Solar cell hits 24 per cent efficiency

I won't buy solar panels till northern canada decides to move to the equator. It's just not economically feasible.

They're talking about 24% _system_ efficiency - counting in the inverter, storage and what-all. The 24% would mean that the system can convert 24% of the sun's total energy (presumably from the visible + infrared spectrum) that hits it. 

People complaining about the efficiency not being high enough, sure higher is better, but it's not like we don't have plenty of room for panels. The real problem is making them inexpensive enough to put everywhere!

I heard a report that in order to solar power Chicago w/ 40% efficient panels that the entire state of Illinois would need to be covered in panels. So that means ..we need to make our appliances more efficient, we need to optimize our panels, we need to rely on more than just solar to do the job. Unless we start putting these panels in space and beaming down the power there just is not enough surface area on our planet to do the job without killing everything under the panels.

for all the doubters- 80% of the most modern internal combustion engines' energy output is waste heat. Kind of puts things in perspective.

We are driving fossils in more way than one.

The efficiency % is the ratio- 
[Output Electrical Energy]/ [Input Solar Energy]. 

Electrical energy is of way higher quality [usability] than Solar energy. Hence achieving conversion of 20% plus is significant. The costs of achieving a higher efficiency go exponentially higher. So, forget about 70%, it will never happen in an economically viable fashion. 

Solar cells become more viable as the amount of electrical energy it produces in its lifetime goes up, with respect to the energy input in producing the solar cell. If [lifetime Electricity O/P ] :: [Energy input to produce cell] is more than 3, we start getting economic viability. A ratio more than 6 will beat coal- the cheapest energy source.

I hear you loud and clear. Considering the prices we have to pay for sun these days. It's a wonder anyone can afford to keep their current cells running at only 15% efficiency.

Here's the deal: claims of 30-40% efficiency are either some crazy cutting-edge expensive new thing, or more likely, multi-layered photovoltaic (PV) panels in which each layer is designed to absorb a certain range of light wavelengths, thus utilizing a wider range of the spectrum.

Is it possible to get over 100% eff. with a bigger magnifier? No. In fact, I don't believe the 24% figure is based on 24/100% of the sun's power per unit area, because that would give an unfair advantage to any magnifier system.

Bit of semiconductor tech: PV cells work by photons exciting electrons to higher energy levels. Some of this energy is essentially absorbed by the material, silicon; this amount is nearly constant regardless of the amount of incident light. So, it makes sense that with higher radiant flux (brightness) hitting the PV panel, it will be more efficient since the electrical "overhead" power needed is smaller compared to the input power when using a lens or reflector. 

If that (rather simplified) explanation still made no sense: here's the consequence of it: by shining more light on your PV panel, you not only get more input & output power (watts), you ALSO get more efficient conversion of that input power. So arbitrarily, let's say you have an input of 100W and an output of 15W. If you put in 200W, as long as there's proper cooling (!), you won't get out 30W- it'll be more like 35 or 40W.

As for efficiency testing as mentioned above, the light bulb thing won't really cut it- even if you get all the light focused on the panel and know the efficiency of all the reflections of your apparatus, the spectrum of light emitted by the light bulb is different than that of the sun (it -is- very similar in that they're both very close to ideal black-body sources of radiation, but tungsten lamps are lower-temp and thus more reddish, and so the spectrum that the PV panel's designed for will differ from that of the input spectrum, and the measured efficiency value will suffer). Also, the point of this article (although the author didn't bother to mention either this or that he/she refers to 1-layer PV cells) was to say that material advances in conjunction with solar concentrators produce greater efficiency- read what I said above ;) The lightbulb has very low radiant intensity/flux and so in addition to the spectrum being off, the input power being very low would also kill efficiency. So, you really have to operate it in the open for a year or so monitoring radiation levels the whole time, and normalize them to calculated averages... or come up with a very good reference source, in order to measure efficiency accurately. Sorry if this is a bit overly technical or lengthy...

Here's the deal: claims of 30-40% efficiency are either some crazy cutting-edge expensive new thing, or more likely, multi-layered photovoltaic (PV) panels in which each layer is designed to absorb a certain range of light wavelengths, thus utilizing a wider range of the available spectrum-- not far off from the former possibility ; )

Is it possible to get over 100% eff. with a bigger magnifier? No. In fact, I don't believe the 24% figure is based on 24/100% of the sun's power per unit area, because that would give an unfair advantage to any magnifier system.

Bit of semiconductor tech to help explain (if you're not a techie skip this paragraph): PV cells work by having photons (light) excite electrons in the panel to energy levels above the ground (lowest) state; given enough energy, the electrons can actually escape their host atoms and migrate toward one side of the cell whilst "holes"- positive charge carriers which signify the absence of electrons- accumulate on the other side. This means, you guessed it, we've got a potential difference!! When a circuit is connected, the potential diff causes a "push" on the electrons so they're forced to flow through the circuit and to the other side-- this is known as current (sorry if you're very good at this stuff and I'm being rudimentary- some people need a simpler explanation than your egg-head can handle! :P). 

