What comes after PERC?

During the current investment cycle of the solar value chain, which picked up in 2014 and is still ongoing, there is one technology that clearly dominates the investment decisions: PERC technology.

With the PERC upgrade cycle still in full swing as several announcements of equipment supplier Meyer Burger for new PERC upgrades in January 2017 highlights the question for the solar industry in general and the PV equipment manufacturers in particular arises, why is PERC technology so popular what will come next?

Based on our extensive research, we can draw six main conclusions: We expect to see significantly less new capacity additions announced from 2017-2019, compared to 2014 through 2016. For the next two years PERC upgrades will dominate the cell technology investments and mono-crystalline cell technologies will continue to gain market share over multi-crystalline cell technologies. It doesn’t look like we will see a significant resurgence of the interest in selective emitter technology any time soon. Other emerging technologies like diamond wire sawing of multi-crystalline wafers and novel/improved metallisation technologies will see steady growth, albeit at a much slower rate than the PERC technology adoption. Heterojunction cell production lines will only start to gain market share once the industry enters a new capacity cycle and PERC technology has fully matured. The arguments for our findings are as follows:

Capacity additions will substantially slow down

The past two years combined saw more than 50 GW of cell capacity additions/technology upgrades announced. These expansion announcements came on the back of unprecedented growth in PV power being deployed in China. According to official Chinese statistics some 34 GW of PV power was grid connected in 2016, more than double the 15 GW added in 2015 and more than tripple the 10 GW installed in 2014.

Global PV power additions in 2016 amounted to some 76 GW up from 55 GW in 2015.

Given the fact that in November 2016 China has lowered it 2020 target for installed PV capacity from 150 GW to 110 GW most market observers believe that China will curtail further PV installations in the next 2-3 years so as not to surpass the reduced target by too significant a margin. As of the end of 2016 China had registered 77 GW of grid connected PV power, leaving only 33 GW of new PV additions for the next three years combined in order to reach the lowered target.

While emerging markets in Latin America, South East Asia and the Middle East are projected to exhibit high growth figures in PV installations in the next three years, none of these markets have the size nor the maturity yet that they could compensate a significant shortfall in Chinese PV demand. Assuming that cumulative PV installations in China in the years 2017 through 2019 will not surpass 70 GW, there is a strong argument that global PV additions in this period will be more or less flat in the range of 70 to 80 GW annually.

This will leave the global PV industry once again facing severe oversupply issues seriously limiting demand for any additional manufacturing capacities. In fact it is very likely that a number of capacity additions announced in the past twelve months will get shelved once again.

 (Mono-) PERC for higher cell efficiencies

By the end of 2016 some 15 GW of PERC cell production capacities were deployed worldwide, of which more than 85% were for mono-crystalline cells. A further 20 GW of PERC upgrades/capacity additions have already been announced so that we expect close to 40 GW of PERC cell production capacities to be available in 2019.

The charm of the PERC technology lies in the fact that it lends itself both to upgrading existing cell manufacturing lines and is also the technology of choice for capacity additions that the industry has implemented during the current investment cycle.

Cell manufacturers benefit from the fact that PERC technology represents an evolutionary upgrade, meaning they can draw on all of their previous production experience and are not forced to massively change their production steps.

What makes the PERC upgrade even more compelling is the fact that efficiency improvements of around 0.8% – 1.0% absolutely can be reached with this technology, and that it does not require any relevant adjustments in the module-manufacturing step.

So far PERC technology has been mainly implemented for mono-crystalline cell production lines, as concerns regarding the “elevated temperature light-induced degradation” that has been observed on multi-crystalline PERC cells still persist.

Interestingly, Suntech made a press announcement in January that, based on a technology co-developed by Suntech and the University of New South Wales in Australia, this degradation can be fully contained. Yet the announcement by Suntech did not disclose if or when the Chinese Tier-1 manufacturer planned on introducing modules with multi-crystalline PERC cells to the market. We have reached out to the company asking these questions, but have not received a reply by the time of going to press.

For the time being Hanwha-QCells and REC remain the only leading integrated solar manufacturers that commercially offer modules with multi-crystalline PERC cells.

Mono-crystalline (PERC)-cells grow their market share

Over the past two years the price gap per W between p-type mono-crystalline and multi-crystalline wafers has fallen considerably, and is now in the range of €0.15-€0.2/W. With the price gap for mono-crystalline wafers narrowing and PERC technology on multi-crystalline wafers still being hampered by serious degradation issues it comes as no surprise that mono-crystalline cell technology has been gaining ground. One of the leading global wafer manufacturers, LONGi Silicon Materials, has responded to the increasing market demand for mono-crystalline wafers, and announced significant capacity additions. In January 2017 LONGi disclosed that it had started construction on a 5 GW mono-crystalline ingot facility in Lijiang, southwestern China, together with its two joint-venture partners Trina Solar and polysilicon maker Yongxiang. Production start for the new facility is scheduled for the first half of 2018.

Similarly to Trina, SolarWorld is shifting more and more of its capacities to mono-crystalline technologies. Even Hanwha Q-Cells, one of the leading multi-crystalline proponents in the solar industry, is adding more emphasis on its mono-crystalline activities, stepping up its investments in this area. In light of these developments we see strong indications that mono-crystalline technologies continue to gain momentum and thus market share.

