Category: Blogs

First Solar’s Series 6 module ramp-up constraint issues, explained!

In reporting second quarter 2018 financial results, First Solar revealed that it was having Series 6 production issues at its lead Ohio factory.

The company took the time in its earnings call with analysts to detail the problems, which centred around a yield issue at a specific tool/process at the upstream cell level, that then impacted downstream assembly throughput, aggravated by insufficient buffer stations. 

Not surprisingly, and I would not blame them, First Solar’s management did not reveal the specific tool and therefore cell process that was giving them yield issues. However, the yield issue led to shutting down of specific tools to diagnose, fix and recalibrate before testing and restarting that production step.

In the earnings call, Mark Widmar, CEO of First Solar said, “We have a single point of failure on the production line today.”
It is also important to note at this time that the Ohio factory is still in initial ramp-up mode with the first iteration of a complete new toolset for the large-area Series 6 modules. This is not mature production line(s) and the ability to make a perfect ramp to full nameplate capacity with everything new would be seriously impressive but unlikely.

Theory of constraints

In the typical linear thin-film module manufacturing flow, rather than batch-based flow where steps are generally isolated and have buffer stations, a single point of failure as Widmar described can create bottlenecks or throughput constraints before and after the point of failure.

“If that tool goes down, everything upstream is shutting down and we’re starting the downstream processing, right? So, that tool is critical because it’s a single point of failure. We have to have the availability and we’ve identified where those points are and what we’re going to do is we’ll buffer it with inventory. So, if the particular tool goes down, we aren’t starving the balance of the production and we can continue to run the production as an example, right?

In other words, the throughput impact of First Solar’s single point of failure (yield issue) at the specific tool can be reduced significantly if a sufficient amount of in-spec modules are produced and stored (buffer) so that the next steps are not starved of semi-completed modules so the final steps to complete modules can operate to their cycle time and throughput requirements.
The impact was and can be significant. Management noted that when the point of failure occurred, capacity would go from almost 100%, straight down to 20%.

This was certainly an unexpected point of failure as First Solar is rushing to provide enough buffer stations in the manufacturing flow to minimize the cycle time and throughput issues it is currently experiencing. 

“The layout at the back-end of the line was built according to the tool set availability specification which resulted in fewer acquired buffers. As we started to ramp the back-end of the line with a tool set availability not yet at a mature state, we realized there were multiple single points of failure in the line that could shut down production. Effectively, the line was not adequately buffered given the current format of the tool sets”, added Widmar.

“We are in the process of installing inventory accumulators to properly buffer the back-end of the line. Once completed in our Ohio factory, we will use our Copy Smart approach to roll out to Malaysia and Vietnam. Over time, as the tool set availability improves, while the inventory accumulators will remain in place, the need for inventory buffers will decline.”

First Solar noted that the resulting constraints, although minimized eventually would lead to an annual shortfall of finished module output of around 200MW. 

This is due in part to the fact that the company would not have all of the buffers in place to hold un-finished module inventory for back-end completion at the end of the third quarter of 2018 or near the beginning of the fourth quarter of 2018, according to Widmar. 

“It is important to note that despite the 2018 volume reduction with the actions we are taking we anticipate to exit the year at the originally anticipated throughput levels and enter 2019 on track to our previously announced Series 6 production volume,” noted Widmar in the earnings call.

Editors note: A short 2 hour video by the guru of ‘Theory of Constraints’ is posted below. 

Read the entire story

Leading GW-plus module suppliers to non-China PV global markets

As module suppliers adapt to the slowdown of Chinese module demand in 2018 and 2019, global EPCs and developers are likely to see new Asian-produced panels being offered for both rooftop and ground-mount installations.

This issue was discussed in a recent PV-Tech blog last week, and forms a key theme of the topics and agenda during the forthcoming PV ModuleTech 2018 event, on 23-24 October 2018, in Penang, Malaysia.

This article reveals who the GW-plus module suppliers are to the global end-market, once we remove module supply to the domestic Chinese market, and identifies some of the chasing pack that are hoping to increase global brand awareness going into 2019.

Twelve module suppliers account for two-thirds of non-China global demand

While there remain hundreds of companies producing modules today, from regional single-production-line start-ups, to the multi-GW capacities of the Silicon Module Super League players, once we remove China market supply channels and all the low-volume suppliers (typically into a small subset of non-China markets), we are left with 12 major global suppliers. This list (shown alphabetically) is displayed below.

In fact, collectively these companies are likely to account for about two-thirds of global PV module installation capacity (excluding China) during 2018, with much of the supply being to utility-scale projects where company and technology are two critical issues that undergo various forms of risk-mitigation, auditing and bankability.

The list of 12 companies can be grouped and discussed here, to illustrate the different profiles and strategies for non-Chinese global module supply.

While having different technology offerings, First Solar and SunPower have relatively similar downstream-driven operations, with SunPower’s rooftop activities being the main differentiator.

Canadian Solar, JA Solar, JinkoSolar and Trina Solar can also be put together, with Canadian Solar having a more arms-length projects business that is not tied to using the company’s branded modules.
Risen shares similarities to the above China-HQ companies, but without its own non-Chinese manufacturing operations. Risen has downstream tactics similar to Canadian Solar.

LONGi and GCL-SI form the next group-of-two, being legacy upstream China-based poly/wafer suppliers, but now having multi-GW cell/module capacity with ambitious non-China module supply aspirations.

LG Electronics and Hanwha Q-CELLS form the Korean-run subset here, with both companies having been focused strongly in the past few years on being leading suppliers to rooftop and ground-mount segments in the US. Not surprisingly, both companies have also revealed module assembly capacity expansion plans in the US, hoping to benefit from the 2.5GW of tariff-free cell imports.

This leaves REC Solar somewhat in a grouping of its own, having a different place in the PV industry, compared to the above companies. While REC Solar has dabbled on and off with projects business operations over the past decade, it retains a somewhat European-run company, despite Asian ownership.

The chasing pack

There are over 100 module suppliers that would love to be included in the above list. Indeed, many of these companies were seen stepping up trade exhibition visibility in Europe and the US during the past couple of months, hoping to connect more with developers and EPCs outside China.

The strongest challengers to the above top-12 are listed now:

Talesun has the potential to be included within the Jinko/JA/Trina grouping above, due to multi-GW status (very close to SMSL inclusion) and non-China cell/module plant operations.

Neo Solar Power remains the surprise package for 2019 possibly, with the collective resources of UREC being potentially available, and at a time when NSP has been repositioning itself with project financing and site acquisition globally.

Note that we have purposely excluded companies (mostly in Southeast Asia) that have been used purely as OEM contract module suppliers.

PV ModuleTech 2018 to explain more about the top-12 market leaders

While strong growth outside China set to be the driver for end-market growth going forward, knowing more about the financial and technical strengths of the above-mentioned module suppliers will be critical to project developers and EPCs, when designing and building out new solar sites.

Even from a list of 12 or 17 however, there is plenty to learn and understand. While the above text offers a basic segmentation, there are many differences between how the companies operate outside China (or Korea), and in the product portfolios offered from each company.

The goal of PV ModuleTech this year is to provide a platform to help in this respect, and many of the module suppliers discussed here (such as First Solar and SunPower) will be explaining how their module offerings are both bankable and performance-leading, backed up with field performance data linked to manufacturing excellence.

To participate in the PV ModuleTech 2018 event in Penang on 23-24 October 2018, please click here for further information.

Read the entire story

How many Chinese module suppliers can compete with JinkoSolar, outside China?

Since Chinese investments into major cell and module facilities started – more than 10 years ago – success ultimately has been driven by overseas market-share gains, above other technical or financial benchmarks that otherwise would be expected.

However, despite the efforts of well over 50 Chinese cell/module makers (where the cut-off is including those with >500MW of capacity), only 5-6 Chinese companies have actually succeeded in establishing global operations with brand-recognition, and can point to a successful overseas business plan.

However, only one of them has evolved into a company that, in several respects, operates as a non-Chinese company outside China – JinkoSolar.

This article looks at where JinkoSolar is placed in the PV industry today, from both a Chinese and global standpoint, and asks the question: can other Chinese companies do what JinkoSolar has done, and if so, can they do it before their Chinese supply pipeline disappears?

Furthermore, is JinkoSolar now unfolding global module supply operations that are setting benchmarks for western-run module suppliers that rely on non-Chinese demand to exist?

Many of the themes running throughout the narrative below are set to be featured in more detail, at the forthcoming PV ModuleTech 2018 event on 23-24 October 2018, in Penang, Malaysia, where the top 5-6 leading global module suppliers will be outlining their company’s module technology, quality, and reliability status.

Resuming overseas module supply much harder than before for Chinese companies

During the last major rush to export modules from China (driven mainly by European FiT demand before 2012), the requirement for a global presence was minimal. This was seen clearly at the time, with Chinese operations being trade-show sized replicas of domestic personnel, and where sales to intermediates (distributors, installers, EPCs) was largely sufficient.

The Asian companies then that sought to have a global presence (Yingli Green, Suntech, Sharp, Kyocera, etc.) ended up challenged by ASP declines and lack of product differentiation, and with global operations that were based largely on copying the Europe third-party sales approach, aside from the occasional frenzy into mainstream product placement activity (read Yingli Green and the 2014 FIFA world cup in Brazil, for example).

As trade barriers crept in (US and Europe), about 90% of the Chinese companies fell back to domestic module operations, in part due to the imminent market pull from local demand, but also arising from lack of desire/cash to have any Southeast Asia manufacturing.

In effect the period of 2012-2016 saw the second major phase of supply-channel consolidation (the first being the eradication of European and Japanese majors on the global stage), and left the only global module suppliers of note, with Chinese HQ operations, being JinkoSolar, JA Solar, Canadian Solar and Trina Solar. By the end of this period, Hanwha Q-CELLS had effectively become a South Korean-run operations, while having legacy Chinese manufacturing arising from the Solarfun days.

Collectively, these four Chinese-run companies (Jinko, JA, Canadian, Trina) have many similarities when it comes to manufacturing operations (China plus one-other Southeast Asia), technologies (p-mono and/or p-multi, PERC migration plans), and business models (aside from Canadian’s downstream activities).

The other feature that differentiates them from the 100+ remaining cell/module makers in China, is related to sales/marketing outside China, and understanding the difference between a Chinese company doing business in China and outside. This remains the key factor determining success rates from the other Chinese companies overseas, regardless of the RMB-levels allocated to being on the global stage.

However, over the past 5-6 years, no Asian company (and very few globally) has been able to come close to the global module supply business strategy of JinkoSolar. In contrast to the other Asian companies that have got to global market-share leadership in the past, JinkoSolar has simply used this as a springboard to move to a different level, and into largely unchartered waters when it comes to a solar module company.

Why JinkoSolar?

This is a question asked by many people globally in the past few years. Indeed, if the answer was simple and prescriptive, others would have done it long before now, or be rolling out 2-3 year plans to get there by 2020/2021.

Of course, one part is being in the right place at the right time (read entering the industry post margin-crippling long-term polysilicon contracts) and being able to juggle tactic and strategy effectively (flexible in-house, OEM, contract supply balancing). But if we were to make a stab at the main reasons for JinkoSolar’s success, it may simply come down to these three points:

Creating non-Chinese operations (sales and marketing) that were geography, culture and sales-pipeline savvy, in exactly the correct global locations at any given time. What is being seen now with the company’s multi-GW pipeline of contractual deliveries is likely a direct result of this being in place. It is the closest thing yet to a Chinese company operating in a non-Chinese way, outside China.

Putting in place a (real) technology roadmap with R&D spending and line upgrades, at the multi-GW level. This is one of the reasons previous market-leaders struggled, as many had effectively bankrupt themselves in getting to number-one with a me-too product, and had no what-next plan that should have been to be a market-leader as well as a technology-leader.

Seeing the Chinese market (especially over the past 2-3 years) as low-priority, and being focused on being either number-one or number-two in all other major end-markets (regional or country-specific).
The first point above can’t be charted as a benchmarked metric (its way more subtle and strategy based), but the others can, so let’s have a look here to emphasize these two points. The graphs below illustrate these points clearly enough.

