Solar Power Industry Analysis

 

(This is a report that I made for a seminar on financial mathematics and investing, and a small investment fund created by the participants of this seminar. This report is not a recommendation to buy or sell anything to anyone else).

Key findings:

* Solar energy is the best source of energy: abundant, free, clean, safe, everywhere. The only reason why solar energy does not yet contribute a significant part in world’s energy consumption is the high cost of making equipment for converting solar energy into electricity. But things are changing fast. Right now, solar electricity already costs less than retail electricity prices in some parts of the world, including Italy and California.

* Total cost per watt peak of large solar photovoltaic (PV) systems will decrease rapidly from $3.5-4/w in 2010 to $1.7-1.8/w in 2015, falling about 15%/year. A price of under $2/w means that the average cost of solar electricity will be under $0.10/kwh, making solar energy economically competitive with electricity produced from other resources (coal, nuclear, etc.). In other words, in 5 years, PV systems will achieve grid parity in most places in the world.

* The solar sector currently depends on government subsidies for its development. However, with grid parity in the near future, it will expand very rapidly even without any subsidies. The photovoltaic (PV) solar sector supplied less than 0.2% of world electricity consumption in 2010; in 2020 it will supply 2-5% or even more, for a CAGR (cumulative average growth rate) of 30-40%/year or more.

* Solar energy is becoming much more reliable than nuclear energy for countries seeking energy independence. Even though nuclear fuel is very cheap, nuclear electricity is actually more expensive than other types of electricity generation, due to high costs of building and decommissioning nuclear plants, maintenance,  disposal of nuclear waste, huge risks, and colossal costs of nuclear accidents like Chernobyl 1986 and Japan 2011.

* Short-term there may be many bumps on the road, e.g. due to reductions in feed-in-tariffs in Garmany and Italy, the two current largest PV markets, but long-term prospects are brilliant. The world PV market grew 100%, from 8GW to 16G, in 2010, and may grow to more than 20GW in 2011, due to cost reductions, high growth in some key markets including USA and China, and the emergence of many new markets.

* Competition in the PV solar sector is high.High-cost producers who can’t make it to grid-parity will fail, while low-cost high-volume producers will prosper and expand. Current tier-1 producers of PV modules, with expected capacity  by end of 2011, are: First Solar (USA, 2.4GW), Gintech (Taiwan), Hanwha Solar (China-Korea), JA Solar (China, 2.2GW cells), Jinko Solar (China, 1.5GW), Kyocera (Japan), LDK Solar (China, 2.5GW), MEMC (USA), Q-Cells (Germany), Sharp (Japan), SolarWorld (Germany), SunPower (USA, 1GW), Suntech (China, 2.2GW), Trina Solar (China, 1.7GW), Yingli (China, 1.75GW). Tier-1 Chinese companies have huge cost advantages, and expand much faster than non-Chinese ones. For example, LDK (China) will have a cost of $1.02/w, while Supower (USA) will have a cost of $1.43/w for crystalline silicon PV modules, by 4Q11.

* From the financial investment point of view, Chinese tier-1 PV companies look very attractive at this moment (13/March/2011): they are very profitable, have strong growth prospects (CAGR above 20% for the next 10 years), have forward P/E under 4 and in some cases even under 3, and price/book around 1. Even though their business is prospering, their stock price have been depressed due to market fluctuations and manipulation based on fears of a slow down.

World electricity demand:

About 20 trillion kwh of electricity generated in 2010. (Source: IAEA). Assuming average electricity price of 0.10$/kwh, this output amounts to 2 trillion $, or a bit more than 3% of world GDP.

Demand is expected to grow to more than 25 trillion kwh in 2020, and more than 30 trillion kwh in 2030. Demand growth is especially high in developing countries. Widespread use of electric cars will also be a major factor.

Electricity generated in 2010, by type of plant:
- Coal: 7.14 trillion kwh
- Gaz: 4.95 trillion kwh
- Oil: 1.35 trillion kwh
- Nuclear: 2.89 trillion kwh
- Hydro: 3.19 trillion kwh
- Other renewables: 0.52 trillion kwh

Picture: world electricity generation by source of energy, 1971 and 2007. (Source: OECD)

The market share of hydro power is decreasing due to limited supply. The market share of oil-generated electricity is also decreasing due to high cost of oil. Nuclear power is getting popular, especially in countries like France and Japan, because it provides energy independence, and nuclear fuel is very cheap. But nuclear power generation is not cheap, because of high capital costs and very high risks (Chernobyl 1986, Japan 2011 are examples of the danger of nuclear plants). The market share of coal remains stable during 1971-2011 because coal is still cheap.

