16.32KW System Requires 2 Inverters?

[Southpasfan]

Well, something else to check out. It seems from a super quick google search of SolarEdge inverters that its not much $ to make sure the inverters can easily handle the DC capacity.

I just signed a contract on Friday and it did not specify the inverter specs.

BrettS, what is your configuration again?

 

[wwhitney]

But for a fixed budget, and for continuously variable inverter size and panel size, the optimum allocation of inverters and panels is going to have some clipping if the cost of inverters is non-zero.

So even though panel sizes and inverter sizes are discrete, you should be getting a system designed for clipping. If you don't, then your kWh/year/$ is not as high as it could be.

Cheers ,Wayne

 

[wjgjr]

But for a fixed budget, and for continuously variable inverter size and panel size, the optimum allocation of inverters and panels is going to have some clipping if the cost of inverters is non-zero.

So even though panel sizes and inverter sizes are discrete, you should be getting a system designed for clipping. If you don't, then your kWh/year/$ is not as high as it could be.

Cheers ,Wayne

Even setting aside other goals (like helping the environment, energy independence, etc.) it doesn't always make sense to try and maximize kWh/$ (over any time-frame.) For example, if you assume a situation where adding more generation capacity will make you more money over time, it potentially makes sense to do so even though you cannot add enough capacity to your roof to get to the optimal kWh/$ ratio for the next inverter size up. And, it may be that the next inverter up is actually less expensive per kW capacity compared to the smaller option.

Similarly, there would be times where, due to various utility/net-metering rules it makes sense to downsize somewhat from the optimal generation to inverter ratio because those last kWs of system size would not be profitible.

All of this is to say that while being aware of the discrete nature of inverter sizing is definitely relevant to sizing the system, it is not the only thing to consider, nor is strictly targeting kWh/$ the best approach for a homeowner (perhaps as opposed to an investor.)

 

[wwhitney]

I don't disagree with any of that. My point was that the "base expectation" should be significant clipping (DC/AC ratio maybe around 1.2) not "no clipping".

Cheers, Wayne

 

[BrettS]

Well, something else to check out. It seems from a super quick google search of SolarEdge inverters that its not much $ to make sure the inverters can easily handle the DC capacity.

I just signed a contract on Friday and it did not specify the inverter specs.

BrettS, what is your configuration again?

I have a 15.12kW system with two 7.6kW inverters. I have 48 315W panels and they put 25 panels on one inverter and 23 on the other, so theoretically one inverter might have a little clipping, but it should be very minimal.

 

[Southpasfan]

BrettS,

I think that is a "no clipping" design, becuase the odds of roughly half the panels (7.57) clipping with a 7.6 inverter assigned to them is pretty rare.

My point was that to up one of your 7.6 inverters to a 10.0 is only $500 retail. On my job, which is coming in at $57K, I don't expect any clipping at all, and if its $500 to make sure of that over 25 years I am going to ask for it.

I probably don't understand something technical, but if the cost of inverters with ratings large enough to make sure no power was lost was like a five figure item, I could get it, but not a matter of a few hundred bucks for a larger rated inverter.

 

[BrettS]

BrettS,

I think that is a "no clipping" design, becuase the odds of roughly half the panels (7.57) clipping with a 7.6 inverter assigned to them is pretty rare.

My point was that to up one of your 7.6 inverters to a 10.0 is only $500 retail. On my job, which is coming in at $57K, I don't expect any clipping at all, and if its $500 to make sure of that over 25 years I am going to ask for it.

I probably don't understand something technical, but if the cost of inverters with ratings large enough to make sure no power was lost was like a five figure item, I could get it, but not a matter of a few hundred bucks for a larger rated inverter.

So just for fun I went to PVWatts to try to see how much of a difference clipping could make. I put in a 10kW system at in my zip location and ran the numbers twice. Once with a DC to AC size ratio of 1.0 and once with a DC to AC size ratio of 1.5. I was actually quite surprised at how little of a difference it really made.

PVWatts calculated that the system with a matching DC and AC size (10kW of panels and a 10kW inverter) would produce 14,754kWh annually, and the one that had a DC size of 1.5 times it’s AC capacity (10kW of panels and a 6.7kW inverter) would produce 14,624kWh annually. The difference is only 130kWh per year, which is less than 1%. And this is for a system with all of the panels facing in the same direction. If the panels are facing different directions then clipping is even less likely.

 

[Southpasfan]

Well, maybe I’m off, but it’s seems like saying “you’re getting 48 panels, we know one of them is defective and only has 50% capacity, but that’s only about 2% or something like that”. I would say, “thanks, but I’d like all 48 panels to work at 100%” And that a just an example, we are talking inverters and the cost differential of the next size of inverter is like a couple hundred bucks wholesale.

 

[MP3Mike]

Well, maybe I’m off, but it’s seems like saying “you’re getting 48 panels, we know one of them is defective and only has 50% capacity, but that’s only about 2% or something like that”. I would say, “thanks, but I’d like all 48 panels to work at 100%” And that a just an example, we are talking inverters and the cost differential of the next size of inverter is like a couple hundred bucks wholesale.

