Since 2018 Chinese solar company JinkoSolar has been running a comparative test of its Swan bifacial module, with transparent backsheet, and monofacial PV technology on a variety of mounting types and ground types. Test sites included white-painted ground surfaces, sand, gravel and cement at locations in China and the US. Some arrays were fixed and others incorporated tracking. Jinko cooperated with solar research company PVEL, the Chinese Academy of Sciences, tracking manufacturer Nextracker and other third parties to study the power generation performance of bifacial modules and factors that affected gains in generation for bifacial technology.

Test conditions and locations

Project location: Haining, Zhejiang, China

GROUND SURFACE PAINTED WHITE: The data gathered between May 2018 and January 2019 show the energy output of Jinko’s bifacial with dual glass (156mm size cell) versus monofacial modules where the surface below the arrays had been painted white. The average bifacial gain is about 16% during the whole testing period, while higher in summer (up to 19%) and lower in winter.

Daily generation per watt of monofacial vs bifacial (LHS) and bifacial gain (RHS) on white surface

SAND SURFACE: The Swan bifacial with transparent backsheet reaches bifacial gain of 10.02%, with the same variation between seasons that is a little bit higher in summer and lower in winter.

Daily generation per watt (LHS) and bifacial gain (RHS) on sand

CEMENT SURFACE: This string type test on a cement surface was performed on Jinko’s factory rooftop. The system was designed similar to a C&I project, with horizontal installation. The lowest point of modules is 0.7m off the ground. The average bifacial gain for the Swan with transparent backsheet is 9.74%. Bifacial modules show better low-irradiance performance. As irradiance become lower, the energy generation per watt become lower but the bifacial gain increases.

Daily yield per watt (LHS) and bifacial gain (RHS) in cement project

Project location: Qiong Hai, Hainan, China

SAND SURFACE: Qionghai is located in the southernmost Hainan province of China and has a humid, tropical climate. Modules were mounted in 2P tracker configuration (two-in-portrait) with module height (flatwise) of 2.7m. The ground type is sand, albedo 19%-20%. The bifacial gain of P type Swan TB bifacial module is significantly high, up to 16.66% against Cheetah monofacial mounted on the same scenario. This high bifacial gain can effectively cut down the LCOE.

Daily generation per watt (LHS) and bifacial gain (RHS) on sand in Qionghai project

Project location: Fremont, California, US

GREY GRAVEL SURFACE: Fremont has a dry and warm summer then a wet and rainy winter. The project is located in the outdoor field test station of solar tracking firm Nextracker. The ground is light gray gravel and the measured albedo of the site is around 20%. Considering the control group is monofacial modules installed on the same tracker and ground albedo is 20%, bifacial gain of 8.34% for the transparent backsheet module is considerable.

Daily generation per watt (LHS) and bifacial gain (RHS) in Fremont on grey gravel


The results reveal that the Jinko Swan bifacial module can exhibit a high gain compared with monofacial technology under different scenarios. In a C&I project with fixed mounting and white-painted ground with high albedo (80%-90%) a gain of 16% was achieved, so the same power generation can be realized with fewer modules and less initial investment to cut down the LCOE of the project. Using ordinary cement ground can also achieve a bifacial gain of more than 9% without additional cost of ground laying.

In the scenario of a large-scale power station, the bifacial modules were tested with a tracking system. When the albedo is only 20%, besides the energy gain expected from tracking the bifacial modules provide another 8% energy generation gain, effectively reducing LCOE. In Qionghai, where irradiance levels are low, the Swan bifacial with a transparent backsheet, the bifacial gain is as high as 16.6%, thus effectively reducing the LCOE of the project.