16 May 2023

Soltec’s bifacial tracking increases production by 0.30%

Soltec’s new algorithm for bifacial modules searches for the optimal position of solar trackers in a PV plant
taking into account the sum of front and rear radiation. This algorithm is in addition to others from the company, such as TeamTrack (+6.2%) or the Diffuse Booster (+5.3%). The gain for a 50 MW plant in Spain is 6200 euros per year.

01. Introduction

The use of bifacial photovoltaic modules has increased significantly since 2017, although the bifacial module technology is not new. The first bifacial cells at a lab-scale date back to the 1960s, whereas the technology was first marketed in 1981 by Isofotón, a Spain-based company ref. 1.

In the 90s, a number of sound barriers made up of bifacial modules in a portrait configuration (90 degrees) were installed in the Swiss city of Zurich, but it was not until the end of last decade that this technology became more popular on the marketplace. Nowadays, it is used in a large number of solar plants worldwide due to its greater energy efficiency, although it has also given rise to discussions and new prospects on how to enhance solar plant and tracker design in order to increase bifacial photovoltaic module performance.

In that sense, Soltec designed in 2015 the first solar tracker with bifacial panels for the La Silla plant ref. 2, carrying out since then multiple trials and tests, both a lab and experimental level ref. 3, with the aim to enhance tracker design and increase bifacial gain. The term “bifacial gain” refers to the amount of energy generated by the backside in relation to the frontside, considering the former is affected by various parameters such as albedo, tracker height, shadowing, module temperature and inter-row spacing. This time, Soltec’s Innovation Team went a step further to design a specific bifacial module algorithm capable of optimizing tracker position, thus increasing plant production ref. 4.

02. Algorithms in Solar Tracking Plants

Solar tracking relies on astronomical algorithms or sensors. Soltec uses an astronomical algorithm developed in-house based on solar positioning data from NREL-SPA ref. 5, This algorithm sends information to the tracker with the aim of placing it at a particular angle depending on the Sun’s position. That is how the basic tracking algorithm works. However, with the aim to increase production, there are other algorithms, such as backtracking, TeamTrack in the case of Soltec, which are used to prevent shadowing between tracker rows in the early and late hours of the day, or the Diffuse Booster ref. 6 which uses weather information (forecast and onsite measurements) to enhance production when diffuse radiation is greater than direct radiation.

03. Bifacial Tracking or tracking Bifacial Algorithm

The bifacial tracking algorithm seeks to find the optimal solar tracker position in a photovoltaic plant by taking into account the total amount of front and rear radiation. Until now, tracking algorithms only considered front radiation to calculate the tracking angle, that is, they followed the sun in order to increase Global Plane Of Array (GPOA) irradiance. However, in the case of a plant with bifacial modules, it is possible that depending on the value of some specific parameters, that position might not be optimal for maximized production. In such cases, estimating Global Rear Irradiance (GRI) on the panel rear side plays a key role. Thus, Soltec’s bifacial tracking constitutes a step forward in bifacial technology implementation, in the sense that solar tracking is adapted to it.

Illustration 1. Source: Soltec
04. Algorithm Functioning

This algorithm is used to determine the solar tracker angle at which the production of a particular bifacial module in a specific solar plant is maximized. To do that, the algorithm estimates power generation for a set of possible angles, choosing the optimal one. This estimation takes into account front and rear irradiance, as well as electrical module performance.

For each angle, the Electrical output Power (EP) is calculated as the sum of Electrical Power in the Front side (EPF) and Electrical Power in the Rear side (EPR), as shown in the following equation (1). EPF and EPR are obtained taking into account power conversion efficiency (η) and module bifaciality (BIF) for each side depending on GPOA and GRI, including an irradiation Mismatch Rear-side Factor (MRF), as shown in the equation below (2).