China end-market-check forcing Chinese companies to globalize, or cease operations

Ironically, the dilemma facing most of the Chinese based module suppliers today has arisen purely from the domestic China market. Almost all companies expanded capacities, production and module supply purely on the expectation of local supply being available.

Indeed, many had been doing this, without any overseas business, or at best having discontinued their focus on export revenue streams. Therefore, in the absence of local supply options, these companies have to export, or effectively cease to operate efficiently with adequate utilization rates.

This largely summarizes the situation many Chinese companies are faced with today: how do they quickly work out the markets to focus on overseas; how can they convince module buyers that they have product quality and reliability to be bankable?

For many Chinese module suppliers that were relying purely on domestic demand, the problem is even greater, in particular those companies that were supplying modules mostly to parent-owned EPCs, themselves undertaking build-out for the same parent entity. The requirement to have third-party factory auditing, module inspection and risk-assessment is clearly not at the level here, compared to shipping half way around the world to a non-Chinese investor.

These questions are going to form a key part of the forthcoming PV ModuleTech 2018 meeting in Penang in October. To allow for this, we have supplemented last year’s agenda with a new session where leading market analysts/developers from all key markets outside China will present on what utility demand will look like in these regions over the next few years and what this means for module suppliers.

Not just China module suppliers, others having to change tactics quickly

The shifting module supply landscape of 2019 is also having an impact on remaining module suppliers headquartered in Japan and South Korea, not to mention Taiwan and Vietnam/Thailand. For many of the module companies based in these countries, the challenge is not so much on meeting demand in their HQ-country (e.g. for markets such as Taiwan and South Korea that are government-created to support domestic sectors), but in working out which of the countries (outside of US and Europe) to focus on.

Perhaps across all these countries however, the what-next for Neo Solar Power (NSP) in Taiwan may be the most interesting to view in 2019. Assuming NSP emerges as the dominant partner in the newly-created consortium of NSP, Gintech and Solartech (with modules rebranded as URE), then NSP may finally make the cell-maker-to-module-supplier/developer transition, on the global stage.

Of course, NSP would face exactly the same issues in terms of creating global sales and marketing, and branding. The advantage NSP would have however is making this transition, having started from a technology (and cell) focused heritage. This is not dissimilar to the path taken by JA Solar and Hanwha Q-CELLS in recent years, and having technology credibility is never a bad starting point when changing operational focus.

PV ModuleTech 2018 to showcase global module supply leaders

Last year, at PV ModuleTech 2017, the audience heard from the most of the top-10 module suppliers to the PV industry, based purely on module supply volumes over the previous 12 months, including both within and outside China.

For PV ModuleTech 2018, the contributions from module suppliers are shifting mainly to the leading module suppliers to the non-China segment of the PV industry, including companies such as First Solar, SunPower, and selected others that are existing (or potential) global suppliers over the next 12-18 months (and the main Chinese companies expected to survive the imminent shakeout).

This select group of companies is likely to form the basis of many of the proposals that will be seen by utility-scale developers and EPCs over the next few years. Therefore, understanding who these companies are, what module technology types they have been supplying recently, what their module supply roadmaps look like, and how their panels have performed in the field so far (from third-party verification standpoints) is vital.

To get involved in PV ModuleTech 2018, on 23-24 October, in Penang, Malaysia, please follow the links at the event website here.

Read the entire story

New module suppliers and technologies to create more opportunities and risks to developers and EPCs

Module selection for utility-scale solar sites in 2019 is likely to see the widespread availability of higher performance products with average selling prices significantly lower than witnessed over the past 12-18 months.

While on the surface, this may appear as a dream-come-true for project developers and EPCs (especially outside China), the challenges in identifying bankable suppliers with quality product offerings are set to increase dramatically, placing far more pressure on making the correct decision to mitigate against risk of plant underperformance over a 20-30 year operating lifetime.

This article explains the background to these imminent changes to module supply, including an outline of the module suppliers’ landscape for 2019, while also identifying the most challenging criteria for developers and EPCs in terms of module supplier and technology-type selection.

The discussion below on module technology (and supplier) selection is perhaps the key takeaway for EPCs and developers, providing top-level selection criteria on modules for 2019 and the areas where increased scrutiny will be required.

Data shown here is taken from PV Tech’s Market Research analysis, included in the May 2018 release of the PV Manufacturing & Technology Quarterly report.

Reference is also made to the themes set to be covered in the forthcoming PV ModuleTech 2018 meeting on 23-24 October 2018, in Penang, Malaysia, where module supplier benchmarking is addressed in detail.

72-cell multi modules met utility demand with few questions asked

Until the end of 2017, the PV industry had depended critically on the availability of p-type multi crystalline silicon (p-multi c-Si) modules somewhat as the default go-to choice for utility-scale deployment by many developers and EPCs.

The main exceptions were projects that used First Solar’s Series 4 thin-film panels, or SunPower’s E-Series c-Si premium (n-type back-contact) panels.

An uptick has been seen in the use of p-type mono c-Si panels over the past few years, with much of the deployment here being within China, using modules where the entire value-chain of manufacturing (poly/ingot/wafer/cell/module) has been domestic.

Otherwise, for the (non-China) global developers and EPCs, module selection has been mostly about which supplier of 72-cell p-type multi modules is successful, and less so on the technology type. It is this category of developers/EPCs that needs to be most aware of the changes set to unfold regarding module technologies (mono driven) and the new paradigm of ASPs on offer.

Why has multi been so dominant for utility-scale solar?

The answer to the subheading above is not complicated. Utility solar has been the main driver of growth in the solar industry in the past 10 years, and this has been accompanied by limited supply of mono-grown ingots (needed for mono wafers used to make mono cells/modules).

In fact, capacity and production levels of multi c-Si wafers and cells has rarely been in short supply, maintaining widespread low-cost product availability for developers and EPCs. Barriers-to-entry in making multi wafers have been low, and having GCL-Poly setting up tens of GW of low-cost/low-price wafers set the benchmark for every other multi wafer supplier in China/Taiwan.

Had the world decided collectively overnight that the choice of utility-scale solar was to be confined to First Solar, SunPower or p-type mono modules, then it would have seen more than 80% of utility-scale solar being removed, due entirely to lack of module availability from this subset.

Developers and EPCs have needed 72-cell p-type multi modules to exist. And many of them have been largely unconcerned with origin-of-manufacture, and whether the maker of the product was in Vietnam or Thailand or chosen ad-hoc from one of China’s many quasi-OEM state-funded institutions.

Pricing was low, often serving the primary goal of lowering site capex and maximising profits when flipping signed-off accredited/PPA-secured plants. Product was available in droves, even during the various trade-related cases in recent memory.

If module suppliers and their respective technologies had been created equal, then this would be the end of the story. However, the raft of underperforming solar plants globally today indicates this is clearly not the case, and anyone thinking that solar modules are a commodity offering needs to spend a few hours with asset owners and O&Ms to get a strong reality check.

For most of last year, for example, looking at many of the 72-cell modules installed globally (or at least reading off the datasheets) showed few if any differences across 50-100 module suppliers, differing only in frame dimensions.

It is therefore little surprise that one of the most frequent questions asked has been: which is the best module supplier – who should I buy from?

If EPCs and developers thought life was tricky in the past few years (having to select which company for 72-cell p-multi modules!), then they are in for a rude awakening by the end of this year, unless they become far more educated in what the GW-scale module supply landscape is set to become shortly. These changes are set to offer major opportunities for plant design, but come with an equal dose of risk should the wrong supplier or technology-type be deployed.

Mono, n-type variants, bifaciality, and Series 6 thin-film panels

Going into 2019, an increasing number of utility solar farms are going to be utilizing p-type mono modules (almost all of which will be PERC based), with ASPs largely at parity with p-type multi offerings and at sub 30c/W (EXW) pricing.

While design and operation of solar farms will reflect the higher powers from p-type mono, the main question for module users is likely to come from the increased number of Chinese based suppliers, and which company to choose for site deliveries.

By now, it is no great secret that the Chinese market is not going to keep growing exponentially, simply to absorb the collective shipment targets of companies that have added capacity (from polysilicon through to modules) over the past couple of years. This single fact will see approximately 20-30 Chinese-based module suppliers seeking to grow export business, adding to the 10-20 existing Chinese companies that have appreciable overseas sales revenues today.

Which of the 30-50 Chinese module suppliers are truly bankable? How many of these companies have a level of manufacturing quality control that is low enough risk for external solar farm investors? Are their PERC modules reliable, with a fully-audited bill-of-materials?

But perhaps more pertinent, how many of these companies will be solvent 2-3 years out and able to honour 20-30 year performance guarantees?

In addition simply to increased p-type mono modules (72-cell PERC), there will be more offers for n-type modules than seen before. Several caveats apply here, as n-type now includes a wide range of company types and performance levels, not to mention strategies.

In theory n-type modules have the capability of higher efficiencies (power ratings) and superior elevated temperature operation, compared to p-type mono and multi modules, as explained below. EPCs and developers should at a minimum absorb these basic facts.

There are three basic types of n-type modules: n-type PERT (using manufacturing processes and equipment closely aligned with p-type mono PERC), n-type heterojunction (HJT), and n-type back-contact (or often assigned as interdigitated back-contact or IBC).

Currently, the efficiencies (STC power ratings) from n-type PERT modules are not that different from best-in-class p-type mono PERC variants, suggesting that the value-added proposition for n-type PERT lies mainly in the temperature coefficients.

At the top-end, LG Electronics version sits as the gold standard today, with several in-house differentiators (front/rear implanting, multi-wire front grid interconnections): however, LG’s priority (excluding its domestic market and a few isolated occurrences) is mainly on rooftops, and not mega-solar ground-mount deployment. Therefore, many global utility-based developers and EPCs will remain somewhat excluded from using these panels still.

Other n-type PERT modules are coming now from new Chinese companies, none of which has any strong heritage in solar cell production (aside from Yingli Green’s legacy PANDA lines). All of these companies have plans to export in volume to global utility projects in 2019, and this certainly points to a greater awareness of module users when it comes to qualifying these suppliers.

The next n-type architecture seeing increased attention is heterojunction. For years, HJT modules were synonymous with Sanyo’s trademarked ‘HIT’ modules, simply rebranded as Panasonic following the acquisition phase. HJT production is fundamentally different to all p-type cell/module assembly, and to n-type PERT variants. These lines are process and equipment tool specific and represent a step-up in terms of manufacturing complexity.

In the past 2-3 years, it is mostly HJT that has seen the investment dollars in China across new entrants in cell/module production that needed to select an advanced technology-type to differentiate from the p-type juggernaut that is already in operation. HJT has also been a convenient technology-transfer route for a number of a-Si thin-film fabs (extending outside China also) that had adaptable deposition equipment/know-how needed to make c-Si HJT cells.

In an ideal world, HJT deployment sits firmly on residential rooftops, commanding ASP premiums that are needed to absorb higher material and production costs compared to the leading multi-GW p-type suppliers today. While LG and Panasonic have been able to manage this mostly during their solar industry participation, these companies benefit from brand association and installer confidence levels built up over many years.

New HJT suppliers are therefore left to focus on ground/utility segments, suggesting again that global EPCs/developers are likely to see new offers from this grouping during 2019, as GW levels of production hits the market with no easy supply channels in China to absorb all produce.

The final n-type (and the most advanced and premium performance) category is based on the back-contact or IBC structure, something that (aside from pilot-line activity in Korea) is the sole domain of SunPower across its Southeast Asia fab operations. This is no indication at all that this will change in 2019 or the foreseeable future, such is the barrier-to-entry level in terms of in-house IP (process and equipment tooling based) that SunPower has meticulously crafted over 20 years of production experience.

This ensures that SunPower’s products will remain the highest-performing (STC and elevated temperatures) and with the highest ASPs in the industry, largely without technology-specific competition. This continues to allow SunPower to be selective (for its IBC product lines) in terms of application segments (residential, C&I and utility split), regional deployment, and shipment to in-house or third-party sales. Applying these factors, and given the GW level of IBC product available (compared to the 10GW mark now common to several of the global p-type module leaders), one can conclude that most of the global EPCs/developers will continue to be forced to make choices confined to the other module technology types.

The final technology type to consider of course is thin-film, with this technology belonging to one company today – First Solar. Yes, there are other thin-film companies in the solar sector still, but they are either seeing declining fortunes in terms of product availability and global competitiveness (e.g. Solar Frontier) or have limited if any market credibility or bankability (e.g. almost all Chinese based thin-film investments going back well over 10 years now).

