Fab & Bifaciality: One small step for Facilities technology, one giant leap for kWh cost Materials

Cell reduction Processing Radovan Kopecek1, Yannick Veschetti2, Eric Gerritsen2, Andreas Schneider1, Corrado Comparotto1, Valentin D. Mihailetchi1, Jan Lossen1 & Joris Libal1 Thin Film 1ISC Konstanz, Konstanz, Germany; 2CEA-INES, Le Bourget du Lac, France PV Abstract Modules The aim of this paper is to dispel the common belief that bifaciality is nonsense as it is not a mature technology, it is expensive and, because in large systems there is limited albedo from the rear side, it only Power serves the niche market. A complete picture of bifacial cell technologies and module concepts is presented, Generation as well as levelized cost of electricity (LCOE) results for present and future bifacial systems. Market Watch Introduction for flat roofs and ground-mounted There are still several technological At a time when high-efficiency solar installations of large systems (e.g. in challenges which are not trivial: for cell technologies are in any case desert areas), bifaciality is extremely example, how to design and where moving towards bifaciality, and many helpful for driving down the levelized to place the junction boxes, how to module producers are changing to cost of electricity (LCOE) and also for build shadowless and stable mounting glass–glass, as well as the fact that increasing the duration of electricity systems and, most importantly, how to high-power modules are necessary generation. standardize bifacial measurements and in order to minimize the balance of how to simulate the bifacial benefit of system (BOS) costs, the authors are very individual installations. However, all confident that bifaciality will become “For flat roofs and ground- of these challenges have already been an extremely important technology in mounted installations of tackled individually for some time by driving down costs per kWh for PV a number of companies and institutes. systems. When rooftops are the main large systems bifaciality is This article summarizes some of these market in a particular location (e.g. the outcomes, which were presented Netherlands), bifaciality is, needless extremely helpful for driving and discussed at the Second BifiPV to say, not of great interest. However, down the LCOE.” Workshop in Chambéry, revealing

Figure 1. Brief history of bifacial cells, modules and systems.

Photovoltaics International 1 the big picture of the status of bifacial and ISC Konstanz were partners in the Electricity costs with bifaciality technology. At that workshop several successful EU-funded FoXy project, ’ share in EU renewable working groups were established, running from 2006 to 2008 within the electricity production could be assigned to work on standards for Sixth Framework Programme (FP6) [4]. substantially increased if its cost bifacial measurements, qualification The project dealt with low-cost structure were to further improve. standards for bifacial modules and the and cost-effective n-type cell and module The cost of solar electricity can be development of dedicated simulations processes for mass production. reduced by increasing efficiency, Cell of bifacial systems. Today, six years later, the n-type implementing low-cost manufacturing Processing processes are being used in several technologies, and improving reliability industrial production lines, under such and sustainability. A fourth powerful Bifacial history and status, designations as n-PASHA (ECN) and approach to reducing cost is to increase and estimations of LCOE BiSoN (ISC Konstanz) solar cells. ECN the energy yield, namely the kWh History and status and European OEMs have successfully produced per Wp module power. The Bifacial solar cells have a very long transferred their n-type technologies to important metric for analysing cost is history. The concept actually originated cell and module manufacturers outside the LCOE, which takes into account the at in 1954 with the very first Europe (Yingli, Mission ), costs and the energy produced over the processed, which was an while ISC Konstanz is currently system lifetime. n-type bifacial interdigitated back transferring the BiSoN process to In southern Germany the current contact (IBC) solar cell. The history MegaCell in Italy. Mission Solar value of LCOE for a state-of-the- is depicted in Fig. 1, as presented Energy and MegaCell are investing in art system is around €ct8/kWh. by Kopecek at the Second bifiPV the installation of additional solar cell Although most of the cell and module Workshop [1]. Russian and Spanish capacity because of bifaciality. manufacturing has moved to Asia, groups made proposals in the 1960s on how bifaciality could be used, and bifacial cells have been employed in space applications in Russian satellites since the 1970s. Around 2000 bifacial modules have also been used in terrestrial applications, for example by Nordmann [2] on Swiss highways. However, all the cell and module concepts employed at that time were extremely expensive, so these systems were strictly niche applications, in which costs were not a concern (e.g. space applications), or they were used for demonstrations. Since 2000, several groups have been picking up the idea of bifaciality again: with the development of, for example, cost-effective n-type solar cells, bifacial technology is being borne in mind in order to benefit from the active rear side. Figure 2. LCOEs of a PV system in southern Germany for different technologies One of the first cost-effective bifacial as a function of the scale of production (GW study of FHG IPA/ISE). A range of mc-Si solar cells is described in the Ph.D. €ct8.2–8.69/kWh is calculated for the production of standard c-Si technology thesis of Kränzl [3]. The institutes ECN on a 0.5–1GWp/a scale.

