The Bifacial Year

Soltec Power Holdings
Download PDF

1. ABSTRACT

This whitepaper is the fourth and last part of a summary presenting the results obtained from September 2018 to September 2019 at Soltec’s Bifacial Tracking Evaluation Center (BiTEC) in Livermore, CA, USA. For a full year, BiTEC studied bifacial gain in photovoltaic modules developed according to the two most common market configurations nowadays: 1-in-vertical (1P) and 2-in-vertical (2P). Full year results show that individual modules in 2-in-vertical configuration obtain a bifacial gain of 15.7% under high albedo conditions (55.6%), that’s 2.1% more than the same modules under a 1P configuration for identical conditions.

This document will also build on information provided in previous whitepapers to better understand the impact of bifacial gain factors, with the aim to enhance knowledge and optimize bifacial module performance.

Keywords

SF7, BiTEC, Albedo, Bifacial gain, Bifacility, Single-Axis Tracker, Monofacial, Photovoltaic, Large-scale solar plant, Bifacial ratio, Irradiation, Diffuse radiation, Radiation, Bifacial projects, Bifacial Tracker Evaluation Center, Sankey Diagram, Energy, Solar module, PVsyst, NREL, RETCHanwha Q CELLS, JinkoSolar, LG, LONGi, REC, Soltec

2. INTRODUCTION: A YEAR AT BITEC (SEPTEMBER 2018 TO SEPTEMBER 2019)

In 2018, with the aim to assess factors affecting bifacial gain and to create models predicting power generation in any bifacial gain plant, Soltec inaugurated its Bifacial Tracking Evaluation Center (BiTEC) of Livermore, California (USA) [1]. After one year researching and gathering data on three different types of albedo, results show that:

  • For high albedo (55.6%), that is, white sand or snow-covered ground, the bifacial gain of a 2P SF7 bifacial tracker is 15.7%.
  • For a medium albedo (29.5%), such as sand, the bifacial gain is around 9.6%.
  • For cultivation areas with an albedo of 19.9%, the bifacial gain of a 2P SF7 bifacial tracker is 7.3%.
  • SF7 Bifacial trackers can generate 2.1% more power than standard 1P trackers under the same GCR and albedo conditions.

For each scenario, bifacial gain assessments were based on an output comparison between monofacial modules and bifacial modules in central tracker positions, where shading and reflection are similar to those of internal modules in a photovoltaic plant. Module optimizers ensure maximum output (MPP) of each individual module. Inverter DC/AC ratio is lower than 1 to prevent inverter clipping. Measurements at a string level were also taken by means of AMP optimizers, with the aim to ensure satisfactory MPP performance.

Figure 1. BiTEC test plant, Livermore (California). Source: Soltec

*Black & Veatch reviewed the data and measurement techniques used by Soltec to produce tables one through six of Soltec’s first white paper on bifacial tracking (the “Technical Tables”). Black & Veatch reviewed the methodology used by Soltec to generate the Technical Tables and believes that the methodology is consistent with accepted industry practice [2]. The methodology, which was used in subsequent papers, can be summarized as follows:

  • Estimated Bifacial Gain was assessed by comparing normalized bifacial module output under various albedo-GCR conditions, divided by the normalized output of a monofacial Jinko JKM320PP-72 module under the same conditions.
  • Normalized output is defined as the division between a module’s power output and the maximum specified power under flash test conditions.
  • Results on this report refer to bifacial module Jolywood JW-D72N-355.

Black & Veatch Management Consulting LLC (Black & Veatch) reviewed the data and measurement techniques used by Soltec Energía Renovables S.L. (Soltec) to produce tables one through six of this white paper (the “Technical Tables”).  Black & Veatch reviewed the methodology used by Soltec to generate the Technical Tables and believes that the methodology is consistent with accepted industry practice.

Black & Veatch understands that Tables seven through eleven of this white paper (the “Economic Tables”) are based upon information provided by Powertis S.A., a developer of large-scale PV projects in Europe and Latin America. Black & Veatch did not independently verify this information and cannot verify its accuracy.  If the data is accurate, Black & Veatch believes the Economic Tables are consistent with accepted engineering practice.

Legal Notice: This statement is qualified in its entirety by the full Black & Veatch Report. The statements contained in this document are for general information purposes only. Black & Veatch makes no representations or warranties of any kind about the accuracy or suitability of such statements for any purpose. B&V shall not be responsible or held liable for any losses or damages related to your reliance on this document and you hereby waive and release B&V from any such losses or damages of any kind or nature, including any consequential, indirect, special, or incidental damages.

