Solar power

From Silent Revolution

Topic: The solar revolution will not be televised until the TVs are already running on solar power

Right now, solar is a bit player in world energy markets. While far beyond its first commercial applications as a pricey scientific curiosity useful only for satellites in the 1950s and 1960s, past an esoteric way to providing expensive power to remote locations in the 1970s, and past being an overpriced power solution for trendy environmentalists in the 1980s and 1990s, it still only makes up a tenth of one percent of the world's electricity generation capacity -- most of that being solar thermal. One can run the numbers on the economics of photovoltaics and see that they're not currently a wise investment in most situations except for people far from the grid.

So, what's the big deal?

Contents 1 Terminology and background 2 The Lesser Revolutions 2.1 The rebirth of solar thermal 2.1.1 Power towers 2.1.2 Parabolic trough reflectors 2.1.3 Fresnel reflectors 2.1.4 Dish reflectors 2.1.5 Indirect solar thermal: solar updraft tower (chimney) 2.2 Sliver cells 2.3 High efficiency chemistries and rooftop concentrators 3 Thin films change the game 4 Nanoantenna arrays for tomorrow

[edit] Terminology and background

• "Solar thermal" means turning sunlight into heat and using that heat to generate power with traditional means
• "Photovoltaics" and "solar cells" means using special materials that use light to create electricity directly.
• Traditional solar cells are made of silicon wafers, similar to computer chips. The high price of crystalline silicon means that these cells are expensive -- currently $3/W or more to make and $4.80/W to purchase
• Whenever you see prices reported in terms of dollars/watts (such as $3/W), this is in reference to how much you'd have to spend to get a solar cell that produces one watt of power in "standard conditions". These standard conditions are 1000 watts of sunlight per square meter at a temperature of 25C (solar cells perform worse the hotter they get). In practice, "standard conditions" are only rarely met; however, they give you a number that you can then adjust to the actual climate and insolation (sunlight exposure) conditions for your given location (as in this calculator). A standard set of conditions is needed because everyone's results from a given cell will vary depending on how and where they were installed.
• Don't confuse "watts" with "watt hours". $3/W does not mean that a watt hour of power from the cell costs $3. Far from it. A 1W cell will typically produce 0.1-0.2W averaged over the course of a day in a sunny location, but will keep running for decades -- and there are 8,766 hours in a typical year.
• Solar cells do not remain at full power forever, and any good economics calculator will take this into account. However, this effect is often overstated. Rarely do cells, clean and with good terminals, produce less than 70-80% of their original power after decades of use. This "long use" benefit of solar cells is tempered by the fact that their purchase costs must be amortized (spread out over time), and even something that gives you a profit forever must still be able to pay back its investment faster than interest on the purchase accrues.

[edit] The Lesser Revolutions

Let's out and state it: the big deal in solar that's changing the world is thin films. But they're not the only player out there. A quick rundown of just some of the competitors, some of which may provide benefits to the masses a bit sooner:

[edit] The rebirth of solar thermal

Solar thermal power has an old history, dating back to the 1700s, with big pushes in the 1800s and early 1900s. It fell by the wayside with the increasing availability of fossil fuels during the 20th century. However, in recent years, a wide range of solar thermal designs have been constructed or are planned, each with significantly reduced cost per watt or cost per watt projections. It is worthy of note how rapidly the expansion of the industry is occurring, both in frequency and scale of new plants. In California alone, right of way requests have been filed for 34 solar thermal plants totaling 24GW of power. Even though thin film is on its way to taking over the future market, for the next ten years, solar thermal will be a big player in producing solar electricity.[1] Below are just some plants that have made news.

[edit] Power towers

Power towers are currently usually not the most economical solar thermal designs in terms of total MWh/day provided. However, a fair amount of investment continues to flow in due to a perceived potential for improvement and easy heat storage for nighttime operation. Whether these reduced costs can equate to grid-rate power production is yet to be seen. Recent improvements have included much simpler/cheaper heliostats and simpler working fluid management/heat storage.

