inversion losses/overheads

Has anyone worked out what the conversion losses are for a GivEnergy inverter?

I have a 3.6kw gen1 inverter & 9.5kw gen2 battery. Ive (back of an envelope) worked out that doing a full charge from 6% to 100% i get ₁₀% in lossess. then when discharging i lose about ₁₀% meaning in total; the conversion losses are 20%

This means that using this system is a waste of time if energy prices are within 20% and its best to pause the battery and run off the grid.

Are my numbers correct, has anyone done a more accurate calculation of the losses? Do GE provide these specs somewhere in thedocumentation.

Not only that, our Gen3 inverter seems to take a base load of about 40W even if it is doing nothing, i.e. no solar and empty battery, so running off grid. That adds about 50% to our base load, something not mentioned in our installers proposal calculations.

Marren2016 Sounds about right.

Marren2016 If you look in the forum there doesn’t seem to be agreement on the losses but most people agree they are in the range of 10-20% overall. There is no free lunch with energy storage and the payback period for batteries can be quite lengthy.

So yes, if you need to be using a time of use tariff that has a decent overnight cheap rate, charge your batteries off that, then run during the day on the cheaper stored electricity.
Lots of approaches for how you automate this.

Do remember that any solar you generate and store in the battery is effectively free.

It is rather difficult to calculate the losses for particular components. The figures vary based on charge/discharge rate.
I have a 3.6 gen1 inverter with 8.2 kWh. The last time I tried to work it out for the battery charging at the maximum rate was a 3% loss, and discharging at my normal rate was 17% so 20% as you say. Discharging at maximum rate was 15% loss. However, any figure on the portal is only accurate to 1 decimal place so looking at a single charge/discharge you could be plus or minus 1% out. Using a larger sample rate could make this more accurate but I doubt it will look much better.
Another way of calculating loss is to do it for the system as a whole:
Input to the system: Smart Meter Import + PV Gen + Battery SoC at the start of the day
Output: Consumption + Smart Meter Export + Battery SoC at the end of the day
The difference between the input and output is the cost of running the system including the conversion losses and power used to run the inverter and battery. Of course Battery SoC is unreliable and may be revised if your battery is idle for a while. This difference is an increase of around 20% on our pre-solar consumption as nophead says there is a large overhead running the system.
As a percentage of the inputs this loss averaged about 12% in December but ranged between 6 and 17%. Difficult to assign a cause - the only time I managed to predict a low loss day when the Battery SoC was revised - the amount of import or PV generation doesn't standout from the figures and I haven't gotten around to more complex statistical analysis.

I presume we should remove the cost of running the inverter (40W someone said above? Or whatever yours is) from the round trip calculation.

Cuz the inverter is going to be on whether or not you're charging it overnight from the grid.

So the round trip losses might not be as bad in reality?

Better to use the inverter rather than leave it idle after taking that into account ?

TX200 No we shouldn't. The inverter losses are proportional to the load/charge rate if running from battery.

I think there are two components to the inverter losses, there is a constant 40W just to run it, which seems very high for an embedded system idling and doing some WIFI comms. Then there is the conversion losses that are probably proportional to the current squared and the battery losses due its internal resistance, again proportional to current squared.

Marren2016 A while ago I did the conversion losses on my Gen1 5kw inverter and 9.5 kw battery. One night when the house was quiet and nothing was running, I did a full 5 min forced charge, and a full 5 min forced discharge, took out the background consumption which was around 150w and it left me with a conversion rate loss of 19.25%. I tried it 3 times and each time the loss was within the same percentage.
The reason for doing it was to see if the Flux rate I was on made it worthwhile downloading at night between 2-5am and letting solar be exported all day or whether it was better to charge the battery using solar and not download 2-5, to export in the high rate 4-7pm.
The upshot was that as the conversion losses made the input price (2-5am) more expensive by 2-3p per kwh than the daytime export price, and I was losing that by downloading at night, during the summer months when I could easily charge fully by solar.

TimHutchings Unfortunately, there's no obvious way to measure round-trip losses of charging battery from solar, rather than just exporting it directly. You get the DC to AC overheads either way, of course, but I wonder what sort of losses you get by first storing the solar in the battery, then discharging it later.

nophead 40W is for the G3, G1 (which @Marren2016 has) takes about 5W when idle - ie running solely from grid.

Any energy that you put in, that isn’t coming out as electricity is coming out as heat. At this time of year this losses are helping keep your central heating bill down! Unfortunately my battery and inverter are outside…

My understanding is that the losses are varied and depend on rate too…

We have
Grid to battery
Battery to house
Battery to grid (should be very similar as battery to house)
PV to battery (higher efficiency I believe?)
PV to grid

Plus grid to battery might be different at 1kW or 3kW.

So many many variables and combinations.

DD Haven't tried that one! However, I would imagine losses will be less PV to battery as both are DC, though export to AC from the battery will obviously be the same losses whichever way you charge it.
As solar charging or export is effectively free electric (except for installation costs) and is variable daily, any conversion losses using solar are not really an issue (unless you're buying a new inverter and comparing which inverter to buy).

GE never published the efficiency curve for PV side (or any side come to think of it)
For Gen 1 one of its cousins did and the sweet spot was around 350V on the PV strings at 50% load got you your max efficency of 97.5% from memory.

TimHutchings I think the PV to battery efficient is really important to know.

If I can import at 5p/kWh it is worth filling my battery even on a really sunny day as I can export excess solar at 15p/kWh.

If the overnight price is 16p/kWh I don’t want to fill my battery.

Somewhere between the two is a price where it crosses over from economical sensible to foolish and without knowing the various different conversion losses it’s hard (impossible) to know exactly where that cut off is.

TimHutchings ps I don’t think any solar as free electricity as I could be exporting it for 15p/kWh

Bizarrely I’ve started to think of any house usage as cheaper in the winter when I’m on agile and a 15p/kWh export rate compared to summer when I’ll likely be back on flux and getting a higher export rate.

naltsta I meant free from the point of view of harvesting electric, as compared to an import price which is supplied by your energy company, whatever price you pay.
As solar energy is so variable each day, the amount of energy received is in the lap of the gods so the efficiency of an inverter from PV is less important and will always be proportionate to the solar collected, and most makes will be similar. You can not control how much the sun shines!
You can however control most other aspects of your solar and battery system. Import, export, timings, quantity, price and usage.

TimHutchings but I can choose to use it to charge my battery, or export it directly for profit. If these things don’t have the same conversion losses then I need to know the differences to be able to calculate what is best

Givenergy said the minimum amount of power my system uses at idle is 40w. I cant see them making a gen3 and increasing its power needs. It would use less. If it does use more thats a stupid design.

Marren2016 Thereby proving that they haven't a clue about their own products.

From the datasheet for Hybrid Gen 1 3.6 - "Consumption Operating (Standby) / Night <5W / <0.5W"