In recent years however, changes have begun to take place in module design. Switzerland is adopting an especially active leadership role in these developments. Together with numerous collaborators, René Schmid Architekten have built Switzerland’s first energy-self-sufficient apartment building in Brütten (fig. 2).1) Its grid-independent energy concept is not the only unusual feature of this building – it also boasts a 485 m2 photovoltaic facade of micromorph thin-film modules. The glass cover panes were retroactively sandblasted and coated with water repellent. Their strikingly matte texture is more reminiscent of fibre cement panels than solar modules. According to Eric Langenskiöld, photovoltaics expert at the engineering firm Basler & Hofmann, the surface treatment did not compromise the power output of the modules, which is about 100–110 W/m2. ”Our laboratory measurements showed no declines in yield, and measurements performed during building operations have indicated no deviations thus far.
Architect Mark Röösli used partial etching of the front glass to achieve the striated texture of the photovoltaic modules that he used on the west-facing gable of a ten-unit apartment building in Aesch, near Lucerne (fig. 3). The process allowed him to visually adapt the modules to the timber panelling of the long facades and the plinth. In this case, the individual cells of the high-performance crystalline silicon modules he employed can barely be distinguished behind the module surface.2)
After being renovated and furnished with additional storeys by architects Viridén + Partner, an apartment building in central Zurich built in 1981, is clad in an unobtrusive shade of grey (fig. 1). Here, too, the appearance of the rear-ventilated glass curtain facade does not reveal its photovoltaic activation. In all, just under 1,700 monocrystalline modules in 18 sizes are incorporated into the facades. The reverse sides of the satin-finished safety glass cover panels are printed with a glass ceramic layer.3) Karl Viridén estimates that the electricity generated by the modules will amortise their additional cost (compared to a non-activated glass facade) over a period of about 15 years.
The colouration of solar modules
In principle, there are three components of photovoltaics that can be dyed to produce coloured modules: the solar cells themselves, the facing glass or the embedding films within the module (fig. 4, 7−13). Coloured crystalline solar cells are not a new phenomenon (fig. 5). The colours arise from slight modifications in the manufacturing process. Silicon wafers are grey in their natural state and only acquire their blue or black hue after the application of a nanometre-thin, antireflective coating, which significantly raises the efficiency of the cells. Changes in the thickness of this coating can produce just about any imaginable colour, at the cost of a 15 – 30 % reduction in energy yields. This technique has been successfully employed for about 15 years and is relatively inexpensive. However, it does not lend the modules a homogeneous appearance, since the characteristic crystalline structure of the cells is preserved. The reflective busbars on the front of the cells likewise remain visible in the facade.
Filter films
A second method of changing module colour is through the use of coloured or filter films that are either deposited onto or integrated into photovoltaic modules. In 2014, the Swiss research centre CSEM achieved worldwide renown when its researchers managed to produce pure white solar cells using such films for the first time (fig. 6). From a physical standpoint this achievement appears somewhat counterintuitive, since white surfaces reflect all visible wavelengths of incident light. The modules therefore absorb only the invisible (mainly infrared) portions of the spectrum and convert these into electricity, resulting in a loss in module yield of approximately 40 %. If, however, highly efficient heterojunction (HJT) solar cells are employed, overall module efficiencies of over 11 % are nevertheless achievable (fig. 6, 17). In principle, this technique can be used to turn all types of crystalline silicon module white – even retroactively. The exterior of the facing glass is laminated with a diffuser and filter film and protected with an additional ETFE foil (fig. 11).
In order to limit the reflectivity to visible wavelengths, researchers have made use of the phenomenon of interference. Numerous nanometre-thin metal oxide coatings, the indices of which reflection alternate between high and low values, are deposited onto the filter film. To ensure that the film appears uniformly white instead of shimmering in all colours of the rainbow, a strongly diffractive surface structure is also utilised. In principle, modifications in the layering on the reflector foil will produce other colours as well, though this process would require a considerable developmental investment. Currently, the Swiss enterprise Solaxess4) manufactures the film. Solaxess’ first industrial partner, the Belgian firm Issol, intends to integrate the films into solar modules and sell them commercially.5)
Glasses with spectrally selective coatings
A colouration technique that has already been proven in practice was developed at the Swiss Polytechnical Institute in Lausanne (EPFL) and is now being marketed by the Swiss start-up enterprise SwissInso.6) In this technique, nanometre-thin oxide coatings, which reflect a narrowly defined bandwidth of visible light, are deposited onto the reverse side of the front glass. An etching process renders the facing side matte (fig. 18). Here too, variations in the number and thickness of the coating layers result in varying colours of glass. The glass is not only suitable for all types of PV modules, but also for thermal solar collectors.7)
The effect of the module colouration is strikingly showcased in the new building of the International School in Copenhagen by C.F. Møller Architects (fig. 15). The facade modules, each 70 ≈ 70 cm in size, are mounted on variously inclined metal cassettes. The inclination angle determines whether they appear brighter or darker, and changes in the incidence of sunlight create colour shifts ranging from light blue to turquoise green.