Today (2014) Solar cells in 2020+ Bifaciality in 2020+ Cz-Si cell Cz-Si cell Bifaciality factor: 0.8 Efficiency: >20% Efficiency: >24% Effective power: up to Power: 5Wp Power: 5.7+Wp 450Wp-e* Costs: €0.3/Wp Costs: €0.2/Wp 60-cell module 60-cell module Yearly system power gain: Cell-to-module loss: 3% Cell-to-module gain: 2% 15–30% (depending on installation) Power: >290Wp Power: >350Wp Costs: €0.55/Wp Costs: €0.4/Wp A conventional 0.7GW factory when upgraded to this technology will produce 1GW System in southern Germany System in southern Germany System in southern Germany Costs: €1.5/Wp Costs: €1.1/Wp (reduction in (running for 40 years/improved Electricity production costs module and system costs reliability with glass–glass): (running for 25 years): €ct8.5/kWh through decrease in area) down to €ct3–4/kWh Electricity production costs (running for 30 years): €ct6/kWh

* Wp-e is the real effective peak power that is generated by a module mounted in a system. For a module with a high BF, Wp-e can be more that 30% higher than the measured front-side power Wp. Table 1. Cost of ownership (CoO) for cells and modules and the resulting LCOE for current and future monofacial and bifacial technologies. (The bifaciality factor – BF – is the ratio between rear- and front-side efficiency; for identical efficiencies the BF is 1.)

2 www.pv-tech.org large-scale production is expected to efficiencies of over 19% on Cz an equivalent cell efficiency of 22%. become competitive in Europe. A recent substrates [5]. This progress is largely As in the case of PERC, the PERT study by Fraunhofer, with other similar due to the development of advanced cell has the advantage of being studies coming to the same conclusion, metallization pastes, which allow the more compatible with existing predicted that multi-GWp production formation of a lightly doped emitter production lines; indeed, only a in Europe based on relatively and reduced shadowing over the front few additional manufacturing tools conventional technology can be cost side of the cell. Nevertheless, it is are required to make the step from effective at this state-of-the-art LCOE expected that the performance will be standard technology. The PERT cell Cell (see Fig. 2). Moreover, if bifaciality limited in the near future because of architecture, depicted in Fig. 3, is made Processing is considered, solar cell production the rear surface (passivation and light up of two diffused layers, namely p+ on an even smaller scale can be cost confinement issues). Different lines of for the emitter and n+ for the BSF on effective. Table 1 shows the results attack are therefore essential in order to each surface. Anti-reflection dielectric of the calculation of LCOE for today’s produce cells of higher efficiencies. coatings (SiN) and symmetrical systems with standard technology and For many cell producers, the contacting grids are deposited on both future systems with higher efficiency passivated emitter rear cell (PERC) sides. and increased power output using concept represents a natural The PERT cell is commonly bifaciality. continuation of standard technology. developed on n-type Cz substrates The cell and module concepts of Such cells are already in production to avoid the LID effect (absence of many institutes and companies aim for on a large scale in Asia, with an the –oxygen complex). This higher cell efficiencies (>24%) and lower efficiency of over 20%, not taking into technology requires three main overall manufacturing cost (including account light-induced degradation additional steps: 1) boron diffusion for silicon material and crystallization, a (LID). Compared with the reference the emitter formation (p+); 2) emitter very important cost factor), leading technology, additional steps are passivation; and 3) the use of specific to a reduction in LCOE to €ct6/kWh necessary, such as rear-surface cleaning, screen-printing pastes for contacting (monofacial application), and higher dielectric layer deposition (Al2O3, SiO2) the boron emitter. As regards the yearly output (using a novel bifacial and laser opening [6]. boron diffusion, a higher thermal high-efficiency module architecture) Alternatives technologies to the budget in the range 900–1050°C is plus longer lifetime targeting down to PERC concept can also be considered, required, depending on the technology. €ct3–4/kWh for most optimal bifacial such as the passivated emitter rear So far, gas diffusion (BBr3, BCl3) is the operations. totally diffused (PERT), the a-Si:H/c- approach most commonly adopted, Si (HJT) and the IBC and is currently used in production (Fig. 3). All these three architectures probably because it is a more mature Bifacial solar cells have one thing in common: they process [8,9]. Ion implantation is an The last few years have seen a steady can be bifacial. When such cells can alternative, as excellent efficiencies improvement in cell and module benefit from an albedo, the gain in with simple process flows have been performance. Currently, the standard yield can be impressive, as stated by reported [10,11]. The use of solid technology – aluminium back- Kreinin [7], who reported that an sources is another option via the use surface field (BSF) on p-type silicon 18.5% cell mounted over a surface of PECVD doped oxides or spin- – represents a very large proportion with intermediate albedo can yield a on [12,13]: for example, the (>90%) of world production, with bifacial gain of 20%, corresponding to technology of PVG Solutions (30MW

Figure 3. Most prominent n-type technologies with bifacial characteristics.

