Asia Solar Cell Producers Rising Amid Pricing War

This is a question. I would like to see a rough sketch from a PV module production expert of the production costs - at least a simplified pie chart. Polysilicon, labor, equipment costs, module assembly, etc. What I really would like to know is __ why is $1 not sustainable, as many are so quick to say. This question does not concern currency fights or politics. I want to see the real breakdown as it is now in 2012. I know some of these costs are different in China vs. elsewhere, but not all of them. Innovation can reduce costs.

brucycle wrote:

"What I really would like to know is __ why is $1 not sustainable, as many are so quick to say."

Several years ago, there were "experts" claiming $10/watt was the cheapest PV would ever get. Obviously they were wrong, and didn't understand the market.

$1/watt is no magical number either, you can buy new PV today retail for 88 cents/watt: http://www.sunelec.com/

The answer to your question "Are module prices sustainable below 1$/W?"is simple: YES.

There are many sources of information but the best source maybe the solar manufacturers themselves. For instance Canadian Solar, which is located in china, shows in their january investor presentation the following:

Wafer (including poly): 0.31 $/Watt
Cell: 0.20 $ Watt
Module: 0.25 $/Watt
Total: 0.76 $/Watt

Two other low cost manufacturers (Trina, Yingli) noted (Q3 conference calls) their silicon to module cost will be at or below 0.60 $/Watt at the end of 2012. With a price of poly at this moment of around 30$/watt and a conservative 6 g/W of poly. This leads to a total module cost of 0.78 $/Watt.

The following examples show clearly that module costs can be below 1$/Watt and still be profitable.

But that is not the most important question. There are several questions which are more vital:
- What is the cost of installation and BOS(Balance of System), or better why is there such a large difference in the installation cost in the US and Germany?
- How can this cost be reduced?

The cost difference between the US and Germany has been noted in some articles lately. One topic has not been dealt with: In the US there is no incentive to reduce cost to improve IRR. Residential customers do not receive FIT just as in germany so there is no financial incentive to install such a system. Since there is no incentive the installers do not have to have an economically reasonable installation and they can absorb all the profit. For large installations this incentive is there, because those investors care about their money, and that is why those prices are much lower.

The reduction of this cost in germany is quite easy: increased efficiency of modules plays a major role here. Increased efficiency modules mean lower labour cost, lower BOS cost, lower land cost, less space, etc.

How much efficiency and cost reduction are possible for crystalline solar?
A LOT!

- Efficiency
- Wafer thickness reduction
- Silver --> Copper
- Other

Well first about efficiency. Everbody probably know Sunpower. They have the modules with the highest efficiency on the market. Their efficiency is over 20% for modules and have cells in development to allow module efficiencies of over 21-22%. So if we translate part of this efficiency gain to the chinese manufacturers what will be the efficiency increase?

The highest module efficiency of a low cost chinese manufacturer is currently 16.8% of JA Solar, the average efficiency of the chinese manufacturers is more in the range of 15%. If the average efficiency could be increased to 18-18.5% without adaptation of the equipment this would translate in a cost reduction of over 20% on the whole module. Is this possible? YES. Yingli solar is researching modules with an efficiency over 18%. They already achieve an average 18.9% efficiency on their PANDA cells which together with MWT technology should translate to modules of 17%. Further increases are are still possible taking the Sunpower case as a benchmark.

Wafer thickness reduction:

Much research has been performed on wafer thickness reduction. Wafer thicknesses as low as 60 micron are currently possible (Silicon Genesis). Reducing the poly cost from about 0.15-0.18 $/W to the range of 0.06 $/Watt. Further reductions are even possible

Silver --> Copper (Imec, Kaneka, Other)

As indicated by several solar companies the price of silver is currently the largest cost in cell processing. With a cost of up to 0.10 $/Watt. Changing it to copper would completely wipe out these costs.

Summing up the total:

0.10 $/Watt reduction through thinner wafers
0.10 $/Watt through copper plating
20% reduction through increased efficiency

Current cost 0.80$/Watt (lets be conservative)

0.8*(0.80-0.10-0.10)=0.48 $/Watt

So in a matter of 2-4 years crystalline solar modules could be sold for under 0.6 $/Watt and that is not even taking into account additional advances such as layers with >100% quantum efficiency (see news), reducted lamination cost (see news), and other developments.

Crystalline PV is definately here to stay...