3. BIFACIAL TRACKERS: SOLTEC’S EXPERIENCE

3.1. La Silla

In 2015, at the European Southern Observatory La Silla, Soltec carried out a number of comprehensive analyses of factors conditioning the radiance absorbed by the module’s rear side, with the aim to design the first bifacial tracking system in the marketplace [1]. This design entailed various construction challenges, including the need to understand which are the determining factors of rear irradiation capture. Results revealed the key factors: albedo, distance between rows (known as “pitch”), height above the ground and shading percentage of the rear side.

In this case, the plant had a proven bifacial gain of 12% under natural albedo conditions [3].

3.2. BiTEC

As abovementioned, Soltec inaugurated in 2018 its Bifacial Tracking Evaluation Center, (BiTEC) in Livermore, California (USA) [4].

The Center aims to improve knowledge on bifacial performance by adjusting models that facilitate the development of both bifacial generation estimates and modern technologies in the area of bifacial tracking. To do that, Soltec collaborates in this research initiative with NREL (National Renewable Energy Laboratory), Black & Veatch and the Renewable Energy Test Center (RETC), as well as with key module manufacturers and stakeholders within the photovoltaic industry, such as Canadian Solar, Jinko Solar, Jolywood, LG, Longi and REC, among others.

The plant is comprised of 18 solar trackers (2P SF7 bifacial and standard 1P) installed under three different albedo conditions and three different pitches, thus favoring performance comparisons among various photovoltaic modules under a wide variety of scenarios.

Figure 2.  La Silla PV Plant (Chile), 2015. Source: Soltec
Figure 3. Trackers under different albedo conditions at BiTEC. Dirt test
Albedo GCR
0.46 0.40 0.33
White 49-65% 2P 2P / 1P 2P
Gravel 24-36% 2P
Seasonal 16-23% 2P / 1P
Table 1. Scenarios analyzed at BiTEC. Source: Soltec
In Blue:
Modules used for
measurements
In Red:
Aisle Pitch in meters:
In White, Brown and Green:
White, Gravel and Seasonal
albedos respectively
2P – 12.0, 10.0 and 8.7 meters
equivalent to a GCR of 0.33,
0.40 and 0.46 respectively
Figure 4. Layout BiTEC. Source: Soltec

4. FACTORS INFLUENCING BIFACIAL GAIN

4.1. Albedo

Albedo is the non-dimensional ratio of radiation flux reflected by a surface. It varies between 0 and 1, with 0 meaning total absorption (black surface) and 1 meaning total reflection (white surface).

Albedo measurement is more complex than assuming a single constant value for a surface, because this changes with soil water content, type of soil, tillage, amount and type of crop-residue cover, etc. [5], with values ranging from 10-20% for rough ground surfaces of dark color, to 40-50% for smooth ground surfaces of light color. The albedo of snow, especially fresh, deep snow can reach values of up to 90% [6].

Approximate albedo ranges for different natural surfaces are shown in the following table.

As seen in the table, ranges available in the bibliography are quite large, meaning it is necessary to perform field measurements for enhanced reliability.

A number of surfaces were measured using the albedo measurement procedure described in the previous whitepaper [7]. Results are shown in Figure 5.

Natural surface types Approximated albedo
Blackbody 0%
Forest 5-20%
Grassland and cropland 10-25%
Dark-colored soil surfaces 10-20%
Dry sandy soil 25-45%
Dry clay soil 15-35%
Sand 20-40%
Mean albedo of the earth 36%
Granite 30-35%
Glacial ice 30-40%
Light colored soil surfaces 40-50%
Dry salt cover 50%
Fresh, deep snow 90%
Water 10-100%
Absolute white surface 100%
Table 2. Typical albedo values for different ground surfaces [6] Source: University of Miskolc.
Figure 5. Albedo values depending on the type of ground. Source: BiTEC

From these findings, three different types of surfaces were selected at BiTEC for research purposes, with the aim to assess the influence of albedo on bifacial module generation. The three types are: white fabric (52-63% of albedo), gray gravel (~34%) and vegetation cover, with seasonal variations (15-27%).

  • White Albedo: A thin layer of white fabric covering the ground, with measured albedo values ranging between 49-65%. The following figure shows albedo degradation in time due to debris accumulation. Figure 6.
  • Gravel Albedo: Grayish gravel on the ground, with measured albedo values ranging between 24-36%. Figure 7.
Figures 6.1 & 6.2. White ground cover at BiTEC. Source: Soltec
Figure 7. SF7 bifacial tracker installed on gravel at BiTEC. Source: Soltec
  • Seasonal Albedo: Fallow land left aside during construction and evolving freely, with measured albedo values ranging between 16-23% and going through various stages: virgin soil, green grass (which was later mowed) and lastly dry grass. Figure 8.
Figure 8. Vegetation evolution in seasonal albedo areas

During the wet season, rainwater puddles caused reflections on the surface, with an unusual effect on bifacial modules: higher rear irradiation due to short-lived albedo increasement. However, such effect did not result in significant power generation gains. In fact, albedo decreased during the following days due to the dark color of wet soil.