Photovoltaics International 3 production line of bifacial cells) is technology. Since then, many players Alternatives such as copper plating and based on the use of spin-on dopants (academics, equipment suppliers or passivated contact concepts are now [14]. The emitter passivation can be cell producers), including ECN, have under investigation to determine their performed by silicon oxide growth reported efficiencies of between 20% feasibility in competing in the near (dry or wet), by Al2O3 layer deposition and 21.3% (see Fig. 5). future with alternative technologies (PECVD, ALD) or by alternatives such Experts in the field are confident like HJT and IBC. Nevertheless, since as the nitric acid oxidation of Si (NAOS) that – for the diffused technologies the Voc value remains limited, the PERT Cell concept (chemically grown oxide) – efficiencies over 21% should be cell is a lot less sensitive to the substrate Processing [15,16]. Dedicated metallic pastes (Ag/ achieved in 2015. The main factor quality. Al) are finally deposited by screen that limits efficiency is the high There are currently four producers of printing to contact the boron emitter. recombination activity at the metal/p+– bifacial PERT cells – Yingli (PANDA), As a consequence, quite different Si emitter interface, which corresponds LG (MonoX NeON), PVG Solutions process flows are possible for the to 40% of the total cell J0 [17]. Most (EarthON) and Neo – fabrication of PERT cells (Fig. 4). The players report a large gap between the and more producers of PERT cells are compromise between cell efficiency implied Voc value measured on the expected in 2015. and process simplification remains cell before metallization and the final A possible evolution of the PERT cell the principal guideline with regard to Voc value. Even if the Ag/Al pastes could be in the direction of IBC; this industrialization. were regularly improved in order to cell structure is well known to have a In 2009 ECN was one of the leaders print narrower fingers, efficiencies very high efficiency potential because in this technology, presenting an nudging 22% will be difficult to achieve of the absence of metallization on the efficiency of 18.5% with its PASHA if the issue above is not resolved. front side. ISC Konstanz has shown that such a device can also be bifacial [18] through its development of ZEBRA technology. Even if an IBC cell yields a better front efficiency than a PERT cell, some studies have reported that this is no longer the case when an albedo is considered. -Panasonic was the first company to market bifacial modules, which were based on its heterojunction with intrinsic thin layer (HIT) technology. This cell concept, grounded on a low-temperature process, is very well suited to the fabrication of high- efficiency bifacial cells. Indeed, the cell structure is made of very thin, hydrogenated, layers (5–15nm) deposited on both sides of the wafer [19]. This technology presents many advantages:

• The band-gap difference between a-Si:H and c-Si leads to an excellent surface passivation, resulting in very high Voc Figure 4. Examples of different process flows for the fabrication of PERT cells. • The complete process is performed at a low temperature: ~200°C

• The cells show an excellent 21.3 temperature coefficient: P ma x 20.7 (–0.3%/°C)

) 20.6 %

( 20.3 20.3 20.3 20.2 20.2 20.2 20.2 y

c 20 20 20 • This symmetrical structure is

n 19.9 19.8 e i compatible with thin substrates c

ffi 19.3 E • The fabrication process requires a limited number of fabrication steps

Finally, the efficiency potential in production is greater than 22%. On a lab scale, Sanyo-Panasonic has reported a record certified efficiency of 24.7% on 100cm2 [20]. * independently certified cells Unlike the PERT approach, this technology is not highly compatible Figure 5. Reported efficiencies for bifacial PERT cells. (The red bars correspond with existing production lines. As the to the performance of a PERT cell in mass production.) efficiency is driven by the very high Voc,

4 www.pv-tech.org the substrate quality is also very critical. water from penetrating the module increased for glass–glass modules, Some studies have reported that a interior over the large area of the solar compared with glass–foil modules, difference of 1% abs. in efficiency could module back side, which will in turn frameless applications become the be obtained over an entire ingot, leading inhibit any degrading effects over time, preferred mounting design. This to a widespread or specific selection of such as oxidation or delamination. The favours direct applications in building- the wafers. This is largely explained by a only region not protected in glass– integrated photovoltaics (BIPV) and consequent effect of interstitial oxygen glass modules is the edge area, which is reduces system costs. Frameless content and related defect distribution, typically sealed by double adhesive tape, designs may also minimize the risk of Cell as well as the effect of thermal donors. a silicone seal or specially designed edge potential-induced degradation (PID) in Processing Although Sanyo-Panasonic has for seal getters. The solar cells themselves systems with a high operating voltage, many years been the only producer of are protected by the large distance as the driving force for PID is the HJT cells, with a production capacity between module edge and cell, and potential between the grounded frame of 900MW, new players demonstrating any water penetration has to diffuse and the cells. impressive results are now emerging: from the edge before degradation can The key challenge for bifacial solar Table 2 summarizes the main players in take place. The advantages of glass– modules is the design and placement of the field [21,22]. glass modules are best exploited when the junction box. Since any placement using encapsulants that do not contain of junction boxes on light-sensitive or produce any chemical components areas on the module back side leads Bifacial module design that degrade cell metallization or to undesired shading, the junction considerations interconnections, such as peroxides, box either has to be reduced in size Apart from a higher energy output, or acetic acids in the case of EVA or must be placed in the edge region bifacial solar modules offer inherent encapsulants. of the module, if module size is to be advantages compared with standard, Another advantage of glass–glass kept constant. At the same time, these monofacial modules. Usually, bifacial modules is their greater flexibility, smaller junction boxes have to handle modules are available as a glass notably when using thin 2mm glass, higher currents because of the extra substrate or, in certain cases, with a as well as their mechanical robustness current generated by the module transparent backsheet foil substrate. as a result of the solar cells being back side. The latter problem can Backsheet foils typically have a certain positioned in the neutral mechanical be solved by cutting the cells in half, water permeability, allowing water to plane of the material sandwich, hence thereby reducing the cell current and, penetrate the backsheet and enter the securing the cells against mechanical at the same time, the cell-to-module interior of the solar module. Glass, tensile or compressive stress. Since losses. Alternatively, these cell-to- on the other hand, will totally prevent the mechanical stability is significantly module losses could be reduced by

