PV technologies: Is there a Buyer’s Advantage?
Solar cells have come a long way from being in the realm of NASA, far beyond the pocketbook of most mortals to an ever cheaper commodity that is quickly becoming a mainstream source of electricity. Throughout the history of photovoltaics (PV) different technologies have been competing for the favors of the market, touting their respective advantages. The main two classes competing for the market today are crystalline silicon and thin film technologies. Is one of these technologies clearly superior to the other?
Sometime ago the notion of different generations of solar technologies took foothold and with thin film technologies having established sizable commercial operations later than crystalline silicon they were referred to as a later generation. This gave rise to the misleading notion that they are inherently superior to crystalline Silicon panels.
There are definitely some advantages to thin film technologies. First and foremost their lower temperature coefficient and how they deal with partial shading. Let’s dive into these two:
A solar cell has properties of a diode. A diode is an electric element that only lets current pass in one direction and in that (forward) direction only above a certain voltage. That voltage is a property of the material used and of the temperature. W
hen the temperatures rise, the voltage over the diode decreases, which for a solar panel implies it will produce a lower voltage and ultimately a lower power. The nature of the material determines how sensitive a solar cell is to this derating due to temperature. These sensitivities are stated by the manufactures as temperature coefficients and are summarized for the main technologies in the table below.
The reason these temperature coefficients are of importance is that the label which sticks on the back of a solar module shows a power rating established at Standard Test Conditions (STC). This set of parameters makes sure that all solar manufacturers test their modules the same way. One of these parameters is the cell temperature which in the rating test is set to 25°C. In real world conditions a solar module is often significantly hotter: If you live in a hot place ambient temperature can regularly be 40°C plus and the module itself can be heated another 20°C above ambient by the sun, especially if mounted close to a roof. In other words a cell temperature in excess of 60°C is not uncommon. With the cell thus 35°C hotter than in the standard test the different technologies derate differently (see table).
The thin film technologies fare considerably better than the standard crystalline silicon (c-Si) technologies. In hot climates, they should have a higher energy yield per Watt installed. Yet the thin film silicon (a-Si and c-Si) manufacturers have mostly disappeared from the market by bankruptcy. Clearly, there is more to the story of what makes the best solar technology.
Let us look at partial shading next. Solar modules behave much like a hose. If you choke a hose anywhere, the flow along the entire hose is compromised. In solar modules, cells are connected in series. As long as all of them produce the same amount of current everything is fine. But when one cell has a reduced output it limits the flow through all cells connected to it in series. The most common c-Si solar panels today contain 60, 6×6 inch square cells, all in series.
In theory, if you held a 6×6 inch cover (like a large fallen leave) over a cell, the entire module would be turned off. Practically, manufacturers have alleviated the situation a bit by introducing bypasses. This way, if you shade one cell, you’ll lose the performance of 1/3 of the panel instead of all of it. This workaround is an improvement but not a fully satisfactory one because by shading 1/60 of the module we may lose as much as 1/3 of the module performance. Partial shading by relatively small objects such as a branch of a plant or a fallen leave can impact the output of c-Si panels significantly.
The manufacturing process for thin film panels is quite different, resulting a very homogeneous looking surface (an esthetical advantage of thin film panels). When looking closely at these panels you see fine lines going from side to side or from top to bottom. These scribes give the panel an aspect of a needle stripe suit and physically separate the homogeneous surface into individual cells. These cells thus run from one side of the module to the other side and are typically about 5-10mm wide.
Consequently, a small object like a leaf may cover parts of many cells but rarely a significant portion of any cell. The thin film modules are relatively tolerant to this. They behave more like a river that flows around a rock, funneling current around such an obstacle so that a partial shadow from a leaf or similar has a much smaller impact than on a c-Si panel. This seems a clear advantage of thin film panels and is regularly used as a prop at tradeshows.
But all is not won: If a cell is entirely (or mostly) shaded the same that applies to a c-Si panel applies to a thin film panel; no current can pass. And because of the way thin film modules are manufactured they often have no bypasses within the module meaning that a shadow of as little as 1cm wide over the entire width can completely turn the panel off. In the constructed environment such shadows are fairly common; a neighbor’s roof or wall or the row in front if there are several rows of PV modules are two common examples. What is thus an advantage in the case of a leaf can quickly turn into a disadvantage if there is a larger obstacle casting shade.
The below pictures are good illustrations:
The first picture shows an installation where partial shading considerations have not been given the attention they merit.
- Vegetation has grown too close to the installation, including some branches growing on the modules. In this example we see thin film panels. The branches on the modules should be cut back but the current loss is much lower than it would be with crystalline panels. 1:0 for thin film.
- The module on the top left is unfortunately installed much too close to the row behind it, casting a shade onto it during large portions of the day, significantly reducing the yield of the two modules behind it. Assuming that the scribes run from left to right in the picture, the losses are about proportional to the area shaded on these modules (and worse for the system if the modules are series connected but that’s beyond the scope of this article). Whether c-Si would have performed better or worse under these circumstances depends on whether the panel would have been mounted in landscape or portrait orientation so we’ll call this one a draw between the two technologies.
- If the scribes are from left to right then the two modules on top of the picture in the long row would pretty much never produce any power. Quite a severe loss due to the way thin film modules work.