• 1981: Solar One (?MW): Daytime only
• 1983: Themis (2MW): Similar to Solar One, but used molten salt to help handle sunlight interruptions.
• 1995: Solar Two (10MW): Expanded from Solar One. Could handle minor interruptions in sunlight due to switching to molten salt, as in Themis.
• Present: PS10 (11MW): The first in a series of towers to total 300MW. The cost for power generation from PS10 is about three times higher than the local grid rate, but the price is expected to go down with each additional tower.
• Under construction: Solar Tres (15MW): Enough thermal storage to run 24-7, thus giving it a very high capacity factor for solar (65%). Simpler design than Solar Two.
• Announced; pilot plant under construction: Ivanpah (400MW): Promises the lowest cost from "photon to electron" of any solar power plant (price per kWh unstated)[2]
• Announced: Uppington, South Africa (100MW): Goal is only to reduce the price of solar down to the price of wind.
• 2009/2010: Cloncurry plant (Australia) (10MW): To be the sole power source for the town of Cloncurry. Heat storage through graphite blocks.

[edit] Parabolic trough reflectors

Parabolic trough reflectors involve a long, curved mirror whose cross section is a parabola with a dark tube of working fluid for heating at the focus. The heated fluid then runs a turbine and optionally stores heat for nighttime recovery. The biggest recent advancements are in the design of the trusses. The trusses need to be rigid (to focus, especially when wind is present), light, and cheap. Computer modelling has made this a much easier task to optimize. Parabolic trough reflectors are approaching grid prices in terms of price per kilowatt hour and credibly have the potential to surpass it in some regions in the near future.

• 1986-1990: SEGS (354MW): A series of parabolic trough plants in the Mojave desert. Solar during the daytime, partially natural gas at night. Produces power at grid rates ($0.10-$0.12)
• 2007: Nevada Solar One (64MW): Daytime only
• 2008: Andasol 1 (Andasol 2 under construction, Andasol 3 starts in '08) (100MW total): 7 hour heat storage.
• 2008-2009: Kuraymay, Egypt Plant (40MW): Part of a combined cycle plant
• 2008-2010: Victorville 2 (50MW): Part of a combined cycle plant
• 2009?: Solel/Negev (150MW/500MW): Delayed from initial construction date (formerly due in 2008) due to delays in approved government funding. Initial version to be 150MW, and later expanded to 500MW -- 5% of Israel's generating capacity.
• 2009: Beni Mathar, Morocco plant (30MW): Part of a combined cycle plant
•  ?: Solar Power Plant One (Algeria) (25MW): Part of a combined cycle plant. Algeria plans to provide 5% of their power with renewables, mostly solar, by 2010.
• 2010: San Luis Obispo, CA (177MW): Ausra has raised funding to develop an unusual parabolic trough design involving fixed (non-tracking) flat mirrors that it says will produce $0.104/kWh power off the bat, $0.079/kWh power in three years, and when they start building 700MW plants, $0.067/kWh. According to Ausra Vice President John O'Donnell, "Once solar-power projects are built at the scale, and with the same financing, as fossil-fuel power plants, we're cheaper. We don't really need a level playing field."[3]
• 2011: Mojave Solar Park (553MW): Large parabolic system for PG&E. Solel is also building a series of 150MW plants in Spain in advance of this project.
•  ?: Yazd (Iran) (67MW): Provides steam for a hybrid solar/gas plant.

[edit] Fresnel reflectors

Fresnel reflectors are similar to parabolic troughs except they use arrays of cheap flat mirrors along the ground instead of raised, curved troughs. A recent advancement, the Compact Linear Fresnel Reflector (CLFR), reduces shading between individual mirrors by having mirrors alternate between aiming at two different tubes. This allows them to use much less land for the same amount of power. Fresnel reflector systems, like parabolic trough systems, are approaching grid rates.

• Under construction: Liddel Power station (95MW): A CLFR under construction will provide an additional 95MW of steam to the existing fossil plant.
• 2010: San Luis Obispo plant (177MW): Under construction
• 2011: Florida Power & Light (300MW): Under construction

[edit] Dish reflectors

Dish reflectors combine the "single point of focus" of a solar tower design with the lack of tower requirements of parabolic and fresnel reflector systems by having heliostats of mirrors constructed like numerous giant satellite dishes. The downside is that the dishes need significant spacing to avoid shading each other. This design is relatively new, and its economics not yet tested.