Selective emitter technology does not appear to be the next big thing

With PERC technology minimising the loss contribution from the rear surface of the solar cell, and gradually becoming the next quasi-standard (mono-) crystalline cell technology as we predict, it is straightforward to assume that R&D in the solar industry once again shifts toward tackling the loss mechanisms attributable to the front surface. In 2010 and 2011 many equipment vendors promoted selective emitter technology as the technology of choice, in order to reduce losses at the front contact of the solar cell. Ultimately the gains (at the module level) proved to be limited as much of the increased short wavelength sensitivity of the solar cell achieved through applying selective emitter technologies did not fully translate into higher energy yields of the modules as both the front glass of the module as well as the encapsulant only exhibit limited transparency for this part of the solar spectrum.

In the meantime, the suppliers of the front side metallization pastes have made significant progress with their recipes, so that the new pastes achieve a good electric contact even on highly doped surfaces, reducing the need to selectively dope the front side surface of the solar cell.

Other technological developments

Among other technologies pushing forward, we at Smart Solar Consulting would highlight two technological developments that we believe stand good chances of capturing increasing traction over the next 2-3 years. One of these technologies is diamond wire sawing of multi-crystalline wafers. For mono-crystalline wafers diamond wire sawing has already established itself as the new standard for state-of-the-art production facilities. The higher throughput of the wafer saws and the lower kerf losses are convincing enough arguments for switching to this technology.

The adoption of diamond wire sawing for multi-crystalline wafers has been hampered by two factors: yield issues based on more frequent wafer damage when cutting wafers from multi-crystalline ingots with diamond wires and costly surface treatments required in order to achieve the desired surface topology for maximum cell efficiency. While improved crystallization processes have been able to reduce the yield issues to acceptable levels, there are still a number of surface treatment approaches to tackling the sawing damages that are being investigated in R&D labs and on pilot lines competing to provide the most cost efficient process step.

In a future issue of pv magazine we will take a closer look at competing approaches for the surface treatment and will highlight which technologies from our point of view have the best chances of reaching the cost and performance requirements set by the industry.

The other field of continued research tackles the front side metallization process. More than six years ago many experts predicted that screen-printing technology would soon be superseded by superior technologies, such as ink-jet technologies, laser deposition or plating technologies. As of today screen-printing is still the method of choice. The competing technologies all proved to exhibit serious disadvantages compared to screen printing when it came to production yield and cost, the ultimate determinants that decide whether or not a technology is suitable for industrial production. The screens, the printers and particularly the pastes have seen such significant improvements over the past six years that the bar for competing metallization processes has been continuously raised.

Yet we have heard of promising results both from plating approaches and extrusion technologies.  At the pilot line stage these appear to yield the required advantages, making them serious contenders to challenge the reign of screen printers for the front side metallisation process.

Heterojunction will have to wait

Besides PERC technology heterojunction cell technology is being promoted as the next big thing. At Smart Solar Consulting we don’t believe its time has come quite yet for a number of reasons:

Given the very limited growth expectations for PV demand in the next 2-3 years, the solar industry is likely to end up facing another oversupply scenario, which may last all the way through 2019. As we have already pointed out, the key determinant to the extent of the oversupply scenario is the market size for PV installations in China in the next three years. Unless China keeps adding more than 30 GW of PV power each year up until 2019 we believe there is hardly any chance for the photovoltaic industry to avoid having to go through another oversupply period. In order for the industry to achieve satisfactory utilization rates of their new production facilities, global annual demand would have to expand to around 90 GW in 2017 and beyond. If Chinese annual PV installations should fall by around 10 GW compared to the level reached in 2016, we don’t expect to see these demand levels globally before 2020.

Image taken at "Trinasolar" in Changzhou, China in January 2016.

In such uncertain times when price competition once again leaves all the manufacturers scrambling for cash the investment risk of adopting heterojunction cell technology appears rather high.

Even established cell manufacturers with many years of experience in mono-crystalline cell production have to make massive changes to their production processes when adopting heterojunction. The switch to heterojunction cell technology requires an investment in completely new production lines, without providing the option to upgrade existing tools. The limited availability of n-type mono-crystalline wafers at this point in time, and the fact that for heterojunction solar cells novel interconnection technologies at the module level (i.e. Smartwire and the likes) are being promoted as necessary to reap all of the advantages that heterojunction cell technology provides. This will require substantial investments at the module production stage, a further argument why we believe that widespread adoption of heterojunction cell technology will not occur before the oversupply woes can be mitigated.

The only other scenario that we could envision leading to a faster adoption of this technology would be if one or two Tier 1 manufacturers would leapfrog the PERC upgrade route and opt to directly invest in heterojunction. As for now we see no indications of anything like this being in the works in the market.

In summary, in terms of business success for the photovoltaic equipment vendors the emphasis in 2017 and 2018 is to undertake everything necessary to minimise the level of order pushouts or even worse order cancellations. We see a high risk that order cancellations might exceed new orders in 2017, in particular once the grace period for PV installations in China at 2016 FIT rates comes to an end by the middle of 2017. While we know of a number of interesting novel technologies that are already being evaluated beyond the pilot line stage, we are not holding our breath expecting a similar breakthrough of any of these technologies within the next 2-3 years as we have seen recently with PERC technology over the past three years.

Author: Götz Fischbeck