Company Cell area [cm2] Record efficiency [%] Status Sanyo-Panasonic 148 21–22 900MW production Chochu 243 22.3 Production line in Q1 2015 Kaneka 171 24.2 Production line in 2015 AUO 239 23.1 Pilot line (BenQ) Silevo 239 23.1 100MW in production Expansion plan 1GW 2016 R&R 239 22.1 Laboratory CEA-INES 239 22.0 Pilot line (35MW)

Table 2. Record cell efficiencies achieved with heterojunction technology.

2 Figure 6. Gain in power Pm with a front-side irradiation of 1000W/m and increasing back-side irradiation from 0 to 1000W/m2, for different interconnection options (as measured in a solar simulator with symmetrical mirrors for simultaneous irradiation of front and back sides and a mesh filter to vary the back-side irradiation). SWCT and three bus half-cells were found to perform best [24].

Photovoltaics International 5 an additional busbar at the cell level. Bifacial cells offer the complexity of the cell interconnection A combination of both options leads “ process. A simplification can be achieved to almost zero electrical losses from potential to realize a by reversing the neighbouring cell so cell to module. This development that the interconnection ribbon does has currently been implemented in significant reduction in not have to make a cross-over from the standard glass-backsheet modules the complexity of the cell front to the back side, as illustrated in in order to further reduce costs by Fig. 7 [25]. This will allow an increase Cell increasing module power, and is an interconnection process. in productivity of the tabbing/stringing Processing attractive option for bifacial modules ” process, a reduction in cell spacing as well. Another interconnection For bifacial solar cells the rear side and, at the same time, an increase in option found to be very beneficial for consists of a similar finger/busbar grid module reliability in withstanding bifacial cells and modules is SmartWire to that of the front side (unlike standard, thermomechanical stresses caused by Connection Technology (SWCT) [23], monofacial cells, in which the cell rear temperature cycling (typically tested which offers the best performance side is fully metallized). This makes the from –40°C to +85°C). This concept of in a comparison of different cells transparent to IR radiation and may ‘planar’ interconnection requires cells interconnection options shown in Fig. lead to lower operating temperatures with a high bifaciality factor (> 98%). 6 [24]. in the field. It also affects temperature The assembly of bifacial modules distribution during soldering and this by cutting the cells in half can also requires modification of the soldering Bifacial systems and optimize the energy contribution time and temperature to obtain an applications of the module back side by making optimal compromise between defect Bifacial modules can be implemented it less sensitive to non-uniform generation (cracks) and adhesion in PV systems in various ways, resulting irradiation of the back-side surface strength of the copper ribbons. in different bifacial gains as a result in ground-mounted systems at low Bifacial cells offer the potential to of variations in the albedo of the elevations. realize a significant reduction in the surroundings and the bifaciality factors of the modules: examples for ground- mounted and flat-rooftop systems are shown in Fig. 8. Systems with a module inclination (slanted) that is optimum in the case of monofacial modules result in the highest total energy production and, accordingly, in the lowest LCOE. Depending on the geographical location of the installation site and on the albedo Figure 7. Top: standard cell interconnection, where the interconnection ribbons of the underlying surface, horizontally (black) connect the cell front side to the neighbouring back side. Bottom: planar mounted systems can also yield a high interconnection process with bifacial cells. energy production.

Figure 8. Different implementations of bifacial modules (landscape pictures taken from PVG Solutions [26]).

