For a long time before, the silicon wafer size was 156 mm. The industry has implemented this standard. The normalized standard is conducive to the reduction of industry costs. In recent years, the competition for single polycrystals has become fiercer. And the front-runner project has power requirements, so everyone has started the 'article' in the size of the silicon wafer, but due to the individual companies fighting each other, the size of the silicon wafer is complicated, 156.75, 157.25, 157.4, 157.75, 158.75 , 161.7 and even 166.7 and other specifications. And different side length of silicon wafers correspond to a variety of sizes of lead angles, a variety of sizes of component versions. Complex and non-uniform specifications increase the company's production costs, enterprise The cost of communication and the design cost of the power station owners. Due to the many production and exchange costs of the silicon wafer standards, the industry does not meet the general trend of PV parity, so the industry needs a new, unified silicon size standard to eliminate The above-mentioned costs, but different specifications have different benefits for different costs. Different companies have different production equipment, and the adaptability to new specifications is different. What size is the size? To meet the needs of everyone, the size that all parties can accept? This is the topic to be discussed in this article.
First, what are the benefits of large silicon wafer size?
The reason for the large size of some silicon wafers in some enterprises is: the intense competition of single polycrystals and the power requirements of the front-runners, so the 'large size of silicon wafers' was deducted by some friends, 'opportunistic' is not very glorious. Hat. However, in fact, the large size of silicon wafers has its realistic economic significance.
1, cell production line capacity is calculated according to 'flux', for example, the latest generation of cell production line can print 5,500 batteries per hour, the production cycle is calculated according to 1.3 seconds / Zhang × 2 (dual-track equipment), Within the allowable range of production equipment, a larger silicon size means a higher power battery output. Although the output of a single production line is still certain, the power of a single battery is improved. Amortization of manpower, depreciation and other consumption is consistent, but the a tile of amortization is lower.
2, the component link is the same, the stringer and laminating equipment are also in accordance with the 'quantity', within the scope of production equipment, larger silicon, battery means higher production efficiency.
3, the accuracy of the old generation of production equipment is not high enough, leaving more gaps between the cells in the standard setting to leave enough fault tolerance for the equipment with insufficient precision. With the leading intelligence, Jiejia Weichuang as the representative A series of excellent battery equipment, component equipment manufacturers supply generations of cheaper, high-precision equipment, we need less fault-tolerant space, the gap between the cells can be reduced, Longji shares launched M2 size specifications ( The size of the silicon wafer is changed from 156mm to 156.75mm). It is achieved by reducing the gap between the cells without changing the size of the components. This is equivalent to the power increase within the unit area. The potential mining in this area seems to have not come to an end. JA Solar recently achieved a silicon wafer size of 157.4mm without changing the size of the components.
Through the above examples, we found that the large size of silicon wafers is by no means opportunistic. After calculation, the silicon wafers are improved by not significantly increasing the cost of the battery, not significantly changing the size of downstream components, and increasing the construction cost of downstream power plants. Dimensionalization can increase the power of components by 1~3% (different silicon specifications correspond to different lifting ranges). Although this improvement is negligible, it is imperative to increase the size of silicon wafers when battery efficiency is becoming more and more difficult. From these perspectives, the large size within a certain degree is a 'micro-innovation' to tap the potential of existing equipment.
Second, what are the obstacles to the large-scale silicon wafer?
Moderate silicon wafers have many advantages in large size, but they are not without end. As the size of silicon wafers increases, the number of obstacles increases. The size of silicon wafers increases to 158.75. It is necessary to replace the flower baskets, fixtures, etc. In terms of battery equipment capacity, the technical transformation cost is about 400,000/100 MW; when the size is further increased to 161.7 mm, the battery production line needs to be replaced more, and the technical transformation cost is increased to 600,000/100 MW; When the size of the silicon wafer is larger than 161.7mm, it exceeds the tolerance of the existing PECVD equipment. At this time, the economical significance of the technical transformation is lost for the older battery production capacity. For the 166.7mm ultra-large size silicon wafer, it needs to be customized. From the perspective of standard setting, a standard should be adapted to the needs of different companies as much as possible. If we think that this round of large-scale process is to tap the potential of the equipment without significantly changing the size of the components. 'Micro-innovation', so too large silicon size is over-innovation.
