What Does SOH Really Mean? Why One Number Isn’t Enough

Everyone talks about SOH. Dealers show it, apps measure it, certificates print it in bold. But what does this value actually tell you — and more importantly: what does it hide? We come from battery repair. We’ve seen broken batteries with 91% SOH. And perfectly healthy batteries with 78%. It’s time to explain that.

What Is SOH? The Short, Honest Answer

SOH stands for State of Health — it tells you how much usable capacity a battery has compared to its factory-new condition. A battery with a 100 kWh nominal capacity and an SOH of 85% can effectively store around 85 kWh.

Technically, the SOH battery value is a simple calculation:

So far so good. The SOH value is real, it’s measurable, and it’s relevant. The problem isn’t what it shows — it’s what it doesn’t show. And that’s quite a lot.

The Problem: Why SOH Alone Is Dangerous

Imagine you buy a used car. The seller shows you a certificate: SOH 88%. Sounds good. You pay €28,000 (roughly $30,000). Three months later the car suddenly shows an error message during fast charging, charging power drops, and the shop tells you: moisture in the battery. Water got in. Repair: €12,000 (roughly $13,000).

The SOH was correct. It correctly showed 88%. Nobody just asked what else was going on inside that battery.

Another example: Two Tesla Model 3s, both 2020, both with SOH 84%. Vehicle A has all 96 cell groups (Bricks) evenly aged. Vehicle B has three Bricks significantly weaker than the rest — one is nearly 18% below the group average. That’s a classic indicator of a Weak Short: a cell with an internal micro-short defect that’s slowly growing. The SOH? Identical in both. The battery? Completely different.

What SOH Battery Values Don’t Show

Here’s the list of things that the State of Health value systematically misses. Not because it’s calculated incorrectly — but because by definition it only captures one dimension of battery health: total capacity.

🌊 Moisture & Water Ingress

Tesla batteries have moisture sensors. Model 3 and Y use so-called floodport sensors, Model S and X actually measure grams of water accumulation inside the battery housing. Water in the HV battery is serious: it corrodes contacts, reduces insulation, and in the worst case can lead to thermal events. SOH doesn’t show this. At all. Not even close.

⚡ Weak Short Detection

A Weak Short is an internal micro-short in a single cell. The cell discharges more slowly than its neighbors, the SOC spread between bricks grows. The BMS compensates for this to some extent — with the result that the SOH value remains stable for a long time while the defect grows. A classic case of „everything’s fine“ until it isn’t. In technical terms: dSOC-drift — the deviation in State of Charge between individual brick groups over time. Also invisible in the SOH.

🔋 Brick-Level Imbalance

In a Tesla Model 3 Long Range there are 96 brick groups. Each group contains several cells in parallel. The SOH value is the average — but averages lie. A pack where 90 bricks are at 95% and 6 bricks are at 70% has the same average SOH as a pack where all 96 are at 90%. The reality is completely different. The first pack has three or four fewer years of life expectancy.

🔌 Internal Cell Resistance

Resistance determines how well a battery can deliver and accept current. High internal resistance means: fast charging slows down, power output drops in cold weather, and the battery heats up more under load. You feel all of this while driving. The SOH tells you: „I’ve still got 87% capacity.“ The internal resistance tells you: „But I can barely deliver peak power.“ Completely different information.

🛑 Active BMS Fault Codes

A Tesla’s Battery Management System logs dozens of fault codes. Some are harmless, some are serious, some are critical. BMS_f107 (internal battery fault), BMS_f123 (moisture), BMS_u018 (communication loss) — none of these automatically show up in the SOH. You can have an active critical fault code and still see a spotless SOH of 93%.

🛡️ HV Isolation

High-voltage isolation protects against dangerous leakage currents from the HV system to the vehicle chassis. Normal isolation: above 1,500 kΩ. Below 100 kΩ it becomes dangerous — electric shock from leakage current becomes possible. Tesla monitors this internally. SOH: doesn’t show it.

How SoHWHAT Does It Differently: 47 Parameters

SoHWHAT is not an app that reads the SOH and displays it nicely. Any OBD app can do that. SoHWHAT reads the battery at cell level — directly via the vehicle’s CAN bus, with a standard OBDLink MX+ adapter, no proprietary hardware, no waiting.

The result: 47 parameters that together paint a complete picture of battery health. A selection of the most important categories:

The difference from solutions like Aviloo is fundamental: Aviloo mainly reads summary values reported externally by the BMS — often an AI-estimated SOH without any real look inside the cells. We read directly from internal BMS registers. That’s not a marketing promise — that’s the difference between an ECG trace and a blood pressure monitor.

A Concrete Example From the Field

The TrueHealth Score: A Number That Actually Counts

We just said we have a problem with one number — and now we’re saying we have a number? Yes. But a different kind of number.

The TrueHealth Score (THS) is our proprietary rating system that combines all 47 parameters into a score from 0 to 100. The difference from SOH: it’s weighted, multidimensional, and honest about priorities. A battery with an active moisture fault cannot achieve a high THS — no matter how good its SOH is.

The six categories in the TrueHealth Score and their weighting:

Notice that SOH only makes up 12% of the TrueHealth Score? That’s not an accident. An 8-year-old Tesla with 78% SOH, perfect brick values, a dry battery, and clean BMS codes is a healthy car. A 4-year-old Tesla with 91% SOH and an active moisture fault is not. Only multidimensional assessment makes this difference visible.

Override Rules: When One Finding Overrides Everything

Certain findings are so critical that the THS is automatically capped — regardless of all other values:

• Active BMS_f107, BMS_f123, BMS_u018 → THS capped at 39/100
• HV isolation below 100 kΩ → THS capped at 49/100
• Active water detection → THS capped at 49/100

This is our way of ensuring that no score average conceals serious problems. A battery with water damage is not „basically okay, just a little worse.“ It’s a problem. The THS says that clearly.

SOH vs. TrueHealth Score — Direct Comparison

Conclusion: Enough With the One Number

SOH is not worthless. It’s one of many important metrics — and it has its rightful place in our rating system. But using it as the sole criterion for the battery health of an electric vehicle is negligent. When buying a used car just as much as when evaluating a repair.

We don’t say this because we’re selling an app. We say this because we’ve opened batteries, measured cells, seen the moisture damage firsthand. Because we know what a battery looks like that appears „fine“ on the outside and is rotting on the inside.

SoHWHAT was built to close exactly this gap. Not with more marketing, but with more data. 47 parameters instead of one number. Because a battery is not a one-dimensional thing.