Panasonic Beats SunPower’s New Solar Module Efficiency Record

65 thoughts on “Panasonic Beats SunPower’s New Solar Module Efficiency Record”

• what’s the problem that makes the efficiency so low? Where to does the rest % go?

  • heat

  • Sunlight hitting the ground exists in a range of frequencies, from ultraviolet to various colors of visible light, down into infrared frequencies. Any given solar cell material can only absorb light and convert it into electrical energy in a certain band of frequencies. Anything above or below that band is lost as heat.
    It’s possible to use several different solar cell materials, stack them together, and build a multijunction solar cell that can absorb light over a larger range of frequencies (and thus be more efficient), but that is more complicated to build and costs a lot more money. It’s usually only used on stuff like satellites where available space and weight are at a premium.

    • Couldn’t have stated it better myself.

    • 100% accurate. It’s mostly about the frequencies.

      The ideal solar panel design would have front layers which convert other frequencies into the frequencies which can be absorbed by silicon solar panels. This is actually something which plants do, but we haven’t figured out how to mimic it.

      There is a second issue: solar panels are best at capturing light which hits them *straight on*. Light coming in at a sharp angle is hard to capture. Efficiency can be improved by, basically, mirrors, to bounce the light in the right direction. This is the principle of “concentrating” solar power, and people have been trying to build cheap surface layers to do this.

      • The standard metric for efficiency is for light at right angles. Unfortunately there is SFIK no standard metric for off-axis performance, which probably means that too little effort is going into improving it. You can always talk in terms of total annual output, but that is site-specific.

        I have suggested using a notional installation at 45 deg latitude, tilt orthogonal at the solstice, and zero cloud, to get a normalised annual derate number, compared to a notional perfect tilt performance for the same module.

  • Thankfully efficiencies below 20% are not a problem, as there is more than enough roof area to produce more power than we need.

    Since there’s about 200 m^2 of roof area per person in a developed country, you are already at 30 kW of PV-module power per person even at only 15% efficiency.

    Average power consumption in Europe is around 7000 kWh per person. So only 7 kW would actually be needed to cover the entire electricity demand (less in sunnier countries).

    • during the day

      • Nights are included in that solar energy calculation.

        Also, Europe has about 200 TWh of hydro storage capacity already.

        With an average demand of 400 GW and 250 GW at night, 200 TWh lasts 800 nighthours or over 60 nights.
        There are no nights nor calm periods which last this long.

    • “…there is more than enough roof area to produce more power than we need.”

      No it’s not enough and it’s not even close. Not when you consider we will need to use electricity to replace all fossil fuels and to cover population increases to year 2100. We are going to need 4-6 times more electricity than we currently generate using all means. Hopefully there will be efficiency breakthroughs, but we can’t bank on that.

      Our electrical future will rely on large-scale energy farms and moving their output along an improved grid as well as rooftop.

      • You don’t need to replace the primary energy leaving the exhaust systems and the cooling towers with renewable energies.
        Also, heat pumps need less than 30% of the energy of a fossil fuel furnace.

        This building in rainy Switzerland produces 4.5 as much energy as it requires (including hot water and heating) with its PV covered surface area alone:
        http://www.ee-news.ch/uploads/articles/images/401852614c88c8b899713c8c430b3ba44ae73781.png

        https://www.solaragentur.ch/images/content/PDF/G-11-09-12%20Solarpreispublikation%20def_Heizplan%20Gams.pdf

        If future populations live in buildings they will need roofs too.

      • You’re forgetting that converting to electric transportation reduces raw energy requirements by ~70%.

  • 58% of the sunlight is converted to heat
    18% is lost to photons being reflected or passing straight through the cell. Some enter the cell and are reflected out again, while others reflect off the surface without entering the cell.
    23.8% is converted into electricity.

    Figures are approximations only.

    • Multijunction (to get additional frequencies) reduces the percentage converted to heat. Concentrators reduce the percentage being reflected or passing straight through the cell.

• One day it might go as high as wind turbines. All the way up to 35% (reportedly) on a very good day. At least they are handy for powering the grow-op. Kidding

  • A few windy states actually average 40% capacity and The London Array did 79% during one month. Wind can vary a lot depending on the season. But solar tends to be more consistent in desert areas so we need both.

    • I think you are confusing capacity factor with efficiency – they are two different things. Efficiency is the maximum amount of a resource that can be captured; capacity factor is the percentage of production energy by a given technology versus what it would have produced if it was running at full output for the complete period.

      For info, there are also onshore wind sites in the UK with annual capacity factors well over 40%, and some offshore sites hit an annual CF of 50%.

