- Email: [email protected]
Financing PV: the fundamentals
Money
Financing PV: the fundamentals IN THE FIRST OF A 2PART COLUMN, RENEWABLE ENERGY FOCUS LOOKS AT THE MURKY WATERS THAT NEED TO BE NAVIGATED WHEN STRUCTURING AND FINANCING A PV PROJECT.
Photovoltaics (PV) offers a relatively simple and low risk technology. While the financing of PV projects should generally be carried out along the lines of conventional project finance structures, particular features of the technology and the market will necessitate a number of important adjustments. They will require banks and sponsors to look at – and take into account – issues that may be unfamiliar to them from other energy (or even renewable energy) projects. PV projects also offer a great opportunity for investment. In 2006, global industry revenues were US$10.6 billion, while capital investment through the PV business chain totalled US$2.8bn. The industry raised over US$4bn in equity and debt financing, up from US$1.8bn the previous year (source: Solarbuzz.com, Marketbuzz 2007, Annual World Solar PV Market Report).
THE PV market The PV market has been undergoing rapid development and growth in recent years. While the PV market still suffers from a lack of sufficient supplies of crystalline modules, which in turn is the consequence of a shortage
Just as with wind, Germany has shown the strongest growth rate in PV. This is almost entirely the result of the favourable and very high feed-in tariff regime (in excess of US$ 0.6/ kWh). Germany has comparatively little irradiance form sunlight so, in normal and market driven circumstances, PV projects in Germany would make little sense and would be wholly unprofitable. Japan – the other major host to PV projects – and Germany together accounted for 72% of the world’s PV installations in 2006. Germany alone accounted for 55%, a figure which, together with about one third of the world’s wind capacity installed, underlines the country’s leading position in renewable energy. The consolidated world production of PV cells increased to 2.204mW in 2006, up from 1,656mW the year before and thus a third higher (source: Solarbuzz.com, Marketbuzz 2007, March 19, 2007). It is estimated that the worldwide annual PV installation rate will reach 3.2gW by 2010; a threefold increase over 2004 market installations. World PV annual turnover is set to grow from US$6.5 billion in 2004 to US$18.5 billion by 2010 (source: Solarbuzz Inc).
of silicon (silicon contributes about 45% of the cost of a crystalline module, which is driven by the costs of the silicon wafer) the situation has improved and is expected to improve even further. This is because of the growing share of thinfilm modules which do not require silicon and which have now – by and large – become accepted, bankable technology. It also appears that the shortage of silicon is about to come to an end fairly soon. More than 70 silicon generation facilities are currently being planned and constructed, and if even half of those were to see completion, the silicon shortage should be overcome before long.
24
renewable energy focus
Today the hottest and fastest growing PV markets are Spain, but also the US, Italy and Portugal. Spain, in particular, has enacted very generous tariff legislation which pays almost as much as Germany did in the heydays of its own PV boom. The Spanish market was up 200% in 2006, the US 33% (Ibid). Expectations for return on investment in Spain are often in excess of 20%. Given the certainty of the off-take regime, and the reliability of the technology this is a very good market for investment (NB – the Spanish government has recognised the dangers of an overheating market – and the costs for the consumers associated with it – and is considering some kind of
September/October 2007
Stefan Schmitz
a cap for further installations. Some of these rumours have strongly affected the Spanish market in the last few weeks but no decision can be expected before the end of the summer break). While it is correct to say that the financing of PV products works along the lines of conventional project finance structures, there are a number of particular factors unique to PV projects which banks and sponsors will need to pay attention to.
Technology of PV modules There are two types of PV modules: wafer based crystalline silicon and thin film modules (see box, cost implications for PV modules I – crystalline, on the opposite page). The latter has an advantage in that the film can be applied very easily i.e. rolled out on a structure; it is also cheaper to manufacture and to install. It does, though, have a disadvantage as far as financing is concerned, in that only limited data on its performance is available. Crystalline silicon technologies accounted for 90% of PV module production in 2006, slightly down from 92% in 2002. The remaining 10% comes from thin film technologies (Source, Solarbuzz, id). PV modules are classified (or rated) by the power they produce under a specific set of standard test conditions. The most common rating used in Europe is the peak rating. The peak rating of a PV system is the sum total of the PV modules’ nameplate power under published standard test conditions (STC). The performance is then stated in Watt peak (Wp). This approach is not too different from that, for example, applied to wind turbines which are rated in mW to denote their peak performance. Continued on page 26
Money
Cashing in their chips: The PV industry has had to compete robustly with the semiconductor, IT and microelectronics industries over the past few years for high grade silicon which has kept prices high. This is set to change next year, and the capacity anticipated to come on line during 2008 will almost certainly lead to a cooling in the price of polysilicon (at least in the spot market). However, the jury is out on whether this increased supply will be able to meet the increased demand for PV from the new markets, such as France, Spain, Italy and Greece.
