Panel voltage matters?

[rdo]

Hi
This is my first post, so HI from Scotland!

I'm about to get a 4kW put on my roof and looking for some opinions from another continent.
I've chosen a reputable installer, who is able to offer a range of quality panels for the same price, what really causes a headache.

The dillema is basically whether to go for the newer technology, gallium-doped 8.33 (aka Pangea, very likely to be from ZNShine). These panels are made of 96 cells and work on a higher nominal voltage (50V rather than usual 30V). Would this mean they would work better in a low-light conditions? Worth mentioning, SunPowers work on a higher voltage as well. The manufacturer also claims that the gallium reduces the Light Induced degradation effect, hence the guaranteed performance is 90% after 25 years. I'm quite unsure whether they're known in the US whatsoever. My alternatives are a better-known REC (Made in Singapur, as for the European market) or another chinese star called Seraphim (top-ten yield scorer in Photon's lab).

Have you got any opinions on this?

[inetdog]

Hi
This is my first post, so HI from Scotland!

I'm about to get a 4kW put on my roof and looking for some opinions from another continent.
I've chosen a reputable installer, who is able to offer a range of quality panels for the same price, what really causes a headache.

The dillema is basically whether to go for the newer technology, gallium-doped 8.33 (aka Pangea, very likely to be from ZNShine). These panels are made of 96 cells and work on a higher nominal voltage (50V rather than usual 30V). Would this mean they would work better in a low-light conditions? Worth mentioning, SunPowers work on a higher voltage as well. The manufacturer also claims that the gallium reduces the Light Induced degradation effect, hence the guaranteed performance is 90% after 25 years. I'm quite unsure whether they're known in the US whatsoever. My alternatives are a better-known REC (Made in Singapur, as for the European market) or another chinese star called Seraphim (top-ten yield scorer in Photon's lab).

Have you got any opinions on this?

The opinions that I am comfortable giving are:
1. For any new formulation of photovoltaic cell, there will be a learning curve as well as unexpected consequences. Being an early adopter rarely pays off for someone who does not have the knowledge to understand fully the new technology and work to make the most of it.
2. The higher panel voltage has no effect whatsoever on low light performance, since over the range of incoming light that produces useful power the voltages (Voc and Vmp) are essentially constant.

[rdo]

Thanks for sharing your opinions.
Ad1 That's what I thought as well. However, These panels are in production for a few good years so far and the techology has some solid foundations in a number of academic research papers that can be found on Google Scholar. Two examples:

http://www.researchgate.net/publication/230601727_Comparison_of_boron-_and_gallium-doped_p-type_Czochralski_silicon_for_photovoltaic_application/file/d912f5107c725ee810.pdf

http://www.sciencedirect.com/science/article/pii/S0927024800000982

Ad2 My initial thought was that it would be easier to induce a voltage that is high enough to start an inverter, because of:
Typical Startup Voltage per panel : 150V/8=19V
19V = is a 60% of a standard panels (Vmp=31V)
19V = is a 40% of a higher-voltage panel (Vmp=50V)

However, I'm not sure if this is the correct way of thinking.
Regardless of that, higher voltage->lower current->lower losses.

[Sunking]

Panel voltage is panel voltage. String or conventional Grid Tied Inverters operate from 250 to 600 volts. Designers select panels based on getting the voltage and power as high as possible in a string with as few panels as possible. It does not make sense or economics to use the highest voltage possible. For example a 70 volt 100 watt panel is not a good choice when you can get a 300 watt 40 volt panels. You can make a 3900 watt string with just 13 panels wired in series using 300 watt panels that yield 520 volts. Amateurs make all kinds of mistake and one of them is going cheap only to fin dout later what a huge mistake they made buying equipment that either performs poorly or not at all because they did not know what they were doing or how to match equipment.

[rdo]

Panel voltage is panel voltage. String or conventional Grid Tied Inverters operate from 250 to 600 volts. Designers select panels based on getting the voltage and power as high as possible in a string with as few panels as possible. It does not make sense or economics to use the highest voltage possible. For example a 70 volt 100 watt panel is not a good choice when you can get a 300 watt 40 volt panels. You can make a 3900 watt string with just 13 panels wired in series using 300 watt panels that yield 520 volts. Amateurs make all kinds of mistake and one of them is going cheap only to fin dout later what a huge mistake they made buying equipment that either performs poorly or not at all because they did not know what they were doing or how to match equipment.

I'm aware of the inverter's operating range and it's maximum efficiency point.
I'm certainly talking about same priced and same wattage panels (265W), but operating at different Vmp (30V and 50V). I'm trying to establish, whether a voltage factor needs to be considered apart from the manufacturer's reputation, etc.

[Sunking]

I'm aware of the inverter's operating range and it's maximum efficiency point.
I'm certainly talking about the same wattage panels (250W) with different Vmp (30V and 50V).

Yeah but how does that play out in a design, layout, and equipment; say a 3000 watt system?

My answer is I would avoid the 50 volt panels like an Eboloa patient.

[rdo]

Yeah but how does that play out in a design, layout, and equipment; say a 3000 watt system?

My answer is I would avoid the 50 volt panels like an Eboloa patient.

My roof would fit 15 panels and I can't exceed 4kW (UK's FIT tariff).

It would be either one string of 15x30V=450V or two strings 7/8x50V=350/400V
The latter is closer to the chosen inverter's max efficiency point of 370V. (emperature coefficiency calculations are within the range).

