Bifacial Photovoltaics Power Plants in UK

Abstract

Bifacial photovoltaics (PV) are able to absorb light from both the front and rear surface, optimising electricity generation, they can be installed both at an angle or vertically and can be used to spread electricity generation throughout the day. This chapter discusses the arrangements which are the most cost-effective in the South-West region of the UK, Optimally Tilted Bifacial (OTB), Optimally Tilted Monofacial (OTM) or Vertically Tilted Bifacial (VTB). The simulations showed that when using the capital (CAPEX) and operating (OPEX) expenditures costs, the OTB arrays were 5–7% more cost effective than the OTM arrays and the VTB arrays were the least cost effective with an LCOE 24% higher than the OTB array. When financial parameters such as interest rates were factored into the calculations, it showed that in certain circumstances monofacial arrays were more or just as cost-effective as the bifacial arrays. This is due to the limited number of bifacial PV arrays throughout the UK, meaning they could be considered a riskier investment. Although VTB arrays were shown to be the least cost-effective, due to their versatility there is still a place for them in future projects. Future trends were also considered where it is predicted that OTB arrays will become more cost-effective than OTM arrays even when financial parameters are considered.

1. Introduction

The photovoltaic effect was first discovered in 1839 by Alexander Becquerel with the first solar cell created in 1883 with a cell efficiency of 1%. Since then, photovoltaics (PV) cells have substantially improved with laboratory-based cell efficiencies reaching up to 42%. In more recent times solar power generation has substantially increased worldwide, in 2005 the global capacity of PV was 2GW whereas at the start of 2022, it surpassed 1 TW. With the rapid expansion of PV industry and ongoing cost reductions it is anticipated that the installed capacity may exceed 10 TW by 2030 and 30–70 TW by 2050, respectively [1].

In a bid to increase cell efficiencies research into bifacial PV cells began in the 1960’s. Bifacial photovoltaics are light sensitive on both sides and can generate electricity by absorbing photons on the front and rear of the panel. This increases efficiency of the cell and improves cost-effectiveness. Uptake of bifacial PV has been slow, in 2017 they accounted for less than 5% of the global market share, yet due to reducing costs this is expected to increase to 40% by 2028 [2] . Although bifacial solar cell and module technologies are mature, they still require research activities to further increase efficiency and production [3].

Bifacial PV can generate a higher electrical output when compared to monofacial PV as they are able to absorb light from both the front and rear surface, however bifacial PV generally require more space to absorb diffuse light and are more expensive. Bifacial PV offer greater flexibility as they can be set up vertically looking east to west or at an ideal tilt angle to efficiently produce electricity. The introduction of vertically tilted bifacial (VTB) arrays will become much more cost-effective as they need less space and can serve two functions. Additionally, because they can produce their peak power output in the mornings and evenings, there will be less need for energy storage, which will reduce energy loss and make them a feasible sustainable option for the future. The adoption of bifacial PV farms in the UK has been slow, with information on their cost-effectiveness being scarce. Bifacial PV farms could be seen as a riskier venture and have a higher weighted average cost of capital (WACC). This would imply that although while bifacial solar farms generate more electricity, monofacial PV arrays might, in some cases, have a cheaper levelized cost of energy (LCOE). As bifacial PV farms become more prevalent in the UK, more data will be gathered on their effectiveness, thus this risk of uncertainty is expected to last only until there is increased confidence in their profitability.

2. Differences between monofacial and bifacial PV

Photovoltaics use the photoelectric effect by converting light into electricity. A photovoltaic cell consists of a layer of toughened glass, often with an anti-reflective coating, a front contact, a semiconductor, a rear contact, and a back plate.

Most semiconductors used in PV are comprised of silicon due to its affordability and natural properties. The atomic structure of silicon has 4 electrons in its outer shell and therefore an insulator in its purest form. Silicon can be doped forming either a p-type or n-type. P-type silicon is produced by adding atoms such as boron or gallium, which has one less outer electron than silicon creating holes for electrons to move into. N-type silicon on the other hand is produced most commonly by doping the silicon with phosphorus which has one more electron in its outer band meaning that an electron is free. When the two types are put together a p-n junction is formed. The free roaming electrons in the n-type fill the holes in the p-type in the area immediately between the junction called the depletion zone. The presence of these oppositely charged ions creates an internal electric field preventing further movement of electrons. When sunlight strikes the cell, enough energy is generated to move the electrons through the semiconductor, generating a flow, this is the main constituent of a PV cell which generates approximately 0.5 V. A PV module normally consists of 60 to 72 cells connected in series and placed in a frame called a PV module. These modules can be connected either in series or in parallel depending on the desired output forming a PV array. The electricity generated by a PV module is in the form of direct current (DC), as households and the national grid require electricity in the form of alternating current (AC), an inverter is required which must be matched to the PV output.

Since the development of passivated emitter and rear cell (PERC) technology there has been increased interest in bifacial photovoltaics. PERC technology introduced a dielectric passivation film to the rear of the cell, this meant that efficiency of the cell was increased as the film reduces electron recombination, increased the amount of absorbed light, and reduced the heat produced in this process. This technology is often used in bifacial cells to increase efficiency. In contrast to monofacial PV, bifacial cells have a transparent back sheet which makes it possible to absorb diffuse sunlight on the rear side of the cell, increasing power output (Figure 1).

Figure 1.

Monofacial and bifacial solar cell construction [4].

There are additional types of bifacial cells such as passivated emitter rear locally diffused (PERL), Passivated emitter rear totally diffused (PERT), interdigited back contact (IBC) and heterojunction with intrinsic thin layer (HIT). However due to the efficiencies achieved and the ability to mass produce, PERC technology forms the majority of the market share of bifacial PV.

