Have I got the wrong Inverter?

[babba]

I'd agree that Power One takes some beating for the price. I think Fronius IG TL range is the best for pure, unshaded yield. SMA are best for features and range. Power One are a good all rounder. The 3.0 model especially. Kostal Piko are a very good alternative.

Ye there are lots out there but some how I just keep comming back to P1. I think they lead, where others follow. They have not revised their inverters in a long while unlike sma, looking forward to the next generation of P1.

 

[babba]

My reading of the data sheets also made me think that the 4000 would be recommended but the Sunny Design software doesn't agree.

If the 4000 is manually entered in then there are no incompatibility issues and the annual estimated output is calculated. I don't know why Sunny Design does not offer the 4000 as a recommendation... but it doesn't.

You cant believe that rubbish...The computer says no...

 

[lewyames]

I definitely did misunderstand you. While it probably won't damage the inverter by connecting too much power to it, I don't think the OP is concerned about this. I think it is the OP concern that he may have an undersized inverter as it will mean that he isn't getting the maximum benefit from his system.

If you check the SMA data sheet the max AC output for a 3000TL is 16A, have a look.....

I don't see how this can be. If the maximum power output of the inverter is 3000w then you would need a grid voltage of around 180v to get 16A.

The maximum current you would expect from the inverter at 230v is 13A - in reality, I'd imagine that the output current is dependant on two things - the DC power and the grid voltage. The output power will never be above 3000w.

Again, I'm always ready to be proven incorrect.[/QUOTE]

Yeah I see where you're coming from! Product specifications can be very misleading, the most common inverter I've fitted is the Fronius IG TL 3.6 which for instance has on its data sheet a max power output of 3680W but they will regularly generate up to 4000W - my assumption is that as you allude to this is down to high grid voltages, but it also leads me to believe that when it comes to product specifications, the most important value is the max output current as this is the real limiting factor, especially for a G83/1 certificated inverter (as per DNO requirements). This is what leads me to believe that the max ac output in Watts can be misleading. SMA are probably the only ones who can really answer this (I've never fitted an SMA 3000TL) if I had an hour to spare to spend on hold I'd give them a call! If you're right and it won't go over 3000W then this guy has definitely had the wrong inverter fitted!

 

[clivep]

I've just downloaded the latest Sunny Design software and there are a couple of changes. The SB 3000TL-20 is still the recommeded inverter (in dark green) but the SB 3600TL-20 and SB 4000TL-20 are now shown as options (in light green).

No other single inverter is shown as compatible and I believe this to be because of yield loss (Min PV voltage for a string of 8 is 198V but inverters need 209V... 1 string of 16 is too high voltage).

Using my data of 16 x Hyundai HIS-S250MG panels, located in Leeds with azimuth of -30 degrees and inclination of 30 degrees gives the following comparison of design suggestions:

SB 3000TL-20 - Annual yield 3348.30 Usability factor 99.5% Performance ratio 83.7%
SB 3800TL-20 - 3327.70 100.0% 83.1%
SB 4000TL-20 - 3356.80 100.0% 83.9%

So the 3800 actually produces less than the 3000 and the 4000 only produces 8.30 kWh more per year than the 3000. Presumably the 3000 is recommended over the 4000 as the extra cost of the 4000 wouldn't be recovered over it's lifespan.

I still can't fathom why the yield of the 3000 and 4000 is so similar. The spec on the 3000 definitely states max AC power of 3kW so I would of thought that in summer there would be a lot more power being wasted.

 

[FB.]

I still can't fathom why the yield of the 3000 and 4000 is so similar. The spec on the 3000 definitely states max AC power of 3kW so I would of thought that in summer there would be a lot more power being wasted.

It's probably because inverters work best when they're working at a medium to high capacity, relative to whath they were built to handle. When there's only a trickle of electricity coming in from the panels (on cloudy days, or when the sun was not in the panels arc, or in winter) the big inverter is well off its "sweet spot" and might only be 80% efficient at converting the DC to AC, while a smaller inverter might be only just off its "sweet spot" and be 90% efficient.
In recent cloudy days my SE-facing 3.75kWp system has only been generating about 3kW total per day, or about 0.6kW per hour at peak late-morning output. My P1 Aurora has a maximum input of 3.75kWp, but with only 0.6kW coming into the inverter (even at the peak in recent days), it was only running at 16% of the maximum that it can handle. Once the sun was out of its arc in the afternoon, it's probably only generating 0.3kW; about 8% of maximum potential.

I extracted this from an ealier posting of mine:

If we had a hypothetical 4kW inverter, it would most likely have efficiency something like this:

0.25kW DC -> 0.2kW AC (80% conversion efficiciency)
0.5kW DC -> 0.45kW AC (90%)
1kW DC -> 0.95kW AC (95%)
2.5kW DC -> 2.4kW AC (96%)
4kW DC -> 3.8kW AC (95%)

In short: the smaller inverter is optimised to deal with smaller DC input, whereas the bigger inverter is optimised to handle higher DC input. To some extent it depends on the exact make/model of inverter as to how well they convert on dull days.

 

[moggy1968]

why SMA produce a 2 string 3.6 or 4kw inverter that can't cope with the voltages of a standard 2 string array is beyond me!!!
Sunny design obviously only recommends SMA inveters and my beleief that the best for a 4kW array is a 3.6 kW inverter from Fronius or Power one. I only use SMA for smaller systems or where I need remote monitoring via bluetooth.