      • You’re absolutely right, I was being sloppy.

  • I thought we had demo cells that were over 40% ?

    • Those are multi-junction; they use different layers to absorb different wavelengths.

      • username relevant

        • lol

        • Well it’s important information since single junction cells cannot reach 35% efficiency or more.

    • I think this refers to single junction silicon cells (which have a theoretical max of 28% iirc)

      • That’s due to the bandgap, yes. Single-junction silicon can only capture certain wavelengths. At 28% efficiency they’re capturing all the wavelengths they can.

        This can be beat by various means such as wavelength conversion (a layer which absorbs one wavelength and spits out the desired wavelength) but that hasn’t been commercialized AFAIK.

        • Have you actually see a wavelength conversion layer which has a net benefit? Often while they can prove proof of concept, it ends up blocking more light than it adds in my experience. (Though I wouldn’t be shocked to find it had been fixed…)

        • I haven’t seen a successful industrial wavelength conversion yet, but I haven’t been looking very hard.

          Plants do wavelength conversion successfully (by a very different method than any of the “industrial” proposals).

• Wahoo! Now lets get this into some consumer panels that everyone can actually afford. Until its something that’s obtainable, it is simply numbers.

  All the best,
  Aaron Lephart

  Techvelocity.com

  • I’m sure it will come in time…..

  • Don’t be such a grinch. PV innovation goes through a standard cycle: university lab cell, hand-made, tiny; industrial lab cell, larger, made in small quantities but using production – line methods; industrial lab module; lead fab production module; standard fab production module. Panasonic’s announcement is at step 3. Their fab efficiencies lag by a year or two. The thing to remember is that improvements are still being made at every stage. Companies that don’t keep pace like Yingli get into trouble. The pipeline shows no signs of drying up.

• Actually the efficiency is largely irrelevant for more then 95% of individuals or businesses as usually one can just place more panels to get desired capacity. The most important is price/watt. How much it costs to produce the energy one needs. Higher efficiencies are important because they drive the technology forward – but they do not always translate directly into lower price.

  The biggest breakthrough in adoption of the solar comes from technology that was cutting edge 4 or 5 years ago but it is now getting cheaper and cheaper.

  • 95%? The NREL concludes that more like 30% of US roofs will likely be suitable using PV technology as configured. Considering that conclusion -change, in the form of rising efficiency, is good.

    • Some roofs are easier to put solar on than others, but here in Australia we don’t really have roofs that we can’t put solar on. Of course there are some cases where it is just a dumb idea. If a house is going to be knocked down in six months then it’s probably not a good idea to install solar. And if a roof is heavily shaded the payback time may not be considered worth it by the people involved. And some people’s roofs are just weird. For example, if you have a thatched roof it is going to be a difficult install. But only 30% of roofs are suitable for for current solar? That sounds like crazy talk to me, unless there is something weird about US roofs I am unaware of.

      Maybe the estimate comes from the days when PV was so expensive, only the very best of locations were considered worthwhile.

      • At least in my neighbourhood, probably only 10% of homes get enough light to be worth solar. Old beautiful trees are just so good at providing shade.

        • Nice area. Here even in our most tree lined suburbs most houses receive a lot of direct sunlight on their roofs during the day. Of course in Australia we have the lowest cost rooftop solar in the world, so losing some output to shading doesn’t mean we won’t find a rooftop system worthwhile to install.

        • If we have shade in the US, all parties simply turn away from the property owner while engaging in polite conversation, on the walk back to their vehicles.

        • That’s a shame. My roof has shade from trees in the morning and evening all year round but careful selection of components still gives me an above average kWh/kWp. Granted, the system wasn’t a bargain but monetary payback time isn’t that important tot me.

        • If one is building a PV system in a location with a lot of shade a small and therefore cheaper solar inverter can be used and panels with a total capacity much greater than the inverter can be installed. The drawback with this approach in Australia is we don’t get our Renewable Energy Target subsidy for panels in excess of 133% of inverter capacity so extra panels will effectively cost us more. And one should also make sure the panels and the inverter are good at handling shade.

          But you should also consider if it might be a better idea to instead put a solar system on your grandmother’s sunny roof. And if she’s at home all day working on her indoor hydroponic herb garden like my gran, she is in a better position to use solar energy than someone who is out of the house most of the day.

        • Granny the ganja grower?

        • No, just cannabis.

          I find it useful to take a bagful of cannabis seeds with me when I go fishing. I sprinkle them on the water and when the fish come up to the surface to eat them I net them. With just a hand net of course. I would never use a large net in an inland waterway. That’s illegal.

