Technology evolution of the photovoltaic industry: Learning from history and recent progress

Abstract

AbstractThis paper reports on the latest advances in crystalline Si cells and modules in the industry and explores the dynamics shaping the silicon PV industry. First, we report on the recent efficiency improvements of passivated emitter and rear cell (PERC) and tunnel oxide passivated contact (TOPCon) cells on 210 mm wafers. At Trina Solar, the best batch average cell efficiency (total area) reached 23.61% for PERC and 25.04% for industrial‐TOPCon (i‐TOPCon). As far as we know, these are the highest values achieved on 210 mm wafers. The best champion efficiency for PERC and i‐TOPCon is 24.5% and 25.42%, respectively, as independently confirmed by the National Institute of Metrology of China in Beijing and ISFH CalTech in Hamelin. We have developed modules with power outputs of up to 660 W by using 66 pieces of these 210 mm cells with 12‐busbar technology in mass production. Besides, the aperture efficiency of the best laboratory PERC module fabricated by Trina Solar is 23.03%, which was independently confirmed by TÜV Rheinland. As far as we know, this is the first commercially sized PERC module with an aperture efficiency of 23% and a power output of over 600 W. Second, we have examined the technological development in the PV industry and summarise some empirical results. A look at the historical data shows that an increase in wafer area of at least 50% is required for a wafer size to become a new industry standard that lasts for 10 years. We find that it typically took about 3 years for the average efficiency of a cell in mass production to reach the efficiency of the champion cell produced in the industrial laboratory. We apply the empirical Goetzberger equation to analyse the module efficiency of c‐Si and thin‐film technologies. Based on our previous work, we update the selling price and manufacturing cost of PV modules and their learning curves. If we restrict the module price learning curve to the years starting in 2015, we find a short‐term learning rate (LR) of about 40%, while the overall LR since 1970 is about 24%. A strong LR is driven by collaboration among industrial players and clustering of the industry, as well as standardisation of the technology, the supply chain, and final product design, which lead to fast equipment development and fast increase in capacity of supply chain. We propose an empirical law to describe the recent evolution of equipment LR, which shows that the throughput of tool increases 100% in every 3 years, so that the investment in cell production lines has decreased by 50% every 3 years since 2015. Finally, we quantify the material consumption and carbon footprint of PV plants today and for the expansion of PV to terawatt (TW) levels. Besides replacing silver fingers with copper and aluminium, saving copper cables in utilities and low‐carbon mining of materials are the most effective carbon reduction measures in the PV supply chain.