A two-cell LiPo was accidentally discharged to just 0.57 V per cell. It initially appeared to recover: the cells charged uniformly, showed no detectable heating or swelling, and retained their original measured AC internal resistance.
Over the next two days, serial measurements documented an initially tiny but progressively widening voltage difference between the cells. Then, after 72 hrs, unexpectedly one cell suffered a catastrophic discharge and dropped to 1.653 V.
The incident is a practical warning:
Even brief extreme overdischarge can be followed by serious internal damage much later. An apparently successful “revival” is not evidence that the battery is safe to operate.
Below, I share my findings and personal conclusions. This does not include any warranties, guarantees, or liabilities. All information is as-is and “best effort”. Information can be wrong, incomplete, inappropriate, and should always be validated independently by yourself.
Overview
During battery testing, a capacity test with a brand new RC LiPo pack was conducted. The battery was fully charged and connected to an electronic load, and a power meter was used to measure the delivered capacity and energy.
Accidental Deep Discharge
The electronic load was accidentally not put into battery mode where it would have stopped discharge at a specified safe voltage. Instead, the load discharged the battery without limit.
The very same can happen in a variety of practical scenarios:
-
Bad Devices:
Unprotected or not sufficiently protected LiPo batteries are used in RC cars, drones, or home improvement equipment that do not reliably cut off at an appropriate battery voltage -
Bad Packs:
2(+)S (multi-cell) batteries are discharged hard to 3.0 V/per cell. The cells may not be perfectly matched, in which case one cell can drop far below the cutoff voltage.Practical example: Assume a 14.8V-4S-LiPo: when discharging this pack hard to 12.0 V, this theoretically stays within the 3.0 V per cell, however practically you cut off the pack at an average of 3.0 V per cell. If cells are not perfectly equal, some may still be at 3.3 V at cutoff, whereas another cell may have already dropped to 2.5 V by that time.
- Always cut off with a safety margin. If you notice the voltage drops much faster than normal, or the pack shuts off unexpectedly/prematurely, always assume a bad cell.
- With RC LiPos, routinely check the individual cell voltages of your pack at the end of discharge, not at the end of charging (when the balancer has equalized the cell voltages and obscured evidence of mismatched cells).
Battery Resuscitation
At the end of the test, the mishap was immediately recognized, and at a cell voltage of 0.57 V, the battery was revived with inital very small currents, then charged to around 40% SOC.
Assumed Success
Recovery seemed successful:
- the pack accepted charge
- no signs of swelling or heating
- both cells in the 2S pack behaved identical
- after charge ended, both cells held their voltage
The reasoning was that the deep discharge
- lasted only minutes
- occured relatively quickly
- cells had a positive rest voltage (no voltage reversal)
- both cells had equal voltage
- mishap of short duration, was treated immediately
- studies that establish cell damage through deep discharge used much more severe deep discharge: longer-lasting, and with voltage reversal
Catastrophic Failure After All
Careful long-term observation after the mishap showed no irregularities over the first 24 hours. Only after 24 hrs did signs of permanent damage begin to manifest, and only in the 48-72 hrs timeframe, one of the two cells suddenly and unexpectedly discharged rapidly.
This example vividly illustrates why any lithium based battery should be retired immediately once it has been subject to deep discharge, even if this loss hurts. An accidental deep discharge that lasted only a few minutes permanently damaged internal battery integrity. The cells seemed to properly recover, and the failure occured with a time delay after three days.
The incident: a 1C discharge ran too far
The battery was a brand new 2S 2700 mAh LiPo with only one charge cycle recorded. The pack was accidentally discharged at approximately 1C (2.7 A) until its terminal voltage collapsed. The load was disconnected quickly after the mishap was noticed. When the pack was connected to a charger, both cells measured 0.57 V — closely matched and still positive.
The available power-monitor screenshot shows a discharge lasting approximately 3,674.6 seconds (61.2 minutes). Its footer reports 2,491.11 mAh and 16,654.34 mWh delivered. The curve has an extended relatively gradual decline followed by a very steep terminal collapse.
The 0.57 V readings were taken after disconnecting the load, so they do not establish the minimum instantaneous cell voltages under load. The graph also cannot exclude transient reversal of one cell during the final collapse. Nevertheless, the measured matching positive voltages and the short interval before discovery initially appeared encouraging.
Apparently Successful Recovery
A cautious recovery charge was initially attempted at 0.1 A, raising the pack voltage to approximately 6.2 V. After resting, the voltage appeared stable. Charging was subsequently continued at 2.2 A, without observable swelling or heating.
Both cells recovered uniformly. Their charger-measured AC internal resistance was 7.8 mΩ per cell, unchanged from the pre-incident assessment. This was encouraging because large changes in bulk cell impedance might have indicated damage. However, a single-frequency AC measurement cannot reliably detect a microscopic leakage pathway or prove that the separator is intact.