Cost of electricity:

Cost of electricity = G + T + O

G= power geneartion cost (plant, fuel, maintenance, and pollution cost)
T = transmission and distribution cost
O = other costs or benefits

The costs may vary greatly from place to place (depending on available resources, infrastructure, etc.), and also from hour to hour. Peak-hour electricity (when the demand is highest during the day) costs up to twice as much as off-peak, because it requires storage and additional electricity generating power which is used only a few hours per day.

Transmission and distribution cost: $0.03-0.04/kwh. About half of capital expenditure on power grids is for transmission and distribution.

Average power generation cost, including pollution cost, by type of plant ($US cents / kwh, in 2010):
- Coal/gaz: 6 – 7
- Wind: 4 – 10
- Nuclear: 7 – 12
- Hydro: 4 – 5
- Solar: 15 – 25

The cost of pollution of coal is estimated at about $0.01-0.015/kwh, assuming $20/ton CO2.

Other costs:

Nuclear plants always pose a great environmental risk, and may be by far the most expensive form of energy if one takes that risk into account. And accidents are a reality, despite technological advances. Chernobyl 1986 accident cost hundreds of billion of dollars (in terms of lifes lost, health problems, etc.). Japan earthquake in March 2011 broke a nuclear plant and posed huge threats to the population.

Besides the fact that hydro power potential is limited, large hydro plants also pose huge environmental risks (floods, drought, irreparable change of environment).

Remarks:

Many reports of cost have misleading estimates about power generation cost, because they don’t properly take into account the cost for building the plants etc. For example, coal fuel is much cheaper than gaz, but coal plants are much more expensive, so the final cost of electricity generation from coal is not much lower than from gaz. Nuclear fuel is even cheaper than coal, but it doesn’t mean that nuclear electricity costs less. Many reports grossly understate the average cost of nuclear power generation, due to underestimation of the cost of new nuclear plants, failures (abandoned projects), nuclear waste disposal, and decommissioning of old plants.

The cost of solar power power generation will fall by more than 50%, to under $0.10/kwh, by 2015. And locally-generated solar electricity doesn’t have to bear the transmission and distribution cost, which make solar power very cost-effective compared to the other sources of alectricity. Moreover, peak-hour effect (the cost of electricity is higher during peak hours) also favors solar energy, because the sun shines during peak hours.

Electricity cost is well-reflected in retail electricity price (= cost + profit margin for electricity companies).

Picture: Retail electricity prices in the world, together with estimated cost of solar (PV) electricity (as a function of solar irradiation level), in 2009. Already in 2009, PV electricity cost is lower than retail electricity price in a number of residential markets in the world.

Solar energy technologies:

There are different technologies to turn solar irradiation into useful energy, which can broadly be divided into 2 types:

- Solar thermal power (STP): concentrate sunlight (using class, mirrors, etc.) to create heat, and then use that heat directly or turn turn it into electricity using a heat turbine.

- Photovoltaic (PV): Using the photovoltaic effect to turn sunlight directly into electricity

There are STP plants in the world, and they are at the moment cheaper per kwh than PV plants. However, this cost advantage will soon vanish. The PV sector is much larger (by a factor of 10), and also growing faster, than the STP sector, probably because PV is more convenient to install at any scale, and is also safer and more environment-friendly. In this report we will concentrate on the PV market.

The PV market can be further divided by type of material/technology used: crystalline silicon (including multi-crystalline and mono-crystalline) is the most popular one (85% of the market in 2010), thin-film (a.k.a. amorphous silicon, 13% of the market), concentrating PV (concentrate sunlight to highly efficient PV cells), and other new technologies. Crystalline silicon is an older technology and is more expensive than thin-film per watt, but is also more efficient in converting sunlight into electricity, which means it requires less space, and hence costs less in supporting hardware and installation than thin-film. That’s why C-silicon remains competitive with thin-film, and thin-film market share actually decreased in 2010 despite predictions of its market share gains.

Bellow is a table of PV technologies currently in use (source: http://www.pvresources.com):