According to the manufacturer, the decline in efficiency resulting from the colour coating is only 8 to 15 %. In practice, however, tests on a refurbished coal silo in Basel indicated slightly greater losses (fig. 16). About a year’s worth of measurements revealed that the yield of the coloured roof modules was 12—19 % lower than that of standard black modules.8)
Printing processes and pigment application
Screen-printing is a long-established method for colouring glass. In PV modules, the print is usually applied to the inner surface of the front glass, facing the solar cells. Depending on the shade and intensity of the colour, the resulting losses in efficiency range between 15 and 50 %, so a white crystalline module will have an efficiency of up to 90 W/m2, while those with darker-hues reach 140 W/m2. Single-colour screen-printing is the least expensive of all PV module colouring methods. The firm Issol uses it to manufacture terra-cotta-coloured modules for roof-integration that it has developed in collaboration with CSEM in Neuchâtel (fig. 21). In principle there are no limits to the range of colours: in 2015, Issol equipped an office building in Drammen near Oslo with a 1,250 m2 photovoltaic facade made up of emerald-green printed modules. According to the manufacturer, the loss in efficiency in this case was only 17 %.
Stephen Wittkopf, a professor at the University of Lucerne, guarantees a maximum efficiency loss of 20 % — regardless of colour — using an optimised ceramic digital printing process that he developed together with his colleagues and the firm Glas Trösch (fig. 20). In this method, the colour is burned into the glass at approximately 600 °C during the normal glass-hardening process. According to Wittkopf, digital printing has cost advantages over screen-printing for multi-coloured applications. In addition, the colour thickness and therefore the module efficiency can be controlled more precisely. “In printing processes with multiple colours, it is critical to ensure that the shading created by the print colour remains constant throughout the entire module surface,” says Wittkopf. “Otherwise, ‘hot spots’ will develop at the contrast points, at which the module can become damaged from overheating. We achieve this uniform translucency by varying the print thickness and the dot matrix spacing as a function of colour.
In contrast, the Solarglaslabor consortium in Basel, which consists of several medium-sized businesses and the architecture firm raumweg, colour-coats modules on the outside.9) In this approach, the module surfaces are sand-blasted, and the resulting pits in the glass are filled with colour pigment using a proprietary process. Subsequently, the surface is sealed to protect against contamination (fig. 19).
This method is likewise usable for every module type. Reflections from the glass surface are largely eliminated by the application of the pigment. An initial series of measurements has shown that the module efficiency declines by 10 to 25 %, depending on the colour. A first reference project has recently been completed: in the spring of 2017, Solarglaslabor outfitted the facades of an apartment building in Basel with yellow modules.
Dye-sensitised solar cells
Dye-sensitised solar cells (also known as Grätzel cells) have a completely different structure than the cells used in the module types discussed thus far. In these cells, a thin layer of dye absorbs the incident light. Since both this layer of dye and the nanometre-thin electrodes are translucent, Grätzel cells can be used to construct homogeneously coloured, transparent solar facades. The colour palette includes red, orange and green, whereas blue and grey modules cannot be manufactured.
This technology saw its first large-scale use in 2014 on the western facade of the EPFL’s Swiss Tech Convention Centre in Lausanne by Richter Dahl Rocha (fig. 14). Its limitations soon became evident: the installed modules have efficiencies of less than 1 %, and the entire installation achieves a peak power of barely one kilowatt. In addition, there is as yet no long-term practical experience with dye-sensitised cells. Though variations in module structure have now made efficiencies of about 4 % achievable, the cell manufacturer Solaronix10), a spin-off of the EPFL, has since turned its focus to another technology: so-called perovskite cells, which have higher efficiencies of 10 —12 %. These are completely black, but can be inexpensively applied to glass via screen-printing. Since they can be used to create patterns and various gradations of grey, perovskite cells open up interesting possibilities in semi-transparent sun shading on glass facades.
Future outlook
The ground seems to have been prepared for photovoltaic roofs and facades that are far better adapted to their surroundings than previously. Nevertheless, the new technologies are likely to remain limited to a niche market for the time being, since they cost at least €40−50/m2 more than standard modules and yield 10 – 50 % less energy. Their use has only become a realistic option thanks to the efficiency increases and price reductions of solar technologies over the past decade. Nowadays it is easier to stomach the loss of a few percentage points of energy generation in favour of an appealing design. In addition, the new photovoltaic array technologies are opening up completely new areas of application. Patrick Heinstein of CSEM illustrates this with his use of terracotta-coloured modules. “They make it possible to install photovoltaics in situations that were strictly off-limits before due to historic preservation concerns. Viewed from this perspective, employing such solar arrays as opposed to none at all can only be viewed as a one-hundred-per-cent win.
