A market in movement

Utility-scale solar is becoming increasingly popular. Investors and energy producers want to build ever larger solar power stations in the U.S. and in many European countries. Two examples are the 60 megawatt Parque Fotovoltaico Olmedilla de Alarcón in Spain and the 25 megawatt Desoto Next Generation Solar Energy Center completed in Florida in October 2009. The goal is to achieve an optimal energy yield at the lowest possible cost. The system selected plays a considerable role. The choice essentially consists of three different systems, each of which has its pros and cons. The least expensive option is a fixed system installed on the ground; during the course of the day, the sun passes over the fixed system. This option costs the least, but at the same time does not utilize the modules’ full capacity.
In the second option, the modules move from east to west along one axis, tracking the sun. As a result, the amount of energy produced increases; however, additional costs arise because of the more complex technology and higher maintenance requirements.
The last option is a two-axis tracker system. The modules are not only aligned to follow the course of the sun from left to right, but also based on the height of the sun in the sky. In other words, the modules can be tilted forwards and backwards. The advantage is that the modules can provide the maximum performance possible for the position; the disadvantage is that the technology is even more complex than the single-axis tracking system and hence more prone to fail. Which technology is best for a given place depends on the geographical position and further factors, for example, module prices and property prices. The local feed-in tariff, of course, also plays a role.
The number of large-scale projects using tracked arrays jumped between 2006 and 2008. According to the study “Large-Scale Photovoltaic Power Plants” (pvresources) roughly a third of them have tracking systems. The proportion of tracked arrays was still relatively small in the U.S. and Germany, whereas in Spain the number of immobile arrays in 2008 was almost the same as those fitted with trackers. More than two gigawatts were connected to the grid on the Iberian Peninsula that year – 1,057 megawatts through tracking systems and 1,230 megawatts from fixed arrays. The cap on the Spanish market and the collapse in module prices ended the boom, but Denis Lenardic, author of the study, which closely follows every movement on the markets, says that “On a smaller scale than in Spain, but in the next few years, the number of tracked arrays will again rise, especially in the U.S. and in China.”
Iker San Sebastián has been implementing photovoltaic projects in Germany and Spain for ten years. He works for Sunedison, a Baltimore-based project planner that finances, installs and operates distributed power plants. With almost 100 megawatts of installed capacity, it’s the biggest solar service provider in North America. Up to now, most of Sunedison’s PV power stations have been fitted with fixed modules. San Sebastián believes, however, that the number of tracker projects will rise in the next few years. “We’ve installed arrays with tilted single-axis trackers in Europe and North America,” he says, “but we’re also open to azimuth trackers and biaxial tracking, and we’re going to build arrays using them this year and next.” But first they have to decide on a manufacturer. Many new producers have flooded the market, yet still suffer from a lack of experience. But even more experienced manufacturers haven’t always satisfied San Sebastián. The wrong decision can quickly become very expensive. A tracker that fails generates repair costs, often in distant locations, while earning nothing in the meantime. San Sebastián has a German manufacturer in mind for an azimuth tracker and a Spanish one for a biaxial tracker, though nothing is decided yet.

Quality control is crucial

Trackers are often viewed critically. Their complex technology, using motors and moving parts, seems to make them susceptible to defects and failure. Nevertheless, yields from all of Sun Edison’s projects using single-axis trackers have been better than predicted. San Sebastián puts that fact down to the comprehensive monitoring systems the firm has installed. All system data are monitored online, and in the U.S. a team of service technicians can intervene immediately if a malfunction occurs. “As a result, our yields are up to 110 percent of the figure predicted.”
The National Renewable Energy Laboratory (NREL), based in Golden, Colorado, publishes data for the U.S. in its Solar Radiation Data Manual for Flat- Plate and Concentrating Collectors (“Red Book”). “Look at Boulder, and you’ll see that there’s on average 5.5 kilowatt-hours per square meter per day from a fixed panel, mounted facing south at latitude tilt.” Matthew Muller, an NREL engineer, gives a rundown of how different systems perform. “For flat panel, single-axis, this becomes 7.2 kilowatt-hours per square meter, which corresponds to an improvement of 31 percent.” The big insolation map in the Red Book says conditions at Bolder, 45 kilometers northwest of Denver, aren’t bad for solar energy – about the same as in Spain, but still well short of the best locations in California. “A 34.5 percent improvement can be reached here with dual-axis tracking,” equivalent to 7.4 kilowatt-hours per meter per day. Again, this is different for every location. “Typical numbers reported are a 20 to 25 percent improvement for single-axis and a 35 to 40 percent improvement for dual-axis tracking,” says Muller.
The efficiency of a system at a given location can be estimated fairly accurately. A project planner’s wishes aren’t necessarily crucial, however. It’s the local authorities which decide in the end whether to authorize a project.