Material Thickness Efficiency Colour Features
Monocrystalline Si solar cells 0,3 mm 15 – 18 % Dark blue, black with AR coating, grey without AR coating Lengthy production procedure, wafer sawing necessary. Best researched solar cell material – highest power/area ratio.
Polycrystalline Si solar cells 0,3 mm 13 – 15 % Blue with AR coating, silver-grey without AR coating Wafer sawing necessary. Most important production procedure at least for the next ten years.
Polycrystalline transparent Si solar cells 0,3 mm 10 % Blue with AR coating, silver-grey without AR coating Lower efficency than monocrystalline solar cells. Attractive solar cells for different BIPV applications.
EFG 0,28 mm 14 % Blue, with AR coating Limited use of this production procedure Very fast crystal growth, no wafer sawing neccesary
Polycrystalline ribbon Si solar cells 0,3 mm 12 % Blue, with AR coating, silver-grey without AR coating Limited use of this production procedure, no wafer sawing neccesary. Decrease in production costs expected in the future.
Apex (polycrystaline Si) solar cells 0,03 to 0,1 mm + ceramic substrate 9,5 % Blue, with AR coating, silver-grey without AR coating Production procedure used only by one producer, no wafer sawing, production in form of band possible. Significant decrease in production costs expected in the future.
Monocrystaline dendritic web Si solar cells 0,13 mm incl contacts 13 % Blue, with AR coating Limited use of this production procedure, no wafer sawing, production in form of band possible.
Amorphous silicon 0,0001 mm + 1 to 3 mm substrate 5 – 8 % Red-blue, Black Lower efficiency, shorter life span. No sawing necessary, possible production in the form of band.
Cadmium Telluride (CdTe) 0,008 mm + 3 mm glass substrate 6 – 9 % (module) Dark green, Black Poisonous raw materials, significant decrease in production costs expected in the future.
Copper-Indium-
Diselenide (CIS)
0,003 mm + 3 mm glass substrate 7,5 – 9,5 % (module) Black Limited Indium supply in nature. Significant decrease in production costs possible in the future.
Hybrid silicon (HIT) solar cell 0,02 mm 18 % Dark blue, black Limited use of this production procedure, higher efficiency, better temperature coefficient and lower thickness.

PV market growth:

On the average, 1w of PV generates about 1.2 kwh a year (it varies from less than 0.9kwh to more than 1.8kwh, depending on the place). So the world demand in electricity is equivalent to more than 13500GW (gigawatt) of PV in 2010, and more than 21000GW in 2020.

The current total installed base of PV in the world is about 27GW, or just 0.2% of the world demand in electricity, of which about 16GW were installed in 2010. Based on price trends, economical value of PV (PV will make more economical sense than nuclear, and will be price-competitive with coal), and environmental and safety concerns, one can expect that PV will grow to at least a few percents of world electricity in 2020, at the expense of coal, gaz, and nuclear.

To supply just 2% of world electricity demand in 2020 (a conservative number), PV will need an installed base of more than 400 GW, for a CAGR (cumulative average growth rate) of 30%/year. In order to supply 10% of world electricity demand in 2020, PV will need to grow at a very fast pace or more than 50%/year. That number is probably too optimistic, but a CAGR of 30-40%/year looks reasonable.

Drivers of growth:

- Grid-parity, i.e. will be cost-competitive with coal-fired plants in a few years (even without government subsidies). Will cost of other types of electricity can only go up, cost of PV electricity goes down.

- Safety and cleanness: much safer and cleaner than other forms of electricity generation. No emisson of CO2. Easily recyclable. This safety and cleanness premium is worth about $0.4/w of PV power. 1w of PV saves about 0.02 tons of CO2 compared to coal-fired plants over its life. Assuming that each ton of CO2 costs $20 in terms of pollution, then 0.02 tons = $0.4, i.e each watt of PV has +$0.4 externality.

- Government policies to encourage green energy: feed-in-tariffs until grid parity is achieved, grants, and tax reductions. Goal set by the EU to get at least 20% of energy from renewable sources by 2020. Similar goals in other places of the world. Governement subsidies are justified by the clean and safe nature of solar energy.

- Energy independence/security: Countries without other energy resources can achieve energy independence safely by developing PV power. (Nuclear power is another option, but it comes at huge environmental and health risks).

- Localized power geneartion which decreases the cost of transmission and distribution: can be easily installed everywhere, especially on roof-tops (doesn’t require any additional space).

- Huge virgin markets, e.g. rural areas without existing power lines (a billion-people market).  The situation is similar to mobile phones for rural areas without existing phone lines. Solar powered boats, cars, phones, etc. will become a reality.

- Integration into new buildings: new houses may integrate PV systems into the construction, thus saving cost on other construction materials (i.e. a roof can be made of PV panels instead of tiles). In the not-so-distant-future, almost all new buildings will integrate PV panels in oder to become “energy neutral homes”.

- The current PV installed base is very small (0.2% of worlds electricity), but it will become the main energy source of the future (more important than coal, wind, hydro, etc., because the usable energy of sun radiation is many times more than all the other sources combined).

- Growth in world’s electricity demand.

Economical value of PV systems and grid parity:

1 watt (peak capacity) PV generates an average of 1.2hwh/year, during 25-30 years. (In sunny places, that number will be more like 1.8kwh/year, while in northen regions its more like 0.9kwh).

