Nextracker recently published findings from its bifacial solar tracker research. It explains some of the advantages inherent within its NX Horizon single-axis tracker design for achieving bifacial gains, but more importantly, it shows how these gains can be accurately modeled within PVsyst. The entire report is worth a deep-dive, but here are some outcomes and tidbits that stood out to us.
Consistency is key
The bifacial + tracker segment is now all the rage, but modeling these projects for accurate / consistent bifacial gains can still be hit and miss. A big goal in Nextracker’s research was to establish reliable PVsyst inputs to improve the bankability of bifacial projections with its NX Horizon system, and it seems to have accomplished this.
Overall, Nextracker is seeing mono-PERC bifacial gains of 5% to 7% under low-albedo conditions (≈20%) and 10% to 12% under high-albedo conditions (≈50%). These results from its Center for Solar Excellence are on par with those measured by PVEL and NREL.
“Insofar as consistency breeds confidence, the general agreement between these field-measured bifacial gains—under both high and low-albedo conditions, as well as across multiple test facilities and states—is encouraging.”
The main discrepancy
PVsyst is reasonably good at predicting bifacial gains on NX Horizon based on manufacturer-recommended loss factors (the data actually has a slight trend toward underprediction). But one discovery from these PVsyst modeled vs. Nextracker field-measured comparisons that will warrant further study is how PVsyst tends to underpredict bifacial gains for sunny months and overpredict bifacial gains for cloudy months.
For the high-albedo case, the PVsyst-modeled bifacial gains are 11.2% over three months, which compares favorably to the aggregated field-measured gains of 12.5%. For the low-albedo case, the PVsyst-modeled bifacial gains are 5.5%, which compares favorably to the aggregated field-measured gains of 5.9%. As a general rule, the offset between the measured and modeled results is consistent for individual months and the cumulative three-month data.
“We believe that the key factors contributing to this modeled versus measured discrepancy are unrelated to the structure shade and mismatch factors that are the primary subject of our investigations. As the accuracy of the PVsyst model improves, we can likewise fine-tune and improve the accuracy of our structure-specific model input parameters. …. In the meanwhile, the general alignment between measured and modeled results indicates that the industry can use standard modeling tools, such as PVsyst, to estimate back-side POA irradiance, and bifacial gains by extension, with confidence.”
NX Horizon-specific advantages
Nextracker posits that its NX Horizon provides an additional 1.02% to 1.67% of DC-side bifacial gains as compared to other 1P single-axis tracker designs. To come to this conclusion, Nextracker modified an experimental tracker design that sat next to its standard NX Horizon design, incorporating all of the same modules, to test the effects of backside mechanical system shading.
The NX Horizon standard design:
• A round torque tube and high-rise mounting rails that result in a tube to- cell height of 90mm.
• A small gap at each bearing assembly, which means the bearings and posts are not located directly below or in close proximity to bifacial cells.
• A gap at each foundation that installers can use as a wire chase.
• The drive system components for NX Horizon are not located directly underneath the bifacial modules but rather in a larger gap at the center of each NX Horizon row.
Differences tested included an octagonal torque tube with a tube-to-cell height of 60mm, as well as foundations and bearings located directly below the bifacial PV cells. These were reconfigured on a monthly basis to represent structural configurations found in alternative one-in-portrait single-axis tracker designs.
The results showed, depending on ground albedo, an NX Horizon-mounted bifacial array generated an additional 0.56% to 1.36% of bifacial gains as compared to the end-of-row spans and an additional 1.28% to 2.09% of gains as compared to the middle-of-row span.
The calculated back-side mismatch loss factor for bifacial systems mounted on NX Horizon came out to be 3.5%.
The 1P vs. 2P debate
Nextracker intentionally designed its two-modules-in-portrait (2P) NX Gemini to eliminate the additional mismatch losses associated with electrical strings split between the east and west panel rows as a way to optimize bifacial gains.
Based on five months of side-by-side test data, bifacial modules mounted on the 2P NX Gemini are generating slightly less bifacial gain than those on the 1P NX Horizon. These empirical data indicate an albedo-dependent bifacial gain advantage for the 1P NX Horizon of 0.6% to 1.2%.
These results conflict with studies done by Soltec to promote its own 2P design, and Nextracker chalks the difference up to a few things, the most important being testing longer rows (the interior-of-row strings exclude edge-of-row effects that tend to exaggerate system-level bifacial gains).
Another issue cited electrical stringing in the 2P design, which has modules from an upper row in the same electrical string as modules on a lower row. In this scenario, the intensity of the reflected sunlight on the back side of the modules varies between the rows based on distance to the ground – increasing mismatch effects. Portrait versus landscape orientation could have a similar impact on mismatch too.
The effect of wiring
The trend toward half-cut cells in bifacial PV modules has general benefits such as lower resistive losses, higher shade tolerance, lower risk of microcracking, and improved durability, but they also benefit 1P tracker designs because the junction boxes are located mid-module, directly over the torque tube.
bifacial, NEXTracker, the Tracker Tracker
To the extent that installers are able to aggregate and route DC source circuit and homerun conductors along the underside of this torque tube, it is possible to minimize back-side shading and optimize bifacial gains. For large wire bundles, installers can add wire hangers or harnesses at each short rail location, providing the necessary support at regular intervals along the torque tube. At each foundation and bearing, installers can use a beam clamp or similar to install wire management hardware that secures and protects the wire bundle while allowing for tracker rotation.