From the component link, the silicon wafer size increases from 156mm to 157.4mm without increasing the component size; 158.75 silicon wafer requires 60-type component length increase of 1.4cm, widening 0.8cm; 161.7 size silicon wafer requires 60-type component length increase 4.5cm; width increased by 2.6cm; size of 60-type component using 166.7mm silicon wafer increased by about 12cm and width increased by 6cm. Historical experience: Asia-Pacific region favors 60-type components, and Europe-Pacific region has 72-components. Behind the phenomenon is the difference in the physique of the installation workers. The current component size is determined by years of experience. We can lengthen the module size by 12CM overnight, but it is impossible to make the average height of the installer grow 12cm at night. Component size Too significant changes will lead to many obstacles in the acceptance process of the end customer. Not only is the terminal acceptance problem, but with the increase of the component size, we need larger photovoltaic glass, larger package frame and film. More expensive shipping costs and installation costs, more robust photovoltaic brackets and a larger footprint. So we can see that the silicon size is large enough It brings the marginal benefit is zero, and because of non-standard components provides increased design costs, when the die size exceeds a certain level, and may even benefit the large size of the marginal negative situations arise.
Third, the choice of next-generation monocrystalline silicon wafers from the perspective of 'micro-innovation'
Familiar with my friends should know that I like to think about problems from the first principle, write some articles on photovoltaic industry research methodology, and also write about the photovoltaic philosophy series. 'Micro-innovation' is the three words that I have repeatedly emphasized today. It is also my understanding of the large-scale nature of this round of silicon wafers. Compared with the previous round of silicon wafer size from 125mm to 156mm and thus revolutionizing the cost of photovoltaic power generation, there is no significant change in wafer size in the industry. Rather, it does not change the potential mining under the existing industrial structure. It is a kind of micro-innovation and micro-progress. When we choose the next-generation silicon size, if we can understand it as 'micro-innovation', then many problems It is also easy to solve, 166mm size silicon is naturally an option to eliminate first, because this standard significantly changes the size of the component, the end customer must have a strong rebound so that it can not become the mainstream standard. Similarly, I am not particularly optimistic about 161.7 This size, after all, the component length increased by 4.5cm has been considered very significant.
Things always have two sides. While the industry can't accept too much change, too small changes are easy to ignore. Since it is 'micro-innovation', to increase input costs, we are in the new generation of silicon wafer size. The choice should at least make the improvement that can be perceived by the end customer. Recently, the size of the polycrystalline component is 157.25, which is only for the higher ratio of components to achieve 275W or more. It is not an innovation. So if the change is too small, it may not be innovative, because the end customer may not have obvious perception, and naturally it will not bring a significant premium on the component side.
In the perspective of 'micro-innovation', the wafer size change is not too big (the industry accepts problems), and can't be too small (the progress is not obvious consumers can't perceive). Under the constraint of both parties, we find that the choice seems suddenly clear, directly Let me draw my own conclusions: I think the next generation of more suitable monocrystalline silicon wafers is 160mm thick, 158.75mm side square monocrystalline silicon wafer. The reasons are as follows:
1, the 158.75mm side length of the silicon wafer is only 2mm longer than the existing 156.75mm silicon wafer, the existing full capacity can be upgraded through technical transformation, and the technical transformation costs are reasonable and affordable. According to a large enterprise battery technology director The data I provide, even for very old battery capacity, 100MW technical transformation costs can be controlled within 400,000.