Eq = EPF EPR(1)
E = η GPOA η BIF GRI (1 – MRF)(2)

The GPOA is calculated using traditional methods ref. 7, whereas the GRI estimation is obtained with the application of ray-tracing techniques ref. 8 and 9 and comprehensive modeling. Global and diffuse horizontal irradiance are used as input data. The inclination angle as well as the tracker and scene geometry (descriptive plant parameters) are also used, including the surface reflectance colors and coefficients of all items. Simulations provide a module surface irradiance distribution which is then used to calculate the average GRI value and estimate the MRF ref. 10.

05. Input Variables

To design the algorithm, Soltec took into account some descriptive plant parameters which influence bifacial module production and, in turn, tracker position.

Soltec also considered other parameters and variables obtained during plant operation, such as the following:

  • Global and diffuse horizontal irradiance: The algorithm uses solar irradiance data to calculate the GPOA and GRI of upcoming intervals. Such data is obtained from weather forecasts and then validated by means of plant sensors so as to enhance accuracy.
  • Albedo (percentage of radiation reflected by a surface in relation to the irradiation it receives): The algorithm uses the plant soil albedo value established beforehand for each month. In the case of soils with a dynamic albedo, the system also allows an external configuration which favors increased efficiency and online updates.
06. Methodology

The algorithm processes descriptive photovoltaic plant parameters as well as operating variables, albedo and irradiation, to calculate the optimal angle at which bifacial generation is maximized. With the aim to optimize algorithm effectiveness, detailed modeling of the plant’s operating mode and geometry is carried out during a preliminary phase known as “Adjustment Phase”, where key performance parameters are calculated. Later, during the Operation Phase, the algorithm is used to calculate the deviation angle in relation to the monofacial tracking position, integrating the obtained value within the remaining plant performance algorithms (e.g. TeamTrack) in order to maximize gain under various environmental conditions.

Illustration 2. Source: Soltec
Illustration 3. Simplified performance diagram of the bifacial tracking algorithm. Source: Soltec

It applies to 1-in-portrait and 2-in-portrait trackers. Due to a highly detailed geometrical plant modeling, it is optimized for application to soils with a dynamic albedo, as well as for Albedo Enhance Materials (AEM), including an update of the reflection surface area dirt level. Since the algorithm was designed for 50-MW plants, in the case of plants with a greater power output, it should be applied to modules or subfields of less than 50 MW. It is compatible with all commercial bifacial modules, regardless of the number of cells (72, 78), transparency, technology, frameless and framed.

07. Anaytical Testing

The performance and gain of the bifacial tracking algorithm depends on the following:

  • Weather conditions: level of irradiation and percentage of diffuse radiation.
  • Accuracy of weather forecast.
  • Site latitude.
  • Soil albedo.

Testing carried out by Soltec aims to quantify the power output gain obtained with this algorithm by combining all above-mentioned variables. Tests were carried out in three different latitudes: Mediterranean, Equatorial and Northern, for a 50-MW plant and albedo levels of 20%, 30% and 40%. Obtained results can be seen below on tables 1, 2 and 3:

Illustration 4. Example of application with AEM. Source: Soltec
07.1. Annual Results

Being Δ Total = annual gain, Δ Period = gain for activation period and Δ Max hour = Maximum value reached in 1 hour.

Table 1. Results at the Mediterranean Site. Source: Soltec
Mediterranen
Albedo %Δ Total kWh/kWp/year(%)Δ Period %Δ Max hour %
20%2.78 (0.12%)0.34%5.40%
30%2.48 (0.10%)0.24%4.89%
40%2.56 (0.10%)0.20%4.49%
Table 2. Results at the Equatorial Site. Source: Soltec
Equatorial
Albedo %Δ Total kWh/kWp/year(%)Δ Period %Δ Max hour %
20%0.62 (0.02%)0.09%3.25%
30%1.22 (0.04%)0.14%2.55%
40%3.22 (0.11%)0.16%4.57%
Table 3. Results at the Northern Site. Source: Soltec
Northern
Albedo %Δ Total kWh/kWp/year(%)Δ Period %Δ Max hour %
20%1.73 (0.12%)0.41%5.77%
30%1.78 (0.12%)0.39%5.67%
40%1.63 (0.11%)0.33%5.24%

Results in the tables show that the algorithm is able to increase plant power generation around 0.10% in Mediterranean latitudes, with an albedo of 30%. As for Northern sites with the same albedo, the increase amounts to 0.12%. In the case of Equatorial sites, there is a substantial increase under conditions of high albedo (40%), with a gain of 0.11%.

