Michel Cauwe serves as Senior Vice-President of Umicore’s Thin Film Products business unit, based in Providence, Rhode Island (USA). Prior to this he was Senior Vice-President for Umicore’s Electro-Optic Materials business based in Olen, Belgium. Mr. Cauwe is a graduate of the University of Leuven.
What are some of the advantages (and disadvantages) of thin film Photovoltaic (PV) technologies over the more conventional crystalline silicon panels that make up most of the market?
Just to put things into perspective, conventional crystalline silicon (c-Si)-based PV now makes up approximately 90% of the market. On average, from 1980-2013 the market share of thin film PV has always been only around 10%. There have been a few cycles, though, where the market share went up to some 15% – most recently around 2008-2009. Most people, including decision-makers, wrongly focus on these short upswings in the market share of thin film PV.
One of the main advantages of thin film PV (TF-PV) technologies is that most of them yield a larger kilowatt-hour (kWh) / kilowatt peak (kWp) output, making their levelised cost of electricity (LCOE) potentially lower. This is in part due to the lower temperature coefficient for the corresponding PV cells, but is also related to the exact sunlight spectrum that is effectively absorbed by the cell – what we call ‘band-gap tuning’.
Secondly, thin film PV technologies have, in general, a much lower material balance (the ratio of input to output) and therefore use fewer natural resources. This is particularly the case for thin-film silicon, where – mainly for rigid, glass based modules – only abundant materials, like silicon and aluminium, are used.
Thirdly, TF-PV technologies can produce PV modules directly, by ‘monolithic integration’ of the cells, where the connections are created in situ and the cells created on a glass substrate or superstrate. This makes the module manufacturing much simpler, and cheaper.
As for the disadvantages – none, or almost none of the TF-PV technologies have reached the economies of scale of their crystalline-silicon counterparts. In c-Si manufacturing, the leading (Chinese) cell- and module-makers have capacities of about 2.5 – 3.5 GWp and are loaded to at least 90%. No TF-PV manufacturing operates in that league. This lack of economies of scale makes it more expensive to produce TF-PV modules today.
TF-PV always requires large-area vacuum deposition systems, to deposit the thin layers. Such systems carry high investment costs, which give rise to low machine costs (in terms of depreciation and maintenance) only when they are fully loaded – and for very large deposition systems. In other words, large glass/film unit areas, or high deposition speeds.
Another disadvantage is that all TF-PV technologies have lower conversion efficiencies than their c-Si counterparts. This means that more area is needed for a given rated power (Wp). This has an effect both on the cost of the module – a greater substrate area and a greater area of relatively expensive encapsulation materials – and on the cost of the balance of systems (BOS), such as module mounting systems. This means that the PV system cost per kWp easily gets higher than for c-Si PV systems. And although the higher energy yield (kWh/kWp) often makes up for this higher system cost, it remains a difficult message to the market, and is not easily understood by most people who are not specialised in the technology.
Thin film technologies are particularly suited for applications in Building Integrated PV (BIPV). Could you say more about these exciting applications?
Our view on this is that only roof-integrated PV really makes economic sense. PV facades intrinsically have a low energy yield (kWh/kWp) due to their limited insolation (low average angle of light incidence). For roof integration, BIPV means that the PV-panel itself is the roofing material (or is intimately connected, or conformal, with it). Only flexible PV modules can really result in such an integration. This limits the available technologies to flexible TF-PV, unless c-Si wafers, on which c-Si cells are made, become so thin that they are flexible.
A roofing material can be flexible or rigid and can, for example, be ceramic tiles or standing seam metal roofing. But first of all it has to provide shelter against the weather. Nobody wants a leaky roof. This, together with the conservatism of the building industry, means that the flexible modules themselves cannot really act as the roofing material as such. They have to be attached, for example by gluing, to safe (i.e. certified) roofing materials that already have a proven track record.
There is still quite some development needed to assure a lifetime of, say, 30 years (typically 50 years in Europe) for such integrated products. This means that, today, BIPV is less than 1% of the overall yearly PV market. Only in advanced architecture, where people are deliberately spending money to have something special, or in areas of the world where there have been more interesting incentives for BIPV (such as France, for a short period of time), has BIPV really been used.
A few years ago thin film technologies for PV applications were seen as a breakthrough for the PV industry. But the market didn’t seem to materialise as expected and several start-ups went out of business. What is the state of play today?
The reality today is that c-Si PV modules are sold in the market at a price of USD 0.60–0.80/Wp,1 even with a small profit. Any new or other technology which aims to enter the market in 2-3 years from now must be able to prove a PV module sales price potential (not cost!) at that time of some USD 0.40-0.50 /Wp.1 This is not so easy to do. To my surprise, new start-ups in TF-PV are still launching product developments without a realistic view of their cost potential, although much less than a few years ago.
But, in my opinion, the real future is in the combination of thin film technology and crystalline Si technology. We are already seeing the success of ‘rearside passivated’ c-Si cells with higher cell efficiencies (so-called PERC, PERL and PERT cells). But all the advantages of both technology groups come together uniquely in what are known as hetero-junction a-Si/c-Si cells. The nice thing about those cells is that they can be produced with a very low thermal budget – the maximum manufacturing temperature being around 200°C. This makes it possible to use very thin silicon wafers, maybe less than 100 µm thick. Modules made with such cells have a cost potential below USD 0.401 /Wp, even if an indium-based transparent conducting oxide is needed.
Also, the temperature coefficient of such cells is lower than for standard c-Si cells, which should result in higher PV-system energy output (kWh/kWp). You will see this family of products developing strongly over the next 3-5 years.
Some critics have said that thin film PV will remain a niche market. Do you agree?
In the short to mid-term (horizon 5 years), we agree. However, there are still a lot of developments on the horizon that can completely reshape the landscape in the long term. Here, we are referring to copper-zinc-tin-sulphide (CZTS) as a replacement for copper-indium-gallium-(di)selenide (CIGS) – an absorber that uses only abundant elements with lower cost – to perovskites or the use of quantum dots. Each of these has the intrinsic potential to completely turn the situation around.
What part do you think thin film PV can play in helping to achieve the EU SET-Plan targets for low-carbon energy for 2020 and beyond?
Referring to what I said earlier, thin-film will be approximately 10% of the total share of PV. TF-PV will be used less in BIPV – as there are almost no flexible TF-PV products left on the market – but mainly in roof-top and ground mounted, grid-connected systems. The higher kWh/kWp potential will become more important, especially in Southern Europe where grid parity is becoming a reality, provided the kWp cost is not too far off from c-Si PV systems.
1 USD 1.00 = EUR 0.82 - December 2014