Effects of clouds on your array? I'm considering SMA's Secure Power Supply inverters.

[home4]

I am looking at getting a solar PV system and am interested in one of SMA's inverters with their "Secure Power Supply" such as the SMA Sunny Boy 6000TL-US-22. This inverter is grid-tied but during periods when the grid is down, it is able to output up to 1500W to an outlet in order to run certain loads.

SMA calls this "opportunity power" and indicates that, since there are no batteries involved, the power could of course go down at any time, like in cloudy conditions. But I was thinking of how this would work in practice, and since I do not currently have PV panels, I thought someone could help me out.

From my reading, it sounds like even quite cloudy conditions usually only reduce the current from a PV array by about 80%. Can anyone tell me based on their own panel, is this fairly accurate? Are there times when your panel output drops all the way to 0?

Because if one assumes that the panel output would not go below 20% of the sunny-day power, even when overcast, that means that a 6000W array could still produce 6000*0.20 = 1200W, which means that the Secure Power Supply feature could still provide 1000W+ even on a cloudy day, which would be really nice.

But perhaps my assumption is wrong, and cloudy days do drop the panel output lower periodically, perhaps all the way to 0? Or maybe 20% of the sunny output is the average, and it fluctuates between 0 and 30% when cloudy on a second-by-second basis? This is what I'm looking to find out from those with real-world experience. Thanks.

[Ian S]

I am looking at getting a solar PV system and am interested in one of SMA's inverters with their "Secure Power Supply" such as the SMA Sunny Boy 6000TL-US-22. This inverter is grid-tied but during periods when the grid is down, it is able to output up to 1500W to an outlet in order to run certain loads.

SMA calls this "opportunity power" and indicates that, since there are no batteries involved, the power could of course go down at any time, like in cloudy conditions. But I was thinking of how this would work in practice, and since I do not currently have PV panels, I thought someone could help me out.

From my reading, it sounds like even quite cloudy conditions usually only reduce the current from a PV array by about 80%. Can anyone tell me based on their own panel, is this fairly accurate? Are there times when your panel output drops all the way to 0?

Because if one assumes that the panel output would not go below 20% of the sunny-day power, even when overcast, that means that a 6000W array could still produce 6000*0.20 = 1200W, which means that the Secure Power Supply feature could still provide 1000W+ even on a cloudy day, which would be really nice.

But perhaps my assumption is wrong, and cloudy days do drop the panel output lower periodically, perhaps all the way to 0? Or maybe 20% of the sunny output is the average, and it fluctuates between 0 and 30% when cloudy on a second-by-second basis? This is what I'm looking to find out from those with real-world experience. Thanks.

I checked my data since May 2012 here in Phoenix for a 6.9 kW system. Worst production day ever was January 26, 2013. On that day the highest power reading was 0.689 kW and the total production was 2.51 kWh. Typical sunny days around that date had high power readings of 5 kW and daily production of 30 kWh.

[home4]

I checked my data since May 2012 here in Phoenix for a 6.9 kW system. Worst production day ever was January 26, 2013. On that day the highest power reading was 0.689 kW and the total production was 2.51 kWh. Typical sunny days around that date had high power readings of 5 kW and daily production of 30 kWh.

OK, that's interesting, thanks. So it sounds like your highest power reading on that day would be about 14% of a sunny day's output, and your overall kWh that day was only about 8% of a sunny day, so it must have been quite cloudy. Do you know how often your data is logged? Do you see times when the power drops down to 0, or can you not see that specifically?

And I should clarify a little bit -- I'm not hoping that the Secure Power Supply will give me power when it is severely storming or foggy or anything, so not really the worst case necessarily. I'm basically just wondering how it would work on a day that is overcast -- would I still be able to power things in the event of an outage or will this basically only work when it is sunny? Since I would only need the array to produce a fraction of a sunny-day power (like 10-20%), I'm hoping that it would work.

[Ian S]

OK, that's interesting, thanks. So it sounds like your highest power reading on that day would be about 14% of a sunny day's output, and your overall kWh that day was only about 8% of a sunny day, so it must have been quite cloudy. Do you know how often your data is logged? Do you see times when the power drops down to 0, or can you not see that specifically?

And I should clarify a little bit -- I'm not hoping that the Secure Power Supply will give me power when it is severely storming or foggy or anything, so not really the worst case necessarily. I'm basically just wondering how it would work on a day that is overcast -- would I still be able to power things in the event of an outage or will this basically only work when it is sunny? Since I would only need the array to produce a fraction of a sunny-day power (like 10-20%), I'm hoping that it would work.