Figure 9. Example of reflection effects caused by puddles in January 2019. Source: BiTEC

Ground surface conditions were subject to continuous changes impacting their albedo. As an example, the experience at BiTEC supports the hypothesis that albedo, and in turn irradiation reflected on the rear side of modules, decreases in humid environments [5].

Measurements for the different types of surfaces analyzed at BiTEC confirm that soil changes in time have a direct impact on the albedo value. The figure 10 shows measurement variability compared to the range of typical values established by NREL for each type of surface.

Figure 10. Albedo measured at BiTEC for each type of soil versus the range of NREL values. Source: [8]

It is important to mention that bifacial gain estimates for long-term analyses requires the use of representative albedo values throughout the period being analyzed.

4.2. Annual Results at BiTEC: Albedo Bifacial Gain

Figure 11 shows full-year results for different albedo combinations.

As seen in the table:

  • For high albedo (55.6%), that is, white sand or snow-covered ground, the bifacial gain of a 2P SF7 bifacial tracker is 15.7%
  • For a medium albedo (29.5%), such as sand, the bifacial gain is around 9.6%.
  • For cultivation areas with an albedo of 19.9%, the bifacial gain of a 2P SF7 bifacial tracker is 7.3%.

BiTEC analyses revealed a linear dependency between bifacial gain and albedo. Besides, the influence of the remaining parameters under analysis, such as height, GCR and temperature, is always affected by the albedo. However, its influence can be assessed separately.

Type of ground Albedo Bifacial Gain
White 55.6% 15.7%
Gravel 29.5% 9.6%
Seasonal 19.9% 7.3%
Figure 11. Dependency of yearly bifacial gain in relation to albedo. Yearly Values Measured ar BiTEC. Source: Soltec 
 

5. GCR: INFLUENCE ON BIFACIAL GAIN

5.1. Pitch and maximum tracking angle

The transposition factor is the ratio between incident irradiation (GlobInc) on the plane and horizontal irradiation (GlobHor), that is, the gain (or loss) obtained when the collector plane is tilted. This is computed by applying a transposition model [9] to the horizontal hourly values. At PVSyst, for instance, this value is reflected in the Loss Diagram, expressed as a percentage, as seen below. Figure 12.

To analyze the influence of pitch (variable in each site) and maximum tracking angle (± 60º for the 2P SF7 model) on this factor and, in turn, on the plant’s power performance, a number of simulations were carried out using PVSyst. Figure 13.

The following figure shows that the smaller the pitch (that is, the larger the GCR), the smaller the transposition gain. It also shows that the larger the maximum tracking angle, the larger the transposition gain.

Figure 12. Example of Loss Diagram, Transposition Factor = +29.7%. Source: PVSyst 
Figure 13. Transposition Factor for different GCR and Maximum Angles. Source: PVSyst

That means pitch is a determining factor in bifacial module generation. Increased distance between trackers implies more hours of tracking and, therefore, increased module front side generation. In other words, decreased generation resulting from tracker deviation during backtracking is lower. This is true for projects equipped with both bifacial modules and monofacial modules. Furthermore, for bifacial module projects, energy absorption in the module’s rear side during tracking hours is higher, meaning pitch impact on bifacial gain is a factor closely related to the albedo, as explained in White Paper 3 [7].

Maximum tracker angle is another key factor. For the same GCR, trackers with a larger tracking angle can operate longer in tracking mode (directly facing the Sun), thus generating more power by direct irradiation. When solar elevation exceeds the limit angle, tracking stops and energy is lost (equivalent to the deviation cosine). Figure 13 shows a scenario with a solar elevation of 30 degrees in which the tracker with a maximum angle of 60 degrees remains oriented but the tracker with a maximum angle of 50 degrees cannot. In the latter case, power reduction is proportional to the difference between shaded areas.

This difference is particularly significant in the case of bifacial plants, where GCR values range between 0.32-0.34 (area marked in grey). This is explained by the fact that the time between the tracker being “disoriented” until its return to an optimal position is longer.

It is also worth mentioning that increased power generation happens during the early and late hours of the day, when electricity prices tend to be higher.

Pitch (m) Tracking Angle (o) kWh/kWp/year Diff
12 60 2046 +0.40%
12 50 2038
Table 3. Specific generation based on maximum tracking angle. Source: PVSyst

Furthermore, as seen in the figure 14, as pitch increases, the surface area capable of reflecting sunlight onto the module’s rear side also increases, thus favoring optimized power generation in the rear side of the module. This effect is more significant for higher albedo values, as explained in the previous White Paper [7].

