Solar Energy After Spain

As a businessman, politician, student, media person or homemaker you don’t do yourself any favors ignoring solar energy and related events. An earlier article of mine spelled out why. This one looks at how turbo-charged photovoltaic (PV) sales peaked in 2008, stumbled at year-end and nose-dived. And at what’s to become of us as a consequence.

The industry’s being targeted by consultants asking $ 3000+ for reports that posit why a seller’s market for photovoltaic cells turned on a dime and threw itself into the buyer’s lap last New Year’s Eve. I’ll save you the money. The reason is Spain.

The global PV market grew 110% in 2008 to 5.3GW with Spain in itself soaring to 2.51GW in 2008, a 285% increase over the previous year. Then, it all went busto as the sun rose on 2009.

Today’s most optimistic projections suggest a 2009 off-take that barely matches 2008. I for one am going to short even that projection and put money on no more than 3GW of PV sales in 2009.

If we deconstruct what led to this perfect storm we find a PV-grade silicon shortage brewing as far back as 2006. In 2003 PV-grade silicon prices were at around $ 32 a kilo. By mid-2008 the price hovered around $ 500/kg.

Sometime in 2007 Spain removed its cap on PV generated wattage eligible for feed-in tariff over the next twenty years. “Feed-in Tariffs” being government mandated rates that utilities have to pay solar power entities patching into their grid. The Spanish Feed-in Tariff was a robust 4x prevailing electricity rates and a good 50% more than what Germany offered. That this encouraged solar panel sales is an understatement.

Spain removed the cap expecting it’s decision to help deliver on their 2010 target of a total of 4GW of installed PV capacity. What they got, was 3.2GW of capacity even before 2008 was out. So they slammed the brakes. PV-grade silicon prices fell to around $ 65/kg, where they hover today.

This Spanish decision to un-cap was the reason for the price spike more than any actual shortage of refined sand. Spanish solar sales soared from 600MW in 2007 to 2,511MW in 2008, building the industry a beautiful sandcastle and the virtual shortage felt very real for a while.

The El Dorado that was Spain-2008 fostered a healthy amount of fraud, leaving the authorities to investigate how much of the 2,511MW was real. To ensure they are not suckered again, they capped 2009 at 500MW thereby handing the global market reins over to its demand-side. The events in Spain dovetailed neatly into the brewing global economic crisis creating for a perfect storm. Spain made the decision in August; Lehman collapsed in September and all hell broke loose soon thereafter.

Then there’s the skewed user demographic. Salesmen rushed from Spain to Germany, France and Japan to unload inventory instead of to where the sun shines in India, China, Africa, and South East Asia. Barring the Mediterranean countries and South West USA, those who can use solar energy the most also remain the most clueless or indifferent. All the announcements coming out of India and China on solar energy, feed-in tariffs etc. are geared primarily to placate overseas concerns over their respective very onerous carbon footprints. China may be the world’s largest producer of solar panels but that’s an export driven story. Like India, they have yet to prove they are even half as serious as Germany about domestic solar energy. Until they do, the market for PV panels remains shy of full potential and the post-2008 gloom deepens.

Looking ahead, one posits that a demand-side shift is an inevitable eventuality and that Spain was just the trigger. If not Spain, something else would’ve come along to kick us in the butt.

In the case of gasoline a limited availability, refining/ processing capacity and cost all decide its price. The same is true converting coal to electricity. In the case of the PV-grade silicon to PV panel vertical though, we need only reckon with processing capacity and costs. There’s no raw material shortage registering on the radar screen, at least not for the next couple of thousand years. As such, except maybe for a few years around 2012 when we expect PV electricity generation to hit grid-parity, the availability of PV-grade silicon may tend towards oversupply well into the future.

Having worked in the telecommunications industry, there’s an analogy at hand. Refined silicon and electronics are as basic to optical fiber as they are to PV manufacture. Optical fiber shortages spooked and concurrently excited investment in telecommunications in the late Nineties. Then came the bust of 2001. Since then, try as we might, the darned selling price of dark and lit fiber remains underwater. The current end-user rate to make a telephone call to China is consequently under 3 cents a minute and telecom executives now travel Economy while flying there.

Point being, why should supply-side economics for PV grade silicon be any different. If a polysilicon refinery takes 5 years to go on steam, so does an optic fiber plant. Refining silicon to PV grade can’t be any more difficult than converting silicon to optical fiber and look how sub-sea the sub-sea optical cable pricing has gone.

The difference however may be in Moore’s Law and its application to the semi-conductor/ IT/ Telecom industry. The Law edicts computing power doubles upward and cost commensurately downward every two years. Hence the current state of the telecom industry and probably why I am now boning up on Solar. One can only fly Economy for so long.

PV pricing may never free-fall like Telecom end-user rates simply because of the base price set by fossil fuels. So, during a long drawn transition spanning decades at the very least, as long as people are willing to pay what they pay for coal-fired electricity and for oil, these fossil fuels set the price of PV solar cells and the market has a buffer protecting it from rock bottom.

Except that we in Telecoms held on tight to a similar blanket with thumb-in-mouth in the Nineties. For us the erstwhile incumbents – AT&T, The Baby Bells, British Telecom, NTT, China Telecom, etc. – would protect end-user rates from rock bottom as long as the emerging technologies and their purveyors worked within the umbrella of incumbent tariffs. What happened though was that the new kids did not play by the rule while at the same time their economies and efficiencies became compelling enough for the incumbents to also aggressively adopt and muddy the competitive fray forever.