6 www.pv-tech.org Vertical mounting has some interesting features. First, soiling is considerably reduced compared with that of modules mounted with a standard inclination, and is of particular interest for desert regions. Second, vertically mounted bifacial modules can be installed, for example, along highways or used for Cell other applications where long single rows Processing of modules are suitable (e.g. anti-noise barriers). Third, vertical modules facing east–west deliver the same energy yield (kWh/kWp front) as south-facing monofacial modules mounted with a standard tilt. In addition, the east–west configuration of vertical bifacial modules has another interesting advantage: during Source: et al. [27]. Joge the day two peaks of energy production are delivered – one in the morning and Figure 9. Electricity generation profile for vertically mounted (90 degrees) one in the evening – with a lower energy bifacial and monofacial modules in east–west and south exposures, compared production at noon. As shown in Fig. with the generation profile of a standard configuration (monofacial, south- 9, a combination of this configuration facing, 30-degree inclination). with standard, south-facing modules contributes to a more homogeneous electricity generation profile during the day (‘peak-shaving’) and is extremely beneficial in terms of integrating more PV power into the electricity grid. As shown in Fig. 10, when bifacial modules are installed over surfaces with good or high reflectivity, bifacial gains (percentage increase in kWh/kWhpfront of bifacial modules compared with kWh/kWhp of monofacial modules) of 15–26% are possible. This has a significant impact on the dimensions of the PV system: assuming a bifacial gain of 20%, and taking a traditional 1MW ground-mounted system composed of 250Wp multicrystalline modules as a reference, bifacial modules with a P (front) of 290Wp under front- mpp Figure 10. Bifacial gain for a PV system mounted on a flat rooftop, considering side illumination enable the number albedos of 0.5 and 0.9 [28]. modules to be reduced by around 30% while maintaining the same yearly electricity production (Fig. 11). Apart from reducing the amount of land required to install a PV plant with a given electricity production capacity, this also results in cost savings in all other area-related BOS costs: mounting structures, cables, and the preparation and maintenance of the installation site. As a result, bifacial PV technology allows a significant reduction in LCOE, compared with monofacial high- efficiency technologies, such as PERC. Figure 11. The use of bifacial modules requires around 30% fewer modules (for the same total kWh/year) and reduces the area-related BOS costs of the PV Standardization of system. measurements, module qualification and system 3, the efficiency can vary depending reflection) are the most suitable. simulations on the measurement method used: The bifacial performance of Standardization of measurements: there is no real ‘true value’ – this a bifacial cell can be estimated cell and module will depend mainly on the module by determining the ratio of the In most laboratories, PERT cells are technology. In the case of bifacial cells measurements of the rear-side measured on a gold-plated chuck, intended for a bifacial module (glass– efficiency and the front-side efficiency which tends to represent the optimal glass), measurements taken in bifacial in standard test conditions (STC): cell performance. As shown in Table mode (probes at the back with no light this gives the ‘bifaciality factor’

Photovoltaics International 7 2 Ref Measure Voc [mV] Jsc [mA/cm ] FF [%] Eta [%] CEA* Conductive chuck 639.5 39.1 79.5 19.9 Bifacial mode 639.1 38.6 79.0 19.5 ISFH* Conductive chuck 658.0 38.6 80.0 20.3 Bifacial mode 657.0 38.2 79.5 20.0

Cell * Measurements at both institutes produced the same results. Processing Table 3. Certified efficiency of bifacial cells measured in two different configurations.

(BF), i.e. BF = ηrear/ηfront. Depending on the technology and the wafer quality, this BF factor can vary from 87% to 95%, according to the range of values reported in the literature. There is current debate in the community whether in fact this BF factor represents a relevant characterization method, since under working conditions, light arrives at both sides of the cells simultaneously, with a light power density of over 1000W/m2. New I–V measurement systems are now being adapted for making bifacial measurements, where the solar cell is illuminated on both sides, with various power densities Figure 12. ISC Konstanz’s measurement site in El Gouna, Egypt. ranging from 0 to 1000W/m2.

surrounding objects is completely how the severity of qualification tests “A major hindrance to masked. should be increased in order to take into the proliferation of bifacial The focus on STC conditions creates account the rear-side current. Ideally the absurd situation that, for a module these tests should reflect worst-case products currently seems to producer selling its products on a Wp operating conditions, which can differ be a lack of standardization. basis, it is advantageous to encapsulate significantly from STC for installations ” bifacial solar cells with a white with high albedo. TüV Rheinland backsheet in order to increase the STC therefore proposed to modify the A major hindrance to the value because of internal reflections current-driven tests to a current that proliferation of bifacial products behind the cell and in the spacing is equivalent to 400W/m2 of additional currently seems to be a lack of between the cells. However, the user rear-side irradiation [29]. standardization. The pricing of a solar of the device would, in almost every The same applies in the case of product is typically done on the basis application, harvest more kWh/year the electrical designer of a bifacial of its peak power output under STC, with the same module if a transparent installation, who must also consider rated in Wp. Other cell and module backsheet were chosen. the increased currents when defining parameters – such as the temperature Some companies compensate for the wire dimensions, the appropriate coefficients or the weak light the above-mentioned deficiency by inverter and protection devices. The performance – play only a minor role quoting the I–V parameters both under designer’s job is at least facilitated by when it comes to pricing. STC conditions and under a range of the fact that the maximum albedo of Bifaciality, until now, has not played conditions with varying additional a specific installation can be easily any role at all in standardization, rear illumination. Others quote measured or estimated from tabulated because it is not considered by state- I–V parameters under illumination values. of-the-art solar simulators with single conditions defined by the resulting Isc An accurate calculation or simulation flashbulbs, when measuring the peak increase as a percentage of the Isc under both of the annual energy yield power under STC conditions. The light STC. Some of these datasheet values gain of a bifacial installation and of source is typically placed far enough seem to be measured, whereas others the performance of the system for away from the measurement subject appear to be extrapolated. Details of the each position of the sun depends on to achieve a sufficiently homogeneous data and their determination are usually the spatial distribution of the rear illumination in the measurement plane. not quoted. irradiance, which is significantly The housing at the sides and behind In summary, it is difficult for a affected by the geometric conditions of the subject is completely black so as consumer to compare these products the specific installation [30]. not to compromise the illumination and even more difficult to contrast Many bifacial test installations homogeneity. For bifacial devices, this them with monofacial alternatives. achieve high bifacial gain in conditions results in very artificial conditions – a characterized by a low zenith angle of solar device in the middle of a black Qualification: module the sun, partial overclouding and high cavity, which is highly unlikely to be Since, in an installation with bifacial indirect irradiance. This explains why the case in any application. Indirect illumination, the maximum operating the annual yield gain is often higher irradiance caused by scattering in current of the module is increased, than the maximum power gain on a the atmosphere and reflections of qualification entities ask themselves sunny day.

