Bifacial PV Systems
Inside this Article
Conventional PV modules are monofacial, meaning that their electrical power output is a function of the direct and diffuse radiation captured on the front side of the module only. By contrast, bifacial modules convert light captured on both the front and back sides of the module into electrical power. Bifaciality improves PV system energy capture—dramatically in some cases—and rewrites conventional system design rules in interesting ways.
This article is an introduction to bifacial PV systems. After briefly reviewing the history of bifacial PV cells and providing a high-level overview of bifacial cell technologies, I summarize the potential benefits of bifacial PV modules and systems. I then focus on best practices and applications for designing and deploying systems that integrate bifacial PV modules. Finally, I consider some challenges to adoption and important efforts under way internationally to unlock the full commercial potential of bifacial PV systems.
Research on bifacial PV cells dates back to the dawn of the solar industry, according to Andrés Cuevas’ oft-cited article, “The Early History of Bifacial Solar Cells” (see Resources). Japanese researcher H. Mori proposed a bifacial PV cell design as early as 1960 and had successfully developed a working prototype by 1966. Russian and Spanish researchers proposed uses for bifacial PV cells around the same time. It was the Russians, however, who first deployed bifacial PV modules in the 1970s, as part of their space program. A major milestone occurred in 1980, when Cuevas and some of his colleagues in Spain documented the ability of light-colored surfaces to direct reflected light (albedo) to the back of a bifacial PV cell and increase its power output by 50%.
Due to the high cost of producing bifacial PV cells, the first terrestrial applications for this technology were relatively late to emerge. One of the best-documented early field applications is a north-south–oriented vertical photovoltaic noise barrier that Swiss researchers deployed in 1997 along the A1 motorway in Zurich using 10 kW of bifacial PV modules. The first signs of commercialization, at least in North America, appeared roughly a decade later when Sanyo introduced its first UL-listed HIT Double bifacial PV modules. Though Panasonic, which acquired Sanyo, subsequently discontinued the bifacial product line, as of January 2017, at least eight manufacturers offer bifacial PV modules certified for use in North America (see Table 1).
Cell technology. While bifacial PV cells currently make up an insignificant percentage of worldwide PV cell sales, the technology is in some ways a continuation or logical extension of standard monocrystalline silicon (mc-Si) cell technology. Depending on whether the semiconductor material contains a relative abundance or deficiency of electrons, the industry broadly categorizes mc-Si cells as either n-type or p-type devices, respectively. It is possible to fabricate bifacial cells out of both p-type and n-type wafers, given high-quality silicon material, although the process requires some additional manufacturing steps compared to producing conventional monofacial cells.
In practice, more than 90% of the PV cells sold worldwide are based on a p-type architecture, while the vast majority of the bifacial products in Table 1 are n-type devices. This underscores the fact that many n-type PV cells, which are primarily found in niche high-efficiency modules from companies such as LG, Panasonic and SunPower, are inherently bifacial. (Some people trace the history of bifacial PV cells all the way back to Bell Labs, since its first practical solar cell in 1954 was an n-type device.) P-type devices dominate the market because they are cost-effective to fabricate at scale. While n-type bifacial cells offer the highest efficiency, companies such as SolarWorld are betting that p-type bifacial cells can provide a good balance between performance and cost.
Regardless of the specific cell technology, the rear side of a bifacial PV cell needs to be able to act as a collector, which requires advanced architectures and manufacturing techniques. The authors of the informative Electric Power Research Institute (EPRI) Bifacial Solar Photovoltaic Modules (see Resources) explain: “Today’s crystalline silicon and thin-film monofacial PV cells commonly use a fully metallized backside. This feature involves a moderately thick metal contact for reduced series resistance and is relatively inexpensive to produce. By contrast, bifacial cells incorporate selective-area metallization schemes to allow light between the metallized areas.”
