The Battery Failures Nobody Saw Coming Often Leave Signals Behind: Understanding Battery State of Health

Two batteries can both show 80% and be in completely different condition. One may be relatively healthy. The other may have lost a significant amount of its original capacity, while one cell is quietly drifting away from the rest. The percentage on the screen cannot tell them apart. That percentage represents battery state of charge (SoC), how much energy the battery has available right now. It does not tell you how much the battery has degraded over time.

That is where the battery state of health (SoH) becomes critical. SoH provides a view of how much of a battery’s original capacity and performance remains, while also helping teams understand how its condition is changing. Combined with battery management system (BMS) data, battery modeling and digital twin technology can reveal degradation patterns that may otherwise remain hidden until they affect performance, reliability or safety.

Why Battery State of Charge Isn't Enough

Almost every lithium-ion battery, from electric vehicles to a large-scale energy storage system, relies on a battery management system to monitor and protect the battery. A conventional BMS measures parameters such as:

• Voltage
• Current
• Temperature
• State of charge
• Operating limits

The Limitations of Conventional Battery Monitoring

These measurements are essential for safe day-to-day operation. However, conventional monitoring is often focused on identifying whether the battery is operating within predefined limits. Battery degradation is different. Ageing is a gradual process driven by factors such as temperature, charging and discharging conditions, chemical changes and repeated cycling. These changes may not be obvious from a single measurement. This creates an important distinction:

“The battery is operating safely right now” does not necessarily mean “the battery is aging normally.”

Understanding that difference is central to modern battery intelligence.

What Is Battery State of Health (SoH)?

A new battery typically starts at close to 100% SoH. Over time, cycling, temperature, heavy loads and calendar ageing gradually reduce its performance. SoH is commonly evaluated through two important indicators: capacity and internal resistance.

Capacity-Based SoH

Capacity-based SoH compares the battery's current usable capacity with its original rated capacity. For example, if a battery originally had a capacity of 60 kWh and can now store 51 kWh: SoH = 51 ÷ 60 × 100 = 85%

This gives a useful indication of how much capacity has been lost.

Resistance-Based SoH

Internal resistance measures how strongly a battery resists the flow of current inside the cell. As resistance increases, a battery can:

• Generate more heat
• Deliver less power
• Charge more slowly
• Experience greater voltage drops under load

The important point is that capacity and resistance do not necessarily decline at the same rate. A battery may retain much of its original capacity while its internal resistance is already increasing significantly. That change can provide an earlier indication that something inside the cell is changing.

Battery Degradation Can Start Before a Failure

Battery failures rarely appear completely without warning.In many cases, changes begin internally, at the cell level, before they become visible as a significant pack-level problem. Several well-documented industry incidents illustrate why this matters.

Aviation, 2013

A lithium-ion battery installed in a parked aircraft caught fire following an internal short circuit that led to thermal runaway, a self-heating chain reaction that can propagate between cells.

Electric Vehicles, 2020–2021

A major automaker recalled all model years of one of its compact electric vehicles after identifying two rare manufacturing defects within the same cell. The risk was aggravated under specific charging conditions.

Grid Storage, 2019

An explosion at a utility-owned battery energy storage system in Arizona injured firefighters. An independent investigation attributed the incident to thermal runaway originating in a single cell, although the cell manufacturer disputed that conclusion and identified an external cause. These incidents highlight a broader lesson that battery health cannot be understood through charge percentage alone. It requires visibility into how individual cells and the battery as a whole are behaving over time.

Why Battery Health Matters Across the Battery Lifecycle

Understanding SoH is not limited to predicting when a battery will reach the end of its useful life. It can influence decisions throughout the battery lifecycle. Changes in cell behaviour can reveal abnormal conditions before they escalate, while declining battery health can impact range, power delivery and efficiency. By tracking degradation trends, predictive battery maintenance can help teams intervene before developing issues result in unexpected downtime. Reliable health data also supports informed decisions around warranty assessments, repurposing, second-life applications and recycling, making battery health a critical part of modern battery lifecycle management.

What Is a Battery Digital Twin?

So, what is digital twin technology in the context of batteries? A battery digital twin is a virtual representation of a battery cell, module or pack that is continuously informed by real-world operating data. Instead of asking only: “What is the battery doing right now?”, a digital twin can help answer: “Why is the battery behaving this way, and where is it heading?”. This is achieved by combining battery data with electrochemical, thermal and degradation models.

A digital twin can account for operating conditions such as:

• Temperature
• Charging and discharging rates
• Depth of discharge
• Usage patterns
• Cycling history
• Cell-level behaviour

This allows battery teams to move from simply monitoring current conditions to understanding degradation trends and potential future behaviour.

SoC, SoH and Digital Twins: Three Different Views of Battery Health

Understanding a battery requires more than knowing how much charge it currently holds. SoC, SoH and Digital Twins provide three different levels of insight into battery behaviour.

State of Charge (SoC) tells you how much energy is available right now. It is a real-time measure used for everyday battery operation, such as determining how much longer an EV can run or how much energy an energy storage system can deliver.

State of Health (SoH) tells you how much of the battery’s original capability remains. It reflects changes in capacity, performance and degradation over time, making it important for assessing long-term battery condition.

A Digital Twin goes beyond measuring the battery’s current state. It helps explain why the battery is changing and where its performance may be heading by combining real-world operating data with battery models and AI.

Together, they create a more complete picture:

SoC: How much charge does the battery have right now?

SoH: How much of its original performance does the battery still have?

Digital Twin: Why is the battery changing, and what is likely to happen next?

How SimLionics Turns Battery Data Into Intelligence

At SimLionics, battery intelligence combines physics-based battery modeling, real-world data and AI to move beyond basic monitoring.

Physics-Informed AI

Physics-informed AI incorporates the underlying science of battery behaviour rather than relying only on pattern recognition. This helps identify changes in parameters such as capacity and resistance and provides greater context around why those changes may be occurring.

Battery Modeling and Simulation

Battery modeling can replicate different operating and ageing conditions, allowing teams to evaluate charging strategies, performance and degradation without waiting for years of real-world cycling. Combined with battery testing, these models can help accelerate engineering decisions and improve understanding of battery behaviour.

Digital Twin Software

SimLionics' battery digital twin software adds an intelligence layer to battery data. It can help teams:

• Track changes in battery health
• Identify cells that are behaving differently
• Estimate remaining useful life (RUL)
• Understand degradation trends
• Support predictive maintenance
• Optimise battery operation

The objective is not to replace the battery management system. It is to add the intelligence needed to understand why battery behaviour is changing and what that change could mean next.

From Battery Monitoring to Predictive Battery Intelligence

Traditional battery monitoring tells teams what is happening now. Advanced battery intelligence goes further.It connects historical data, current operating conditions and physics-based models to build a more complete picture of battery behaviour. That shift, from monitoring to understanding and from reacting to predicting, can help battery teams make better decisions across electric vehicles, stationary storage and aerospace battery applications.The result is a more informed approach to safety, performance, maintenance and long-term battery value.

Conclusion

SoC tells you how much charge a battery has today. SoH tells you how much of its original capability remains. But neither tells the complete story on its own.The real opportunity lies in understanding how battery health is changing, why it is changing and what that change could mean in the future. By combining real-world data, physics-based models, AI and digital twins, battery teams can identify hidden degradation earlier, understand cell-level behaviour and make more informed decisions before a trend becomes a failure.

Want to understand where your batteries are heading? Talk to the SimLionics battery intelligence team.