Bifacial PV have tempered glass on both the front and rear which increases durability and efficiency. Unlike monofacial PV they are not suited to roof installation as this would not provide the reflectivity required to make them a suitable alternative, instead they should be installed on ground-mounted racks and elevated to allow the rear panel to absorb reflected light. As bifacial PV are able to absorb reflected light the electricity output can be increased by increasing surrounding albedo levels thereby increasing the amount of reflected light. Table 1 showing the differences between monofacial and bifacial PV.

Monofacial	Bifacial
Structure; Glass top-layer; Front contact; Anti-reflective coating; Silicon P-N junction; Aluminium Back Plate	Structure; Glass top layer; Front contact; Anti-reflective coating; Double P-N Junction; Rear Anti-reflective coating; Rear Glass Panel
Higher operating temperature	Lower operating temperature
Greater availability	Higher energy yield
Beneficial for both utility and residential market	Greater benefit to the utility market and ground-based structures
Single sided	Double sided, front, and rear
Output based primarily on direct irradiance	Output based on direct and diffuse irradiance therefore reliant on albedo
Cheaper	3–7% more expensive
Optimal tilt 30°–40°	Usually optimal tilt 10–15° higher
Efficiency 10–25%	Bifacial gain 5–20%

Table 1.

Comparison between monofacial and bifacial PV.

It has been estimated that flat panel bifacial photovoltaics could provide 50% more electrical power than a monofacial equivalent [5], although, this study was undertaken in a laboratory-controlled environment and the practicalities of day-to-day use such as soiling and shading were not considered. An experiment carried out by Sanchez-Friera [6] showed that it is more likely that bifacial PV produce 5–20% more output than monofacial PV also known as the bifacial gain. Chiodetti (2015, cited by Aghaei et al. (2022)) discusses the development of large-scale PV power plants using the bifacial c-Si PV modules including the bifacial PV power plant with a capacity of 1.35 MWp installed in Hokuto city, Japan. The performance data collected from the plant showed a 21.9% gain compared to a monofacial PV plant of a similar capacity [7].

Bifacial solar cells increase the power density of PV modules resulting in the reduction of the PV module and balance of system (BOS) costs and levelized cost of electricity (LCOE) [8].

3. Levelised cost of electricity

The levelized cost of electricity (LCOE) provides a simple measure that can be used to compare the cost-effectiveness of electricity generation techniques. The LCOE demonstrates the relationship between the total cost of the installation and the total energy produced and is calculated in its most basic form using Eq. (1). The LCOE can be used to compare the cost of energy generated of PV with other renewable and non-renewable methods of energy production:

$$\mathit{\text{LCOE}}=\frac{\mathit{\text{Total cost}}}{\mathit{\text{Total energy yield}}}$$E1

If calculating the LCOE over a broad area over a number of countries and continents it is difficult to obtain precise figures as different areas will have different policies, land costs and taxes such as those studies carried out by Sun et al. [9] and Rodriguez-Gallegos et al. [2]. LCOE calculations are undertaken for each specific project as its cost-effectiveness will be dependent on the region the farm is constructed. Table 2 below shows the different factors that need to be considered when calculating the LCOE for a solar farm.

Cost	Electricity generation
Types of PV modules	Geographic Location (albedo, soiling, sunlight)
Balance of System (mounting racks, cabling, inverters)	Module technology (efficiency, temperature behaviour, degradation, bifaciality factor)
Installation Costs	System configuration (mounting, light tracking)
Groundworks/preparation	Number of PV and how arranged
Project development costs
Number of PV purchased
Optimal use of land
Operational and Maintenance (O&M) costs
Land Costs
Subsidies
Financing

Table 2.

Factors that affect the cost of a solar installation.

A complete LCOE assessment will consider that the money spent in the future will have a lower value than the money spent today, this can be factored into the calculation using the net present value (NPV) calculation as shown in Eq. (2):

$$\mathit{NPV}=\frac{\mathit{Ct}}{{\left(1+d\right)}^t}$$E2

Where, t = year of plant lifetime; Ct = total expenditure in year t; and d = discount rate.

Financing of the PV farm will also depend on the amount of debt and equity in the system. Although the interest rates on the money borrowed will depend on the economy of the country, they will also depend on the amount of risk and so project specific. If this is a first project being undertaken by a particular company the lender will consider there to be greater risk. These factors contribute to the WACC. The WACC is often used as the discount rate in the above calculation and can be calculated using Eq. (3).

$$\mathit{\text{WACC}}=\left(e\times \mathit{ie}\right)+\left(d\times \mathit{id}\right)$$E3

Where, e = equity; ie = Interest on equity; d = debt; and id = interest on debt.

PVsyst (references) uses the calculation in Eq. (4) to calculate LCOE and this has been used to calculate the LCOE in this chapter [10].

$$\text{LCOE}=\frac{{\sum }_{d=1}^n\frac{\mathit{It}+\mathit{Mt}}{{\left(1+r\right)}^t}}{{\sum }_{d=1}^n\frac{\mathit{Et}}{{\left(1+r\right)}^t}}$$E4

Where, It = Investment and expenditure for the year (t); Mt. = Operational and maintenance expenditures for the year (t); Et = Electricity production for the year (t); R = discount rate that could be earned in alternative investments; and N = Lifetime of the power plant.

The initial costs of PV have seen considerable decline in the recent years causing a reduction in LCOE, the International Renewable Energy Agency (IRENA) estimate that there was a 13% drop between 2017 and 2018 [11] and between 2010 and 2020 the global LCOE has reduced by 85% [12]. This is supported by The Department for Business, Energy and Industrial Strategy (BEIS) who have published a study into LCOE based on UK PV installations, which predicted the LCOE is to decline further to £44/MWh in 2025 from £68/MWh in 2016 [13] . This pattern has not only been seen in the UK, but globally, the graph below shows the international reduction in LCOE and therefore payback period (Figure 2).

