SolarPVInvestor Forums

Quoted from "Klothilde"

Thanks for your insights, things are clearer now: There are two sources of value for a cell/module manufacturer who switches from mc wafers to hpmc wafers:
1) higher ASP of cells and modules because of the monetized benefit of lower BOS cost
2) lower production cost at cell and module level

Yes, this is the benefit. So you currently have lower cost (because it is roughly same as standard process once you've figured out the formula) and higher premium on adding the value in the wafer than adding the same value in cell step.

Quoted from "Klothilde"

One issue I see is that it will be increasingly harder for SOL to stay ahead of the efficiency game and reap the associated value since I suspect that further cost-effective efficiency improvements for mc (or hpmc) wafers will be hard to achieve.

Yes, this is the risk.

Quoted from "Klothilde"

I imagine SOL will increasingly direct R&D funds (which are substantial) to other segments of the value chain, as can be suspected by their incursion into microinverters for instance. What do you think?

SOL has a quite interesting seed approach. There's a lot of small PV R&D companies in ReneSola's Zhejiang province. So they have both their own R&D center where the partner with R&D teams of their main equipment suppliers and also do a lot of own R&D on material consumption efficiency etc., but they also have this ecosystem of small nisch R&D companies that they seed and take in those reaching breakthrough (I'm speculate a bit on how they work with this ecosystem, which has been described in articles a few years ago, but I think they seed and prune many tracks and then buyout on success). This allows them to have this wide R&D focus at low cost.

Since SOL is R&D focused they have higher degree of proprietary technology than those who source technology (like Yingli which source n-type mono and hpm technology). I'm not convinced this protects them in the long-term, but should enable head start to collect the higher margins in the early adoption periods and allow lower cost in mainstream adoption periods.

Interesting view Klothilde. Odyd's not gonna like it.

Honestly I think only SOL is transparent enough with costs of different technologies, so it is hard for me to say what value addition the Honey technology really has. I know that there are a couple benefits of SOL technology though. One is that all existing lines have been converted to it, confirming major added value benefit under mass production conditions, and achieving this with low capex through upgrades instead of replacement/expansion (Honey capacity was all through new line additions). Another one is as I mentioned before that it is early in the value-chain and thus the added value benefit covers a larger part of the value-chain. A third one is that the production of cells is relocatable as standard cell processing is used and the HPM products can thus not be blocked by tariffs as easily as they are done for Trina's specialized Honey cell lines.

Regarding what SOL's tolling partners do with their cell lines, they don't need to change anything for Virtus II. That's all in the wafer and standard screen printing gives the excellent cells. For Virtus I it was not as easy, since it was quasi-mono. Same cell lines could be used, but they needed to be reconfigured for a different process. Both wafer and cell processing steps were thus a bit more costly and less directly available and the resulting cell looked much less aesteticly attractive. Virtus I was what they converted all furnances to in 2011 and that product was a great success for a year, but as they discovered that the Virtus II process could be used with the same quasi-mono enabled furnaces they transitioned to fully focus on this in 12Q4 by producing only Virtus II (Virtus A++) wafers that quarter on the non-mono side. This together with guiding shipments at 140% capacity utilization indicates management confidence in the value of their new product.

You might want to consider Monocasting as the ultimate for of processing Poly adn the end goal of Virtus wafers and why SOL has gotten rid of their pullers but still sell mono. That the seeding and pulling technology that created mono wafers from cylinders is being replaced with a casting version of creating mono wafers. This is reducing the costs of mono wafers significantly and you are getting into square mono wafers. The poly capacity is able to converted relatively inexpensive from Poly. This casting is what is leading to the high end greater powered casted wafers.

http://www.sciencedirect.com/science/art…876610212012490

here is a nice reference site for technical information

http://www.sciencedirect.com/science/jou…66102/27/supp/C

Thanks that looks like a good technology reference site.

It looks like SOL first started to move away from mono (but not 100%) and now move away from monocasting too (Virtus I). HPM wafers seem to currently add maximum value.

Quoted from "explo"

Interesting view Klothilde. Odyd's not gonna like it.

Few points :Trina
"MG: The silicon used for the Honey cells is different. The ingots are mono-seeded, meaning you can’t just convert an existing cell line using standard wafers. It needs to be a complete process from ingots through the panels."
source : https://solarpvinvestor.com/spvi-news/265…-the-sweet-spot

I am not sure about the nasty mind Klothilde but official and unofficial what we call now GDAP (global declared average price) points to premium for Trina modules, exception one month, recent one.
While processing cost was higher by a cent, it was explained as due to under utilization in Q4.

We have no understanding what is underutilized, but since we have the US market in the spectrum and cell-technology is within the Trina's value chain, honey is not going to the US. Using Panda and Yingli as an example of lack of need for efficiency is also wrong. Yingli is driving scale based on market demand needing to absorb large amount of modules. This means heavy Chinese market output, low processing is still the driver greater than efficiency for scale sales, lastly Yingli also has the DSS tech.
I used this article before to explain HP wafers. This article has been written with assistance of wafer company insider. As you can read this technology is not unique, in fact LDK M3 wafers are apparently superior to any wafer out there. Anyway the point is that SOL has that advantage, real, marketed or otherwise but other companies are either quiet about their abilities or are on the track but certainly, many use it as successfully in more or less quantities within their products.
https://solarpvinvestor.com/spvi-news/388…stalline-wafers

odyd, the M3 is likely as good as Virtus A+++, but they have not been able to control their costs, which makes LDK a troubled supplier shipping less than half the third party wafer volumes that SOL did in Q4 and only 35% of total volumes (including internal shipments to module business).