07.2. Monthly Results

Results vary considerably between months, although there is a clear trend to improve under lower solar angles, with a greater percentage gain in winter months.

Graph 1. Monthly bifacial gain for the three latitudes. Source: Soltec
07.3. Climate Sensivity

The differentiating element of this algorithm is its capacity to adapt to climate conditions based on weather forecast information. Results show that, with a few exceptions in the Equatorial case, the lower the percentage of extra diffuse energy generated, the higher the bifacial tracking, as shown in the following graph, where penalties due to frontside disorientation are lower. However, this absolute increase is not so evident when the percentage gain is assessed, due to lower irradiation levels under these conditions.

Graph 2. Bifacial gain based on the diffuse rate. Source: Soltec
07.4. Bifacial Gain
Table 4. Bifacial gain for the three latitudes compared to standard tracking (facing the sun). Source: Soltec
*Standard Tracking
**Bifacial Tracking
Bifacial gainAlbedoSTDT*BfT**Δ
Mediterranean20%8.30%8.43%0.13%
30%11.71%11.82%0.12%
40%15.10%15.22%0.12%
Equatorial20%8.30%8.34%0.02%
30%11.85%11.89%0.05%
40%15.35%15.47%0.12%
Northern20%9.65%9.79%0.13%
30%13.30%13.43%0.14%
40%16.87%16.99%0.13%

Table 4 compares the impact of standard and bifacial tracking on the bifacial gain, the latter being the extra percentage of production of the bifacial module compared to that of a monofacial module. The bifacial gain increase translates into an economic gain for the plant, which in this particular case was calculated for a price of €50 per MWh. The results obtained are shown in the following table:

Table 5. Economic gain of bifacial tracking. Source: Soltec
Bifacial gainAlbedoTBFEUR
Mediterranean20%2.786,945
30%2.486,192
40%2.566,400
Equatorial20%0.621,550
30%1.223,055
40%3.228,041
Northern20%1.734,321
30%1.784,449
40%1.634,068
08. Combination with Diffuse Booster and Teamtrack

The bifacial tracking algorithm is compatible with other Soltec algorithms, more specifically, the ‘Diffuse Booster’, an algorithm capable of maximizing module generation when the diffuse radiation percentage is greater than the direct radiation percentage, as well as TeamTrack, capable of increasing plant generation in up to 6.2%.

09. Experimental Testing

Upon completion of analytical testing, Soltec launched some experimental testing on its test field in Murcia (Mediterranean latitude) with the aim to doublecheck simulation results and determine practical algorithm performance. To do that, two solar trackers were compared, one using the algorithm and the other a traditional tracking mode of monofacial modules. The difference in tracking angle behaviour is shown below.

Graph 3. Comparison between BfT and STDT tracking angles. Source: Soltec

Using the operational algorithm for an entire month contributed to an overall improvement of 0.183%, with a mean of 1.48%, a percentile 90 of 4.95% and a maximum value of 9.33%. Gain during the activation period was 0.37%.

Bifacial tracking was also tested at BiTEC California, with AEM [poster ref. – point 4] obtaining an algorithm gain compared to regular solar tracking of 1.64%. Another conclusion reached is the excellent sensitivity to weather forecasts of the algorithm.

After a reliability analysis carried out by the weather specialist contractor, the average diffuse ratio error was measured at 5.7%, with a hit rate of 89%, translating into an average angle error of 1.57º.

In that regard, the annual gain estimate is also sensitive to the selected meteorological year, especially considering that the impact of small diffuse variations is not negligible.