Regards,,

Olaf de Swart
Long time follower of the solar industry

Thank you Olaf and others, this is just the sort of info I was looking for. My opinion has been - YES, prices can and will fall sustainably- but that was based on less information and more gut. Another thing I have been thinking of is the PV equipment costs. As more of the equipment is designed and made in Asia, I think those costs will fall too.
And I have been really thinking about the BOS, since this falling module pricing turns up the heat on BOS price reduction. But I did not follow your thinking in comment 6 above. I know the great differences between German FIT and USA, but why are there not incentives here for the installers to get efficient ? For example, here in CA we have the PPA and a plethora of installers. Some are poor business people, some are good. There is also a lack of ROI thinking and sophistication in the US customer base. These customers must be educated and convinced. Could you elaborate on your previous comments on IRR and the US ?

Hi Olaf -

I am currently writing a study on the German-American installation price difference (focusing on the residential sector) and would be interested in the articles you were referring to. Could you post some names or links?

Thanks
Jo

Dear Jo,

There are not many if any good articles written about the difference in approach between the US and Germany. In germany the residential producer of electricity can make a profit with the electricity it produces by sending it to the grid. However since this year it is for german producers of PV electricity better to use as much of the produced electricity theirselves because the FIT rates are below the retail electricity prices. In the US the size of the market is small. This means that only a few installers exist and there is not much competition. All the rebates that are available lower the cost for the residents who install a PV installations but do not reduce the cost of the residential PV installation. In other words, while residents pay a relatively small amount (In US standards) the cost of the PV system is still realy high. There was no incentive to reduce cost. Luckily the 30% investment tax credit expired which means the cost of residential PV systems should also drop (Current PV system prices unaffordable for residents without rebates). PV in the US is still seen as much more expensive than retail electricity prices becuase of the high cost before rebates.

So it is currently a balance between market size and rebates (and lack of competition). In the near future the market will grow, competition will increase, rebates will vanish and the same IRR thinking as in Germany can take a hold for residents too. The US generally has the potential to have much lower kWh cost as in Germany due to lower labour cost, more sunshine, more available cheap land, etc.

I would like to inquire what the average cost per watt is of installed residential PV arrays in Germany compared to that in the US.

Brucycle: the cost equation is simple: / of the module.
1. Module efficiency is the big knob: if one can increase the efficiency of a module from 14% to 15% at no cost, $/Wp drops 6.7%. Things that affect efficiency:
- cell efficiency
-- actual photovoltaic conversion efficiency
-- front surface active area (e.g. occlusion by front contact structure, geometric shape (square vs rounded), etc)
-- improved emitter/collector contact structures (e.g. selective emitter)
-- reduced semiconductor defect
- optical gain/loss
-- glazing reflectance & transparency
-- encapsulant refractive index & transprency
-- recycling within the cell as well as light that doesn't directly impinge on the active area
-- cell light capture
- resistive losses
-- e.g. front contacts built up or moved to rear where they can be heavier
-- number / gage of interconnections
-- geometry i.e. increase the ratio of active area to total area of the module
-- more cells per module
-- square cells
-- reduced edge seal area
- improved cell sorting/matching
- etc.
2. Per area costs:
- glass, encapsulant, back sheet: all generally driven down by economy of scale and commodity market but also some improvements/changes to material
- cells/wafers/silicon(the topic of this article)
3. Linear costs:
- framing: probably the least movement here as conventional aluminum extrusion is ~75% energy cost but this can change.
-- lowest $/W modules are, in fact, frameless
- edge seal materials
4. unit costs:
- junction box w/ leads, connectors, seals, etc.:
-- economy of scale and commodity market
-- larger modules / more cells per module
- corner connectors, lables, etc.
-- molded parts, no-fastener systems, etc.
5. Labor
- always a small factor
- greatly reduced by high automation but this is only justified by high volume e.g. >120 MW/a production lines.
6. Production yield
- improved process knowledge
- more SPC
- more automation

check google and search for PREISINDEX BSW
Then you get an overview of the average cost of small pv systems in germany

Olaf - good to know (there is English if you hunt around a bit).

'... when it comes to constructing PV projects, bureaucracy and highly complex procedures and requirements (i.e. notification, registration, licensing, environmental impact assessment, etc.) hamper installation processes significantly. As a result, the development and the sustainable establishment of solar markets in many European countries is unnecessarily delayed.' Apparently, the US hasn't cornered the market on this capability ... they've just perfected it :) Only government can make 2 quarters cost a dollar.

