Why Generation Report differences

Hi

Install: 10 x 445 AIKO panels SE; 10 x 445 AIKO panels NE; TIGO TS4-A-O on all; GE Gen 3 5KW hybrid; AIO

Prior to installing my solar I used the PVGIS tool to estimate expected solar generation for the year for each panel group and then combined the results.

However, now that I am generating I am seeing differences on the amount generated depending on which source I use.

For example, the TIGO report says for May I generated 1.06MWh; The GE portal from the hybrid says 830 KWh, whilst the portal from the AIO says 808KWh.

For my purposes I am comparing that AIO report vs expected as that is power I get to see and use/export.

I can understand (I think) the difference between the hybrid and AIO views as presumably the AIO is showing the power after the hybrid has converted to AC and so is showing after conversion losses. But I just can't understand why the big difference between what TIGO says and the Hybrid.

Anyone have any ideas?

Know very little about Tigo, but am guessing that their figures may be taken from their optimisers connected directly to the panels. There will inevitably be losses in the cable runs to the Inverter, so that could be a reason why the inverter shows less. This is pure speculation and I have very limited electrical knowledge, so may well be talking rubbish! I'm sure someone will come along shortly with a vast breadth of knowledge to answer your question .

CW For example, the TIGO report says for May I generated 1.06MWh; The GE portal from the hybrid says 830 KWh, whilst the portal from the AIO says 808KWh.

There are always measurement errors but nowadays these should be low percentages eg <5%. The discrepancies are likely to be down to the measurement points. The optimisers measure the raw power coming off each panel probably before the DC to DC conversion that happens in the optimiser. Then you have the inverter from DC to AC which will gives its own view of the power. Then you have the current clamp on the AC output from the inverter which will probably be marked as solar somewhere. Then you have the current clamp on the grid feed which might be marked as import/export. Finally your smart meter does its own measurement. So the 5 measurement places are unlikely to agree due to the losses that happen along the way. In addition the GE does calculations to fill in the gaps that it can't measure directly eg measured solar minus measured grid export minus measured battery charging current must be the house load. When solar, grid and battery values are high the errors give a wild figure if the house load is actually very low. All these things contribute to the different values you see.

1.06MWh x 80% efficiency would be 848KWh which is not far from the GE value so the efficiency of your optimisers and the inverter appears to be around 80% - not unreasonable.

In my own system (17 panels facing SW), for May's solar measurement, my Solaredge optimisers/inverter said 765KWh, my GE portal said 710KWh and my independent Emporia system said 737KWh. I trust the Emporia system the most but the other 2 measurements are within +- 5% so these could just be down to measurement accuracy.

Windy Miller I trust the Emporia system the most but the other 2 measurements are within +- 5% so these could just be down to measurement

Agree with all you say about measurement inaccuracies and no ‘single version of the truth’

I’ve compared my inverter import and export readings from GivTCP to that measured by Octopus at the smart meter and its within a kW over a month. Other people have found much bigger variances and if you look at the standards for smart meters they’re all ‘within tolerance’

I have no personal experience of the Emporia energy meter but I have three Shelly EM1’s on my system. One of them is on the FIT panel output right after to the in-line meter that you have to have. The Shelly is consistently around 7-9% higher than the meter reading on a month to month basis. But for the purposes of calculating generation I have to just treat it as ‘good enough’

Windy Miller Thank you for your reply - I found that interesting, epecially the optimiser part, because if I understand correctly then those figures from TIGO are the actual power from the panel rather than what I get down the line (from an interest point of view I could compare these figures with the predicted solar from PVGIS)

Just checked my May import/export figures
Import from grid GE app - 1.66 kWh
Import from grid Octopus bill - 5.56 kWh

Export from GE app - 636.41 kWh
Export from Octopus - 626.16 kWh

It kinda makes the idea of having two decimal places a bit of a joke.

CW They probably just multiply DC voltage from the panel with DC current from the panel to give you the panel power. I would expect the optimiser to be in the 90-95% efficiency range as it alters the voltage and current to maximise the DC power available to the inverter (along with all the other panels and optimisers called MPPT - Maximum Power Point Tracking). Then the inverter converts to AC with a similar 90-95% efficiency. So you can see that approaching 20% of what the panels create is lost (mostly as heat) before becoming useful energy.

Optimisers are good though as any panel imbalance (caused by dirt or shading) does not affect the whole string of panels with the optimisers always ensuring you get the best from each individual panel regardless of how much light is falling on it.