Policy has a part to play

Every district applies its own regulations. If, for example, some local mayor is an enthusiast for solar energy, large arrays will encounter few obstructions. But in certain areas, there are also quite clear restrictions, as in Greece, where a photovoltaic system must not exceed 2.50 meters. On the other hand, such conditions can offer a selective advantage to some market competitors.
“Our arrays were predestined to be used there,” says Philipp Steinhöfel sales leader at Deger, a company based in Horb, Germany, that makes single-axis and biaxial tracking systems. Deger claims to be market leader for sensor controlled tracking systems that orient the modules towards the brightest spot in the sky. They were able to score points in Greece with their low single-axis system. The same was true in Italy, where large megawatt projects are built. Steinhöfel sees a number of lucrative markets in Europe, including the Czech Republic with its generous feed-in tariffs. Nevertheless, he feels many operators won’t risk investing there because political conditions aren’t very stable. “It’s a similar gamble in Greece,” says Steinhöfel. “We started up some big projects, but they ground to a halt because of government bankruptcy.” Deger’s orders in Greece and Italy helped it get through the financial crisis. Its main product is biaxial trackers, and Steinhöfel believes in biaxial technology even though the worldwide sales figures weren’t all that good last year. His experience confirms the NREL figures. “We achieve a genuine 40 percent in the south.” Deger is in opposition to the idea that dual-axis tracking gets uneconomical when the module prices are very low. “The project price was around 3.20 euros per watt-peak last year,” says Steinhöfel, “but it’s now between 1.50 and two euros. And the tracking systems keep on selling.”

An uncomplicated calculation

Steinhöfel finds the calculation simple: a biaxial system costs 20 percent more, but in return delivers 40 percent more. In purely economic terms it’s the best system. He likes the American mentality that wants to know about benefits first, then costs. This is in contrast to Germany, where the price per kilowatt peak installed is the most important criterion.
Sunpower is among the manufacturers that make the best modules on the market. In approximately 90 percent of the solar plants it builds, it fits single-axis tracked arrays. Matt Campbell, Sunpower’s director of utilities, takes a pragmatic view of this: “Trackers are delivering the best cost of electricity,” he says. “If the fixed structure was better, we would just use fixed structures.” He says single-axis trackers have the advantage of not being much more expensive than fixed mounted systems. For biaxial systems, however, he sees little room. “Dual-axis trackers were most popular in Spain because panel prices were very high and people tried to maximize every bit of energy from their systems.” But he says the market has shifted. “Two-axis trackers once made sense and now don’t.” Besides, for single-axis systems he still sees room for cutting costs, with lower cost structures and cost controls: one possibility would be to preassemble trackers, reducing installation cost and installation time.
Or to manufacture them on site. Exosun wants to squeeze into this fiercely contested market by keeping infrastructure costs as low as possible. Calling themselves pioneers in French tracking systems, Exosun uses biaxial trackers developed in-house. What makes them special can’t be seen at first glance: the trackers aren’t made of the usual aluminum and steel, but of prefabricated concrete structures. “From an economic perspective, the price of concrete is less sensitive than steel to fluctuations in the price of oil,” says Frédéric Conchy, Exosun’s president. Their stands could even be manufactured at the construction location to save infrastructure costs. Conchy is convinced of the concept; the company recently built the first two megawatts of a solar farm slated to grow to 76 megawatts, making it Europe’s largest. To complete the company's portfolio, the single-axis version of the tracker will shortly reach market maturity.