2, the area of the square single crystal silicon wafer of 158.75mm side length (25,197mm2) is 3.14% larger than the area of the current M2 specification single crystal silicon wafer (24,429mm2). According to my understanding from Ai Xu, now M2 Under the size of 2019, the expected mainstream power can reach 315~320W, then the cell area increases by 3.14%. The theoretical increase can reach 10.05 watts, which is equivalent to two components of the module. At the level of 'significant', consumers can clearly perceive this improvement. If this size of silicon wafer can be promoted and superimposed on battery technology, we are expected to see the mainstream power of 60-type single crystal perc components by the end of 2019. Close to even reach the level of 330 watts.
3. The package size of the 158.75mm side-length battery pack is only increased by 1.4cm, which is equivalent to an increase of 1% according to the percentage, and due to the application of the 160um thinner silicon wafer and the corresponding thinner EVA, Although the component area is increased by 1%, the overall weight of the component remains unchanged. The adverse effects are reduced to a minimum for installation and transportation.
4. The cost of pulling the crystal link and the silicon material of the 158.78mm square single crystal silicon wafer has increased, but the cost reduction from the thickness of the silicon wafer from 180um to 160um is more significant. Therefore, the comprehensive calculation: production of 158.75mm, The cost of a square silicon wafer with a thickness of 160 um will be 8 cents lower than the cost of a 180 um thick M2 silicon rod (which will be measured in detail later).
To sum up four points: 158.75mm side length, 160um thickness, 224.5mm diagonal square monocrystalline silicon wafer has the advantages of lower production cost, constant component weight, two power boosts, and high battery end acceptance.
Fourth, square monocrystalline silicon wafers are a must for 'micro-innovation' perspective
This section first spreads some basic knowledge, and professionals can skip it directly.
The size of the silicon wafer involves not only the length of the side but also the shape. So the topic discussed today involves not only the side length of the silicon wafer, but also the shape of the silicon wafer. In the past, the long crystal cost of the single crystal silicon rod was very high, silicon The price of the material is also high, so in order to minimize waste, the silicon wafer will have four 'conducting angles'.
These four lead angles are ultimately reflected in the component is the package blank.
The whitening of component packaging is essentially a waste. In the era of high cost of pulling crystals and expensive silicon materials, the white space of the package is economically rational, and the time comes to the present, the price of silicon material has been compared with last year. It has fallen by half, and the cost of single crystal pulling has also dropped significantly. The promotion of diamond cutting and flaking has significantly reduced the cost of slicing. Can we eliminate the monolithic wafer lead angle and eliminate the encapsulation of single crystal components? What is the historic moment of white?
We further quantify the problem: To produce a 158.75mm square monocrystalline silicon wafer, the growth diameter of the silicon rod needs to increase from 215mm to 228mm, which increases the cost of the crystal growth and silicon. But at the same time eliminates the monocrystalline silicon wafer. The lead angle can increase the effective utilization area of the component, and the downstream battery and component links can increase the power of the component without adding any cost, which brings value. So to answer whether to use the square single crystal is essentially to answer: square single Is the value of crystal power increase significantly greater than the cost of pulling crystals and silicon material? In fact, this can be broken down into two problems:
1. How much will the cost of producing square monocrystalline silicon wafers increase?
2. What is the value of power boost caused by square monocrystalline silicon wafers?
A principle before calculation, two hypotheses:
Rationale: Future-oriented analysis is based on future-oriented assumptions.
Two hypotheses: Since I think that the future P-type monocrystalline silicon wafer will transition to 160um thickness, and according to the analysis in the third section of this paper, the silicon wafer side has the potential to become 158.75mm, so we are all based on 160um thickness in the following calculations. And two future-oriented assumptions of 158.75mm side length.
Answer to question 1: The cost of producing square monocrystalline silicon wafers will increase by 3.78 yuan / 60 pieces.
The cost of producing a diagonal length of 210mm, a side length of 158.75mm, and a thickness of 160um with a single-angle silicon wafer is: 2.358 yuan / piece, the calculation process is as follows:
The cost of producing a diagonal length of 224.5mm, a side length of 158.75mm, and a thickness of 160um square monocrystalline silicon wafer is: 2.421 yuan / piece.
