By Iain Wilson, BloombergNEF editor. This article first appeared on the Bloomberg Terminal.
Oxford Photovoltaics Ltd., a U.K. company developing solar cells that use a crystalline material called a perovskite to achieve higher conversion efficiency, says it is aiming to have a 250-megawatt-a-year production line operating by the end of 2020.
Frank Averdung, chief executive of Oxford PV, told BloombergNEF in an interview that the line will be making “tandem” cells. These fuse Oxford PV’s cells made of a perovskite compound that has the same crystal structure as calcium titanate, with conventional silicon cells.
Averdung said: “Conventional silicon technology is limited in efficiency and reach.” The idea of the tandem cells is to be able to absorb more of the blue part of the spectrum from sunlight, and so reach higher efficiency for converting solar energy to electricity.
Oxford PV announced in December that it had achieved a 28% conversion efficiency for its tandem perovskite-based solar cell. This compares to 22% at present for an average silicon cell.
In March, Switzerland-based Meyer Burger Technology AG said it would take a stake in Oxford PV of as much as 18.8%, with an option to double the investment. The equipment supplier also entered a strategic partnership and signed an agreement to accelerate the development of mass production for the U.K. company’s tandem cells.
Also in March, Oxford PV raised 31 million pounds ($40 million) in the first phase of a Series D funding round led by Chinese wind turbine maker Xinjiang Goldwind Science & Technology Co.
But skepticism remains over how quickly commercialization of perovskite solar technologies can be achieved.
“I will get excited about perovskite when there is a disclosed commercial partnership with a major crystalline silicon module manufacturer,” Jenny Chase, head of BNEF’s solar analysis, said by email. “Until then, I note that the major companies best placed to evaluate the technology and bring it to mass market consider it to be at the R&D stage, with many questions around manufacturing feasibility and lifetime to be answered.”
Read the Q&A with Oxford PV’s Averdung and Chief Technology Officer Chris Case below.
Q: What is perovskite and what is its relevance to the solar industry?
Case: Perovskite is the name of the mineral calcium titanate, first discovered in 1839. The perovskites that are used in solar are not mined but have the same structure as that mineral. Different low-cost elements are substituted for the calcium, titanium and oxygen, while maintaining the same crystal structure as the mineral to deliver very good solar-absorbing properties.
Perovskite photovoltaics offers the opportunity to increase the performance of silicon solar, which will enable the dramatic cost reductions that are critical for accelerating the adoption rate of solar.
Q: Can you explain how perovskite photovoltaic technology differs from silicon and where you are in terms of development?
Case: A fundamental difference between silicon, as it’s used in solar cells, and perovskite is that perovskite is used in thin-film form. It only has to be a micron or less thick; about 1/200th the thickness of silicon. Compared to silicon, it is much more effective as a solar absorber and its properties can be altered by changing its composition. If you break down the industry into bulk materials like silicon or gallium arsenide and thin-film materials such as cadmium telluride, silicon has 95-plus percent of the market and thin-film the rest.
Averdung: Conventional solar silicon technology is limited in efficiency and reach. The theoretical limit is 29% and the practical limit is approximately around 25%. Beyond that point, volume production becomes very expensive. Today, if you buy a good silicon cell at volume production, they have an efficiency of around 22%. The best silicon solar cell ever has reached 26.7% efficiency on a lab scale, so that means the silicon-based industry is reaching its limit. Perovskite has a theoretical efficiency of 33% and a perovskite on silicon tandem device has a theoretical efficiency of 43%.
We have now developed our perovskite solar cell technology to a point where it is stable, has a record efficiency, has been scaled up to commercial solar cell size and is ready to move into its commercial phase.
Q: What are the challenges getting perovskite from the theoretical to the practical?
Averdung: We decided that the best way of getting into the market and delivering something very beneficial to the industry and to the world is by combining perovskite solar cell technology with silicon solar cell technology.
Solar cells work by absorbing sunlight and converting it into electricity. The part of the solar spectrum that can be used by silicon (the red part) is fixed. But on top of the silicon solar cell, you can add a perovskite solar cell. The thin-film perovskite solar cell our company has designed has a different bandgap from silicon and uses light from the blue part of the spectrum. The resulting perovskite on silicon tandem solar cell can achieve an efficiency not possible with silicon alone.