Cost implications for PV modules I – crystalline The crystalline solar sector, which accounts for over
future production capacity to fund expansion), combined
years to complete, come on line. However, predictions of
with the ability of the IC sector to support far higher poly-
when the market will return to supply/demand equi-
silicon prices (due to the higher-value applications per kg
librium vary greatly due to the uncertainly surrounding
of the raw material).
production estimates – which vary from 97,100 tonnes to 140,500 tonnes in 2010.
90% of the solar market, continues to be affected by the
The tight raw material supply has led to a contrac-
availability of high-purity silicon feedstock (nicknamed
tion in the solar market growth rate, and an industry-
The capacity anticipated to come on line during 2008 will
polysilicon). The shortfall in the supply of polysilicon to the
wide margin squeeze. Companies that are involved in
almost certainly lead to a cooling in the price of polysil-
solar market is an effect still being felt from the implosion
upstream polysilicon production activities have been the
icon (at least in the spot market). However, the jury is out
of the technology bubble; polysilicon producers such
most resilient to these effects, and the stock market best
on whether the increased supply will be able to meet the
as Wacker, Hemlock and Tokuyama were encouraged
performers include fully-integrated companies such as
increased demand for solar from new markets (such as
to invest in additional capacity to meet the perceived
REC and MEMC. In comparison, some cell manufacturers
France, Spain, Italy and Greece, all of whom have intro-
increase in demand from the integrated circuit (IC) sector,
without long-term supply contracts, such as Sunways
duced subsidy mechanisms similar to that in Germany)
only to find the market fall away.
AG, have been highly exposed. However, most cell
and lead to significant price decreases across the board
companies entered into strategic long-term supply rela-
to more normalised levels around US$30/kg.
Few realised that Germany’s renewable energy law (EEG, 2000), which introduced the first feed-in tariff subsidies for solar power generation, would transform the solar and polysilicon markets so dramatically. The German solar sector expanded rapidly, benefiting greatly from
tionships (such as Ersol’s 10-year contract with Hemlock) to protect feedstock supply. Other companies, such as PV Crystalox Solar, have decided to move upstream into polysilicon production and others into silicon recycling.
For the longer term we expect a combination of factors – including the creation of low-cost, solar-grade silicon (SGS) purification plants (solar cells require lower purity silicon than integrated circuits) and the increased
a supply of polysilicon at low prices, much of which
Recycling polysilicon scrap from the IC and solar indus-
commoditisation of the end product – to enable crystal-
had been stockpiled by the aforementioned chemical
tries has provided a cheaper and plentiful feedstock
line solar prices to decrease towards levels at which solar-
companies, and the sector grew rapidly – some estimate
supply, as solar and IC industries can create as much as
generated energy will become cost competitive with
over a 40% CAGR up to 2004. By mid-2004 most silicon
25% and 10% recyclable scrap respectively. ReneSola
fossil fuel-based generation.
stockpiles had been depleted and the solar sector began
has been an outstanding example of this opportunity,
competing with the IC sector for polysilicon supply. This
as its low-cost, secure supply of polysilicon feedstock has
led to a steady increase in polysilicon contract pricing
enabled it to grow at an astonishing pace. We predict that
(from around US$25/kg in 2003 to around US$60/kg
ReneSola’s solar wafer capacity could exceed 500mW by
in 2007) and a significant squeeze on the price of poly-
the end of 2008, placing it as one of the world’s largest
silicon on the open market (some recent spot market
wafer suppliers. This is an astonishing feat considering
For Cost implications for PV modules II – other
trades have been estimated at over US$200/kg).
that ReneSola only began operations in 2004.
technologies, see next month’s Money column.