Figure 2.

A graph showing the international reduction in LCOE and payback period for PV [14].

Optimally tilted bifacial (OTB) modules are more cost-effective above 40°, but below monofacial PV are more cost-effective [2]. Cornwall is located at a latitude of 50.26°N and therefore bifacial modules should have a lower LCOE. Optimally tilted monofacial (OTM) solar modules were shown to be more cost-effective when compared to vertical PV, except when the panels are located at the poles, therefore it is expected that the monofacial panels will be more cost-effective in the UK.

It is thought that when the albedo is 0.2 or less, monofacial modules would be more cost-effective, as this would cause the bifacial gain to be lower than 10% reducing electrical output. When the albedo is artificially doubled and the elevation heightened, the bifacial gain was shown to increase up to 30% [9]. Raising the elevation and artificially improving albedo will increase the overall cost causing a change in the LCOE, this shows that there needs to be a trade-off between improving the bifacial gain and increasing costs.

In countries with high land costs it is more cost-effective to orientate bifacial modules horizontally and pack densely. In areas with high module cost, but low land cost, the modules should be optimally tilted. In areas with high land cost and high module cost in locations over 30° latitude bifacial PV should be tilted 10–15° higher reducing LCOE by 2–6% when compared to monofacial PV [15].

4. Solar farms

Solar cells can be connected together in either series or parallel which form solar arrays. These PV arrays can be grouped together to form a power plant which can distribute electricity for commercial, residential and industrial purposes. As of 2022 the largest solar plant in the UK was Shotwick Solar Park in Wales producing 72.2 MW of electricity per year, providing energy to approximately 18,000 homes [16] (Figure 3).

Figure 3.

Shotwick solar farm.

Many solar farms in the UK produce 5 MW as this was the upper limit that feed-in-tariffs (FiT) would be issued. The UK government introduced FiT in 2010 in an effort to increase renewable energy production. The FiT meant the owners of the installations would receive quarterly payments for the electricity produced. However, this scheme was replaced by the smart export guarantee in 2019.

The development of bifacial solar farms has been slow in the UK with the first being created in Warrington in 2019 by Gridserve, which produced 34.7 MW. The slow uptake of bifacial PV in the UK is most likely due to the relative infancy of the technology and a lack of evidence showing their productivity. The aim of this chapter is to model 3 types of solar farm optimally tilted bifacial (OTB), vertically tilted bifacial (VTB) and optimally tilted monofacial (OTM) and to compare the cost effectiveness between each of the three throughout the South-West of the UK. To understand whether it is more cost effective to instal an OTB, VTB or an OTM solar farm in the South-West of the UK simulations were undertaken using PVsyst to determine the lowest LCOE. Six separate locations were assessed to understand the variation throughout the region, these were: Camborne, Truro, Torquay, Bridgwater, Bristol, and Weymouth. To run the simulations a realistic and comparable 5 MW solar farm was designed within PVsyst to establish a realistic LCOE. To calculate the LCOE the following factors were assessed.

1. System Equipment

2. Losses

3. Economic Evaluation

4. Financial Parameters

4.1 System equipment

In the South-West 5 MW solar farms are the most common as they qualify for government subsidies, simulations were therefore carried out using this as the nominal power. The modules that were selected were high-spec monofacial and bifacial silicon monocrystalline JA Solar 390Wp PV, these modules are directly comparable to reduce the variation in the data received. The modules although expensive are good quality with a cell efficiency of 20% and a high cell output. They are readily available in the UK and are durable providing a 25-year warranty which would drive down the costs of repair.

Solar modules generate DC electricity, before it can be used it needs to be converted into AC, therefore an inverter is required. The ABB PVS800-57 1000 kW inverter was selected as these are high efficiency, long-lasting products that are common within the PV industry. To ensure the required capacity is met 5 inverters need to be installed.

To provide the 5 MW capacity and meet the requirements of the inverter the monofacial and bifacial PV are arranged with 19 modules connected in series forming 675 strings, requiring 12,825 modules covering an area of 25,471m².

4.2 Detailed losses

PV systems experience performance losses which can be difficult to model and are often expressed as a percentage loss called the derating factor. The derating factor displays the difference between laboratory test conditions and the real-world, derived from on-site measurements and recorded data.

Some of these factors are embedded as default within PVsyst V7.1, these values may slightly vary from those used by other software such as NREL. PVsyst V7.1 requires the user to assess the losses caused by the following factors, where default values exist those were used (Table 3).

No.	Losses
1	Thermal Parameter
2	Ohmic Losses
3	Module Quality, Light Induced Degradation (LID), Mismatch
4	Soiling Losses
5	IAM losses
6	Auxiliaries
7	Ageing
8	Unavailability
9	Spectral Correction

Table 3.

PVsyst V7.1 system losses.

Within PVsyst V7.1 the thermal behaviour is characterised by the thermal loss factor, without any further information the default data within PVsyst V7.1 was used. For free standing modules that have full air circulation, the thermal loss factor is 29 W/m²K.

In a PV system cables are used to connect solar modules and inverters. Resistive energy losses are proportional to the cable gauge and length, the greater the cable gauge the lower the resistance and the longer the cable the higher the resistance. Losses can be split into DC and AC losses; DC losses are caused between PV modules and the inverter, whereas AC losses occurring between the inverter and the point of injection into the grid. In this study the PVsyst V7.1 default was used setting the DC ohmic losses at 1.5% and the AC losses at 1%.