Manufacturing exclusivity aside however, First Solar’s successful roll-out of its Series 6 panels (coupled with multi-GW of new fab builds across three countries now) is set to provide higher performance products in far greater volumes than seen in the past. However, this is not simply confined to product coming off the production lines, but the amount that is now being shipped or sold to third-party developers and EPCs.

To put this into context, during 2019, MW-levels of shipments to third-party developers/EPCs are forecast to increase by approximately one order of magnitude, compared to third-party shipments just 4-5 years ago, and could easily exceed the 4GW mark next year.

Given also that First Solar’s product is – for all purposes – utility-segment specific, this effectively propels First Solar to a new place in the PV industry, and will bring a whole bunch of new EPCs/developers into contact with thin-film products for the first time, or simply re-engage those that had been champions of thin-film panels for utility deployment in the past but were forced to rely on p-type multi-modules to fulfil build-out plans over the past few years.

What does this really mean for EPCs and developers?

In short, the companies and product types being considered for utility-scale solar in 2019 will be different to what has been seen for most of the past 2-3 years (where almost everything was 72-cell p-type multi).

Many of these companies have minimal track-records in exporting supply out of China, some are new to cell/module production and are trying to ramp up GW-levels of new process flows for the first time, while others (JinkoSolar, Canadian Solar, JA Solar, First Solar, for example) are firmly established with global EPCs/developers and will release new module versions with improved performance and reliability.

It is probably fair to say that any developer/EPC currently planning a utility-based solar site for 2019 based on 72-cell p-type multi-modules should pause, and ask whether this is the best option in terms of investment ROI or secondary site valuation figures in three years from now.

While not wanting to complicate the issues, seasoned campaigners may well be asking why there is no talk above of bifacial/half-cut/shingles. The reason for this is relatively simple.

Right now, these are still options, not necessities in the market. The benefits are not in doubt. It is just that they are more additive to existing plans (which are not yet fully implemented and qualified in production).

Bifacial remains a curiosity more than a must-have, with widespread confusion about just what is on offer in terms of yield deltas, and how it is possible at all to predict performance over 20-30 years. While the easy argument is to say that anything extra is always good, this is as dangerous a statement for investors and O&M’s as is anything impacting site underperformance.

How do you value the worth of a site, if you can’t forecast yield over 20 years? How can you set performance ratios for O&Ms or even dare to include upside payments based on over-performance relative to a fixed (unknown) reference line?

Either you bit the bullet and have some carefully-worded clauses into supply arrangements (power guarantees) or have highly-flexible O&M contracts (especially during the first 2-3 years), while baseline parameters (likely almost all site/environment specific) are established. Or you just wait a few years until the industry has worked out how to deal with double-sided absorption from solar modules.

Half-cut modules are less of an unknown or a differentiation and EPCs/developers have less reasons to fear them, other than diving into module supplier selection from an unknown entity. Until now, half-cut cell module design has been a key focus from REC Solar: it is unclear still how much the Chinese sector will fully embrace. Will China solar want to laser cut all its cells in two and re-assemble them all across its various cell/module supply-chains? Taking this one step further to multi-cut – or singulated cells designs – and potentially we enter more niche-status manufacturing today.

PV ModuleTech 2018 to provide clarity for EPCs and developers

Going into its second year, PV ModuleTech 2018 will focus specifically on utility-scale module supply and demand for the 2019/2020 period, in particular for all countries/regions outside China.

Therefore, this two-day conference should provide EPCs and developers with the tools they need to assess and benchmark module suppliers and product technologies for sites in preparation or going into planning/approval phases over the next 12 months.

The event will include a non-China specific geographic module supply session, where the demand for modules outside China in 2019/2020 will be explained, including company and technology market-shares in key regions globally today and how this may change going forward.

Leading module suppliers will then outline product availability, and what measures are in place to ensure bankability and product quality, and how these companies are placed to honour 20-30 year performance guarantees.

PV ModuleTech 2018 will again hear from leading independent engineers, test and inspection organizations, certification labs, factory auditors, and module assembly materials and equipment suppliers.

Findings from the event will be invaluable to companies (EPCs, investors, developers, O&Ms) whose business models are contingent on the correct module type and supplier being chosen.

Details on how to attend PV ModuleTech 2019 can be found here, including the event agenda. Speakers are by invite-only from the PV Tech team, and will be revealed by us in the coming weeks.

Read the entire story

Solar manufacturing capacity expansion announcements in Q1 2018, reached 24.8GW

PV manufacturing capacity expansion announcements in the first quarter of 2018 continued to follow the strong trend set in the fourth quarter of 2017. The quarter also represented a revival in thin-film expansion plans as well as the return of PV module assembly outpacing solar cell announcements. Also notable was the return of India and the US as major destinations for new capacity plans.

January review

Total expansion announcements were 11,450MW, down from 16,100MW in December 2017 and down from 20,800MW in November 2017. However, January 2018 was responsible for setting a new ‘mini’ record for capacity expansion announcements, compared to other January months, since the beginning of 2014. 

The majority of expansion plans came from the PV module assembly segment, which topped 8,600MW. Only one month (November 2015) had exceeded this figure when 11,180MW of module assembly plans were announced. 

Solar cell expansion plans in January 2018 were 2,850MW, down significantly from 7,350MW in the previous month and down substantially from 20,000MW announced in November 2017, which set a new record for monthly solar cell expansion announcements. 

Therefore it should be a surprise that after just two months when a total of 27,000MW of new cell capacity plans had been announced, January would experience further declines. In fact, 2017 stands out for breaking the trend since 2014 that solar cell expansion plans closely tracked those of module assembly. However, solar cell expansion in 2017 accounted for more than 65% of the total. 

Notable announcements included LONGi Green Energy Technology and newly created UREC, a joint venture consolidation of Taiwan-based PV manufacturers Gintech Energy, Solartech Energy, Neo Solar Power. 

As a ‘Silicon Module Super League’ (SMSL) member, we will cover LONGi later in the SMSL review but the company accounted for 5,000MW of module assembly expansion plans in January in China and a further 1,000MW of cell and module plans in India. 

Soon after its formation, UREC was cited in media reports as being interested in establishing cell and module manufacturing operations in the US, post the Section 201 trade case, as high tariffs were imposed. Some reports indicated a 500MW to 1,000MW nameplate capacity that could be implemented in phases. 

Other notable plans included the possible expansion at Photowatt, a subsidiary of EDF Energies Nouvelles in France to meet the growing French Government tenders and in-house projects with effectively a 450MW module assembly expansion using mono-cast wafers and possibly a JV involvement from SMSL member, Canadian Solar. 

Leading SMSL JinkoSolar also confirmed plans to build a highly automated module assembly plant in Jacksonville, Florida, USA, post the Section 201 trade case.

February review

The month of February was in total contrast to the previous month as only a combined total of 850MW of new expansion plans were announced. 

Only 500MW of solar cell expansions were announced coupled to only 350MW of module assembly. 

Notable was a proposed 150MW module assembly expansion at Recom-Sillia in France and plans by Mission Solar in San Antonio, Texas, USA to double module assembly capacity to 400MW, which was after the Section 201 trade case tariff decision. 

Although February announcements did not top 1,000MW, 2017 was notable for having two months (August and September) when announcements did not reach 1,000MW. 

March review

After the collapse in announcements in February, March bounced back stronger than January, accounting for a combined total of 12,570MW of new cell, module assembly and thin-film expansion plans. 

Indeed, March 2018 was the second highest for announcements since 2014, the highest March so far record was in 2016 (13,325MW). 

Once again module assembly announcements led the way, totalling 6,620MW, compared to 3,810MW of solar cell expansion plans. 

However, thin film module expansions, primarily CIGS (Copper-Indium-Gallium-Selenide) from Hanergy Thin film Power Group, totalled 2,140MW. 

Hanergy would seem to have created a completely new business model in 2017 that provides new industrial parks a selection of portfolio of a-Si, CIGS, GaAs and c-Si heterojunction (HJ) turnkey production lines to provide local government bodies access to solar technology and attract other hi-tech companies to new industrial parks.

The company announced for the first time in its 2017 annual report, issued at the end of March that it had already secured contracts from three newly formed industrial park project companies in Mianyang Sichuan, Datong Shanxi and Zibo Shandong, who had purchased thin-film production lines from the company valued at approximately RMB 11.3 billion (US$1.79 billion). 

Unrelated to the industrial park business model, Hanergy also highlighted a contract signed in October 2017 with Huafengyuan (Chengdu) New Energy Technology Co.,Ltd., for the purchase of 600MW of nameplate capacity of automated and integrated ‘High Efficiency Silicon heterojunction (SHJ) solar cell’ production lines and technology transfer, valued at RMB 1,39 billion (US$222.5 million) and RMB 175.9 million (US$27.9 million), respectively.

Hanergy TF noted that it had delivered the equipment for the first 120MW production line during 2017, with an advance from the customer of US$4.05 million. Total equipment orders outlined in its annual reported reached 2,740MW. 

India was also notable for around 28 dedicated PV module assembly firms planning small-scale expansions that reached around 4,000MW, while several cell and module producers planned a total of over 200MW of cell expansions. 

Leading SMSL JinkoSolar also provided expansion plan updates totalling 2,500MW in March on top of the 400MW module assembly plans for the US, which were announced in January, 2018. 

Quarterly review

The first quarter of 2018 is almost identical to the first quarter of 2017. Combined capacity expansion announcements reached 24,879MW, compared to 24,745MW in the first quarter of 2017. 

However, the Q1 2018 breakout by segment is more biased to module assembly expansion plans, while the Q1 2017 bias was towards solar cell expansions. 

A key trend consistent from the beginning of 2014 has been that the first quarter of each year has been strong for expansion plans and in the last three years exceeded or came close to reaching total combined announcements of 25,000MW. 

In Q1 2018, module assembly capacity expansion plans topped 15,570MW, the second highest on record, only exceeded in the first quarter of 2016 when plans announced topped 16,000MW. 

Thin film activity increased quarter-on-quarter, due solely to Hanergy and totalled 2,140MW in the quarter, up from 1,200MW in Q4 2017. 

Geographical review

On geographical basis, Q1 2018 replicated the return of China as the number destination for new capacity expansion announcements seen in 2017. China accounted for a combined segment total of 14,240MW in Q1 2018, or 61% of the total. China accounted for over 71,000MW in 2017, or 73% of the combined segment total.  

However, Q1 2018 saw the re-emergence of India as the second largest destination for planned expansions. India accounted for 6,210MW in the quarter, or 27% of the combined segment total.  

As already noted, plans from domestic module assembly companies totalling around 4,000MW were a key driver, while China-based LONGi announced 1,000MW of both solar cell and module assembly plants to be built in the country. 

The resurgence of India is believed to be driven by threats of anti-dumping duties in India as well as momentum building, despite challenges in the downstream utility-scale sector. Indeed, with the US imposing further anti-dumping duties in early 2018, India becomes even more important to PV manufacturers located in China. 

As previously noted in these reports, PV manufacturing capacity expansion announcements in India have proved significantly difficult to translate into ‘effective’ capacity. 

In 2014, expansion plans totalling over 1,400MW were announced for India, which increased significantly in 2015 to 7,850MW, peaking at just over 17,000MW in 2016. Planned expansions in India collapsed to only 2,790MW in 2017. 

In total, planned expansions in India since 2014 to the end of 2017 had reached over 29,000MW.

In contrast, the total of planned expansions in India that have translated into effective new capacity since the beginning of 2014 is around 4,500MW, which includes around 1,700MW of new effective cell capacity and around 2,750MW of new effective module assembly capacity. 

However, adding to the challenges in developing effective new capacity in India are the low utilisation rates of existing manufacturing facilities.

SMSL update

Typically, in the first quarter, the majority of SMSL members (JinkoSolar, Trina Solar, Canadian Solar, JA Solar, Hanwha Q CELLS, LONGi Group, GCL Group), provide capacity expansion updates when releasing fourth quarter and full-year financial results.  

However, at the time of this report only JinkoSolar, Canadian Solar, Hanwha Q CELLS and LONGi Group have provided updates. Since going private, Trina Solar has not provided updated information on capacity expansion plans.


