Yesterday I caught up with Dr Pierre Verlinden, the chief scientist at the world’s biggest producer of solar modules in 2014 – Trina Solar of China.
No, he didn’t tell me that we should expect the price of solar power will suddenly drop by half in the next two years to blow everything else away in its wake (see note at bottom for why this differs to analysis from Deutsche Bank). In fact his view was that prices for modules are likely to barely change until around 2017 when we might again expect a resumption of price reductions.
Is this bad news?
Our chat in many respects boiled down to a tussle between me being the typical journalist after a simple, single answer on the price per watt of a solar module we can expect in the future, and him the scientist who felt a single metric was too simplistic. Instead he wanted to talk more about the process and drivers for how we might see the price of power from solar evolve and improve over time.
The weird thing is that the conversation ended up landing on an anecdote about a surf lifesaving club in the beach holiday town of Tathra that I visited on a recent camping holiday.
The Tathra Surf Club has a great big billboard on it emblazoned with the logo of the Australian Government’s Greenhouse Office. This large billboard dating back to late 2006 informs people of how the surf club happens to have this amazing 1 kilowatt solar system on it as well as a wind turbine generator that can miraculously feed electricity back into the grid. The billboard contains a series of diagrams explaining how this technological marvel all takes place.
The installation ended up costing northwards of $10,000 with the government chipping in a very large proportion of the cost.
The Tathra Surf Life Saving Club, with solar system on right hand side
Photo from http://cleanenergyforeternity.net.au
Today the billboard comes across as incredibly quaint. There are houses all around the town and surrounding area with systems on their roofs many times the size of the one on the surf club. No one would be remotely impressed with this puny 1kW system today. And no one more would find the concept of feeding power from a rooftop solar system back into the grid as novel.
In addition, I couldn’t help myself boring my holiday companions by pointing out that the system on the surf club had six solar modules, whereas now you’d only need four panels thanks to significant advancements in the efficiency with which a given area of solar panel can convert photons of sunlight into electricity.
How’s this relate back to what Trina’s chief scientist sees in the future?
The first thing is that the incredible speed and size of price reductions we’ve seen since that system went in at Tathra were a function of events that we can’t expect to be repeated.
In 2006, the solar industry was almost a token, bit player reliant on the silicon offcuts of the computer industry. It has now developed large-scale silicon production facilities purely dedicated to solar power and 90% of polysilicon production is consumed by the solar sector compared to just 27% back in 2001. Silicon prices today are around $20 per kilogram instead of $300, thanks to dedicated solar grade production facilities.
Also, costs are heavily influenced by improvements gained via learning by doing, which is a function of cumulative production. As a rule of thumb you can expect a 25% reduction in cost for a doubling of cumulative installed capacity. Back in 2006 volumes of production were tiny relative to the size of the overall power sector, so it didn’t take much volume to see significant gains. Now solar represents the largest proportion of new incremental power generation capacity being installed in the developed world.
Also, the industry back around 2006-2007 had very large profit margins which then completely evaporated by 2013 such that the industry has been barely covering its costs. Over 2014 global module prices have since been pretty much stable, reflecting a recovery of profit margin. According to Verlinden, gross margins currently being achieved of 15% are still insufficient to deliver a reasonable return on capital. So while panel manufacturing cost reductions will continue, these shouldn’t really translate into price reductions until margins improve further. Indeed, a review of Trina’s 2014 third quarter results reveals that after operating expenses, interest and income tax they made a net margin of just 1.7%.
The second message is we shouldn’t obsess with a single figure of the price per watt of solar panel modules. The prices for modules to Australia (which tend to be lower than Europe, the US and Japan) for large wholesalers can be as low as 50-60 cents per watt, on a total installed price for a system of closer to $1.70 to $2.50 (pre rebates). To drop prices for the module by 10 cents per watt while maintaining a reasonable profit would represent a major achievement for module manufacturers. But for the end consumer it’s a saving of just 5%.
Verlinden argues improving the energy conversion efficiency of panels should be worthy of more attention because it has an important influence over not just the cost of the module itself but also the economics of the two-thirds or more of system costs outside of the production of the module. If you can manage to reduce the number of panels for a given amount of power output it can make a critical difference in reducing installation labour costs per watt, as well as shipping and warehousing costs. Also, panels whose output degrades less over time maximises the bang you get for a given installation.
Verlinden believes that to get a 10% reduction in the overall cost per kilowatt-hour of energy (what we ultimately want from the solar panel is energy, not the panel itself) he would have to reduce the module cost by 40% – an incredible feat. But he could get the same outcome by improving the conversion efficiency from 18% to 20% which he believes is far more achievable. Verlinden also said that we might ultimately expect multi-junction solar cells from Trina that could take this further. But they were likely to be around a decade away from commercial production.
Ultimately, I managed to squeeze a module price per watt prediction from Trina – expect not much change to 2017. And by 2020 something close to 45 cents. But given everything else I was told, there will be bigger gains available if we think in terms of cost per kilowatt-hour rather than just module cost per watt.
*A note on Deutsche Bank claims of a 40% reduction in solar system installation costs.
Yesterday Renew Economy carried the report, Why solar costs will fall another 40% in just two years, citing Deutsche Bank analysis which appears to contradict some of the points above. However, Deutsche Bank’s 40% reduction is off the base of installations costs in the United States which it estimates at US$2.90 per watt in 2014 and which it expected to fall to US$1.77 per watt in 2017. Yet in Australia a range of solar PV retailers are quoting solar systems fully installed at prices below US$1.77/watt (pre rebate) and prices well below US$2.50 are easily obtainable. This is a function of a vastly more efficient solar supply chain in Australia, less red tape involved in grid connection and accessing government subsidies and the unusual fact that Australia can obtain modules at lower prices than Europe, the US and Japan due in part to an absence of tariff barriers and possibly that Australia is used to clear excess stock and also is more willing to accept modules from less established and lower quality manufacturers. For example, some large Australian wholesalers are reported to be able to obtain solar PV modules at prices of US50 cents to US60 cents per watt, yet Deutsche Bank assumes such prices won't be achieved until 2017.