low output until all panels in sun

[Gavin A]

Had the guy out today and to be fair to him he was really decent and is keen to get this sorted. He's said we can have a different inverter and has said sma sunny boy and aurora power one both have technology to deal with shading. I think sma has software to track the global peak whereas aurora has 2 mppt.

So the question is which one is best to go for? Any thoughts?

As worcester says, power-one has the same function, but it's just called something like 'full scan', contrary to what Worcester says though, it's actually on as standard on the power-one, but only set to operate every 15 minutes - in high shading situations we'd set it to every 5 minutes - the SMA actually will only go down to every 6 minutes, power-one I think can be set as low as you want, but I'd not advise any less than 5 minutes really.

monitoring is better with SMA, however you can use something like this current cost system for maybe £100 installed to upload the output to a web portal for you which will give you reasonable monitoring, it just won't allow
the installer to interact remotely with the inverter... but unless they know what they're doing with that (as worcester does), then this is a little pointless anyway, and prone to cause as many problems as it solves IME.

Current Cost Dashboard :: device 114779


ps This is the system I screenshotted earlier in the thread (I think) with hard shading coming across the system rapidly from around 5:40pm - it's 12 x 285Wp panels and uses power-one 3.0 Out-D, but specifically doesn't use the dual tracker function as we calculated it was more efficient to operate it single string with the shading function enabled at 5minute intervals, due to the higher efficiency at higher voltages, and the fact the shaded string would rapidly end up below the MPPT voltage as it cut out shaded cells. The performance is working pretty much as expected.

 

[Worcester]

@GavinA, "contrary to what Worcester says though, it's actually on as standard on the power-one, but only set to operate every 15 minutes"

The last one we installed was 'disabled' by default, so definitely worthwile checking it takes less than 30 seconds to do so. - (page 28 of the installtion manual)

Also we tend to leave the scan at 15 minutes, in general the sun doesn't move that quickly across a panel, and each time it scans I believe (GavinA you're welcome to correct me if I'm wrong) that it will lose a little output as it re-scans. And I would hope that they have giving some though to setting at 15 minutes as the default.

 

[Gavin A]

@GavinA, "contrary to what Worcester says though, it's actually on as standard on the power-one, but only set to operate every 15 minutes"

The last one we installed was 'disabled' by default, so definitely worthwile checking it takes less than 30 seconds to do so. - (page 28 of the installtion manual)

hmm could have sworn it was on as standard, wonder if that's changed or I've had a memory glitch.

Also we tend to leave the scan at 15 minutes, in general the sun doesn't move that quickly across a panel, and each time it scans I believe (GavinA you're welcome to correct me if I'm wrong) that it will lose a little output as it re-scans. And I would hope that they have giving some though to setting at 15 minutes as the default.

yeah, think SMA reckon it loses about 0.2% minutes, I assume that's for 15 minutes setting. set it for 5 minutes will lose 0.6%, so it's up to the installer to determine if it will gain more than 0.4% in improved shading performance vs the 15 minute setting - which is mainly down to how fast the shading moves across the array, and how much of the array is impacted for how long.

On the system linked to the shading is from the next door house, so the shading moves fast when it happens, and there were noticeable drops in output between the 15 minute optitrac sweeps, so I think it was worth it in that case.

This is SMA's take on how effective it can be

Also worth pointing out that it only works if the designer has designed the system with the bypass diodes orientated correctly for the shading - no use if the shade is running all the way across the short side of the panels, so the bypass diodes can't do anything about it.

 

[davey_b]

Our installer has ordered an sma sunnyboy 3600tl which I am a bit surprised at since we have 16 x 250w panels giving 4000 watts max output. Wouldn't we be better off with the 4000tl or is there a good reason why we might be better off with a smaller inverter?

 

[SolarCity]

A couple of reasons. For a start, oversized inverters will under perform. Inverters, generally, work more efficiently when they are working at somewhere near their capacity. Therefore, a 3kWp inverter will work better when the panels are producing, say, 2.9kW compared to a 3.6kWp inverter with the same panel production. The other reason is that 3.6kW inverters, generally, don't require prior notification to the DNO as they tend to be rated at less than 16A.