That data is logged every 5 minutes and there were a couple of those periods where production was less than 0.01 kWh and show as zero. At that time, I was not getting full detailed data on power at 5 minute intervals. Here's another cloudy, probably wet day, November 22, 2013: just under 5 kWh production and max power of 1.46 kW. The 5 minute data is plotted below. No zero power readings that day.

ScreenHunter_02 Sep. 19 12.40.jpg

[inetdog]

As long as the amount that you draw from the SPS is low enough, it should be able to continue producing even on an overcast day.
Based on the SMA documentation, including a webinar for their dealers (Check out SMA University), all that will happen if the demand exceeds what the panels can support is that the SPS will shut down for 5 minutes and then try to restart.

But note that the 1500W rated output of the SPS is a lot lower than the rated GT output/panel input.
Running the SPS to produce a certain average power output over one cycle requires more than that in peak power from the panels. So if your PV input goes below a certain level (greater than 1500) the output power available will be reduced accordingly.
Using Ian's numbers and a guess as the required input to output ratio (not necessarily the efficiency, since the SPS may not be putting a constant load on the PV) of 50%, then at a time when your panels are producing ~500W, the SPS will be able to deliver 250W output.

Notice that the same SPS module with the same maximum available output is used in all of the TL-X000-US series, so with a larger inverter and panel array you will have a lot more headroom on cloudy days before the output power starts to decrease.

[home4]

Thanks guys. It seems like it could be a good product, even with its limitations, as it isn't that much more expensive than any other inverter and there are times when having this output could be useful. I don't know if anyone reading this has one of these inverters at their house, since they are pretty new, but if you do, I would be really interested in hearing how well they work in practice. If I had one I would go out on a cloudy day and flip the AC disconnect and put a known load like a space heater on it and see if it can maintain the load for an hour without disconnecting, and see how big of a load it can support during the cloudy weather. Perhaps someone has tested that already?

[inetdog]

No way would you be able to power a space heater on a cloudy day, IMHO. Although they do make some pretty small (200W) space heaters.

[home4]

No way would you be able to power a space heater on a cloudy day, IMHO. Although they do make some pretty small (200W) space heaters.

I don't really want to power a space heater Just giving an example of a fixed load, like 200W. That's what I'd like to know, would a 6000W array be able to power a 200W load on a cloudy day without interruption? Or would this not be possible?

[inetdog]

I don't really want to power a space heater Just giving an example of a fixed load, like 200W. That's what I'd like to know, would a 6000W array be able to power a 200W load on a cloudy day without interruption? Or would this not be possible?

If the panel output never dropped below 400W, then you could probably run a 200W resistive load off the SPS.
The amount that your array output drops on a bad day depends on your local weather, panel type and orientation, and so much else that I am not willing to make any predictions.

[home4]

If the panel output never dropped below 400W, then you could probably run a 200W resistive load off the SPS.
The amount that your array output drops on a bad day depends on your local weather, panel type and orientation, and so much else that I am not willing to make any predictions.

Thanks

[peakbagger]

There is a lot to be said for finding a UPS that could be powered by the SPS and having that plugged in. That should stabilize the output somewhat on days with passing clouds

[home4]

There is a lot to be said for finding a UPS that could be powered by the SPS and having that plugged in. That should stabilize the output somewhat on days with passing clouds

That's basically what I was thinking of too -- it seems like the cheapest way of having some sort of small "off-grid backup" system, charge the ups with utility power when available, switch to the secure power supply when that goes down, if the UPS gets low and secure power supply isn't enough then get out the generator, that way you don't have to run a generator for small loads. I'm just trying to estimate how much power the secure power supply would actually be able to give.

[home4]

That's basically what I was thinking of too -- it seems like the cheapest way of having some sort of small "off-grid backup" system, charge the ups with utility power when available, switch to the secure power supply when that goes down, if the UPS gets low and secure power supply isn't enough then get out the generator, that way you don't have to run a generator for small loads. I'm just trying to estimate how much power the secure power supply would actually be able to give.

To clarify, I meant to charge the UPS from the secure power supply when the grid is down, unless it couldn't keep up in which case you could fire up a generator.

[Mike90250]

If I recall, the SPS outlet, has to be turned on manually every time it resets.

[control4userguy]

If I recall, the SPS outlet, has to be turned on manually every time it resets.

That's right- switch-over is completely manual on a SMA TL. I like the UPS idea even though it probably wouldn't work for those with exterior GTIs.