I see that Mars has changed the post title. Not a problem, Mars, but I would like to confirm that Tesla have advised me that the self-consumption I’m experiencing is completely ‘normal’ and is not a fault with my system.

I’m interested to find out whether this self-consumption is typical of all batteries / inverters (Tesla assure me that it is) and also how widespread knowledge of this self-consumption is. I can find no information on data sheets & installers appear unaware or in denial. Food for thought and discussion ……….

Posted by: @caron

↑

I see that Mars has changed the post title. Not a problem, Mars, but I would like to confirm that Tesla have advised me that the self-consumption I’m experiencing is completely ‘normal’ and is not a fault with my system.

I’m interested to find out whether this self-consumption is typical of all batteries / inverters (Tesla assure me that it is) and also how widespread knowledge of this self-consumption is. I can find no information on data sheets & installers appear unaware or in denial. Food for thought and discussion ……….

My 3yo 9.5kWh GivEnergy battery imports 9.3kWh from empty  and fully discharges 8kWh 

@caron, I'm looking into this for you... from my own personal vantage points, 3-4kWh would seem high, but I don't know much about batteries.

I have done some investigation into this previously, I have 3 data sources - Tesla measurements (instantanious), Tesla App History and also some high level independent monitoring.

Noting that the powerwall 3 datasheet has a reported round-trip efficiency of 89% @ 25C ambient.

If you take my history in the Tesla App battery tab for this month, its reporting 96% efficiency - however in my testing, the "instant" figures within the App and the historical graphs don't align.
For example, the "fast" Charge rate shown at the CT Clamps as "instant" consumption report 5.0kW but compared to the historical data it's only showing 4.8kW.
So based on my independent monitoring i see a round trip of 86% at around 6C ambient.

Battery Discharge was closely aligned in my testing, but it didn't match historical Home Load
Similar to you, i saw "over consumption" reported by the Tesla App history, but not instant.

Over the month of November, Tesla history reported 10% more consumption than actual

In summary, i think the Tesla App is making it look worse in some places (home load) and better in others (charge/discharge) - but i don't think in reality it's as bad as your quoting.

Hope this is helpful towards your own observations

@f1p Might be normal

my numbers 3+ year old powerwall 2, last 12 months:
3.24MWh in, 2.85Mwh out, ~0.88% round efficiency trip, around ~1.07kWh per day

efficiency depends on the rate of charge and discharge, for example 3.3kW charge rate should be the most efficient (based on what Tesla was telling everyone 3 years ago). but now tesla uses 5kW for charging almost all the time.

Running costs is kind of fine, main issue for me with Tesla Powerwall is control, Tesla regularly changes firmware to introduce unpredictable behaviour which will change what you can and can't do and which tariff to use.  

@caron, I heard back from @saenergy (a top battery installer and expert) and here's what he said:

"It’s not uncommon but there’s factors to consider. The inverter and BMS will use power to keep alive. Depending on charge state and reserve, it should take this from solar or battery, sometimes from grid.

RTE is a thing so there’s always less out than in.

The other thing, and the most likely cause, will be thermal effect on the battery packs.

As capacity is, at a very basic level, a function of voltage, as temp changes, so does the resistance of the cells and the voltage moves.

In a battery, when it charges and discharges, it will generate heat. Remember the RTE, well losses come out in the form of heat.

As the electronics attempt to compensate for that, using cell temps, sometimes per cell, sometimes in groups, it kind of loses track.

This is because the system measures kWh in an out, uses that to adjust alongside the volts, temp changes and a really complex charge curve equation.

It’s inevitable as we use electronics to manage a variable chemical reaction. It’s not a truly closed loop signal.

This is why calibration cycles exist. They ignore a lot of the adjustments and equations and just drive the volts right down to minimum, then charge to max volts.

Most batteries do this every now and then automatically. Call it cell balancing or calibration, whatever, it’s doing the same job.

I do this manually on my van for example. As the fully charged range drops off, usually to around 165miles, I’ll run the van to basically dead, then slow charge it to full. Hey presto, range back up to 185miles.