Leading SMSL member JinkoSolar is planning further capacity expansions across wafer, cell and module assembly in 2018, including a module assembly plant in the US, after strong capital expenditures in 2017 that totalled US$480 million.

The SMSL reported that in-house wafer capacity went from 5GW in 2016 to 8GW in 2017, a 3GW increase, year-on-year, while solar cell capacity increased by 1GW in 2017, reaching 5GW. 

Module assembly capacity was said to have increased from 6.5GW in 2016 to 8GW in 2017, a 1.5GW increase, year-on-year. 

In 2018, JinkoSolar has set plans to add 1GW of in-house wafer capacity in the first quarter, bringing total nameplate capacity to 9GW. By the end of the year a further 500MW expansion of wafer capacity is expected. 

The company is also adding a further 1GW of solar cell capacity through the year, bringing in-house nameplate capacity to 6GW by year-end, while in-house module assembly capacity is being expanded by a further 1.5GW in 2018. This includes a 500MW increase in the first quarter of 2018 and therefore a further 1GW by year-end. Total module capacity is therefore expected to reach 10GW in 2018.

The difference between 2017 and 2018 expansions, apart from a slowdown in wafer capacity expansion plans, is the establishment of a 400MW module assembly plant in Florida. 

Canadian Solar

Third ranked SMSL, Canadian Solar surprised by announcing a slowdown in capacity expansions and lower nameplate capacity plans than given in late 2017. Having adjusted manufacturing capacity expansions throughout 2017, Canadian Solar continued to tweak plans for 2018.

The SMSL noted that its wafer manufacturing capacity at the end of 2017 stood at 5.0GW, a 3GW increase from 2016. However, the company has not announced new wafer capacity expansions for 2018, keeping capacity as 5GW. 

Solar cell manufacturing capacity stood at 5.45 GW at the end of 2017, up from 2.44GW in 2016, in-line with previous upwardly revised guidance.

However, Canadian Solar has revised its cell capacity expansion plans again, noting that it expected nameplate cell capacity to reach 5.6GW by mid-2018, compared to 6.20GW in its previous update. The SMSL also noted that cell capacity at the end of 2018 was expected to reach 6.35GW, compared to previous guidance of reaching 6.95GW. 

A similar adjustment has been made to in-house module assembly capacity expansion plans. The SMSL noted module capacity reached 8.11GW by the end of 2017, up from 6.17GW in 2016. The company said that module nameplate capacity was expected to reach 8.31GW by mid-2018, compared to its last update of reaching 9.06GW in that time frame. 

Total module assembly capacity by the end of 2018 is targeted at 9.81GW, compared to 10.31GW guidance, previously given. Canadian Solar has not issued its annual report and therefore has yet to disclose capex figures for 2017. 

Canadian Solar’s management noted that it had recently experienced under-utilization rates at its module assembly plant in Canada and its manufacturing plant in South East Asia, due to the Section 201 tariff decisions by the US government.

Hanwha Q CELLS

The fifth ranked SMSL, Hanwha Q CELLS had already restricted capital expenditure throughout 2017, all except for a the JV manufacturing plant in Turkey planned in response to building a 1.3GW (DC) PV power plant in the country which is expected to be operational in 2021. 

The company had previously guided capital expenditures in 2017 to be around US$50 million, which would be allocated to manufacturing technology upgrades and certain R&D related expenditures. However, the SMSL’s capex in 2017 was US$66.1 million, while R&D expenditure was down 51.2% to US$24 million, compared to US$49.2 million in 2016. 

The SMSL had an in-house name plate capacity of 4,300MW for solar cells and modules at the end of 2017, unchanged from the previous year.  

In 2018, Hanwha Q CELLS expects a slight increase in capex, due to initial spending on its new integrated manufacturing operations in Turkey. The company guided capex of US$90 million in 2018 and an allocation of around US$37 million to the new plant in Turkey. 

In early April, so technically outside the scope of this report, the SMSL reported a fourth quarter loss of US$50.5 million, primarily attributed to the asset write down of its entire wafering operations, which were based at dedicated facilities in Lianyungang, Jiangsu Province, China. 

The company had multicrystalline ingot nameplate capacity of 1,550MW and 950MW of multicrystalline wafer capacity. The SMSL cited that the wafering operations were unprofitable as well being impacted by a downward trend in wafer prices.

However, the JV in Turkey requires Hanwha Q CELLS to establish wafering operations not just solar cell and module assembly to comply with the downstream project tender win. 

LONGi Group

Leading integrated high-efficiency monocrystalline module manufacturer and seventh ranked SMSL member LONGi Green Energy Technology via is subsidiary LONGi Solar (cell/module) manufacturer had executed on an aggressive capacity expansion strategy in 2017. 

Mono wafer capacity went from 7,500MW in 2016 to 12,000MW by the end of 2017, a 60% increase, year-on-year. Mono solar cell capacity went from 2,500MW in 2016 to 5,000MW by the end of 2017, a 100% increase, compared to the prior year. 

However, module assembly capacity increased at a relatively lower pace, going from 5,000MW to 6,500MW by the end of 2017, a 30% increase, year-on-year. 

The company announced in Q1 2018 that mono wafer capacity would be expanded to 28,000MW by the end of 2018, more than a 133% increase over the previous year. 

Although solar cell capacity is expected to remain at 5,000MW, LONGi will expand mono module assembly to 8,000MW by the end of 2018, a 60% increase, year-on-year. 

However, separate to the expansion cited, LONGi Group announced in the first quarter of 2018 that it would invest US$309 million, including around US$240 million in constructing a new facility in Andhra Pradesh, India, with an initial nameplate capacity of 1,000MW of monocrystalline solar cells and expand its mothballed 500MW module assembly plant (previously announced) to 1GW. 

The new solar cell facility is expected to be operational in January 2020, while the expanded module assembly plant plans are expected to be completed and production ramp occur by the end of August 2019.


The capacity expansion announcements in the first quarter of 2018 remained strong, driven primarily by China and India on the back of domestic downstream demand. The US benefited from Section 201 tariffs, but only in respect to module assembly expansions and relatively small new assembly plant plans from the leading SMSL. 

A significant increase in thin-film expansion plans, specifically in China, driven by one company, Hanergy, provided a surprise revival, notably for CIGS technology. 

The return of module assembly announcements that far outpaced those of dedicated solar cell expansion plans was a key highlight. 

Read the entire story

Bizarre demand forecasting for 2018 overshadows real impact of China’s confused investment strategy

The past few weeks has seen some of the most dramatic knee-jerk, naïve and misinformed PV market reporting seen in recent times, with the headlines often resembling nothing more than tabloid sensationalism.

In fact, for most of 2018, the confusion across the board within the market research community has been evident, perhaps reflecting the gap today between those that ship the product and those that seek to track from afar on a spreadsheet.

This article discusses why the fascination on these tabloid rumblings is creating a false narrative that is neglecting the fundamental issues today in the PV industry. The sections covered below seek to highlight the real issues that need to be considered to fully understand manufacturing supply-chain supply/demand, cost/pricing and technology progress at the shipped module level.

I start by stepping back from the May madness post-SNEC, and put the first half of PV activity in 2018 into perspective. This takes us neatly onto the topic of China’s investment time-bomb that has been brewing for about 3 years but is now ready to create changes, ending with what can realistically be expected with cost and pricing at the end of 2018 and the technology-question of the day (mono versus multi).

The narrative here is very much based on qualitative reasoning, with the focus on trying to explain clearly in words what is going on globally. The quantitative argumentation behind the themes and conclusions will be covered in a PV-Tech webinar I will deliver online on 27-28 June. Details to register to hear the webinars can be found here.

PV forecasting is a thankless task, at the best of times

Forecasting global PV demand has long been challenging, largely due to the sheer volume of companies shipping modules to the end-market. The PV industry continues to have a highly fragmented route-to-market for modules, with the top-20 module suppliers making up about two-thirds of demand of whom only a few deliver guidance that can be tracked with any accuracy. It creates a virtually impossible task for outside observers.

The default route for forecasting demand has routinely been to look at country-specific PV policies, auctions or long-term aspirations. However, with many of these being government advocated, it then places supreme trust in the will of the policy makers coming to fruition.

Markets that are attractive and uncapped tend to be flooded with product, and compared to all other new energy sources, PV can be deployed at a phenomenal rate. Almost always, it takes policy makers months to react to unexpected exploitation by return-driven investors, by which time installations can vastly exceed original forecasts.

These issues are largely the main reasons why demand forecasting has been so hard to undertake until now in the PV industry. In the past couple of years though, it has hit new heights as the culprit in this case has been a country accounting for half of global demand – China. And for the whole of 2018, it is China that has single-handedly created the confusion in terms of forecasting global demand, and will continue to do so in the short-term.

Before May 2018, it is fair to say that no-one appeared to have a good handle on what demand was going to be in China this year. Forecasts ranged from 30GW to about 70GW, which is such a massive range that the global numbers for 2018, from 90GW to 130GW, simply emphasized that the current means of market forecasting are perhaps needing an overhaul.

Now, when we factor in the announcement from China’s National Development and Reform Commission, the Ministry of Finance and the National Energy Administration outlined in the “2018 Solar PV Power Generation Notice” – suggesting the real possibility of imminent deployment caps, a reduction to domestic feed-in tariffs rates, and setting rules at the central government level for utility-scale projects – then we definitely have the basis for confusion on a level not seen before.

Within 48 hours, 10-30GW was slashed from previous 2018 demand forecasts, with disastrous consequences being advocated that would ‘shake-out’, ‘consolidate’ and restructure the PV industry, as millions of workers in China were sent back to their villages as factory production at hundreds of PV plants ground to a halt.

It almost suggested the industry had been in a harmonious supply/demand existence, and this dramatic press release from China had come out of nowhere, and the future would never be the same again.

How wrong, naïve and misleading these summations may end up in just a few months from now. In the next section, I start to explain why.

Arrival of the next downturn has been unsubtle and depressingly inevitable

When annual deployment of PV in China hit the 10GW mark several years ago, analysts across the board concluded that this was an upper limit to demand in China and drew a flat line with a ruler going forward forecasting a stable 10GW per year in China almost ad infinitum.

The reality was a 10GW to 50GW domestic market surge that fuelled Chinese manufacturing investments from dozens of companies spanning the polysilicon to module value-chain stages, and even saw new entrants keen to jump on the bandwagon with multi-GW factory additions.

By the end of 2017, it seemed to reach feverous proportions, with new 50,000 MT polysilicon factories being accelerated, cell makers proclaiming 3-year plans to add tens of GW of new capacity, and ingot pulling plans that were destined to go ahead, come what may.

The question was not if this would create a problem for manufacturing, but simply when.

Behind the optimism was often the justification to come out of a shakeout as a top-5 player in China, where the government would then assign favoured-status and secured investments would be bankrolled at zero-risk. Sadly we have been here before. The only difference today is that the 1GW threshold has been replaced by 10GW, as a metric for long-term competitiveness.

If such aspirations were restricted to just a few players, then maybe things would not have been so bad. But when you have dozens of Chinese companies across the c-Si value-chain wanting to move from 1GW to 5GW, 5GW to 10GW, and 10GW to 20GW, there is clearly a problem brewing that will inevitably come to a head. (Add in the capacity being largely me-too and not technology differentiated and we have further problems on hand.)

The timing of when this juggernaut of capacity expansion investments would result in chaos was largely governed by the point at which China finally realised that it had a supply-driven deployment environment that was positioning to move the 50GW annual domestic number to 100GW during 2019.

In the absence of government intervention, it would simply have self-imploded, by virtue of module supply levels growing faster than developers and EPCs could cope with. That tipping point was likely to arrive in the second half of 2018 regardless of the May announcements from China concerning policy adjustments.

Domestic China supply/demand has been walking on a tightrope for the past 18 months, supported only by the lack of raw material to make wafers and cells. Once polysilicon hits oversupply, we get to the point that downstream parts of the value-chain collapse with the demand simply not being there to absorb modules at the rate they are being produced.

The net result is downturn, and the basic factors are no different to what we saw in 2012, except we have moved from problems at the GW-level (back in 2012/2013) to ones at 10GW or more.