Again, I understand Olaf's points about US market _only_ partially. Here is what I mean : here in CA we have tons and tons of installers. No lack of competition! And although the rebates may cover most of the system, customers here care about money! why pay $4000 for a system if you can get it for $2000 ?! I do agree reducing the rebates would heighten price competition. But I suppose the bigger factor is all of the paperwork and a lack of standardization/modularization of the desgign and install process. I think prices just have to fall further.

I know these price declines will set the installers crying, but they have to get leaner and meaner. I am a green energy advocate, but the industry does not exist in order to provide lots of jobs. It exists to serve a real function in our energy economy, and to provide an appropriate level of employment. The tax credits and gov incentives serve a useful purpose, but these must naturally be gradually phased down and out.

Thanks to Gerald for the particulars - interesting point in the whole debate is the relatively low level of labor (at least pre-install.) The price of labor is not the main thing about China. I assume most in this forum understand this. My background is semiconductor so this whole debate sounds so deja-vu : RAM history with US and Japan.

Arizona has First Solar and Sunteck, along with the most sun of anyplace on earth. We will prosper even against subsidied fossil fuel and Chinese PV.

Now if we can just get to keep our SRECs and get Feed In Tarrif since solar is clean and during the Peak Time of day.

What you say is there are actually only small companies which do not have enough instalations or power to reduce the installation costs further?

brucycle.... Do you really have any idea about what you're talking about? I think you need to stick with semi-conductors and leave solar to those how are involved in, and understand, the industry. Installers as well as the other businesses that support the industry have a right to make a living, just as you do in whatever it is that you do.
The equipment may be a commodity but the labor, design, construction etc. are not. I have 30 years of experience in the construction industry and I take a dim view of people who harp about the cost of BoS and the installer's costs without having aclue about what actually goes into a complete turnkey installation. It is not rocket science but nor is it a simple deck project. A customer gets a lot more than just a pile of equipment when they enter into an agreement with my company and we are worth the money they are investing. Two final questions: Do you earn and support your family on McDonald's wage for whatever it is that you do? Any holes in your roof?

Malba1.. Design? What is there to design? Most installations in Europe are standard packages. They have standard packages for flat roofs, inclined roofs, etc in all kinds of sizes. The only thing that has to happen is that somebody goes on the roof and installs the stuff and the electrician has to make a connection to the electricity network. Installation cost are typically about 0.4 euro/Watt for an installation of 5kWp.

I would like some honest, technical opinions about the products from Solar Frontier ( especially would like to hear from Gerald and Olaf, but all welcome.) They seems to have a quality product with really good total cost of ownership. Is this for real? Are the advantages of low thermal coefficient and 'light soaking' really significant ? Is Cadmium really worth fearing ? They have some good utility announcements lately and new products.

Dear Brucycle,

I don't know many details about the light soaking effect, but I know the thin films generally have a higher performance (kWh/kWp) at higher temperatures. Before 2010/2011 the costs of thin film solar panels were much lower than crystalline solar panels. Their efficiencies were also much lower 7-13% but they could still be sold easily. From 2010 the picture changed. Drastic reductions in both manufacturing cost and polysilicon prices reduced the prices of crystalline solar cells to the price of the thin films, however these solar cells have an efficiency of 13-16%. So what happened first was that the low efficiency amourphous crystalline solar manufacturers are almost pushed out of the market.Because of the drop of solar panel cost the BOS cost become more and more important, even so much that panels with efficiencies of 12% are less sold (first solar) unless they have their own utility scale projects. So far for this story. Now more about the technology itself?

During the same time startups such as nanosolar, abound, and many others claimed solar panels with high efficiencies at a very low cost. But if you follow the industry carefully there is actually only one company able to put thin films on scale on the market (Solar Frontier) next to first solar. It seams that most of the statements made by these companies could not be fullfilled and still yet have to fullfill their promise. What we actually see is that the records on cell efficiencies on small (research) scale go up but the efficiency of their products is low and the efficiencies of the crystalline competition continues to go up with enough headroom for the coming years making it very difficult for the thin film (CIGS, CIS, CdTe) to compete. Only time will tell what will happen.

Maybe I should also mention something about the cost reductions and efficiency improvement possibilities of crystalline silicon solar cells:

thinner wafers: 200-->20 mu
Higher efficiency modules: 16%-->20-22% (See SunPower)
Lower poly prices: 30 --> <20 $/kg (see REC)

Currently lowest module cost: 0.75 $/Watt (Canadian Solar)
Short term lowest module cost (1-2 years): 0.65-0.70$/Watt
Future (3-5 years): 0.50$/Watt

"Future (3-5 years): 0.50$/Watt"
was this announced by canadian solar?