The only way to truly know the in/out energy on the battery is to meter it, log it and compare. Best to do when you can fully charge/discharge it."

@gruff2001 Our Growatt 13 kWh system charges about 11.7 kWh (from 10% to full) and then discharges about 10.4 kWh back to 10%. Given that the HP is on 24 hours, it invariably gets to that 10% and starts using the grid by early evening. I was doing some research as to why and this seems to be fairly normal. Good to see that others with different battery systems observe similar patterns, and not just a Growatt thing.

I suspect a good chunk of the "losses" are power consumption of the inverter itself.

Our Solis inverter consumes somewhere between 50-100w depending on the load, so over a 24hr period that's going to add up.

As the inverter is built into the Powerwall, I suspect this appears as a "hidden" drain on the battery capacity.

I wonder if the loss is similar for one Powerwall and Powerwalls with the extended capacity?

@editor Thank you so much for looking into this for me and to everyone else that’s replied.

i’m sorry I’ve gone quiet since posting. I have a lot to say about this subject but I’ve realised that I’m completely burnt out after spending almost 4 months chasing Tesla and disappearing down a rabbit hole of data collection & analysis.

I need to consolidate my findings into some kind of report/review but, at the moment, my brain has turned to mush so I’m going to concentrate on enjoying real life for the next week or so. 

@caron I can confirm that the power usage of the PW3 is quite high, and I was initially quite shocked at how high. Others have noted similar in discussions around the internet.

The weeks before our July installation, our average daily usage was fairly constant at around 9kWh per day. The weeks immediately after installation, daily consumption came in at the 11-14kWh range, so a definite increase of around 2-3kWh per day was immediately noticeable (and it's a lot more more obvious in summer where our usage is fairly low and stable versus winter where it's all over the place and would probably just go under the radar)

Previously we had a basic solar inverter, which was self powered by the connected solar (it didn't switch on until there was solar), so we never saw any 'consumption' from the grid. The PW3 has an inverter plus other electronics that are powered continuously so it's not unreasonable to attribute maybe 50-100W to that (1.2-2.4kWh/day). Then there's the thermal management of the battery - cooling in summer and heating in winter, which again uses power. That's before we consider any round trip conversion losses.

I'm not sure how/where the Tesla app records various usage/losses. I keep a spreadsheet where daily 'usage' = 'grid import' + 'solar generation' - 'solar export' (i.e, everything in minus everything out). Grid import and solar export come from the smart meter, and solar generation is from the Tesla app as I have no independent solar metering.

I will be interested to read any further details you have managed to discern.

@old_scientist Thanks for your reply. My initial observations were similar to yours. I had solar & PW3 installed in mid-August when my consumption had been low & consistent so it was very apparent that there was an increase of about 30%. If it had been a month later, I would have blamed the heat pump & been frantically trying to work out why it was consuming an extra 3 to 4 kWh per day.

I have now been monitoring my Powerwall 3 for almost 5 months and ascertained that the Tesla App falsely adds battery losses to 'home use' instead of 'battery'. This results in the battery appearing super-efficient and makes impossible to check actual battery performance without additional energy metering equipment. My installer came & fitted a standalone meter after I noticed the increase in home consumption and this has enabled me to see the following:

• battery losses average 3.8 kWh/day
• energy consumption has increased by an average of 18% due to battery losses (ranging from 32% at low use to 14% at high use)
• average battery efficiency is 71% (ranging from 58% at low use to 80% at high use)
• there is a constant parasitic loss of 2.4 kWh in addition to anticipated round trip losses of 11%
• battery losses are applied during battery discharge so the usable capacity is not 13.5 kWh
• actual usable battery is relative to discharge period due to constant parasitic loss 
• usable battery capacity is only 12.2 kWh over 6 hour discharge period & 11.4 kWh over a 15 hour discharge period 

I'm really not sure what to do next. The Powerwall 3 is marketed as 89% efficiency (solar to battery to home/grid), 97.5% efficiency (solar to home/grid), 13.5 kWh usable battery & comprehensive App for control & monitoring. That's why I chose it. My battery is obviously failing on all points but am I completely misunderstanding the data or am I just expecting too much?