The half chip module using G12 silicon chip (edge distance 210mm) can achieve nearly 600W module power. The disadvantage of this product is that the internal loss caused by high current is high, and the reliability risk of hot spot and junction box is significantly improved. The advantage is that the power of single string modules is significantly increased, thus reducing the cost of the system side. However, through the specific analysis of the system end bracket, cable, inverter, etc., it will be found that the G12 half chip module is difficult to bring value to the system end compared with the M10 (edge distance 182mm) module, which is more mature in technology. This paper will reveal the reasons behind this.
The advantage of M6 module system end is that its length, width, current, etc. are slightly improved compared with M2 and G1 modules. The equipment and design of the system end can be compatible with the original form. The design of M10 module is also based on the idea of this small increase to drive the inverter and tracking support to be slightly upgraded and downward compatible. For G12, because of the obvious changes in size and current, it is necessary to establish a supporting system, which will cost a lot in product manufacturing and product management. The following is a detailed analysis of each component of the photovoltaic system.
Support and foundation
For the fixed support system, the support cost is related to the total support area, and the increase of the total support area within a certain range is conducive to the reduction of the support cost. Specifically, for components of the same size and the same support design, the cost of two strings of components loaded on the support is significantly lower than that of one string of components, but the cost of the support tends to be flat as the number of strings loaded continues to increase, which mainly reflects the impact of the support length. In terms of support width, widening can not only reduce the support cost, but also increase the bearing power of each pile foundation, thus reducing the foundation cost of single W (assuming that the pile spacing is the same). Therefore, the BOS savings brought by widening the support are greater than lengthening. However, considering the convenience of installation, the total width can only be increased in a limited way.
According to the above background knowledge, the change of component size shows that the change of length and width of M6 component is less than 4.5% compared with M2 component, which can maintain the design of single fixed support bearing 2 or 4 strings of components. The length and width of the bracket have only increased slightly, so M6 components can fully replace M2 components in various application scenarios.
The length and width of M10 component are increased by about 13% compared with M2 component respectively. Considering that the length of single support may be limited under mountainous conditions, this product is more suitable for use in super large ground power stations with flat terrain. In flat terrain, it can also be considered to continue to lengthen the support to increase the number of strings loaded on the support to further reduce costs.
The G12 module has two designs: 55 cells and 60 cells. The string power is greatly increased by 80% compared with the M2 module. For the components bearing the same string, the bearing area of the bracket needs to be increased by 80%. Theoretically, if the field conditions support an increase of 80% in the load-bearing area of the single support, why can’t the M2 component double the support and double the number of strings?
The essence of the above problem is that the huge changes in G12 components’ parameters will inevitably affect the initial boundary conditions of BOS calculation and comparison, and it is inappropriate for the support to bear the same number of strings. Take the horizontal installation of double-sided components as an example, as shown in the following figure: G12 component (2172) of 60 batteries × 1303mm) four rows of transverse mounting brackets have a bearing width of 5.27m, which will bring difficulties to the installation of components. If it can be used to make four rows, M6 and even M10 components can be installed in five rows to reduce the cost of brackets and foundations.
Specific calculation of support and foundation cost: if the same four rows of horizontal installation, G12 components can indeed save 4-5 points/W compared with M10 components, but if the M10 components are installed in five rows, they can also save 2-3 points/W compared with G12 components. In comparison with M6 components, five rows of M6 components and four rows of G12 components have the same width for horizontal installation. The G12 component can reduce one purlin, but the specification of purlin steel needs to be improved, and the cost of the support is not high.
Flat single axis tracking support
The cost of the flat single axle support is divided into the electrical control part and the mechanical part. The analysis of the support cost of different size components can mainly focus on the mechanical part. Unlike the fixed support, the length of the main shaft of the tracking support is relatively fixed, which is difficult to adjust in a large range.
The mainstream 1P tracking support can support three strings of components (1500V system). If it is equipped with M2 components, the length of each string of 28 component supports is about 83m; The length of the support with M6 components is increased to 88m; For M10 components, the support length is increased to 95m. Because the G12 module of 60 cells is too wide, the support length of three string modules (33~34 cells/string) is more than 130m, and the corresponding support length of the G12 module of three strings of 55 cells is also more than 120m. The existing 1P support structure design is difficult to support three strings. If only two strings of modules are carried, there is no advantage in the cost of the support. In addition, if the 1P bracket is upgraded to a length of more than 130m in the future and can support three G12 components, it can also support four M6 and M10 components.
Similarly, the 2P tracking support can usually bear four strings of components, and the length of the support corresponding to the M6 component is 60m; The support length of matched M10 components must be increased to 65m; For G12 components, the same bracket length can theoretically carry three strings of components, but in this case, one string of components is arranged on both sides of the main shaft of the 2P bracket, which will lead to obvious series mismatch, so it can be considered that G12 components are difficult to match the 2P tracking bracket.
Series inverter
In recent years, with the appearance of double-sided components and M6 specifications, the single string current of inverter used in large ground power stations has increased from 11A to 13A. After the appearance of M10 specifications, the inverter current will increase to 15A. The inverter current from 11A to 15A can be realized without changing IGBT and other key components, so this upgrade can be an overall upgrade of the product line.
To adapt G12 components, an inverter with a single string current of more than 20A is required. Customizing higher specification IGBT chips means establishing an additional product system. In the future, additional costs will be required in product management, component procurement and decades of life cycle maintenance of inverters. Therefore, mainstream inverter manufacturers are in a wait-and-see state in the development of new systems. If G12 components are only a few GW level niche market, it is sufficient to make a temporary matching scheme through parallel IGBT chips.
Cable cost
The increase of module current leads to the increase of string power. Therefore, the consumption of photovoltaic cable decreases with the increase of string power, and the decreasing trend gradually slows down; The increase of current also brings about the loss of power generation (that is, the increase of line loss). The line loss itself is proportional to the square of current. In addition, considering the reduction of cable consumption, the cost brought by the increase of line loss is roughly linear with the current. As shown in the figure below, using 4mm2 cable, the best comprehensive cost point is 13~15A. For G12 double-sided module, considering the back power generation gain, its working current often exceeds 20A at noon. It is no longer suitable to use 4mm2 cable, but must use 6mm2 cable. Considering the line loss, the comprehensive cost of photovoltaic cable can be equivalent to that of M10 module.
Manual handling and installation
In principle, two people are required to carry photovoltaic modules in large power stations, and the impact of large modules weighing no more than 40kg on the flat ground is not significant. It is mainly necessary to assess the impact of module area and width on manual handling and installation. The M2 module of 72 batteries has an area of about 2m2 and a width of less than 1m; The area of M6 module is increased to 2.17m2 and the width is 1.04m. Due to small size change, M6 module can replace M2 group in various scenarios, such as mountain, water and agricultural photovoltaic. The M10 module has an area of 2.56m2 and a width of 1.13m, which is 30% larger than the M2 module. It is suitable for flat terrain. The work efficiency and installation damage rate in block terms are still guaranteed. The G12 module of 60 batteries has an area of 2.83m2 and a width of 1.3m, which is far greater than the natural spread distance of human hands. It is likely to bring difficulties to manual handling and installation.
Summary
This paper specifically analyzes various factors that affect the cost of BOS in photovoltaic power plants, and finds that G12 specification cannot bring obvious value in the system side. The reason is that the introduction of M6 and M10 specifications has considered the compatibility with the original system system, used the margin of the original design system, and denied that all G12 specifications need new systems to match. Considering that G12 specification can not save money at the manufacturing end, but also brings reliability problems such as component load, hot spot, junction box, etc., it is not a technical route worth choosing.