• 2008-2012: Pisgah/Victorville plant (500MW-850MW): Dishes with hydrogen gas stirling engines for nearly double the efficiency of solar trough designs. Daytime only.
• 2010?: Imperial Valley plant (300-900MW): Also stirling engines, like above.

[edit] Indirect solar thermal: solar updraft tower (chimney)

Solar updraft towers, also known as "solar chimneys", are a relatively old design (1903) that has largely remained obscure until recent years. In a solar chimney, sunlight heats up what is effectively a gigantic greenhouse and is funnelled toward the center, where the stack effect draws it up and out a tall tower. The motion of the air drives turbines to produce power. Solar energy can be stored in the ground or water pipes for nighttime power generation. The "greenhouse" could be a literal greenhouse near the edges where the winds are calmer. The economics of such towers is still highly speculative.

• 1982: Prototype in Manzanares, Spain (50kW): Encountered severe structural instability due to induced vortices near the tower and was closed in 1989.
• Proposal (dates uncertain): Ciudad Real Torre Solar (Spain) (40MW): would stand over twice as tall as the tallest structure in the EU
• Proposal (dates uncertain): Solar Tower Buronga (Australia) (200MW?):

To deal with the problems of this design, a number of modifications to the design have been proposed. The "floating solar chimney" proposes a lightweight chimney kept aloft by balloons full of lighter-than-air gas. In another design steep mountainside could be used to support the chimney. The tower could be run in reverse as a downdraft tower, spraying water at the top to cool the air like a swamp cooler.

Perhaps most interesting, a theoretical method to eliminate the chimney altogether exists through the creation of a vortex, as an artificial dust devil or tornado. A so-called "vortex engine", if it proved workable, would eliminate most of the capital costs of a solar tower. Even better, the heat from a large vortex would reach high into the stratosphere, thus radiating more effectively and contributing to cooling the planet. A vortex engine would also provide benefits to existing power plants by providing an alternative to expensive cooling towers, and could even be used to tap oceanic heat. The technology is currently being studied by the University of Western Ontario with seed money from OCE's Centre for Energy

[edit] Sliver cells

Sliver cells are a silicon photovoltaic cell-based technology designed to use thin "slivers" of crystalline silicon instead of a whole wafer for each cell, thus reducing the amount of expensive silicon consumed tenfold. The downside to sliver cells is that the process to make them is itself expensive. Origin (the company making them) expects industrial-sized modules to produce "at around market prices" for electricity when released.

[edit] High efficiency chemistries and rooftop concentrators

The efficiency of the most efficient solar cells has been taking off. While commodity silicon solar cells have efficiencies around 20%, higher efficiencies have been achieved by combining silicon with other chemistries, such as gallium arsenide. While more expensive, the higher efficiency numbers help offset this cost increase. To improve the numbers even more, some cells are starting to include additional layers, such as fresnel (thin) lenses to concentrate, beam splitters to allow parts of individual parts of the cell to be optimized for specific frequencies, materials that produce two electrons for every one photon of sufficient energy, and phosphors to turn solar energy of undesirable wavelengths into energy of desirable wavelengths. To drive home how fast the efficiency numbers are advancing in a field where a 1% improvement is typically viewed as a leap forward, here are the last three advances:

• October 2006: Sharp: 36%
• December 2006: Boeing-Spectrolab: 40.7% (target production goal: $3/W)
• July 2007: University of Delaware/DuPont: 42.8% (splits light into three wavelength groups for optimal conversion via a built-in variable concentrator; $100M invested in commercialization)

With further beam splitting, efficiencies could potentially reach over 60%

Even if these cells don't become economical enough on their own, their extreme efficiency combines nicely with another rapidly advancing technology: solar concentrators (especially rooftop-mountable concentrators). These are typically small modular devices, easily mass-produceable, which focus light on a single small high efficiency solar cell. Sizes have gone down enough that the weight is low enough to be realistically mounted across the entire roof of a typical home or building; a modern solar concentrator is nearly flat. Since the cell is small, it contributes little to the total cost. The downside to this concept is that the intense light heats the cell, which lowers its efficiency. Nonetheless, concentrators have the potential to significantly lower solar power costs, especially for home installations. The technology ranges from passive focusing, such as holographic tubes that keep sunlight focused on cells in the center no matter what angle the light strikes from, to active systems with arrays of mirrors that share a common drive system but use clever gearing to keep them all focused on the same point. No prices have been disclosed yet, but the companies involved (Prism, SolFocus, Greenvolt, Practical Instruments, etc) all promise "significant" reductions in prices, and have attracted quite a bit of venture capital.