Bibliography

ref. 1 Eduardo Lorenzo, On the historical origins of bifacial PV modelling, Solar Energy, Volume 218, 2021, Pages 587-595, ISSN 0038-092X, https://doi.org/10.1016/j.solener.2021.03.006. Cuevas, Andrés. (2005). The early history of bifacial solar cells.

ref. 2 A.Di Stefano, G. Leotta, F. Bizzarri. “La Silla PV plant as a utility-scale side-by-side test for innovative modules technologies”. 33rd European Photovoltaic Solar Energy Conference and Exhibition EUPVSEC2017 September 2017 Amsterdam. Proc 6CO.14.1 p.p 1978 – 1982, online reference: https://www.eupvsec-proceedings.com/proceedings?paper=44211

ref. 3 J.Guerrero-Pérez, I. Muñoz Benavente, J.Navarro, The Bifacial Year, Soltec White Paper, online reference https://soltec.com/wp-content/uploads/2019/11/BiTEC-whitepaper-4_en.pdf

ref. 4 J. Guerero-Perez, A.Ros Gomez, Tracking Algorithms for Improving Bifacial Modules Energy Yield with Albedo Enhance Materials and Agrivoltaics, Poster THU – E – 06 SiliconPv 11-14 April 2023 Delft

ref. 5 Reda, Ibrahim; Andreas, Afshin, SPA: Solar Position Algorithm, Astrophysics Source Code Library, record ascl:1504.002, April 2015, Bibcode: 2015ascl.soft04002R. https://www.nrel.gov/docs/fy08osti/34302.pdf / Link tool: https://midcdmz.nrel.gov/solpos/spa.html

ref. 6 Magnus Herz et all, 5.3% Diffuse Booster, Soltec White papers, online reference https://lab.soltec com/5-3-diffuse-booster/#02

ref. 7 R. Perez, P. Ineichen, R. Seals, J. Michalsky, R. Stewart. Modeling Daylight Availability and Irradiance Component from Direct and Global Irradiance. Solar Energy 44, no 5, pp 271-289, 1990.

ref. 8 Stein et al. 2019. Bifacial Photovoltaic Performance Optimization Using Ray Tracing and High Performance Computing, Keynote presentation at Photonics North, Quebec City, Canada. https://pvpmc.sandia.gov/download/7213/

ref. 9 Ayala Pelaez, C. Deline, P. Greenberg, J. S. Stein, and R. K. Kostuk, “Model and Validation of Single-Axis Tracking with Bifacial PV”, IEEE J. Photovoltaics, vol. 9, no. 3, 2019. https://ieeexplore.ieee.org/abstract/document/8644027

ref. 10 Chris Deline et all, Bifacial PV System Mismatch Loss Estimation and Parametrization, PVSEC2019 Marsella September 2019. https://www.nrel.gov/docs/fy19osti/74831.pdf

Authors
Javier Guerrero-Pérez holds a Ph.D. with honors in Renewable Energy. His professional activity spans over fifteen years in the solar industry within multinational EPC operations. He has published several papers on modelling electrical behavior of both PV modules and inverters, oriented to large scale simulation. With Soltec since 2015, he has work on standardization and optimization of bifacial PV trackers and managing Soltec’s Bifacial Tracker Evaluation Center in Livermore, California. Current research lines are focused on tracking algorithms for energy yield optimization at Soltec Innovations in Spain.
Irene Muñoz Benavente holds an International Ph.D. with special honors in Renewable Energy from the Technical University of Cartagena, working in the renewable energy sector since 2016, specifically in Photovoltaic. Focused in optimizing tracking algorithms for solar trackers.
Samir Chaouki-Almagro holds a B.S. degree in Automatic and Industrial Electronics Engineering, M.S. degree in Electrical Engineering and M.S. degree in Renewable Energies and he is currently working toward the Ph.D. degree. His research interests are isolated communications for power multistage converters, photovoltaic grid connected inverters and since 2015 working on PV tracker control systems.
Antonio Fabián Ros Gómez holds a bachelor’s degree in Industrial Electronics and Automation Engineering from the Technical University od Cartagena. Focused in optimizing tracking algorithms and perform solar plants modelling, simulations and data analysis.
background--light_fall