8 www.pv-tech.org In order to achieve a standardization of certification measurements, data sheet declarations and test conditions, and to connect these to maximum current and yield simulations, it was decided by a group of institutes, test laboratories, equipment manufacturers and producers of bifacial solar products Cell during the latest bifacial workshop to Processing establish four working groups addressing all standardization topics [1]. The first meeting took place at the EU PVSEC, in which the participants agreed on a roadmap for the next few steps.

Simulations: systems It has also already been demonstrated at large-scale solar power plants that the energy yield of a solar system can be significantly enhanced by the use of bifacial modules [14]. In the desert at El Gouna (Egypt), ISC Figure 13. Percentage gain in kWh/kWp energy yield in 2014 of the BiSoN n Konstanz compared a bifacial module bifacial module over the Solar monofacial module. with a monofacial module [31], both containing n-type screen-printed solar cells of similar technologies [32,33]. For both modules, the tilt angle was 20 degrees, the lower edge was 1m high and the front sides were facing south (Fig. 12). Fig. 13 shows the percentage gain in energy yield in terms of kWh/kWp in the first eight months of 2014 of the BiSoN bifacial module over the nSolar monofacial module. The overall average gain was as high as 22.3%, and in August an average monthly gain of 25.6% was recorded. A second bifacial module, namely the nSolar bifacial module, with a BF of only 55%, was also installed at El Gouna. Fig. 14 shows its power output on May 15th, 2014, along with the irradiance throughout the day; a peak power output as high as 426W was recorded. Figure 14. Irradiance and power output of the nSolar bifacial module at El bSolar has developed a simulation th tool for its bifacial module technology, Gouna on May 15 , 2014. whereby the electrical gain is calculated as a function of installation height, packing density and albedo, as shown in Fig. 15. For example, a very densely packed system, with an installation height of 1m and an albedo of 50%, can yield a yearly electrical gain of more than 20%.

Commercialization and outlook As already mentioned, glass–glass modules are rapidly entering the market, as they offer several advantages over standard monofacial modules with white backsheets. Module manufacturers using this technology (SiModule, Apollon Solar, etc.) are therefore screening the market for bifacial solar cells which can be manufactured the same way as Figure 15. Simulations by bSolar of yearly electrical gain of south-facing bifacial standard cells. Currently, there are only PV installations as a function of installation height, packing density and albedo.

Photovoltaics International 9 (a) (b)

Cell Processing

Figure 16. (a) Aerial view of the world’s largest bifacial PV plant; (b) monthly energy contribution from the rear.