Consider a PV watt like a long-term bond which gives 1.2kwh/year. Applying an real interest rate of about 3% (5% interest rate minus 2% inflation), the value of 1w PV is equivalent to about 17 times its generated electricity in 1 year, or about 20kwh. (The maintenance of of PV systems is small, and ignored here for simplicity). If the price per kwh is $0.10, then it means that 1w PV has economical value of approximately $2. In other words, at $2/w, 1.2khw/year/watt, and electricity price $0.10/kwh, PV achives grid parity, if one uses 3% real interest rate. If one uses real interest rate of 5% instead of 3%, and the other parameters remain the same, then 1w of PV power will be worth $1.70 instead of $2. If one also takes into account the value of safety and cleanness of PV compared to other energy sources (which may come in the form of lower taxes or tradable emission permits), at $0.40/w, then the value of PV will become $2.1-2.3/w.

In sunny places, and in places with high electricity prices, it may be worth much more. For example, in Italy, 1w PV may have an economical value of $5-6. In many other parts of the world, including Japan, Spain, Germany, Australia, California, PV systems have an economical value of $3/w or more, and PV systems there can achieve grid parity as early as 2012.

For new houses, one can use PV panels instead of tiles to cover the roof. PV systems will be more valuable in such cases, because not only they generate electricity, but they also replace tiles, i.e. one can deduce the cost of saved tiles from the cost of PV systems. This tile value may add about $0.1-0.2/w to the economical value of PV.

The above esitmates show that, at the cost of $2/w installed, PV will achieve grid parity in most places in the world. Even at $3/w or $4/w, PV is already at grid parity in some markets. For example, in rural areas in India without power lines (400 million Indian people have no access to electricity as of 2010), PV may be a good economical proposition right now.

Remark: To estimate economical value of large grid-connected solar farms, one should not use retail electricity price but average wholesale cost of electricity, which is only about 2/3 of retail prices. On the other hand, electricity cost will not remain constant, but will rise during 2011-1015, due to inflation, rising demand, rising cost of new plants, rising cost of fuel, phasing out of nuclear energy in many places (e.g. Germany and Japan), and environmental costs. In many parts of the world, wholesale price of electricity will surpass $0.10/kwh ?

Price and cost-of-goods-sold trend for PV:

A typical crystalline-silicon-based PV system consists of solar modules (panels for converting solar into electricity), and the rest (called BOS — balance of system). BOS consists of: inverter (for converting DC to AC), wiring, other supporting hardware, installation work, etc.

Roughly speaking, modules account for about 60% of the cost (or value) of a PV system, while BOS accounts for about 40%. These number may change, depending on the technology used, and other variables. If one adds in margins by installers (say for small PV systems), then PV moduleswill be less than 1/2 of the total cost.
In 2010, total PV system cost per watt for large system is $3.5-4, only 1/2 of which is the cost of modules.

In 2011, the cost of PV systems may fall 15-17% from 2010 levels. Wholesale price of PV module may go down to $1.4-1.5/w in 2011, and $1.3-1.4/w in 2012, but still leaving low-cost producers, who can make modules for less than $1/w, a nice gross profit margin of 25-30%.

Low cost producers like LDK Solar or Jinko Solar indicated that, their COGS of modules (based on cristalline polysilicon technology) will be lower than $1.05/w in 4Q11. LDK cost: silicon $0.27 + wafer $0.25 + cell $0.20 + module $0.30 = $1.02 by 4Q1. JKS cost: wafer + cell + module = $0.75 in 4Q10 already.

In 2015, it’s possible that COGS of modules (of low-cost producers) will fall below $0.75/w. Their prices will fall below $1.10/w, making $2/w total system cost a reality.

A scenario in 2015 with $0.75/w polisilicon-based module:

- silicon: $0.20
- wafer: $0.15
- cell: $0.15
- module: $0.25

Factors influencing cost:

- Energy cost. PV production, in particular silicon production, is energy-intensive. According to various estimates (see: http://www.pvresources.com/en/economics.php), energy packback time, i.e. the time necessary for a photovoltaic panel to generate the energy equivalent to that used to produce it, is 1 to 3 years for current PV systems.

- Silicon cost (or silicon price for producers who don’t make silicon): currently the silicon cost of each PV watt is about $0.25-0.30, but producers who don’t make silicon may have to pay up to $0.50/w in silicon. Spot price of silicon ingot is around 100$/kg in 03/2011, long-term contract prices around 70$/kg. COGS to silicon producers about 25-45$/kg, depending on the producer. GCL is the low-cost leader (around 25$/kg). Daqo New Energy also has low costs (its gross margin in 4Q10 is about 55%). LDK Solar has a cost basis of 39$/kg, and that number will decrease to about 32$/kg by 4Q11. GCL indicated that their cost of silicon ingot will decrease to about 20$/kg in the near future.

- Cost of other materials.

- Labor cost: this industry is also labor-intensive. China currently has a huge labor cost advantage. In the future, labor cost in China will increase.

- Advances in technologies: higher energy conversion efficiency, thinner materials, less needed energy, etc. will make cost per watt fall. For example, in 02/2011, JA Solar achieves 18.2% efficiency for multi-crystalline solar cells.