The module quality loss is a parameter that can be adjusted to express the confidence in the module achieving the performance in the specification. PVsyst V7.1 calculates the module quality loss by dividing the module tolerance by 4. For the JA solar modules, the module quality loss is 0.8%. Light-induced-degradation occurs to the modules in the first few hours of sun exposure and is therefore excluded from calculations. Mismatch losses are caused either when two PV modules or cells do not have the same properties, or by modules that experience different conditions such as shading. Mismatch losses are also caused by the way the modules are interconnected and is affected by the number of series and parallel connections, commonly bypass diodes are used to reduce the impact of mismatch losses and PV degradation. Mismatch losses cannot be completely eradicated, both NREL and PVsyst V7.1 state that typical mismatch losses account for approximately 2% of overall losses.

The amount of dust that accumulates on a panel can, block light absorption reducing electrical output, and, vary depending on the geographical location and PV orientation. As the UK is a temperate environment it is generally less affected by soiling when compared to dusty drier locations. Studies have shown that tilted modules in rainy environments can have annual soiling losses less than 1% [17]. In this study 2% will be used for the tilted modules and 0.5% for the vertical modules to account for dirt build-up around the frame and bird-droppings.

The incidence angle modifier (IAM) relates to the amount of light utilised by the PV cell. When light strikes a cell a portion of light is usually reflected by the glass cover, the amount of light that is reflected increases as the incidence angle increases. At normal incidence, the reflection caused by the glass of the PV is approximately 5%, however this is accounted for in laboratory testing and included in the module specification, the IAM only therefore concerns the angular incidence. PVsyst V7.1 uses Fresnal’s law to calculate the IAM, this default value was used within the calculations.

Auxiliary consumption relates to the losses caused by system management equipment such as inverters, air conditioning and lighting. This data is commonly taken from the manufacturer’s specification. In this instance the ABB PVS800 inverter has been used and the losses have been characterised within the specification.

Overtime the PV modules degrade reducing the overall efficiency and power output, this is defined in PVsyst V7.1 as ageing losses. These ageing losses are often provided by the manufacturer within the manufacturer’s warranty, although it should be noted that this figure is the lower limit for any individual PV, the JA solar modules that are used state a maximum degradation rate of 16% over 25 years, in this study a 12% loss is used.

It is unlikely that the PV system would remain online throughout the entire year and would need to be switched-off for repair and maintenance accordingly, this is unpredictable, however PVsyst V7.1 allows this to be factored into the calculation, PVsyst V7.1 uses a 2% derate factor which will be used in this study.

Spectral correction losses take into consideration the changes in the solar spectrum caused by light scattering and absorption. The spectral correction is set in accordance with the PV technology used, in this case the technology used is silicon mono-crystalline modules, the default value in PVsyst was used.

4.3 CAPEX and OPEX expenditure

To understand how cost-efficient a PV farm is an evaluation of both capital (CAPEX) and operational expenditure (OPEX) is required. Typically, the capital costs are considerably higher than the operational costs and once installed PV farms require limited maintenance when compared to other renewable sources. BEIS Electricity Generation Report [13] states there are two aspects that need to be taken into consideration when identifying CAPEX costs:

1. Predevelopment costs

2. Construction costs

Pre-development costs include pre-licencing, technical design and development, economic analysis, finance, legal costs and stakeholder engagement. BEIS estimate that pre-development costs in 2025 will be £50/kW, this figure provided by BEIS is used in this study. For a 5 MW solar farm pre-development will account for £250,000 of the overall cost.

Construction costs include procurement and installation of PV modules, mounts, inverters and LV and AC equipment. BEIS have estimated the construction cost for a monofacial PV farm will be approximately £400/kW in 2025, totalling £2 m for a 5 MW farm. The land costs for each area have been calculated using data from The Ministry of Housing, Communities and Local Government [18]. For a 5 MW plant 8Ha of land is required.

OPEX costs consider the annual running costs, these are generally lower than the capital costs, but are still significant and can impact the rate in which the PV farm starts to make a return on the initial outlay. OPEX expenditure includes the fixed operational and maintenance costs including preventative and reactive maintenance, consumables, and spares. A recent report undertaken by BEIS has estimated that the fixed O&M cost to be approximately £6700/MW/Yr, insurance costs to be £1900/MW/Yr and connection and system charges to be £1600/MW/Yr. These values have been used in this study as the most accurate for a 5 MW PV farm, totalling £51,000 per year.

PVXchange undertook a market analysis and concluded that bifacial modules are 3–6% more expensive than highly efficient monofacial modules [19], this is supported by a report in PV-magazine that states that bifacial modules are 5–7% more expensive [20]. In this study it has been assumed that that the price of construction for a bifacial farm is 5% more expensive than a monofacial as shown in Table 4. Due to trends which show that the cost of bifacial technology is reducing at a faster rate than monofacial, when undertaking simulations for future trends post-2030 the construction cost difference is reduced to 3% which is the lower end of the PVXchange estimation.

	Monofacial				Bifacial
	2025	2030	2035	2040	2025	2030	2035	2040
Pre-Development (£ × 1000)	250	250	250	250	250	250	250	250
Construction Cost (£ × 1000)	2000	2000	1500	1500	2100	2060	1545	1545
Fixed O&M (£ × 1000 /yr)	33.5	32	30	28.5	33.5	32	30	28.5
Insurance (£x1000/yr)	9.5	9	8.5	8	9.5	9	8.5	8
Connection and System Charges (£x1000/yr)	8	8	8	8	8	8	8	8
Land Costs (£/Ha) (Location Dependent)	21,000–25,000

Table 4.

CAPEX and OPEX costs of a 5 MW solar farm [13, 18].

4.4 Financial parameters

To make a full LCOE assessment, the financial climate was considered. These factors change annually and therefore assumptions were made based on the current economic climate. The LCOE is assessed over a typical 35-year lifespan, to account for the changes in prices of goods overtime it is important to consider the inflation rate in LCOE calculations. The inflation rate in the UK fluctuates depending upon the national and global economy, in 2017 the inflation rate increased to 4% however in recent years it has seen a reduction to 0.8%.