Though thin-film manufacturers are still working out the material science issues necessary for bifacial thin-film modules, many mc-Si manufacturers have successfully produced bifacial cells, which often incorporate thin-film layers, such as the rear passivation layers of amorphous silicon in Figure 1. The next challenge is adapting these technological advances for mass manufacturing. The EPRI report continues: “The lower amount of metal changes how cell performance is optimized, potentially requiring tighter (more expensive) specs on the silicon and thin-film material used and also increasing series resistance concerns. Furthermore, bifacial cells may employ different metals, such as copper and nickel, and/or deposition methods, such as plating or inkjet printing, which, in part, requires different equipment and entails a potentially more complex manufacturing process. Consequently, the backside metal represents a nontrivial impediment to manufacturing bifacial cells with high performance and low cost. This added complexity and cost needs to be offset by the performance gains from increased light collection.”
BIFACIAL VALUE PROPOSITION
The rapid growth of the solar industry in recent years has been largely premised on significant up-front cost reductions, especially lower costs for PV modules. Bifacial PV modules run counter to the grain in the market since they are inherently more expensive than conventional monofacial modules. Fabricating bifacial PV cells requires not only high-quality mc-Si wafers, but also anywhere from two to six additional manufacturing steps compared to conventional cells.
The crux of the bifacial value proposition, therefore, is improved production and performance over the life of the system, which is a function of both bifacial energy gains and improved durability. Because bifacial modules offer high conversion efficiencies, they also have the potential to lower BOS costs, which make up an increasing percentage of up-front system costs. The ultimate goal, of course, is a lower levelized cost of energy (LCOE).
Increased energy generation. Unlike PV systems deployed with monofacial modules, bifacial PV systems can convert light that shines off the back of the module into electricity. This additional back-side production increases energy generation over the life of the system. Ongoing research and side-by-side testing suggests that a bifacial PV system could generate 5%–30% more energy than an equivalent monofacial system, depending on how and where you install the modules. Moreover, the manufacturers’ linear performance warranties for bifacial PV modules are some of the best in the industry.
Improved durability. To allow light to shine on the back-side of a bifacial cell, module manufacturers need to use either a UV-resistant transparent backsheet material or an additional layer of solar glass. In most cases, as shown in Table 1, manufacturers have opted for a glass-on-glass package that generally improves field durability as compared to glass-on-film options. Not only is a glass-on-glass package more rigid—which reduces mechanical stress on cells during transportation, handling and installation, or from environmental conditions such as wind or snow—but it is also less permeable to water, which may reduce annual degradation rates. Moreover, many bifacial modules are frameless, and eliminating the aluminum frame effectively reduces opportunities for potential-induced degradation (PID).
Reduced BOS. As prices for modules and interactive inverters have fallen in recent years, BOS costs—specifically, the costs associated with mounting systems—have come to make up an increasing percentage of total PV system costs. An interesting side effect of this trend is that commercializing higher-module efficiencies is beginning to look like one of the best opportunities to squeeze additional value out of PV systems. Higher-efficiency modules not only reduce the area of the mounting system on a per kW basis, but also allow a developer to increase system capacity and energy harvest at a given site with fixed development costs.
Lower LCOE. The LCOE for a power generation asset is found by dividing the total life-cycle costs—both the up-front construction costs and the operational costs over time—by the total lifetime energy production. In the field, bifacial PV modules outperform their nominal power and efficiency ratings, which addresses the energy-generation side of the LCOE calculation. Factoring in the bifacial energy gain, a 19% efficient bifacial 300 W module might harvest energy in a field application equivalent to what a 21% efficient 335 W monofacial module produces. From the manufacturer’s perspective, meanwhile, it is could be more cost-effective to add bifaciality to a 20% efficient mc-Si cell than to mass-produce a monofacial one that is 22% efficient. This balance between performance and cost can make bifaciality an attractive feature for a module manufacturer’s technology roadmap.
OPTIMIZING BIFACIAL DESIGNS
Though bifacial PV modules can convert both front- and rear-side irradiance to electrical power, they nevertheless put their best face forward, in the sense that front-side efficiencies are invariably higher than back-side efficiencies, whether due to semiconductor properties or the amount of back contact metallization. The bifacial ratio quantifies the STC-rated power of a bifacial module’s back side in relation to the front-side power. For the products in Table 1, bifacial ratios range between 55% and 95%, which obviously suggests something about the relative energy production for different products in equivalent applications.