There is a limit as to how much cost per watt can fall, because of the prices of primary materials, capacity limits, and a huge demand once grid parity is achieved. Thats why we belive that once the $2/w mark is achieved, prices will stabilize. (They may continue to fall, due to technological advances, but at a much slower rate than before that).

Chinese companies will have the best cost strructure during the period 2011-2020, due to cheaper labor costs, cheaper prices for materials, etc. In 2010, the cost advantage of Chinese vs. European producers of modules was as high as 0.50$/w, of which 0.25 is due to labor costs, 0.25 is due to materials and electricity prices.
European and American companies can’t compete cost-wise with Chinese companies, unless most of their production is also in Asia. That’s why most of world PV production (80%) is in China, and Chinese companies dominate the market and expand fast, while European and American producers are struggling.

Remark: The cost of thin-film based modules is lower than crystalline silicon based modules per watt, mainly because it needs less material. But thin-film modules have lower conversion efficiency rate than crystalline silicon modules (10-11% vs. 15-17% as of 2010), which means that thin-film need larger surface (about 50-60% larger) than crystalline-silicon for the same power generating capacity, resulting in higher balance of system cost. So in the end, these two technologies have approximatively the same total cost per watt. First Solar in the only tier-1 leading PV maker which uses thin-film.

Picture: estimated costof PV modules, by producer, and available capacity, 2011 (high-cost capacity will not be used). Bankability means certified and can get loans from banks if buying these modules. Note that First Solar has the lowest cost per watt, but its thin-film modules require higher balance of system costs, so they dont really have cost adtavtage over c-silicon-based producers like Trina and LDK. Low-cost producers (especially the Chinese ones) will dominate the market, while high-cost small players like Evergreen Solar and Energy Conversion Drives will find it increasingly difficult to compete in this market, and will eventually be bought out or go bunkrupt: when the cost difference is $0.50/w, the low-cost guy may earn $0.40 while the high-cost guy loses $0.10 for each watt produced. Note also that the estimates in the picture  are not very accurate, and may give an impression that vertically-integrated producers achieve much lower costs than non-vertically-integrated ones: the cost estimates for non-vertically-integrated ones include the profit margins of the other steps in the process while the cost of vertically interagted ones do not include those profit margins. In reality, Jinko Solar and Jaso Solar are cost-competitive with Trina and Yingli.

Subsidies:

Before grid parity can be chieved, the solar sector has to rely on government subsidies for its growth. The subsidies can come in many forms, including cheap loans, tax breaks, and especially feed-in-tariffs (FIT).
In a FIT scheme, the governement agrees to buy back the electricity produced by PV systems at much higher prices (say 0.30-0.60$/kwh) than current market prices, for a long period of time (10-20 years). Most major countries have some kind of FIT for the solar sector. The FITs can become expensive for the governements, especially when the number of PV installations become very large, and the tariffs are too generous. Thats why governements have to change their FIT scheme every year, and in some case put a cap on the amount of PV capacity qualified for FIT. Its reasonable to reduce FITs when the costs per watt are falling rapidly. On the other hand, the governments will not want to phase out or cut FITs too abruptly, because doing so would kill the solar sector.

Once the grid parity is achieved, the governements will not need to subside the solar sector much anymore (except for an ecological bonus of say 0.4$/watt). So even though during the period 2010-2015, the subsidy costs (to the governments) for the solar sector can be quite large, the total subsidy will not be too high (a few hundred billion $ in total), for a cleantech sector which will eventually generate a trillion $ early in electricity.

In order to estimate the total amount of subsidy, we will use the notion of economical subsidy, which is equal to max(cost – economical value, 0). For example, for each installed watt whose total cost is 3$ and whose economical value is 2$, we say that the economical subsidy is 3$ – 2$ = 1$.

Year / Economical subsidy (estimates)
2010 / 50B$ (16GW, subsidy of 3$/w)
2011 / 40B$ (20GW, subsidy of 2$/w)
2012 / 30B$
2013 / 30B$
2014 / 25B$
2015 / 15B$
2016 / 0B$
Total = 190B$

The faster the sector grows in terms of volume, the faster it will get to the grid-parity point. So the best way for the governements to finetune their incentives is not to put a hard cap on the number of watts qualified for subsidy per year (doing so would harm the sector, pushing the grid-parity date further away), but to reduce subsidy per watt gradually in function of the total installed volume and market prices.

PV market in 2011:

After growing more than 100% in 2010 to 16GW, the world PV market may have some difficulties in 2011due to FIT reductions in some major markets including Germany and Italy. But the market will still grow, and may surpass 20GW in 2011. Most importnant markets in 2011are still Germany and Italy. High-growth markets this year include: USA, Japan, Canada, China, India, Australia, Greece, UK, etc.

Picture: World PV market, from 1995 to 2010.

