Slashdot Mirror
New Solar Panel Design Traps More Light
GoSun wrote in with an article about new solar panels that opens, "Sunlight has never really caught fire as a power source, mostly because generating electricity with solar cells is more expensive and less efficient than some conventional sources.
But a new solar panel unveiled this month by the Georgia Tech Research Institute hopes to brighten the future of the energy source." The new panels are able to produce sixty times the current of traditional models.
Do you have any references to back up that claim with?
The fact that you closed with an ad hominem barb leaves me doubtful. More referenced research and less willfully ignorant babble please.
Energy pay-back time and CO2 emissions of PV systems
"energy pay-back time was found to be 25-3 years for present-day roof-top installations and 3-4 years for multi-megawatt, ground-mounted systems. [...] This leads to the conclusion that in the longer term grid-connected PV systems can contribute significantly to the mitigation of CO2 emissions."
(found by typing 'photovoltaic payback time' into google)
But it doesn't matter to me that some hydro-electric plant far from my house is making power at $0.02 per kWh, what matters to my economic reality is that my local power company charges just over $0.08 for the first dozen kWh delivered each day and then has a sliding scale that goes up to $0.36 kWh for increased amounts of power.
Before I installed solar panels a high percentage of my power was costing me that top rate. So the relevent economic calculation for me is the cost to install my panels divided by the expected number of kWh that they will generate across their lifetime. This number comes out at about $0.16 per kWh. So I'm better than breaking even now, and assuming that energy prices continue to rise, I'll do even better in years to come.
The final kicker in the equation is that I've switched to a time-of-use tariff so across the summer the power company will credit me with $0.209 for excess power that I generate in peak hours (between 1pm and 7pm), and $0.112 for partial-peak (10am-1pm + 7pm-9pm).
If I'd taken the capital that I used to install the panels and invested it instead, I'd have to maintain a >19% annual pre-tax rate of return to beat the panels. Possible, but extremely unlikely (especially with my stock-picking track record!).
Maybe a power engineer can answer this... the obvious way to build a solar power plant is to take a whole slew of lenses and focus them on a water tank, and then turn a turbine. Given that heat -> power is a fairly mature technology, wouldn't that be more efficient than solar cells?
Sometimes it's best to just let stupid people be stupid.
Actually, even in the Airzona desert, which receives on average 250 watts/m^2 daily, more than most places on Earth, the watts per square foot is only about 19. If the best panels on the market produced 13 watts per square foot, that would be quite outstanding.
Your maths are wrong somewhere.
When Mark Pinto of Applied Materials spoke at Stanford in EE380 two weeks ago, he said that the current energy payback time on their solar panels is two years, and they're trying to get that down to six months. Some of the fab steps borrowed from semiconductor processing, where the areas aren't so large, can be improved.
Hey, if you have solar panels on your roof, how often to you have to wash them? Do they develop a film that reduces their efficiency?
Computers are useless. They can only give you answers.
-- Pablo Picasso
Recently, I was having a conversation about the upper limit on solar power. I hadn't done the math then, but I just trotted out a fresh napkin to satisfy my curiosity. The earth is 12756 km in diameter. That presents a 127.8 million km^2 cross section to the sun. With the napkin-math estimate of 1kW/m^2 incident at the earth's surface, there's an upper limit of 127.8 million MW of power available from the sun. Okay, so that's an absolute ceiling for terrestrial solar collection - you can't collect more energy than is incident in the first place.
Okay, now for a more practical limit. Let's put the solar collection grid on land - that's a reduction to 30%. Let's also go with solar cells that are 20% efficient - that's not too shabby, but not bleeding-edge-expensive either. (127.8 * 0.3 * 0.2) = 7.67 million MW.
Finally, how much of the available global land mass are we willing to pave over with solar cells? If I use a residential rooftop model, a 1500 sq.ft. house on a 1/4 acre (~10000 sq.ft., sorry for the non-metric-unit shift) property would be about 15%. I think that's probably a bit high, considering that houses aren't aligned for optimal solar collection, but I'm looking for the practical upper limit of solar collection opportunity. Using 15%, the available harvestable power limit becomes 1.15 million MW.
Let's compare that to current consumption stats in the US (no pun intended.) If I read this chart correctly, December of 2006 had 335.6 million MWh of power generated across all industries. There were 744 hours in December, so that equates to 451 thousand MW average continuous power generation. So the maximum solar harvest potential is only about 3x our current consumption rate? Damn, that's sobering.
Actually, boosting the current is just the wrong way to go since they are having trouble with resistance. So, they do want to get the voltage up (not churned out) to help reduce the Ohmic losses (I^2R). With detectors, you usually put on a bias to help get the defects that are causing the resistance filled up, but for power generations you need to rely on the dopant gradiant alone which is probably pretty ragged after they fabricate their nano-posts.s -selling-solar.html
--
Eat the reflectance and get it now: http://mdsolar.blogspot.com/2007/01/slashdot-user
TFA says they increase the surface area without increasing the dimensions of the panel. But that's not enough.
/. That would double the efficiency of both panels, without the drawback of using nanoscale structures. The panels would have to track the sun for this to work, though.
Let's say that the 3D panel has 10 times the surface area of a flat panel, with the same dimensions. It still receives the same 1400 W/sq m as a flat solar panel, so the amount of solar power going into each sq cm of the panel has to drop to 1/10. It seems to me that the 3D panel wouldn't produce any more power than the flat design.
So there has to be a second effect at work. Let's see if we can find a better article than the information-starved FA? this article claims that the efficiency is increased due to reflections, i.e. each photon has more than one chance of being caught by a PN junction. Ah.
I wonder if this would work on macro scale, by placing two panels at a 45 degree angle to the sun, and 90 degrees to each other, like this \
That's really not true (about wind). It's entirely conceivable that humans could use almost all available (near-ground) wind power, if we chose to make that our only power source. And long before we even get to harnessing 10% of the available wind power, you're going to see big changes, like climate shift, thanks to the reduced power of the winds.
That's completely, totally, laughably wrong. MOST light that hits the ground is STILL reflected outward, back into space. And a significant amount of the light that is absorbed, is STILL radiated back out into space, shortly thereafter.
The rest isn't necessarily converted into wind... You don't just need high temperatures, you need significant temperature *differentials* to generate appreciable amounts of wind.
Complete nonsense. You don't need near 100% efficiency, much lower efficiencies will do a perfectly good job reducing the temperature of the deserts. And you certainly don't need to absorb heat... The deserts get most of their heat from the sun hitting the ground, not from some magical source of "hot" in the atmosphere.
Did you have a point, here, other than baselessly brushing off his concerns? "[Having] an effect" could potentially be very bad.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
1) The cost of the land would have to be taken in to account
2) The cost of maintenance would be taken in to account
3) The cost of legal fees and vet bills for treating spontaniously combusted neigbours pets would be taken in to account.
The parent's point still holds, the important factor is the total cost of a PV system (installation, land space, maintenance and enclosure costs included) divided by the power it produces.
Hmmm. I read the linked article, and from the truly minimal amount of hard information, it suggested that: 1) They're using some SOI process 2) They're using a laser ablation/etch process to define the cell boundaries 3) Each cell then has to be glued & connected, perhaps on a flexible substrate. Brilliant! I mean, why bother using cheap, OTS silicon wafers, when you can use expensive SOI, slice it up into tiny fragments with a laser, and then throw cash at the packaging! Yes, I am an OE engineer.