Why would you avoid it, that's interesting? (These are the gallium ones actually)

[Sunking]

Why would you avoid it, that's interesting? (These are the gallium ones actually)

Because it would take two strings. Two strings would be lower voltage, higher current, combiner/fusing required, or a more expensive inverter that has two MPPT trackers operating at lower voltages. In other words a bad design. What I am driving out from a design POV, if you start by being biased with your mind made up what panel you are going to use, you have put yourself behind the 8 Ball with 1 hand tied behind your back trying to make a bank shot. It forces to make compromises to make it work.

A good designer usually has a lot of experience and software that aid him/her in matching the equipment to play nicely with each other from many manufactures. Look a the inverter manufactures. Let's say you want a 4000 watt system. You first look at the Inverters that can do the job with the features you want, then check the manufacture recommended panels. They already know which panels work best with their product. If you start by selecting a panel, and you eliminate a lot of options. You end up with a square peg trying to find a round hole to hammer it in.

[rdo]

Good points, thanks.

I've already considered inverter (either Samil 4400 or Aurora 3.6) - both have their maximum efficiency at around 370V per string (both can operate two strings).

It's this gallium technology that caused the dillema.

[Sunking]

It's this gallium technology that caused the dillema.

I don't have any skin in the game so I really do not care what you do. Makes me no difference if you succeed or fail.

What I am trying to tel you is be careful especially with some new tech. Nothing really new about Gallium Thin Film Cells and every product on the market has failed and went bankrupt such as Nanosolar and Solyndra. Thin films have come in various disguises like roof shingles made by very reputable manufactures like Corning. None have been successful so far. So the question to you is: Are you willing to place a bet on a product with a ZERO batting average as of to date? Personally not me as I prefer to stick with proven technologies like mono-silicone with a 60 year proven track history. Which one will allow you to sleep at night knowing your investment is safe?

[inetdog]

Thanks for sharing your opinions.

Ad2 My initial thought was that it would be easier to induce a voltage that is high enough to start an inverter, because of:
Typical Startup Voltage per panel : 150V/8=19V
19V = is a 60% of a standard panels (Vmp=31V)
19V = is a 40% of a higher-voltage panel (Vmp=50V)

However, I'm not sure if this is the correct way of thinking.

Well, trying to get to a higher voltage for a grid tie inverter (GTI) with a smaller number of panels could be an argument for the higher voltage panels. But the watts per area will come out in the same general ballpark and if you need to use the higher voltage panels to hit the sweet spot of a GTI, then you probably do not have enough panels and power to make it worthwhile.
And I think that you did not get my point that in low light the panel Vmp will not change more than a percent or two from what it is at full power. So no advantage whatsoever in terms of starting up sooner in low light. Also, even it did work the way you thought, you would still be sandwiched between a high enough starting voltage and a Voc that does not exceed the maximum rating of inverter input, especially in cold weather.

[PVAndy]

Well, trying to get to a higher voltage for a grid tie inverter (GTI) with a smaller number of panels could be an argument for the higher voltage panels. But the watts per area will come out in the same general ballpark and if you need to use the higher voltage panels to hit the sweet spot of a GTI, then you probably do not have enough panels and power to make it worthwhile.
And I think that you did not get my point that in low light the panel Vmp will not change more than a percent or two from what it is at full power. So no advantage whatsoever in terms of starting up sooner in low light. Also, even it did work the way you thought, you would still be sandwiched between a high enough starting voltage and a Voc that does not exceed the maximum rating of inverter input, especially in cold weather.

Personally I'd stick with more conventional panels. Regarding panel voltage, I regularly use SunPower 96 cell panels with a Vmp of 54.7
Regarding inverters I use to use a lot of Power One (now ABB) Aurora Inverters until they had very serious issues with false Arc Fault tripping. Now I use mostly SMA SunnyBoy x000TL-US-22. They start at 150 VDC, run down to 125 VDC and have a very flat MPPT curve allowing them to produce earlier and later in the day than most if not all other inverters.

Andy

[russ]

It's this gallium technology that caused the dillema.

Meaningless but cute.

[rdo]

And I think that you did not get my point that in low light the panel Vmp will not change more than a percent or two from what it is at full power. So no advantage whatsoever in terms of starting up sooner in low light. Also, even it did work the way you thought, you would still be sandwiched between a high enough starting voltage and a Voc that does not exceed the maximum rating of inverter input, especially in cold weather.

That's the explanation I was looking for, as I didn't know if this is the case.
Therefore my conclusion is that higher voltage might be considered as 'better' only if it helps to match inverter's voltage highest efficiency point (370V for Aurora and Samil). Thanks.

Personally I'd stick with more conventional panels. Regarding panel voltage, I regularly use SunPower 96 cell panels with a Vmp of 54.7

This is contradictory, as I consider Vmp=54V as 'high'.
Conventional voltage is 30V, while the panels I mentioned have Vmp=50V (and are also 96 cell).

Regarding inverters I use to use a lot of Power One (now ABB) Aurora Inverters until they had very serious issues with false Arc Fault tripping. Now I use mostly SMA SunnyBoy x000TL-US-22. They start at 150 VDC, run down to 125 VDC and have a very flat MPPT curve allowing them to produce earlier and later in the day than most if not all other inverters.

Andy

I've heard about this recent issue with Aurora.
Lot's of installers in the UK started using Samil nowadays (and claim they're happy with them). It offers 10 years warranty, while a bit cheaper than SMA. Australians say it's a good one, as it's more popular down there.

[russ]

This is contradictory, as I consider Vmp=54V as 'high'.
Conventional voltage is 30V, while the panels I mentioned have Vmp=50V (and are also 96 cell).

0.5 volts per cell - 60 cell panel = 30 volts

no high or low about it.