Regardless of its specific bifacial ratio value, the field performance of any bifacial PV system is highly dependent on back-side irradiance. Generally speaking, back-side irradiance is light reflected off an adjacent horizontal surface. Therefore, you can optimize bifacial PV systems by following a few simple guidelines: Install bifacial arrays above surfaces that reflect as much light as possible, increase array height or tilt angle to collect more reflected light and avoid shading the back side of the array.
Surface reflectivity. A bifacial PV system will generate more energy when installed over a light-colored rather than a dark-colored surface. This is because the former will reflect more light onto the back of the array, whereas the latter absorbs more of the incident irradiance. Albedo is a dimensionless quantity, usually expressed as a percentage, that describes this ratio between light reflected off a surface and the original incident irradiance. The higher the albedo value, the higher the surface reflectivity.
Table 2 provides representative albedo values for a variety of common ground surface types, as documented in the SolarWorld white paper “How to Maximize Energy Yield with Bifacial Technology” (see Resources). These values suggest that white roofing membranes, which reflect roughly 80% of the incident light when new and unweathered, are an ideal ground cover surface under a bifacial PV array. By contrast, the measured albedo value for raw concrete is only 16%. While the albedo for concrete increases dramatically when it is painted white, SolarWorld’s research indicates that not all light-colored surfaces are created equally. White gravel, for instance, has a relatively low albedo due to an “open-pored structure [that] causes a large amount of light to be lost within the voids.”
While the additional rear-side power output in a bifacial system is clearly proportional to ground surface albedo, the authors of the EPRI article note that this simple relationship “belies the fact that, in practice, energy gain depends on a number of complicated installation-specific factors.” For example, white surfaces reflect light of all colors, whereas other surfaces reflect light preferentially, absorbing some colors and reflecting others. Grass, for instance, absorbs blue and red light and mostly reflects green light. PV cells, meanwhile, vary in their ability to collect and convert different wavelengths of light into electrons.
Height and tilt angle. The closer you install a bifacial array to the ground or roof surface, the more self-shading occurs. Flush mounting, for example, effectively blocks any reflected light from reaching the back of the array. Increasing the height of the array or its tilt angle increases reflected light collection and enhances the bifacial contribution. Generally speaking, the higher you can install a bifacial PV array, the better its bifacial energy gain. However, this does not mean that bifacial modules are suited for carports and awnings only.
SolarWorld simulations suggest that a significant bifacial energy boost is possible with a relatively modest height increase. Not only is the energy boost curve in Figure 2 steepest between 0 and 0.2 meters (7.9 inches), but also the inflection point occurs somewhere around 0.5 meters (19.7 inches), after which point the curve begins to flatten out; the saturation point occurs around 1.0 meter (39.4 inches), meaning that additional energy gains are negligible above this height. These data suggest that bifacial modules are potentially well suited for just about any ground-mounted application, as the leading edge of these arrays is often 18 inches–36 inches above grade.
It is also possible to adapt conventional flat roof– mounting systems for use in bifacial applications. In its bifacial system design guide, for example, Prism Solar recommends a minimum height of just 6 inches above the reflective surface. To facilitate a slight increase in array height in low-slope–roof applications, the company has worked with mounting system manufacturers, most notably Opsun Systems, to develop structural solutions optimized for use with bifacial modules. In addition to a modified ground-mount system, the Bifacial SunGround, Opsun Systems also offers the SunRail Structure Bifacial, a higher-elevation version of its standard commercial rooftop mounting system.
Back-side shading. To optimize bifacial energy gains, system designers also need to avoid shading the back side of the array. Most racking systems have rails that run across the module’s backside, which an opaque white or black film usually covers. These structural components, especially support rails, are potential sources of shade in a bifacial system. As a result, mounting systems optimized for bifacial applications locate mounting rails at the perimeter of the modules, orienting these in parallel with rather than perpendicular to the module frame or the edge of the glass.