The squeeze was clearly exacerbated by the expansion
Polysilicon production is set to rise significantly over the
plans of many solar companies (some even pre-selling
next three years as new facilities, which have taken several
Predictions for when this pivotal point will be reached range from 2010 to 2015. Thus, there are exciting times ahead for those keen to invest in the crystalline solar sector.
John-Marc Bunce, Ambrian
renewable energy focus
September/October 2007
25
Money
Looking for the new Germany: Today the hottest and fastest growing PV markets are Spain, but also the US, Italy and Portugal. Spain, in particular, has enacted very generous tariff legislation which pays almost as much as Germany did in the heydays of the PV boom there (image shows one of the buildings in the Solarsiedlung am Schlierburg, Freiburg, Germany).
Continued from page 24
Performance guaranty / degradation One of the most positive aspects of PV technology is their reliability and low maintenance. Only 2.1% of the projects had a problem with modules and their cabling (Source: Gute Noten vom Handel für die boomende Solarbranche, Erneuerbare Energien 4/2005, page 53). With no moving parts, solar cells should be able to operate reliably for 25–30 years, with virtually no maintenance. The energy yield of PV modules can be substantially enhanced if put on tracking systems which follow the sun during the day and during the year (vertical and horizontal tracking – 20%-50%, depend-ing on location). In PV module supply contracts, the manufacturer usually guarantees that in the first 10 years of the PV module’s operating life the performance of the PV modules, under STC, will not be lower than 90% of the specified minimum peak performance. These contracts typically accept that peak performance is subject to a general variance of +/- 5%, which means that the acceptable minimum peak performance would be 95% of the peak performance under STC. In relation to the first 20 or 25 years of the PV module’s operating life, the guarantee is usually for 80% of the minimum peak performance. The reason for this model is that it is generally believed that the performance of the PV modules would decrease over time. However, long term studies have shown that after 10, or even 20 years, there is no evidence of general degradation and the performance generally remains within the tolerance laid down by the manufacturer.
26
renewable energy focus
It would be impracticable to check every module to be installed in a project for its
The two measures of sunlight most commonly
compliance with the specified peak perform-
irradiance and insolation, of which irradiance
ance.
agreements
is the more important factor. The higher the
usually contain a clause – according to which
irradiance, the more sunlight hits the PV cells,
the purchaser is entitled to have a certain
and the more electrons get an energy boost
percentage (usually less than 1%) of modules
to flow through the PV system circuitry. The
checked by their own engineers or an inde-
more electrons that flow, the higher the current
pendent expert. It can also be agreed that the purchaser has the right to send its engineers to inspect the production facility and process.
(amperage); and the higher the current the more
Therefore,
purchase
used to evaluate PV system performance are
electricity (kilowatts) is produced. Although irradiance varies from moment to moment during the day, the total energy received by the system from the sun in a given year remains relatively
Energy yield study
constant from year to year, varying between 5%-
Similar to other renewable energy projects,
10% of the average.
the financing banks or equity investors will require an energy yield report from one or more reputable experts. Due to the relatively short history of the PV market in the context of project finance, as well as the demanding nature of such reports and the work required, there are few institutions that can provide such reports. Since large PV projects have only been undertaken in relatively few countries, there are many countries where no expert institutions have emerged, and there has not yet been a strong movement towards international institutions, which has happened for example in the wind market. Like in other projects, the energy yield study will assess the likely electricity output of a
Irradiance levels are affected by the angle of the sun, passing clouds, hazy weather, smog and other air pollution, all of which affect the amount of sunlight available at any given moment. Hence, these are the factors which have to be looked at by the experts and applied to their calculations. They use satellite data as well as data coming from local meteological stations. As with wind energy projects, the longer the duration of measurements, the less variation and the more accurate the predictions for energy yield can be made. While the variations can amount to about +/- 16% after the first year, they can come down to less than 10% after three years and less than 5% after 10 years (Source: Meteocontrol).
project, based on the technology, location and other circumstances. For PV projects, the most important factor to be considered is sunlight and how the applied technology performs with the sunlight available.
September/October 2007
About the author Stefan Schmitz is a partner at Squire Sanders & Dempsey in London, UK. Part 2 of his article will be published in the next issue.