It has been estimated that the inflation rate will average at 2% over that 35-year period. The discount rate also requires consideration, this is the interest rate used to determine the present value of future cash flow, the WACC is often used as the discount rate. The WACC will vary from country to country and will relate to the specific risk levels of the project, so it may be presumed that bifacial modules, a newer technology, are considered higher risk and therefore will incur a greater WACC. Generally, the UK has a moderately stable economy and currently has a credit rating of AA, the WACC in this study is set at 6% when undertaking simulations for both monofacial and bifacial PV. It is assumed that 50% of the project is financed by the purchasing company and the remainder is funded through commercial loans with an interest rate of 4%. The annual tax also needs to be removed from total earnings, in the UK commercial tax is 20% (Table 5).

Financial parameter	Rating (%)
Interest rate	4
WACC	6
Debt vs. equity	50/50
Inflation	2
Tax	20

Table 5.

Financial parameters.

5. Results

This section summarises the results observed in the simulations of the three PV arrangements. The results are split into the four stages to determine the most cost-effective solution whether it be the OTM, OTB or VTB arrays. The first stage shows how the arrangement of the PV modules was established ensuring optimal output. The second stage shows the results obtained for the electrical output of each simulation, the LCOE and results from each of the locations specified in the methodology. The third stage investigates whether adjustments to the ground coverage ratio (GCR), albedo and financing parameters improve the LCOE so the array can be most cost-effective. The final stage uses the predicted future CAPEX and OPEX figures within the BEIS Electricity Generation Report [13] to understand how the LCOE changes in the future and whether this increases the differences between each LCOE.

5.1 Stage 1: Array results

The aim of this stage is to establish the optimal equipment arrangement, to ensure peak LCOE values for each simulation. This was achieved by implementing a generic setup as shown in Table 6. Three aspects of this arrangement were adjusted to determine the effect on electrical output: azimuth, tilt angle and elevation. If the output increased this would confirm that the change was beneficial.

Azimuth	South (0°)
Plane Tilt	40° (discounting vertical orientation)
Pitch	6 m
Elevation	1.5 m
Portrait/Landscape orientation	Portrait
Albedo	0.3
Solar Farm Size	5 MW
Number of Modules	12,825
Module Size	1 m × 2 m
Module Area	25,471 m2

Table 6.

Arrangement characteristics.

The first assessment carried out determined the orientation in which provides the greatest output, simulations were run using the information from Table 6 whilst changing the Azimuth to face each direction. The simulation showed that the OTM and OTB modules should face South, whereas the VTB arrangement should face East/West, this ensures the most direct sunlight is achieved throughout the day. This is as expected, typically, to ensure the highest irradiance on a panel surface, modules in the Northern hemisphere, such as the UK, should face South whereas those in the Southern hemisphere should face North. The VTB show little variation between North, South and West orientation, however as the electrical output was low when facing North, this means that optimum orientation is East/West. This is because they can better utilise the direct irradiance in the morning and evening. Overall, the VTB array produces the lowest electrical output of the three simulations with a peak output of 3806MWh/yr., despite this, they still have practical benefit, as they are able to generate high proportions of electricity in the mornings and evenings spreading the output throughout the day and supplement demand when it is most needed.

The second assessment carried out determined the optimum tilt angle to be used for the OTM and OTB LCOE calculations. The results showed that highest electricity production was at 30° tilt for OTM with an output of 4572MWh/yr., however the optimal tilt angle for the OTB was 35° with an output of 5078MWh/yr. Results from this assessment show the tilt angle for bifacial modules should be 5° higher than the monofacial modules. As the bifacial modules absorb light on their rear surface, they benefit from being installed at a steeper tilt angle as the rear panel will have a higher diffuse view factor. Patel et al. [15] support this argument, however their research claims that bifacial modules should be tilted 10–15° higher as opposed to the 5° suggested by the results in this study. Patel’s research was undertaken to provide a global picture regarding optimal orientation and therefore generalisations may have been made, a lower tilt angle may be required in the South-West as the area still receives a high amount of direct irradiance and a 10°–15° increase may be more beneficial in areas that receive less direct light.

The final simulation was run on the bifacial arrays to determine optimum elevation; in this scenario the tilt angle was set to that established above at 35°. Simulations were not undertaken on the monofacial panels as a limited amount of the electrical output is generated from reflected light. The results in Table 7 showed that the elevation has a greater effect on the OTB than the VTB array, the output of the OTB peaks at 2.5 m with an output of 5085 MWh/yr., after 2.5 m the effect of the elevation plateaus. In contrast the VTB array rises sharply between 0 m and 0.5 m, after the output increases slowly with elevation, however this has a negligible affect overall. In the LCOE simulation the elevation was set to 2.5 m for the optimally tilted module and 1.5 m for the vertically tilted modules. The OTB array showed the greater sensitivity due to elevation, this is to be expected as they need to be located high enough to reduce self-shading, it was found that 2.5 m produced the optimal output for the OTB arrays. The VTB array were less affected by elevation as they were orientated in the East/West direction, so a higher proportion of their output is based on direct light absorption rather than diffuse.

	OTB Array	OTM Array	VTB Array
Azimuth	South	South	East/West
Tilt Angle	35°	30°	90°
Elevation	2.5 m	N/A	1.5 m

Table 7.

Orientation of PV arrays.

5.2 Stage 2: Electrical output and LCOE comparison at different locations

A direct comparison was made between the electrical output and LCOE of a 5 MW OTM, OTB and VTB mounted PV array. The assessment uses CAPEX and OPEX data from the BEIS Electricity Generation Cost Report [13]. Results showing electricity output at Camborne throughout the year are shown in Figure 4 and a location comparison is shown in Figure 5.