Figure 14. Graphic representation of reflected irradiation variation depending on the pitch. Source: Soltec

6. DC/AC RATIO

The DC/AC (also known as ILR, “Inverter Load Ratio”) is an important parameter to be considered when designing a solar plant. For instance, a 6 kW DC system with a 5 kW AC inverter would have a DC/AC ratio of 1.2 (6 kW / 5 kW = 1.2). The key is to design the system accurately to prevent power losses caused by inverter clipping, a phenomenon occurring when DC power supplied to an inverter exceeds inverter power intake capabilities. In such cases, the power surplus is “cut” and lost [10].

With PVSyst, for instance, these losses are reflected in the Loss Diagram, expressed as a percentage, as seen below.

Figure 15. Example of Loss Diagram, Clipping = -0.66%. Source: PVSyst

The following table and graph shows PVSyst simulation results obtained at BiTEC. They present the inverter loss over nominal inverter power for different DC/AC ratios and two different albedo conditions (53% and 19%).

Inv. Loss over Nom. Inv. Power
Bifacial Monofacial
DC/AC 

Ratio

Albedo 

53%

Albedo 

19%

Albedo 

19%

1 0 0 0
1.1 -0.14 0 0
1.15 -0.28 -0.01 0
1.20 -0.85 -0.07 0
1.25 -1.50 -0.33 -0.3
1.30 -2.40 -0.71 -0.11
1.35 -3.73 -1.39 -0.48
1.40 -5.25 -2.34 -1.06
1.45 -6.41 -3.47 -1.83
1.50 -7.93 -4.60 -2.81
Figure 16. Inv. Loss over Rat. Inv. Power. Source: PVSyst simulations

As can be observed, albedo is a determining factor for inverter losses, because the higher the albedo, the higher module performance and therefore, the higher the power output. That also entails a reduced need to oversize DC power (in other words, lower DC/AC ratio), thus preventing clipping-induced losses.

A comparison between both graphs reveals that for a DC/AC ratio under 1.20, inverter losses for a 19% albedo are negligible, whereas for a 53% albedo they reach up to 1%.

Ideally, there should be no plant losses caused by the “cut” effect. Having said that, inverter power sizing does not only consider this requirement, but also the need to strike an economic balance based on multiple factors, such as inverter cost vs. kWh or grid power limitations.

7. COMPARISON BETWEEN 1P TRACKERS VS. SF7 BIFACIAL 2P TRACKERS

BiTEC findings reveal that a bifacial module installed on a SF7 Bifacial 2P tracker has significant operational differences with the same module installed on a 1P tracker. Such differences would explain the 2.1% power generation increase measured on a SF7 bifacial tracker versus a standard 1P tracker. To quantify this increase, the Bifacial Gain¹ of bifacial modules Jolywood JW-D72N-355 installed on the SF7 Bifacial 2P tracker (horizontal module height: 2.35 meters) was compared to the bifacial gain of a standard 1P tracker (horizontal module height: 1.35 meters) with a Ø 100 mm round tube. Both trackers were under white albedo (around 55.6%) and GCR 0.4 conditions. Figure 17 shows comparison results for these modules in each season and for the entire year.

1P Standard tracker Measured Bifacial Gain 2P SF7 Bifacial
16.8 Fall 19.2
12.6 Winter 14.3
11.2 Spring 13.1
13.7 Summer 15.8
13.6 Year 15.7
Figure 17. Bifacial Gain calculation for modules on trackers in 1P & 2P configurations for one year (September 2018-September 2019). Source: BiTEC
¹Bifacial Gain

As can be observed, the enhanced annual power generation of a SF7 bifacial tracker is consistent in all seasons, with an upward trend during periods of increased irradiation. Below we present how increased height and reduced restraint system interference favor increased rear side irradiation, as well as better irradiation distribution and more effective module cooling during SF7 2P bifacial tracker operation.

7.1. Increased rear radiation with height

Bifacial module height has a direct impact on power generation because as module distance from the ground increases, bifacial module rear side irradiation does too [11]. Module distance from the ground facilitates diffuse radiation penetration and increased view factor [12]. Optimal bifacial module height ranges between 2 to 2.5 meters, as explained in more detail in the first White Paper [2].

The view factor model is used to calculate reflected irradiation in one point as function of the reflective surface area, and the surface angle in with respect to that point. In the case of bifacial modules, height causes rear irradiation from ground reflection to increase because modules are sensitive to ground areas reached by sunlight with a better angle.