Back-side shading is also a concern for bifacial module manufacturers. The junction box on many monofacial modules, for example, is located directly behind one or more PV cells. By contrast, most bifacial modules have a low-profile junction box located at the perimeter of the module to minimize back-side cell shading. Though testing indicates that back-side shading from junction boxes or mounting structures will not damage a bifacial module, it does result in yield losses.
FIELD APPLICATION AND PERFORMANCE
Data from initial test beds and performance simulations—some of which are summarized later in this article—suggest many potential applications for bifacial PV systems. These include most conventional applications such as flat roofs and free fields, where installers deploy monofacial PV modules today, as well as niche applications such as building-integrated PV (BIPV) carports and awnings, where they typically deployed early bifacial modules. Back-side power collection also rewrites the rules that apply to traditional PV system design and performance, which could enable new markets and business models.
Sandia test results. Sandia National Laboratories recently published a report (see Resources) documenting the side-by-side test results for Prism Solar bifacial modules in comparison to reference monofacial modules. Sandia installed modules at the test bed in five orientations over two surfaces at its New Mexico Regional Test Center. Data collected over a 6-month period (between February 15 and August 15, 2016) indicated that the bifacial modules were outproducing the monofacial devices by anywhere from 18% to 136%, depending on the orientation and ground cover. Figure 3 provides the average daily power output curve for each test condition.
The report’s authors draw some interesting conclusions from these data. First, they note that bifacial gains vary throughout the day, depending on the angle of the sun or whether conditions are clear or cloudy. The impacts of sun angle are somewhat intuitive when you consider that the sun is closest to the horizon early in the morning and late in the afternoon, which not only decreases the available incident energy but also increases the amount of reflected light. As a result, the percentage of the instantaneous power output resulting from the bifacial contribution is highest at these times, and the bifacial gains are relatively lower at or around solar noon. The impacts of direct versus diffuse irradiance are similar. During cloudy conditions, the incident energy is relatively low, which increases the percentage of bifacial gain due to reflected light. Under sunny conditions, by comparison, the bifacial contribution is higher in absolute terms (back-side power) but lower in relative terms (percentage of bifacial gain).
The authors also note that bifacial modules are relatively insensitive to changes in array azimuth. As you rotate a bifacial array east or west of true south, the bifacial boost increases, effectively offsetting some of the losses that a monofacial array experiences in non-optimal orientations. As a result, “west-facing bifacial modules tilted at 15° produced a similar amount of energy as south-facing, 15°-tilted bifacial modules and surpassed the energy production of all of the monofacial orientations considered.” Not only did the west-facing, 15°-tilted bifacial array outperform the optimally oriented monofacial arrays, tilted at 15° and 30°, but also the west-facing, vertically oriented (90° tilt) bifacial array “outperformed monofacial modules at any orientation.”
Not surprisingly, the bifacial gains were also greatest in a west-facing, vertically oriented application, which creates an effective collection area for bifacial modules literally double that of monofacial modules. As a result, the bifacial power curve in this application has two peaks, one in the morning and one in the afternoon, whereas the equivalent monofacial power curve has one peak only. An east-west facing array is also effective at shifting solar power production later into the afternoon, when electric demand is often greatest. This configuration is likely well suited to take advantage of certain time-of-use rate structures and could provide additional value to utility operators. (The downside of an east-west vertical orientation is its high susceptibility to horizon shading losses.)
BARRIERS TO ADOPTION
On the one hand, bifacial PV arrays require specialized modules and mounting systems, as compared to conventional PV systems, which invariably increases up-front system costs. On the other, side-by-side field tests, such as those Sandia conducted, clearly reveal a bifacial energy boost. It is entirely possible, therefore, that bifacial PV systems could provide the best value, in terms of LCOE or return on investment, in certain applications. Making that case and taking it to investors, however, remains a barrier to widespread market adoption.
