Battery SoH: The Number That Doesn’t Tell the Whole Story [The Battery Cycle #5]

In 2021, Claudius Jehle, CEO of volytica diagnostics GmbH, and Sustainable Bus launched a series of articles around the “The Battery Cycle”. They shed light on the complexities of Li-Ion batteries and provided valuable insights for anyone involved in electric mobility. In 2025 and 2026, the articles were updated with real-world data and lessons learned to understand how battery chemistry affects real-world operations. (The article was first published in February 2022, this is a revised version.)
For questions and contributions please reach out to us via contact@volytica.com.

Terminology – What is the State of Health?

The term “State of Health” or “SOH” is easily the most controversial term in the battery industry. No other – spoiler: un- or ill-defined – term is used so liberally and in so many contexts and for so many purposes as SOH. If we could turn back time, we would try to get rid of it altogether, hence we often tend to put it in “quotes”. For reference: Almost all financial warranty risk of the whole industry is tied to “SOH”, but the industry lacks a universal definition of this: Unprecedented in the younger industrial history.

But as a matter of fact, it is there, so what is it, and what is it not? Let’s find out.

To set the stage, we will use SOH broadly speaking as a relative (i.e. in %) measure for battery capacity – and not as a measure for resistance, performance, safety or so. But trust me, this is complex enough.

Phenomenology – How Degradation Makes Itself Noticeable. Or is Degradation Actually So Relevant?

SOH is widely used in degradation contexts, i.e. how much is left after so-and-so-many years, and this is not wrong, but certainly also not the whole truth. The fact alone that the amount of dis/chargeable energy depends on the conditions, and that disbalance between cells has a significant impact on the system performance, suggests that not only long-term effects play a role here.

In the most general and widely accepted definition, State of Health is a momentarily available capacity, measured in a balanced (!) battery by discharging (!) from BMS 100% SOC to BMS 0% SOC, using a constant current (or power), and that result divided by a reference value, e.g. a nominal type-plate value.

You might recognize that this sentence – and we even simplified it – offers ample room for interpretation, argumentation and dispute. Words like “momentarily”, “available”, charging vs. discharging, constant current vs. constant power, “capacity”, and “reference” allow different perspectives. Let’s start with the most stunning: “available”.

It might not come as a surprise that the BMS has a word to say about how it controls the battery and e.g. the voltage limits. We learnt in recent articles that it maintains an “operational” State of Charge, the one we see on our displays and tools, and that 0/100% does not “technically” mean empty/full – it’s just what it makes available to you, retaining some technical reserves.

“Eat the Buffer” – When Degradation Stays Hidden

So far, so obvious. But imagine the BMS always making the same amount available to you, year over year? You bought an asset with 300kWh capacity, and you perform a test at delivery: Voila, 300kWh. And you use it heavily, and in year 5 your test yields: 300kWh. Some might be proud (“Look how careful and battery stress-aware we are!”), some might be skeptical (“How can that even be, why does it not degrade?”).

This little thought experiment is very relevant and various battery manufacturers employ this “Eat the buffer” logic: Excess capacity is installed, and constantly “eaten away” by degradation – but shielded from the user as long as possible. Only when the technically available capacity has degraded so far that it falls below the programmed threshold, the operational capacity is reduced.

The capacity that you “experience” in daily operation, and during capacity tests, is first-and-foremost software-defined; it is not necessarily affected by degradation, depending on the implemented BMS strategy.

You need proof? Here you find both the technical (reddish) and the operational/ net capacity (blueish) for a ca. 500kWh e-bus for 1 year, analyzed by volytica’s continuous monitoring engine. You will clearly observe at least 2 things:

1. The BMS of this e-bus employs such “eat the buffer” strategy: The operational capacity stays absolutely constant, …
2. … while the “real” capacity does, albeit slowly, show a downward trend.

In ca. 7 years from now, the buffer will have been consumed, and the BMS will need to reduce the operational capacity in line with the rate of degradation. Without such advanced and continuous analysis, owners and operators are left with the very labor-intensive manual capacity checkups, that never show the technical capacity, but only the – here always constant – operational.

