How to get maximal power output from solar panels in reality

[Mayan Fung]

I learned that solar panels have their maximal power output at a particular voltage and current level. If we want to get the max power, we have to connect a load with a specific resistance. However, in reality, how does the solar plants solve this problem? I guess they won't allow the power output to be dependent on the end-users. And I wonder how solar plants integrate their power generation to the grid?

(from PVEducation)

 

[anorlunda]

That's correct, a simple resistance won't do it. They use active circuits called maximum power point tracking(MPPT) that actually search for the optimum point.

https://en.wikipedia.org/wiki/Maximum_power_point_tracking

Edit: For not much money, you can buy a MPPT controller for home installations.

 

[Mayan Fung]

That's correct, a simple resistance won't do it. They use active circuits called maximum power point tracking(MPPT) that actually search for the optimum point.

https://en.wikipedia.org/wiki/Maximum_power_point_tracking

Edit: For not much money, you can buy a MPPT controller for home installations.

Thanks! This questions had been puzzling me for a few years since my undergraduate study.

 

[Windadct]

The point and need for MPPT is due to the variability of the curves you have shown - Light Intensity, temperature, and panel degradation - factors can have significant impacts on the curve, and we are looking to maximize POWER transfer ( harvest) from the system.

So, yes - thinkinging terms of resistance over simplifies the issue, especially since the "load" in a grid connected system is also very complex.

So this controller applies some type of a power converter circuit converter that can typically control the output voltage (buck/boost/inverter), which will drive the current into the load, basically Optimizing Ipv x Vpv (@Mpp-Max Power Point) = By controlling Vout * I out.

 

[Tom.G]

And of course the other two almost as important items: Keep them clean, and track the Sun position.

 

[sophiecentaur]

And of course the other two almost as important items: Keep them clean,

Well - yes and that can be very hard to achieve with some installations. 'Wasting' a continuous couple of hundred Watts because of bird poo is very bad value when your situation is marginal.

Something that I would like to know ( I guess I should be able to do the sums but you know the problem) and that is how much real advantage can be gained from a MPPT, in terms of Wh over the day. The VI slope (post #1) is not as relevant as a VI vs incident light flux. The attached Wiki diagram may be a bit more indicative of the true situation in low illumination, I think. It looks as if the steep black curve represents what you actually get as the illuminance varies. Have I got this wrong?

 

[anorlunda]

how much real advantage can be gained from a MPPT, in terms of Wh over the day.

Gained compared to what? The curve you posted is correct, including that wiggly black line. Without MPPT, the actual operating point will be someplace else away from the black curve but still on the VI curve for the actual light flux.

The limiting cases are of course zero V (the vertical axis) or zero I (the horizontal axis). Those produce zero WH in the day.

If you consider a fixed resistance load R, then just draw a straight line extending from the origin sloping up. Where that line intersects the curve for the actual light flux is the operating point.

A constant power load gives a hyperbola on that plot. Where the hyperbola intersects is your operating point.

In both cases, the power at those operating points will be less than the MPPT point.

Another way to view it is that the MPPT controller is an active optimizer. For any value of light flux, it experimentally moves you up and down the curve to find the point delivering maximum watts. That really doesn't depend on the shape of those characteristic curves, or the underlying physics or the variability in light flux. Some but not all, MPPT controllers do indeed use the hill climbing optimization algorithm, so in theory they could optimize devices that are not solar panels.

 

[sophiecentaur]

That really doesn't depend on the shape of those characteristic curves, or the underlying physics or the variability in light flux.

I understand your post but it's the variation of illuminance that you have to apply to that curve (important practical detail) in order to work out how much more Ah you get during a 'typical' day / year, compared with just using a normal regulator. I wouldn't have though that cannot be estimated, once the statistics of illuminance are available for a particular site and installation. I'd make do with a 'typical' example. Thing is, if you're always operating under dim conditions then MPPT could give significant advantage but it does nothing much for you in bright or long enough sunlight. I was after getting a feel for that. Plenty of installations use conventional regulators but they do the job.

I realize that the cost of MPPT technology may be low enough to justify its use anywhere and that's probably the best answer to the 'is it worth it?' question. Sometimes, a bit of smart tek really does it for you - like the joule thief allows LED garden lights to operate with small PV units which are struggling to get to the wanted battery charging volts.

 

[anorlunda]

Let's say you mean constant V when you say normal regulator. The V set point can be adjusted to be optimum at noon on bright days. But you want the panels to generate throughout the day. The noontime setting can be very non-optimum when the sun is low in the sky in morning and afternoon.

But yes, it is also true that the cost of MPPT is low compared to the cost of "smart" 3 stage battery charge controllers.

Friends who had solar on their boats and later upgraded to a MPPT controller, report that their batteries are fully recharged by 10:30AM, rather than 2PM with non-MPPT controllers. One friend even added a 2nd refrigerator to his boat without changing the panels because of MPPT.

But to really get the maximum, you need to use MPPT, plus 2-axis tracking.

I once posted the following from a friend's installation (on land at 45 degrees north). Note how "square" his power-versus-time curve (green) looks compared to the typical sine curve for a "normal" fixed-orientation installation. We can even see a difference between morning and afternoon. That is because trees block more of his western sky than eastern sky.

 

[sophiecentaur]

Note how "square" his power-versus-time curve (green) looks

Very impressive. It seems to indicate that you can almost double up on what you would get without any help. How much of that would be due to panel tracking and how much to MPPT treatment, I wonder. MPPT seems to sell itself on cost / benefit but panel tracking always sounds costly, compared with just doubling up on panel area. (On a boat etc. there may well be other constraints) - plus boats tend to move themselves about twice a day, which may be just what you want or just what you don't want - depending on what the tide does to your direction on your mooring. Where to put a big PV area is always a problem on a boat.

 

[anorlunda]

I wonder. MPPT seems to sell itself on cost / benefit but panel tracking always sounds costly,

Yup, exactly right. You need to do a careful cost benefit analysis on the whole system, including MPPT, 1-axis or 2-axis tracking, and battery storage. But the solar PV part roughly halves in cost every 3 years, tipping the balance.

I also expect the cost-benefit result to be highly sensitive to latitude and local weather patterns.

When I lived at 65 degrees north latitude, the sun just circled us in summer. A 1 axis tracking system up there should be much easier to justify. It was also grey and drizzly 200 days per year.

In the tropics, a fixed nearly-horizontal installation is probably optimum.

My friend does need to shovel snow off the panels in winter, and if they get ice coated, he just has to wait for it to melt. If he was trying to live off-grid, that would be highly significant.

 

[Mayan Fung]

Let's say you mean constant V when you say normal regulator. The V set point can be adjusted to be optimum at noon on bright days. But you want the panels to generate throughout the day. The noontime setting can be very non-optimum when the sun is low in the sky in morning and afternoon.

But yes, it is also true that the cost of MPPT is low compared to the cost of "smart" 3 stage battery charge controllers.

Friends who had solar on their boats and later upgraded to a MPPT controller, report that their batteries are fully recharged by 10:30AM, rather than 2PM with non-MPPT controllers. One friend even added a 2nd refrigerator to his boat without changing the panels because of MPPT.

But to really get the maximum, you need to use MPPT, plus 2-axis tracking.

I once posted the following from a friend's installation (on land at 45 degrees north). Note how "square" his power-versus-time curve (green) looks compared to the typical sine curve for a "normal" fixed-orientation installation. We can even see a difference between morning and afternoon. That is because trees block more of his western sky than eastern sky.

View attachment 267965

Really impressive to see a real-life small-scale example. Really wish that we can switch to renewables before we mess things up.