Powerwall Degradation

[jasonacox]

How are you Powerwall battery capacities holding up over time?

The graph of my two powerwalls since I started measuring:

I want to add a plot line on there for the "expected degradation" to see how I'm trending. The warranty states that it should have at least 70% of the original capacity remaining at ten years. Anyone know how to add that?

[jasonacox]

Also - saw this great question on Reddit: Does the charging efficiency stay (relatively) constant no matter the maximum capacity, or is the efficiency reduced when the maximum capacity is reduced?

It would be interesting to see if we could determine and graph charing efficiency (e.g. $\Delta level/t$ per $kW$).

[jgleigh]

Do you have a Powerwall+ and a Powerwall2? I've got two of each. How do the 1-4 correspond to the actual units?

[jasonacox]

Yes, I do! You can see what telegraf is pulling by looking at: http://localhost:8675/pod

In your graph, it does look like 1 & 2 swapped with 3 & 4 last July? The pypowerwall code assembles and assigns those PW numbers in sequence from the http://localhost:8675/vitals (protobuf) payload. It assumes the protobuf sequence is always the same, which in general it is, but now your example has me wondering.

You can see the code here:
https://github.com/jasonacox/pypowerwall/blob/c9f20d7c20f99aae102fea9b70c4549109cf1e56/proxy/server.py#L218-L241

[jgleigh]

1&2 (at least now) are the Powerwall2 and 3&4 are the Powerwall+. Probalby no way to know how they reported last June.

[zip90]

Hello
I finally settled on a raspberry pi 4 with the rasp OS to run the dashboard. I am not a soft ware programmer. please let me know how to work out the PW 2 capacity degradation overtime from the dashboard. I like to thank to all those who have made it possible to get the PW function information to the end user.

[jgleigh]

@jasonacox I've been meaning to ask, what's the difference between PWx and PWx Charge?

[jasonacox]

Hi @jgleigh -

• The "PWx Charge" line is the current charge - energy remaining (POD_nom_energy_remaining) which varies as you charge/discharge the Powerwall.
• The "PWx" line shows the full charge capacity (POD_nom_full_pack_energy) which is what we see degrade over time.

You can see the "charge" cycles for my system here with the "capacity" line at the top.

[zip90]

@jasonacox

Thanks for the quick reply ,i have uploaded the PW capacity and charge. To monitor degradation over time do i follow the PW charge and see if it drops below 70% of 15.1Kwh . How come the PW1 in your image is 13.5Kwh

[jasonacox]

Use "PW1" not "PW1 Charge" to see the degradation. It will show the max "capacity" which is what degrades over time.

[zip90]

thank you

[zip90]

PW capacity monitoring what is Max and min. In my thinking the KWH will be gradually reducing over time ,so what is the min. Is it that the Max is the day one charge and over time when the capacity reduce it is visible as Min , thanks for the reply

…

On Thu, Aug 17, 2023 at 12:15 PM Jason Cox ***@***.***> wrote: Use "PW1" not "PW1 Charge" to see the degradation. It will show the max "capacity" which is what degrades over time. — Reply to this email directly, view it on GitHub <#264 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AZ2L524Z3WQS2PJCAEVYJSTXVV5EZANCNFSM6AAAAAAXSQ34MU> . You are receiving this because you were mentioned.Message ID: <jasonacox/Powerwall-Dashboard/repo-discussions/264/comments/6746497@ github.com>

[jasonacox]

The Powerwalls have an algorithm to instantaneously compute their "fully charged" capacity in kWh (think of it as the size of the fuel tank). That is what the dashboard is pulling for the "PW1" values. The Powerwalls also have a way to report "current charge" which is the amount of kWh (think of it as how much fuel is in the tank).

• PW1 - Max is the maximum "fully charge" capacity (size of tank) that the Powerwall reported since you have been measuring (Min is the minimum). The capacity will vary (up and down) but general trend is down over time (degradation)
• PW1 Charge - Max is the maximum "current charge" (how full the tank got) and will cycle as you charge/discharge over time. Min is the minimum amount.

[zip90]

Thank you

[ThePnuts]

I was installed April 2022, but didn't get grid approval until Sept so I used the battery heavily in this time to maximize the solar since I couldn't send to grid.

Once I got grid approval, the price difference for my peaks is not huge, so tesla suggests the battery sits full for backup if needed, and thats what I do.

I didnt get this setup to log until March of this year, so my data doesn't go that far back. I do have an interesting increase of capacity after a FW update though back in March.

I'm interested to see how this goes over time.

[MattMake]

Mine has only been measuring for a few months, max was 15.1, min 14.5. I use my powerwalls every day/night with 20% reserve.

Granted, the specs for Powerwalls are 13.5kWh each, starting out at 15.x is exceeding published specs. I wonder if the 70% capacity after 10 years is for the stated amount (13.5) or the actual amount (15.1)? At 13.5, that means I should have 9.45kWh for each Powerwall in September, 2032. I should achieve ROI on the solar+batteries in just about 9 years, so anything I get beyond that is just printed money :).

It will be nice to track this over time, thanks for taking the time to put this app together!

[jasonacox]

I do the same thing (use the Powerwall heavily each day/night). I always wanted to add a "Warranty Trend" line to the PW capacity graph to show if my Powerwalls were within the estimated degradation. I never had time to get that done. If anyone figures it out, it would be a cool feature to add. :)

[jgleigh]

My four units started at 14 kWh. The Powerwall2 are down to 13.5 kWh and the Powerwall+ are down to 13.7 kWh after two years of use.

[youzer-name]

I do the same thing (use the Powerwall heavily each day/night). I always wanted to add a "Warranty Trend" line to the PW capacity graph to show if my Powerwalls were within the estimated degradation. I never had time to get that done. If anyone figures it out, it would be a cool feature to add. :)

A horizonal warranty threshold line should be fairly easy to show in Grafana, but since you refer to it as a 'trend' line, I assume you mean a sloped line that goes towards the warranty level over the warranty time. but that would only be useful if the degradation is linear.

If the PWs degrade to their warranty level in 12 months and then hold steady, they would be well below a linear degradation line but also still not subject to a warranty claim. I don't know how much of a parallel can be drawn between the car batteries and the PWs, but my understanding of the car battery degradation is that most of it occurs early, and then it levels off.

[MattMake]

I completely agree that the line may cross 'expected' ahead of time. I personally would use it as a 'where am I' with battery degradation over time. It wouldn't be of any practical use until closer to warranty expiration, but having the data over time would be instrumental in the event it is below the 70% threshold. I'd have more ammunition going to Tesla with 10 years of data than a final kWh count near the 10 year mark.

[jasonacox]

My latest 2yr graphs show a gradual then clear increase in Powerwall energy capacity, after a year+ of degradation. I don't know if this is due to the recent firmware or the fact that I adjusted my system to go to grid power during the summer after peak (9pm) to handle a burst air conditioning load (our zone system activates to cools down the bedrooms at 9pm). In anycase, I'm happy to see this.

Example "on-grid for cooling" cycle:

[richardklam]

My April 2022 install was began logging on January 2023, and initially it showed 14.8 kWh, and then 2 months later bumped to 15.1 kWh (approximately 10 months old). I'm presently lat 14.2/14.3 kWh. What's interesting is the large capacity drops seem to be happening during the summer months when there is a lot AC usage. Most of the the year I only use 8-12 kWh during the peak, but summer it can be as much as 20 kWh.

Apologies for the data gaps but that was during the Tesla API change. Of the measured energy there has been approximately 5.3 MWh charged, 4.6 MWh discharged (approximately 200 cycles per PW).