When my TaiyangNews colleague Shravan Chunduri and I were working on our first PERC report that was published in April 2016, it was still about describing the big picture, the potential and challenges of this nascent technology, which end of 2015 had an installed production capacity of less than 5 GW. Now that we are publishing our "PERC 2017 Edition – The Lastest on Solar's Hottest Cell Technology" things are much different. With an estimated 12 GW of production (new lines and upgrades) capacity added by end of 2016, the answer is clear. PERC is fully established.
Looking at the Q1/2017 order-books of the leading manufacturers, it is obvious that this trend will continue. We estimate around 20 GW of capacity will be reached by year-end, which would be equal to nearly a quarter of total installations forecasted for 2017. It is unlikely anyone adds a new p-type mono line that is not capable of producing PERC; and very soon any of those p-type mono lines that are still good enough will be upgraded to PERC.
And what about multi PERC ? It might tell a lot that the first cell makers – SolarWorld, NSP – have turned their back completely on solar's cell workhorse. SolarWorld's situation is different – it is smaller than the top large Chinese competitors and it does not produce in low-cost countries with local access to low-costs processing materials. But also the world's largest wafer manufacturer, GCL, which originally was only producing multi wafers, has started mono-wafer production. And even QCells, a pioneer and largest manufacturer of multi PERC cell technology, is expanding into the mono variant. But there is a lot of activity among wet-etching equipment companies working on – and the first even offering – process solutions for black silicon, which would enable usage of diamond wire technology on a large scale for multi and reduce cost significantly. If black silicon untangles the knot, wafers might compete again against mono wafers on cost – and that would boost multi-PERC again.
The material of choice for PERC passivation is without doubt aluminum oxide. The equipment of choice is time-tested PECVD, which continues its successful PERC march. The market leader, Meyer Burger, just introduced a new PECVD in-line tool, which offers passivation for the front and rear in one go, combined with anti-reflection coating. PECVD tube furnaces, mainly used in the beginning of PERC commercialization for silicon oxynitride deposition, is experiencing a second spring in the AlOx field. And even ALD, the newcomer technology in solar cell processing, has been finding a notable number of customers. But it's likely that the International Technology Roadmap for Photovoltaic (ITRPV) estimate comes true, which forecasts ADL's share of usually less than 10% in the coming decade.
It's not that the sky is the limit for PERC efficiency potential – but there's still room towards the 24% efficiency ISFH sees if a number of not too complicated improvements will be applied. Trina Solar improved its world record for a commercial-size champion cell to 22.61% in December 2016, that's almost 0.5% absolute better than its previous record from 22.13%.
Although mass-produced PERC is still young, there is already the second generation in production. SolarWorld uses selective emitters for 5-busbar cells that have elevated efficiencies beyond 22%. That's the second step of the cell improvements proposed by ISFH to reach 24%. We are also seeing PERC as a means of propelling bifacial glass-glass module technology, which panel suppliers are now increasingly starting to advertise.
Does that mean PERC is the upper end of traditional c-Si cell evolution? For sure not, but the more this rather low-cost high-efficiency technology is improved – and this will happen for sure, the longer it will take for others to get decent shares in the market, including 'simple' n-type as well as heterojunction cell technologies.
The PERC 2017 Report can be downloaded free of charge here.