Analyst: First Solar Reaches Grid-Parity Milestone

By Ucilia Wang

First Solar (FSLR) has made it to grid parity, according to at least one analyst.

A 12.6-megawatt system installed by First Solar for Sempra Generation showed that the system can produce electricity at below the price of conventional power in the United States, said Mark Bachman, an equity analyst at Pacific Crest, in a research note Tuesday.

The solar power plant, located in the Nevada desert, costs $0.075 per kilowatt hour to install without any subsidies, Bachman wrote. Conventional power fed into the grid costs $0.09 per kilowatt hour.

"In our view, the industry leaders will be those companies that can deliver electricity at or below grid parity pricing without the aid of subsidies while also delivering superior return to shareholders," Bachman said. "Currently, only First Solar can claim these achievements, in our view."

Bachman's cost calculations, of course, are impacted by a number of factors and others will likely come to different conclusions. Part of the calculation relies on what others are achieving in other locations with different kinds of panels. Nonetheless, it underscores the progress the industry is making toward the important milestone.

And First Solar isn't the only narrowing in on it. On Monday, Cypress Semiconductor (CY) CEO T.J. Rodgers told a group of reporters that power from crystalline silicon solar panels will be cheaper than coal power by 2012 when transmissions lines, utility bureaucracy and other factors are added in.

"We are zeroing in on parity," Rodgers said. ""We're going to match PG&E (Pacific Gas and Electric) by 2012. Within a couple of years, the price of solar will be just as cheap."

Rodgers invested in SunPower (SPWRA) in 2000 when it had 40 employees. He turned the shares over to Cypress later. SunPower now sells billions worth of panels a year.

First Solar, based in Tempe, Ariz., makes thin-film solar panels using cadmium tellurium as the key ingredient to convert sunlight into electricity. It's one of a handful of thin-film companies to be producing panels in high volumes and the only one turning out cad tel panels in volume.

Most of the solar panels today use crystalline silicon, which is able to convert more sunlight than materials used by thin-film makers. Next year, a handful of manufacturers will start making copper indium gallium selenide (CIGS) solar cells.

First Solar's claim to fame for the past several years has been in its ability to churn out large numbers of panels and a fairly low cost. Last month, the company said it was able to produce panels at $1.08 per watt. The figure, however, is a blended average of all of the company's factories. First Solar's cost out of its Malaysian factories is lower, closer to 75 cents.

The $40 million system at Sempra is comprised of 168,300 panels, which First Solar installed at a cost of $3.17 per watt, Bachman wrote. (The installed cost is higher because it includes frames and installation, not just the solar module.)

He then used SunPower's installation of a 14.2-megawatt system at the Nellis Air Force base in Nevada for comparison. Bachman said SunPower's crystalline silicon panels cost $7.04 per watt to install.

After figuring out how much electricity the system is generating, Bachman determined that it costs $0.164 per kilowatt hour.

He concluded that for the SunPower system to generate electricity at the same rate (in kilowatt-hour) as First Solar's, SunPower would have to cut its panel prices by 52 percent and sell them at $3.4 per watt.

But Bachman went on to argue that the solar industry cares too much about the cost of producing and installing panels, and not enough about the how much a system costs in terms of its power generation, in kilowatt-hours. He noted that financial analysts should do so when dissecting the average selling price of a company's panels.

"By focusing on the cost/kWh calculation, we can compare competing

business models on a defined metric that is independent of technologies," he said.

Bachman also figured that First Solar's system make more efficient use of real estate than SunPower's, something that contradicts SunPower's claim and the estimates of others.

SunPower's system needs more space because it uses a tracking system to point the panels to the sun throughout the day. Crystalline panels perform the best when facing the direct sun.

By using the tracker, however, the panels need to be farther apart to avoid the shadow cast by other panels as they follow the sun, Bachman said. That means it would take more space to fit a SunPower system than a First Solar system at comparable capacity, he said.

First Solar tilted its panels at a fixed angel instead. Its cadmium-telluride panels might not be as efficient as SunPower's panels under direct sun, but the thin-film panels do better at converting diffused light.

SunPower, however, achieves a far higher efficiency. It sells panels that can convert 20 to nearly 23 percent of the sunlight that strikes them into electricity. First Solar's cad tel panels wallow around the ten percent range.

Comments

[GMI1982:]

This article is pure marketing BS. If First Solar sold this deal for $3.17/W, it was either a loss leader or fraught with miscalculations. This would be more like be First Solar's total cost with modules at $1/W plus another $2/W+ for installation. Given the nature of the technology, FSLR has higher installation costs as many more modules are required per acre given the lower efficiency. Also, the modules are tied together using DC, and this further increases the losses. Other than a-Si, CdTe is the worst performer from an area/MW and install cost perspective.

