State-of-Health Assessment of the Molicel P45B: How Long Will My eVTOL Battery Last?
- Dr. Miles Robertson
- 6 days ago
- 5 min read
Battery ageing is a major challenge for eVTOLs due to high-power demands during take-off and landing. These aggressive duty cycles can drive rapid degradation, especially with fast charging and cooling. This blog assesses the Molicel P45B’s performance under such conditions using validated models.
Ageing: A key challenge within the eVTOL industry
The ageing of lithium-ion battery cells presents a significant technical hurdle to the commercial success of electric Vertical Take-Off and Landing (eVTOL) aircraft.
A characteristic feature of eVTOL operation is the requirement for high-power pulses bookending any flight to drive the vertical hover used during both take-off and landing. Whilst providing superior manoeuvrability and reduced footprint compared to its more conventional cousins, eVTOL duty cycles are significantly more severe than most other aerospace and automotive applications.
This point is most pertinent when it comes to battery cell selection and pack design; pulses of up to 5C risk rapid cell degradation over time - often compounded by the desire to fast-charge and cool upon landing to meet turnaround-time and thermal requirements.
Therefore, eVTOL developers are often posed with tough questions such as:
1. At which severity of ageing should we be concerned about failing our requirement cycles?
2. When are we likely to reach that point - how long, and how many cycles?
The inability to estimate these factors may lead to compromises to aircraft range and/or payload, as the requirement for safety forces excess conservatism in the face of uncertain data. These are the questions where access to high-quality data and simulations, early in the design process, can provide clarity to engineers and managers.
Cell model validation at About:Energy using Molicel P45B
In this example we will be focusing on a popular cell for eVTOL designers, the Molicel P45B - with measured power and energy densities of and 2314 W/kg and 231.4 Wh/kg respectively, it presents a commercially available solution which balances performance and cycle life within a familiar 21700 cylindrical format:
Prior to this example, an Equivalent Circuit Model (ECM) for P45B performance was characterised in our laboratory, using our Peltier Element controller to accurately set surface temperatures. The parameterised cell model was then re-tested against previously unseen data, representative of an eVTOL cycle, with high-current pulses at take-off and landing, followed by a short on-ground rest.
Good agreement was seen in a comparison of cell voltage between experiment and model, with tests being conducted at a range of enforced surface temperatures. RMS errors were in all cases constrained below the ~20 mV range, despite some larger errors for short durations during the highest discharge rate events:
Predicting eVTOL mission viability as cells age
Given a well-characterised model, we can start to look at how State-of-Health (SOH) impacts on performance across a duty cycle. Typically in an ECM, we break SOH into two constituent effects, which change as the cell ages - these are represented as linear scaling factors to (a) increase cell resistance and (b) decrease cell capacity, such that:
Before delving into how our models and outputs change over time, it is prudent to first look at how cells stand against requirements at beginning of life (BOL). For the purpose of this analysis we will define success based on (a) maintaining the maximum cell temperature below a value of 65°C, and (b) keeping cell voltage above the minimum of 2.5 V at landing. We have also applied ‘warning’ ranges to provide a safety margin, at 60°C and 2.8 V.
Running the baseline case shows that we are passing - good news! Cells reach 53°C and 3.4 V, which are within specification - but given the currently unknown ageing profile they will be subjected to, we maintain an interest in analysing how this changes during operation:
Often when running a cell or pack model, we will want to perform sensitivity and response analysis - this is the domain of Design of Experiments (DOE). We will typically vary one or more input parameters, run simulations, and then observe how output parameters of interest change in response. This can be for multiple purposes:
A sensitivity analysis to determine how outputs are statistically correlated to inputs - often to determine which inputs we would be best placed to observe/change/modify to order to control or optimise the given output.
Generating a response surface in order to determine how the output changes with regard to input(s) - a response surface is a model in itself, which can then be used for other purposes (such as a lookup within a wider study).
A precursor to numerical optimisation, where we seek to algorithmically minimise an output variable by choosing a certain set of inputs, whilst optionally constraining other parameters.
In this instance we are varying the two SOH parameters, with an interest in how our two responses (landing voltage and maximum temperature) respond. To do this, we have generated a Full Factorial experiment, which provides a regular grid of data points across the two inputs:
After the experiment is generated, the values from each data point are successively applied to the About:Energy ECM and run across the eVTOL duty cycle. Each response is recorded into a grid (matrix) and finally plotted as a heat map.
Results show that:
Peak cell temperature is primarily correlated to SOHR - an unsurprising result given that that the irreversible heat generation I²R is dominant.
Minimum cell voltage is strongly correlated to SOHC - again this is intuitive given the known relationship between voltage and State-of-Charge.
Given the above results, we can apply a pass/warning/fail to the different regions of our 2D SOH map, to get insight into to what extent ageing will impact our ability to meet our requirement cycle - in this case, SoHC = 0.625 and SOHR = 1.4 appear to be the points at which, in isolation, each of these mechanisms begin to cause concern to operation. Of course, there is a locus points between these extremities which can be extracted.
It is worth bearing in mind that the above analysis is conservative in that it is a 2-parameter study which does not account for variation in the innumerate eVTOL parameters which may experience statistical variation - this is a far larger and more complex question!
But, we have begun to bound this problem and provide valuable insight which - using About:Energy’s data and simulation models - can provide answers on viability early in the design phase, when decisions such as cell selection are still on the table.
What next?
Knowing the severity of ageing is an important part of the puzzle - it lets us put constraints on the level of acceptable ageing for a given cell during operation. However, this must be coupled with in-context predictions of degradation, allowing determination of how many cycles, or how long, can be expected before reaching these critical conditions.
At About:Energy, we specialise in cell degradation testing, and simulation. Our data and models allow accurate forecasting of capacity fade and resistance growth during operation - allowing customers to design Battery Management Systems, size thermal systems, and ensure their performance requirements are met across the lifetime of their products.
If you’re interested in learning more about how our models can help you, book a demo here.
Further reading:
Debating between the Molicel P45B and P50B?
Want to know more our Molicel models?
Learn more about our Molicel models series
Reduced development time by 70%: McMurtry Case Study
Read our Forbes Feature
