By Antonio Ruffini, Editor, ESI Africa magazine
One of the world’s largest integrated solar photovoltaic groups has begun to focus on South and southern Africa, thanks to the size of SA’s renewable energy independent power producer procurement programme (REIPPPP). South Africa’s renewable independent power producer programme of 3,725 MW and its extension during 2012 by another 3.0 GW is our frame of reference,” vice president and general manager for Africa and the Middle East at Canadian Solar, Jos van der Hyden, says. “We are using that to estimate the sizing for South Africa’s market and see it as the driver for the solar industry in the region.”
While South Africa is definitely not among the leading nations in solar photovoltaic (PV) developments globally, its well-structured government programme does put it among the leaders of a second tier of countries emerging in the sector. Saudi Arabia has announced ambitious plans for solar capacity to be in place by 2032. Morocco is in a similar phase with plans for a 2.0 GW solar programme and ran the tender for the first 165 MW in 2011. It selected concentrated solar power (CSP). Brazil is ahead of South Africa in terms of its wind programme, and at the end of 2012 was creating a model for grid connected solar systems and power purchase agreements for larger systems. “However, South Africa is further advanced than all these countries with its programme for large scale grid connected solar projects. Tenders have been awarded and a proper legal structure is in place.”
Canadian Solar is one of the top three manufacturers of PV panels that have emerged globally over the past few years and is vertically integrated throughout the whole solar PV chain.
During 2012 the company established an office and legal entity in South Africa. This was too late for the company to feature in the first two bidding rounds for REIPPPP projects, but Canadian Solar is certainly interested in round three. “We want to be a player in South and sub-Saharan Africa. We like to work with local partners and have signed up the Romano group and Powermode as authorised resellers. Both have a presence in a couple more countries in the region. Globally our local partners are typically engineering procurement construction (EPC) entities or distributors that can sell into off grid markets, and our partnerships in South Africa reflect that. The Romano group has EPC capability in the rooftop segment and our belief is that there will be more rooftop PV programmes in South Africa.” Shopping centres and large warehouse buildings could be prime locations for rooftop PV installations.
The Canadian-based multinational is also an EPC group in its own right, but only in selected markets, mainly Canada and China. “We are even a development company but only in Canada and possibly other very specific markets. For South Africa we have a 50/50 joint venture with SkyPower, a Canadian solar park developer, and this joint venture could be a bidder in round three of the REIPPPP.”
“We are also watching the potential of other segments of the PV sector in the region, including the off-grid market, the use of solar PV by the mining industry which has a specific strategy, and the rooftop market segment. Eskom itself has some PV projects in which it is interested and that could become a separate market segment. There are the other southern African markets. We are looking at all these,” van der Hyden says.
The opportunity for solar PV in the mining sector across Africa exists as such operations are often far from the grid and require a lot of power. Typically they have their own generator sets or gas plants and the cost per kWh is usually high. “It could be an economical option if part of their power requirements were generated from solar energy. A lot of people are investigating this, though the problem has not yet been solved since a mining operation usually runs for 24 hours a day and if a solar solution provides power for only 10 hours it becomes less attractive. We think the first pilot projects will be instructive, and are looking at this market segment in countries such as South Africa, Chile, and Australia.”
Van der Hyden observes that a significant number of people in South Africa are not yet connected to the grid. Canadian Solar has a solar home system and is working on small systems that could use a combination of diesel and PV.
Regarding rooftop PV, in general worldwide, feed-in tariffs and subsidy schemes for this are higher than for ground mounted systems. This confirms that it is more expensive to build rooftop systems. “That said, every country and every situation is unique. Some rooftops are easier to build on than others. New buildings should be obliged to include PV as, if this is integrated at the construction stage, the cost is much less than if it is a retrofit. Large logistics buildings with large rooftops could be suitable, but one also has to take into account that areas with heavy rain could see water penetrating the roof. The PV industry will have to learn what the proper approach is regarding rooftop PV in South Africa, and the answers will typically be site specific.”
Overall, South Africa’s renewables programme is achieving its objective of attracting interest from major groups across the world. But as Brian Dames, CEO of Eskom points out, the new renewables capacity comes in at a price that is five times higher than the country’s existing installed capacity’s cost of generation. Will the PV plants being established under the auspices of South Africa’s REIPPPP, some of which will be 75 MW in size, be an expensive anomaly when one looks back from the future at the country’s power sector history? From the perspective of future history, will it have been a wise move by South Africans to fund and build such power plants?
Over the past four years a revolution has occurred in the solar PV sector, which has seen the emergence of manufacturing facilities that operate on a scale that was previously unknown. Canadian Solar owns eight of these plants in Canada and China with a capacity of over 2.0 GW a year, which covers the whole chain from silicon wafers to solar cells and completed modules using large scale manufacturing principles. “Five years ago this manufacturing chain would have required four or five different companies,” van der Hyden says.
This scaling up has led to the dramatic drop in the cost of production of solar panels. A study from McKinsey indicated that between about 2009 and 2012, the price of solar PV dropped from about US$4/Watt to US$1/Watt.
In this light van der Hyden notes that while there is an element of job creation and local content in the South African programme, the harsh reality is that in the solar PV sector, local manufacture of panels would only come at a high cost. “In the solar industry cost reduction and scale increase are directly linked and South Africa is too small a market to establish a 2.0 GW factory, which is what is required to be competitive in the PV market. Of course we understand the ambition of South Africa’s government to ensure that jobs are created locally, and this will be achieved in the construction, engineering and maintenance of the facilities.”