 

[Gavin A]

why SMA produce a 2 string 3.6 or 4kw inverter that can't cope with the voltages of a standard 2 string array is beyond me!!!.

how do you mean?

They cope fine with every 2 string array we've installed on them, although we would drop the start up voltage down to it's lowest level (125V IIRC).

It's true that they're operating below their optimum voltage point when using 30V panels in 2 strings, but then so would the Aurora (not sure about the Fronius).

I'm looking forward to the TL-21 inverters coming our way though as the higher voltage range would give more options.

 

[Gavin A]

re the OP

According to sunny design, there's little to choose between the 3000TL and 4000TL, because the 3000TL will be performing closer to it's optimum efficiency point more of the time than the 4000TL and this largely offsets any limiting of output from a standard 4kWp system.

There is a problem with Sunny Design and sanyo's though as for some reason it doesn't (or at least it didn't) take account of the 1/3 lower reduction in high temperature output of the Sanyo's vs standard panels. In it's calculations it's still working on the basis of 0.45% reduction in output per degree of increased panel operating temperature over the STC temp of 25 degrees that the panels are rated at.

The Sanyo's only have 0.3% reduction in output per degree, so in full sunlight when the panels are something like 50-60 degrees, the output from the Sanyos will be around 4.5% higher than standard panels or in 1000W/m2 of sunlight would be 3494W DC input vs 3321W DC input for standard crystaline panels. The additional 170 odd watts of input power from the sanyo's in full sunlight would essentially be lost via the inverter limiting, which effectively negates the advantage the Sanyo's would usually offer in high temperature performance.

I raised this as an issue with SMA last year, but don't think they've resolved it, which is really annoying given that Sunny Design already has all the data incorporated within it that would be needed to make these calculations for the temperature performance.

 

[SolarCity]

To be fair to SMA, they have pointed out to me in the past that their software shouldn't be relied on for yield calculations and that PV Sol should be used for that instead.

Obviously this mean that, as Gary mentions above, they may not necessarily choose the right inverter for the job.

 

[Jaycee]

Has anyone used the Eversolar range of inverters, they are ip65 as standard, and perform very well from what I can see.

They also have inverter selection software called ever plan which is simple to use.

 

[Gavin A]

To be fair to SMA, they have pointed out to me in the past that their software shouldn't be relied on for yield calculations and that PV Sol should be used for that instead.

true, but the software ought really be able to be used to compare performance between one SMA inverter and another on any given system. That's really most of the point of the software. I agree though about not entirely relying on it, which is how we managed to pick up on it not factoring in the Sanyo's temperature performance advantages, and we'll over rule the softwares recommendations if there's a good reason for it.

Mostly it's pretty good at this, and it's really useful to see the variation in the inverter efficiencies for differents combinations of panels and inverters, which we then use to refine our more detailed performance estimates using PVGIS.

 

[lydart]

Just a quick update. After several phone calls I finally managed to speak to the Sales and Marketing Director who asked me to provide proof of the problems with my invertor. So I sent him some csv files and I finally heard back from him today, but only after I asked him for an update. He emailed me this reply.

“I have spoken toxxxxxx at SMA about this and they state that the 3000TL is an appropriateinverter for the panel combination you have (at the time of install). While thetop end will be “clipped” slightly on a sunny day the overall annualperformance figures provided by the 3000TL is greater than the 4000TL.

4000TL Units wereimpossible to source at the time of your install, 3600TL units were not inproduction so I contacted SMA to discuss at the time and they stated then asthey state now that the 3000TL is an appropriate inverter for your panelcombination.”

This is what’s happening.



Am I just worrying about a few kWh or will this become significant over the next few months?

 

[clivep]

We've just had our best result on Sunday 1st April with slightly over 20kWh produced.

3kW was reached at 11.00 and continued until 13.00. The SB3000TL may have slghtly clipped the output but it didn't seem as if the 4kW Hyundai panels were capable of putting out much more than 3kW. This seems to support their view of the inverter being the correct one.

The chances are that in summer any increase in air temperature will tend to limit the array's capability to generate at any higher power so the amount of clipping will still be small.

 

[loskie]

Our 3.6kwp system(18xKioto 200 panels and Aurora 3.6OUT) peaked today at 4040 watts. This too is the lifetime peak.
The weather was very dull and a gentle rain shower. The sun however was glaringly bright through the haze at 11;30am and I noticed the generation meter flashing rapidly. So went into loft and saw peak readout on inverter.
Generally today production was poor as I said it was/is constantly cloudy. Amazing the power of the sun through the cloud.
Our system was fitted mid October last year.
My calculations being right our Landis/Gyr meter should flash once every second when system is generating 3.6kw. Is that right?
We are in SW Scotland facing sse on about 40deg roof.

 

[flyman]

Loskie,


Not sure on your meter but on my Elster the pulse is for every watt produced so once a second is 3.6kwh.

Our 3.6kwp system(18xKioto 200 panels and Aurora 3.6OUT) peaked today at 4040 watts. This too is the lifetime peak.
The weather was very dull and a gentle rain shower. The sun however was glaringly bright through the haze at 11;30am and I noticed the generation meter flashing rapidly. So went into loft and saw peak readout on inverter.
Generally today production was poor as I said it was/is constantly cloudy. Amazing the power of the sun through the cloud.
Our system was fitted mid October last year.
My calculations being right our Landis/Gyr meter should flash once every second when system is generating 3.6kw. Is that right?
We are in SW Scotland facing sse on about 40deg roof.