End in Sight for Silicon Shortage in Solar industry

Severe shortages of silicon have plagued the solar photovoltaic market over the past few years. According to a Frost & Sullivan press release a turnaround can be expected this year with polysilicon catching up with the demand . Quoting from the press release:

It was estimated that the demand for silicon feedstock neared 26,000 tonnes in 2004. In 2005 there was a rise in wafer production by nearly 7 per cent. However, this increase was not sufficient to keep up with the market need. In 2006 the shortage of feedstock reached a critical point affecting the production of solar panels and, consequently, the industry growth.

However, things are about to change. "We expect polysilicon supply to catch up with the demand already in 2008" – says Alina Bakhareva, Renewable Energy Programme Manager at Frost & Sullivan. "The majority of the new quantities will be supplied to the market by top 4 producers that are expanding their existing production capacities."

In fact, four top polysilicon producers are expected to add more than 17000 tonnes of capacity in 2008. This would represent over 50% increase over their current capacities.

On the demand side, demand from the semiconductor industry is expected to grow at steady one-digit rates. Demand for solar-grade polysilicon is expected to reach over 50% of the total demand for high purity silicon in 2008-2009.

This should bode very good news for the silicon solar PV cell manufacturers. With silicon supply no longer a constraint manufactures can ramp up production to meet demand and as a market driven supply chain develops the price of silicon should eventually settle out at a lower price. Then competitive prices and technical merits can let consumers choose what type of cells they prefer, rather than having to use what is available.  This more importantly means that solar PV can be a significant (10s of gigawatts) source of renewable energy in a shorter time period -- perhaps in as short as five years.

Comments

[Fifi]

End in sight ...

[Lunken]

Good news... :-)

[Leonard]

Who are the 4 major silicon suppliers mentioned in this article? Thanks.

[Charles Barton]

The real problem of solar power is not the supply of polysilicon or the cost of producing PV modules. It is the cost of installing the things in massive solar arrays. And then, of course, once you get them installed, the only produce power for five and a half hours a day,

[Alex]

Good news and good prospects for the PV industry and market in an increasingly volatile environment. To answer Charles Barton's skepticism concerning PV cost and inadequecy: 1. film will solve that.
2. Not here mate in the Med!!

[Tristan Wibberley]

Attempting to turn off bold ...... Did it work?

[Tristan Wibberley]

One more by the looks of it. There :) that's better

[Sascha]

it is going to be interesting to watch a possible further market consolidation among the poly suppliers over the next few years. with supply growing overproportionally, prices will drop and ultimately force some suppliers out of the market. this is assuming that TF can make further progress and seriously challenge traditional PV.

[Paul Jorgensen]

So what does this mean in terms of cost per watt a year from now. Will it be down to $1.00 as Nanosolar claims they can achieve?

[Kevin]

@Leonard:

My guess would be:

1)Hemlock
2)MEMC
3)Wacker
4)REC

[steve selverston]

IF the photovoltaics get too expensive, people will probably move more towards solar thermal collectors, which they'll probably do anyway.
steve selverston

[mds]

Thank you Mr. Barton. Detractors of PV have for many years pointed out that PV just costs too much. Now this is no longer the case, as you point out. I am pro-nuclear, but PV is going to beat Nuclear precisely because of lower installation costs. Building Integrated PV (BIPV) is already starting to be used by home builders in CA, AZ, and FL. Expect this trend to continue and the installation cost of PV to drop precipitously.

Answer to the energy problem: Nuclear AND PV
(...and CST, wind, waves, geothermal, clean coal, and conservation/efficiency)

Answer to oil/transport problem: E-REV

Answer to CO2 problem: Both

No real problems here, just a technological transition that needs to occur.

[Kit P]

mds, Charles said that solar does not produce energy when we need it. It also does not work very well especially on the roofs of houses. Therefore, the environmental impact of solar thermal is worse than coal.

