Inverter Output Higher than Panels and Inverter Spec

jimjames

Our 3kWp panels were installed in 2011 and we had a 3000W SMA SunnyBoy inverter fitted as part of that. The output regularly reached 3000W over the summer months but never exceeded that. Our inverter failed in March and has now been replaced with a Growatt 3000TL. I've noticed on a number of occasions the output showing on the Growatt portal is over 3000W, on a few occasions substantially so. I initially thought it might be a measurement or rounding issue but the difference is pretty significant (3500W) that I don't think it can be. Even today when the sun comes out after cloud it's going above 3100W on a 5 min basis.

I can understand that panels can output above their rated amount as that's based on standard temperature etc but it seems odd that an inverter is able to produce so much above the rating. Has anyone else found this with their setup? 

Waywardmike

I see this flagged a lot in the Facebook groups, from all makes of inverter.

I always assumed it was because they had hybrid inverters and the extra from the panels was being fed (DC) into the batteries whereas the amount changed to AC and fed to the house would never go above the inverter rated capacity, in your case 3kW.

Maybe I'm mistaken though as yours isn't a hybrid with batteries right?

*edit - Maybe it's just a better monitoring platform?

I never realised how much clipping I got until I got Solar Assistant and my AC coupled Growatt inverter. 7th April is a great example, inverter tops out at 3.57kW according to AC side, but that's not even visible on the SMA output.

SMA monitoring:


Growatt/Solar Assistant monitoring:

zeupater

Hi

Probably depends on how the data collection/monitoring software is written and how it registers the power .... I'm pretty sure that ours somehow averages the power over a period (seconds), yet registers it and shows it as a peak within the defined interval (minutes) where others may simply register the peak power as the peak seen at some time within the defined interval ....

With the combination of cloud edge effect, low temperatures, direct sun, clear sky & clean panels we've often seen our AC output of ~5%-10% over nominal panel DC rating for a short time on our remote display before the inverter ramps down and/or the 'perfect' combination starts to dissipate, yet this level of power output has never been been reflected in the historical time-slice data despite well over a decade of records ... that's why I suspect that the peak power is an averaged figure within the standard allowable reaction time of an inverter operating according to it's certification required performance levels.

Which is the correct way to do it is open to debate, however, reflecting the maximum power according to how the certification defines the allowed maximum power may have it's advantages if the array/inverter was operating close to it's maximum allowable threshold and questions were to be asked, whereas somehow missing this element within the data could well open up a data-led can or worms, whether warranted or not.

Just my thoughts .... Z      

jimjames

Waywardmike said:

Maybe I'm mistaken though as yours isn't a hybrid with batteries right?

No, mine is just a pure inverter/panels, no battery fitted. It's not something I ever saw with SMA so wonder if it's a poor design from Growatt that doesn't limit output very well but it sounds like others are the same too.

jimjames

zeupater said:

With the combination of cloud edge effect, low temperatures, direct sun, clear sky & clean panels we've often seen our AC output of ~5%-10% over nominal panel DC rating for a short time on our remote display before the inverter ramps down and/or the 'perfect' combination starts to dissipate, yet this level of power output has never been been reflected in the historical time-slice data despite well over a decade of records ... that's why I suspect that the peak power is an averaged figure within the standard allowable reaction time of an inverter operating according to it's certification required performance levels.

I think the surprise for me is that the output is beyond the inverter rating rather than the panels. My understanding was that the inverter was a hard limit hence the clipping mentioned above by @Waywardmike It's not something I've ever seen in 11 years of data from my previous inverter but like you say might be a different measurement method.
Presumably if you have a 4kWp system and use a 3.68kW inverter to meet grid requirements they'd want to know that you're not going to suddenly be exporting a lot more than 4kW because the inverter doesn't cap the output.

70sbudgie

I don't know the details for all the different inverters (and in fact don't know if this would apply to my Solis inverter), but I wonder if it has something to do with the panels outputting DC and the inverter outputting AC. It isn't impossible for there to be discrepancies arising during the DC-AC conversion, due to things like grid synchronisation and the way AC is measured. 

I have a vague memory of my installer giving me a warning that the output from my inverter and that measured by the generation meter wouldn't agree. I can't remember the reason he gave, though.

Edit to add: though the new inverter outputting 3.5kW vs the old inverter outputting 3.0kW sounds more like a difference in specifications / ratings.

jimjames

70sbudgie said:

I have a vague memory of my installer giving me a warning that the output from my inverter and that measured by the generation meter wouldn't agree. I can't remember the reason he gave, though.

Thanks, that's definitely the case. There was a sizeable difference between inverter and generation meter for my old inverter but that was to be expected and it wasn't as a result of the peak values but more down to inverter not being calibrated to the same level as the gen meter.

zeupater

jimjames said:

zeupater said:

With the combination of cloud edge effect, low temperatures, direct sun, clear sky & clean panels we've often seen our AC output of ~5%-10% over nominal panel DC rating for a short time on our remote display before the inverter ramps down and/or the 'perfect' combination starts to dissipate, yet this level of power output has never been been reflected in the historical time-slice data despite well over a decade of records ... that's why I suspect that the peak power is an averaged figure within the standard allowable reaction time of an inverter operating according to it's certification required performance levels.

I think the surprise for me is that the output is beyond the inverter rating rather than the panels. My understanding was that the inverter was a hard limit hence the clipping mentioned above by @Waywardmike It's not something I've ever seen in 11 years of data from my previous inverter but like you say might be a different measurement method.
Presumably if you have a 4kWp system and use a 3.68kW inverter to meet grid requirements they'd want to know that you're not going to suddenly be exporting a lot more than 4kW because the inverter doesn't cap the output.

Hi

The issue is that in most cases the inverter is software controlled and the MPPT and other algorithms effectively perform the requisite logic on a lagging basis as it cannot work on data that hasn't been received yet ... the standard therefore allows for a reaction/correction period to bring inverter performance within specification thresholds and that's what we see, a transitionary condition whilst self-correction is being performed in which a peak only lasts a few seconds and therefore doesn't register in the overall performance statistics.

Effectively, I suppose that the hardware controlled alternative to this would be some form of performance limiting trip circuit (fuse) which would either need manual intervention or software reset after a random period, however, this would likely have detrimental impacts on inverter lifespan as the 'hard stop' frequency increases, which is probably why there's a need for a form of managed self-correction ... after all, it would have been representation of the industry responsible for the investment, maintenance & upkeep of such equipment that were responsible for the equipment operating standards in the first place & it would be against their own interests to add unnecessary cost.

As mentioned, it's more-than-likely that both inverters you mention function well within the requirements of the applicable standards and it's just the way that the historical data is calculated, registered & presented that differs ....

HTH - Z

Niv

I have been given the impression solaredge inverters can give a larger output then their stated spec, it almost used as a feature. I have not seen it myself yet but the year is young.

ABrass

I've seen my6kW Growatt claim to be outputting 6.6kW. No way to validate it though.

Tryinghardtosave

On some occasions my 5.2kwp system is happily doing over 5.4kw for fairly long periods of time.