[edit] Thin films change the game

All of that being said, the ball is now in the court of thin films -- Cadmium Telluride (CdTe) being one, but perhaps more notably, Copper Indium Gallium Selenide (CIGS). Both promise incredibly low prices per watt -- $0.50-$1.50 -- but by avoiding cadmium, a toxic element, CIGS has a "greener" reputation. It's also currently the cheapest. Despite the use of some rarer elements in thin films, such as tellurium or indium, such tiny amounts are used that it's not a major factor. And contrary to some indium scare that's been reported, indium is more common than silver, which is produced in quantities 40 times greater. Indium's current high prices are more due to a lack of indium recovering circuits at zinc and other mines due to low prices back before LCD displays increased demand. Future indium production can easily meet CIGS needs.[4] CIGS cells are not only incredibly cheap, but they're also very radiation and degradation resistant to boot, [5][6][7] and can sometimes even increase in output after accumulated light exposure[8].

Leading the pack of the dozen or so competing CIGS companies is the privately held company that everyone seems to want to have a piece of: Nanosolar. Their still-under-development plant is designed for a capacity of 430MW/yr, which would put it in with the biggest manufacturers on the market today. The company claims an incredible production cost of only $0.30/W[9], making the company profitable with market prices of less than a dollar per watt. Their first ~600 MW of capacity is going to freefield deployment in east Germany for a price of $0.90/W, and Nanosolar has already received their first payment for delivery of product for the first deployment (to peak at 1MW). Nanosolar CEO Martin Roscheisen is quoted [10]as saying that his company's cells operate in the "9 to 10 percent range", with expectations of 15 percent in the future.

It's hard to overstate the significance of these numbers: with current solar prices at ~$4.80/W, being able to make cells for $0.30/W is like having a license to print money. Assuming Nanosolar, or really any of its many competitors, can achieve numbers even close to this, they'll have all the money they could ever want to scale up to orders of magnitude greater production. $1/W means that solar power is cheaper than coal even in Alaska. And with as many companies in the thin film field as there are, Nanosolar need not carry all of the burden. One CIGS company, ISET, has complained about Nanosolar getting all the press. They expect to soon have CIGS cells on the factory floor available for $0.65/W, down to $0.50/W as they scale up, and consider the technology scaleable to $0.40/W. Just a small fraction of the other companies include Miasole (CIGS), Konarka (polymers), Uni-Solar (thin film silicon), HelioVolt (CIGS), SoloPower, PrimeStar (CdTe), Innovalight (liquid silicon ink), and Solexant (broad spectrum thin film; secretive)

Speaking of broad-spectrum...

[edit] Nanoantenna arrays for tomorrow

Solar cells that generate power at night. The whole concept sounds crazy, and it doesn't sound any less crazy when you mention that these cells could have efficiencies as high as 80%. But that's just what we're looking at[11].

Researchers at the Idaho National Laboratory have worked out a way to produce nanoantenna solar arrays that pick up solar energy in the same way that an antenna picks up radio waves. Due to the extremely short wavelength of light, the antennae must be extremely small. Nonetheless, these thin-films can be produced in bulk -- likely even with a "roll-to-roll" process (one of the reasons why NanoSolar's cells are so cheap) using incredibly cheap polyethylene as the substrate. The incredibly high efficiency numbers are also why they can run at night; numbers like that are only achieved by absorbing energy from a wide part of the electromagnetic spectrum. Since the earth absorbs sunlight during the day and continues to radiate its heat at night as infrared light, the cells can also harness this infrared to generate power.

The only unsolved problem is that these cells produce extremely high frequency alternating current, cycling at about 10 THz. This is far too fast to be usable. The team next needs to produce equivalent cells that contain nanoscale rectifying diodes to convert the current to DC. Undoubtedly, one should expect additional small losses from the diodes. The team expects to need only a few years before they have a viable commercial product.

While only a fraction of technologies that work in the lab ultimately end up on the market, this is an exciting development, to say the least. If we were to assume an overall module efficiency of 70%, it could even make partially solar cars realistic.