The pack was then brought to an approximately 60% indicated state of charge for passive observation. At 45 minutes after charge termination, both cells read 3.757 V and had shown no change at the displayed millivolt resolution over the preceding ten minutes.
At this stage, the findings were superficially reassuring:
- No visible swelling or measurable abnormal heating during recharge.
- Matching cell voltages after extreme overdischarge and during recovery.
- Unchanged 7.8 mΩ charger AC resistance for both cells.
- Stable initial resting voltages.
The relevant question now was whether the internal structure could still retain charge safely, and whether the cells had truly recovered and were safe to use again.
Subtle Early Earning, Followed by Rapid Self-Discharge
Early spot readings suggested that the eventually damaged cell declined slightly faster than the other, however changes over time initially occured in the low mV region. This includes benign effects such as post-charge relaxation, temperature variation and meter uncertainty. So within the first 24 hours, there was no clear indication visible that either cell suffered permanent damage. All looked normal.
The initial asymmetry became meaningful only in retrospect, when the same cell later suffered severe voltage loss. It was no useful assessment indicator.
At Day 1, 05:00, the cells differed by just 22 mV; by Day 2, 20:30, the difference had widened to 166 mV. At Day 3, 21:30, the healthy cell was still at 3.712 V while the damaged cell measured 1.653 V — a difference of 2.059 V. This confirms a severe damaage of one of the two cells in the pack.
Recorded Cell Voltages
The original measurements supplied after the incident are reproduced below. Days are numbered consecutively; the values are in volts.
| Observation | Cell 1 (stable) | Cell 2 (failed) | Difference |
|---|---|---|---|
| Day 1, 05:00 | 3.745 | 3.723 | 22 mV |
| Day 1, 06:30 | 3.734 | 3.709 | 25 mV |
| Day 1, 08:50 | 3.732 | 3.705 | 27 mV |
| Day 1, 11:15 | 3.732 | 3.697 | 35 mV |
| Day 1, 14:40 | 3.728 | 3.691 | 37 mV |
| Day 1, 16:40 | 3.726 | 3.689 | 37 mV |
| Day 2, 09:00 | 3.728 | 3.683 | 45 mV |
| Day 2, 10:30 | 3.718 | 3.666 | 52 mV |
| Day 2, 20:30 | 3.714 | 3.548 | 166 mV |
| Day 3, 21:30 | 3.712 | 1.653 | 2,059 mV |
- Early Observation (24-36 hrs):
The 3.726 → 3.728 V rise in the stable cell between Day 1 afternoon and Day 2 morning is within the range of meter resolution, thermal effects, and voltage redistribution; it is not evidence of spontaneous charging. - Delayed Collapse (48-72hrs):
The exact timing of the final collapse remains unknown: The cell was disconnected from the charger and external circuitry during the last interval; the voltage could have declined progressively or rapidly at any point within those 25 hours.
Even though the available data cannot clearly point to a specific failure mechanism, the outcome is unambiguous: one previously deeply discharged cell developed severe, asymmetric self-discharge while disconnected from external circuitry.
Where Did the Energy Go?
A fall from about 3.7 V to 1.65 V in an unconnected cell represents severe internal charge depletion (defect):
-
If the loss resulted from an internal electronic leakage path or a soft short, much of the cell’s available electrochemical energy would be released as heat through resistive and electrochemical losses.
-
Parasitic chemical reactions offer another possibility: some energy could be retained in new reaction products, with some also appearing as heat. There was no external load during the final 25-hour observation gap to account for external discharge.
A relatively resistive internal path could dissipate energy over hours with little visible change in surface temperature. Conversely, heating localized at a microscopic defect could be much greater than indicated by the pouch temperature. A cool exterior would not establish that the cell was safe.
Suspected Mechanism: Copper Dissolution and Internal Leakage
In conventional graphite-anode lithium-ion cells, the negative-electrode current collector is copper foil. Severe overdischarge can raise the negative-electrode potential to a range in which copper corrodes and enters the electrolyte.
Copper may subsequently redeposit elsewhere during discharge or recharge. If deposits or other damage create an electronic path through the separator, an internal short circuit can result.
Scientific copper dissolution experiments generally involve prolonged zero-voltage exposure or cell reversal, which in the presented case was not present. Other internal defects or parasitic reactions are therefore more likely to explain the late self-discharge.
What matters practical is not the suspected mechanism but solely the observed severe self-discharge in one cell after a prolonged unsuspicious delay.
Hazards
24 hours of innocent observation and supposingly properly behaving cells made it tempting to perform a full-charge test or put the LiPo back in service. In hindsight, that would have been a dangerous decision as the cell suffered rapid discharge much later.
- Charging a compromised cell to its maximum voltage increases its stored energy and exposes it to additional electrochemical stress. If an internal leakage path were to become more conductive, more energy would be available for internal dissipation.
- During charging, the charger supplies additional energy into an evolving defect.
