Author: Dominic Pascal Förstermann, Akkuracy
Battery chemistry choices affect everything from product safety certifications to insurance and procurement decisions — yet the common viewpoint remains “LFP is safe, NMC is dangerous.” But is that actually true? This article looks at the empirical data.
In order to describe the safety aspects fully, it is split into two parts. The first looks at the temperature sensitivity of the cell chemistry, to describe the cell behaviour leading up to a thermal runaway event. The second part looks at the event itself and how to describe its severity.
Sensitivity to Excessive Temperature and Propagation
Understanding the sensitivity of a cell to excessive temperature is crucial, as it determines how carefully the thermal environment of the cell needs to be managed to prevent overheating.
All cell chemistries ultimately start to decompose at elevated temperatures, releasing additional energy in the process. If this leads to the cell self-heating, it will ultimately enter thermal runaway itself.
This can occur, e.g. during a thermal runaway of an adjacent cell. Here, the released heat from one cell can cause adjacent cells to be triggered. To describe the sensitivity, two temperatures are of importance.
Self-Heating Onset Temperature (T1)
The self-heating onset temperature is the temperature at which a cell begins self-heating under adiabatic conditions. It describes the lowest temperature at which a cell will begin to heat up, if it cannot pass any heat to its environment. It is therefore a theoretical, worst-case lower value, as cells in real-world battery packs can readily exchange heat.
NMC/graphite cells begin self-heating at approximately 90 – 110 °C, while for LFP/graphite it occurs around 150 – 170 °C.
Critical Self-Heating Temperature (T2)
The critical self-heating temperature describes the threshold beyond which thermal runaway becomes almost certain. At T2 the cell is self-heating so rapidly that the reaction cannot be stopped anymore.
For NMC/graphite cells this occurs at around 170 – 190 °C, compared to 230 – 260 °C for LFP/graphite.
The data for T1 and T2 show that LFP is significantly less sensitive to temperature. At a temperature where the first reactions occur in LFP cells, NMC cells are already close to a full thermal runaway event.
Data for a thermal runaway test of a 21700 cell with NMC-811 cathode and graphite anode. The cell was overheated in steps of 5°C until self-heating was achieved at T1. The critical self-heating temperature T2 is highlighted right before the thermal runaway event occurs and the maximum temperature of the test T3 is observed.
Adapted from Jan Schöberl et al [3].
Severity of the Thermal Runaway Event
Once the cells have reached a point past T2, they will likely enter a thermal runaway event. The cell releases significant amounts of energy very fast, often leading to fire or explosion.
The most objective metric to describe this phenomenon is the total energy released by the reaction (Qtotal). The larger Qtotal, the more energy the cell releases, and therefore the more violent the reaction.
NMC/graphite cells release approximately 20 – 25 kJ/Ah during thermal runaway, while LFP/graphite cells tested under the same conditions released only approximately 10 – 15 kJ/Ah – approximately half the energy.
| Self-Heating Onset Temperature (T1) [°C] | Critical Self-Heating Temperature (T2) [°C] | Thermal Runaway Energy Release (Qtotal) [kJ/Ah] | |
| LFP | 150 – 170 | 230 – 260 | 10 – 15 |
| NMC | 90 – 110 | 170 – 190 | 20 – 25 |
What Does This Mean in Practice?
LFP is the safer chemistry at the cell level – the data supports this clearly. Safe battery systems, however, can be built with both chemistries. NMC simply requires greater development effort to achieve the same safety margins, with the trade-off being higher performance.
It’s also worth noting that aspects such as cell size, ageing history, state of health (SOH), reaction atmosphere, and – critically – anode material all have a significant influence on the parameters discussed here.
About Akkuracy
Akkuracy Battery Testing is focussed on providing the highest quality battery safety data possible. We are passionate about understanding thermal runaway and provide detailed insights into it using our DATRC calorimetry method and conventional testing. Our laboratory is located in Munich, Germany. Find out more at akkuracy.com.
References
- Akkuracy Battery Testing GmbH. Internal DATRC calorimetry data.
- Jan Schöberl, Stefan Schaeffler, Linus Grahl, Christoph Bach, Sebastian Ohneseit, Dominic Förstermann, Xuning Feng, Carlos Ziebert, Andreas Jossen, Markus Lienkamp, Modeling the variability of thermal runaway behavior in large-format cylindrical cells with NMC-811 and LFP cathodes: A simulation study, Journal of Power Sources, Volume 668, 2026
- Jan Schöberl, Sebastian Ohneseit, Stefan Schaeffler, Dominic Förstermann, Linus Grahl, Andreas Jossen, Carlos Ziebert, Markus Lienkamp, Thermal runaway characterization of cylindrical lithium-ion and sodium-ion batteries with various sizes and energy contents, Journal of Power Sources, Volume 648, 2025