 

[davey_b]

SMA Sunnyboy installed this morning. Was giving about 0.7kw with half the array in shade. Switched on OptiTrac and it went up to 1.7kw. This picture below shows what it's doing in pretty much the same situation as the original post so it seems OptiTrac works. Many thanks guys. Now I just need to work out why my power monitor is showing about 10% less output than the SMA software, there can't be that much loss in the cables. I take a note of the days output each night on the generation meter at the consumer unit so I'll leave it a few days and compare that with the daily output as logged on the inverter and see if there's much difference.

 

[moggy1968]

has you installer split the install into 2 separate strings? it should be split 8 left hand panels, 8 right hand so the shading doesn't drop the output more than necessary. Is there any reason they didn't use the left side of the roof (as we are looking at it) to minimise the shading from the trees?Have you tried using SMAs free download on a Bluetooth equipped computer. Ideally move the 4 right most panels somewhere else. even onto the west roof if you have room there will be better than this but otherwise, although mixing formats, you would be better having 2 landscape and 2 portrait around the solar thermal. As for the rest of the install, it's already been said, it's shabby. looks like it was probably put in by an electrical subby with little/no knowledge of the requirements of solar PV and nont much interest in what they were doing or doing a decent tidy job.

 

[Gavin A]

has you installer split the install into 2 separate strings? it should be split 8 left hand panels, 8 right hand so the shading doesn't drop the output more than necessary. Is there any reason they didn't use the left side of the roof (as we are looking at it) to minimise the shading from the trees?Have you tried using SMAs free download on a Bluetooth equipped computer. Ideally move the 4 right most panels somewhere else. even onto the west roof if you have room there will be better than this but otherwise, although mixing formats, you would be better having 2 landscape and 2 portrait around the solar thermal. As for the rest of the install, it's already been said, it's shabby. looks like it was probably put in by an electrical subby with little/no knowledge of the requirements of solar PV and nont much interest in what they were doing or doing a decent tidy job.

disagree, that would look terrible, the current install looks pretty good on the roof anyway.

Splitting the array left to right through would have been sensible, as would switching the optitrac fucntion on - surprising how few installers even know about it / how to do that. We had a british gas trained guy worked for us briefly with 3 years experience working with SMA inverters who had know clue about that.

 

[SolarCity]

Agree with the OptiTrac comment. I imagine there are some installs where installers are switching them on without considering the shade and analysing whether it would actually help or hinder performance.

 

[moggy1968]

disagree, that would look terrible, the current install looks pretty good on the roof anyway.

Splitting the array left to right through would have been sensible, as would switching the optitrac fucntion on - surprising how few installers even know about it / how to do that. We had a british gas trained guy worked for us briefly with 3 years experience working with SMA inverters who had know clue about that.

it might look good on the roof, but it doesn't work!!yes, it won't look the best (although I disagree with terrible) but at least it will work!

- - - Updated - - -

this is one of those arrays where less really would be more. A 3kW array would have been a better choice than 4kW

 

[Gavin A]

it might look good on the roof, but it doesn't work!!yes, it won't look the best (although I disagree with terrible) but at least it will work!

- - - Updated - - -

this is one of those arrays where less really would be more. A 3kW array would have been a better choice than 4kW

but done right, optitrac effectively switches off the shaded panels anyway, so different rules apply when designing with optitrac. You may be right, hard to know without a detailed analysis. You also have to factor in that if you split the string into 2 on a 12 panels system each string is then right at the bottom of the efficiency range for the inverter voltage wise, which knocks off 2-3% across the entire year, and would disable optitrac as minor shading would take the system out of the MPPT voltage range.

There could even be a case for running it in single string to optimise the efficiency of the inverter, and allow the optitrac to operate to maximum affect while the string voltage is still at the sweet point for the inverter on the unshaded panels, so the unshaded kit would be getting 2-4% more output than the unshaded panels plus optitrac on the split MPPT option.

it's all a bit complex, which is why the MCS shading method is such a nonsense, you can't apply a one size fits all policy to analysing shading - not properly anyway, and it certainly doesn't encourage good design.

 

[moggy1968]

I was assuming a change in inverter. not familiar with optitrac though. My recommendation initially would to have been enphase TBH if a 4kw system was wanted. If the price was too much for the customer then probably 3kW with 2 panels to the left of the thermal system, you might even squeeze 3 in there to give a 3.2kW system