issues contribute. Compare cell behavior at similar temperatures and currents, and check connections. A cell with a different resting voltage as well as a load-dependent gap may carry both an SOC and a resistance problem, because the causes are not mutually exclusive.
9. When Is Balancing Enough, and When Do You Replace a Cell?
The decision for LiFePO4 cells rests on repeatable evidence, not one reading.
| Observed behavior | Likely interpretation | Appropriate next step |
|---|---|---|
| Spread narrows after proper balancing and stays small | SOC imbalance was the main issue | Continue monitoring |
| Spread is large under load, small at rest | Resistance, polarization, or connection effects | Compare controlled current response, check connections |
| Same cell repeatedly reaches a limit early | Reduced capacity, resistance growth, or persistent imbalance | Run cell-health testing |
| One LiFePO4 cell stays different after proper balancing | Persistent SOC difference, self-discharge, capacity loss, or sensing error | Investigate the cause instead of repeating balancing |
| Readings contradict pack voltage or jump without cause | Measurement or configuration problem | Validate the BMS sensing system |
| A LiFePO4 cell violates limits, runs hot, or shows damage | Potentially unsafe condition | Stop, follow the approved safety procedure |
LiFePO4 cell replacement should never rest on a single voltage reading. It is justified when controlled testing confirms materially lower capacity, abnormal resistance, or persistent self-discharge, or when the manufacturer prescribes it. Replacement cells must match the pack chemistry, specification, configuration, and BMS. Introducing a mismatched cell can reproduce the original imbalance or create a new one. For larger moveable packs, that matching matters even more. The 20kWh moveable LiFePO4 battery and the 15kWh moveable design run large series strings, and one mismatched LiFePO4 cell sets the whole pack limit.
10. Common Troubleshooting Mistakes to Avoid
- Treating every LiFePO4 cell voltage difference as imbalance. A spread is a symptom, not a root cause. Establish conditions, compare resting and dynamic readings, then pick the fix.
- Expecting zero LiFePO4 cell spread at full charge. LiFePO4 cells are not identical, and their voltages respond differently near the top. A higher reading at the end of charging is not proof of damage.
- Raising the charge voltage to force balancing. This exposes the highest cell to excessive voltage before the others reach the target. A protection event is a signal, not an obstacle.
- Assuming balancing restores capacity. Balancing redistributes charge; it does not repair degraded electrodes. Test capacity and resistance instead of repeating the same procedure.
- Ignoring temperature and sensing errors. A cold cell behaves differently under load, and faulty sensing makes a normal cell look abnormal. Correlate voltage with temperature, current, and fault history before replacing anything.
11. Quick Fix Table
| Symptom | What to measure | Most likely cause | First fix |
|---|---|---|---|
| Charging stops at 70-90% | Cell voltages at absorption | One high LiFePO4 cell trips over-voltage protection | Top-balance at absorption for 2-4h, repeat 2-3x |
| BMS warning "cell difference" | Resting spread after 1-2h | Persistent imbalance above 100mV | Full charge, hold, re-measure |
| Spread appears only under load | Cell voltages during discharge | Resistance mismatch, loose busbar | Torque connections, re-test at rest |
| Spread grows back every week | Resting spread over several cycles | Drift outpacing passive balance | Lower absorption target, consider active balance |
| One cell below 2.0V permanently | Single cell voltage | Failed cell | Stop use, warranty service |
| Pack voltage disagrees with cell sum | BMS data vs measured values | Sensing fault | Validate BMS before balancing |
12. Prevention: Keep the Spread Small From the Start
Prevention is cheaper than recovery. Six habits keep the LiFePO4 cells in a pack balanced over years:
- Charge to full regularly. Passive balancing only works in the absorption window, so a LiFePO4 cell pack that never fully charges never balances.
- Let the BMS finish its job. Do not interrupt absorption early, or the LiFePO4 cells never get their balancing window.
- Keep the module temperature even. Avoid putting one side in direct sun or against a cold wall, so every LiFePO4 cell ages at the same rate. Temperature gradients quietly drive spread.
- Torque and clean the busbars at the interval the manual specifies.