- Germany: 7-9GW in 2011, vs 7.7GW in 2010, still the most important PV market in the world. Germany has finacial strength (so they can afford it), and the public opinion is very favorable of the solar sector (so politicians wont dare to change policies abruptly). Germany has a gradually decreasing FIT schema: FIT decreases 3% for each 1-1.5GW additional installed capacity, and this year may see FIT reductions of more than 15%. Germany’s market is maturing, and may slow down to about 5GW in 2012.

- Italy: 3GW in 2011. Despite recent fears and confusions about a change of FIT policy, Italian market is still looks attractive, because it is already at grid parity, due to a combination of very high electricity prices and high solar irradiation. Previous FITs are too generous, giving investors an internal rate of return up to 20%/year. Such FITs are not sustainable and they will have to change them. Italy is likely to adopt Germany’s policy of gradual FIT reductions. Even with significantly reduced FITs, new Italian PV systems will still be quite profitable. The 3GW number assumes no growth from 2010.

- USA: 2GW, more than doubling from 2010, according to First Solar. Many new mega-projects signed there recently. Huge potential market. California (the lagest market in the US) is already at grid parity.

- Japan: 1.6GW in 2011, up from 1GW in 2010. Good market, due to governement support and scarcity of other energy resources. Government subsidy of $0.85/w, plus tax breaks. Japan currently relies heavily on nuclear power (30% of energy consumption), but earthquakes will make nuclear plants very dangerous, so they need solar and wind to offset that. After the earthquake in 03/2011 which damaged a nuclear plant posing very serious radiation risks, Japan will push the solar initiative even harder.

- China: 1.2GW, double from 2010, according to Suntech. Huge potential market. China market is growing fast even without FIT, because of the cost advantage. The government has a program to promote PV installations in China, with a target of at least 5GW by 2015 (which will be easily surpassed). On its way to becomes world’s largest PV market.

- Canada: 0.6GW in 2011 (Ontario FIT program for PV). Big potential market.

- France: 0.5GW in 2011, down from 0.6GW in 2010. France recently anounced reduced FITs, and an annual cap at 0.5GW. Land-based projects are not attractive now, but small roof-top systems still get attractive FIT.

- Spain: 0.4GW in 2011. Spain was No.1 PV market in 2008 with 2.6GW installed that year, but fell down to nearly 0 in 2009 amid fraud allegations and attempts to make retroactive subsidy cuts for existing installations. Finally in 2010 they finished fraud investigations, and decided not to make retroactive cuts. This decision will help installers regain confidence in the Spanish market. In 2011 there will be new tariffs and caps in Spain, which will favor roof-tops over ground-mounted installations.

- Australia: 0.4GW, up from 0.3GW in 2010. Growth driven by residential market. Big potential.

- Greece: 0.3GW, double from 2010. Has high FIT of 40 Euro cents per kwh. The Greek government approved 1,793 megawatts of solar photovoltaic projects by the end of 2010, compared to 393 megawatts a year earlier, according to the environment ministry. Targets 2.2GW solar power by 2020, but applications for projects totaling 9.5GW have been submitted.

- India: 0.2-0.3GW. Huge potential market. In July 2009, India unveiled a $19 billion support plan to produce 20,000 MW of solar by 2020. 1.3GW target by 2013.

- UK: 0.2GW in 2011, doubling from 2010 (attractive incentives in UK).

- Czech Republic: 0.2GW in 2011, significant slowdown from 2009-2010, due to subsidy cuts. Roof-top installations will still get subsidies (as in many other countries).

- Korea: 0.2GW. Has a governmental PV market creation plan since 2009 to encourage the development of PV market. Hosted ExpoSolar (international exposition of PV industry) in 02/2011.

- Other countries: 1-2GW

Remark: There are lots of new markets in 2011, which almost didn’t exist in 2010. For example, Vietnam market may grow from nothing in 2010 to maybe 30MW in 2011. (They signed a 300M$ project with First Solar in December 2010). These new markets, and high growth in some key markets including USA and China, will more than make up for any possible shortfall in old markets like Germany and Italy.

Solar vs. Nuclear:

France is successful with its nuclear program, and can produce nuclear electricity at a cost of about $0.07/kwh. However, in the other countries, e.g. USA, nuclear power is much more expensive, due to many factors, including:
- Huge cost overruns: On the average, nuclear plants in the US costed 3 times as much as initially estimated. It’s also true for recent projects: $7B for a 1.5GW plant vs. initial estimates of $2-4B.
- Long construction times: A nuclear plant takes about 5 years to build minimum, but in many cases there are long delays of 2-3 or more years. There was even a plant whose construction took 14 years.
- Rising construction costs: prices increased more than 2 fold during 2000-2010.
- Safety concerns. There is even a plant built in Long Island, which was never allowed to operate. There are always risks of radiation from nuclear plants and nuclear wastes. (People still remember the Chernobyl disaster).
- High cost of waste disposal and decommisioning (more than $2B for a 1GW nuclear plant)

According to various estimates, the cost of nuclear electricity is north of $0.20/kwh in 2010 in the US (vs. optimistic estimates of $0.05-0.07/kwh). Right now solar power is already less expensive than nuclear power. And solar power costs are still coming down, unlike nuclear.