What does a 2018/2019 downturn look like?

When the industry had its 2012/2013 downturn, capex was hit almost instantly with equipment suppliers seeing new order intakes reduced instantly, manufacturing gross margins moved from 10-20% to 0-5%, and module pricing gravitated to a baseline level that allowed loss making for the market leaders to be absorbed without devastating consequences.

Should we expect anything different this time? Probably not.

However, perhaps the most relevant metric to explore here is where baseline module ASPs end up. This is not actually too difficult to do, and I will attempt to undertake this now.

The supply tightness at the polysilicon level, coupled with the market dominance of leading multi and mono wafer suppliers in China, over the past couple of years, has resulted in reported accounts from the leading upstream (poly/wafer) Chinese companies claiming gross margins ranging anywhere from 20 to 40%, or even higher on a quarterly basis.

In fact, the imbalance of profits seen at the polysilicon and wafer level within China has been a source of annoyance for cell/module producers for some time. Why should they have 40% gross margin, when cells and modules are generally single-digit? It has only been a matter of time before Chinese upstream poly/wafer margins were eroded, and we could be at that point now.

So, in looking at the near-term baseline module ASPs, the first thing we need to do is take down $/kg poly ASPs and per-piece wafer prices, such that the major players (not the loss making second tier producers) have gross margins eroded to the 5% level. Then we assume that cell and module production continues somewhat as a zero-sum-game, with the likes of JinkoSolar and Canadian Solar being able to command brand/performance premium levels a few percentage points above this.

This exercise takes about 15 minutes on the back of an envelope, but is potentially remarkably accurate in the grand scheme of things. So, here goes….

For reasons that will emerge in the following section of this article, it is actually more useful to adopt the silicon and non-silicon cost breakdown segmentation, when looking at overall module costs and resulting ASPs (as close as we can to a full COGS type analysis that has some form of commonality with accepted Chinese and western accounting practices).

Taking polysilicon ASPs to 10% above the best-in-class costs (cash plus depreciation) would flatline poly at about 10$/kg. Driving down wafer margins in China (again across the likes of GCL and LONGi) would then result in wafer ASPs sub-10c/W. Adding in realistic cell and module costs, and we end up with silicon costs at sub-5c/W and non-silicon costs at 19c/W, resulting in all-in (blended) module production costs at 24c/W.

Assuming then that module makers are posting 5% gross margins, we end up with a module ASP baseline of about 25-26c/W. This then has multi modules selling at the low 20c/W level (potentially heralding the first news of sub-20c/W sales from ad-hoc fire sales), mainstream mono PERC modules at say 27c/W and premium brand p-type mono-PERC modules at 28-29c/W.

Scaling rooftop and n-type variants is then performed purely on a pro-rated basis, with virtually nothing selling anymore north of 40c/W before long.

If it turns out that loss-making is more prevalent across greater parts of the value-chain, then we can possibly take another 5-10% off the above ASPs, but this would be a rather depressed sector getting saddled with debt and having to focus on a further round of draconian supply-chain cost reductions to return to break-even operating status.

The main question is when this occurs, and this will ultimately depend on what happens in China – on both supply and demand fronts. It may be late 2018, or delayed into 2019, but is does seem a path now on module ASP erosion that is simply unavoidable.

Who are the winners in the next downturn solar world?

The natural conclusion is to state that developers, EPCs, investors and asset owners will benefit the most, not to forget the humble private homeowner of course. This has been widely proclaimed by media outlets, almost like some kind of eureka-moment in time.

It is not rocket science, but there are secondary issues that impact on new developments and project financing when manufacturing stocks start to collapse, and investor confidence in the sector as a whole softens or concerns arise that assets are being populated by product coming from fragile loss-making producers.

In the upstream, there are no winners, unless you had the foresight and wisdom to have a secure long-term sales pipeline and the ability to lock customers into legally-binding ASPs that were not index-linked in any way to third-party Asian spot market price reports, nor at risk from force majeure.

Interestingly it would appear only one company in the solar industry succeeded in doing this, and this company does not produce in China and does not make c-Si panels, and may soon look at Section 201 as being a gift of a lifetime.

But there are some relative winners across the upstream segment that will largely by luck find themselves in the right-place at the right-time; so let’s explore this now as it is potentially more critical in terms of manufacturing competitiveness, in particular from a China standpoint.

If you scan back through the archives of PV manufacturing then many will remember the days when there was a rush to be fully vertically-integrated, as far back as polysilicon, but at least covering the ingot-to-module stages. When polysilicon ASPs then enacted their major price decline from the days of 300$/kg to 30$/kg, companies that were free to purchase outside of long-term contracts found themselves in pole position.

Then during the 2012/2013 downturn, companies that were reliant on buying in cheap wafers (for cell/module operations) or simply cells for module-only activities had a key advantage over the ingot-to-module companies.

The term ‘flexible-buying-strategy’ then became a marker when comparing fully-blended prices that were often hampered by in-house costs being higher than third-party component availability. (Expect the ‘flexible manufacturing’ phrase to return with a vengeance very soon!)

The pending downturn is likely to see the final scenario unfolding, where companies that are top-heavy with cell/module capacity have a better cost structure when buying in wafers, than relying on their own in-house supply channels.

This creates an interesting dynamic then going forward, as there have been significant moves from some of the multi-GW module suppliers of recent to bolster ingot/wafer capacity. It may simply be that the companies that invested in adding mono pulling end up the winners within a multi oversupplied wafer climate. More on this below now as we shift to the ramifications for the mono-versus-multi balance.

Mono adoption almost certain to be fast-tracked now with muted objections

Just over a year ago, we ran a feature on PV Tech talking about the rate of market-share shift from multi to mono, and the factors behind this; see Mono and multi production 50:50 in 2018, but mono is the future.

The thesis was based on the assumption that the move to mono was happening at a rapid rate, and would before long see mono c-Si production levels at parity with multi, and then moving to a more dominant position out to 2020.

In this piece, I laid out the factors that would ultimately dictate the rate of market-share gains from mono, with two main caveats, copy and pasted from the May 2017 article as follows:

a) If the industry contracts, or even remains static, it will only increase the rate of mono market-share gains over multi, as mono is tight in supply and has the scope to be competitive with multi now. In this scenario, multi is wiped out faster than expected.

b) Conversely, if the market over-performs, and ends up over 85GW [for 2017] (don’t discount for one second), then multi has a lifeline due to the supply constraints of mono. And potentially more time to get its act together for low cost wafering and cell efficiency improvements.

What unfolded in 2017 fell firmly into scenario b) above, and the final annual production level of more than 100GW in 2017 meant that mono ended up with just 33% of the c-Si market-share. During the first half of 2018, mono managed to nudge up a few more percentage points, reaching 38%, and pending a real shift in deployment of modules in China during 2H’18, was heading for an overall market-share in 2018 of about 40%.

However, what the downturn now does is flip the mono/multi equation firmly over to scenario a) above, and will accelerate mono levels closer to the 50% tipping point during 2019, or at the very latest 2020.

Despite all the ambitions of those left with multi-GW of ingot casting, or recently-installed diamond wire sawing of multi wafers, multi has only cost to play with to stay competitive until enough mono pullers exist in the world to fully eliminate multi from the production mix. This may not happen for a further 5 years however, so plenty of time for multi to still have one final say in the PV industry, by setting lower and lower module ASPs that pull down the pricing on anything mono being shipped.

But maybe it’s a good idea to revisit this hypothesis in 12 months! Something will almost certainly happen that tweaks the mono/multi equation a bit more either way. However, the two scenarios above look likely to remain the basis of mono/multi share splits for some time.

What China should do next to control its PV sector

So, let’s leave the best until last in the article, with a sub-title that could occupy academics and industry panel discussions for years to come, and try to answer this with twitter-like brevity.

Before laying out what I think needs to happen, let’s set the record straight on one thing. The impending downturn is not happening because of a PR from public bodies in China in late May 2018. It is the unavoidable consequence of 2-3 years of chronic over investment across the c-Si value-chain within China that was a time-bomb waiting to explode.

The PV industry operates on cyclic investment phases, and after every upturn, there is a downturn. It is not scaremongering to say a downturn is coming!

Global capex across the c-Si value-chain (ingot-to-module) doubled between 2015 and 2017, with China alone investing more than US$10 billion, and accounting for between 60-70% of global manufacturing investment here. Add in here the >US$3 billion ploughed into polysilicon factories during this period, and we can see the problems now ready to impact margins.

But this is just the starting point. The even bigger issue now comes from the 2017-stimulated euphoria in China from the 5-10-20GW expansion plans, some of which are done, midway through, and are set to happen, come-what-may!

It almost seems incredible to the outside world that a company would choose to increase production of anything by a factor of 10 within a year, with no sales pipeline to fulfil! Or have an ambitious global expansion strategy, when the company is unknown outside its own HQ-location and has no overseas staff or route-to-market!

This happened back in 2010-2012 and ushered in the downturn of 2012-2014, and was accompanied by the reality of a European market that was not ready to grow by a factor of 10 overnight! The difference today is only in the numbers, except for the fact that it is a China-inflicted problem that will impact first on its domestic companies, and then have its secondary ripple globally.

Before any tinkering has to happen with policy and deployment of PV in China, it needs one major action to happen first: turn-off-the-credit-lines, and instruct the companies not to build any more factories until they are allowed to. Without this, it matters not what they do with policy, because it is runaway investments, and not local demand, that is the real problem.

Having efficiency targets to weed out the technology laggards doesn’t do the trick. Ordering firms to show >1% of revenues going into the R&D-spending box during accounting is also not a route to removing the also-rans (or actually being innovative!).

Stopping the credit lines for new factories has to be the top priority. Then see what the lay of the land is, and adjust expectations on deployment (policy) to prevent an excess of factories to slow down or be shut down. In the meantime, perhaps the time has come to finally draw up that mysterious list of who the top-10 government-favoured manufacturers in China really are, and work out how much of the capacity residing with the other 200 companies should be folded into each of these top-10 producers. (Like what South Korea and Taiwan did in other sectors over the past few decades, but in the totally opposite way!)

It is almost unimaginable that China will leave its manufacturing segment to shut-down, and this may cushion the impact, while they convince the government that they are getting their act together collectively.

So in contrast to much of the sensationalist output coming from other media outlets in the past few weeks, the sky is not going to fall in on China’s PV manufacturing segment overnight, and the industry will continue to grow globally, have lower production costs and increased panel efficiencies and performance levels.

I conclude the discussion now in hope that the next shakeout of manufacturing will be one driven by technology, and not merely by the ability of one country to access endless cash to add me-too production lines. This may be closer than we think, especially if wafers move to 120 microns as standard, or n-type finally takes chunks out of p-type dominance. Commenting on this type of activity will be far more stimulating for sure!

Read the entire story

PV manufacturing capacity expansion announcements collapse in Q3 2017

After the significant upwards revisions made to global solar PV manufacturing capacity expansion announcements in the first half of 2017, which we reviewed in a previous blog, the third quarter was characterised by much more tempered plans. The ‘Silicon Module Super League’ (SMSL) continued to execute on previously announced plans with some adjustments, while others in emerging markets such as Turkey and India retained grandiose nameplate targets but initial ramps remained small.

July review

The month of July proved to be the most active for new capacity expansion announcements in the third quarter of 2017. Total plans reached 3,001MW, which included 1,000MW of new solar cell capacity, a total of 2,000MW of dedicated module assembly and a nominal 1MW of advanced integrated manufacturing 4.0 R&D facility in California, opened by SunPower Corp. 

Turkish solar company Smart Energy Group was reported to have established a Joint Venture with China-based Phono Solar (part of SUMEC) to build and operate an initial 400MW module assembly plant with future plans said to take capacity to 1,200MW. The assembly plant would be established the Gebze Organized Industrial Zone (GOIZ). 

Also of note in July was plans by India-based Premier Solar Systems to build a 1,000MW solar cell plant with an overseas business partner. The company had also announced that it had opened a 200MW fully automated solar module manufacturing facility in Sangareddy, Telangana, India. The module assembly expansion takes nameplate module capacity to 375MW. The company has 50MW of solar cell capacity. 

