Careful optimization of laminate pre- and post-heating temperature enables fabrication of bubble-free thin-film Gen 8.5 PV modules
Han-Wen Chen, Jeff Ye, Ofer Amir, Applied Materials, Santa Clara, CA, USA
From the March/April 09 issue of Photovoltaics World magazine
Photovoltaic (PV) module packaging is crucial in preventing degradation of module efficiency due to environmental exposure such as heat and humidity. The major challenge in encapsulation for large-sized modules is de-airing across locally stepped topology from cross- and side-buss ribbons that collect power from individual solar cells on the front glass.
Full-surface lamination technology, which seals the entire photo-electrically sensitive film stacks, is currently being implemented for 5.7m2 (Gen 8.5) sized modules. Initial lay-up temperature, laminator heating/pressing uniformity, and glass alignment effects were optimized for the process to achieve bubble-free results on solar module laminates after the autoclave process.
|Figure 1. Lamination lay-up.|
Key components of solar module lamination lay-up (Fig. 1) start with front glass at the bottom with photo-electrically sensitive films, back reflective coating, and cell-defining scribes. Then, power-collecting ribbons are bonded to the glass, followed by pairing of polyvinyl butyral (PVB) sheet and back glass. Various encapsulation techniques exist already in automotive and construction material industries. However, Gen 8.5 modules present major challenges due to the large size (2.2m × 2.6m) of the panels and the cross- and side- buss wires on the glass that make the de-airing process particularly sensitive. A newly developed laminator use a two-stage heating-pressing approach. A pre-nip oven heats and softens the PVB sheet in preparation for the subsequent de-airing process by a first set of nip rollers. Then, the laminate is further baked to a higher temperature in the main heating ovens and pressed by a second set of nip rollers to seal the front and back glass panels. Autoclave processing subsequently is performed to evenly dissolve residual air into PVB. Ideally, the PVB layer should be transparent without bubbling after autoclave.
Innovative solutions are required for large-size lamination applications due to the increased degree of technical difficulty arising from scaling up of the conventional sized 1.1m × 1.3m (Gen 5.0) process. Improved full-surface lamination technology enhances the durability and long-term reliability of large-size commercial solar modules.
Annealed back glass (with a 50mm hole at center) and 3.2mm-thick Gen 5.0 and Gen 8.5 transparent conductive oxide (TCO) front glass were used for lamination process development. Insulation tape, and cross- and side-buss wires were applied on the TCO side of the front glass panel. Then, 45gage (1.14mm-thick) PVB was cut and put on top of the front glass, followed by pairing with a back glass panel. Excessive PVB around edges and at the center hole was trimmed before the lay-up was sent to the laminator. Laminates were further processed through the autoclave.
Various factors, including initial lay-up temperature, PVB heating and nip roller pressing uniformity, air bubbles in PVB resulting from high process temperature and premature sealing, and laminate breakage due to misalignment between front and back glass, were experimentally investigated to characterize the laminator’s process window.
Initial lay-up temperature. Table 1 shows actual laminate top/bottom surface and PVB temperature under the same lamination process conditions, with different incoming temperatures. Data shows 7°–12°C higher laminate surface temperature on the lay-up with 9°C higher incoming temperature. However, corresponding peak PVB temperature during the lamination process is only 2°–3°C higher. The initial temperature of a lay-up has more impact on laminate surface temperature than on actual PVB temperature, which should be taken into account for lamination process optimization.
PVB heating uniformity. PVB temperature uniformity is crucial to the de-airing press after pre-heating and also to the sealing press subsequent to the main-heating. Temperature stripes were applied to the surface of a PVB sheet between front and back glass panels to record the highest temperature experienced at designated spots over the entire laminate during the oven heating process. Figure 2 shows the PVB temperature distribution profile across the laminate width. PVB temperature variation of <±3°C, which is within the resolution of the temperature stripes, was recorded in both the pre- and main-heating ovens.
Figure 2. Direct PVB temperature measurement with temperature stripes on a Gen 5 solar panel: a) Temperature dot set-up and reading after pre-heating chamber; b) PVB temperature distribution across the width of the Gen 5 laminate after pre-heating the oven.
Heating oven temperature. The impact of overheating the PVB before the de-airing process in the oven is shown in Fig. 3. Using excessively high temperature in pre- and main-heating ovens causes more air to be dissolved in the molten polymeric material, and the edges of the laminate seal early during subsequent pressing. This prevents air from being squeezed out and results in the non-uniform lumps seen in Fig. 3.
Figure 3. A Gen 8.5 panel laminated after overheating.
Nip roller pressing uniformity. Pressure-sensitive tapes were used to check nip roller imprint span across width of a glass panel under different line pressures (pressing force per unit length across laminate width). Table 2 displays nip roller footprint results normalized by nip roller diameter on a) Gen 8.5 and b) Gen 5 glass panels, with pressing profiles summarized in Fig. 4. For Gen 8.5 lay-ups, line pressure ranging from 200N/cm to 400N/cm showed ~40% better pressing uniformity across the width, while a wider process window is indicated from the data on Gen 5 panels.
Figure 4. Normalized imprint profile across Gen8.5 and Gen5 panel width under various line pressures.
Glass alignment. Glass alignment is closely linked to edge breakage during nip roller pressing under high module temperature in the lamination process. Edge misalignment was susceptible to breakage during the main-heating and nip-roller-pressing processes. In Fig. 5, a 5mm extrusion of the back glass relative to the front panel on the laminate trailing edge caused the breakage shown in the figure.
|Figure 5. Laminate trailing edge breakage due to glass misalignment.|
In bulk areas without thickness variation, post-lamination panels showed a uniform stripe-like pattern while cloudy regions were associated with the cross-buss, where a topological step on the front glass exists (Fig. 6). Quality of the de-airing process can be estimated visually by the width of the region around the cross-buss, with narrower regions being an indicator of a better de-airing process. However, the quality of the laminate can be confirmed only after the panel has gone through autoclave.
Figure 6. General appearance of laminated panels with buss ribbon attachment.
Full-surface lamination of PV modules prevents degradation of module efficiency due to exposure to environmental conditions, such as heat and humidity. Although the automotive and construction industries have used various encapsulation techniques for a while, large-sized solar modules present a major challenge due to the locally stepped topography from cross- and side-buss wires on the glass. For the 5.7m2 modules, there is an increased degree of technical difficulty arising from scaling up of the process from the smaller, conventional-sized modules.
Fabrication of bubble-free thin-film Gen 8.5 PV modules was accomplished by careful optimization of laminate pre- and post-heating temperature, nip roller line-pressure profile along the module’s length, and conveyor speed at leading edge, cross-buss area, and trailing edge, for effective de-airing process, which is crucial in enhancing durability and ensuring long-term reliability of the modules.
The authors thank Fang Mei, former technology development engineer of the Thin-film Solar Business Group at Applied Materials for contributions to this article. Next-generation laminators used in the study are SunFab Laminators from Applied Materials.
Han-Wen Chen received his PhD in mechanical engineering from the U. of Michigan, Ann Arbor, and is a key account technologist in the Thin Film Solar Business Group at Applied Materials, Santa Clara, CA USA; email firstname.lastname@example.org; www.amat.com.
Jeff Ye has a PhD and MSc in mechanical engineering from Queen’s U., Kingston, Canada. He is a senior engineering project manager in the Thin Film Solar Business unit of Applied Materials.
Ofer Amir received a BS in mechanical engineering from the U. of California, Berkeley, and is a mechanical engineer in the Thin Film Solar Business unit of Applied Materials.