In order to get low nuclear electricity prices, a country has to achieve nuclear economy of scale like France (with its 63GW nuclear capacity). Otherwise nuclear power will be too expensive compared to other sources of electricity, even for a country like USA. For many countries, it will be much better (both economically and safety-wise) to go solar than nuclear.

Remarks: There are reports which claim that nuclear power costs as low as $0.02-0.03/w. But that number doesnt take into account the construction cost for new plants, the cost of waste disposal and decommissioning, the risk of disasters, nor the cost of failed nuclear projects.

Solar vs. Wind:

Both wind and solar will play important roles in the energy mix. Wind is still much cheaper than solar (about 1/3 to 1/2 price per kwh), and cumulative installed wind power is much larger than solar at the moment. Howerver:

- Cost of wind power cant go down much more, while solar cost is going down fast, making solar power almost as cheep as wind power in the near future.

- Wind power can be installed only in some places (with lots of wind), and cant be installed in residential areas, while solar power can be installed almost everywhere.

- Wind may generate lots of power during the windy nights when people sleep and dont need much electricity, while solar panels produce electricity during daytime, especially during the peak time when electricity is needed the most.

So wind does not pose a threat to solar. The two sources complement each other.

Competition in the PV sector:

Fierce competition among PV producers. (For installers, competition is not much of a problem). The sector requires economy of scale and very high level of capital expenditure. Due to price competition and decreasing prices, low-cost producers which achieve economy of scale (tier 1 companies) and have access to big loans will prosper and make most of the profits. Many high-cost producers (e.g. Energy Conversion Devices, Evergreen Solar) will bearly survive, and may go bankrupt. Even borderline tier 1 – tier 2 company like Canadian Solar will have difficulties competing with high-volume low-cost Chinese producers.

The sector will consolidate: most big players will become vertically integrated to save costs. Smaller players will be bought out, or will operate only in some nich markets where they can compete. Currently the sector has hundreds of players, but only a handful of tier-1 companies, with revenues in the billions $, will dominate. Most of the top-10 companies (in terms of capacity and revenues) are Chinese.

It’s difficult for new entrants to jump into the game, because it requires huge financial backing, in $billions, which is usually much more than what venture capital is ready to put up.

Chinese Companies:

Were especially bullish on Chinese companies in the PV sector, because of their growth (much higher growth rate than European and Amarican companies, due to high growth and financial strength of China, and their huge cost advantage), and very low valuations et this moment: many companies have forward P/E < 4, and some are even selling under book value. German companies (e.g. SolarWorld) and US companies (e.g. First Solar) may also be good investments, however Chinese companies look much more attractive to us at current levels.

Some stocks that were following (all of them trade on the US stock market, data are of early March 2011):

* Daqo New Anergy (ticker: DQ). Makes silicon used in PV, and also some PV wafers and modules. Current price 12.3$/share. Made 0.95$/share in 4Q10. Likely to make more than 3$/share in 2011, for a forward P/E < 4. Current market cap 246M$. Current book value 260M$, for a P/B < 1. Gross margins > 50% in 4Q10, and will remain very healthy in 2011. Revenues may hit 1B$ in 2011, for a forward P/S of 0.25. Has 200M$ cash and 150M$ debt (positive cash – debt).

* LDK Solar (ticker: LDK). Vertically integrated (from silicon to wafer to cells to modules to installation) low-cost producer of PV. Current stock price 12$/share, market cap $1.75B. Capacity in 2011: silicon 18000 tons, wafers 3.6GW, cells 1.2GW, modules 2.9GW. Revenues $3.6-4B in 2011. May achieve EPS = $4 in 2011 (4Q10 EPS > $1), for a forward P/E = 3. Gross margin 22-25%. Will IPO its silicon business (2011 revs > $1B, gross margins > 40%) in Hongkong later this year.

* Jinko Solar (ticker: JKS). Fast growing low-cost producer of PV wafers, cells and modules. Very profitable. Sold 156MW wafers, 56MW cells, and 268MW modules (total = 480MW) in 2010, revenues = 705M$. Expects to double the revenues in 2011, to 1.4-1.5B$ range, with module shipment of about 950-1000MW. Current price $24/share (American depository share), market cap 480M$ (number of ordinary shares = 80M, but each ADS = 4 ordinary shares). Equity = 400M$. Debt = 220$. Earned 2.36$/share in 4Q10, 6.62$/share in 2010 (trailing P/E = 3.6). Expected to earn 7.3$/share in 2011 (probably conservative estimate, given the fact that revenues are expected to double), for a very low forward P/E of 3.3. JKS had net earnings of 133M$ in 2010. May earn 200M$ in 2011. With 20M shares (American), it means it may earn 10$/share in 2011, for a forward P/E of 2.4.