When comparing 2P and 1P trackers, it is necessary to make several comments on view factor application and its geometric impact, especially considering that 2P trackers are twice as wide. For an identical GCR, the shading of 2P trackers and aisles is two-fold that of 1P trackers, meaning the sunlit ground area reflecting radiation is the same, although the vision angle of such area would change with a different tracker height proportion. In this type of comparison, it is common to use normalized height (height/width), as a view factor generation factor. The design of tracker sizing usually includes a standard minimal clearance between tracker and ground. That means normalized standard 2P tracker height tends to be lower, which in turn means that reflected radiation capture in lit areas is also lower.

Figure 18. Graphic representation of irradiation capture in a 1P and 2P tracker. Source: Soltec

The design of tracker sizing usually includes a gap between tracker and ground. That means normalized standard 2P tracker height tends to be lower, which in turn means that reflected radiation capture in lit areas is also lower.

2Bifacial gain in each case, estimated in relation to standardized monofacial module Jinko JKM320PP-72 generation, in line with the methodology described in WP1 [4].

SF7 bifacial design overcomes this view factor inconvenience thanks to increased standard 2P tracker height (2.35 m) and includes a gap above the torque tube. Such gap enables irradiation to enter and clear shading in the central part of the tracker. This allows reflected irradiation to be fully captured, favoring a 1.7% radiation gain, as described in White Paper 2 [13].

Comparative PVsyst simulations for a full year show that bifacial SF7 can use up to 8.2% of ground reflected radiation (view factor estimate), whereas the 1.35-meter high 1P tracker can use up to 9.2%. View factor modeling neglects reflections on shaded ground areas. However, irradiation reflected by shaded ground areas (estimated with bifacial radiance [14] shows that 2P trackers capture 0.84%, whereas standard 1P trackers capture 0.09%.

In addition to directly reflected energy, estimates calculated using the Pérez model [9] make it possible to quantify diffuse irradiation at 17.5% of energy captured in the SF7 bifacial tracker rear side, whereas rear diffuse irradiation for 1P trackers is only 15%. In short, SF7 bifacial trackers capture 2.5% more of diffuse radiation in the rear side.

There are other determining operational factors that impact on bifacial generation. Module height should be measured in relation to the reflecting surface area. In an operating plant with seasonal ground, the reflective surface area is no longer the ground but the grass, as shown in the examples of figure 14. That means that the relative height of both 1P and 2P trackers decreases, although significantly more in the former when grass grows former when grass grows, as it is illustrated on figure 20.

Figure 19.  Reduced effective height of 1P trackers. Source: RES
Figure 20.1. Influence of useful tracker height in relation to vegetation growth 1P Vs. 2P. Source: Soltec
Figure 20.2. Growth 1P Vs. 2P

7.2. Rear Interferences Impact

Similarly to monofacial modules, shading has a negative impact on bifacial module performance. Some analyses suggest that module mounting system shading could result in a rear irradiation loss of up to 20% [15].

In the case of standard 1P solar trackers, the torque tube is a significant module shading source.  A number of analyses [16] point out to the fact that projected 1P tracker torque tube shading proyection varies along the module depending on the time of the day, potentially reaching an irradiation decrease of up to 15% with a distinctive peak, as seen in the figure 21.

Figure 21. Torque tube shading behavior [17].  Source: NREL

Specific interference location does not only favor reduced rear radiation, but also increases mismatch losses and may contribute to overheating of some cells, and as a result of a premature module degradation.

SF7 bifacial trackers are designed to prevent torque tube shading by having a gap between modules over the torque tube, as shown in the figure 22. With the aim of representing these effects in PVSyst simulations, the externally calculated shading factor needs to be included because this software cannot simulate designed tracker features. As described in White Paper 2, [2] the tube shading factor recommended by NREL for the SF7 bifacial tracker is 0.0%, because comparative simulations using Bifacial Radiance 0.3.0 show that the central gap generates a reflection with a 1.7% radiation gain. For 1P trackers, the same evaluation yields a torque tube shading factor of 5.6%, accounting for a total power loss of 0.7%.

Figure 22. Shading factor estimation 1P and 2P. Assessment with Bifacial Radiance 0.3.0 during one year for Livermore (CA), GCR=0.4 Albedo 38%. Source Soltec/NREL.

7.2.1 Module Mismatch

Various analyses [111819] suggest that irradiation distribution on the rear side of modules closer to the ground, is more heterogeneous. As explained in White Paper 3 [2], this height-induced irregular radiation distribution implies mismatch losses when modules are closer to the ground. In addition to this mismatch, many authors have explained that heterogeneous irradiation distribution is even more intense when interferences are present [15192021], causing an additional mismatch loss. The behavior of tracker torque tubes is quite interesting, as shown in the figure 23.