Quoted from "Klothilde"

I personally don't think this is s.th. to write home about, but SOL probably will write one letter per day.

The quality edge is mainly important when there's oversupply. In shortage times everyone gets bloated. When there was a lot to choose from they chose SOL, even with 10% price premium. And they were all quality customers. Look at the huge AR provisions taken by most peers. SOL never has to take any AR provisions.

If you look at the beating wafer makers were expecting going into 2H2012, then LDK and GCL came out as expected all bruised up (LDK -300.8% operating margin), while SOL fared better (-7.8% operating margin) than even the module focused companies who were at least enjoying great wafer to module pricing spreads.

Quoted from "cfeng"

140% utilization and no AR provisions speaks volumes, when the competition is struggling to get to 70-80% speaks volumes

How do you know they are 140% utilizations? They could be at 80%. You do not know the ratio of products produced and sold. What you know is they have a claimed ingot manufacturing capacity of 2GW which is based on normal throughput.

"We also measure our ingot manufacturing capacity and production output in MW according to the solar wafers in MW that our current manufacturing processes generally yield."

You also know they buy ingots from third parties for Mono Crystalline.

" we also had certain monocrystalline ingot manufacturing processing arrangements with third-party manufacturers."


What you do not know is they have stated that the extra modules will be outsourced from third parties

" We will not expand our solar module capacity, but may instead outsource some production to third parties under ReneSola's supervision and brand and in compliance with localization rules in some jurisdictions."

What you do not know is the amount of cutting capacity they have not do you know as suggested in the con call the use of third party to create virtus wafers.

From these comments, one can presume that SOL is planning on buying externally 600MW worth of third party products in the form of ingots, cells, and or modules or 20-25% of their shipments.

They may be running at or near 100% capacity for ingots but you do not really know do you?

So what is the split and ratio of outside purchased products? Did they sell 50MW of mono of 200MW of mono modules? If there is overcapacity is not SOL taking advantage by procuring this low cost oversupply for what looks to be 20-25% of their manufacturing needs? That is the same that CSIQ does, that CSUN does, that Yingli is starting to do and TSL decided to do a year ago which is to buy low cost products from thrid parties and brand under their own name.

SOL is actually more risk exposed than the other companies for price fluctuations by relying on third party for Poly to Ingots to wafers and cells and to modules. They are exposed at every single point in the value chain.

This s nothing new as LDK during their Hyper growth was buying ingots and doing some tolling for sawing for others.

SOL is at 100%,the rest is outsourced!! (wafer and modules)

Yes, I guess Q4 put them to the test selling modules to China for first time and 50% of their modules sales that quarter.

God forbid SOL produced just a wafer, without anything to write home about; they would be swept to non-existence.
So the question is how durable is the property of this operational marvel (Virtus II)?
How safe is to operate without own cell technology?
How does the company want to make a “brand” deals with OEM arrangements?
I think, considering market today, SOL may have short-term successes in last two cases. Not having own cell production allows SOL to purchase cells built on own wafers and introduce them to OEM model. As long as module offers GM, this would be a success. However short-term successes have long-term repercussions. SOL is selling own wafers to others. Either you lose the module benefits from wafers sold to clients with superior cell technology (eventually or now), or you lose clients if you sell them, sorted out, poorer efficiency wafers. The wafer, no
matter what kind, is the one without GM as well, SOL knows module makes money. This model (of a wafer as a retail product) is truly temporary. The model of OEM is also temporary for others. This means that some swift, significant decisions will result in the allocation of supply chain globally.
Then what is left is the proprietary aspect of its wafers, which makes that argument for SOL today. However, has anyone even flinched about it? Not really. Here and there, companies say that have the DSS, sort etc.
If technology is not uncommon, offers no particular advantage of profitability (company is going away from it), then what is the cornerstone of the value-added to SOL? I think the lack of obvious answer is the market evaluation assigned to a stock. I honestly think that efficiency is going to be moving equally through surviving businesses. As processes cost less, new materials like copper, instead of silver, re-emergence of n-type mono, will promote efficiency up and costs down. The only potentially useful setting for SOL is that, cell companies will have to refit own lines with modern novel methods, but at the same time those cell companies may no longer be around, when the new order is revealed. I think SOL not having much to say in this area (cell) makes the company model vulnerable today.

SPWR is a cell technology company; it makes biggest waves as produces a lot better cell on wafer which is easily accessible in China, with a lot better efficiency and only slightly higher pricing. Today cell companies are in Taiwan. In the near future, the first time Yingli will create or buy a cell company outside Mainland, the purpose of a “cell company” will die with it. The globalization of the future value chain will result in same ownership of global locations, allowing
for technological separation (in-house or unique equipment contracting) perhaps for the very first time.

As virtually all major module makers in China buy a good portion of wafers from GCL, there is very limited benefits or product differentiation. That is why this is a Commodities business. The only differentiation will be cost and warranty for those that care. The technology is proven to last 30 years if manufactured properly. I always listen for the inventory returns provisions as these are generally failed modules.

Odyd, if we should summarize what gives good efficiency in ingot and cell step it's

Ingot: mono
Cell: back contact

That is the SPWR solution. The problem with both these and reason why not popular by solar11 is cost. For now I would say that the cost revolution is in ingot, where DSS is much cheaper than CZ. This combined with closing efficiency gaps between mono and multi with the new controlled DSS process seem to be one point of value and as a result we've seen more clear trend here than for cell approaches. The cell process still has ahead of it to find out how to add conversion efficiency at low cost. I'm not sure it is a bad thing to wait for that development to unfold before placing the cell equipment bet.