[mds]

Yes, Charles doesn't like solar. Clearly in the sunbelt of the USA, the most populated part of the most energy intensive country, PV produces power exactly when we need it most. My point is even solar detractors like Charles are starting to realize PV has dropped in price. When PV production catches up with demand it will drop radically. In the mean time, BIPV is also reducing installation costs. At 48% per year growth rate since 2002:
http://thefraserdomain.typepad.com/energy/2007/12/fyi-solar-cell.html
...PV will be one of the larger energy providers in 10 to 15 years when new nuclear plants are starting to come on line.
3.8 GW per year world production for 2007
2017 @ 48% per year = 3.8 x 50.4 = 191.6 GW
2021 @ 48% per year = 3.8 x 241.9 = 919.2 GW

191 GW per year PV production in 10 years
919 GW per year PV production in 14 years

Please don't tell me it's going to slow down. Why would it? Price of energy is high, PV prices are dropping, new PV technology already in production can support huge further drop in price, and BIPV (building integrated photo-voltaic) is increasingly being used. If that isn't enough the raw Si shortage is coming to an end this year:
http://thefraserdomain.typepad.com/energy/2008/02/end-in-site-for.html
(Can you imagine 48% growth rate DURING the raw Si shortage? Holy smoking techno-revolution batman!)

PV works fine on roofs of all kinds and in the future will more often BE THE ROOF, ...and the walls, ...and the windows. BIPV is here now.

CST worse than coal? What have you been smoking? ...deisel extract?

[mds]

Sorry, didn't mean to reference back to the same site. Forgot where I was for a minute.

[bigTom]

mds: good post. Of course technological revolutions go through an S curve. If you plot LOG of capability versus time, the curve reaches some rapid slope, which it sustains for a while before leveling off. There will be some level at which PV begins leveling off. This will likely be when it has reached some degree of saturation in the best markets, whereby the costs of intermittency begin to outweigh the benefits for an incremental increase in capacity.

I understand where Charles is coming from. I used to work in the high performance computer industry. When some new related paradigm comes along and captures nearly all of the public mindshare, but you KNOW that it is not a complete solution (i.e. your own solution is needed as well), you develop such a mindset. Such is the case for Nuclear. Needed for baseline power. Perhaps not everywhere, but in enough markets to make it a crucial component of the energy future. Having too much of the public mindset on any one power source is not healthy.

[Kit P]

People like mds have been telling me for 30 years about solar. Still waiting. I am not opposed to solar just those who waste resources in the name of protecting the environment.

First is is necessary to to define what 'need' is. You need energy to keep your pipes from freezing in winter. You do not need electricity to keep your hot tub or pool warm.

In the Southwest, solar peak precedes peak demand. The capacity factor for utility owned and maintained PV systems is between 10% and 19%. Most other systems with a capacity factor of systems between 0% and 5%. How stupid are those who brag about putting in solar systems that do not work.

The net result is the dieoff of solar equipment because it does not work and apparently no one cares.

For those who want to talk about how much solar capacity has been or will be built, I am waiting for them to start talking about how much electricity is produced. Right now the number is zero or very some small number that looks like zero to the grid operator.

[Charles Barton]

Charles doesn't like solar. - mds

It is not that I dislike Solar. I dislike the overhyping of PV and ST technology. I think solar hot water heating is a great ideal, with mature technology behind it. Solar space heating works too, at least in the Southwest.

PV point of sales prices are not dropping, and in fact have gone up during the last four years. (See
http://www.solarbuzz.com/moduleprices.htm) In the mean time the cost of materials used to build solar installations has doubled. The construction cost for PV and ST installations run higher per nameplate KW than nuclear construction costs. Comparison of capacity factors reveal that Nuclear power plants produce 3 to 5 times as much power per equivalent nameplate rated capacity as solar installations do. Solar advocates provide misleading information when they talk about the price of solar panels dropping, or how rapidly solar installations are increasing.

The truth is that per KWh of electricity generated, solar generating facilities are far more expensive than nuclear plants, and the rapid inflation of materials costs is making this increasingly so.

