Powerwall Efficiency

[PBBear]

Thanks mikemillar, my initial reading based on 4 days and a total export of 7.3 kWh says my efficiency was only 75%.

After a year-and-a-half, and correcting for the current SOC of 1.8 kWh stored, my stats are
"energy_exported": 6319520,
"energy_imported": 7035558.2, efficiency 89.82% (close enough to 90% for my purposes).

 

[Briar_Reed]

My typical discharge is around 600-800W, when charging during the day it will charge as high as 5kW from solar. I see around 85% round-trip efficiency with my single PW (~10kWh in to get ~8.5kWh out). It's harder to determine long-term averages since the app stopped reporting total kWh in to the PW beyond the single-day view (at first that sounded like a bug, but they've done a bunch of app updates since then and the measurement has never returned), but any time I've calculated it the answer has been right around 85%.

I just reviewed my data from the Tesla app, and for the months Mar thru May, 2020, my round trip efficiency was 84.2%, 84.9% & 85.7% respectively, much like your observations. And easy to see that it improves as the weather warms. Tesla advertises 90% RT efficiency on the PW2, so wondering where the additional 4-6% losses may be coming from, any thoughts?

 

[woferry]

Probably just not being at the optimal charge/discharge rate, which seems to be ~3.3kW on charging, no idea on discharge. May also matter how deep the discharge goes, e.g. going from 100%-50%-100% vs. 100%-20%-100%.

 

[bkp_duke]

I just reviewed my data from the Tesla app, and for the months Mar thru May, 2020, my round trip efficiency was 84.2%, 84.9% & 85.7% respectively, much like your observations. And easy to see that it improves as the weather warms. Tesla advertises 90% RT efficiency on the PW2, so wondering where the additional 4-6% losses may be coming from, any thoughts?

Indoor or outdoor installation?

PW2s like all Li-Ion batteries have an "optimal" temperature range, and I'm betting those installations inside a garage in a near 70F environment have higher average round-trip efficiencies.

 

[power.saver]

Interesting. I just checked on mine, 2 PWs outside, temps 80F-90F daily, with typical 50% DOD each day.

For past 3 months, efficiency (energy out/energy in) has been 88%, 87%, 90%. But last December was 83%. Lifetime is 86%.

 

[bkp_duke]

Interesting. I just checked on mine, 2 PWs outside, temps 80F-90F daily, with typical 50% DOD each day.

For past 3 months, efficiency (energy out/energy in) has been 88%, 87%, 90%. But last December was 83%. Lifetime is 86%.

That corresponds with what we know about the cells in these, they prefer warm over cold.

Same reason we get a lot less mileage on the cars in colder weather.


How do you query efficiency over different time periods? My 3 PW2s are in the garage (insulated, temperate climate), and I would like to see what their efficiency over the past 2+ years has been.

 

[power.saver]

That corresponds with what we know about the cells in these, they prefer warm over cold.

Same reason we get a lot less mileage on the cars in colder weather.


How do you query efficiency over different time periods? My 3 PW2s are in the garage (insulated, temperate climate), and I would like to see what their efficiency over the past 2+ years has been.

I have a program running that polls the meters (https://powerwall/api/meters/aggregates) once per minute, and I log the data. At the end of each day, I save the cumulative energy_exported and energy_imported, as well as some other stuff) so I can see how the system is performing. That way I can see the trends over time.

You can make that call from your browser (substitute "powerwall" with your Gateway IP) and it will show you the lifetime readings in the Battery section. If you save the response, you could compare it to another response sometime in the future. There is a timestamp in it too.

 

[bkp_duke]

I have a program running that polls the meters (https://powerwall/api/meters/aggregates) once per minute, and I log the data. At the end of each day, I save the cumulative energy_exported and energy_imported, as well as some other stuff) so I can see how the system is performing. That way I can see the trends over time.

You can make that call from your browser (substitute "powerwall" with your Gateway IP) and it will show you the lifetime readings in the Battery section. If you save the response, you could compare it to another response sometime in the future. There is a timestamp in it too.

Here is what I found with that, looks like things are broken apart by "site", "battery", "load", and "solar":

{"site":{"last_communication_time":"2020-06-14T14:12:57.689276178-07:00","instant_power":-7841.820068359375,"instant_reactive_power":
930.4500122070312,"instant_apparent_power":7896.827160938746,"frequency":60,"energy_exported":9036467.870800734,"energy_imported":
2222033.749967402,"instant_average_voltage":123.13999938964844,"instant_total_current":0,"i_a_current":0,"i_b_current":0,"i_c_current":0,
"timeout":1500000000},

"battery":{"last_communication_time":"2020-06-14T14:12:57.69014984-07:00","instant_power":0,"instant_reactive_power":
970,"instant_apparent_power":970,"frequency":59.974000000000004,"energy_exported":13330450,"energy_imported":15489510,
"instant_average_voltage":246.16666666666666,"instant_total_current":-0.6000000000000001,"i_a_current":0,"i_b_current":0,"i_c_current":0,
"timeout":1500000000},

"load":{"last_communication_time":"2020-06-14T14:12:57.689276178-07:00","instant_power":1520.2871897619902,
"instant_reactive_power":-221.83954852212784,"instant_apparent_power":1536.3872964337186,"frequency":60,"energy_exported":0,"energy_imported":
36065024.83527778,"instant_average_voltage":123.13999938964844,"instant_total_current":12.346006149889512,"i_a_current":0,"i_b_current":0,
"i_c_current":0,"timeout":1500000000},

"solar":{"last_communication_time":"2020-06-14T14:12:57.690293173-07:00","instant_power":9374.58984375,
"instant_reactive_power":-2127.97998046875,"instant_apparent_power":9613.076174451986,"frequency":60,"energy_exported":45038562.16583423,
"energy_imported":43.209723122677474,"instant_average_voltage":123.1349983215332,"instant_total_current":0,"i_a_current":0,"i_b_current":0,
"i_c_current":0,"timeout":1500000000}}

Under the "battery" section I'm seeing:
"energy_exported":13330450
"energy_imported":15489510

What are the units? What is the timeframe? (day?)

