In the solar industry, no efficiency gain is too incremental to matter.
When building big projects, every cent per watt is important to developers. About 80% of total project costs are now tied to the number of modules in an installation and how much power they produce, according to Wood Mackenzie.
As declines in module prices slow over the next decade, gains in efficiency will become increasingly important to push down solar costs. If producers can also keep production costs down, higher wattages will mean that more solar-powered electricity is in use in 2030 than today.
Manufacturers are already relying on new engineering techniques, larger sizes, and efficiency gains to better power output. The technologies the industry will be building in 2030 may not even exist yet, but those becoming popular today foreshadow where the market could move.
Increasing the power output of solar modules
Solar manufacturers currently have three main ways to increase power output, according to Xiaojing Sun, a senior solar analyst at Wood Mackenzie.
Panel makers can increase efficiency, which explains the evolution in market preference from multi-crystalline silicon modules to the current standard monocrystalline. Producers can simply make the panels bigger; larger wafer sizes are becoming increasingly common. And manufacturers can change how the panels themselves are engineered, manipulating the flow of electric current through cells to produce more power.
All three techniques are already in use.
China's Jinko Solar currently holds the record for commercial efficiency in p-type mono PERC technology, the mainstream standard, at 21.4%. Using silicon technology, the industry could produce panels for the utility-scale market with 25% to 27% efficiency by 2030, said Jenya Meydbray, CEO at PV Evolution Labs, a solar reliability testing lab.
And while p-type mono PERC is currently the mainstream cell technology choice, many analysts and companies also expect the more efficient (but more expensive) n-type to gain traction in the coming years.
Bigger panels are already becoming more ubiquitous. In 2012, 125 mm wafer sizes dominated. Over the years, that has increased to over 166 mm, and now even bigger sizes are gaining in popularity. Increases in wafer area size, up 80.5% for 210 mm wafers from the standard 156.75 mm of years ago, mean fewer panels and lower hardware costs for many installations.
“Projects using large modules are going to have lower capex and lower [levelized cost of electricity],” said Sun. “I think large modules are going to become very, very popular within the next five years.”
Companies such as Jinko, Longi and JA Solar are already pushing 182 mm wafers. Other producers are going bigger. Tianjin Zhonghuan Semiconductor, a partial owner of SunPower spin-off Maxeon, introduced the 210 mm wafer last year. Canadian Solar is “jumping over” 182 mm wafers and going right to 210 mm form factors, where the company will focus in the coming years, said Thomas Koerner, vice president for modules and system solutions.
Meydbray thinks the use of smaller wafers is likely to fade long before 2030.
There is, however, such a thing as too big. Koerner told Wood Mackenzie he believes panel sizes have already reached a “certain limit” beyond which further growth may be more of a hassle than it’s worth.
Beyond just growing the panels, manufacturers also must focus on making them work better. That is where engineering changes come in. Half-cuts (literally cutting cells in half) and multi-cuts add minimal cost and reduce resistance losses, increasing output. Adding more busbars or “ribbons” (the wires that conduct electricity across cells) does the same.
“Those easy physical manipulations are already happening, and they’re going to stick around,” said Sun. “These are very low investment, and in the case of multi-busbars, it’s even cost-saving.”
As the years' pass, it is not unlikely that two or more of these output boosters will be combined in single products.
“That may not happen this year or next year, but eventually, I do expect to see an n-type made on a large wafer made with multiple cuts and multi-busbars,” said Sun.
It is difficult to predict which solar technology will dominate in 2030. More likely than not, it has not even been invented yet, Meydbray says.
“We don't know what the technology is going to be, frankly,” he told Wood Mackenzie. “When looking at the leading technology ten years from now, it probably doesn't exist yet. It's probably not on the horizon yet.”
Downstream implications of design changes
The swift evolution in panel sizes and efficiency has divided the industry. Some companies, such as China’s Tianjin Zhonghuan Semiconductor, have led the changes. Others are lagging behind or choosing to stick with technologies like 166 mm wafers to maximise investments they have already made.
That strategy runs the risk of making some capacity obsolete as the industry pushes ahead, said Sun.
Koerner pointed to that logic as one reason Canadian Solar has chosen to head straight toward larger wafer sizes. “I’ve seen companies making [research and development] and technology decisions which only lasted for a year before they realised it was the wrong decision,” he told Wood Mackenzie.
Changes to panel sizing and structure often mean upgrading manufacturing lines, but not every company has the money to do that continuously.
Lack of a single standard also challenges balance-of-system producers to keep up. Inverters need to be able to handle higher output, while trackers must be able to physically support larger panels. Different modules may also require different electrical designs when installed.
“When module design starts to change a bunch, it starts to have pretty big implications on other systems design costs that you can’t ignore,” said Meydbray. “You can't just look at module dollar-per-watt [costs] and have that be sufficient anymore; you have to look at the other downstream implications.”
Size also has implications for shipping.
“That’s why so far, we only see G12, and we haven’t seen G15,” said Sun, referring to another name for 210 mm wafers. “If you get that large, you are going to waste a lot of your container space.”
Sticking to a single size is not the answer. But Meydbray said some standardisation is critical, keeping project developers from needing to completely redesign a system based on which type of module they choose.
Overall, the challenges around PV design are marginal compared to the obstacle of scaling up solar in the coming decade to account for a far more significant part of electricity generation.
Pushing solar power beyond currently conceivable limits will require even greater efficiency gains, with new technologies like perovskites, which we will cover later in this series.
A lot has changed for solar power in the last decade, and Koerner expects the 2020s to be no different. Though he has “no clue” about the form panels will take in 2030, he adds: “We have not reached the limit at this point on what’s possible.”