Figure 23. Rear irradiation distribution on Single Axes 1P tracker. Source:  University of Ottawa.
Figure 25. Example of rear radiation distribution in tracker 2P and 1P. Assessment with Bifacial Radiance 50 sensors. Source: BiTEC
Figure 26. Increased mismatch losses in 1P tracker as a result of height, tube inclusion. Source. NREL [22]

Therefore, in the case of 1P trackers, it is necessary to consider both the 5.6% power reduction mentioned in the previous section, as well as the mismatch loss. NREL has proposed a method for mismatch quantification based on standard rear irradiation deviation [20]. The following formula can be used to evaluate the “M” parameter entered into PVSyst as mismatch rear radiation.

Figure 24. M parameter estimate in Pvsyst. Source: NREL

Table 4 shows a comparison between mismatch calculations for 1P and 2P SF7 Bifacial trackers. To do that, Bifacial Radiance by NREL [23] was used to estimate the typical deviation of distributed tracker rear side radiation, reducing performance to only a half day and differentiating between lower and upper modules of the SF7 bifacial tracker. The evaluation included in table 4 shows that the lower tracking module of SF7 Bifacial 2P trackers has values nearing 18%, accounting for an energy loss of 0.74%. As observed in the table, these values are similar to 1P tracker values, which had its module at a similar distance from the ground for a half day. However, standard rear irradiation deviation in the upper SF7 bifacial module is reduced to values of about 6%, accounting for an energy loss of about 0.31%. The 2P tracker mismatch improvement is caused by tracker modules being less time near the ground.

2P SF7 Bifacial Standard 1P
σ UPPER module 6.30%
σ LOWER module 17.80% 18.50%
σ Average 12.00% 18.50%
M Irradiation Mismatch PVSyst param 4.11% 6.36%
Energy lost 0.31% 0.47%
Table 4. Rear irradiation mismatch and rear standard deviation comparison for one year. Source: BiTEC

Results obtained at BiTEC are within the range of analyses presented by other organizations. As seen in the figure 26, mismatch losses calculated by NREL when taking the torque tube into account, are approximately a fifth larger than when the torque tube is not considered.

On the other hand, 2P tracker mismatch occurs at a string level due to different module heights. As explained in WP3 [7], the mismatch at a string level was less than 0.06%, meaning it was much lower than actual mismatch losses caused by rated power tolerance according to the manufacturer.

7.3. Influence of Height. Cooling.

The electrical conversion efficiency of a PV cell decreases as temperature increases, meaning the power generation of a module, both bifacial and monofacial, increases as module cooling improves. The downside of bifacial modules is that they generate more heat due to rear side irradiation capture, meaning they tend to work at higher temperatures and, therefore, need additional cooling.

As pointed out in the first White Paper [2], measurements performed at BiTEC show that according to temperature distribution along the tracker, temperature decreases as distance from the ground increases. This trend applies to both 1P and 2P tracker, meaning the average temperature of the lower module in a 2P tracker is similar to that of the 1P tracker module. As for the upper module, temperature values are lower. Measurements show that the average difference of 6 degrees Celsius translates into a power gain for the entire plant of about 1.2%.

Figure 27. 2P SF7 Bifacial Vs. 1P standard tracker cooling. Source: Soltec

Enhanced cooling is caused by three factors. On the one hand, 2P tracker modules are further away from the ground, hence favoring air flow circulation under the tracker. On the other hand, aisles are twice as wide, hence favoring improved air inlet between rows. Lastly, SF7 Bifacial tracker design allows cool air to flow through the gap in the central part of the structure.

These effects can be simulated with PVSyst by including parameters Uc=31 W/m2 k and Uv= 1.6 W/m2 k/m/s.

7.4. Differences analysis

Table 5 proposes a theoretical breakdown of the bifacial gain difference between a SF7 bifacial tracker (BG=15.7%) and a standard 1P tracker (BG=13.6%), measured throughout the year in Livermore (∆=2.1%) for white albedo conditions (albedo=55.6%) and the same GCR=0.4. According to these findings, the main generation differences are caused by lower module operating temperatures of SF7 bifacial trackers (+1.2%) and absence of torque-tube shading (0.7%).

The remaining 0.1% is caused by other factors comprising multiple effects. That means that previously mentioned view factor simulation results (-1%), which indicate less reflected irradiation associated to lower relative height, are actually compensated by higher shaded area reflection (+0.75%). In addition to these differences, it is important to consider the mismatch effect (+0.16%), diffuse irradiation (+0.24%) and radiation penetration through the gap in the central part of the tracker (0.17%).