        • We have a LOT more trees in the US. The Great Northeastern Forest is actually pretty impressive.

      • Does anyone have thatched roofs in Australia?

        • Oddly enough: http://thatch.com.au/

          A real problem is asbestos in older buildings, but this rarely prevents solar installation, as even in the 50s people weren’t fond of filling their roof space with fluffy, loose asbestos.

        • Cool! Our roof is natural slate which caused all installers who quoted to roll their eyes and mutter 🙂 But pretty much every other house in the street have 1920’s asbestos tiles. None have solar panels but I wonder what installers would make of them.

        • In Australia most roofs are either galvanized iron sheeting or open ceramic tiles. The “tin” roofs as we call them are the easiest to install solar on, while ceramic tile roofs require a little more work. But I imagine the asbestos tiles you mention in the UK would be sealed instead of having gaps in them to let the air circulate like a proper Australian roof. And that would make access to the roof beams problematic. And drilling through that risks spreading asbestos around the place. I don’t know what is done in these situations, but at the very least you’d need someone who knows how to deal with the roofing material. In Australia, a lot of installers, particularly “cowboys” who specialize in selling cheap and nasty systems, only know how to do two kinds of roofs – tin and ceramic tile, and won’t be interested in you if you have something different.

        • I’ve emailed an installer here in the UK and asked the question.

        • I’ll be interested to find out what they say.

        • Slate’s a big pain.

          Asbestos tiles would simply need to be removed. You cannot drill into them. Not safe.

        • Bloody Victorians 😉

  • On the the biggest cost in installing solar is the installation cost. So the more efficient the panels are the less you need bringing down this cost (labor, support structure and inventory space). Also when you make a panel more efficient but a little more expensive it generally brings the price/watt down as less material is used (silicon and process cost).

    Now that is for everyone, but for utility scale it also brings down land cost as you get more watt/foot of land. This also makes tracking panel more viable as the added cost gets split over more watts with the same amount of mechanical cost for the tracking mechanisms.

    • Exactly. More efficient modules reduce installation cost with no innovation on the installation front. Even a 1% improvement is north of 4% less panels needed. Repeat a couple times and it’s noticeable

      • Why does 1% improvement in efficiency result in 4% fewer panel needed?

        • Because people are bad at talking about percentages. A one percentage point increase in efficiency raises the efficiency of the most efficient available modules by about 4.5%. So a model that is 23% efficient is 4.5% more efficient than a module that is 22% efficient. Yeah, I know, confusing.

        • Ah, a 1% absolute change (22% to 23%). I think Steve and I were thinking a 1% relative change (22% to 22.22%).

          Thanks for the clarification.

        • Oh, thanks. I’m getting stupid in my old age.

        • People who have been installing solar systems for years trip up over relative and absolute percentages. I wouldn’t worry about it. When I wrote that people are bad at talking about percentages I meant human beings as a whole.

        • Your comment was fine, I understood it was nothing personal. Thanks.

        • And to continue: If your utility investment is just break even, as they are in many places right now, an increase of your total income from electricity production with 4.5% might turn the investment into a very profitable one.

  • Efficiency is not completely irrelevant to final costs of domestic PV systems. In this case, a very large segment of such costs are so-call “soft costs” including, racking, wiring and installation labor. The higher the cell efficiency the smaller the ratio of soft costs per installed watt. You get more bang for your buck, so to speak, and a smaller, cheaper system can be installed that still delivers the same power as a larger, less efficient system.

• Nerd question: these are “heterojunction” cells, suggesting more than one active layer. What is the relevant theoretical limit to efficiency for this type?

  • Well, heterojuction cells should just have a single junction, but it is made from the interface between two different materials. Normal silicon PV is homojunction because the junction is made from the interface between the same sort of material, but it has been doped differently so it functions as a junction. So heterojunction cells are not multijunction cells, but they could be used to make multijunction cells.

• Every increment takes us nearer to the holy grail of self-powered EVs.

  • Nearer to never?

  • No. The theoretical limits to conversion efficiency create an absolute ceiling for the output you can possibly get from a car covered entirely in solar cells. It won’t be enough, assuming fairly conventional design. The Solar Impulse plane has enormous wings and is built out of super- lightweight materials. The all-solar cars entered in competitions are gossamer tricycles.

    • The theoretical limit to conversion efficiency is 100%, obviously. But we’re nowhere close to that technologically.

      Unfortunately, even the theoretical limit of 100% is actually too low to move a sizeable weight in a vehicle with a small surface area. So, batteries.

Comments are closed.