- Store at a moderate SOC when idle, around 50%, because LiFePO4 cell self-discharge differences show up fastest at high SOC.
- Watch the resting LiFePO4 cell spread trend, not a single reading. A pack that holds under 50mV for months is healthy even if one reading looks odd.
Moveable batteries add one more habit. Rolling a 15kWh pack shifts mechanical stress, so re-check busbar torque and the resting spread after every relocation. The 15kWh battery with wheels and the 20kWh battery with wheels benefit from this check after every move. The full range of portable home storage is collected in the moveable solar battery collection. The BMS guide for LiFePO4 explains how protection and balancing circuits keep LiFePO4 cells aligned in detail.
FAQ
What is an acceptable cell-voltage difference for LiFePO4 batteries?
There is no single value for every pack and condition. The limit depends on cell specification, BMS balancing strategy, current, temperature, and whether you measure during charging, discharging, or rest. At rest, a healthy LiFePO4 cell pack typically holds 20mV to 50mV. Use the manufacturer's documented limits instead of a generic millivolt threshold.
Why is the voltage difference large during charging but small after resting?
Charging current adds resistive voltage rise and dynamic polarization across LiFePO4 cells. Those effects relax after charging stops, so the spread narrows. The pattern does not prove equal capacity. Repeat the measurement and, if needed, run controlled testing to separate normal relaxation from a real cell-health problem.
How do I fix out-of-balance LiFePO4 cells?
Diagnose first, then balance. Record the operating context, measure the resting spread, and check whether the same cell stays highest or lowest under load. If the cause is an SOC difference and the cells are healthy, run the approved top-balance. Charge to full, hold at the absorption voltage for 2 to 4 hours, and let the BMS bleed the high cells down. Then rest and re-measure. Repeat until the spread stays stable.
Why does my LiFePO4 battery stop charging at 80-90%?
One cell reaches its 3.65V ceiling before the others, so the BMS cuts charging to protect it. That is the weakest-cell effect, not a charger fault. Top-balance the pack and keep a regular full-charge schedule so the same cell stops tripping over-voltage protection.
Can LiFePO4 cells balance automatically?
Yes, if the battery has a working balancing circuit for its LiFePO4 cells and the required conditions are met. Many BMS designs use passive balancing; others use active balancing. The function may only run during specific charging stages or when set voltage, spread, and temperature conditions hold. Automatic balancing is not guaranteed to correct every severe mismatch.
Can a battery be out of balance if all cell voltages look similar?
Yes. The flat voltage plateau of LiFePO4 limits how accurately voltage reveals SOC in a LiFePO4 cell through most of the range. Cells can show similar voltage while holding different usable capacity or internal resistance. Compare behavior under current and over time, not just the resting numbers.
How long does cell balancing take?
It depends on the charge mismatch, balancing current, time in the balancing region, and BMS algorithm for the LiFePO4 cells. A small mismatch can correct during a normal charging period. A larger one can take many hours or several sessions. Documented field procedures for one 48V system report one to two days at a 50mA alternating balancing current.
Is 0.1V difference between LiFePO4 cells bad?
It depends on when you measure the LiFePO4 cells. At rest, 0.1V (100mV) is in the action zone and needs top-balancing. During heavy current, a temporary 0.1V to 0.15V spread can be normal. Measure after 1 to 2 hours of rest to judge the real balance, and compare with the pack's own history.
Does cell imbalance mean the battery needs replacement?
Not necessarily. If the cause is an SOC difference and all cells are healthy, proper balancing may resolve it. Replacement becomes likely when testing confirms a persistently weak cell, substantial capacity loss, abnormal resistance, self-discharge, or physical damage. A BMS alarm must be read against its actual operating limits, not treated as automatic proof of failure.
Is it safe to charge a battery with a large cell-voltage spread?
Not automatically, because pack voltage does not cover every LiFePO4 cell. A large spread can indicate an unsafe condition, and pack voltage does not guarantee every cell is within its limits. Stop routine troubleshooting if a cell is outside its permitted range, the BMS reports a serious protection fault, or the battery shows abnormal heat or damage. Follow the battery-specific instructions or get qualified service before charging again.