Variable Factor Respect Total Energy
Rear 

Irradiation

-0.02%

 

Interferences

/ Design

+0.89%

 

Effect 1P SF7 2P Bifacial Parameters
View Factor Affection -1.00% 9.20% 8.20% Height
Shadowed area Reflection 0.74% 0.09% 0.84% Albedo-model
Diffuse 0.24% 1.51% 1.75% Height
Torque Tube Shading 0.56% -0.56% 0.00% Shading Factor
Radiation through Gap 0.17% 0.00% 0.17% Module Transparency
Mismatch 0.16% -0.47% -0.31% Rear G Mismatch
Operation 

+1.2%

Temperature 1.20% -5.10% -3.90% Uv & Uc
TOTAL 2.07%
Table 5. Bifacial gain of SF7 bifacial tracker vs. Standard 1P tracker, differences in % relative to total energy

PVSyst can be used to simulate BiTEC-measured performance by adjusting parameters in the table. These effects are also visible in the daily tracking curve measured at BiTEC. As observed in Figure 28, the difference is minimized in the early hours of the morning when air temperature is lower (zone A), tends to remain constant in the central part of the day (zone B) and increases late in the evening (zone C). This behavior shows that as temperature increases, the link between improved SF7 bifacial tracker cooling and generation gains is clear.

Figure 28. Daily measured power curve of 2P & 1P tracker under the same conditions. Source: BiTEC

8. HOW TO SIMULATE BIFACIAL PROJECTS IN PVSYST

Most photovoltaic simulation software doesn’t model radiation and reflection behavior with the level of detail needed for accurate Bifacial Gain estimation, especially in the case of solar tracking systems. The most popular software, PVSyst®, features a bifacial tracker simulation algorithm that includes electrical modelling for long-term simulations. This program uses a simplified model which assumes that photovoltaic modules are an air-suspended continuous plane. It also uses manually-adjusted parameters to adjust rear irradiation effects, namely: Structure Shading Factor, Shed Transparent Fraction, Field Thermal Loss Factors and Mismatch Loss Factor. These variables directly impact Bifacial Gain, and how they can be simulated on PVSyst® for a correct representation of the SF7 is presented in the table below.

Table 6 presents a summary of the adjustments that PV designers should consider when simulating in PVSyst® with Soltec’s SF7 Bifacial tracker.

Parameters Standard 1P trackers SF7 Bifacial
Max Angle -60º +60º
Height 1.35 meters 2.35 meters
Structure Shading Factor 5.6% 0%
Shed Transparent Fraction MT* (MT + 3.75) x 1.017
Thermal Loss factor Thermal factor (Uc) 29 W/m2 k 31 W/m2 k
Thermal Loss Factor (Uv) 0 W/m2 k/m/s 1.6 W/m2 k/m/s
Mismatch Loss Factor 10 % 3.1 %
Table 6. SF7 Bifacial parameter PVSyst Adjustment. Source: BiTEC
*Module Transparency

9. CONCLUSIONS

After a year of measurements promoted by Soltec at the experimental plant of BiTEC, results confirm that SF7 bifacial trackers generate 15.7% more power under high albedo conditions (55.6%) than monofacial trackers. In the case of vegetation soils, which vary seasonally, and an average albedo measured at 19.9%, the bifacial gain would be 7.3%.

Another goal of BiTEC was to evaluate the most optimal configuration of 1P and 2P (SF7 Bifacial) solar trackers for bifacial modules. In that sense, results remained constant in all cases: SF7 Bifacial trackers generate 2.1% more power than 1P trackers. A detailed analysis confirms that, although view factor models show that 2P trackers capture less energy from ground areas exposed to direct irradiation, as shown by view ray-tracing models, increase contribution of  shadow areas reflected and diffuse radiation, combined with a structure design preventing rear shading and reducing mismatch losses , SF7 bifacial 2P tracker improves the rear radiation usage on +0.87% .

It is also worth mentioning that the SF7 2P bifacial tracker has a lower upper module temperature (6 degrees) as it operates farther away from the ground, where cooling conditions are better. This temperature difference with modules operating in 1P trackers accounts for a power generation increase of 1.2%.

Lastly, as published in other documents, the affection of all these factors (height, temperature, etc.) should be considered by simulation models to obtain accurate generation data from simulations. In this White Paper, Soltec provides the adjustment of recommended parameters to simulate photovoltaic projects with SF7 Bifacial trackers.