* Jaso Solar (ticker: JASO). Low-cost producer of PV cells. Currently its volume of cells is largest in the world. (The top 7 in cells are, by decreasing order of volume in 3Q10: Jaso, Suntech, First Solar, Sharp, Yingli, Trina, Motech). 2010 shipments 1.46GW, revs 1.78B$, EPS 1.61$. Already sold out for 2011. Expected shipments 2.2GW, revenues 2.3B$, and EPS 1.7$ in 2011. Current price $7, book value = 1B$, market cap = 1.1B$. Forward P/E = 4.

* Trina Solar (ticker: TSL). Large low-cost producer of PV modules. Good gross margin at 31%. Forward P/E = 5.

* ReneSolar (ticker: SOL). Weaker gross margin than other players. Forward P/E < 4.

* Yingli Solar (ticker: YGE). Large low-cost producer of PV modules. Forward P/E = 6.

* Suntech Power (ticker: STP). Large low-cost producer. 2010 revs $2.9B, EPS $1.4, gross margin 17%. Uses both c-silicon and thin-film. Capacity at end of 2010: cells and modules 1.8GW, 500MW silicon ingot and wafers. Current price $9.2, market cap $1.7B. Expects 2011 gross margin 20-22%, cell and module production capacity 2.2GW (up 20% from 2010), revs $3.4-3.6B. Debt $1.4B. Forward P/E = 7.

Remark: Even though these Chinese companies are low-cost, they also invest strongly in R&D, have research labs at universities, and hold many patents and/or applications for patents.

Some data sources used in this analysis:

Lux Research (www.luxresearchinc.com)

http://www.pvmarketresearch.com/solar.php (IMS Research)

http://www.solarserver.com/

http://www.energy.eu/

http://www.greenrhinoenergy.com/

http://www.interpv.net/

http://www.pv-magazine.com

http://www.pvresources.com/

http://www.solarfeeds.com/

http://www.suncentricinc.com/

http://www.steelguru.com/

http://integrating-renewables.org/

http://www.world-nuclear.org/

DOE

European Photovoltaic Industry Association

isuppli

http://www.world-nuclear.org/

Financial reports and presentations of the companies

Print Friendly
 

17 comments to Solar Power Industry Analysis

  • Bảo Huy MonsterID Icon Bảo Huy

    Xin phép GS Dũng cho cháu copy bài này qua một diễn đàn về điện được không ạ?

  • admin MonsterID Icon admin

    Bảo Huy có thể copy lại được. Nếu các bạn nào quan tâm thì có thể ra đây thảo luận thêm về thông tin và phân tích cho ngành này. Chú ý là các số liệu trong bài tôi viết có thể không hoàn toàn chính xác.

  • Bảo Huy MonsterID Icon Bảo Huy

    Xin cảm ơn GS ạ.

  • Uwe Zimmermann MonsterID Icon Uwe Zimmermann

    I don’t know where the data for the grid-parity diagram comes from, but I have seen it at several places now. And there seem to be a lot of mistakes in it. For example the consumer price for electricity here in Sweden is about 0.10USD/KWh and the energy yield for PV is about 1000kWh/kWp. This would place the Sweden mark next to France. And with this obvious flaw I do not trust the rest of the data.

  • Hey There Zung,
    In addition to your post I was wondering, I would like to know about the likelihood of by making use of photovoltaic power in a furnishings factory. Is it truly worth?
    Great Job!

  • admin MonsterID Icon admin

    It depends on the place, the electricity cost, the storing technology, etc. Lots of variables. With module prices under $1/w, solar energy is becoming very attractive.

  • Christophe W. MonsterID Icon Christophe W.

    Interesting presentation, but it’s too bad you chose not to explore the option of Solar Thermal Power. From what I’ve been told, PV technologies are as of now mostly based on doped silicium cells which includes rare earth elements. The way I see it, geopolitical issues over these elements are accute, their market is already tense, and overall much too small if we ambition to replace tiles by PV…
    On the other hand STP seem to rely on simple and safe physics that have been known for decades, but on which the research seems completely under the needs and the potentials. I think it would be interesting to make a comparative study between the two : PV clearly has advantages in flexibility, logistic and integration into the existing grid that STP do not have. Yet, it seems clearer every day that a solution, if there is any to come, will imply the use of multiple sources of energy.

Leave a Reply

  

  

  

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Spam Protection by WP-SpamFree