Analysis of Powerwall Battery Retention

Ziga Mahkovec, Netzero Labs
Mar 1, 2025

Powerwall is a rechargeable lithium-ion home battery, manufactured by Tesla. It stores energy for backup power, solar self-consumption, and time-of-use load shifting[1]. Powerwall 2 entered mass production in 2017.

Like most lithium-ion batteries, the capacity of a Powerwall gradually diminishes due to its aging and usage (also known as calendar and cycle degradation[2]). Although various Powerwall 2 warranty versions exist, one common variant ensures 70% of the 13.5 kWh rated capacity for 10 years following its initial installation.[3]

Information about the real capacity and degradation of a Powerwall battery is not available from the Tesla app. However, this information can be accessed by connecting directly to the device for diagnostics. This analysis utilizes anonymously gathered statistics from Powerwall owners via Netzero, an app for managing home energy systems that also provides easy access to Powerwall diagnostic data.

Background Information

Powerwall Models

The Powerwall 2 Datasheet[4] lists the relevant specifications:

Powerwall 2 is believed to use Nickel-Manganese-Cobalt (NMC) battery cells.

Note that there are other Powerwall models:

Diagnostic Data

By connecting directly to a Tesla Gateway unit on the local network[5], diagnostic data can be retrieved from each installed battery. The data collected for this analysis includes:

Methodology

A random sample of 2,000 Powerwall 2 batteries was selected, along with the manufacturing date, part number, nominal full pack energy, and lifetime energy discharged. Powerwalls were limited to the ones manufactured in Tesla Gigafactory 1 (serial numbers starting with TG1). This excluded around 4% of samples, and was done to avoid outliers (e.g. refurbished batteries).

Some random noise was introduced into the manufacturing date, nominal full pack energy, and lifetime energy, to prevent individual battery identification. This random noise does not impact the observations or conclusions.

Analysis

Battery Retention by Manufacturing Date

The first scatter plot shows samples with manufacturing date on the X axis and nominal full pack energy (capacity) on the Y axis. The X axis is reversed to show the most recently manufactured Powerwalls on the left. The oldest Powerwalls in the sample set are from the end of 2017, making them 7 years old as of 2025.

Initial observations:

Further analysis is required to understand this outlier cluster by incorporating part number (hardware revision) as an additional dimension.

Battery Retention by Part Number

The scatter plot, now color-coded by part number, reveals that the high-degradation cluster consists of Powerwalls with part numbers starting with 2012170, manufactured between 2018 and 2020.

With a clearer understanding of the time degradation and outliers, the next analysis examined battery utilization.

Battery Retention by Discharged Energy

In this scatter plot, the third dimension is replaced with lifetime discharged energy, a measure of battery utilization. Discharged energy varies based on the number of cycles and the depth of each discharge. For example, a Powerwall owner who keeps their battery at 100% backup reserve—e.g. for grid outage protection—will have very low utilization and minimal discharged energy; on the other hand, keeping a low backup reserve and cycling the battery daily—e.g. for solar self-consumption—can result in high utilization. The highest discharge values in the sample set are at around 35 MWh. This amounts to over 2,590 cycles (assuming 100% depth of discharge at the nominal 13.5 kWh capacity), or a full cycle every day for 7 years.

This plot uses a sequential color scale, ranging from 0 MWh (purple) to 33 MWh (bright yellow) to represent lifetime discharged energy.

As expected, the plot shows that the Powerwalls with higher utilization also have higher degradation. For example, in the cluster of Powerwalls manufactured at the end of 2017, the difference in capacity between a low-utilization Powerwall and a high-utilization Powerwall is around 2 kWh.

There is no evidence of higher utilization for the outlier cluster with high degradation.

Powerwall 3 Battery Retention

Powerwall 3 launched in 2023 with higher power output and a built-in solar inverter. The capacity specifications are similar to Powerwall 2: 13.5 kWh nominal capacity[6]. Powerwall 3 is believed to use Lithium iron phosphate (LiFePO 4 or LFP) battery cells, which have longer cycle life compared to NMC battery cells[7].

A comprehensive analysis of Powerwall 3 battery retention requires more historical data. However, Tesla's modifications to Powerwall 3 have made diagnostics access more challenging[8], potentially limiting data availability compared to Powerwall 2. The scatter plot below shows a random sample of 500 Powerwall 3 batteries, analyzed using the same methodology as Powerwall 2.

Initial observations:

Conclusions

This analysis provides insights into the long-term performance and degradation trends of Powerwall 2 batteries. It presents evidence of reasonable performance consistency 7 years post-manufacture. However, some manufacturing batches exhibit higher-than-expected degradation. The battery utilization plot confirms the expected link between higher usage and increased degradation.

Future research directions include:

References

  1. Tesla Powerwall: Time-Based Control.
  2. Ageing mechanisms in lithium-ion batteries.
  3. Tesla Powerwall Limited Warranty (USA).
  4. Powerwall 2 Datasheet.
  5. Netzero: Powerwall 2 Diagnostics.
  6. Powerwall 3 Datasheet.
  7. Navigating battery choices: A comparative study of lithium iron phosphate and nickel manganese cobalt battery technologies.
  8. Netzero: Powerwall 3 Diagnostics.

Contact

For questions or comments contact Ziga Mahkovec at ziga@netzero.energy.