August review

The weakest month for capacity announcements in the third quarter was August with only one verified announcement. India-based Heavy Engineering Corporation announced plans to build a 150MW integrated cell and module assembly plant using both monocrystalline and multicrystalline wafers. Initially, the plant will have a nameplate capacity of 150MW. Modules will be used initially for its in-house downstream PV power plant projects. 

There were no dedicated cell or module assembly plant announcements in August and none for thin film. 

September review

New announcements rebounded slightly in September. Total new capacity plans were around 900MW, which were dominated by China-based PV module manufacturer Sunport Power, which had officially started production at a 1GW module assembly plant using Eurotron’s MWT equipment for back-contact PV modules that reduce cell-to-module losses and boost module performance. The company had initially deployed a 200MW line so the new expansion accounted for 800MW of the total in September. 

Also included in September was the first fully-automated and unmanned 100MW monocrystalline solar cell line officially opened by Tongwei Group in Hefei, China as part of its most recent 2,000MW new solar cell plant. 

Quarterly review

Total third quarter 2017 capacity expansion announcements reached only around 4,122MW, compared to 28,000MW in the previous quarter.

The subdued environment was driven by dedicated module assembly plans, which totalled 2,870MW, while integrated cell and module plans, absent so far in 2017, totalled 151MW. No new thin film expansion plans were announced in the third quarter. 

India review

With a history of solar manufacturing, albeit small scale, India has held the promise of becoming a major power house for solar manufacturing, second only to China. 

With a downstream PV market in the 6GW range in 2017 and promise of much higher installation rates through 2022, the gap between capacity expansion announcements and effective nameplate capacity continues to be the one of the widest. 

Indian government data released in the second quarter of 2017 put solar cell capacity in India at just 3,164 MW, yet only 1,667MW as deemed to be operational. A similar situation existed with module assembly capacity. A total of around 8,400MW of capacity had existed in the country, while only around 5,500MW was deemed to be operational. 

As seen in the chart below 2017 has seen a significant reduction in new announcements compared to the last two years with 2016 peaking at over 17,000MW. In total we have tracked over 27,800MW of announcements in India since 2014. 

Challenges in building a manufacturing supply chain base in India that even comes close to meeting domestic demand has proved elusive. A key perennial challenge has been the capital markets but much of the projects tracked were JV’s with China-based companies as well as plans from US and Japan, which have also stalled despite the ability to tap low-cost finance in those countries. 

A key emerging challenge that has been cited for the manufacturing plans being stalled is related to the low prices tendered on multi gigawatts of downstream power plant projects. Simply put, the winning bids are lower than potential manufacturing costs in India, not least due to the lack of a highly efficient low-cost manufacturing supply chain in the country that could match that of China. 

Having depended on low cost modules produced in China, India is challenged to compete and JV’s with major Chinese producers such as Trina Solar and LONGi Group remain suspended. 

However, new efforts by the Indian government to support domestic content requirements at the manufacturing sector through a new wave of government led downstream projects, could become the catalyst required to kick start more effective capacity in India. Uncertainties in trade cases in the US could also make India attractive to supply modules to the US market in 2018 onwards. However, further reforms, a complete and low-cost manufacturing supply chain, coupled to rational tendering all need to be in alignment before the imbalance between capacity expansion announcements and effective capacity is closed. 

Solar manufacturing 4.0

Although the third quarter of 2017 was subdued for capacity expansion plans, it has signalled an important milestone in PV manufacturing. Several facilities were opened in the quarter that relate to the concept of Manufacturing 4.0, which includes fully automated manufacturing lines and operated remotely. 

In July, Silicon Module Super League’ (SMSL) member GCL System Integrated Technology (GCL-SI) announced the establishment and operation of a module assembly workshop that was completely unmanned to test intelligent fully automated manufacturing tools and software systems. The workshop is expected to undertake tests for around two years.

The company noted that it was cooperating closely with Chinese domestic equipment manufacturers, and has independently developed and researched and developed a series of intelligent equipment and systems, which included high-speed automated tabbing machine a high-precision layout machine and robot palletizing system, In all, GCL-SI said that 26 separate systems so far developed were industry firsts.

A key aim of the tests is to achieve a 50% improvement in efficiency, a 60% reduction in online manpower and reduced processing costs by 30%. Product quality improvement targets were being set at a 21% overall improvement. The intention is to implement the improvements across its volume manufacturing operations.

In August, SunPower Corp said it had invested around US$25 million in the last 12-months on a new US R&D and pilot line facility located at its headquarters in San Jose, California. 

SunPower noted that the new facility included several high-volume production-sized manufacturing tools and high levels of automation, and specialized testing equipment, designed to support its next-generation of high-efficiency N-type monocrystalline IBC (Interdigitated Back Contact) solar cells and modules, which are being designed with greater emphasis on lower cost manufacturing.

According to SunPower, over 30 parts suppliers and equipment manufacturers located in the US supplied the facility, which is housing over 100 SunPower engineers and support staff.

In September, as already noted, Tongwei Group opened its completed high-efficiency solar cell plant (S2), which included the world’s first technically unmanned 100MW monocrystalline solar cell production line under the intelligent manufacturing term, 4.0. 

The S2 plant in Chengdu, China has an initial nameplate capacity of 2GW, which brings Tongwei’s monocrystalline cell capacity to around 3.4GW. The company also has around 2GW of multicrystalline solar cell capacity. The company also has around 2GW of multicrystalline solar cell capacity and recently completed a 5,000MT polysilicon plant expansion, bringing nameplate production capacity to 20,000MT. 

Tongwei is investing around RMB 12 billion (US$1.8 billion) in total constructing new cell manufacturing facilities at Hefei Solar’s facilities in the Hefei High-tech Industrial Development Zone in Chengdu City to provide nameplate capacity of 10GW, while a further 10GW of capacity will be housed in the Southwest Airport Economic Development Zone of Shuangliu District, Chengdu City. Construction on the new projects is expected to start in November, 2017 and production ramped in phases over the next three to five years. 

Tongwei has taken the early lead in China in investing in Manufacturing 4.0 capabilities, however much is being done behind the scenes at other major manufacturers and the learning curve is expected to take several years.

SMSL Q3 manufacturing update


Leading SMSL member JinkoSolar reported that its in-house annual silicon wafer capacity stood at 7GW at the end of the third quarter, up 1GW from the prior quarter.

Solar cell capacity as expected was 4.5GW, while module capacity did increase by a further 500MW in the third quarter, reaching 8GW. These are the expected nameplate capacities exiting 2017. Although the company claimed its next wave of expansions had yet to be determined and would be based on market demand dynamics it is highly likely new plans will be announced in the next two quarters. 

JinkoSolar is expecting to hit record shipments in 2017, having guided just short of 10GW, indicate almost a 10% global market share of module shipments and is sold out through the first half of 2018

Canadian Solar

SMSL member Canadian Solar has made four revisions to capacity expansion plans in 2017 and has also provided initial new expansion plans for 2018.

The SMSL member noted that it had completed the ramp up of a new multicrystalline silicon ingot casting workshop at Baotou, China at the end of the third quarter of 2017, with a total annual capacity of 1,100MW, which included capacity relocated from its plant in Luoyang, China. 

The company noted that it expected debottlenecking to push capacity to 1,200MW by the end of 2017, which is in line with the last two updated plans. 

Canadian Solar said that it had plans further increase its ingot capacity to 1,720 MW by June 30, 2018, and may expand to 2,500MW if market conditions justify.

Wafer manufacturing capacity had reached 3GW in the third quarter of 2017. The company had previously guided that it expected wafer capacity to reach 4GW at the end the year. 

However, Canadian Solar noted that its shift to diamond wire-saw technology, which is compatible with the Company’s proprietary and highly efficient Onyx black silicon multi-crystalline solar cell technology, helped to significantly offset the recent impact of polysilicon price increases that impact margins and so was planning to add a further 1GW of wafer production to end the year at 5GW.

The company said that its solar cell manufacturing capacity reached 4.7GW at the end of the third quarter of 2017, which was the target in its third revision to its capacity expansion plans.

However, Canadian Solar said that it planned to add additional cell manufacturing capacity at its Funing and South East Asia plants by year end, bringing 2017 cell nameplate capacity to 5,450MW, a 750MW increase. 

Subject to market conditions the SMSL said it planned to add another 1.5GW of cell capacity in 2018 to reach approximately 7GW by the end of 2018.

With respect to PV module manufacturing capacity, Canadian Solar is adding almost 1GW of nameplate capacity more than its third revision made in the second quarter of 2017, which would have led to a 2017 capacity of 7,190MW. 
The company expects that its total worldwide module capacity would exceed 8,110MW by the end of 2017.

Subject to market conditions again, the SMSL member said it planned to add another 1,250MW of module capacity by the end of 2018, bringing nameplate capacity to 10.3GW. Canadian Solar is the first manufacturer to guide nameplate module capacity to reach over 10GW.

JA Solar

SMSL member JA Solar confirmed that it expected to achieve both cell and module  nameplate capacities of around 7,000MW by the end if 2017. JA Solar is on track to achieve full-year module shipments in the region of 6.8GW in 2017, but did not provide an update on 2018 capacity expansion plans. 

Hanwha Q CELLS 

SMSL member Hanwha Q CELLS said that it was starting to migrate solar cell capacity at its China-based facilities to PERC (Pasivated Emitter Rear Cell) technology, highlighting the shift away from standard BSF (Back Side Field) technology for higher conversion efficiencies. 

Hanwha Q CELLS noted in its third quarter earnings call that capital expenditure was being focused in its manufacturing facilities in China to enable the company to have PERC cell production capacity of 1.4GW, while retaining around 1.2GW of BSF production. Hanwha’s lead manufacturing facilities are in Malaysia and are already 100% PERC. 

Its affiliate, Hanwha Q CELLS Korea is currently adding 1.6GW of cell and module production, which is expected to provide a nameplate capacity of 3.7GW by the end of this year.

Hanwha Q CELLS in-house cell and module capacity has not increased in 2017 as the company keeps tight control on spending to return to sustainable profitability. However, with Hanwha Q CELLS Korea expansions in 2017, the group will have access to 8GW of cell and module capacity starting in 2018, up from 6.4GW at the end of the third quarter of 2017.

Hanwha Q CELLS recently reiterated that it expected module shipments in 2017 to be in the range of 5.5GW to 5.7GW.

GCL System Integrated Technology (GCL-SI)

SMSL member GCL System Integrated Technology (GCL-SI) reported it had ramped its solar cell and module assembly JV plant in Vietnam to around 800MW in the third quarter of 2017.

Critical to the market, the solar cell capacity ramp has been PERC (Passivated Emitter Rear Cell) technology with the flexibility to produce P-type multi and P-type mono cells for the residential, commercial and utility-scale markets.

The company said that its total solar cell capacity would reach 2GW by the end of 2017, which would be completely PERC-based technology based.  Currently, around one third of production is P-type multi using ‘Black Silicon’ texturing after wafers (S2 size) are cut with diamond wire. Around a third of production is P-type mono PERC, while a further third of production is flexible to customer demand. 

LONGi Group

SMSL member LONGi Group, which is leading the wafer transition to monocrystalline high efficiency wafers, cells and modules has actually increased the pace of some of its previously announced plans for 2017. 

At the beginning of the year its capacity for wafers had reached 7.5GW and would reach 12GW by the end of 2017. In the third quarter of 2017, LONGi surpassed the 12GW mark and was planning to add further capacity to meet continued strong demand. 

LONGi still expects to meet expansion goals of 5GW for solar cells and 6.5GW for modules by the end of the year. However, it does not plan to provide updated plans until issuing its 2017 Annual Report. 


Despite the slowdown in new plans in the quarter, execution on existing plans has been a key theme throughout the year, notably for China-based firms and the majority of SMSL members. Having continued to gain market share in 2017, SMSL members are all expected to announce record annual module shipments in 2017. 

Overall, expansion plans simply collapsed in Q3, hardly surprising due to significant level of new plans announced in the first half of the year and increased concerns over the potential imposition of restrictive trade practices in the US and India.

Read the entire story

Second major capex cycle underway as PV industry enters new phase of 100GW-plus annual deployment

Solar PV capital expenditure (capex) covering the midstream segments of the industry (c-Si ingot-to-module and thin-film) is now well into its second major upturn in spending, going into 2018, at a time when the industry is just about to move to a new phase in annual deployment levels of greater than 100GW.