What This Case Establishes — And What It Does Not
| Directly observed | Not established by the evidence |
|---|---|
| Both cells at 0.57 V directly after accidental deep discharge | Severe deep discharge without voltage reversal |
| Initial uniform charging and recovery | Absence of microscopic damage |
| Unchanged charger AC impedance: 7.8 mΩ each | Intact separator or copper collector |
| Widening voltage difference over ten measurements | Early-warning threshold based on millivolt differences |
| One disconnected cell fell to 1.653 V after 72 hrs | Severe and potentially catastrophic defect was present despite initially encouraging observations |
Conclusion: Retire Deeply Overdischarged Batteries
Once a lithium cell has been deeply discharged — however short this incidence may have been — there is a real risk of catastrophic failure at any time in the future, even if the cells seem to behave normally within 24-48 hrs after careful recharge.
Apparently, damages may be obscure at first, and rapid and potentially catastrophic internal discharge can occur any time; days, weeks, or months after the discharge incident.
Apparently reassuring short-term electrical measurements, and cells accepting charge witout signs of swelling or heat, both do not exclude serious delayed failure after extreme overdischarge.
Practical Consequences
The practical response to a LiPo (or any other lithium-based cell) found at extreme undervoltage should be:
- Avoid further charging and further use, even if cell recovery may look promising, and retiring a previously good LiPo may be a considerable economic loss.
- Isolate the cell in a fire resistant bag, preferrably outside your home, until it can be properly disposed.
- Arrange proper disposal through a facility equipped to accept damaged lithium batteries, and disclose the deep-discharge history.
The most dangerous assumption would have been:
“It charged normally, the resistance is unchanged, and the early voltage changes look small — therefore it is safe.”
This case illustrated why that inference is not valid.
Risk Assessment
Only one of the two cells showed delayed failure. So apparently, not every deep discharge event leads to permanent damage, and you may be “lucky”.
However, it was clearly observable that damages may be very subtle at first. While it is increasingly unlikely that the second cell was damaged, it cannot be excluded. It frankly seems almost impossible to safely predict whether a cell truly has not suffered a microscopic damage.
The observation that such a damage leads to unexpected catastrophic discharge much later must be seen as a grim warning:
If you keep previously deep discharged lithium cells in service, you must be prepared to suffer a sudden internal discharge at any time in the future, whether under load or resting, with the associated severe risks. These risks outweigh by far the economic gain of not having to retire a costly LiPo.
If you still want to test your luck, and resuscitate deeply discharged LiPos, at the minimum, you should carefully observe cell voltages for at least a week until you make any assumptions on cell health, and isolate the cell in a fire-proof environment during that time.
Technical references
- Ouyang et al., Mechanism of the entire overdischarge process and overdischarge-induced internal short circuit in lithium-ion batteries, Scientific Reports (2016). Open-access article.
- Flügel et al., Cu Dissolution during Over-Discharge of Li-Ion Cells to 0 V: A Post-Mortem Study, Journal of The Electrochemical Society 168, 020506 (2021). DOI.
- Dayani et al., Quantification of the Deep Discharge Induced Asymmetric Copper Deposition in Lithium-Ion Cells by Operando Synchrotron X-Ray Tomography, Advanced Materials Technologies (2023/2024). DOI.
- Hendricks et al., Copper Dissolution in Overdischarged Lithium-ion Cells: X-ray Photoelectron Spectroscopy and X-ray Absorption Fine Structure Analysis, Journal of The Electrochemical Society (2020). DOI.
Reporting note: This is a single-pack observational case study, not a controlled experiment. The ten resting-voltage readings are reproduced from the user-provided log; the discharge values come from the original monitor screenshot. Lines between points are graphical interpolation only, not recovered measurements. The figures distinguish observations from illustrative mechanisms.
Slow Website?
This website is very fast, and pages should appear instantly. If this site is slow for you, then your routing may be messed up, and this issue does not only affect done.land, but potentially a few other websites and downloads as well. Here are simple steps to speed up your Internet experience and fix issues with slow websites and downloads..
Comments
Please do leave comments below. I am using utteran.ce, an open-source and ad-free light-weight commenting system.
Here is how your comments are stored
Whenever you leave a comment, a new github issue is created on your behalf.
-
All comments become trackable issues in the Github Issues section, and I (and you) can follow up on them.
-
There is no third-party provider, no disrupting ads, and everything remains transparent inside github.
Github Users Yes, Spammers No
To keep spammers out and comments attributable, all you do is log in using your (free) github account and grant utteranc.es the permission to submit issues on your behalf.
If you don’t have a github account yet, go get yourself one - it’s free and simple.
If for any reason you do not feel comfortable with letting the commenting system submit issues for you, then visit Github Issues directly, i.e. by clicking the red button Submit Issue at the bottom of each page, and submit your issue manually. You control everything.
Discussions
For chit-chat and quick questions, feel free to visit and participate in Discussions. They work much like classic forums or bulletin boards. Just keep in mind: your valued input isn’t equally well trackable there.
(content created Nov 08, 2026)