The numbers do not appear to reflect system inefficiencies either. The calculation of Wp are at the module level, and there are significant losses at the system level as well as yearly degradation. Pacific Crest does not know how to do calculations on LCOE, that is for sure. The real LCOE needs to include land costs, O&M expense, management costs, grid connection fees, etc. If FSLR has truly reached grid parity, you'd see MANY more plants being built today...

The acreage comment is totally misleading. If you look at the original estimate for the project, it was about 8 acres/MW. See this link: www.semprageneration.c.... c-Si with a tracking system is closer to 6 acres/MW and other leading utility-scale technologies, e.g. CPV, are below 5 acres/MW.

Bottom line, in terms of grid parity, CdTe and a-Si are well behind c-Si and CPV particularly in areas of high DNI (solar irradiance) like the Nevada desert. Flat plate c-Si produce 160W/m2, CPV = 250W/m2, and a-Si is well below 100W/m2. With these kind of inefficiencies at the cell level, you cannot possibly produce systems near grid parity today...

2008 Dec 17 09:10 AM

[Road Runner:]

It's so frustrating that there is so much spin going on with the performance numbers of solar verses grid that it is painfully difficult to compare them. Every time I look at a number I have to disect the true meaning. Like installed cost - what does that include? Seems that rarely the installed cost is a true "turnkey" cost.

2008 Dec 17 11:58 AM

[Road Runner:]

Also on the subject of spin. When these articles convert the solar system fixed cost to a per unit cost like kWh to compare to eletric rates, is financing included, and at what interest rate?

2008 Dec 17 12:01 PM

[Keef Wivanef:]

Did I hear someone say spin?


Herald-Sun

May 15, 2007 12:00am

PLANS for a private $30 million, 15-megawatt solar farm in one of South Australia's most arid regions are likely to be signed off today.

The public will be allowed to invest in the solar farm in a similar way that investors buy units in forestry companies.

Chief executive of Green and Gold Energy, Greg Watson, will today meet state electricity officials to put the finishing regulatory touches to a project set to save thousands of domestic users more than 50 per cent off their power bill without having to install expensive solar panels on their roofs.

"I'm confident there won't be any issues," he said.

Mr Watson, inventor of the award-winning SolarCube technology, told BusinessDaily yesterday the 200,000 sq m facility was likely to be located at either Leigh Creek or Woomera, near the old rocket range, or Roxby Downs, near the Olympic Dam uranium mine.

"They are the furthermost areas that the existing power grid extends to," Mr Watson said.

The site, which will be leased from either the state or federal government, does not need to be cleared for the farm.

"The land is already so arid, that we do not need to grade, scrape or bituminise it," Mr Watson said.

He said the project had not received any government funding and at this point was not expecting any subsidies like those provided to coal-burning companies that generate electricity.

Mr Watson said he was using his own capital which was generated from licence fees he collects from overseas distributors of SunCubes.

The company plans to erect 50,000 SunCubes initially and will sell a quarter to the public.

The offer, which is open to individuals only and not to commercial groups, will allow customers to buy up to four $1,500 SunCubes each.

"We will have measures in place to stop some rich guy on a hill coming along and snapping up all the 12,500 that we are selling," he said.

The solar farm could be complete and hooked up to the grid by the end of the year.

Mr Watson said he planned to add units to the farm in stages and had the manufacturing capacity to create a solar energy facility producing up to 500MW of base-load electricity.

Mr Watson expects to receive licences within a fortnight to set up two power companies - GGE Generation and GGE Retail.

GGE Generation will build the solar farm and infrastructure to link it to the power grid and sell the electricity to GGE Retail, which in turn will sell the electricity on to the wholesale market.

GGE Retail customers will benefit by having their bills' total kilowatt hours use reduced in line with the size of their SolarCube investment.

GGE will own all the carbon offset credits the farm produces and its customers will be able to boast that their electricity was entirely produced from a renewable resource.

SPIN ON THIS WATSON!
WOOF

2008 Dec 17 02:57 PM

[vitamin_j:]

I agree with above two comments. While it is in the interest of individual companies and investors in those companies to claim their own measures of "grid parity," it is in the interest of the solar industry in general to come up with a uniform, accepted standard for describing total cost of a solar kwh and comparing that to the cost of grid power in a given region.

2008 Dec 17 02:57 PM

[SolarDave:]

I kind of enjoy the spin in these stories because I know that other people will come in more and join the discussion.

2008 Dec 17 03:57 PM

[Doc 224899:]

If somebody tries to tell you that thin film photovoltaic cells will be too expensive to achieve grid parity because you can't get cadmium and tellurium and copper and indium and and gallium and selenium easily and inexpensively, what are you going to tell them? YES WE CAN !

If somebody tries to tell you that rare metals in thin film solar cells will only be more expensive and difficult to obtain in a contracting global economy due to sharply reduced base metals production, and that there is no economically feasible way to extract rare metals other than as byproducts of mass-produced base metals, what are you going to tell them? YES WE CAN !

2008 Dec 18 10:57 AM