Financing PV: the fundamentals IN THE FIRST OF A 2PART COLUMN, RENEWABLE ENERGY FOCUS LOOKS AT THE MURKY WATERS THAT NEED TO BE NAVIGATED WHEN STRUCTURING AND FINANCING A PV PROJECT.
Photovoltaics (PV) offers a relatively simple and low risk technology. While the financing of PV projects should generally be carried out along the lines of conventional project finance structures, particular features of the technology and the market will necessitate a number of important adjustments. They will require banks and sponsors to look at – and take into account – issues that may be unfamiliar to them from other energy (or even renewable energy) projects. PV projects also offer a great opportunity for investment. In 2006, global industry revenues were US$10.6 billion, while capital investment through the PV business chain totalled US$2.8bn. The industry raised over US$4bn in equity and debt financing, up from US$1.8bn the previous year (source: Solarbuzz.com, Marketbuzz 2007, Annual World Solar PV Market Report).
THE PV market The PV market has been undergoing rapid development and growth in recent years. While the PV market still suffers from a lack of sufficient supplies of crystalline modules, which in turn is the consequence of a shortage
Just as with wind, Germany has shown the strongest growth rate in PV. This is almost entirely the result of the favourable and very high feed-in tariff regime (in excess of US$ 0.6/ kWh). Germany has comparatively little irradiance form sunlight so, in normal and market driven circumstances, PV projects in Germany would make little sense and would be wholly unprofitable. Japan – the other major host to PV projects – and Germany together accounted for 72% of the world’s PV installations in 2006. Germany alone accounted for 55%, a figure which, together with about one third of the world’s wind capacity installed, underlines the country’s leading position in renewable energy. The consolidated world production of PV cells increased to 2.204mW in 2006, up from 1,656mW the year before and thus a third higher (source: Solarbuzz.com, Marketbuzz 2007, March 19, 2007). It is estimated that the worldwide annual PV installation rate will reach 3.2gW by 2010; a threefold increase over 2004 market installations. World PV annual turnover is set to grow from US$6.5 billion in 2004 to US$18.5 billion by 2010 (source: Solarbuzz Inc).
of silicon (silicon contributes about 45% of the cost of a crystalline module, which is driven by the costs of the silicon wafer) the situation has improved and is expected to improve even further. This is because of the growing share of thinfilm modules which do not require silicon and which have now – by and large – become accepted, bankable technology. It also appears that the shortage of silicon is about to come to an end fairly soon. More than 70 silicon generation facilities are currently being planned and constructed, and if even half of those were to see completion, the silicon shortage should be overcome before long.
24
renewable energy focus
Today the hottest and fastest growing PV markets are Spain, but also the US, Italy and Portugal. Spain, in particular, has enacted very generous tariff legislation which pays almost as much as Germany did in the heydays of its own PV boom. The Spanish market was up 200% in 2006, the US 33% (Ibid). Expectations for return on investment in Spain are often in excess of 20%. Given the certainty of the off-take regime, and the reliability of the technology this is a very good market for investment (NB – the Spanish government has recognised the dangers of an overheating market – and the costs for the consumers associated with it – and is considering some kind of
September/October 2007
Stefan Schmitz
a cap for further installations. Some of these rumours have strongly affected the Spanish market in the last few weeks but no decision can be expected before the end of the summer break). While it is correct to say that the financing of PV products works along the lines of conventional project finance structures, there are a number of particular factors unique to PV projects which banks and sponsors will need to pay attention to.
Technology of PV modules There are two types of PV modules: wafer based crystalline silicon and thin film modules (see box, cost implications for PV modules I – crystalline, on the opposite page). The latter has an advantage in that the film can be applied very easily i.e. rolled out on a structure; it is also cheaper to manufacture and to install. It does, though, have a disadvantage as far as financing is concerned, in that only limited data on its performance is available. Crystalline silicon technologies accounted for 90% of PV module production in 2006, slightly down from 92% in 2002. The remaining 10% comes from thin film technologies (Source, Solarbuzz, id). PV modules are classified (or rated) by the power they produce under a specific set of standard test conditions. The most common rating used in Europe is the peak rating. The peak rating of a PV system is the sum total of the PV modules’ nameplate power under published standard test conditions (STC). The performance is then stated in Watt peak (Wp). This approach is not too different from that, for example, applied to wind turbines which are rated in mW to denote their peak performance. Continued on page 26
Money
Cashing in their chips: The PV industry has had to compete robustly with the semiconductor, IT and microelectronics industries over the past few years for high grade silicon which has kept prices high. This is set to change next year, and the capacity anticipated to come on line during 2008 will almost certainly lead to a cooling in the price of polysilicon (at least in the spot market). However, the jury is out on whether this increased supply will be able to meet the increased demand for PV from the new markets, such as France, Spain, Italy and Greece.