[Ender]

Charles - "Comparison of capacity factors reveal that Nuclear power plants produce 3 to 5 times as much power per equivalent nameplate rated capacity as solar installations do."

The problem is that that they produce most of this energy precisely when we do not need it. At this time electricity practically has to be given away because running a nuke at less than full capacity is less than economic.

Furthermore nuclear like thermal coal is baseload power which is difficult to integrate with renewables as it cannot be stopped and started quickly to make the most of the savings possible with renewables.

We need more flexible power generation sources and solar PV along with solar thermal and wind can provide this. Mind you all renewable operators should be forced to provide storage to get a license to generate. This could stimulate V2G as the renewable operators could add storage and provide transport at the same time.

[Charles Barton]

Ender, you appear to have little concept of the orgization of energy generation. Nucks are run at fule power, not because it is impossible to run them at lower rates, but the produce power at lower costs than other energy sources. It costs about $.02 per KWh. But given the price structure of the electrical industry, electricity from Nuks sells for $0.05 per KWh. The problem is not with nuks but with renewables, that produce power in erratic, unpredictable fashion. Renewables, with the possible exception of ST power, are unreliable. You need more reliable sources of power than renewables to keep power systems going.

[Kit P]

It is important to understand the order that power plants supply power to the grid in the US and Canada. Nukes and renewable energy supply electricity first but there is never enough. Fossil fuel plants come on line next depending on the incremental cost. Efficient coal plants supply power next. The last to come on line are inefficient oil, coal, and natural gas.

When considering the economics of new generation, renewable energy is competing with inefficient oil, coal, and natural gas. New nukes are competing with new coal. The farther a coal plant is from cheap Powder River Basin, the less econmical it is because of transportation cost.

The only power plants that can be built fast are natural gas which are being built not because it is a good economic choice but because they are the only choice when better projects are delayed in court.

What will a carbon tax do. It will change the order that inefficient oil, coal, and natural gas plants come on line. It will not reduce ghg emissions much.

[disdaniel]

Kit P says "The net result is the dieoff of solar equipment because it does not work and apparently no one cares."

Hehe yeah that is exactly what we are seeing...oh wait...I haven't heard of anyone that returned a solar power system because it doesn't work. In case you haven't been paying attention solar market was 10 times bigger in 2007 than in 2001.

Perhaps people do care and solar actually works?

[Ender]

Charles - "Ender, you appear to have little concept of the orgization of energy generation. Nucks are run at fule power, not because it is impossible to run them at lower rates, but the produce power at lower costs than other energy sources."

So you are trying to say that you could run at a profit with a load following nuke plant? I don't think so. It fully realise that it is possible to run them at part load however would you make any money. Additionally can you quickly start them up and stop them in response to demand - again I do not think so.

"The problem is not with nuks but with renewables, that produce power in erratic, unpredictable fashion."

As do nuclear power plants. How often do they break down like in Florida? How often does the transmission line from the distant power station get interrupted? Renewable power is to a certain extent predictable as short term weather forcasts are getting better and better.

"Renewables, with the possible exception of ST power, are unreliable. You need more reliable sources of power than renewables to keep power systems going."

No you need advanced weather prediction and storage in electric cars to keep both the transport sector going and the electricity going when we start running short of oil. What runs the machinery that mines and transports uranium????

[Clee]

http://www.uic.com.au/nip28.htm
France's nuclear reactors comprise 90% of EdF's capacity and hence are used in load-following mode and are even sometimes closed over weekends, so their capacity factor is low by world standards, at 77.3%. However, availability is almost 84% and increasing.

[Clee]

Profit, you asked? From the same URL
"The cost of nuclear-generated electricity fell by 7% from 1998 to 2001 and is now about EUR 3 cents/kWh, which is very competitive in Europe. The back-end costs (reprocessing, wastes disposal, etc) are fairly small when compared to the total kWh cost, typically about 5%.

That would be about 4.6 cents/KWH in US currency. Which is competitive even here, even if it's not as cheap as if they were run at 90% capacity factor.