The dramatic growth of PV in the 2000s and 2010s was achieved on the back of multicrystalline silicon wafer technology. So-called multi wafers and cells, while not as high-performance as their monocrystalline cousins, were significantly cheaper to make than mono equivalents due to use of low-cost directional solidification technology in the ingot crystallisation process. In PV manufacturing, the metric that matters is cost per watt of rated power. The significant cost-per-watt advantage enabled by multi led to it command a global crystalline PV market share of 80% in the early 2010s, and for the Chinese manufacturers that adopted multi to dominate the global industry.
By 2015, when Exawatt was founded, substantially all of the mainstream PV industry was dedicated to multi and most leading PV manufacturers were convinced that multi would remain the leading technology for many years to come. However, at Exawatt we disagreed. By combining our founders’ years of experience in PV manufacturing with detailed research into PV supply chains and technology roadmaps, Exawatt came to the then-radical conclusion that mono would swiftly come to dominate the PV industry, achieving 50% annual market share in five years. Our thesis was based on the following critical changes:
- Significant improvements in the productivity of the Czochralski mono ingot growth process, driven by increases in crystal growth rates and the use of larger crucibles, combined with a move from batch growth (one ingot grown per crucible used) to recharge crystallisation (multiple ingot “pulls” per crucible), significantly reducing the cost per watt of mono ingots
- A change in the wafer sawing process, moving to diamond-coated wires from the traditional combination of braided steel wire and slurry. This would increase saw productivity by raising saw speed and reducing kerf loss, leading to significant decreases in the cost per watt of mono wafers. At the time, diamond-wire sawing was only possible on mono ingots, as the traditional multi wafer texturing process was unviable for diamond-sawn wafers, giving mono a window of opportunity
- A move from the traditional and near-ubiquitous back surface field (BSF) cell technology to passivated emitter and rear contact (PERC) cells, which would confer a disproportionate performance benefit to mono cells, reducing manufacturing cost per watt at the cell and module levels
The result of these changes, Exawatt argued, would be that by 2020, the manufacturing cost per watt of mono modules would be equal to that of multi modules, but that the significantly higher efficiency of mono modules would lead to lower LCOE (levelized cost of energy) at the system level, prompting downstream developers to choose mono modules, giving mono a market share in 2020 of at least 50%.
Almost all the solar manufacturers we met with in 2015 disagreed with our forecasts, arguing that multi would remain dominant due to its significantly lower cost and “good enough” performance. In the end, we were half-right. While mono did grow its market share from 20% to 50%, it achieved this goal in mid-2018, more than two years ahead of our forecast.
For some Exawatt customers, the size and growth of mono mattered – and still matters – a great deal. For example, in 2015 the prospective market for high-purity quartz (HPQ), used to make mono ingot growth crucibles, depended critically on whether mono would be widely adopted, since multi crucibles could be made from lower-grade quartz. Exawatt was the first strategic consultancy to correctly forecast the growth of the market for HPQ in PV, and has since established a significant HPQ practice, helping customers across the HPQ sector with their strategic planning. Our tried-and-tested forecasting methods are now in demand across the PV supply chain, from upstream materials manufacturers to large-scale PV developers, which rely on Exawatt’s understanding of module cost and performance to support their strategic planning.