Cost implications for PV modules I – crystalline The crystalline solar sector, which accounts for over
future production capacity to fund expansion), combined
years to complete, come on line. However, predictions of
with the ability of the IC sector to support far higher poly-
when the market will return to supply/demand equi-
silicon prices (due to the higher-value applications per kg
librium vary greatly due to the uncertainly surrounding
of the raw material).
production estimates – which vary from 97,100 tonnes to 140,500 tonnes in 2010.
90% of the solar market, continues to be affected by the
The tight raw material supply has led to a contrac-
availability of high-purity silicon feedstock (nicknamed
tion in the solar market growth rate, and an industry-
The capacity anticipated to come on line during 2008 will
polysilicon). The shortfall in the supply of polysilicon to the
wide margin squeeze. Companies that are involved in
almost certainly lead to a cooling in the price of polysil-
solar market is an effect still being felt from the implosion
upstream polysilicon production activities have been the
icon (at least in the spot market). However, the jury is out
of the technology bubble; polysilicon producers such
most resilient to these effects, and the stock market best
on whether the increased supply will be able to meet the
as Wacker, Hemlock and Tokuyama were encouraged
performers include fully-integrated companies such as
increased demand for solar from new markets (such as
to invest in additional capacity to meet the perceived
REC and MEMC. In comparison, some cell manufacturers
France, Spain, Italy and Greece, all of whom have intro-
increase in demand from the integrated circuit (IC) sector,
without long-term supply contracts, such as Sunways
duced subsidy mechanisms similar to that in Germany)
only to find the market fall away.
AG, have been highly exposed. However, most cell
and lead to significant price decreases across the board
companies entered into strategic long-term supply rela-
to more normalised levels around US$30/kg.
Few realised that Germany’s renewable energy law (EEG, 2000), which introduced the first feed-in tariff subsidies for solar power generation, would transform the solar and polysilicon markets so dramatically. The German solar sector expanded rapidly, benefiting greatly from
tionships (such as Ersol’s 10-year contract with Hemlock) to protect feedstock supply. Other companies, such as PV Crystalox Solar, have decided to move upstream into polysilicon production and others into silicon recycling.
For the longer term we expect a combination of factors – including the creation of low-cost, solar-grade silicon (SGS) purification plants (solar cells require lower purity silicon than integrated circuits) and the increased
a supply of polysilicon at low prices, much of which
Recycling polysilicon scrap from the IC and solar indus-
commoditisation of the end product – to enable crystal-
had been stockpiled by the aforementioned chemical
tries has provided a cheaper and plentiful feedstock
line solar prices to decrease towards levels at which solar-
companies, and the sector grew rapidly – some estimate
supply, as solar and IC industries can create as much as
generated energy will become cost competitive with
over a 40% CAGR up to 2004. By mid-2004 most silicon
25% and 10% recyclable scrap respectively. ReneSola
fossil fuel-based generation.
stockpiles had been depleted and the solar sector began
has been an outstanding example of this opportunity,
competing with the IC sector for polysilicon supply. This
as its low-cost, secure supply of polysilicon feedstock has
led to a steady increase in polysilicon contract pricing
enabled it to grow at an astonishing pace. We predict that
(from around US$25/kg in 2003 to around US$60/kg
ReneSola’s solar wafer capacity could exceed 500mW by
in 2007) and a significant squeeze on the price of poly-
the end of 2008, placing it as one of the world’s largest
silicon on the open market (some recent spot market
wafer suppliers. This is an astonishing feat considering
For Cost implications for PV modules II – other
trades have been estimated at over US$200/kg).
that ReneSola only began operations in 2004.
technologies, see next month’s Money column.