- Equally, if the scribes run left to right, the shadow at the bottom of the picture is worrying. As the sun sets further, this shadow will move onto the module on the bottom of the picture. As soon as the first cell which measures only less than 1cm is covered that panel’s output goes to zero! With its 6 inch wide cells and bypasses c-Si wins that.
It’s quite hard to tell which technology won that match in real life, isn’t it?
The second and third pictures show another unfortunate design error: PV solar panels installed under the overhang of a roof. Inevitably, the zone on the top is going to be shaded almost permanently.
Although we can’t see it from the picture, I have to assume the scribes run from top to bottom on these modules. In this case it is lucky that thin film modules were used as the losses will be roughly equivalent to the relatively small fraction of the surface under the ledge. Would the scribes run from left to right, the output of the system would be close to zero and the installation would surely have been revised. This example shows well how a marketed advantage of thin film modules can quickly turn sour if design errors are committed.
Sometimes thin film modules are marketed as performing better in low light (cloudy) conditions. The evidence of this is thin and the generalization of it being applicable to any thin film technology is hard to justify. For starters: To get to a c-Si cell the light has passed through a glass and a plastic sheet before hitting the cell. The refraction indices of all materials involved and the surface structure of the layers define how well the light will pass into the cell.
In thin film technology there are modules where this passage glass-plastic-cell is the same but there are others where it is a direct passage from glass to cell. Similarly, other factors that could explain a low light advantage vary from one manufacturer to another. Generalizing thus seems ill advised.
Another often heard argument is the lower embedded energy (the energy needed to produce the panel) in thin film modules. There are a number of academic studies generally affirming this point. The solar industry develops quickly while academic papers and their data lag considerably because of the time it takes to compile the research and to go through the publication process. Also, they are often quoted 5 or 10 years after they are published which bears little relevance for the present products manufactured.
In essence, c-Si or thin film modules are remarkably similar with the exception of the cell. Both use glass and plastic encapsulants in comparable quantities. The blame on c-Si cells is that ultrapure silicon is energy intensive to produce. At its peak a kilogram of refined silicon cost $400. Today it trades for about $15/kg. With about 5g of silicon needed per W of solar panel and panels selling for about $0.55/W on average silicon manufacturers are under intense pressure to lower their manufacturing cost. Because the energy insensitivity of silicon refining is a cost driver, the manufacturers are continuously improving their processes to become less energy intensive.
The classical Siemens process has reportedly come down a good 50% in energy intensity in last generation reactors versus older reactors and they face competition from fluidized bed reactors and upgraded metallurgical grade silicon. So economics is driving the energy intensity of all solar panels down quite aggressively.
Ultimately, to argue whether one product takes a couple months longer to recover its embedded energy for a product that lasts 25 plus years seems to miss the point. Both technologies generate vastly more energy than their production consumes.
After having given these typical thin film talking points a critical (and hopefully fair) review I’d like to make a couple of additional considerations:
- PV modules are supposed to perform for more than 25y. There are minimum requirements as to how much performance degradation is acceptable during the life of the module. But every manufacturer is free to go beyond this minimum warranty. Which manufacturer is willing to guarantee the lowest degradation depends not only on his technology but also on his commercial strategy. Is he claiming a lower degradation to sell more, or does he have actual confidence in his product? And in case it doesn’t perform as guaranteed, will the manufacturer honor the warranty? Such points are usually evaluated by lenders for large projects and there are “bankability” rankings of the PV manufacturers. In these rankings c-Si and thin film manufacturers rank in a way that shows no clear technology advantage. But because there is only 2 large thin film manufacturers, banks will often have no data on smaller players and will regard them with suspicion.
- Thin film panel specs are mostly bespoke to a manufacturer. If a panel needs to be replaced and the series is no longer in production it can be difficult to find a compatible replacement whereas for c-Si modules there are some de facto standards with many modules possibly being able to plug a gap in a c-Si system
- More than 90% of the PV market is c-Si and thin film has been losing market share in recent years. There is only two large thin film manufacturers left worldwide. The larger one is First Solar, producing CdTe panels with the lowest temperature loss. Containing small quantities of Cadmium – a carcinogen – First Solar made the strategic and responsible decision to sell only into large scale projects where they could assure the logistics for proper recycling of the product at the end of its useful life. This way the Cadmium will not end up in our waterways. It means that for most of us this panel is not available. For a residential project your thin film module will thus most likely be a CIGS solar module. While these panels have a better temperature coefficient than c-Si, the advantage becomes smaller.
- Heterojunction solar cells (HIT) are crystalline solar cells with thin film layers added to each side. This hybrid also has the temperature advantages of the thin film technologies. In addition, these panels are often bifacial, thus able to receive light from the front and the back resulting in higher energy yields per Watt.
The technological race between thin film and c-Si will continue. But is this really what matters to the buyers? I would argue not. The bottom line is written in dollars: what we are all looking for is the solar system that gives us the cheapest solar energy. All the technological advantages of one technology or another are but a small consideration compared to what price you are getting for your system. And that price has been changing so quickly that if I could tell you which product is your best choice today, it most likely wouldn’t hold true tomorrow.