The 3 biggest challenges in EV battery safety: SOH, SOC, and thermal Runaway

Battery Safety Intelligence is quickly becoming a must-have for electric fleets, not because battery fires are common, but because the downside is massive when they happen. Most battery safety conversations eventually come back to three technical challenges. If you understand these, you can make better decisions on charging policy, maintenance, procurement, and depot risk management.

Here are the 3 biggest challenges in EV battery safety:

SOH (State of Health): slow degradation that quietly removes safety margin
SOC (State of Charge): daily operating stress that can accelerate risk
Thermal runaway: a fast-moving failure mode that’s hard to stop once it starts

Below, I’ll break each one down in plain terms, then show how they combine in real fleet environments and what you can do about it.

EV battery safety is now an operational continuity issue

For fleets and energy operators, the biggest cost of a battery incident isn’t just the vehicle. It can include:

damaged charging infrastructure or depot facilities
multi-day downtime and rerouting
investigations, reporting, and insurance friction
reputational impact with customers, regulators, and employees

Battery fires are low-frequency events, but they’re high-consequence events. That’s why prevention matters.

Challenge 1: SOH (State of Health) — the risk that builds quietly

SOH describes how a battery has aged compared to when it was new. It’s often treated as a “capacity” topic, but it’s also a safety topic.

As batteries age, you typically see:

capacity fade (less usable energy)
higher internal resistance (more heat during charge and discharge)
uneven cell aging (one weak cell can become the problem)

Here’s the key point that’s easy to miss:

As SOH declines, safety margin shrinks.

A pack that handled stress comfortably in year one may become far less tolerant in year four.

Where fleets get exposed

Fleet operations are tough on batteries: high utilization, frequent fast charging, and long dwell times at high SOC. Even if each action is “within spec,” the accumulated stress changes the battery’s internal condition over time.

Why traditional monitoring isn’t enough

Most BMS approaches estimate SOH using voltage behavior and charge throughput. That’s useful, but it doesn’t always catch early-stage instability. A degraded battery doesn’t always announce itself with a big temperature spike or a clean error code. Sometimes the earliest warning is subtle: small voltage drift, unusual heating under normal load, or patterns that only become obvious when you compare one asset against a fleet baseline. In practice, the hardest part of SOH risk is that it’s easy to underestimate until it isn’t.

Challenge 2: SOC (State of Charge) — daily stress that can amplify fire risk

SOC is the battery’s “fuel gauge,” but it’s also a stress dial.

High SOC means higher voltage and more electrochemical stress.
Very low SOC can create its own problems, including damage that shows up later during charging.

The problem isn’t that high SOC automatically causes fires. It’s that high SOC increases exposure when other risk factors are present. Here’s a quote-friendly version: SOC doesn’t just affect range. It affects how hard the battery is being pushed in that moment.

The compounding effect: SOC + degraded SOH

This is where things get real for operators. When SOH is lower (older packs, harder life), the battery has less headroom. Holding those batteries at high SOC, or repeatedly pushing them toward extremes, can narrow that headroom further. So the risk isn’t “SOC” in isolation. It’s SOC interacting with battery age, duty cycle, and cell-to-cell imbalance.

The operational blind spot

Many teams assume: “If we’re inside SOC limits, we’re safe.” That’s an understandable assumption, but it’s not always true. A battery can sit inside normal SOC bounds while its chemistry is moving toward instability. Thresholds catch obvious violations. They don’t always catch early failure signatures. This is one reason battery fire prevention is getting more attention in fleet settings. It’s not enough to manage charge levels. You also need to understand the battery’s condition and behavior.

Challenge 3: Thermal runaway — hard to stop once it starts

Thermal runaway is the event fleets fear most: a self-accelerating reaction inside a cell that can lead to ignition and, in some cases, propagation through the pack. Two realities make thermal runaway especially difficult to manage:

the reaction can generate its own heat and oxygen as it progresses
suppression is often reactive because the root reaction is inside the cell

Once a cell enters runaway, your options shift from “prevent” to “contain, protect people, and minimize damage.”

What testing has shown (and why it matters)

Large-scale EV fire testing has highlighted recurring gaps in the market:

warning information that isn’t clear or actionable
insufficient warning for people nearby when a hazardous event begins
inconsistent performance in delaying propagation and protecting the cabin

For fleet operators, the takeaway is simple: If you only learn about risk when smoke appears, you’re already late.

That’s why thermal runaway prevention increasingly focuses on early detection, not just stronger suppression.

How these three challenges combine in the real world

SOH, SOC, and thermal runaway aren’t separate topics. They’re connected.

A clean way to think about it:

SOH sets the baseline safety margin (how much stress the battery can tolerate)
SOC influences daily stress exposure (how hard you push it operationally)
thermal runaway is the failure mode when conditions cross a critical line

In fleets, this combination shows up in very normal scenarios:

an older vehicle on a demanding route
repeated fast charging to keep utilization high
high SOC overnight at a depot
a single weak cell that’s drifting from the rest

None of these on their own guarantees an incident. But together, they can raise risk in ways that traditional dashboards don’t make visible.

What forward-looking fleet operators should do now

You don’t need to “panic-proof” your fleet. You need a practical plan. Here are five moves that consistently reduce risk without killing uptime:

Treat SOH as a safety KPI, not just a maintenance metric.
Track degradation patterns, not just averages. Outliers matter.
Reduce unnecessary time spent at high SOC.
Avoid full-charge dwell time when you don’t need it operationally.
Build exception handling into your operating model.
Define what happens when an asset behaves oddly, even if it’s still “within limits.”
Evaluate your warning stack.
Ask: “How early would we know if a battery was trending toward instability?”
Add a safety intelligence layer that works with your existing BMS.
This is where Battery Safety Intelligence comes in: software-led, data-driven detection that looks for early patterns that threshold alarms miss.

For EPTTAS, that means an AI-powered early warning and battery intelligence platform that can detect the earliest signs of thermal instability, often up to 60 minutes before ignition, so operators can intervene early and protect people, assets, and operations.

Closing: the future of battery safety is early detection

As electrification scales, the stakes change. More vehicles and more storage means higher energy density in more places: depots, garages, ports, warehouses. You can’t suppress your way out of that. You have to detect risk earlier. The three biggest challenges are clear: SOH, SOC, and thermal runaway. The operators who manage them best will be the ones who combine strong operational policy with modern, AI-driven early warning. If you’re evaluating your approach to battery fire prevention, it’s worth a direct conversation.

Contact us to discuss how Battery Safety Intelligence can fit into your fleet safety program and work alongside your existing BMS and hardware controls.