Data-driven battery insights: assessing second life value

Connected Energy has recently worked with a bus operator, using their data set to demonstrate these principles. We set out with four goals:

1. Integrate telematics data into our platforms
2. Offer fleet operators new insights into how their fleets currently operate – and help them use this data to optimise charging cycles
3. Accurately forecast state of health for each battery pack
4. Provide a realistic valuation for each battery pack in second-life applications.

Our analysis revealed that the batteries were at 95% state of charge for over a third of the time – demonstrating high vehicle readiness but leading to increased battery degradation. This identified immediate optimisation opportunities for the operator.

But most importantly, access to data from the vehicle telematics and the battery management system gave us the ability to accurately predict the state of health at end of vehicle life, forecast the pack’s value and estimate the optimal timing for sale into second-life energy storage applications.

We have developed four methods for estimating battery degradation, depending on the data available.

1. Capacity extrapolation

By examining the discharge cycles, we can see the total energy output and the depth of discharge. This is ideal for large datasets, though sensitive to SOC accuracy. In our pilot we found annualised degradation of 3 – 4.5% in line with expectations.

2. Incremental capacity

This method tracks charge curve inflections after deep discharges. This provides higher accuracy than capacity extrapolation but requires specific charging conditions which are not always possible in a fleet environment.

3. Voltage shifting

Even without incremental capacity analysis, we can still look at voltages over time. The SOC vs voltage curve shifts in three ranges. Plotting these enables us to accurately calculate degradation.

4. Voltage imbalance

This is particularly useful for older batteries, because degraded packs begin to see more instability in cell voltage. This can be used to reinforce or confirm evidence from the SOC vs voltage curve.

Together, these methods can give OEMs and operators powerful insight into real-world battery use.

A key challenge for other SOH modelling providers or OEMs is that access to second life performance data is not widely available to them. Unlike these providers, Connected Energy operates across the entire battery lifecycle – from first-life monitoring through to second-life deployment and operation.

This gives us unique insights into how different battery characteristics translate into real-world second-life performance, enabling more accurate valuations than theoretical models based solely on first-life data.

Our valuation methodology combines the battery health insights with second-life market data to produce accurate pack valuations. By knowing a battery’s current state of health, degradation rate, and usage patterns, we can predict its remaining useful capacity when it reaches end-of-life in the vehicle – typically when it retains 70-80% of original capacity.

This predicted capacity directly correlates to the battery’s value in second-life applications. A pack with 75% remaining capacity will deliver more energy storage and have a longer operational lifespan in a BESS than one with 70% capacity. We factor in additional parameters such as cell voltage balance, thermal history, and cycling patterns to refine our projections. Combined with current market pricing for second-life batteries of different capacities and conditions, this enables us to provide OEMs and fleet operators with concrete financial valuations – turning abstract battery health data into tangible asset values that can inform end-of-life planning and total cost of ownership calculations.

Most importantly, what this data can identify is the optimal time to replace the battery and move it into second life use. For fleet operators, it may seem logical to delay replacement for as long as possible to maximise first-life use. But this overlooks a critical factor – the relationship between battery degradation and second-life value isn’t linear and there is a cut-off point where the battery no longer has value in second life. We’ll be covering this topic in more depth in a future blog.

For battery owners, the insights that we can gain from first life use can help in a variety of ways:

• Extending vehicle working life through better maintenance and planning
• Providing concrete valuations for end-of-life batteries
• Providing insight into the timing of battery swaps to maximise second-life revenue
• Reduce costly pre-BESS testing, saving logistics costs and accelerating deployment

For OEMs, this approach de-risks their entry into the second-life value chain by providing accurate insights into battery pack performance, health and degradation.