This article discusses why this is happening, the companies and technologies driving this change, and what can be expected in 2018 and beyond.

The underlying themes and outcomes will form a key part of the forthcoming PV CellTech 2018 event in Penang, Malaysia on 13-14 March 2018.

Defining the methodology

PV capex is best analysed by removing spending on polysilicon plants, as the capex here operates on fundamentally different timelines and phasing, compared to capex being allocated to the midstream segment covering ingot-to-module spending. In addition, capex for polysilicon plants is more weighted to plant construction, than discrete process tooling.

PV capex covers factory build costs (often partially financed in Asia by autonomous regions or inward investment vehicles), infrastructure design and build out, and (mostly) production equipment tooling. Therefore, PV capex largely defines the total addressable market for PV equipment suppliers, and offers a key metric to assess market-share trends and share-levels.

The data and graphics contained within this article are generated bottom-up by PV-Tech’s in-house market research business unit that gathers the data for use within its portfolio of syndicated and bespoke research products, in addition to guiding the content and production of the PV CellTech and PV ModuleTech events.

PV capex encompasses also new manufacturing capacity and upgrade equipment spending, in addition to routine maintenance and equipment replacement spending.
Understanding the past

We have gone back 10 years to understand some of the historic trends and drivers that offer some perspective on where the industry stands moving into 2018, and that shape some of the factors that can be used to predict how PV capex will trend over the next 3-5 years.

The figure below shows PV capex trends over the period from 2007 to 2017, referenced to the annual revenues recognized by the leading PV equipment supplier of the past decade, Meyer Burger.

PV capex is going through its second major growth phase, driven by a combination of new capacity expansions, new technology introductions and upgrades to existing installed capacities.

Analysis of PV capex before 2007 is somewhat academic, as spending levels then were minimal. The first real capex cycle for the industry started in the 2007-2009 period, and was driven by a diverse range of issues.

China and Taiwan were then starting their first significant expansion phases for c-Si cell and module capacities, stimulated by demand that was coming from Europe (mostly Spain, Germany and Italy). However, while this was driving huge growth in new order intake from c-Si based tool suppliers, it was thin-film excitement that ultimately drove PV capex levels above USD$10B during 2010 and 2011.

Thin-film market-share gains were being advocated as a done-deal before 2010, and investments of USD$100M at a time were being issued to credible turn-key equipment suppliers, mostly on the back of producing panels that were well below 10% in efficiency. The numbers added up very quickly, and turn-key thin-film production lines quickly had a major influence on overall PV capex levels. Today, none of these lines are in operation.

The spending peak occurred in 2011 and saw literally hundreds of new companies making cells and modules emerge all over China. Each one of these companies saw Europe as a bottomless pit of government subsidy hand-outs that would see crates of Chinese modules shipped in volumes to ports in mainland Europe for years to come.

This didn’t happen. The industry went into chronic overcapacity and oversupply mode, resulting in module pricing collapse. Investor confidence plummeted. Operational losses were widespread. And the result was a collapse in capex in early in 2012.

It took two years for supply and demand to largely get into balance, and this period was all about cost reductions, existing capacity optimization and debottlenecking.

The downturn lasted about two years, and by 2015, green shoots were emerging everywhere with the first signs of technology (not capacity volumes) being the new impetus for the rebound cycle of manufacturing capex.

Technology driven capex rebound

The focus on new technology, both due to upgrades and when incorporated in new fab builds, since 2015 looks very different to the technologies that drove PV equipment spending above $10 billion during 2010 and 2011.

At the c-Si stage, the most obvious change has come from PERC, and by 2019, most of the capacity for both p-type mono and multi will have shifted to include rear passivation deposition, with many of the companies having a clear roadmap to bifaciality. Moving into 2020, this will become mainstream for the industry. Indeed, anyone making c-Si cells with efficiencies below 20% is likely to be left with low-cost selling options then.

However, some of the other key drivers of the current capex rebound are new technology-driven initiatives that have not been seen before, including First Solar’s shift from Series 4 to Series 6 panels and the n-type spending boom that is emerging now in China especially.

It is fair to say that n-type has always seen high levels of interest, with no shortage of roadmaps from major cell producers that tended to be largely wishful thinking. While SunPower and Panasonic largely sat back and watched many n-type start-ups come and go, only Yingli Green and LG Electronics succeeded in having n-type cell capacities in excess of 500 MW that were being operated at mass-production utilization rates.

This all changed about two years ago in China, and the first GW fabs are currently ramping up, with n-type variants spanning n-PERT and heterojunction being the main focus. Regardless of the success rates of these initial movers (many of whom have little if any experience making solar cells until now), what is certain is that we are going to see more investments during 2018 and 2019 with government initiatives in China offering the backdrop for investment security.

In fact, while previous n-type activities in the solar industry were company and R&D lab specific, the China efforts today are more collaborative in nature, or simply less prone to the commercial confidentiality that is the norm for manufacturing outside of China and Taiwan.

Moreover, efforts to use domestic equipment suppliers for n-type tool supply in China further helps to establish (albeit indirectly) an open pool of knowledge that could be shared somewhat freely by other players in China wishing to jump on the n-type bandwagon should it become a competitive threat to p-type mono in coming years.

No sign of Chinese capex slowdown

On so many counts today in the PV industry, not simply for PV capex, there are no obstacles suggesting any imminent cooling in China. The security of a 50GW-plus domestic end-market exclusive to Chinese-produced ingots/wafers/cells/modules is unprecedented, and the fact that we see new technology-driven initiatives now for n-type differentiation should not come as any great surprise.

Investments in a few gigawatts here and there for new n-type projects may seem like reckless extravagance to most outside China struggling to make the numbers balance monthly on established GW-level fabs, but within China this is almost the new norm, and not entirely out of context if one regards the entire c-Si capacity in China as a single integrated manufacturing unit.

The graphic below spells out just how important equipment spending on new capacity and technologies within China has been to the capex rebound since 2015. Remove this contribution, and we are back essentially at 2012 levels.

Equipment spending across the c-Si ingot-to-module value-chain has been stimulated by domestic manufacturing within China, with 2017 and 2018 contributions seeing new efforts to scale up n-type GW fabs.

PV CellTech 2018 to address the competitive threat from n-type activity

Going into its third year, PV CellTech 2018 has been structured to address the threat coming from the new n-type investments, and whether in particular n-PERT and heterojunction can truly offer value-added propositions versus p-mono PERC.

This issue will become one of the key themes for solar manufacturing during 2018-2020, as important as the ‘what-next-after-PERC’ question for p-mono. And behind all of this is cost and wafer supply, not to mention potential wafer thickness reduction possibilities not currently being driven by the p-type community.

Could the threat from a few GW of n-type heterojunction capacity in China, coupled with multi-GW of sub-140 micron n-type wafer supply, be the catalyst that finally drives the entire c-Si industry to make a step-wise silicon material consumption decrease that would make major inroads into the silicon component of blended module costs, and cause havoc with polysilicon expansion plans?

PV CellTech 2018 takes place on 13-14 March 2018, in Penang, Malaysia. The relevance of the event, and in mapping out the real technology trends over the next 12-18 months, is almost definitely set to move to a new level of importance.

Read the entire story

JinkoSolar poised to hit 10GW annual shipment target, but what next?

When JinkoSolar released its third quarter results last week – and guided full year 2017 module shipment figures – the company remained on track to overachieve on final quarter shipments, thereby becoming the first ever PV supplier to ship more than 10 GW of modules in a calendar year.

If this landmark figure is reached, JinkoSolar will effectively have 10% market-share, and it will have achieved one of the key goals set internally 12 months ago.

This article tracks how JinkoSolar has managed to move so quickly in the past four years, from having 5% market-share in 2013 (with module shipments less than 2GW), to 2017 where 10% and 10GW has been the goal for its global sales team.

The question of ‘what next’ is also discussed, in view of JinkoSolar having relied heavily on third-party cell and module supply during the past four years to support its module shipment aspirations; and from an audited reporting perspective, being essentially a pure-play module revenue operations.

The 10GW dream

In February 2017, JinkoSolar gave its first guidance for 2017 shipments, at 8.5-9GW. However, it was generally understood that JinkoSolar had set an internal goal for its sales team closer to 11GW, providing the first indication that hitting 10GW was seen as the real target, and one that would yield significant marketing kudos during 2018 and beyond.

When first half shipments in 2017 exceeded 4.9GW, further signs emerged that 10GW would be achieved rather easily. Even during 2016 (with the China slowdown in Q3’16) second-half shipments exceeded first-half by more than 10%. Previous years for JinkoSolar have been significantly higher than 10% half-year growth.

When 9-month shipment figures were reported by JinkoSolar last week, the annual level of 2017 had exceeded 7.3GW, almost perfectly on track to reach the 10GW shipment goal. Guidance for 2017 was given at 9.6-9.8GW.

However, the final quarter has generally been the strongest quarter for JinkoSolar, with approximately 30-40% of annual shipments occurring in Q4 during the past few years. If 2017 follows this trend, then the 10GW annual figure will be met, and explained simply by final shipment figures coming in just above the high end of Q4’17 guidance.

Therefore, when JinkoSolar reports its final 2017 shipment figures (during February or March 2018), don’t be surprised at all if the company begins its press release with a simple and clear message, based firmly on being the first ever PV company to ship more than 10GW of solar modules in a single year.

Expect no shortage of marketing collateral and fanfare should this occur!

Doubling market-share in a high-growth segment

Any company that doubles its market-share in four years – in any sector – will generally be accompanied by its chief executives shouting from the rooftops, happily receiving plaudits from the investor community in recognition of a strategy successfully enacted.

However, to do this in a sector that has seen more than 25% compound annual growth during the same period, there has to be more than just a few stars fortuitously coming into alignment to help things go your way.

And for good measures, let’s add in a bunch of unforeseen trade cases impacting place-of-manufacture and shipment-destination, and a 40% reduction in market pricing due to chronic oversupply.

This is indeed what has happened in the PV industry during the past four years, while JinkoSolar has moved its market-share from 5% to 10%, is likely to hit the 10GW module shipment figure this year, and will be miles ahead of any other module supplier as the leading and single most important module supplier to the PV industry going into 2018 and beyond.

Nothing like this has been seen before in the PV industry, when it comes to module shipments. So let’s have a look at what has enabled this, leading finally to the big question now – what next!

How did JinkoSolar do it?

There are many reasons behind the sub-header question here, but four issues prevail in dissecting JinkoSolar’s rise from a 1-2GW module supplier to a 10GW player in 2017:

  • Flexibility in cell and module supply to meet market-growth opportunities
  • Having a competitive module product offering at any given time
  • Maintaining gross margins in the 10-20% range
  • Having a global sales operations capable of winning opportunities in all key end-markets

While other companies have laid claim to meeting the first three bullet points shown above, it is the final issue that probably differentiates JinkoSolar from other competing global module suppliers in the past few years.

Indeed, having a truly global sales presence and operations is something that most PV module manufacturers have struggled with over the past decade, dating back to efforts from Sharp Solar, Suntech (pre-Shunfeng), Trina Solar, Yingli Green, and many others headquartered in China, Japan and South Korea.

Parallels exist today with other Chinese companies that are driven to play on the global PV stage, including LONGi Solar, GCL, BYD, Astronergy, Risen, Talesun, etc., and those that once had an overseas route-to-market but are now constrained to domestic efforts: Kyocera, Solar Frontier, Panasonic and a myriad of Chinese cell/module makers that had briefly flirted with the legacy FiT days of mainland Europe.

Chinese companies that manage to create global brands, accompanied by sales operations that, to the outside world, operate as local entities, are few and far between – in any sector. However, this is something that JinkoSolar appears to have managed to do in the past few years, and the company is certainly the first and only Chinese module supplier to get this right.

The facts and figures clearly support this strategy in 2017. If we split the global PV end-market into six regions/counties, this comes through.

During 2017, JinkoSolar will rank number two in the Chinese and Japanese markets, third in India, fourth in the US, and top-rank in Europe and the all-other category (less the countries/regions just mentioned). There are very few companies that can claim to be top-4 suppliers to every key region globally in the industry today.

