Battery capacity

Hi,

Does anyone know if the kWh discharged and charged for the battery on the giv app include conversion losses?

I'm going to run a capacity test because I feel as though I have something weird going on with the total amount that is charged and discharged from the battery.

At midnight yesterday my battery was on 7% so approximately 0.665kwh of capacity including the 4% reserve.

It charged 10.06kwh and discharged 5.84kwh.

0.665 + 10.06 - 5.84 = 4.885kwh I should have left approximately.

My battery was showing as 29% remaining so 2.75kwh. That's a difference or 2.1kwh.

I'd expect it to be out slightly and not to be bang on perfect, but that's quite a difference.

Today, before my charge kicked in my battery was on approx 10% so roughly 0.95kwh of capacity left. It's charged 8.38kwh from the grid plus the 0.95kwh capacity which I had at the start of the charge cycle which gives 9.33kwh. But battery is showing as 100%.

I regularly take my battery down quite low and let it idle below 500w for at least 30 minutes.

One further point to add is that previously the figures have pretty much been bang on and within a few % of what I'd expect.

This issue has started occurring since I've started using predbat and I'm force discharging the battery a lot more to take it down to the reserve.

Qyaunos1 What battery do you have, is it an All in One?

The AIO mis-represents its capacity to GivTCP as being the total capacity battery as 15.9kWh not the available/usable capacity of 13.5kWh. This will throw all the % figures off and you need to set battery_scaling to 0.85 in apps.yaml

Qyaunos1 Does anyone know if the kWh discharged and charged for the battery on the giv app include conversion losses?

There are various conversions that can result in losses:

• AC to DC
• DC step-down to the correct voltage
• electrical to chemical in the battery itself.

I think you have a hybrid inverter, which can store power from solar directly to the battery, so I assume that any measurement of power to battery would be at the DC stage, and so will be after any AC losses. So difference between energy in and out is probably primariy chemical losses ? (I think DC to DC conversion is relatively efficient ?)

I'm going to run a capacity test because I feel as though I have something weird going on with the total amount that is charged and discharged from the battery.
...
My battery was showing as 29% remaining so 2.75kwh. That's a difference or 2.1kwh.

The GE firmware seems to have a great deal of trouble inferring SoC, so possibly best to totals between charges to 100%.

In https://community.givenergy.cloud/d/4019-api-energy-flow-data-is-this-all-measured-on-the-ac-side-of-inverterss I discovered that my inverter doesn't appear to correctly record cumulative energy in and out of the battery, but records an average. Which I find very strange. On my TODO list is a script to fetch each day's total in and out meter data, and then sum those - over a long enough period, the start and end SoC become unimportant.

DD On my TODO list is a script to fetch each day's total in and out meter data, and then sum those - over a long enough period, the start and end SoC become unimportant.

I keep an eye on the Inverter app’s Total Charge and Discharge figures for this. Dividing one by the other gives me an average efficiency since inception last Sep of 85% for my AIO. Sounds about right.

geoffreycoan I have a 9.5 with a gen 3 5kwh hybrid inverter

PianSom Yes, that's what I was expecting, but from that other thread, you'll see that the hybrid inverter seems to have completely messed up the total charge and discharge counters (firmware 3012).

DD my battery is on 3015.

The issue I have is that prior to installing predbat the numbers were pretty much bang on but since installing they've gone out of line and the discrepancy is driving me nuts.

Qyaunos1

DD do you think I'm over thinking it or do you think there might be an underlying issue with my figures?

Qyaunos1 thanks for confirming it’s a 9.5, so battery scaling isn’t the issue.

I’ve given up worrying too much about the individual sensors. Some of the figures in the GE dashboard are just wrong for me, but I found the direct inverter readings (the BBC IOS app) and the values read from GivTCP seemed correct and were at least consistent.

But to contradict that statement of happiness, the battery charge and discharge data in my Home Assistant dashboard is built up from the sensor.givtcp_xxxx_battery_charge_energy_today_kwh & corresponding discharge sensor, wrapped into a utility meter, and all seems fine, but if I look at total charge and discharge over a period, e.g. since start of 2024 it says that total charge and total discharge are within a few kW of each other - so as if there was no battery or inverter conversion losses as at all ?

So, stopped worrying about it

Just ran a manual discharge for around 1.5hours battery capacity decreased by 25%, and the exported figures total 2.37kwh, 25% battery capacity is 2.375kwh so pretty much 100% bang on

Going to see if it charges up again by roughly the same amount

DD Sorry - didn't read your post properly!

PianSom Out of interest, does the web portal show you the total meter readings for charge and discharge? Just wondering if it selectively hides them for hybrid inverters, or for everyone.

DD
The web portal for me shows Charge Total and Discharge Total readings - which are identical. And the number shown does not match either of the numbers from the Inverter app.

The portal numbers do match the app for Import Total and Export Total though

Qyaunos1 There is definitely something odd about the values in the app. I wonder if they did try to be clever and factor in losses.
Looking at my historical data very little in the app seems to match the portal and Home Assistant. Most values appear lower in the app than they should be.

Between Jan 1st and Feb 9th I had a difference (loss) of 1.6% comparing the charge/discharge figures taken from Home Assistant (GivTCP) - values taken at 4% SoC.
Using the figures in the app I got a 16% loss which is an order of magnitude greater than the actual charge/discharge loss.
I would say that it cannot be conversion losses as the average loss for my whole system is only 12% (Gen 1 3.6kW inverter, 8.2kWh battery) but who knows what is being done to the figures.

My advice is ignore the app and get your figures from HA, or failing that the portal (accepting that you might be out by a snapshot).

SJB I've been wondering where in the app these numbers were. I'm guessing we're looking at the energy chart, and then getting totals from there, daily, weekly or monthly?
The other thread mentioned was a comparison of energy data vs meter data, with the (my) suggestion that the meter data is possibly more accurate.
Eg comparing the totals from the app for yesterday with the meter data on the portal view,
app says 5.23kWh charge, 5.71kWh discharge
meter says 5.6kWh charge, 6 kWh discharge.

DD Yes the energy chart. I agree the meter data is probably more accurate,

SJB Thanks for that it does give me some comfort knowing other people are also having similar issues.

Its just really frustrating when you're a numbers person and you want things to balance and they dont.

I do use home assistant but its to monitor daily performance only. Going to have to amend some dashboards and try creating monthly views so that I can see daily as well as monthly.

Qyaunos1 you want things to balance

With this app the numbers balance, but that doesn't mean they are accurate.