(Note: The calculation assumes that the 215 diameter round bar costs 34 yuan per kilogram of pink crystal (pink portion), while the 228 diameter round bar 1kg has a long crystal cost of 32 yuan, so this assumption is due to the 228 diameter round bar. The cross section is larger, and the contact surface of the crystal growth is larger, which can bring about a reduction in the cost per unit length of the crystal)
Answer: Introducing the size of a square single crystal on the 158.75 size will increase the cost per single crystal wafer by 2.421-2.358=0.063 yuan, and the cost per component will increase by 0.063×60=3.78 yuan. This is the need to introduce a square monocrystalline silicon wafer. the cost of.
The answer to question 2 is the value of the power boost brought by the square monocrystalline silicon wafer?
The effective area of the 158.75 square monocrystalline silicon wafer (25,193mm2) is 0.84% larger than the effective area of the 158.75 angled single crystal silicon wafer (24,983mm2), corresponding to the power of the main crystal perc 315 watts next year, 0.84% The increase in area can result in a power boost of 315 × 0.84% = 2.65 watts. Component power boost has two aspects of value, one aspect is the more effective amortization of the area-related cost of larger power components (again, Higher power components can be sold more expensively; on the other hand, the added power itself has value. The two values are:
Under the assumption of an area-related cost of 500 yuan, the component-related amortization value of the component power increase of 2.65 watts is: 500÷315 — 500÷317.65=0.0132×315=4.17 yuan.
According to the expected price of 1.9 yuan perper component in 2019, the value of 2.65 watts is 2.65 × 1.9 = 5.03 yuan.
Therefore, the value brought by the elimination of the lead angle is 4.17+5.03=9.2 yuan.
Conclusion of this section:
For a 60-type component, the cost of eliminating the lead angle of the single crystal silicon wafer is 3.78 yuan, and the value of the power boost caused by eliminating the lead angle is 9.2 yuan, which brings the value significantly greater than that of the Increased cost, rational analysis, the development of a new generation of monocrystalline silicon wafer size to eliminate the single crystal silicon wafer lead angle, the comprehensive promotion of square single crystal is a must. And from a long-term perspective: square monocrystalline silicon wafer is conducive to future promotion stack Tile technology is the standard preparation for the next round of component packaging technology revolution.
Going back to the topic of this section, I learned from the technical leaders of the leading companies in PV shipments such as Jinko, Tianhe, etc.: For the battery and component parts, there is no obstacle to using square monocrystalline silicon wafers, no need to increase any cost. It can bring power boost. The only thing that needs to change in the promotion of square monocrystalline silicon wafers is the growth of the thicker silicon rods. From Longji, there is no technical difficulty. The changes that need to be made in the industry are very Less, the value is significantly greater than the cost, and the square monocrystalline silicon wafer is a typical 'micro-innovation' activity. In turn, it is the most unobstructed choice under the 'micro-innovation' approach, the easiest change to make. Therefore, I will say in this section: Square monocrystalline silicon wafers are a must for 'micro-innovation'.
Conclusion of the article:
The next-generation monocrystalline silicon wafers are 158.75mm in length, 160um thick, and 224.5mm diagonal monocrystalline silicon wafers. I have come to a conclusion based on a series of thoughts and summary information. Under this specification: Wafer cost No increase or even slightly lower than the current 156.75 180um thickness of the single crystal silicon wafer, the component size does not change greatly, the component weight remains the same but the component power is expected to increase two gears (10 watts). If this size can Popularized in 2019, the power of the 60-type single crystal perc module is expected to reach 330 watts by the end of 2019, thus continuing to maintain the high speed of component power of more than 10 watts per year.
At the same time, I also want to emphasize that this specification is only my own thinking, and it does not mean that it must be the most reasonable, the best, I have detailed the analysis process and basic data in this article. Conclusion is not the key The process and method of thinking are the key. I hope that I can give you some reference on the selection of the next-generation silicon size standard in thinking methods. If this article can be used to think about the future choice of monocrystalline silicon size. Ding points, I am satisfied.