What Is Overcurrent?

Overcurrent means the battery is being asked to deliver or accept more current than it was built for. Simple as that. The cell heats up, chemistry gets stressed, and things go sideways if nobody steps in.

I see this problem constantly in field returns. Packs come back with burned traces, melted tabs, cells that look fine on the outside but measure dead internally. Nine times out of ten, someone pulled too much current or the protection failed.

Lithium cells have hard limits. A cell rated for 10A continuous will tolerate 10A all day at room temperature. Push it to 15A and the internal temperature climbs. Push it to 30A and you have maybe seconds before something gives.

Heat generation in a lithium cell follows I²R. Internal resistance on a typical 18650 sits around 15-30mΩ depending on the cell and its age. Run the numbers on a 25mΩ cell.

At 10A: 10² × 0.025 = 2.5W At 20A: 20² × 0.025 = 10W At 40A: 40² × 0.025 = 40W

That 40W has nowhere to go inside a steel can. The cell temperature spikes. Electrolyte starts breaking down around 80°C. Above 130°C, the separator can fail. After that, thermal runaway becomes a real possibility.

Charging overcurrent follows the same physics but with an extra problem. Lithium plating. Push charge current too high and lithium metal deposits on the anode surface instead of intercalating properly. That plating is permanent capacity loss. It also creates dendrite formation risk.

Overcurrent

External shorts happen more than manufacturers like to admit. A loose screw in an enclosure. A damaged wire rubbing against a frame. Water ingress creating a conductive path. I worked on a recall where the pack connector had a design flaw that let the positive pin contact the enclosure ground during insertion. Thousands of packs went out before anyone caught it.

Load faults show up in power tool applications constantly. Brushed motors stall at 6-8x running current. Brushless with FOC control handles stalls better but still spikes hard. The battery sees that spike whether the motor controller does anything about it or not.

Charger problems tend to be subtle. A charger designed for a 4S pack plugged into a 3S pack. The voltage limits might still work, but the current profile is wrong. Or the charger sense line breaks and the charger just pushes full current until something trips.

Cell mismatch in series packs is insidious. Four cells in series, three of them 3000mAh, one of them actually 2700mAh because it sat in a hot warehouse for six months. That weak cell hits full charge first. The other three keep pushing. That weak cell gets overcharged, generates gas, heats up, and now you have a local overcurrent condition happening inside a single cell while the pack voltage looks fine.

The immediate consequence is heat. Sustained overcurrent at 2x rating will push most cells past 60°C within a few minutes. The cell survives, probably, but cycle life takes a hit.

Repeated overcurrent events create cumulative damage. The SEI layer on the anode thickens. Internal resistance creeps up. Capacity fades. A cell that started at 3000mAh and 20mΩ might be 2400mAh and 35mΩ after 200 abusive cycles. It would have made 800 cycles under proper use.

Severe overcurrent-short circuit conditions-can end a cell in seconds. The current peaks at whatever the cell's internal resistance limits it to. A fully charged high-drain 18650 into a hard short can see 200-300A instantaneous current. The tab welds are often the first failure point. If the tabs hold, the jelly roll heats so fast that venting follows within seconds.

I have seen cells that vented so hard the positive cap launched across the lab. Safety glasses are not optional when testing short circuit response.

Overcurrent

Good pack design uses multiple protection mechanisms. No single device handles every failure mode.

PTCs sit inside most cylindrical cells from reputable manufacturers. They trip based on temperature, not current directly. A PTC might be rated to hold 7A and trip at 15A, but the trip mechanism is thermal. Response time is slow-hundreds of milliseconds to seconds. PTCs will not save you from a hard short. They handle moderate overcurrent and give the pack time to cool down.

Fuses blow once and stay blown. Pack-level fuses are sized above normal operating current with margin for inrush and transients. A 10A continuous pack might use a 15A fast-blow fuse. The fuse clears a hard short faster than a PTC, typically under 100ms at high fault currents. But it also kills the pack permanently. Warranty claims follow.

Protection ICs monitor current through a sense resistor. Common parts from Seiko, TI, and others provide overcurrent thresholds programmable through external resistors or hardcoded internally. Detection delays run 8-24ms typically. Short circuit detection is faster, often under 500µs. The IC drives external FETs to disconnect the pack.

The sense resistor value matters. A 5mΩ sense resistor gives better resolution but drops more voltage and dissipates more power at high currents. A 2mΩ resistor wastes less power but needs a more sensitive front end. Most consumer packs use 3-10mΩ depending on current class.

BMS in larger packs adds intelligence. Active current limiting rather than just trip/no-trip. Temperature-compensated thresholds. Logging of events for diagnostics. A good BMS reduces overcurrent limits as cell temperature rises, keeping the cells in their safe operating window even under dynamic loads.

