Bifacial solar modulesâ€”those that can absorb light from both the front and back of the panelâ€”offer numerous benefits for photovoltaic farms, the most noteworthy of which is more energy production from a given space. Although these modules have a slightly higher price tag than their â€śone-sidedâ€ť counterparts, that extra cost is negated in the U.S., where bifacial technology is exempt from recent trade import tariffs. This makes the levelized cost of energy (LCOE) of a bifacial PV array lower than the LCOE of an equivalent-sized monofacial solar farm. As they become more mainstream, I expect the LCOE of bifacial modules will continue to drop, regardless of tariffs.
Tariffs may increase the cost of solar, but the bifacial exemption negates the added cost. (Image courtesy of NREL.)
Before utilities invest in this technology, theyâ€™ll need to accurately predict its production capabilities. Tools like PVWatts and PVSyst give designers a reasonable estimate of the energy production at a specific location, based on climate data and other factors, but bifacial technology introduces a slew of new variables. With a handful of bifacial solar farms currently in the field, researchers now have real-world data to help them develop better models. Studies from the National Renewable Energy Laboratory (NREL) and CFV Labs (an independent PV equipment testing facility) suggest that previous estimatesâ€”those that examined single panels or very small solar arraysâ€”were over optimistic in their estimates of bifacial gain. Nonetheless, solar farms made with bifacial modules still outperformed monofacial arrays, so itâ€™s worth quantifying their performance. Since the benefits of bifacial solar are seen primarily at the utility scale, both groups of researchers examined large arrays with single-axis tracking systems.
Bifacial gain is defined as the amount of additional energy that a bifacial array generates compared to a monofacial array of the same size. For example, if a monofacial array generates 10kWh in a certain period of time, and an equivalent bifacial array generates 11kWh under the same conditions, the bifacial gain would be 0.1 or 10 percent. As a formula, itâ€™s:
Bifacial Gain = [(Bifacial yield – Monofacial yield) / Monofacial yield] x 100%
Two components of bifacial gain: beam reflection and diffuse light. (Image courtesy of NREL.)
Since much of the bifacial gain comes from reflected light, itâ€™s no surprise that the albedo, or surface reflectivity of the material under the array, is a significant factor. NREL, currently in the first year of a three-year study, is using natural grass as its testbed. The grass is less reflective when green, but its albedo increases as it turns brown in the autumn. Winter conditions are even better, with snow providing very high reflectivity, resulting in the best bifacial gains. Next year, NREL plans to test crushed rock and may use a reflective fabric in future studies.
How albedo impacts performance. (Image courtesy of NREL.)
Albedos of different surfaces. (Image courtesy of NREL.)
Modeling and simulation can help PV system designers accurately predict production levels under various conditions, which minimizes the risk to investors. Computer models also give engineers insight into optimal configurations, including rack height, row spacing, and balance of system components. NRELâ€™s high performance Eagle computer can simulate an entire yearâ€™s worth of productionâ€”something that would take a typical laptop nearly a week to performâ€”in less than a minute. Its open-source simulations â€śStart at the panel and fly photons back to where they intercept the skydomeâ€”itâ€™s backward ray-tracing and evaluates irradiance at the modules,â€ť said Chris Deline, an engineer and researcher at NREL. â€śAll sorts of geometries are possible.â€ť They can also simulate topography and structures, as well as scenarios such as vehicles parked beneath solar modules that are used as carport roofs.
When bifacial solar was introduced, preliminary studies suggested performance improvements of 20 percent or higher. It turns out that those estimates were a bit optimistic. Many were conducted with very small arrays or even individual panels. If you consider a row of solar panels, the modules on each end of the row will receive more reflected light than the modules in the center, because of the unshaded area adjacent to the row. As a result, the outer modules can experience bifacial gains up to 50 percent higher than those on the interior. Both NREL and CFV Labs used longer rowsâ€”20 and 30 modules, respectivelyâ€”to account for this effect. In fact, CFV went as far as to disregard the data from the end modules and to include only results from the interior panels.
NRELâ€™s test setup. (Image courtesy of NREL.)
CFV researchers wondered whether the cell-to-cell irradiance mismatch would cause hot spots in certain cells or trigger bypass diode activation, either of which would decrease production. IR camera measurements showed no hot spots, even under worst-case conditions. Bypass diodes are built into solar modules to short-circuit underperforming cells so that a partially shaded or defective module doesnâ€™t drag down the production of an entire string. Testing showed no bypass diodes being activated under mismatch conditions.
NREL evaluated several kinds of PV technology under different albedo conditions and found that bifacial versions of heterojunction technology, a hybrid of crystalline and amorphous silicon, fared a little better than their PERC counterparts.
PERC vs Si-HJT performance. (Image courtesy of NREL.)
While using somewhat different techniques and designs, both studies drew similar conclusions. First, a utility-scale PV farm with single-axis tracking should expect 5 to 10 percent more energy from a bifacial array than from a monofacial array.
Average bifacial gain with all bifacial technologies. (Image courtesy of NREL.)
Second, albedo changes with weather, seasons and other factors. To accurately model this, designers should divide up the year into climate-dependent segments, using the appropriate albedo for each time slice, and then combine the results to estimate annual production. NASA has year-round satellite information available for that purpose.
Third, tools like System Advisor Model (SAM), PositsÂ and Bifacial Radiance can help engineers design bifacial systems and accurately predict their performance. These models will continue to be tweaked as more field data becomes available.
Finally, the Levelized Cost of Electricity (LCOE) of bifacial systems is on par with monofacial systems, especially in the U.S. where the bifacial import tariff exemption makes up for the slightly higher cost of the modules.
With single-digit improvements in production, bifacial solar isnâ€™t exactly a game-changing technology, but with businesses trying to maximize their return on investment, it certainly sweetens the pot a bit. As designers gain experience and researchers continue to shed light on the situation, we may see some creative ways to squeeze even more production out of these two-faced photon harvesters.