Figure 4.

Annual electricity output for Camborne.

Figure 5.

Electrical generated for each location.

Figure 4 shows that the OTB array produces the highest electrical output consistently over the course of the year when compared to the VTB and OTM arrays. The total output of the OTB array is 5114MWh/yr. this is higher than both the OTM and VTB arrays which have an electricity output of 4605 MWh/yr. and 3805 MWh/yr. respectively. If the Smart Export Guarantee (SEG) yearly payments are 5.6p/kWh, the OTB array has the potential to earn £28,000 more than the OTM and £72,000 more than the VTB arrays. The VTB output in June exceeds that of the OTM due to the longer days in the summer, meaning more direct light on the modules in the morning and evenings, outweighing the intensity of the direct light on the OTM array during the day.

Figure 5 graphically compares each location, it is shown that the VTB displays the lowest output whereas OTB the highest with an average difference of 1346MWh/yr. Despite this, the variation between locations is slight for each arrangement, with OTM showing the lowest variation and OTB modules the highest, this can be attributed to the variation in diffuse light between certain areas.

The financial parameters of each project were initially considered to be identical when the LCOE simulations were undertaken as shown in Table 5, this was to ensure a direct comparison could be made between the bifacial and monofacial arrays ignoring the impact of project financing. The figures for 2025 from the BEIS Electricity Generation Costs Report [13] were used to establish the cost of the array as shown in Table 4. The LCOE results are shown in Table 8.

	Camborne	Bridgwater	Bristol	Truro	Torquay	Weymouth	Average
Electricity Output MWh/Yr (OTB)	5114	5359	4795	5277	5244	5111	5150
LCOE £/MWh (OTB)	42	40	45	41	41	42	41.8
Electricity Output MWh/Yr (OTM)	4572	4786	4379	4723	4677	4501	4606
LCOE £/MWh (OTM)	45	43	47	44	44	45	44.6
Electricity Output MWh/Yr (VTB)	3806	3813	3508	3909	4016	3778	3804
LCOE £/MWh (VTB)	54	53	60	53	52	55	54.8

Table 8.

Stage 2 summary of results.

Table 8 shows that in the South-West the OTB array produces approximately 10% more electricity than the OTM array when the albedo is 0.3. Sun et al. [9] concluded that bifacial gain should be no more than 10% across the globe when there is an albedo of 0.25, this corresponds with the results in this study. Due to the high levels bifacial gain calculated it can be concluded that the bifacial PV are highly beneficial in this region. When calculating LCOE the BEIS Electricity Generation Costs [13] were used, and the construction costs of the bifacial arrays were considered to cost 5% more than the monofacial array. Results show that OTB arrays in the South-West are 5–7% more cost-effective than the OTM arrays with an average of £41.8/MWh compared to £44.6/MWh. This confirms the conclusions drawn by Rodriguez-Gallegos that angled bifacial arrays are more cost-effective than monofacial angled arrays at latitudes above 40° [2]. The simulations that were run in this study, however, were slightly more cost-effective than those predicted in Rodriguez-Gallegos although this is to be expected as the price of solar panels have reduced since 2018.

The conclusions drawn by Rodriguez-Gallegos [2] also concluded that the VTB arrays do not provide a cost-effective alternative to either the OTM or OTB arrays. The vertical modules in this study were approximately 23–28% less cost-effective than the OTB arrays and 17–22% less cost-effective than the OTM arrays. The results of the simulation agree with predicted LCOE from the BEIS Electricity Generation Costs Report [13] which projects that the LCOE will average to be £44/MWh in 2025 and shows that energy from PV is more cost-effective when compared to renewable alternatives with the LCOE for offshore and onshore wind estimated to cost £57/MWh and £46/MWh respectively and CCGT will cost considerably more at £85/MWh.

The simulations between each area displays little variation and shows that the bifacial arrays are the most cost-effective throughout the region. The locations on the North coast show variable cost-efficiency with Bridgwater the most cost effective of all six locations and Bristol the least, ranging between £40/MWh-£45/MWh. In contrast, the LCOE is more consistent on the South coast varying between £41/MWh-£42/MWh. The difference in cost-effectiveness varies depending on the amount of irradiance, Table 9 shows the comparison between global horizontal irradiance (GHI), HDI (diffuse horizontal irradiance) and LCOE for each simulation.

	Camborne	Bridgwater	Bristol	Truro	Torquay	Weymouth
OTB LCOE (£/MWh)	42	40	45	41	41	42
OTM LCOE (£/MWh)	45	43	47	44	44	45
VTB LCOE (£/MWh)	55	53	60	53	52	55
GHI (kWh/m2/yr)	1071.8	1132.6	984.6	1103.5	1108.4	1055.4
HDI (kWh/m2/yr)	604.7	591.1	567.8	599.6	563.1	565.5

Table 9.

Correlation between irradiance and LCOE.

Results show a direct correlation between LCOE and GHI, as the GHI increases the LCOE decreases showing improved cost-effectiveness, this is to be expected, as the more light is available the more electricity can be produced. There is less of a correlation between HDI and LCOE, it can therefore be concluded that the direct irradiance has the overriding impact on LCOE and the diffuse light supplements electricity generation. Another factor that may contribute to higher LCOE in Bristol is the higher land costs amounting to £25,000/Ha whereas those further in the South-West were £23,000/Ha, these are small variations and so will only have a slight impact on LCOE when all considered.