Moving on to points relevant: Some of the electrons' kinetic (valence-state) energy is essentially absorbed by the material, silicon, as the e-'s vibrate; this amount is nearly constant regardless of the amount of incident light so long as it is above a certain minimal level. So, it makes sense that with higher radiant flux (brightness) hitting the PV panel, it will be more efficient because the electrical overhead or "wasted" power increases at a slower and slower rate as the intensity of light is linearly increased. There may be other factors at work here, but this is the primary & most important phenomenon that I'm aware of.

If that (somewhat simplified) explanation still made no sense: here's the consequence of what I just said: by shining more light on your PV panel, not only do you have more input and get more output power (watts), you ALSO get more efficient conversion of that light input. So arbitrarily, let's say you have an input of 100W and an output of 15W. If you put in 200W, as long as there's proper cooling (!), you won't get out 30W- it'll be more like 35 or 40W.

Not only is this good because practical (read:huge)-scale solar power plants, if they ever become economically feasible, will be able to take up less space and generate more power. This means less of an eyesore, less $ for land, fewer cells to maintain, less wire, and last but definitely not least, less semiconductor material can be used. That is probably ****THEMOST IMPORTANTPART OFTHEARTICLE**** 
Semiconductors are expensive to make & process. It takes natural resources and a TON OF POWER, lots of nasty toxic chemicals for processing too. So I think that even if efficiency weren't increased at all with by concentrating the light, it should be done whenever possible since a piece of metal/glass reflector is much much much much cheaper and friendlier to Ma Earth than making more semicon products.

As for efficiency testing as mentioned above, the light bulb thing won't really cut it- even if you get all the light focused on the panel and know the efficiency of all the reflections of your apparatus, the spectrum of light emitted by the light bulb is different than that of the sun (it -is- very similar in that they're both very close to ideal black-body sources of radiation, but tungsten lamps are lower-temp and thus more reddish, and so the spectrum that the PV panel's been designed for will differ from that of the light source, and the measured efficiency value will suffer). Also, the point of this article was to say that material advances in conjunction with solar concentrators produces high efficiency- read what I said above for why ;) (although the author didn't bother to make clear either what the eff. increase means or that he/she refers to 1-layer PV cells) The lightbulb has very low radiant intensity/flux and so in addition to the spectrum being off, the input power being very low would also kill efficiency. So, you really have to operate it in the open for a year or so monitoring radiation levels the whole time, and normalize them to calculated averages... or come up with a very good [expensive] reference, in order to measure efficiency accurately. However, even then it would be difficult because the concentrator has a strong effect on efficiency (increasing it by up to about 25%, I believe) and it wouldn't work properly with the non-parallel rays of a light bulb; also, the reference would be designed for a certain area so you'd have to make a smaller version of the concentrator-cell assembly to fit the standard reference size... and THEN you have to account for that eff. increase due to irradiance rise isn't linear... so best go outside and start testing!! Just don't stare at the sun... yes I know it's pretty, just don't do it anyway hehehe... alright maybe for just a bit : )

Sorry if this is a bit overly technical or lengthy

Sorry for posting twice, but I forgot to mention...

Pascal, if it were possible to make a 70% efficient solar cell, it'd cost so much... probably more than buying four times as many 20% efficiency ones. I actually think a 70% efficient cell would be possible with today's technologies- multi-layered cells bringing it up to the low to mid 40's, and possibly recovering the wasted radiation and re-feeding it back in... see, the cost is already becoming astronomical with all the components that'd be necessary to accomplish that.

By the way, that car which you probably have? The engine is only ~30-40% efficient at most- not including drivetrain and wheel losses, making cars and solar cells approximately equal. Ever wonder why solar powered cars, though they look really weird, can actually work? It's because even though the PV panels aren't great at conversion, they employ efficient electric motors instead of a smog-belching behemoth of a V8 Stag-Smoosher SUV engine that's not only inefficient in an absolute sense but also has way more punch than necessary.

Since this is a japanese project should it then not be watts per yen? for the dollar has an ever declining efficiency of 78% tops currently I hear ;)

Oh and incidentally, how much room a panel takes can mean a lot for the people in cities, especially japanese cities.

You can't have more than 100% efficiency without violating the laws of physics. Even 100% efficiency is probably impossible.

How is the 24% calculated? Could they get 500% efficiency if they just place the panel in the focus of a giant satellite dish?
by : Simen1, 23 October 2007 

The efficency is calculated how much light can be converted to electricity. When that figure reaches 100%, solar cell becames like black hole, you will not be able to see it, as the cell will "eat"(actually convert) all light that reaches it.

@Jax - Your proposed method (power in and out) would include the efficiency of the reference light bulb and it's mirror system. It's not a workable method. Also, the correct way to measure the efficiency is already well defined in the industry.