a few possibilities, such as those offered and operation of two bifacial efficiency on large area n-type PERT by PVG Solutions and NeoSolarPower photovoltaic systems on Swiss cells by ion implantation”, Energy (NSP); the newcomers MegaCell, roads and railways”, 1st BifiPV Procedia, Vol. 55, pp. 437–443. Motech and Mission Solar Energy Workshop, Konstanz, Germany. [12] Blevin, T. et al. 2014, will offer bifacial solar cells in Q1/Q2 [3] Kränzl, A. 2006, “Industrial solar “Development of industrial 2015. Panasonic, Silevo and Sunpreme cells on thin multicrystalline processes for the fabrication of currently market bifacial modules, and silicon: Diffusion methods, new high efficiency n-type PERT cells”, First Solar will most likely follow next materials and bifacial structures”, Solar Energy Mater. & Solar Cells, year as well. Ph.D. dissertation (in German), Vol. 131, pp. 24–29. Manufacturing equipment suppliers University of Konstanz, Germany [13] Rothhardt, P. et al. 2014, and technology transfer companies [http://kops.uni-konstanz.de/ “Characterization of POCl3-based who offer bifacial cell and module handle/123456789/4935]. codiffusion processes for bifacial technologies are the n-PASHA Alliance [4] FoXy Project [details available n-type solar cells”, IEEE J. Photovolt., (Tempress, RENA, ECN), BiSoN online at www.sintef.no/ Vol. 4, No. 3, pp. 827–833. Alliance (centrotherm, ISC Konstanz), Projectweb/FoXy/]. [14] Goda, S. 2014, “Cell mass French companies supported by INES [5] SEMI PV Group Europe 2014, production and array field (ECM Greentech, SEMCO Engineering), “International technology demonstration of n-type bifacial Schmidt and Meyer-Burger. roadmap for photovoltaic (ITRVP): ‘EarthON’”, nPV Workshop, Results 2013”, 5th edn (March) ‘s-Hertogenbosch, The [http://www.itrpv.net/Reports/ Netherlands. “Large bifacial power plants Downloads/]. [15] Benick, J. et al. 2008, “Surface will be an important part [6] Metz, A. et al. 2013, “Industrial passivation of boron diffused high performance crystalline emitters for high efficiency solar of worldwide PV electricity silicon solar cells and modules cells”, Proc. 33rd IEEE PVSC, San generation in the future. based on rear surface passivation Diego, California, USA. ” technology”, Proc. 3rd SiliconPV [16] Mihailetchi, V.D., Komatsu, Y. & Conf., Hamelin, Germany. Geerligs, L.J. 2008, “Nitric acid To summarize, it has been [7] Kreinin, L., Bordin, N. & pretreatment for the passivation shown that it is now time to take Eisenberg, N. 2012, “Industrial of boron emitters for n-type base the step towards bifaciality and production of bifacial cells: Design silicon solar cells, Appl. Phys. Lett., that standardization and system principles and achievements”, Vol. 92, p. 063510. simulations are necessary in order 1st BifiPV Workshop, Konstanz, [17] Edler, A. et al. 2014, to support a sustainable market Germany. “Metallization-induced penetration. The authors are confident [8] Rominj, I. 2012, “20% efficient recombination losses of bifacial that large bifacial power plants will be screen-printed n-type solar cells, silicon solar cells”, Prog. Photovolt.: an important part of worldwide PV n-PV Workshop, Amsterdam, The Res. Appl. (DOI: 10.1002/pip.2479) electricity generation in the future Netherlands. [forthcoming]. – plants similar to the largest one at [9] Veschetti, Y. et al. 2011, “High [18] Mihailetchi, V. 2014, “Zebra, present from PVG Solutions in Japan, efficiency on boron emitter n-type bifacial IBC technology”, 2nd BifiPV shown in Fig. 16. Cz silicon solar cells with industrial Workshop, Chambéry, France. process”, IEEE J. Photovolt., Vol. 1, [19] Kinoshita, T. et al. 2011, “The References No. 2), pp. 118–122. approaches for high efficiency HIT [1] Kopecek, R. 2014, “Bifacial world [10] Tao, Y. & Rohatgi, A. 2014, “High- solar cells with very thin silicon – monofacial modules: Why do we efficiency large area ion-implanted wafer over 23%”, Proc. 26th EU compromise the system power?”, n-type front junction Si solar cells PVSEC, Hamburg, Germany, pp. 2nd BifiPV Workshop, Chambéry, with screen-printed contacts and 871–874. France [www.bifiPV-workshop. SiO2 passivated boron emitters”, [20] Yano, A. et al. 2013, “24.7% record com]. Proc. 40th IEEE PVSC, Denver, efficiency HIT® solar cell on [2] Nordmann, T. 2012, “15 years Colorado, USA. thin silicon wafer”, Proc. 28th EU of experience in construction [11] Lanterne, A. et al. 2014, “20.5% PVSEC, Paris, France.