5.3 Stage 3: Sensitivity analysis

5.3.1 Ground coverage ratio

This section investigates the changes made to the pitch and how this influences LCOE. The simulations have been undertaken at Camborne assuming that an area of 8Ha is available. Reducing pitch will mean there will be an increase in PV modules so increasing the potential of electrical output, however self-shading will also increase reducing direct and diffuse irradiance. The longer the pitch the greater the potential to absorb diffuse light, however there will be fewer PV modules. The GCR simulations vary depending on orientation of the PV module; whether in portrait or landscape and are also bespoke to the size of the PV module used. In this instance calculations have been made with the modules in the portrait orientation. The simulation was carried out assessing the pitch between 1–8 m increasing the increments by 1 m each time, the results are shown in Table 10 and Figure 6.

Pitch (m)	Nominal PV power (MWp)	OTM LCOE (£/MWh)	OTB LCOE (£/MWh)	VTB LCOE (£/MWh)
1	31	109	111	154
2	15.5	52	56	83
3	10.3	47	45	67
4	7.8	44	44	61
5	6.2	43	43	56
6	5	45	42	55
7	4.45	51	45	53
8	3.9	53	47	52

Table 10.

GCR comparison.

Figure 6.

LCOE vs. pitch.

Table 10 shows that the OTM array is most cost-efficient when there is a 5 m pitch, in contrast the OTB requires a longer pitch of 6 m. Monofacial PV can perform better at a shorter pitch as they rely less on reflected irradiance and as the modules are cheaper it is more affordable to instal more monofacial than bifacial modules. The bifacial PV require more space to reduce shading from adjacent modules and so a longer pitch is required. Despite the OTM array requiring a shorter pitch which means more PV modules can be installed, the OTB still has a lower LCOE of £42/MWh compared to £43/MWh. The LCOE of the VTB array reduces with pitch which continues beyond 8 m, simulations were not carried out to determine an optimum level and this could be an area for further research. The increased pitch reduces LCOE as there is less shading and the modules are able to absorb more direct and diffuse light increasing electricity production. The OTM array has an optimal LCOE at 5 m, however the nominal output of the array is 6.2MWp, the SEG agreement is only permitted for an output of up to 5 MW and so any additional electricity created would not be available for the scheme and therefore sold for a lower rate meaning it will have a longer payback period.

5.3.2 Effect of albedo

The albedo can affect the electrical output of bifacial PV modules as the higher the reflectivity of a surface the more diffuse irradiance is available. Using a highly reflective surface immediately beneath the PV modules can therefore increase electrical output. These simulations investigate how artificially improving the ground reflectivity will affect LCOE bearing in mind that changes to the surface will be costly. The calculations so far have used the albedo of grass as 0.3, further calculations have been carried out using ground surfaces of white paint, concrete, and aluminium. Results in Figure 7 show how changing the albedo can affect electricity output.

Figure 7.

A graph showing electricity output vs. albedo.

In all scenarios the OTB array produced a greater output than the VTB array, however for both OTB and VTB arrays increasing albedo increased electrical output. Aluminium and white paint showed the greatest increase of 771MWh/yr. and 586MWh/yr. respectively for the OTB array, and an increase of 1159MWh/yr. and 908MWh/yr. for the VTB array. This shows that improving the albedo has a greater impact on the VTB modules with a 30% increase for aluminium compared to a 15% increase for the OTB array. As a high value of albedo was used for grass, to understand what the worst-case production would be simulations were undertaken with grass using an albedo of 0.15. Due to the changing albedo of grass an output between 4882 and 5114MWh/yr. could be achieved for the OTB panels, in contrast a range of 3501–3806MWh/yr. for the VTB array.

Despite an increase in electrical output, the cost of the ground works to increase albedo can mean the changes are less cost-effective. A simulation was carried out using the electricity output generated in Figure 7 and comparing these values against estimated costs using PVsyst V7.1, the results are shown below (Table 11). It should be noted that the ground being built upon initially is assumed to be grass and therefore this comes at no additional cost. Approximate prices were calculated using quotations from UK companies, these prices are only for the raw materials and do not include installation costs.

Material	Albedo	Additional Cost (£)	OTB 35° Output (MWh/yr)	VTB 90° Output (MWh/yr)	OTB 35° LCOE (£/MWh)	VTB 90° LCOE (£/MWh)
Grass (PVsyst)	0.15	0	4882	3501	44	57
Fresh Grass	0.3	0	5114	3806	42	55
White Paint	0.75	690,000	5700	4714	44	54
Concrete	0.35	2,040,000	5196	3884	64	87
Aluminium	0.85	2,720,000	5885	4965	65	79

Table 11.

The effect of albedo on electrical output and LCOE.

Figure 8 shows that despite the increase in electrical output caused by increasing albedo, the LCOE of grass is still most cost-effective even when the worst-case is considered, however this is on the proviso that the array is built on grassland originally. An aluminium surface showed the highest electrical output however the cost of installation would mean that the LCOE would rise by £23/MWh and £25/MWh for the OTB and VTB arrays, respectively. By installing the concrete base for the OTB and VTB array the LCOE increases to £64/MWh-£87/MWh which is the least cost-effective solution for the VTB, whereas the aluminium surface is the least cost-effective for the OTB. In contrast of the changes made, the white paint showed the most cost-effective alternative to grass and can be more cost-effective when grass has a lower albedo. However, the price estimations have only considered the cost of the reflective paint and has not considered the price of the surface in which to paint. The paint will need to be applied to a surface which will increase cost, further to this the paint may need to be reapplied regularly to ensure it maintains its reflective properties. This is not a practical alternative surface for a solar farm and would be better employed on roof-mounted PV.

Figure 8.

A graph showing LCOE vs. albedo.

It can therefore be concluded that if building a PV array on an agricultural site it is more cost-effective to maintain grass rather than undertaking groundworks to increase surface reflectivity, this also provides the advantage of being able to use the field for pastoral farming allowing animals to graze around the PV array keeping the grass short at no additional expense.