In general, van der Hyden argues in favour of a free trade environment, as trying to force localisation will increase the cost of modules produced, and increase the cost of power produced. South Africa is not unusual in its desire for localisation. “It is the same situation we encounter in countries ranging from Brazil to Saudi Arabia. Our argument is the same when all these places attempt to force localisation of manufacture of panels. If done, the asking prices for modules will be much higher than those commonly found on the internet. Governments must think twice as to whether this is the right condition to try and impose on the sector.”
The rapid expansion of the global solar industry, which has seen it grow at 50% a year over the past two decades, has resulted in short term overcapacity. This has put pressure on pricing. Many of the large solar PV companies are listed on NASDAQ or similar exchanges, which creates a great deal of transparency. Currently their gross margin range is about 10%, which indicates they can’t lower prices much without going out of business. “Those companies with high production costs have gone or are going out of business, and some of the extra capacity was removed in 2012,” van der Hyden says. “Right now, going forward, the cost reductions won’t be as dramatic as in the last four years. At a certain point, we are impacted by the basic materials cost; glass, aluminium, silicon.”
Solar PV does in part obey Moore’s Law, where electronic processing power or processing speed doubles every two or so years and hence price per capacity reduces accordingly, but only in part. “In solar we have half of Moore’s law, because in micro-electronics one can reduce scale. In solar one cannot miniaturise as one needs surface area to catch the rays of the sun. In fact we would like to enlarge the surface. That part of Moore’s law does not apply. The other part, that of scaling up production and increasing efficiency, does apply though.” This means, in spite of overcapacity, the price reductions are sustainable and won’t be reversed, though the downward cost trajectory is likely to be slower from now on.
The coming of age of PV is illustrated by the rapid growth in the world market. In 2011 the global installed capacity was about 25 GW, and this grew to about 60 GW in 2012. Even more exciting for this technology sector is that up until now solar PV has relied on government incentives, but it has reached the point where it is achieving grid parity in certain parts of the world.
There are places with lots of good quality sunlight coupled with high electricity prices that enable PV to compete. Among these are southern California, southern Italy and Japan. “That is exciting for the solar industry. Power plants are being built to an unprecedented scale, up to 500 MW. In Canada a 250 MW facility is being built. Coupled to the increasing pricing parity this type of scale is helping people understand that we can build systems using PV technology that are the size of gas or coal plants,” van der Hyden says.
In South Africa, historically, Eskom successfully used mine mouth coal fired technology to produce low cost electricity. In addition, the electricity price was kept artificially low over the past decades, and even with the much decried increases of late the price of electricity in South Africa remains low in a global context. The country has good sunshine, but PV projects are not yet in a position to achieve grid parity.
Solar PV technology is also likely to evolve but at the moment 85% of all the panels being produced globally feature polycrystalline cells. That is the workhorse of the industry. “While not the highest in terms of efficiency, they can be produced at low cost and at the moment this provides the lowest cost per kWh. At the same time mono-crystalline and some thin film technologies are bankable.” It is expected that the majority of South African solar PV projects are likely to stick with polycrystalline cells.
Van der Hyden says that there isn’t any bias against thin film technology. But silicon technology has a very strong competitive edge. “We need to encourage all the technologies to compete. That’s the only way for the industry to grow. The greater focus will be on the strength of the company involved. Canadian Solar has one of the better balance sheets and our warranties are backed by leading insurance companies that underwrite this warranty.”
The company does offer a new technology called efficient long-term photovoltaic solution (ELPS). ELPS features metal wrap through (MWT) cell technology that delivers one of the highest mono-crystalline cell (P-type) efficiencies of up to 21.1%. However, currently this technology remains more expensive than polycrystalline cells.
Some in South Africa favour concentrated solar power (CSP) over PV as the solar technology of the future, with its conduciveness to storage, hybridisation with gas, and possibly even with coal generation, and its familiar steam cycle. Morocco is pursuing that route, and South Africa has very high direct normal irradiance (DNI) as well as gas potential in the Karoo.
However, PV is very far from losing the battle against CSP, and if anything the reverse seems true at the moment. “Because we are a PV company we have taken a lot of time to understand competing technologies in depth. We wanted to know if PV is simply enjoying a brief window of opportunity or whether we are in business for the long term,” van der Hyden says. The outcome of this investigation favours PV.
Four years ago CSP seemed to be on a lower cost curve trajectory than PV, and that information, though now out of date, remains in the reports used by some consultants to the power sector. “People missed the 75% cost reduction in PV, and during 2012 there was a further 25% to 30% cost reduction. At this moment, the actual cost for PV is much lower than CSP, and the cost curve going forward is trending lower for PV than CSP. PV has beaten CSP,” van der Hyden believes.
In the end the market will decide, and PV currently holds a big advantage with some 50 GW to 60 GW installed compared with CSP with under 3.0 GW. But more telling is that about three to four years ago in California some 8.0 GW of CSP projects were announced, and subsequently 75% of these have been converted to PV. The market is the ultimate arbitrator; PV holds ascendency and appears unlikely to relinquish that to competing solar technologies in the near future.