Chinese PV industry stays in PV world at a position of producing 140% of worldwide demand (110GW vs.80GW esp. cell and module), consuming 40% of WW demand (30GW vs. 80GW), crushing competitors in US/EU even with 30-50% AD/CVD tariffs.
• 2017 China PV installation range 30-35GW (calendar frame). 2018 likely at same range.
• Demand crumbles post 630 but likely less destructive than 2016, as government intends to promote FiT independent projects (less sensible to deadline)
• Price pressure and overcapacity remain severe. Market purge underway pushing out low efficiency producers.
Chinese PV industry stays in PV world at a position of producing 140% of worldwide demand (110GW vs.80GW esp. cell and module), consuming 40% of WW demand (30GW vs. 80GW), crushing competitors in US/EU even with 30-50% AD/CVD tariffs. There is no any other way where we can outline PV demand supply metrics without closely looking into Chinese market.
It’s time to look to market trend 2H of the year. For China PV the scrutiny is more meaningful, given the fresh memory of last year when the Chinese market froze right after Jun., and a worldwide tumble ensued. A variety of market reports and producers have given outlook of 2H2017. Many agree on that trajectory lines of main market elements, demand, supply, and price, would appear similar to past year in both China domestic and global market. If anything notable, the line of demand is likely modest of fluctuation in China compared to 2H2016, while the line of price would remain precipitate, driven by still fervent expansion among Chinese factories.
Taking a closer look at China PV installation, we can find more details along several time points: 630 (30th Jun. ), 930, and 1230.
630 marks end of rush effect of China domestic installation (FiT for ground project will be 15% lower afterwards and new quota to be released long after). NEA officials recently gave estimates, citing China installation 1H2017 would be around 23GW, vs. 22GW of 1H2016. Many industry people have come to the point that this year the rush effect would appear to be a modest one compared to 2016. Take a look at the 23GW installation:
Generally, 7GW for Q1, 16GW for Q2.
Breakdown of project variations.
i) 12.6GW ground project (FiT based) set up by NEA (at Jun. 2016) as 2016 ground project target. But 4-5GW is estimated to be completed by 2016, so around 8GW will be installed by 1H2017.
ii) Anti Poverty Project (APP) 5.16GW (issued at Oct. 2016, 2.98GW ground mounted, 2.18GW village based) should be completed by 630. Notably, mono products have got strong position of supplying APP projects which require high efficiency and tendering assessment.
iii) Rooftop: no limit on RT capacity but those having full grid transport would comply with 630 deadline. NEA estimates it to be around 4GW by 630.
iv) Supplementary Project. On 22nd Dec. 2016 NEA provided a list of additional projects on top of 2016 ground target, in order to abate financial strains in some provinces where excessive projects were installed before receiving approval. It’s become sort of NEA routine operation in previous years but this year the capacity came out surprisingly large, around 11.4GW. Records showed that 8GW or so has been installed in 2016, resulting in around 3.4GW left upon completion before 630.
v) Front Runner Project (FRP). 2016 target capacity was set as 5.5GW (announced at Nov. 2016). In a disputed situation, FRPs are almost 100% supplied by mono products. There is no official voice in regard of deadline of FRP project. So at beginning people thought FRP was deadline independent (due to its tender structure based on $/KWh). However more and more development companies winning FRP projects have come up citing that there were firm deadline terms in the contract, mostly pointing to 930 of 2017. As result, a sense has prevailed that FRP has got deadline of 930. However experts suppose around 2GW of FRP has been installed by 630.
In summary, around 23GW is supposed to be installed by 630 time point. Q2 alone will see around 16GW installation, which explains several market movement at present, such as, rush order effect, tight supply (esp. mono), and price uptick along the whole supply chain.
930. It mainly applies for Front Runner Project (FRP). As above noted the time period between 630 and 930 will mostly be seeing installation of FRP projects, which accounts for 3.5GW as reminder of 5.5GW 2016 target.
Moreover, this week NEA published a draft guidance for 2017 FRP project, aiming to annual capacity target of 8-10GW. Therefore a significant increase of FRP installation is expected for 2H2017 and 2018 throughout. This new target should provide at least 2GW installed for Q3.
So between 630 and 930 there should be around 5.5GW installation, mainly FRP projects, which would provide a substantial cushion to post 630 market fall.
1230: Affected by weather and holidays, Q4 is traditional slow season for PV installation. Also as previously addressed, no calendar year end is applied for any deadline mechanism. In general sense, 2GW is supposed to be installed in Q4.
