The advantages of module products based on M10 (182 mm) wafers was the topic on the first day of the TaiyangNews Virtual Conference on High Efficiency Solar Technologies held on Dec 1, 2020. LONGI Solar, key promoter of this wafer format, presented an elaborated comparison of 182 mm wafers based modules over the ones using 210 mm. Hongbin Fang, director of product marketing at LONGi Solar, in his talk titled Value Proposition of High-Power Bifacial PERC Modules Based on M10- 182 mm Wafers methodically explained the rationale behind selecting the 182 mm wafer size as optimal and provided a cost comparison of M10 technology with G12.
Fang stressed on the benefit of high module power realized in terms of BOS savings by increasing the power per string that is deduced from savings on the trackers and electrical components. At the same time, he emphasized that other aspects of the LCOE equation such as module reliability and energy yield cannot be neglected. “Otherwise, the value of high-power modules diminishes he noted. Therefore, a new module design should ensure minimum reliability risk, controlled manufacturing cost, superior energy yield and, of course, low BOS costs.
Historically, significant module power improvement has been achieved through cell efficiency improvement, such as the shift from multicrystalline to monocrystalline and switching the cell technology from BSF to PERC. Advanced module technologies such as half-cut and MBB have also helped in increasing module efficiency. Approaches such as improving the AR coatings on glass and using white EVA have helped in improving module performance. The industry then explored other avenues such as increasing wafer size progressively from M0, M2, G1, M4 and M6.
Discussing the introduction of M6, Fang said that LONGi has analyzed the existing cell and module production line compatibility and evaluated the upgrade costs. “We came to the conclusion back then that M6 is a good size for the whole industry,” said Fang. Modules based on M6 are also suitable for both residential and commercial scale C&I applications as well as utility scale power plants. LONGi has successfully ramped up its manufacturing for M6 and shipped more than 10 GW of Hi-MO4 modules, since their introduction in 2019.
Explaining the evolution of G12, Fang emphasized that the wafer size is derived from the semiconductor wafer diameter of 300 mm. LONGi did analyze if G12 would be optimal, keeping only large scale power plants in view, and not taking into account the constraints from the side of existing production tools and non-compatibility with the rooftop applications, which are important. He first explained the difference between PV and semiconductor in a slide titled, Is it necessary to unify semiconductor and PV silicon? In semiconductor manufacturing, a wafer is cut into multiple chips, whereas a PV wafer more or less retains its size even as part of the module. Thus, there is a need to evaluate the whole process of manufacturing and module deployment to ensure that there are no compatibility issues.
According to LONGi’s analysis of first generation module products based on G12 wafer format, the 5-row and triple-cut design is a workaround for application challenges, but is still a compromise with reduced benefits. Slicing cells into three pieces reduces module current, but the higher voltage limits the BOS savings. The 5-row design brings module width in line with the industry norm, but the extra returning ribbon to complete the circuit results in additional power loss and module efficiency reduction. Moreover, slicing the cells into three pieces introduces additional complexity into the production process.
Size matters: In a boundary condition analysis, LONGi found that the container opening height
determines the ideal module width – and that’s why it opted for 182 mm wafers. (source: LONGi)
Then came the G-12 half-cut module with a six-row configuration, said Fang. But it has higher currents exceeding18 A and the same module in a bifacial configuration has even higher Isc exceeding the 20 A level. Such module has higher resistive losses, increases the risk of hot spot and may cause junction box failures, according to Fang. He further emphasized that the width of these modules is about 1.3 m, which reduces the packing density of a 40- feet container, has limited glass supply apart from increased concerns of mechanical loading strength of the modules.
“Looking at these limitations, we try to evaluate what is the optimal module size when designing from scratch,” said Fang. LONGi started examining the whole manufacturing process, including production tools all along the value chain, materials supply and manufacturing yield. Mechanical loading and hot spots were the topics of interest on the reliability side.
