As a technology, HJT has made remarkable progress, not referring to its infinite past, but just in the last 1.5 years. The latest reference point is our previous report on the topic that was published end of 2020. There was a long list of challenges for the technology and serious skepticism about if and when it will overcome those issues (see Heterojunction Solar Technology 2020 Edition).
As things stand, the majority of these issues have been addressed or, at the very least, there exists a commercially viable path to a solution. In fact, in the recent list of our TaiyangNews Highest Efficient Commercial Solar Modules ranking a HJT product has reached the second place with 22.53% efficiency (see here). However, our intention is not to say the technology has fully established itself, but the progress we have seen in the recent past is commendable.
One such challenge was the scalability to larger wafer formats such as G12. There were even questions on possible benefits. Today, production equipment suppliers have come out with innovative tool platforms that support high-throughput processing of G12 wafers. Going a step ahead, the tool platforms are even compatible with half-sliced wafers throughout the production line, starting from cells to finished modules. Processing half wafers safeguards from edge losses, which can be seen only at the module level. While this large form factor was used for PERC modules in the beginning, we are now seeing the first HJT panels with power ratings of 500 to 700 W based on 210 mm wafers too.
High efficiency is the most important attribute of HJT cell technology. While average cell efficiencies were about 24% as of our previous report, there are several companies today that are operating at about 24.3% today, with a clear roadmap to go even further. The industry has realized that the current is the major limiting factor due to the absorption losses in the window layer of amorphous silicon. Replacing doped amorphous silicon films with microcrystalline on the front side is expected to bring in the next efficiency boost. Replacing the rear side doped amorphous silicon, again, with microcrystalline films would also add up to it eventually. Hitting the 25% number with this so-called next generation seems practical. Equipment platforms are also ready to support such a switch.
Switching to microcrystalline also seems to be a key enabler for thinner wafers, a topic that invariably drags the discussion into costs. In fact, the wafer costs contribute about 50% to the equation of HJT cost of ownership, while n-type wafers are about 5% more expensive. Employing thinner wafers is one way to compensate for the high price. Referring to a recent price list from one of the wafer manufacturers, the reduction in thickness from 150 um to 130 um leads to a reduction in price by about 4.5%, and the price difference between p-type and n-type wafers is about 6%. HJT, with its symmetric structure and low temperature process, supports aggressive wafer thickness reduction.
Metallization constitutes the next major chunk of HJT costs. A deeper understanding of low temperature cured pastes, in addition to enabling several operational benefits such as quicker curing, ability to support fineline printing and high printing speeds, has also enabled reduced paste consumption while providing enhanced conductivity. An increase in the number of busbars per layout has also contributed a lot to the paste layout reduction. After 12 and 15-busbars have already been implemented, manufacturers are seriously pursuing 18 and even 24 busbars. The paste consumption for HJT cells is about 180 mg for M6 cells, realized by printing through a 28 μm opening and finger width of 35 μm. There is still some potential left.
On the other hand, the industry is seriously evaluating silver coated copper paste. A good blend of silver and copper have resulted in comparable performance figures, while its reliability is under testing with satisfactory first results. When implemented in large scale, the silver coated copper paste has high potential to reduce the costs. And whatever the progress made in terms of fineline printing with silver paste can also be adapted to silver coated copper paste.
The third most important process consumable is indium, which has limited global reserves and, hence, availability. Thus, when consumption goes up a lot, the costs will likely skyrocket. Reducing or eliminating indium is also a hot topic for TCO steps of the HJT process. Companies are working on alternative materials, and AZO seems to be a promising candidate. Companies and research centers are working on different configurations — AZO only on the rear side and stacks of AZO and ITO — and the initial results seem to be good. Whether these results can be replicated in mass production is something that needs to be assessed by further testing.
Next to OpEx, not necessarily in order, high CapEx is an equal concern with HJT. There is also a significant development in this area as well. Back in 2019, a GW HJT line was priced at 1 billion RMB. Within a year, by 2020, it was brought down to 450 million RMB per GW. As of today, prices sit slightly lower than 400 million RMB per GW. This is still 3 times higher compared to PERC. However, equipment makers are working hard to reduce the machines’ CapEx, especially through increased productivity and throughput.
On the production capacity front, HJT was never modest when it comes to expression of interest, but not without its share of criticism. Still, the installed capacity is merely at about 6 GW end of last year, while announcements exceed 80 GW. However, there are some serious players and their true stories. One example is Huasun from China, having successfully finished its second phase of expansion of 2 GW for a total installed capacity of 2.7 GW. There is several more working on HJT in China. But also outside the world’s largest PV production hub there is a lot of activity in HJT as companies see this high-efficiency technology as an entry point to compete with China’s gigantic solar manufacturing complex. India’s industrial conglomerate Reliance Industries with its deep pockets recently acquired early HJT cell/module producer REC from Norway/Singapore to establish GW-scale HJT capacity in India. In Europe, Meyer Burger wants to reach over 1 GW HJT capacity this year. And after Italian utility Enel in early 2022 was awarded a huge European Union grant that enables it to accomplish its longtime plan of expanding its first small-scale HJT line to a multi GW scale in its home country, REC was also selected in a second round of the same program for a GW level HJT factory to be built in France.
Well, HJT is just at the beginning of its commercial journey – and some of the large Chinese companies have yet to decide if and what’s the role of HJT in their future technology portfolio. One of them, LONGi Solar, just reported a new world record of 26.5% for a M6 sized cell in June, that’s over 1 percentage point higher than the best large-scale TOPCon cell. If HJT can keep its efficiency advantage there will be certainly always quite some space in the rooftop segment, but when it comes to utility-scale applications every penny counts both on the module sales price and LCOE.
However, the HJT cell architecture provides an attractive platform for further improvement in the long run. A well-known method is adapting the back-contact cell architecture to the HJT platform, called HBC, which still represents the highest lab efficiency of 26.7% attained by any crystalline silicon solar cell. As this is close to the theoretical limit for single-junction silicon cells, many are looking at next-generation tandem cells, where much hyped perovskite technology is expected to play a major role – and HJT is the best fit for the bottom cell of a perovskite-based tandem cell. Almost all HJT stakeholders are working in this direction.
But this is still a long way and to be discussed in a future edition. For the moment it’s great to see that HJT is Entering High-Volume Production – and that’s why we’ve picked it as the tagline for this report (see Progress Of Heterojunction Technology).
If you are at REI India this week, you can also pick up your printed copy of our HJT, which is being distributed during the show.