All numbers through the three time points should come up with a summary of 30.5GW. We should note this calculation has ignored a larger portion of rooftop projects in pattern of self-consuming or partly sold to grid. These projects are not subject to centralized FiT and deadline policy. Central government gives no limitation on its capacity, encourages city level authorities support it with local finance and policy administration. In spite of numerous issues around this kind of project, there is definitely a boom in this field across the country. People are quite confident to see 4-5GW for 2017 installation.
After all, a more comprehensive number of around 35GW, registered and connected, is highly likely to be accomplished in China PV installation throughout 2017.
2018 China installation outlook. Many analysts are expecting a flat or declining year 2018 compared to 2017. The main declining factor comes from normal FiT projects. There are signs suggesting that 2018 FiT project could be slashed by half from 12.6GW of 2017 down to around 6GW for 2018. FiT rates are supposed to be cut off too as it should be every year until power generation becomes grid parity in natural course. PV manufacture cost reduction kept escalating over past years, the time before grid parity arrives has greatly shortened (some places around the world already have). It’s reported that central government is considering it’s time to get rid of FiT mechanism, once and for all. Those special projects, FRP, APP and so on, should stay on upward course as they provide consistent impetus to innovations like efficiency up, LCOE down, and market self correcting (tendering structure). Some experts are suggesting government take equal capacity between FiT, FRP and APP, for 2018 installation, with 6-8GW of each. The only exception is rooftop project. With no limit on capacity and supported by local financial incentives, it’s definitely heading to blossom for 2018 and years beyond. 2018 should guarantee 10GW for rooftop installation.
Therefore, 2018 should bring up around 28GW installation in conservative scenario, and 35GW bullish. This is regardless of all those time points or whatsoever deadline.
And again we should note that there is overlapping effect when counting real installation volume between 2017 and 2018, as that between 2016 and 2017. That said, quite often policies and target are announced at middle of the year, while most installations are made and registered by first half of next year. After all, the above stated figures should stay feasible as far as the fashion remains rolling that annual policy and target are announced at middle of year. For example, 2017 is affected by 2016, it will affect 2018 too.
1) Post 630 fall of demand might be migitated in 2017, by 930 effect as above mentioned.
2) High efficiency technologies see increasing boom among Chinese factories. NEA officials seems to be still implementing in favor of mono products, resulting in constant price strength of mono products going forward. Late May NEA published 2018 guidance of FRP and APP projects, reinforcing requirement of high efficiency throughout cell and modules (multi cell 19.5%, module 17%, mono cell 21% module 18%). Clearly, traditional cell and module of multi type, say 4BB and double print, are almost totally out of league for FRP and APP projects. This would expedite the course of PERC becoming a standard process and, more value added process becoming ubiquitous (MBB, bi-facial, half cell, lapping cell).
In previous posts, we talked about multi-featured HE process BS (black silicon), and the fact that its combination with PERC process was stagnant due to nature of LeTID effect. Now that multi cell factories seems to have no choice but make PERC a standard process.
It’s interpreted as Chinese government’s ambition to lead PV world in technology rather than scale and price.
3) 2017 marks a substantial change that high efficiency replaces cost as key survival strategy for PV factories. Market is increasingly centralized toward HE products. Regardless price pressure and overcapacity, high efficiency producers always see strong demand and good GM. Low efficiency factories, or those not quick enough to upgrade technologies, are faced with increasing risk of being pushed out of business and investment going down the drain.
4) Price pressure remains fierce after Q3. Clearly expansion remains steadfast and fervent among Chinese factories, as many citing they must be ready for surging demand driven by fast approaching grid parity. In spite of government efforts to smooth down demand curve, people still expect great price squeeze in 2H. Because demand is almost flat YoY while capacity’s got another 15% increase. Some common thoughts of component prices in Chinese market after 630, lowest point around Aug. and Sept.:
Poly price down to $11/kg, multi wafer $43c/pc (DW), multi cell $17.5c/w, multi module $30c/w.
Top makers are supposed to have 10-15% GM from above prices. Mono products are still motivated by increased volume of FRP and APP projects. Tight supply could continue until large new capacity comes online (supposed to be 1H2018). Therefore mono price trend remains low predictability.
By Xiaodong(Bond) Wang. Xiaodong is a solar energy enthusiast and a regular contributor.
Disclaimer: The views above are strictly those of the author and do not reflect those of any other organisation or individual or this publication.