LONGi found the key limiting factor in module transportation to be the container opening height of 2,570 mm. The current best-known method for loading the modules in a container is to pack the modules on a pellet vertically in landscape orientation, while a double stack of pellets is placed in the container. Eliminating the margins for operation and pellet packing gives the ideal module width, which is 1,130 mm. Considering a 6-row module design, the back calculation leads to 182 mm as the most fitting wafer size. This is how we derived at the ‘best wafer size of 182 mm,’ emphasized Fang. LONGi also checked the compatibility of this wafer size through the whole value chain, manufacturing feasibility and reliability and found no challenges. The company also placed special emphasis on hot spots by measuring hot spot temperatures for modules based on M2, M6 and M10 wafer formats. The hot spot temperature for these PV panels increased from 118.4 °C for M2 to 156.6 °C for M10 and corresponding power loss is given as -63 W and -89 W. Extrapolating these results to the 210 mm based module raises serious concerns about high hot spot temperatures and losses, according to Fang.
Fully convinced about the advantages of the M10 wafer size, LONGi has introduced its Hi-MO5 module series. LONGi, in cooperation with other leading integrated module makers such as JinkoSolar and JA Solar, has started an initiative to adopt the M10 silicon wafer size as a new standard. As part of this effort, the company has addressed all those concerns raised earlier with 210 mm wafer based modules. These 3 manufacturers with substantial capacities have chosen M10 as their main product and are also vouching for its future. This should minimize the supply risk at the wafer, cell and module levels. M10 and M6 can coexist serving different applications, LONGi pointed out. At the system integration level, string inverters with minor upgrades to 15 A are compatible with 182 mm modules and, as mentioned above, the tracker length is also ideal fit for the M10 wafers, according to LONGi. There is also ample glass capacity to support the 1.13 m width of the module with new capacities coming online, while the old glass facilities can also be upgraded, according to Fang.
Fewer losses: Solar modules based on M10 wafer size feature lower resistance losses compared
to G12, according to LONGi Solar. (source: LONGi)
LONGi has evaluated the costs of M10 wafer processing against G12 at all manufacturing stations. According to Fang, wafering costs are slightly higher for G12 by about 0.15 to 0.3 US cents/W due to relatively slower pulling rates required for growing ingots with larger diameters and also due to the complexity involved in controlling oxygen content. The wafer thickness needs to be slightly higher to maintain mechanical strength due to a larger surface area.
G12 provides a better scale of manufacturing, meaning the throughput per tool (in MW metrics) is relatively high. When the stage is all set up, the additional costs incurred in wafering can be made up at the cell lines in processing 210 mm wafers with nearly the same benefit of about 0.15 to 0.3 US cents/W. But the scenario again reverses at the module level. Due to the high currents, the resistive losses are relatively high with G12 modules, negatively impacting efficiency, said Fang. Higher module width, requiring higher frame strength, also increases frame costs. And junction boxes to handle higher currents from the larger wafer based modules are not only expensive, but also not mature. According to Fang, such a module in bifacial configuration requires a junction box of at least 30 A, for which the technology is not yet mature and has very little safety margin. Overall, Fang claimed the costs of G12 modules are higher by about 0.6 US cents/W.
Echoing nearly the same benefits with BOS costs as emphasized by JinkoSolar’s Yu, Fang mentioned handling advantages of M10 modules during the installation. The 1.1 m module width is still within the stretching limit of installation personnel and the lower weight also aids in ease of handling.
Assuming the same module price and degradation, but factoring in the BOS benefit of 0.1 US cents/ W and 1 °C higher module operating temperature, according to LONGi’s LCOE calculation, 182 mm wafer based modules reduce energy generation costs by about 0.5% or 2.75 US cents/kWh in absolute terms compared to 2.76 US cents/kW with 210 mm wafer based modules.
The text is an excerpt from the TaiyangNews Report High Efficiency Solar: Featuring PERC Modules with M10/182mm Cells Towards Very High Efficiency Products –A Conference Summary, which can be downloaded for free here.
A comprehensive overview on Value Proposition of High Power Bifacial PERC Modules Based on M10-182mm Wafers was provided by Hongbin Fang, LONGi Solar, during the TaiyangNews High Efficiency Solar Conference; the recording can be viewed here.