Solar PV Efficiency Real World Jump to 26-27% in 2021

what is the current efficiency?

 

I.e. two-junction cells? Vs. the single junction cells that most cost-constrained solar installations have now, as multi-junction products have historically been reserved for uses where highest efficiency is more important than device cost. Getting these products down to the general consumer market will be a big boost.

That depends on just which ones you use for comparison. I need to go back to the old datasheets, but seem to remember the ones on my roof, monocrystalline silicon vintage 2013-15, as 17-ish %.

I was needing some better solar tech for a work product proposal in the 2010-early-2011 timeframe. Oh My! how things have changed since, as several on-the-horizon concepts have come to fruition and made big jumps. Too bad that internal investigation line was one of many to be defunded and cannibalized. But at least the line that took all the available resources, did do quite well.

 

That's impressive! We still get useful power out of some Atlantic Richfield panels I helped my father to install in 1987, on that chart at about 10%.

 

That much is consistent with multi-junction devices. Different layers have different bandgaps to harvest different wavelength ranges.

The indoor energy harvest is what I was exploring, for a specific use. I was trying to collect and store enough energy, while others where slashing the device load and had already vastly increased battery life. When the effort was disbanded / cannibalized, we had narrowed the gap to just 3X, which was really very close compared to where the first battery devices had started. Close enough that we could envision reaching the goal reasonably soon, if management had remained interested. Close enough that natural diffuse skylight coming in next to non-sun windows already did the job, but strictly artificial light in interior offices remained substantially short.

At least the broader energy harvest field continues to greatly expand.

 

A couple of old URLs discussing harvesting E from very long wavelengths:

Researchers boost solar energy by tapping infrared light | Engadget
This Solar Cell Can Capture All Wavelengths of Solar Spectrum - ExtremeTech

Our learned associates may know what has become of such ideas since.

==
As I understand it, some extremely efficient harvesters in post #3 are also extremely expensive. Used in Earth orbit and (donated by manufacturers) to solar car-racing teams.

 

Very old solar-energy converters (plants) are not spectacularly efficient. They 'bounce' green light because they have no trap for it. Earth is a green planet because that's what we waste around here.

Finest 'greenhouses' actually look purple because they have red* and blue* LEDs. Which of course has nothing to do with semiconductor photovoltaics and I expect to be sternly scolded for mentioning it

*food photons

 

I went on a tour of a Korean plant which manufacturers panels for Utilities. They showed us 3 different grades of production, 20%, 22%, and 25%. Small and highly efficient, this manufacturer made the whole process seem simple.

 

Considering the extremely high cost-per-mass-unit of just getting anything into orbit, they prefer to launch only the very highest efficiency PV devices (non-concentrating) in production. Back when I was looking around, that was triple junction devices, i.e. three different photon energy bands, one band of which doesn't work well at ground level due to the attenuating atmosphere above.

At that time, I was seeing roadmaps to 6 junctions, but the chart at #3 shows measured results so far only to 4 junctions. Not that any of this multi-junction tech would be within plausible reach of our mass-market consumer product budget during our careers.

I had missed that recent "up-conversion" path proposal. It could certainly help, by converting two low energy photons (longer wavelengths) into one higher energy photon (shorter wavelength).

Fluorescent light tubes create visible light by "down-conversion", using a phospor material to take a high energy photon (typically UV) and convert it to a lower energy photon (visible light) plus heat. Blue LEDs can also do this to make white light. This is vastly easier than up-conversion, as it can be done singly, not needing to get two photons together.

Current silicon PV materials have a bandgap energy of about 1.1 eV, which corresponds to the near infrared wavelength of 1130 nm. Normal visible light ranges from red at very roughly 740 nm (1.7 eV) to blue at 380 nm (3.2 eV).

Photons with energies below the bandgap cannot be collected by this process, absent that separate up-conversion idea. Photons with higher energies can be collected, but only the bandgap energy can be used, the rest must be thrown away as heat. Thus, greater efficiencies can be achieved by having a ladder of bandgaps at different energies (each in a separate junction layer), so that more of the energy of the various shorter wavelength photons can be kept and delivered to the user, instead of wasted in the capture process.

 

"Fluorescent light tubes create visible light by..." As do white LEDs.

As my scolding for OT has yet to arrive, I'll mention that putting LEDS inside a small thermoelectric (Peltier) cool box generates too much heat for the box to discard. 70 or 80% efficiency is just not enough. Photosynthesis people try to do such things and I (a decomposition person) got roped into that project. We're gon na need a bigger box

(I will describe that project elsewhere when the kinks are resolved. It's actually quite clever)

 

Up conversion of red light energy in plants is pretty nifty, but requires cells with physiology loops that seem not compatible with semiconductor architecture. Makes food though. Can't complain.

All this talk of bandgaps reminds me of what Einstein got Nobel Prize for. His even finer work was beyond what Nobel-givers could understand 110 years later we still cannot, mostly

 

Yup. None of these white LEDs actually create white at the semiconductor junction. They must create blue (or UV now too), then use phosphors to convert to white.

Then there is also the RGB route, three separate junctions or even close-packed devices to create the separate primary colors.

Either way, we were waiting a very long time for someone to figure out how to make the first blue LED. We had all the other primary colors in LEDs for a full generation, but blue was stubbornly missing. Zillions of uses for white LED lights were sitting on hold, waiting in the wings for someone to finally fill in that missing blue element.

Now the white LED market is ginormous.

 

Cell efficiency and panel efficiency rarely go hand in hand, it is the actual construction that often let the panels down. Just because they are built with exceptional quality cells that have a greater than average harvest potential doesn't mean the panel itself will be any better than the quality gear already out there.
I looked at the Panasonic site and their panels are rated up to 85*C, fine for Europe and probably most of the USA, but panel temps in Australia exceed that figure during summer, especially the summer just past.
If we continue to see such high temps (and there is no reason to expect it will cooler as yrs go by) research into wet back panels will be more important than multi junction layered panels for the Australian market, the rest of the world will catch up a few yrs later.
The rate things are going, actual energy harvesting might be the long term aim with high tech heat energy electrical generation working hand in hand with solar voltaic farms. Start adding in heat storage so the heat energy takes over the generation after the sun goes down and the whole "base load power" claims that the various thermal fuel generators use to prop up their industry goes out the window. None of them can compete with the cost of establishing the power generation plant or maintenance costs now, once the base load power requirement is met by renewables it will mark the end of coal and gas fired power station and eventually nuclear power stations.
I doubt I'll still be above ground by the time that forces its way through the resistance to change, but I'm confident it will happen.

T1 Terry

 

In college, a tech rep from Kodak (yes, it's been a while) showed our department a prototype 3" display with green LEDs. They were large, about 1cm across, and were not very bright. He probably told us how they worked, but all I can remember was that they were extremely expensive, and I never saw one again.