If it really only is 130kWh/year then over 25 years that is only a total of $390 worth of energy. (At $0.12/kWh.) So it wouldn't make sense to pay $500 for a larger inverter to avoid that clipping.

 

[BrettS]

Well, maybe I’m off, but it’s seems like saying “you’re getting 48 panels, we know one of them is defective and only has 50% capacity, but that’s only about 2% or something like that”. I would say, “thanks, but I’d like all 48 panels to work at 100%” And that a just an example, we are talking inverters and the cost differential of the next size of inverter is like a couple hundred bucks wholesale.

It’s kind of hard to wrap your head around. It’s not like one panel isn’t working at full capacity or even that your system is working at less capacity. 99% of the time (maybe even more) your system will be working at full capacity and you will get the full output of your panels. But occasionally, at peak production time, if there aren’t clouds in the way, your system may not operate at full capacity for a short time. But again, the “lost” production is minimal and quite possibly not worth the extra $500 you would pay to get it.

 

[DrSmile]

So just for fun I went to PVWatts to try to see how much of a difference clipping could make. I put in a 10kW system at in my zip location and ran the numbers twice. Once with a DC to AC size ratio of 1.0 and once with a DC to AC size ratio of 1.5. I was actually quite surprised at how little of a difference it really made.

PVWatts calculated that the system with a matching DC and AC size (10kW of panels and a 10kW inverter) would produce 14,754kWh annually, and the one that had a DC size of 1.5 times it’s AC capacity (10kW of panels and a 6.7kW inverter) would produce 14,624kWh annually. The difference is only 130kWh per year, which is less than 1%. And this is for a system with all of the panels facing in the same direction. If the panels are facing different directions then clipping is even less likely.

Isn't this showing the exact opposite? It shows how little difference the extra panels make for the time it is NOT clipping, since you make the same power when it's clipping regardless. The correct calculation would be to compare your 10KW with 15KW panels system to a 15KW system with 15KW panels. The difference would be very large.

 

[wwhitney]

For PVWatts, you input the system size in DC, and then input the DC/AC ratio. So the number of panels was fixed, and the inverter was getting smaller and smaller.

In re "spend $500 on a bigger inverter," if you aren't space limited, then spending that $500 on more panels will usually get you more energy production. That's the point.

Cheers, Wayne

 

[BrettS]

Isn't this showing the exact opposite? It shows how little difference the extra panels make for the time it is NOT clipping, since you make the same power when it's clipping regardless.

That doesn’t make sense. I compared a system with a 7.66kW inverter with 10kW panels to a 10kW inverter with 10kW panels. One of those systems will have clipping and one will not. The systems will make the same amount of power when there is no clipping because that will be based on the panel size. However the system with the smaller inverter will make slightly less power when it is clipping. The system with the larger inverter won’t clip. The point is that for most of the time the two systems will produce the same amount of power because even the smaller inverter won’t be clipping during most of the day.

The correct calculation would be to compare your 10KW with 15KW panels system to a 15KW system with 15KW panels. The difference would be very large.

That is essentially what I am comparing, but with slightly different numbers. Instead of a 10kW inverter with 15kW panels compared to a 15kW inverter with 15kW panels. I am comparing a 7.66kW inverter with 10kW panels to a 10kW inverter with 10kW panels. The difference just isn’t as much as one would think because clipping can only happen during a short time during the day, and when it does happen there isn’t a significant amount of power lost.

 

[DrSmile]

That doesn’t make sense. I compared a system with a 7.66kW inverter with 10kW panels to a 10kW inverter with 10kW panels. One of those systems will have clipping and one will not. The systems will make the same amount of power when there is no clipping because that will be based on the panel size. However the system with the smaller inverter will make slightly less power when it is clipping. The system with the larger inverter won’t clip. The point is that for most of the time the two systems will produce the same amount of power because even the smaller inverter won’t be clipping during most of the day.

Once you go much above a 1.3 ratio the system will clip pretty much the entire day. The ramp up gets steeper and steeper as you add more panels. Granted this only happens on sunny days in Spring/Summer months, but that's when you make most of your solar power. And it does assume all South panels, with different roofs overpaneling makes way more sense.

 

[BrettS]

Once you go much above a 1.3 ratio the system will clip pretty much the entire day. The ramp up gets steeper and steeper as you add more panels.

Absolutely not. This is an example of the output from my system on a nearly cloudless day. In my case I have a 15.12kW system and two 7.6kW inverters, so the inverter size is essentially equal to the panel output. Other than that brief spike from the cloud edge effect in the middle the max output was about 11kW. So if my system had a 1.3 ratio then I would have 11.6kW of inverters and the system wouldn’t have clipped at all. Even if I had 9kW of inverters, which would have been nearly a 1.7 ratio it only would have clipped for about 2 hours in the middle of the day.