The squeeze was clearly exacerbated by the expansion
Polysilicon production is set to rise significantly over the
plans of many solar companies (some even pre-selling
next three years as new facilities, which have taken several
Predictions for when this pivotal point will be reached range from 2010 to 2015. Thus, there are exciting times ahead for those keen to invest in the crystalline solar sector.
John-Marc Bunce, Ambrian
renewable energy focus
September/October 2007
25
Money
Looking for the new Germany: Today the hottest and fastest growing PV markets are Spain, but also the US, Italy and Portugal. Spain, in particular, has enacted very generous tariff legislation which pays almost as much as Germany did in the heydays of the PV boom there (image shows one of the buildings in the Solarsiedlung am Schlierburg, Freiburg, Germany).
Continued from page 24
Performance guaranty / degradation One of the most positive aspects of PV technology is their reliability and low maintenance. Only 2.1% of the projects had a problem with modules and their cabling (Source: Gute Noten vom Handel für die boomende Solarbranche, Erneuerbare Energien 4/2005, page 53). With no moving parts, solar cells should be able to operate reliably for 25–30 years, with virtually no maintenance. The energy yield of PV modules can be substantially enhanced if put on tracking systems which follow the sun during the day and during the year (vertical and horizontal tracking – 20%-50%, depend-ing on location). In PV module supply contracts, the manufacturer usually guarantees that in the first 10 years of the PV module’s operating life the performance of the PV modules, under STC, will not be lower than 90% of the specified minimum peak performance. These contracts typically accept that peak performance is subject to a general variance of +/- 5%, which means that the acceptable minimum peak performance would be 95% of the peak performance under STC. In relation to the first 20 or 25 years of the PV module’s operating life, the guarantee is usually for 80% of the minimum peak performance. The reason for this model is that it is generally believed that the performance of the PV modules would decrease over time. However, long term studies have shown that after 10, or even 20 years, there is no evidence of general degradation and the performance generally remains within the tolerance laid down by the manufacturer.
26
renewable energy focus
It would be impracticable to check every module to be installed in a project for its
The two measures of sunlight most commonly
compliance with the specified peak perform-
irradiance and insolation, of which irradiance
ance.
agreements
is the more important factor. The higher the
usually contain a clause – according to which
irradiance, the more sunlight hits the PV cells,
the purchaser is entitled to have a certain
and the more electrons get an energy boost
percentage (usually less than 1%) of modules
to flow through the PV system circuitry. The
checked by their own engineers or an inde-
more electrons that flow, the higher the current
pendent expert. It can also be agreed that the purchaser has the right to send its engineers to inspect the production facility and process.
(amperage); and the higher the current the more
Therefore,
purchase
used to evaluate PV system performance are
electricity (kilowatts) is produced. Although irradiance varies from moment to moment during the day, the total energy received by the system from the sun in a given year remains relatively
Energy yield study
constant from year to year, varying between 5%-
Similar to other renewable energy projects,
10% of the average.
the financing banks or equity investors will require an energy yield report from one or more reputable experts. Due to the relatively short history of the PV market in the context of project finance, as well as the demanding nature of such reports and the work required, there are few institutions that can provide such reports. Since large PV projects have only been undertaken in relatively few countries, there are many countries where no expert institutions have emerged, and there has not yet been a strong movement towards international institutions, which has happened for example in the wind market. Like in other projects, the energy yield study will assess the likely electricity output of a
Irradiance levels are affected by the angle of the sun, passing clouds, hazy weather, smog and other air pollution, all of which affect the amount of sunlight available at any given moment. Hence, these are the factors which have to be looked at by the experts and applied to their calculations. They use satellite data as well as data coming from local meteological stations. As with wind energy projects, the longer the duration of measurements, the less variation and the more accurate the predictions for energy yield can be made. While the variations can amount to about +/- 16% after the first year, they can come down to less than 10% after three years and less than 5% after 10 years (Source: Meteocontrol).
project, based on the technology, location and other circumstances. For PV projects, the most important factor to be considered is sunlight and how the applied technology performs with the sunlight available.
September/October 2007
About the author Stefan Schmitz is a partner at Squire Sanders & Dempsey in London, UK. Part 2 of his article will be published in the next issue.