10 www.pv-tech.org [21] Hernández, J.L. et al. 2012, “High Kopecek has also been teaching the D r . Va l e n t i n D . efficiency copper electroplated basics of PV at the DHBW in Mihailetchi joined ISC heterojunction solar cells”, Proc. Friedrichshafen since 2012. He received Konstanz in 2008 and is 27th EU PVSEC, Frankfurt, his master’s degree from Portland State c u r r e n t l y a s e n i o r Germany, pp. 655–656. University (Oregon, USA) in 1995, and scientist, leading the [22] Chen, F.-S. et al. 2014 then obtained his diploma in physics n-type solar cells group “Optimisation of heterojunction from the University of Stuttgart in 1998. in the Advanced Cell Concepts solar cells on 6 inch wafers He completed his Ph.D. in 2002 in department. He studied at the Cell with high efficiency”, Proc. 29th Konstanz, with a dissertation topic of University of Groningen in the Processing EU PVSEC, Amsterdam, The thin-film silicon solar cells. Netherlands; in 2005 he received his Netherlands. Ph.D. in physics, with the device physics [23] Söderström, T. et al. 2013, “Smart Dr. Yannick Veschetti of organic solar cells as his thesis topic. Wire Connection Technology”, joined CEA-INES in 2005 After that, Dr. Mihailetchi worked as a Proc. 28th EU PVSEC, Paris, France. a n d i s c u r r e n t l y research scientist on [24] Joanny, M. et al. 2014, “Cell re sp onsible for the at ECN Solar Energy in the interconnection challenges for homojunction silicon Netherlands, where he developed bifacial modules”, 2nd BifiPV solar cells laboratory. He n-type-based solar cell processes for Workshop, Chambéry, France. studied at Strasburg University and industrial applications. [25] Kopecek, R. et al. 2006, “Module received his Ph.D. in physics in 2005. Dr. interconnection with alternate p- Veschetti specializes in the field of Jan Lossen studie d and n-type Si solar cells”, Proc. 21st crystalline silicon PV, with his main R&D physics at the University EU PVSEC, Dresden, Germany. work focusing on the development of of Freiburg and the [26] PVG Solutions, EarthON™ [details solar cells technology on n-type silicon. University of Cologne, available online at http://www. graduating in 2003 with pvgs.jp/en/earthon.html]. Dr. Eric Gerritsen has a d i p l o m a t h e s i s [27] Joge, T. et al. 2004, “Basic application been a project leader in concerning microcrystalline silicon technologies of bifacial photovoltaic PV modules at CEA-INES layers. He worked on the production solar modules”, Electrical Eng. in since 2008, working on and development of PV products for Japan, Vol. 149, No. 3. module reliability and more than 10 years at ErSol Solar [28] Eisenberg, N. 2014, “High efficiency performance. Before Energy AG and later Bosch Solar industrial PERT p-type bifacial joining INES he spent 23 years in various Energy AG. Since June 2014 Jan has cell and field results”, 2nd BifiPV positions with Philips Research been a project manager in the Workshop, Chambéry, France. Laboratories, Philips Lighting and Industrial Solar Cells department at [29] Hermann, W. et al. 2014, “IEC Philips /NXP in the ISC Konstanz, where he is responsible qualification testing of bifacial PV Netherlands, Germany and France. He for transferring BiSoN technology modules – Test conditions and received a Ph.D. in the field of ion from laboratory to industrial test requirements”, 2nd BifiPV implantation from the University of production. Workshop, Chambéry, France. Groningen in the Netherlands. [30] Yusufoglu, U. et al. 2014, “Analysis Dr. Joris Libal works at of the annual performance of Dr. Andreas Schneider I S C Ko n s t a n z a s a bifacial modules and optimization received his diploma in r e s e a r c h e n g i n e e r, methods”, IEEE J. Photovolt. [in p h y s i c s f r o m t h e focusing on business press]. University of Freiburg in d e v e l o p m e n t a n d [31] Comparotto, C. et al. 2014, 1999 and his Ph.D., with a technology transfer in the “Bifacial n-type solar modules: thesis topic concerning areas of high-efficiency n-type solar Indoor and outdoor evaluation”, crystalline silicon solar cells, from the cells and innovative module technology. Proc. 29th EU PVSEC, Amsterdam, University of Konstanz in 2004. He then He received a diploma in physics from The Netherlands. worked at the University of Konstanz, the University of Tübingen and a Ph.D. [32] Libal, J., Mihailetchi, V.D. & where he was responsible for the in the field of n-type crystalline silicon Kopecek, R. 2014, “Low-cost, high- development of crystalline silicon solar solar cells from the University of efficiency solar cells for the future: cells. From 2005 to 2011 he was employed Konstanz. Dr. Libal has been involved in ISC Konstanz’s technology zoo”, PV at Day4Energy, first as the head of R&D R&D along the entire value chain of International, 23rd edn, pp. 35–45. and then as the director of the company’s crystalline silicon PV for more than 10 [33] Kania, D. et al. 2013, “Pilot line quality management department. At the years, having held various positions at production of industrial high- beginning of 2011 Dr. Schneider worked the University of Konstanz, at the efficient bifacial n-type silicon for a short while at Jabil, before joining University of Milano-Bicocca and, more solar cells with efficiencies ISC Konstanz as the head of the module recently, as the R&D manager at the exceeding 20.6%”, Proc. 28th EU development department. Italian PV module manufacturer Silfab PVSEC, Paris, France. SpA. Corrado Comparotto About the Authors obtained his M.Sc. in Enquiries Dr. Radovan Kopecek, electronic engineering ISC Konstanz one of the founders of ISC i n 2 0 0 8 f r o m t h e Rudolf-Diesel-Straße 15 Konstanz , has been University of Brescia in 78467 Konstanz working at the institute as Italy. Since March 2011 Germany a full-time manager and he has been working on bifacial researcher since January n-type solar cells in the Advanced Tel: +49-7531-36 18 3-22 2007, and is currently the head of the Solar Cell Concepts department at Email: [email protected] Advanced Solar Cells Department. Dr. ISC Konstanz. Website: www.isc-konstanz.de

Photovoltaics International 11