Q: What is the compromise when you approach it from this direction?
Averdung: There’s an overlap between the two cells and that overlap means you don’t get the combined efficiency of both. What you see is that you get 100% of the efficiency of the top cell and you get half of the efficiency of the bottom cell. This leads into efficiency areas that are completely impossible for silicon solar cells alone.
Initially we developed this in small area samples in our R&D organization in Oxford. Eventually we scaled up to the full commercial standard wafer size and that way we were able to achieve several things. We could demonstrate process efficiency and last year in relatively short succession we announced two certified world records. The top number is a 28% efficiency solar cell.
High efficiency is great, as it plays a major role in reducing the levelized cost of energy, but equally important is stability. Our company has spent the last five years on improving the composition and improving the manufacturing method to design stability into the solar cell. We are now using and have passed the IEC accelerated stress tests, developed by the PV industry for standard silicon solar cells in order to gauge reliability.
By subjecting our tandem cells to those tests, we get a very good understanding that the reliability of the product is up to the level where it has to be in order to become a commercial product.
Q: So are we talking about the next stage of development for the solar industry?
Averdung: It’s all about how we accelerate a market introduction. Our view is that we’ve developed the technology to a level where it can move to production. We’ve set out a pilot line — a low-volume production line — in our manufacturing site in Germany and there we generate pilot volumes. We already have pilot capability and with the new funds available to us we are now going to set up a full completely integrated manufacturing line. Starting from the raw silicon wafer to making the heterojunction silicon bottom cells, to adding the perovskite top cell, and then having an output from that line of 250 megawatts of nominal capacity.
Q: Is that a change in strategy from when you were considering a licensing model?
Averdung: Exactly. Oxford PV will become a manufacturing company. It’s an expansion of our business model. The licensing model can be an attractive model but the problem we saw is that you’re not in control of the timing. For us, the consideration is how fast we can get the best solar cell into the market. That means we need to go another route than the conventional solar cell manufacturer.
When we started the funding round, we had several options. We said it depends on how much money we can get for the company, to what level we extend our business plan. Very early, we started talking with Goldwind and what we realized during the funding round was that pretty much all the major wind companies are looking at extending their business to solar. Goldwind was one of the first to approach us during that funding round, outlining their strategy and looking at Oxford PV as a path to enhancing their solar strategy. For Goldwind, the investment into the company on the one hand is to bring the technology to the production level, and at the same time it’s a potential partnership for future installations.
Q: How far are you from commercial production?
Averdung: We expect to have our 250-megawatt line up and running by the end of 2020. Our intention is to only make the cell. Panels are a different thing. The business model that is evolving in the industry at the moment is that you have large cell manufacturers mostly in Asia with several gigawatts of cell capacity, and then you have a local module maker that takes those cells and makes the modules locally. Then you have a mix where the people that make the cells also make the modules.
Q: What is the lifetime for the perovskite portion of the tandem cell?
Case: Since we are combining these two technologies together and we intend to sell into the same market, we have to deliver the same performance and warranty expectations as the silicon cell – 25 years at a minimum.
Q: How much more expensive is it to produce a tandem cell compared with a conventional silicon-based cell?
Averdung: Obviously we increase the cost per cell because the additional perovskite cell on top adds cost. The rule of thumb would be that it may increase the cost of the solar cell by 15% to 20% and we generate an efficiency increase of above 20%. That is roughly the framework we are operating in cost-wise.
Q: Will you expand the pilot line in Germany?
Averdung: We have a fairly large facility that used to be a thin-film manufacturing site and that site is large enough to house up to 500 megawatts. The first step is to put into that factory 250 megawatts. That would be manufacturing perovskite on silicon tandem solar cells.
Q: How quickly before you ramp up production beyond the 250-megawatt level?
Averdung: Once we have demonstrated manufacturing capability, the logical next step would be to go ahead and build the next volume. That depends on the funds available to us and it’s of course something our shareholders will determine at that point in time. It could be that we have to do another larger funding round, which would give us the funds we need to get to the gigawatt-capacity level. It could be that we do an IPO [initial public offering] at that point of time. Those are things that will need to be considered in 2020 after we see that we are on track with our planning and that we see the boundary conditions support additional funding.