CIDs inside cells provide last-resort mechanical protection. The Current Interrupt Device activates on internal pressure buildup. By the time a CID trips, the cell has already experienced significant stress. CID activation usually means the cell is scrap.

Datasheet ratings assume specific conditions. A Samsung 30Q rated for 15A continuous assumes 25°C ambient and adequate cooling. Stick that cell inside an insulated enclosure at 35°C ambient and 15A will push it past safe temperatures.

Pulse ratings look attractive but come with strings attached. A cell might claim 30A for 10 seconds, but that assumes the cell started at 25°C and has time to cool before the next pulse. Back-to-back pulses without recovery time accumulate heat just like continuous current.

Charging rates are often more conservative than discharge rates. A cell that handles 20A discharge might only tolerate 4A charge. Lithium plating risk is the reason. Some newer cells with silicon-doped anodes are even more sensitive to charge rate.

Cell age changes everything. A new cell with 20mΩ internal resistance handles 20A better than a year-old cell at 30mΩ. Pack protection should account for end-of-life resistance, not just new cell specs.

Every pack design needs overcurrent validation. The test equipment matters.

Electronic loads need to sink current faster than the protection responds. A protection IC with 10ms detection delay needs a load that reaches target current in under 1ms. Slow load ramp-up lets the protection trip early and gives false confidence.

Current measurement needs bandwidth. A 10mΩ sense resistor at 100A gives 1V signal. Capturing the actual peak requires at least 10kHz bandwidth, preferably more. Scope probes with proper grounding avoid the noise problems that give garbage data.

Temperature testing catches design problems that room temperature testing misses. Protection IC thresholds drift with temperature. FET Rds(on) increases at high temperature, adding voltage drop. Sense resistor TCR matters when the resistor heats up from passing fault current. Test at -20°C, +25°C, and +55°C minimum.

The test matrix gets large. Charge overcurrent, discharge overcurrent, short circuit. Each at three temperatures. Each at multiple SOC levels because cell impedance varies with charge state. Multiply by sample size for statistical confidence. A thorough validation runs hundreds of tests.

Overcurrent

UL 2054 covers portable battery packs for North American markets. Short circuit testing applies a dead short for 10 seconds. The pack must not catch fire or explode. Temperature is monitored but no specific limit is called out. This is a minimum bar.

IEC 62133-2 applies internationally. External short circuit testing uses less than 100mΩ total circuit resistance held for one hour or until temperature stabilizes. More stringent than UL 2054 on duration.

UN 38.3 governs shipping. Test 5 requires a short through less than 0.1Ω. Cells or batteries must not disassemble or catch fire. This one matters because failing UN 38.3 means your product cannot be transported legally.

SAE J2464 covers EV applications with tighter requirements. The short circuit resistance drops to 5mΩ and thermal/mechanical criteria are more specific.

Passing these tests does not mean the protection design is good. It means the protection design is adequate for certification. Real-world abuse can exceed test conditions.

Wire gauge comes up constantly. Undersized wire adds resistance and creates heat at connections. The wire itself might handle the current, but crimped terminals or solder joints become hot spots. Packs returned for "battery failure" often show burned wire terminals while the cells are fine.

Connector selection is another weak point. XT60 connectors are rated for 60A continuous, supposedly. That rating assumes perfect crimps and clean contacts. In practice, derated by 30-40% for reliability. Anderson Powerpoles have similar considerations.

Tab welding on cells needs attention. A good weld has low resistance and mechanical strength. A cold weld looks fine but fails under vibration or thermal cycling. A burned weld damaged the can and creates a weak spot. Weld schedules need validation with pull testing and resistance measurement.

Thermal management ties into overcurrent capability. A pack with active cooling can sustain higher currents than a sealed pack with no airflow. Some designs use thermistors on cells to trigger current reduction before temperatures go critical.

Overcurrent kills lithium batteries. Sometimes fast, sometimes slow, but always damaging. Current limits exist for reasons tied to physics and chemistry, not arbitrary conservatism.

Good protection requires multiple layers because each protection type has weaknesses. PTCs are slow. Fuses are one-shot. ICs depend on sense accuracy. BMS adds cost and complexity. CIDs mean the cell is already damaged.

Testing needs to reflect actual use conditions and abuse scenarios. Passing certification tests is necessary but not sufficient.

Pack design choices-wire, connectors, thermals, margins-determine whether the protection actually protects or just looks good on paper.

The cells I trust come from manufacturers with consistent quality and published data. The packs I trust use protection circuits from reputable vendors with proper application design. Everything else gets treated with caution until proven otherwise.