Lithium-Ion Battery Internal Short Circuits: Causes, Clues, and Prevention

A cell completes formation without an obvious problem. The next morning, its open-circuit voltage is lower than the rest of the batch. It may also feel slightly warmer during charge. Calling it an "internal short" is tempting, but that conclusion is still one step ahead of the evidence.

An internal short circuit is an unintended electronic path between regions that should remain electrically separated inside the cell. It can begin as a weak, intermittent leakage path or appear as a low-resistance hard short. Either can matter. What changes is the rate of energy release and how much time the lab has to notice the warning signs.

In practical terms: abnormal self-discharge, local heating, or poor coulombic efficiency can be consistent with an internal short, but none of those signals proves one by itself. Side reactions, temperature drift, incomplete wetting, formation differences, and test-channel problems can look similar.

A short is a current path, not a single failure mode

The phrase "internal short circuit" covers a wide range of behavior. A high-resistance path may leak current slowly enough that the first symptom is an outlier in storage voltage. A low-resistance contact can release energy much faster, producing intense local heating and, in the worst case, thermal runaway.

The outcome depends on more than the names of the two materials that touch. State of charge, contact area, short resistance, local heat dissipation, cell chemistry, format, and nearby separator behavior all change the response. NASA and NREL developed an implantable test device partly because common crush and penetration tests do not reproduce every field-relevant internal short in the same way. Their work also showed that different contact combinations can produce different severity in a particular cell design, but it should not be turned into a universal four-step danger ranking for every battery.

Where an internal path can begin

In laboratory cells, the initiating defect usually falls into one of three groups.

Assembly and material defects

A metal particle, a rough electrode edge, a foil burr, a damaged separator, or poor electrode alignment can create a path across the insulating layer. The risk is not solved by a "high-purity" label alone. Edge quality, dimensions, handling, storage, and cleanliness all need to be controlled. Our earlier guide to coin-cell assembly mistakes covers the practical checks that are easiest to miss at the bench.

Current collectors deserve particular attention after cutting or punching. A clean-looking edge can still contain a raised burr or loose fragment. For more context on foil selection and handling, see the guide to copper foil in lithium-ion battery research.

Conditions created during cycling

Lithium plating can develop when lithium reaches the anode surface faster than it can intercalate. Low-temperature charging, aggressive charge rates, poor electrode balancing, and overcharge can all make plating more likely. Deposited lithium does not always become a separator-piercing dendrite, but it is a warning that the intended electrochemistry has changed.

Severe over-discharge can also be hazardous. Copper from the anode current collector may dissolve and later deposit elsewhere in the cell, creating another possible route to an internal bridge. Charge and discharge limits therefore belong to the safety design of the experiment, not just the cycling recipe.

Mechanical or thermal damage

Crushing, bending, impact, excessive stack pressure, and high temperature can deform electrodes or compromise the separator. Ordinary aging, however, should not be described as if it automatically creates a micro-short. Capacity fade has many causes. An aged cell with unusual self-discharge needs diagnosis, not a shortcut from "old" to "internally shorted." The distinction matters when interpreting lithium-ion aging data.

What the data can tell you

Early internal shorts are difficult to confirm because their electrical and thermal signatures can be small. A useful investigation looks for agreement between several observations rather than relying on one threshold.

• Rest-voltage decay: compare the suspect cell with matched controls at the same state of charge, temperature, and rest time. A repeatable excess decay is more informative than one isolated reading.
• Coulombic efficiency: persistent charge loss can indicate a parasitic path, but side reactions and measurement resolution must be considered.
• Localized temperature: a repeatable hot spot during a controlled protocol is more concerning than a small whole-cell temperature difference.
• Cell-to-cell divergence: in a pack, one cell drifting in voltage or temperature relative to comparable neighbors can support a fault diagnosis.
• Imaging or teardown evidence: X-ray or CT imaging, followed by a controlled teardown where appropriate, can help locate deformation, debris, or contact. These steps require suitable facilities and procedures.

A simple sequence works well in R&D: stop the test, quarantine the cell, verify the channel and temperature history, repeat non-invasive measurements under a defined protocol, and compare with controls. Do not recharge a suspect cell simply to "see whether it recovers."

Prevention starts before formation

The most effective controls are usually ordinary, repeatable ones:

• inspect punched electrodes and cut foil for burrs, loose particles, folds, and edge damage;
• keep the separator clean, flat, correctly sized, and compatible with the electrolyte and operating window;
• define electrode overlap and alignment instead of relying on visual centering alone;
• use consistent spacer, spring, stack, and crimping conditions for coin cells;
• control electrolyte quantity and wetting time across the batch;
• record material and component lot numbers so an outlier can be traced;
• set conservative formation and cycling limits, including temperature and current constraints;
• build enough replicate cells to separate a process issue from a single-cell anomaly.

Ceramic-coated and shutdown separators can add useful thermal or mechanical protection in suitable designs, but they do not remove the need for clean assembly and correct cell geometry. Likewise, a BMS can limit damaging operating conditions and detect some abnormal behavior, but it cannot guarantee that every nascent internal short will be caught before it develops.

Abuse tests answer a different question

Nail penetration, crush, indentation, and impact tests intentionally create severe mechanical conditions. They can be valuable for comparing cell response and validating safety features, but they are not routine diagnostic tests for a questionable lab cell. They may breach the enclosure, alter the failure path, and create fire, toxic gas, and projectile hazards.

Such testing belongs in a purpose-built facility with remote operation, containment, ventilation, emergency procedures, and a protocol matched to the applicable standard or research objective. A benchtop puncture is not a substitute for an internal-short investigation.

Sources and further reading

• NASA Technical Reports Server: On-Demand Cell Internal Short Circuit Device
• NREL/NASA Internal Short-Circuit Instigator in Lithium-Ion Cells
• Analysis of internal short-circuit in a lithium-ion cell
• Detection of soft internal short circuit using open-circuit-voltage behavior

For prototype work, the goal is not to buy a component described as "safe." It is to specify the component, control how it is handled, and keep enough records to explain an outlier. Flux Battery Hub supplies current collectors, battery separators, and coin-cell hardware for small-batch R&D. Match the dimensions, material specification, and cell design before ordering.