10. REFERENCES

[1] Enel Green Power, La Silla: donde la innovación hace que el sol más eficiente, Viernes, 28 abril 2017, online reference
https://www.enelgreenpower.com/es/historias/a/2017/04/la-silla-donde-la-innovacion-hace-que-el-sol-mas-eficiente]

[2] Soltec. Bifacial Trackers, the real deal. BiTEC results – Fall 2018, online reference:
https://lab.soltec.com/bifacial-trackers/

[3] 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

[4] PILAR SÁNCHEZ , Soltec to launch world’s first bifacial tracking evaluation center PV-Magazine MAY 24, 2018.
https://www.pv-magazine.com/2018/05/24/soltec-to-launch-worlds-first-bifacial-tracking-evaluation-center/

[5] J.L. Hatfield, T.J. Sauer, J.H. Prueger,
RADIATION BALANCE,
Editor(s): Daniel Hillel,
Encyclopedia of Soils in the Environment,
Elsevier,
2005,
Pages 355-359,
ISBN 9780123485304,
https://doi.org/10.1016/B0-12-348530-4/00407-0
https://www.sciencedirect.com/science/article/pii/B0123485304004070

[6] Endre Dobos
2005
DOI: 10.1201/NOE0849338304.ch15
University of Miskolc, Miskolc-Egyetemva´ros, Hungary
Encyclopedia of Soil Science
http://www.uni-miskolc.hu/~ecodobos/14334.pdf

[7] Soltec. Albedo: Sun gathered from the soil. BiTEC Results 3 – Fall 2018-Summer 2019. Online reference:

[8] Helmholtz Alfred-Wegener Institut and the National Renewable Energy Laboratory (NREL)

[9] 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.

[10] https://www.solarpowerworldonline.com/2016/07/solar-inverters-clipping-dcac-inverter-load-ratio-ideal/

[11] C. Monokroussos, X. Y. Zhang, M. Li, S. Chen, A. Zhou, S. Liu, J. Zhu, N. Wang, E. Lee, J. Bonilla Castro, W. Herrmann, M. Schweiger, V. Feng, Y. Zhang and C. Zou. ‘Additional Requirements for Supplementary Rating and Qualification of Bifacial PV-Modules’, Conference at 12th International Photovoltaic Power Generation and Smart Energy Exhibition & Conference, SNEC 2018, Shanghai May 29,2018 on-line link:
http://www.snec.org.cn/Download.aspx?dir=SNEC2018-Presentation+Slides%5CS8+PV+Module+Quality+Assurance&file=6-ADDITIONAL+REQUIREMENTS+FOR+SUPPELEMNTARY+RATING+AND+QUALIFICATION+OF+BIFACIAL+PV-MODULES.pdf

[12] Isidoro Martinez, RADIATIVE VIEW FACTORS, Universidad Politénica de Madrid online reference:
http://webserver.dmt.upm.es/~isidoro/tc3/Radiation%20View%20factors.pdf

[13] Soltec. BiTEC: How to simulate bifacial projects? BiTEC Results 2 – Fall 2018-Winter 2019. Online reference:
https://lab.soltec.com/bifacial-trackers-ii/

[14] https://www.radiance-online.org/

[15] Pelaez, S.A., C. Deline, P. Greenberg, J. Stein, and R.K. Kostuk. 2018. “Model and Validation of Single-Axis Tracker with Bifacial Photovoltaics: Preprint.” Golden, CO: National Renewable Energy Laboratory. NREL/CP-5K00-72039.    On-line link:
https://www.nrel.gov/docs/fy19osti/72039.pdf

[16] 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

[18] S. Wang, O. Wilkie, J. Lam, R. Steeman, W. Zhang, K.S ingKhoo, S. Chun Siong, H. Rostan. ‘Bifacial photovoltaic systems energy yield modelling’. 5th International Conference on Silicon Photovoltaics, SiliconPV. March 2015, Konstanz. Elsevier, Sience Direct Energy Procedia 77 p.p 428 – 433, online reference: https://www.sciencedirect.com/science/article/pii/S1876610215008280

[19] Annie Russell et al. Modelling Bifacial Solar Energy Yield for Single-Axis Tracked Systems with Racking .  NUSOD2019,  8-12 July 2019, IEEE
https://www.nusod.org/2019/nusod19paper31.pdf
DOI: 10.1109/NUSOD.2019.8807097

[20] Annie Russell et al. Modelling Non-Uniform Irradiance and Annual Energy Yield for Single-Axis Tracked Bifacial PV Systems with Racking.  BifiPV worshop 2019. Amsterdam 16-19 September 2019. Poster season.

[21] Thomas Capelle et al. BIFACIAL MODELLING A comparison of bifacial PV system modelling tools BifiPV worshop 2019. Amsterdam 16-19 September 2019. Poster season.
http://bifipv-workshop.com/fileadmin/layout/images/bifiPV/presentations2019/bifiPV2019-CEA_INES_Capelle.pdf

[22] Chris Deline et all, BIFACIAL PV SYSTEM MISMATCH LOSS ESTIMATION AND PARAMETERIZATION, PVSEC2019 Marsella September 2019
https://www.nrel.gov/docs/fy19osti/74831.pdf

[23] Ayala Pelaez et al. 2019. The subtle art of bifacial performance modeling, 12th PV Performance Modeling Workshop, Albuquerque, NM.
https://pvpmc.sandia.gov/download/7213/