>Surprisingly SOL seems looking really good. YGE looks like they lost the most.
With latest guidance I have YGE saying their shipments will be down more than 25% from record Q2 (est. 592 MW) to Q4 (guidance est. 434 MW) and SOL saying their shipments will be up more than 60% from record Q2 (504 MW) to Q4 (810 MW using mid range Q3 and FY guidance).
These are significant suggestions about the relative direction of the business these two companies get.
Since I've put some money on these indications, it is important for me to understand if there's anything that fundamentally supports these suggestions.
For me Virtus II looks like a game changer.
Virtus I has per watt costs to make the wafer much closer to multi than to mono. Then it requires some cell processing adjustments causing slightly more expensive cell processing. All-in-all Virtus I still has significantly lower per watt cost to make a cell and power of the cell will be close to that of mono due to it being mono like and having 3% more active area from full square shape.
So Virtus I has a cost edge over standard mono, but inconveniently needs cell line adjustment which partially offset this advantage in the current multi/mono adapted marketplace. Aside from that it has a high temperature advantage (-0.39%/C vs -0.43), a light induced degradation advantage (-2% vs -3%), not too much low irradiation disadvantage (-1.9% vs -0.6% at 200 W/m^2) and an aesthetic disadvantage (uneven pattern).
This is enough to change the game a bit since cost saving per watt of from efficiency gap has squeezed as total costs have come down and on top of that efficiency gap is closing with Virtus.
Looking at Virtus II I see a much more game changing opportunity. It does not require any cell line changes. Standard multi cell processing configuration works. The per watt cost to make the wafer is lower than multi. This also holds for cell, due to higher power per piece. Now the extraordinary thing is that these cells also achieve the same mono like power as Virtus I. SOL's standard non-square mono, Virtus I and Virtus II all come in the same 250-260W range, with the average at the mid is my understanding. Translating the mid 255W to cell efficiencies this is 18.2% for the square shaped Virtus I and II with 4% CTM loss and 18.9% for mono with 4.5% CTM loss and 3% inactive corner loss. So 18.2% from 6 inch casted ingot wafer is on peak power par with 18.9% from 6 inch CZ rod ingot wafer. So Virtus II has same power as mono at per watt costs below multi. Aside from this Virtus II vs mono has a high temperature advantage (-0.40%/C vs -0.43), a light induced degradation advantage (<-2% vs -3%), not too much low irradiation disadvantage (-1.3% vs -0.6% at 200 W/m^2) and no aesthetic disadvantage (even pattern). All these 4 other properties except temperature coefficient is better than Virtus I. Mono only beats by slight margin on low irradiation performance and possibly slightly on pattern aesthetics depending on preference.
I'm not saying that SOL owns the game changer (although they have like GCL and contrary to LDK taken the proprietary technology path), but this seedless controlled nucleation technology that they are applying (with all their furnaces upgraded to support both Virtus I and Virtus II configuration) seems to outperform other technologies on PV value-chain economics, since it at the end of the day (cloudy/hot or not) allows lower cost per kWh produced.
This could be why we see demand for SOL products disconnecting Chinese peers.
What would be very interesting for me to hear in your potential interview with SOL are:
1. The connection between their wafers and modules. Are Virtus II modules achieving its 255W average simply using Virtus A++ wafers and standard cell processing? Since I believe wafer purity set the base level for cell efficiency, but then more advanced cell technologies can enhance/reduce loss in cell activity by capturing more light for conversion (texturing, fine line etc.). Do SOL modules use any non-standard cell processing techniques to achieve its high power? If not one can understand the high Q4 demand for Virtus A++ wafers suggested. For those with advanced cell processing technologies taking their modules from 240W to 255W, the SOL wafers would take them to 270W, giving them a great edge in the marketplace through the combination of own cell technology and SOL's cheap high power wafer technology. For those with cheap standard cell processing technology, they can suddently compete in the more premium product range, thus gaining brand value at low cost.
2. Are their square mono from high-grade mono casted wafers or from growing bigger CZ rods? The black square mono have a lower power range than non-black, 250-260w instead of 260-270w. Why?
3. What's their mono casting (seeded controlled nucleation) quality yield, i.e. how much can be processed with standard multi configured lines, how much requires the special quasi configured lines and how much can use standard (except handling shape) mono configured lines? Are some of the multi parts below the Virtus A++ quality gained from seedless growth? How much?
4. Why do seeded controlled nucleation when Virtus II economics beats Virtus I? Is it because getting some highly calue square mono too on the cheap along with Virtus I balances the economic outcom of seeded vs seedless controlled nucleation?
5. Basically a complete break down of their target for how they use their capacity and its yield would be nice. First they have 400 MW for CZ and the 1600 MW for ingot casting that can be configured for seeded or seedless controlled nucleation. The seedless random nucleation configuration option for standard multi will no longer make sense I assume since multi wafers have no benefit over Virtus A++ wafers? So of the 1600 MW how much seeded vs seedless controlled nucleation are they targeting? Of the seeded controlled nucleation how much of each of the four quality levels (V+, V, M and A) are they getting? Does one (or more) of these four grades correspond to square mono or is that a separate grade, actually how do these grades of wafers coming out of a seeded controlled nucleation grown ingot map to the three types of cell processes required. Finally of the different types of wafers they are targeting for their 2000 MW capacity production, how will these be allocated for external sales vs use for own modules? Will external sales of wafer for examples have a high-end bias (as indicated by premium wafer ASP of 30 cents Q3 guided vs multi at 22 cents on spot market now, i.e. 35% diff) or low-end bias (as indicated by high power module portfolio)? All this is interesting for the understanding of the volume of high quality at low cost their capacity park is capable of yielding, since for others we've seen that high purity wafers either comes at low yield or high cost. It's this dilemma that SOL seems to have found a solution to. They might not be alone on this path, but at least it seems they're on the right path and maybe a quarter or two ahead of the competition.