 

[miimura]

Here is what I found with that, looks like things are broken apart by "site", "battery", "load", and "solar":


Under the "battery" section I'm seeing:
"energy_exported":13330450
"energy_imported":15489510

What are the units? What is the timeframe? (day?)

If I had to guess, I would say it's lifetime Watt-Hours.

 

[power.saver]

Here is what I found with that, looks like things are broken apart by "site", "battery", "load", and "solar":


Under the "battery" section I'm seeing:
"energy_exported":13330450
"energy_imported":15489510

What are the units? What is the timeframe? (day?)

The units are Wh, and this is cumulative since install. To get daily or hourly values, you have to poll the GW at the interval you want, and subtract from the previous reading.

There is a lot of good information in that request, you can also see cumulative solar along with instant_power, frequency and voltage.

Yours have a lifetime efficiency of 86% too.

 

[bkp_duke]

The units are Wh, and this is cumulative since install. To get daily or hourly values, you have to poll the GW at the interval you want, and subtract from the previous reading.

There is a lot of good information in that request, you can also see cumulative solar along with instant_power, frequency and voltage.

Yours have a lifetime efficiency of 86% too.

86%, bit lower than I expected, but acceptable. Would have been nice to trend the data over time if I knew, see if / how efficiency degrades.

 

[LesMoss]

But, you need to know your system's daily loss without power draw (i.e. in Backup mode). My 2 PW system seems to have a daily usage of 0.5 KWh. So, I need to reduce that "energy_imported" number by 500 per day of operation.

Using that methodology I see >90% efficiency. But, I have only 2 months using time shiting. Most of my time has been spent in Backup. Mine are mounted outside, but in a mild west coastal climate.

 

[Redhill_qik]

I have only recently turned on my system, so I don't have a lot of data right now, but I do have concerns about the efficiency of the Powerwall.

My limited data for two PWs installed in the garage.

• Pre-PTO - Charged 2x PW to 100% then turned them off (and solar) for 12 days and they drew 3.5 kWh to get back to 100%, so 0.292 kWh/day leakage or 1.08%/day
• Post-PTO - Left 2x PW at 100% on and in Backup only mode for 17 days and they periodically charged a total of 6.1 kWh, so 0.359 kWh/day or 1.33%/day for leakage and operation.
• Post-PTO - 2x PW on an in Cost Saving mode for 6 days, discharged 15.5 kWh and charged 21.4 kWH which is 72.4% efficiency or if the operational power is subtracted (15.5/19.25) it would be 80.5% which is a bummer.

Earlier in this thread there was a mention of the 90% efficiency being for charge/discharge rates in the ~3.3kW range. If this is accurate then why would the control software be configured to start charging at first light when it is 0.2kW instead of waiting until later with more power is available? I've seen a mention that the system "learns", does it really adjust over time?

 

[LesMoss]

I now have completed my first year with Powerwall and my original estimate (90% efficient after taking .5Kw per day overhead) has held up after 6 months of time shifting. Now I am in backup only mode until mid next year.

I don't see evidence that the charge rate has significant impact on efficiency.

Yes, I have seen evidence that the system "learns" what energy it can expect from the solar panels and adjusts its output to maximize the economic benefit.

 

[CrazyRabbit]

New powerwalls (8x), I’m seeing a round trip efficiency of 80%. 20.5 kWh discharge, 25.5 kWh charge. First time using Powerwall in self powered mode, so possibly still balancing the cells.
Powerwalls in garage, temperature between 60 and 70F.

 

[Vines]

New powerwalls (8x), I’m seeing a round trip efficiency of 80%. 20.5 kWh discharge, 25.5 kWh charge. First time using Powerwall in self powered mode, so possibly still balancing the cells.

Also, each Powerwall consumes approximately 250 wh/day to keep itself alive and healthy, and more when it is cold. While this power is not used to discharge to the grid or loads I do not believe it is counted against efficiency in the classic definition. It is used to run the cooling pump, heater as needed, and communication to the gateway.

 

[CrazyRabbit]

I didn’t account for overhead, but no evidence of heater kicking in. Garage average temp is 65F.
Plus I’m not running a laboratory load, balanced resistive load of 3kw per Powerwall.

 

[Redhill_qik]

I didn’t account for overhead, but no evidence of heater kicking in. Garage average temp is 65F.
Plus I’m not running a laboratory load, balanced resistive load of 3kw per Powerwall.

I track my Powerwalls closely and this is basically a constant drain in the system. Some of it is the electronics and thermal management, but some of it is LiON battery leakage loss that would occur if it was just sitting there powered off.

 

[CrazyRabbit]

I track my Powerwalls closely and this is basically a constant drain in the system. Some of it is the electronics and thermal management, but some of it is LiON battery leakage loss that would occur if it was just sitting there powered off.

I see occasional draw from solar, which is changing the cells; and I see an occasional draw from grid which must be a pump and or heater running.
Screen grab from day at 100% backup:

 

[Redhill_qik]

I see occasional draw from solar, which is changing the cells; and I see an occasional draw from grid which must be a pump and or heater running.
Screen grab from day at 100% backup:

That's strange to see the PW charging sporadically from the grid, weird that it isn't reported in the summary unless it is included in Solar number. I never see any PW grid charging in my app and I don't have that option enabled.