Macroeconomic conditions. In the short term, the low costs for conventional monofacial PV modules represent one of the biggest challenges to the commercialization of bifacial products. Module prices are at an all-time low, largely due to downward price pressure caused by global oversupply. As a result, many manufacturers are operating at low to negative operating margins, which hinders investment in new manufacturing tools and product lines.
The authors of the EPRI report note: “It is financially difficult to sustainably grow manufacturing capacity of existing products, let alone a more innovative concept such as bifacial PV modules. This issue is exacerbated by the more expensive manufacturing tooling and processes required to produce bifacial modules today. The high capital expense and low returns on cell and module production is a bottleneck for adoption by manufacturers.”
Module nameplate power rating. Today, STC ratings for bifacial modules are based on front-side performance only, which obviously fails to capture the effects of bifaciality. To reflect the fact that bifacial electrical properties vary in proportion to back-side irradiance, manufacturers will also provide some version of Table 3, detailing performance characteristics at different levels of bifacial gain. The manufacturers leave it to the designer to decide how to apply these data. Since back-side irradiance has no impact on open-circuit voltage and has a negligible impact on voltage at maximum power, the real design consideration is the potential for higher currents.
Industry stakeholders around the world are actively developing a consensus on standard testing procedures for rating bifacial PV modules that the International Electrotechnical Commission (IEC) will eventually publish as IEC 60904-1-2. Researchers at the National Renewable Energy Laboratory (NREL), for example, have proposed flash-testing both sides of bifacial PV modules and using these flash test data to derive a compensated short-circuit current value. Additional indoor and outdoor testing is under way at NREL and Sandia to determine the accuracy of this approach.
Production modeling. Perhaps more important, the industry needs bankable methodologies for modeling bifacial system energy production in the field, a requirement complicated by the fact that field conditions have an inordinate impact on bifacial system performance. Performance models need to account for rear-side shade effects associated with mounting structures and adjacent rows of modules, which will vary considerably both over the course of a day and from one application to the next. Ground-surface albedo is another consideration. This can change seasonally, when snow covers grass or dirt, or over time, due to soiling effects. The albedo for a white roof membrane, for example, might be 80% when the membrane is newly installed but only 50% after it has spent a few years in the field. Research also indicates that rear-side irradiance is also nonuniform, meaning that it varies across the back of the array.
Because of all these factors, field test results are essential for developing and verifying the accuracy of bifacial performance models. Unfortunately, many laboratory test beds consist of only a few rows of modules, which are often spaced out to minimize self-shading. These results tend to overestimate performance in larger systems, especially in applications where rows are more tightly packed together. This creates a chicken-and-egg scenario. To optimize design variables, such as ground-cover or dc-to-ac ratios, you need a sophisticated production-modeling tool. But to develop an accurate production-modeling tool, you need field data—and the more of it, the better.
David Brearley / SolarPro / Ashland, OR / solarprofessional.com
Cuevas, Andrés, “The Early History of Bifacial Solar Cells,” 20th European Photovoltaic Solar Energy Conference (EU PVSEC) Proceedings, 2005
Electric Power Research Institute (EPRI), Bifacial Solar Photovoltaic Modules, September 2016
Lave, Matthew, et al., “Performance Results for the Prism Solar Installation at the New Mexico Regional Test Center: Field Data from February 15 to August 15, 2016,” Sandia National Laboratories, SAND2016-9253
SolarWorld, “How to Maximize Energy Yield with Bifacial Technology,” white paper, 2016
LG / 855.854.7652 / lg.com/us/business/solar-panel
Lumos Solar / 877.301.3582 / lumossolar.com
Mission Solar Energy / 210.531.8600 / missionsolar.com
Opsun Systems / 581.981.9996 / opsun.com
Prism Solar Technologies / 845.883.4200 / prismsolar.com
Silfab Solar / 905.255.2501 / silfab.ca
SolarWorld USA / 503.844.3400 / solarworld-usa.com
Sunpreme / 866.245.1110 / sunpreme.com
Yingli Solar / 86.312.8929.800 / yinglisolar.com