@Pascal - what if the 70% efficient panel costs a trillion dollars per watt? It appears that you have missed my point. 

Dollars per watt folks. Or watts per dollar.

Even for satellites, they should probably be looking at watts per kilogram (for the whole system). If a panel was 100% efficient, but had a very high mass, it might be a very bad choice. There's actually plenty of room in space, provided that the panel folds up.

I can't even think of any application where photon-electron efficiency would be the first consideration. It would have to be something where space (area) was THE limiting factor. For most solar panel applications, you'll run out of money long before you'll run out of space.


The efficiency measure involving the light bulb reminds me of Homer Simpson buying bacon so he could sell the grease.

nasa made a solar panel that gets more of the color spectrum, compared to only the red of conventional panels. Also solar panels should be able to acomodate 60 percent of the power needs of a typical house hold. What is the point of getting solar panels if youre not willing to do a life style change? you can start with your 1k+ watt power supply that you use to browse the inquirer

http://www.pbs.org/wgbh/nova/solar/program.html

For 99% of the Earth-bound applications, who cares if the solar cell is 24% or 12% efficient. What I want is maximum watts PER DOLLAR (not per square unit area).

Philiosphers have argued about the existence of infinity. Have they bothered to calculate the payback period for solar panels? In many cases the payback period is several infinities.

Let's get the Chinese ramped up and start selling large panels in the One Dollar stores.


Okay, so these panels hit 24% efficiency. What's the typical efficiency for a solar panel? Without that information we can't really judge how much of a progression that is.

I still remember there's higher efficiency cells around the globe, some thing in 30es or 40es I think, but dunno if these are good to use in the wild

One year ago, the US Department of Energy already announced that it achieved a 40% conversion efficiency with solar cells.

Here is the according news:
http://www.energy.gov/news/4503.htm

It's not easy to do a valid financial efficiency comparison, Jeffy. Electrical rates vary from one area to the next. So do amounts of sunlight. And if you're storing to battery, there are quite a few types of batteries, with a wide range of costs. When I bought a 12 volt, 1 amp solar panel in 1979, there was supposed to be a 20 year payback period...but that was eons ago in terms of power rates, so I probably, "broke even" in 10 years. And that was with a lower efficiency panel that Exxon was making at that time. 
I think I read of a much higher efficiency panel being developed with inexpensive materials.

Yes, only Watt/Dollar really matters.
For example, Thin-film Solar Panels can be build relatively cheaply but they only get around 8%. Regular Solar Panels get around 15%.

Since these 24% Cells use optical concentrators (lenses) the higher efficiency per Cell is not surprising but it also means higher production cost. 

I think NASA developed some multi spectrum Solar Cells that get up into the 40's.

It's trivial and does not require anything more than this:

1. How many watts (under the usual standard conditions)?

2. How many dollars (retail)?

Divide one by the other.

Payback period is certainly a "YMMV issue".


How is the 24% calculated? Could they get 500% efficiency if they just place the panel in the focus of a giant satellite dish?

If so, i would call the efficiency claim false or at least not comparable to how i normally think of efficiency.

It shouldn't be that difficult to come up with an unbiased efficiency calculation. Here's one simple idea that should work:

1. Take a reference light bulb
2. Setup mirrors around the bulb to direct the light to the solar panels
3. Power the bulb with a known amount of power (X)
4. Take your solar panel of a reference size
5. Rig up your magnifying lenses
6. Generate power (Y)
7. You lose power Z = X - Y
8. Calculate the efficiency from X and Y.

Power X goes in, Power Y and Z comes out, efficiency is trivial to calculate based on any 2 of those 3 numbers. Sure, Z will be dependent upon how well they setup their mirrors, lenses, etc. But that's fine by me. If they half-ass it, then their claims are worse for their product.

I'll buy solar when a panel gives me a 70% ratio or better.

Simen1... regarding your question "Could they get 500% efficiency if they just place the panel in the focus of a giant satellite dish?"... No. You can't get over 100% efficiency... which is to say that you can't have less than 0% wasted energy. To say something has 500% efficiency is like saying there is 0% waste plus another 400% magically appeared.

I've researched solar panels for my home. 24% is hugely better than anything you can buy now. And while they have made individual cells with efficiencies as high as 40%, the problem is that they are just *individual cells* in a laboratory environment, rather than an array of cells that develop a usable amount of power. 

There is also the space factor. Not outer space. The space on my roof - which is not infinite. A 24% efficient array will give twice the power of a 12% efficient array. Or in other words, I could get the same power from half of the panels. While the panels may be more expensive, you end up saving via lower installation costs and probably less weight on the roof.

So efficiency matters up to a point. 

And at 24% efficiency, it matches the efficiency of a gasoline engine, with the difference being the fuel is widely available, non poluting and costs nothing.


Watts per dollar and watts per surface are very important. But the energy that went into producing the solar cells also is. If it takes 40 years to generate the energy that went into its production, then it's a bad deal no matter the other metrics.