No clear winner

NREL’s Matthew Muller regrets that a lack of long-term data prevents reliable assessment of which system works best. “I don’t have the data to say what operation and maintenance costs will be over thirty years.” He believes that more transparency on the part of system operators would be helpful, but the information remains behind closed doors: “Sunedison owns the systems, making the data internal to that company.” The question of which system will win out in the long term leaves him unruffled. If module prices fall still further, he can even imagine the end of tracked photovoltaics, whether single-axis or biaxial.
“Tracking won’t be economically viable, even though it produces more energy,” says Muller. Then it would be cheaper to simply buy more panels. “It would be cost-effective to have arrays knowing that you can walk away without having to maintain them.” But he says it probably won’t be so simple, as many factors are in flux, with the cost of land just as important as module prices. “In areas where land is most expensive, you might find that, even though panels are really cheap, it’s more cost-effective to have a tracker and get more energy density out of your land,” says Muller. This will differ across regions. “You can’t just put a number out there and say, when a module gets to this price tracking is no longer viable,” he points out. If the interest in large-scale arrays inflates the cost of land, or if arrays are so remotely sited that maintenance costs become too high, then tracking will be worthwhile again. Plus the cost of the land is only one of the parameters that decide whether spending extra money on a tracker is worth it. Political measures such as feed-in tariffs and module efficiency are just some other variables that influence any calculations, and these also vary regionally and over time.

The extra yield isn't worth the extra cost

Interview: The market for tracking systems is changing. While some start-up companies keep trying to bring biaxial tracking systems up to market maturity, Solon is preparing to ditch the technology. Lars Podlowski, Solon's chief technical officer tells us why.

At PV Powerplants in Las Vegas last December, you said that the end had come for biaxial trackers. Why so?

The cheaper solar modules become, the less useful biaxial tracking systems are. If I want 30 percent more electricity, nowadays I can simply add more cheap modules or hang them on a single-axis tracker.

Biaxial trackers are technically complex, but promised to make up for that with comparably higher performance.

Biaxial trackers were good while installed photovoltaic output per kilowatt-peak was expensive. 30 percent extra yield was genuinely worthwhile then. But the cheaper modules become and the further prices fall, the less return you are going to get in extra yield for the extra money that has been invested. After a certain point, it’s cheaper simply to add on to it 30 percent more modules.

It’s said that, in the sunnier parts of the world, up to 40 percent more yield can be obtained.

No. That’s a myth. Trackers give you more when there’s a lot of sun, that’s true. But the benefit mainly comes as you move further away from the equator. Right on the equator, you can mount the modules almost horizontally, and a tracker will add little more yield. In the Southwestern U.S., a biaxial tracker delivers little more than a cheaper single-axis tracker.

If, for example, we set up a system in California, with a fixed installation, a single-axis tracker and a biaxial tracker, how much will their performance differ?

In California, if a fixed array delivers 100 percent, a single-axis tracker will do about 122 and a biaxial tracker about 130 percent. Naturally I will get more output with a biaxial tracker. But the giant leap between modules fixed on stands and a single-axis tracker is far more interesting looked at from an economic perspective.

And what’s the proportionate cost of the tracker in a module-and-tracker configuration?

If modules cost only 1.60, as at present, the cost of the tracker is considerable. A biaxial tracker certainly still costs between 70 and 75 cents per watt. A single-axis tracker costs approximately 20 cents.

Why do biaxial trackers involve high costs?

They have a large number of components. The steel construction alone is relatively elaborate, expensive and space-consuming, so the logistics are expensive as well. Plus, of course, they are assembled on site. You have to have these central concrete stands, and they have to be cast. And yet only a fifth or a sixth of the space is exploited. As I said, when the overall cost of photovoltaics was higher, the extra 20 to 30 percent yield justified everything. Nowadays, however, the extra yield isn’t worth the additional investment.

Is the price difference between single-axis and biaxial trackers so great that it never repays itself?

Exactly. You can lay out single-axis trackers in long rows, making relatively full use of the area. Biaxial trackers are such large, isolated objects that it is necessary to leave relatively large gaps between them in all directions. So the area utilization with a biaxial tracker is poorer than a single-axis tracker.

Then you think that the single-axis tracker system is the one that will win out in the long term?

Definitely. Many large-scale projects now being built in the U.S. have been planned around single-axis trackers, because single-axis trackers currently result in the most cost-effective system.

In the end, it’s sales figures that count.

Yes. And I believe the sales figures are only a result of economic calculation.

Has Solon already begun converting its production process?

Yes. We’re still selling biaxial trackers this year – approximately ten megawatts of them – in countries where projects using this system have already been applied for. After that, it’s over. We’ll stop making this product.

Solon said it was the market leader in tracking systems in 2007.

That’s true.

Surely you won’t enjoy giving up such an important position?

Of course not. We certainly don’t look forward to losing it. But if general market conditions mean it’s no longer worthwhile, there’s no sense hanging on to something as a matter of principle. We’re reacting along with changed circumstances by building single-axis trackers as well. At a certain point, it became clear that the time for biaxial trackers had gone by. So we began swinging round to single-axis trackers relatively early, and are now building such systems.