Solar installed then consumption apparently increased by ~60kw per month

petalouthi

Thanks everyone. I will go through all the comments and suggestions in the next day or so and reply in detail when I've had time to understand all the comments properly. Thanks.

wrf12345

Out of interest, are you better off financially on the bills despite the increased usage, if you are getting lots of solar/battery export money then the cost of running them does not really matter?

petalouthi

QrizB said:

Current clamps aren't particularly accurate.

You should have a generation meter as well as your import/export smart meter. How does the generation meter compare to the generation totals in the app?

Ah yes - the Generation meter, and also another extra meter labelled "Energy meter" and also referred to as CHINTS meter by QCells. Both meters started at zero on installation.

We thought this might be a good independant means of looking at generation. But we can't make sense of readings on these meters, they seem all over the place. Some examples:

from installation (22/11/23) to 07/09/24

QCells app total generation=3863kw

Gen. meter import reading=3718

Gen. meter export reading=546

CHINT import reading=26.86

CHINT export reading=59.63

from installation (22/11/23) to 07/01/25

QCells app total generation=4735kw

Gen. meter import reading=4919

Gen. meter export reading=1168

CHINT import reading=42.57

CHINT export reading=63.30

from 07/10/24 to 07/11/24 - one month

QCells app generation for month=193kw

Gen. meter import reading increased by 272

Gen. meter export reading increased by 152

CHINT import reading increased by 4.28

CHINT export reading increased by 1.77

from 07/11/24 to 07/12/24 - one month

QCells app generation for month=174kw

Gen. meter import reading increased by 277

Gen. meter export reading increased by 173

CHINT import reading increased by 4.87

CHINT export reading increased by 1.71

We've asked for more info about what these meters are showing, initially from installer, and then, when we realised that they were referring the questions on to QCells, then just punting answers back to us, we had a lengthy correspondence direct with QCells. Various responses, sometimes asserting that Gen meter measured generation, others saying they didn't measure anything useful! We've also read elsewhere that Gen and CHINTS meters can measure different things according to how they were installed, and that they are a legacy from FIT tariff days and irrelevant to more recent installations. This is one reason for turning to MSE forum in desperation!

petalouthi

tim_p said:

Has your annual consumption actually increased by the 720kWh?

Consumption measured from meter 07/12/21-07/12/22=2598

Consumption measured from meter 07/12/22-07/12/23=2580(includes 1st few days after installation on 22/11/23, when measured from app)

Consumption from QCells app 07/12/23-07/12/24=3255

So it's increased by 675 from '23, and 657 from '22 figures

petalouthi

matt_drummer said:

Battery charging and discharging losses probably account for a lot of it.

If you are charging from the grid at night and then using batteries when the sun isn't shining you will consume more electricity than if you just took it from the grid as you used it.

We wondered if this might be the case. It's set to charge from grid at night during winter, but not during summer. The extra consumption continued throughout the year. But perhaps the inverter is also working hard in the summer exporting lots??

petalouthi

wrf12345 said:

You need to look at daily consumption, see if turning everything off except for solar results in the high usage, 2kw a day is roughly 80W an hour, which could be that inverter and associated electronics permanently on. I had relatively simple solar/inverter system more than ten years ago and never had any kind of excess consumption even in winter but with more powerful modern systems and their constant monitoring and battery electronics who knows?

---------------------------------------------------

We did do this. We turned off everything except router (15w/hr according to BT) and the solar installation.

The QCells app told us we were consuming 50w. So 35w unaccounted for. It's possible we did miss some other item when turning off, but looked very carefully and couldn't find anything. It was in Sept. and after dark, so that there was no generation to possibly muddy the waters. The inverter wasn't set to import from grid at the time, and power for the house was coming entirely from the battery which was gradually discharging.

35w/hr works out at 25.2kw/month, so wouldn't account for the whole increase in consumption if this was a consistent power usage.

QCells told us that the inverter should use 10w/hr.

petalouthi

jimjames said:

What does your actual electricity meter show? If that isn't increased by 60kWh a month then it's the app giving erroneous data

All the readings are from the meter itself. But I check each month and the app always matches the meter readings well for export and import totals.

petalouthi

QrizB said:

jimjames said:

What does your actual electricity meter show? If that isn't increased by 60kWh a month then it's the app giving erroneous data

So I read it as "I know my import and export, but my app thinks my household's consumption has increased by 60kWh a month".

Reading the import meter won't directly help with that.

However, assuming OP has a generation meter, they can check:

Consumption = Import - Export + Generation.

See above for problem with Gen meter. I can get import and export figures from Smartmeter. I can get a generation figure from the app. I can get a consumption figure from the app, but I think it's calculated using the formula you state rather than measured directly. If I do the calculation myself I can see that the answer matches what the app says for consumption.

petalouthi

Scot_39 said:

Anything you feed into a battery - you will lose some on the way out. The chemical process isn't 100% efficient electrically.  It often generates quite a lot of heat for instance.

Maybe as much as 10% recovery loss for lead acid - still used in small ups today - bigger battery packs ? - maybe sub 5% for li ion.

The invertor electronics maybe another few percent in both convertor ac to dc and invertor dc to ac modes.

One q-cell setup quotes 96% conversion efficiency battery to grid.  

So likely losses plus simply powering the electronics 24/ 7 could eat up your 2 kWh per day.

If charge 7-8kWh a day losing even 10-15% is 0.8-1.05 kWh.  

Tge other 1.0 - 1.2kWh a day averages just 40-50W continuous load - does the electronics loading get built into the  96% type conversion or is there a "constant" vampire type load too.

It's a li ion battery. This is just the kind of efficiency loss we were wondering about, so thanks for useful information.

petalouthi

Qyburn said:

QrizB said:

Current clamps aren't particularly accurate.

You should have a generation meter as well as your import/export smart meter. How does the generation meter compare to the generation totals in the app?

That's what we find with Solaredge. The app overestimates generation, compared to the generation meter. And underestimates both import and export compared to the smart meter. Taken together that gives the appearance of consuming more than we actually do on days with net export.

As noted by others the house load and consumption is not directly measured. Looking at the inverter via Modbus it doesn't appear that solar generation is directly measured either, the values for both AC and DC power include both solar and battery power. 

Thanks - useful to get other people's experience. As mentioned in one of my other responses above, I'm fairly confident that import and export to grid is being measured accurately by the app, as correspond well with smartmeter readings. But we don't have an independant means of checking generation measurement on app - see above for problem with gen. meter.

There are 12 panels each with theoretical max generation of 405w, so at 100% efficiency that would be 4.860kw. As an example of high generation, on a sunny day (29/06/24 12.00-13.00) app says the system generated 4.2kw. This seems very reasonable. The panels are very well sited with no shading.