“Keep the Buffer” – When Degradation Appears Early

Of course, there is another such philosophy, let’s call it “Keep the buffer”, which is also frequently employed and feels more natural to most people – because it, at least apparently, does degrade: The operational capacity is not kept constant by software, but it decreases more or less parallel to the technically available capacity.

This requires much more sophistication from the BMS, as it must have means to closely track the actual, electrochemical capacity, much more frequently and precisely than in the other case. If it can’t, because it is technologically a challenging endeavor, and it assumes a rate of decrease that is steeper than the actual degradation, the perceived loss of performance and value is unnecessarily high!

• The BMS defines the capacity you get, and it cannot (and legend has it, in rare cases does not want to) tell the entire truth about the full technical capacity
• BMS systems follow shades of grey between the 2 fundamental strategies: “Keep or Eat the Buffer”; the latter leaving a “blind” operator with the impression of no degradation, infinite lifetime and perfect value retention
• It is recommended that owners have a robust understanding of different interpretations of, and a continuous and holistic picture of, the SOH/ residual capacities of their fleet. Manual capacity tests won’t give these insights.

SOH, “End of Life” and Safety

We might dedicate a separate article on this, but as the topics are too closely related, let’s speak about it now: A hard-to-eradicate rumor has it that a battery is at its “End of Life” once “the” SOH reaches an OEM-defined threshold, say 70%.

There is already a kind-of obvious catch here, namely that “the” SOH is the operational, day-to-day available net capacity, which is programmed and decided by the BMS. As we learned, a battery might have plenty of technical reserves left, even if the BMS-released SOH already reached said threshold. We have seen cases of operational SOHs that reach 70.0% exactly on the last day of an 8-year warranty, but our analysis yielded plenty of remaining technical capacity reserves – evil to him who evil thinks.

But even if: Speaking of End of Life is grossly misleading – it might be the end of the warranty, but for sure a battery is not dangerous or in other forms “dead”. As we will learn in later articles, safety for instance is, if at all, only weakly linked to residual capacity! Many other factors are way more important and significant to judge the end of life!

We strongly advocate to replace the notion of “SOH low = End of Life” with “End of Warranty”; to determine the true “End of Life” of such a precious and versatile component, one better takes a few looks and checks more!

State of Health at a Glance

• The capacity you experience in daily operations and during capacity tests is primarily software-defined. Depending on the implemented BMS strategy, it is not necessarily affected by degradation.
• The BMS defines the capacity you get (and, legend has it, in rare cases does not want to) tell the entire truth about the full technical capacity.
• BMS systems follow shades of gray between the two fundamental strategies: “Keep or Eat the Buffer.” The latter leaves a “blind” operator with the impression of no degradation, an infinite lifetime, and perfect value retention.
• Manual capacity tests might yield very optimistic results that are not per-se representative for daily operations. The outcomes can only be used for operational dispatching with a huge grain of salt.
• We strongly advocate replacing the notion of “SOH low = end of life” with “end of warranty.” To determine the true “end of life” of such a precious and versatile component, it is better to take a few looks and perform more checks!

Looking Ahead in the Series

In the next article on Stress Level, we will examine degradation stress factors in-depth to understand what actually “hurts” your batteries in daily operations.

All knowledge articles of the battery cycle:
Intro – The Battery Cycle – opening the black box
1 – NMC, LFP, LTO: What’s the Difference in Battery Chemistry? – energy density, safety, lifetime, cost
2 – State of Charge: Why It’s Harder to Measure Than You Think – about really knowing how full your battery is
3 – Fast Charging Explained: Why More Power Doesn’t Mean Less Time – how to keep a battery healthy
4 – Why an LFP Bus Can Suddenly Stop: The Battery’s Weakest Cell – why imbalances define the limit
5 – Battery SoH: The Number That Doesn’t Tell the Whole Story – State of Health is mostly misunderstood
6 – Stress Level: The Key Drivers of Battery Degradation – what really hurts a battery
7 – Battery Data: Are You Seeing the Full Picture? – use and interpret your data correctly