Game Theory cautions us there is no going back once the protective umbrella is rent. It’s a reasonable bet the same happens to energy generation once solar achieves / surpasses grid parity, too many new producers surface and their technology becomes too compelling for the incumbent energy establishment to ignore. Having said that, it’s not necessarily a bad thing if solar does for energy generation what IP, the Internet and wireless did to traditional telecommunications. In such a scenario for the solar industry, how one puts silicon to play decides one’s fate.

It’s reasonable then to assume that post grid-parity it’s going to be a demand-side market at least to the extent telecommunications is today, varying from government managed to free market depending on the country.

Caught up in the silicon shortage scare, I put a lot of faith in triple-junction and thin film technologies as alternatives to PV grade polysilicon in my last article. I was wrong so here comes the back-pedaling. If we go with the thesis that production of PV-grade silicon will tend by-and-large towards over-supply then producers who made the shortage a basis for investing in alternative technologies or production facilities may be headed towards a reef.

Easing up on planned excess capacity is relatively simple as we saw in the case of Chinese solar panel maker Trina Solar who recently canceled plans to build a billion dollar silicon refinery. Almost all the cost of PV-grade polysilicon is in power once the factory is built. You can simply turn the factory off if demand crashes. There is no capacitance in the system.

What though, if you’re already in PV cell production using non-Silicon based technologies like thin film cadmium-telluride, copper-indium selenide, etc? Such manufacturers must feel justified in the course they’ve charted but I for one would worry every time I went to the beach, saw sand and did the math.

The other nice thing about silicon besides availability is that it’s the best studied material in the world; it is very reliable- well suited for the 25 year life that is needed for a power plant and the cell is very efficient. None of the other technologies even come close and are at half the efficiency of silicon. Besides, other than some CdTe plants, all other technologies are still science projects.

In the new post-2008 demand-side world being shaped, power resides with low-cost mass producers, the project developers/installers and in markets that are on-grid and have intelligently thought out feed-in tariffs.

The winners in the solar play-offs shall be the PV-grade polysilicon producers who use the least amount of power to generate a gram of PV grade silicon and PV wafer manufacturers who produce the most power from a gram of the same silicon. It’s all about how well you can slice & dice silicon wafers and how many of them you can produce.

In the immediate short-term though every, and I mean every, existing PV production line is in trouble and at the mercy of project developers as the rest of the world is asked to take up the slack from a Spain that is capped at 500MW for 2009, down from the 2511MW of 2008. Faced with limited access to costly capital and waning incentives in major markets, project developers are demanding higher returns from fewer projects. They are passing these pressures on to manufacturers, forcing them to bring down module prices by 25 percent in 2009. Fewer projects lead to slackening demand – projected to increase by only 13 percent in 2009 – and, ultimately, to an industry-wide 15 percent revenue collapse this year alone.

The corollary then is a hugely sunny outlook for any silicon-based PV cell producer who goes into steam post-2009 free of legacy and able to plot inventory flows without having to worry about the Spanish aftermath.

Before signing-off, lets save you $ 3000 and list some of what the consultants say along with the odds against their accuracy.

The $ 3000 Forecast: “Access to projects and financing has replaced access to poly-silicon and modules as the market’s key gating factor. With this dynamic unfolding, we forecast module average selling prices to fall below $ 2.50 per Watt in 2009 and $ 2.00 per Watt in 2010 as demand-side financing pressures force manufacturers to cut prices”.

My Take: Spot on. The odds against this happening are low. However any producer who goes into steam post 2009 and is able to mass produce silicon-based PV cells would have factored for this wrinkle and will coast, relative to those forced to weather the current storm. Also, financing models like the Power Purchase Agreements (PPA) are already pivotal to deciding the size of the market superseding production capacity as the deciding factor to market growth. Read my earlier article on Muni-solar for more on PPAs .

The $ 3000 Forecast: “The industry’s market size will contract 15 percent to $ 12 billion in 2009, and will remain relatively flat through 2012. Low cost and high performance module manufacturers able to retain margins and sell products generating high internal rates of return for project developers will gain an increasingly large share of this shrinking pool of capital”.

My Take: Again, correct. Odds against this happening, low. Note they predict market remaining flat until 2012. There is a consensus of sorts that around 2012 the PV wafer will achieve true-grid parity without help from feed-in tariffs and other subsidies. With that, the market will spike along with demand. The supply-side will beef up but only for a couple of years max before the market again tends towards oversupply. Needless to say, in such demand-governed markets the relationships that manufacturers build with project developers decide nature of their success.

The $ 3000 Forecast: “An industry dominated by Asian multicrystalline and CIGS – with solid market share for CdTe and Super monocrystalline – will emerge in the near-term. Significantly, by 2012 we forecast thin film modules will comprise 50 percent of incremental demand”.

My Take: This forecast of course flies in the face of the empiricals laid out in this article. Given the preponderance of silicon on the planet and the fact that refining it is not rocket science, the markets will tend towards an oversupply of refined PV-grade silicon. The edge devolves to those who can do their magic on this refined product to reduce costs and amp up volumes. Under this scenario the odds against alternatives like CdTe occupying any significant market share appear high, more so given that unlike silicon, stuff like cadmium is highly toxic and banned from use in PV modules installed in large markets like Japan.

Article Source:

Leave A Reply

Your email address will not be published.