Supply enabled, with flexibility

The previous section of this article flagged up the need to have flexible supply. This is key in the PV industry today, as the seasonal demand profiles from the key markets have often been out of phase, and have been impacted by pending or enacted trade-related deadlines for shipments.

Being able to manage these cycles has been a massive challenge to companies seeking to serve the market globally, and has been amplified for companies confined to cell and module manufacturing operations based purely in mainland China.

This has impacted on how much midstream capacity has been brought online within China or at overseas plants across Southeast Asia. Ideally, these sites should be run to serve the baseline demand coming from the regions that the manufactured product is intended; Japan, China and India for Chinese-produced product, and US and European supply coming from Southeast Asia.

Upside – and market-share gain – therefore has to rely upon outsourced cell and module supply/production when needed. Managing this from quarter to quarter has been one of the biggest challenges for global module suppliers in the past few years.

The figure below consolidates JinkoSolar’s in-house and third-party cell and module supply levels across the period 2015-2017, shown against the total module shipments from the company over the same time period.

Using third-party suppliers may have the benefit of flexibility in supply (assuming you are the dominant party in the contractual agreements), but the downside is a hit of 2-3 pennies per Watt on the cost side. (In-house manufacturing ideally is a lower cost operations compared to buying in components from outside.)

JinkoSolar’s strong use of third-party cell supply can actually be traced back to the company’s initial entry within the PV industry, when it positioned itself as a low-cost module assembly company in China before integrating upstream to cells and ingots/wafers.

However, all this has changed in the past three years and is discussed in the section below.

Technology leadership could be the key differentiator

During the past decade, many module suppliers have been rather fixated on claiming the mantle of being the number-one module supplier to the PV industry, based purely on having shipped the highest MWp-dc of company-branded modules.

The driver from any marketing department is simple; it enables legitimacy on claiming to be the market leader based on a universally accepted metric – module shipments.

However, any historian of the solar industry will instantly flag the fact that most companies that had achieved number-one status (in modules and cells) were subsequently left exposed from a what-next strategy standpoint, and within a few years, had sadly fallen from grace: Sharp, Q-CELLS, Suntech and Yingli Solar being names that spring to mind here.

However, JinkoSolar’s tenure as number-one module supplier appears to have a few key differences that are not widely known or discussed, and this may suggest that we are potentially seeing the first major globally-dominant module supplier to the PV industry.

Looking at the Chinese and Japanese companies that held leading market-share status in the past, this accolade was not the result of having a differentiated product offering; and once these companies become market leaders, they failed to invest heavily in R&D or made the wrong decisions with regards cell manufacturing technology. Very quickly, competition overtook them in terms of product performance.

What we see with JinkoSolar is very different to the behaviour of the other aforementioned companies, and come over clearly when we dissect the company’s operations since 2013, and forecast 12 months out to the end of 2018. This is shown in the graphic below:

Since becoming the leading module supplier in 2016, JinkoSolar has doubled investments into R&D, and by the end of 2018, will have shifted completely its in-house technology mix, aligned with having performance leading p-multi and p-mono module offerings.

If one single graph explained why JinkoSolar is different from other market leaders in the past, then the one above would take some beating. But this is probably not the main thing, as discussed now below.

Being non-Chinese outside China on the global stage

Of course, having strong R&D allocations and GW’s of differentiated technology-capacity only really takes one thing: cash. Selling this globally requires a winning sales and marketing strategy.

Perhaps indeed this is the hidden gem that JinkoSolar has in its arsenal, and what really differentiates the company when compared to all other module suppliers headquartered throughout China, Japan, Taiwan, Korea and Southeast Asia – with the exception only of Canadian Solar, JA Solar, Trina Solar, Hanwha Q-CELLS and REC Solar.

But the gap between JinkoSolar and all other companies mentioned just above is not insignificant, and has certainly grown in the past few years. Indeed, on the global stage today, the leading Chinese companies that truly portray global-brand status can be counted on one hand: JinkoSolar, Canadian Solar and Trina Solar.

It is actually the above two factors combined that make JinkoSolar’s place in the PV industry right now different. The solar industry has not seen this before from a Chinese headquartered company, and during 2018, there is not a single reason why JinkoSolar will not further increase its market dominance.

What happens in 2019 is too far to speculate – there are way too many variables in the PV industry that can change quickly, impacting fortunes overnight.

Whether we are truly seeing the first signs of a global manufacturing powerhouse emerging in the industry, or just another cycle that will be replaced by another Asian company in a couple of years, is going to be a fascinating dynamic to observe and track over the next 12 months.

Read the entire story

‘All modules are not created equal’ – PV ModuleTech 2017

At the inaugural PV ModuleTech 2017 in Kuala Lumpur, it became clear that PERC (Passivated Emitter Rear Cell) module technology is fast becoming the industry standard and bifacial modules are no longer seen as a niche product, however, both technologies were scrutinised from the perspective of bankability and quality assurance studies throughout the event.

The words ‘PERC’ and ‘bifacial’ were repeated multiple times during the two days in Malaysia, reflecting a steady shift in industry focus.

There is a need for advanced testing of these emerging technologies to allow investors, developers and EPCs to get to grips with the real advantages and risks. Finlay Colville, head of market research at Solar Media, said that until recently the PV module technology had not significantly changed for a number of years and he still speaks to developers who don’t know of anything beyond standard 260W multicrystalline aluminium BSF (Back Sise Field) modules, which they have deployed in huge numbers. However, these players are now beginning to ask how bifacial modules and PERC can benefit them. For this, they require clear data and assurance from both manufacturers and the range of third-party bodies examining these new technologies.

For example, when it comes to PERC, developers have to grapple not just with light-induced degradation (LID), but with light and elevated Temperatures Induced Degradation (LeTID) which can occur in multi and mono p-type PERC modules. A number of presentations covered the LeTID issue in depth. For example, a Hanwha Q CELLS study found that LeTID could hit performance over three years by up to 8% (Cyprus) and 3% (Germany).

Lab vs outdoors

There is also a tension between ‘real world’ data, achieved through outdoor testing over months and years, versus experiments in the lab. The suitability of accelerated conditions testing was a hot topic, and as PV Tech reported from day one of the conference, a representative from one of the world’s largest developers stressed that financiers, who really drive project development, are looking for more than five years of field data. Thus in the absence of this field data, manufacturers with new technologies need to carry out as many multiple tests, assurances and simulations with independent parties as possible to give the finance community confidence.

Jenya Meydbray, VP of solar technology at Cypress Creek Renewables, a US-based company that has developed more than 5GW of solar projects, said: “We are very open to new technology […] and we are willing to take risks on 2MW of new stuff to start getting field experience with it, but the manufacturer probably has to participate in some way depending on what the specific risk is.”

Throughout PV ModuleTech, presentations showed a number of different testing methods, ranging from temperature tests to First Solar demonstrating the effects of dropping a hand tool from a height of one metre onto various module architectures.

25 years

Having visibility and traceability of one’s bill of materials (BOM) was also in the spotlight. One of the key takeaways was that the entire value chain needs to come together to ensure quality and reliability. For example, Lou Trippel, vice president, product management at First Solar, said that module manufacturers tend to be the first stakeholder called up and blamed when a PV project starts to underperform. He then showed a number of images including mounting structures broken to the point where modules were engulfed in flooding and even a picture of a bush growing freely on top of a solar array. 

These were extreme examples, but the point made was that the EPC players, installers and operators have just as much responsibility in ensuring a successful project for 25 years. This was indeed one of the aims of PV ModuleTech 2017, to bring together all players across the value chain to learn about the requirments of each party. It was also a platform for those in the middle ground, the third-party testers, engineers representing banks, and quality assurance houses to show how they test, what they look for, and how emerging module technologies are performing in simulation at present.

Colville added: “Variation in quality control is massive.”

Frederic Dross of DNV GL also noted an “asymmetry of information” out in the market which can create mistrust, hence the need for the middlemen to provide unbiased data.

One of the values of PV ModuleTech was indeed that it had manufacturers, buyers and independent assessors all under one roof with the ability to scrutinise the new module technologies that are deemed most likely to go mainstream in the near future.

Bankability vs new tech

Delegates did raise concern about whether some of the testing regimes were not conducive to progressing innovation into the mainstream.

George Touloupas, Clean Energy Associates, asked: “Is bankability an instrument to block the introduction of new technologies into large deployment, maybe not by intention but by effect, and if it is not, how can a new technology get through this filter.”

Ralph Romero, senior managing director at Black & Veatch, said that his firm, for example, works with start-ups and new entrants at an early stage offering workshops to identify as early as possible what requirements they will need in order to pass these bankability studies. They unpick the nuances of the independent assessment and then develop a roadmap to be successful from a bankability viewpoint.


Adding to the general perception that PERC is becoming standard, Zhu Qiangzhong of LONGi Solar said: “We think that the PERC and bi-PERC is coming. […] By 2020 modules with mono-cells will take up half of the whole capacity and most of the modules will be PERC.”

‘Bifacial believers’

Many different claims were made about the possible energy yield improvements brought about by bifacial modules. Concerns that there needs to be a more standardised figure to help the industry compare bifacial with monofacial modules were brushed aside given the huge range of factors that affect bifacial performance, including height of installation and the albedo of the ground surface.

We heard several times that bifacial modules combined with trackers was a highly successful solution, providing significant boosts to module and plant yields. The industry has wondered for sometime whether trackers could add actually add value to bifacial modules given the importance of reflections hitting the backside of the modules.

Lim Cheong Boon of Trina Solar said: “Bifacial plus tracker is a win win combination for our customers.”

Similarly, Radovan Kopecek of ISC Konstanz said that bifacial with trackers could give even more than 40% advantages to the system.

However, Colville added the caveat that EPCs and developers need to prepare for a significant change in project layout.

He said: “You have to completely rethink the way you’ve been designing solar parks for the last five years. There’s those that are about to do that very quickly and understand all the factors and those that are just going to be literally caught at a certain point two years, 18 months down the line not even knowing what’s coming round the corner.”


PV ModuleTech closed with presentations from Indian manufacturers Adani and Indosolar, offering a new way of perceiving cell and module production in India. They presented the transition from ‘Make in India’ to ‘Made in India’, offering up the idea of India’s manufacturers possibly having a new strength in exporting, particularly with some of the trade cases looming worldwide, with Section 201 in the US and anti-dumping investigations in India.

Offering a downstream perspective on the Indian solar market, Colville added: “India is probably the most stable long-term growth market I’ve ever seen in the solar industry and that’s because it is a prime minister who wants it to happen, genuinely wants it to happen. and there’s a need almost more than anywhere else in the world. It’s a question of the rate of deployment and infrastructure and so on.”

Biggest Investor concerns

To wrap up a panel discussion on lifetime and return-on-investment, Colville asked what the speakers felt was the biggest concern aired by investors:

  • Rahul Khatri, technical specialist at E.I. DuPont India, said: “The biggest concern we have heard in India is whether the plant would work for 25 years or maybe 20 years and secondly is the kind of risks that are not controlled in terms of Distribution company reliability, whether they will pay you on time.”
  • Andrea Viaro, head of technical service, EU, JinkoSolar, said: “Reliability. The fact that the normal IEC standards evaluate very well the infant mortality of modules but there is no way to practically estimate a 20 years’ lifetime of modules. From the customers point of view would be nice to have a reliable estimation of the real performance of the modules in the very long term.”
  • Miguel Laburu, business field manager, TÜV Rheinland Singapore, said: “Banks and investors; what they are looking for is reduced uncertainty so they can really calculate what is the actual performance of the power plant in 25 years so they can really bring back to investors or to the owner of the pension fund what they have promised.”
  • Paul Wormser, vice president of client solutions at CEA, said: “Today it’s Trump and that could have implications that go well beyond the US market, and I think in 25 years, it’s going to be how do I decommission this?”
  • Jenya Meydbray, VP of solar technology at Cypress Creek Renewables, said: “It’s risk, in that of all the things, reliability, performance, there’s never going to be a happy area where we’re comfortable even with aluminium BSF 72-cell poly with SMA central inverters and fixed tilt racking and a plain vanilla system. If it’s plain vanilla then all the developers are going to get more aggressive with what they are trying to push. There’s always going to be a natural tension between the developers pushing and what the banks are willing to accept.”

Read the entire story

Copyright 2017 Xevent. All rights reserved