It’s also worth noting that here is no hard and fast number that applies to all systems. Panel orientation plays a big role in this as well. If all your panels are facing due south then they will peak much higher and you will probably want a lower ratio. If all your panels are facing due north then even during peak times they won’t produce as much and you could get away with a smaller inverter. If your panels face multiple directions then they won’t all peak at the same time, so again, you can use a smaller inverter without clipping.

 

[DrSmile]

Absolutely not. This is an example of the output from my system on a nearly cloudless day. In my case I have a 15.12kW system and two 7.6kW inverters, so the inverter size is essentially equal to the panel output. Other than that brief spike from the cloud edge effect in the middle the max output was about 11kW. So if my system had a 1.3 ratio then I would have 11.6kW of inverters and the system wouldn’t have clipped at all. Even if I had 9kW of inverters, which would have been nearly a 1.7 ratio it only would have clipped for about 2 hours in the middle of the day.

It’s also worth noting that here is no hard and fast number that applies to all systems. Panel orientation plays a big role in this as well. If all your panels are facing due south then they will peak much higher and you will probably want a lower ratio. If all your panels are facing due north then even during peak times they won’t produce as much and you could get away with a smaller inverter. If your panels face multiple directions then they won’t all peak at the same time, so again, you can use a smaller inverter without clipping.

I'm assuming your system falls into the latter category. Also the panel quality obviously matters. I'm overpaneled by about 1.15 and I clip for about 8 months a year on sunny days. Note I'm clipping for 3 hours in the graph below:

The graphs shown on websites for overpaneling are also misleading. They always show one size inverter with more panels and the larger sides in the curve make up for the peak, but in reality the larger inverter system with the same size panels would have BOTH the sides and the peak in the production. Granted the inverter will be slightly less efficient at a lower wattage.

 

[BrettS]

I'm assuming your system falls into the latter category.

Yes, that’s true. I have panels facing predominantly E, W, and S. Although I also have two inverters, so one inverter has all the E and some S panels and the other inverter has all the W and the rest of the S panels.

I'm overpaneled by about 1.15 and I clip for about 8 months a year on sunny days. Note I'm clipping for 3 hours in the graph below:

Even in your case as it does clip it’s not affecting production significantly. It looks like you’re clipping at around 11.25kW. The curve tapers off pretty quickly at the top, so even without clipping I’m not sure it would get much above 12kW, but even if we assume it would have gotten up to 12.5kW at the very peak then you would have ‘lost’ around 1.5kW of production on that day, which is less than 2% of the day’s production. And given that it only happens 8 months of the year, and only on sunny days during those 8 months it’s probably not much more than 1% of your total annual production that’s ‘lost’.

But really my point was that this is something that needs to be considered on a system by system basis. You can’t just take a number and say that any system with a ratio of 1.3 will have unacceptable levels of clipping. On my system 1.3 would be completely unnoticeable and on your system it could definitely be approaching unacceptable.

 

[VslavZman]

I'd just reiterate that it's not a great idea to overpanel too much unless you have odd shading or orientation issues. If your panels are producing 15 KW and you only have a 10KW inverter you'll experience clipping, meaning that the excess energy is lost and you're only producing 10KW. You can't use it in the house or feed it back to the grid. You're "maxing out" when you hit the A/C rating, not the DC. Just because it's rated at 15.5 KW DC doesn't somehow mean you'll make more A/C power than the rating. My 11.4 KW Solaredge inverter delivers below 11 KW in the Summer when the line voltage drops. I'm overpaneled at 13.2 KW DC and wish I had two 6KW inverters instead.

That's really helpful, thanks. I don't really have any shading issues, but the bulk of the panels will be east-facing, with the balance north-facing, so it's definitely not ideal. I suppose I will experience some clipping in the summer-time, but is that offset enough for you by the added production in the winter months? If not, I'll go with full 16KW system and the two inverters.

 

[wwhitney]

PVWatts is going to give you actual answers, rather than just the speculation you are getting so far. So:

What's your zip code?
For each group of panels, what is the DC size, the tilt (roof slope), and the azimuth (compass direction)?

Cheers, Wayne

 

[VslavZman]

PVWatts is going to give you actual answers, rather than just the speculation you are getting so far. So:

What's your zip code?
For each group of panels, what is the DC size, the tilt (roof slope), and the azimuth (compass direction)?

Cheers, Wayne

Thanks - I appreciate the offer to do the analysis as I'm not yet familiar enough with PVWatts to navigate this. Here's the layout:

Zip code: 01951

Group 1: 27 Panels = 9.18KW
Tilt: 20 degrees
Azimuth: 80 degrees
Shading: 93% unshaded yearly average

Group 2: 16 Panels = 5.44KW
Tilt: 20 degrees
Azimuth: 350 degrees
Shading: 93% unshaded yearly average

Group 3: 5 Panels = 1.7KW
Tilt: 20 degrees
Azimuth: 350 degrees
Shading: 95% unshaded yearly average