5.3.3 Financial parameters

Comparisons of the arrays have been undertaken ignoring financial parameters that may affect cost-effectiveness such as interest rates and the practicalities of funding the initial outlay. As bifacial farms are relatively rare in the UK and the initial outlay will be greater than monofacial farms, a higher amount of borrowing may be needed, it may also be required that the project has more equity. It is possible that those who have invested in the project will want a higher return on their investment which is likely to increase the WACC. New technology inherently has an increased risk as there is there is limited data to determine whether it works and whether it is likely to earn a profit, bifacial modules lack this track data and so there will be a higher perceived risk. Calculations were undertaken to determine how this increase in perceived risk will affect affordability by adjusting the WACC. Table 12 shows a comparison between LCOE and WACC. In this simulation it was assumed that the project was 50% financed from own funds and a 50% loan was provided at an interest rate of 4%.

WACC (%)	OTM LCOE (£/MWh)	OTB LCOE (£/MWh)	VTB LCOE (£/MWh)
4	43	39	50
4.5	43	40	52
5	44	40	53
5.5	45	41	54
6	45	42	55
6.5	46	43	56
7	47	43	58
7.5	48	44	59
8	49	45	60

Table 12.

WACC analysis.

Table 12 shows that if the WACC is reduced the LCOE is also reduced improving cost-effectiveness of the array. As monofacial technology is more established it is likely that the WACC will be lower in the range of 4–6%, with an LCOE ranging between £43/MWh-£45/MWh, in contrast as the bifacial technology is newer the WACC is likely to be higher between 6 and 8% with an LCOE ranging between £42/MWh-£45/MWh, this could mean that the bifacial and monofacial technologies almost produce an identical cost of electricity and in some cases an OTM array would be more cost-effective. However, in both instances the cost of electricity will be more cost-effective than installing a VTB array. A VTB array may even be considered a greater risk than the OTB, therefore incurring a greater WACC and may be more expensive still. This shows that if the appropriate subsidies are applied to bifacial farms it will benefit the UK economy and strive towards a more sustainable means of energy production as bifacial farms do produce more electricity per unit than monofacial farms.

As there is a greater risk associated with installing a bifacial PV array a higher deposit maybe required which will affect the LCOE. In the scenario below, the WACC for a bifacial array is set at 8%, whereas the WACC for a monofacial farm 6%. A lender may request that the monofacial solar farm is funded 50% by the owner and 50% by the bank at an interest rate of 4%, in contrast the bifacial farm maybe seen as greater risk and the lender may require the project to be funded 75% by the owner which may mean external funding needs to be sourced. An assessment of how the funding can affect cost-effectiveness is demonstrated within Figure 9.

Figure 9.

Comparison between loan amount and LCOE.

Under the circumstances stated above if the installer is required to provide over 50% equity in the PV array it is more cost-effective to instal an OTM array, whereas if a lender is willing to provide more than 50% of debt the OTB will be more cost-effective. However, if the project is funded by a large corporation who can afford the initial outlay without external investment it would be beneficial to instal the OTB. As can be seen, OTB arrays are more cost-effective when considering electricity output and costs alone, the financial agreements put into place to fund the project can strongly influence the overall cost-effectiveness. It is possible that in the future as bifacial farms become more common place funding for monofacial and bifacial arrays will be on a level playing field and there will be a clearer distinction between the most cost-effective solution.

5.4 Stage 4: Future trends

The CAPEX costs of solar farms are reducing annually as is the cost difference between monofacial and bifacial modules. Using the prospective monofacial solar costs within the BEIS Electricity Generation Report [13] and reducing the cost comparison between monofacial and bifacial equipment to 3%, a future LCOE analysis was carried out. Calculations made in stages 2 and 3 of this study used a 5% mark up in construction costs of a bifacial array, however some commentators such as PVXChange put the difference in cost around 3% which would indicate that the figures in this study are conservative and the LCOE for bifacial panels will be lower and therefore more cost-effective still. BEIS [13] shows that between 2025 and 2030 the CAPEX costs of the construction stage will see no change, however the OPEX costs are expected to reduce slightly. Between 2030 and 2035 BEIS expect that the greatest reduction in CAPEX costs will be seen. Within the analysis the assumption was made that the bifacial technology has become more commonplace in line with the International Technology Roadmap for Photovoltaic [21] report which states that by 2028 bifacial modules will form 40% of the PV market. This would therefore reduce the WACC to a similar level to monofacial farms at 6% and it has been assumed that the funding requirements are the same at 50% equity and 50% debt.

Table 13 presents simulation results which show that by 2030 the LCOE of bifacial modules could be at £41/MWh, in contrast the LCOE of the OTM array would remain higher at £45/MWh making the bifacial farm more cost-effective. The VTB array will see a slight reduction in overall cost; however, they are still unable to compete with OTM and OTB arrays. Table 13 shows that by 2035 it is expected that the LCOE of the optimally tilted modules will break the £40/MWh mark in the South-West, showing a significant reduction in LCOE, the reason for this drop is due to perceived enhancements in technology with BEIS predicting a reduction in construction costs of the plant by 25% in this period. Overall, BEIS predict a 21% reduction in CAPEX costs for a solar farm between 2025 and 2040 and a 15% reduction in OPEX costs throughout this period. These calculations may be conservative yet as IRENA have shown that between 2010 and 2018 the LCOE in the UK has declined 77% and this predicted to further reduce by 50% by 2050 [22]. It can be concluded that by 2030 following current predictions OTB arrays will be the most cost-effective arrangement taking in consideration financial parameters, in contrast the VTB array will be the least.

	2030	2035	2040
OTB	£41/MWh	£34/MWh	£34/MWh
OTM	£45/MWh	£38/MWh	£37/MWh
VTB	£55/MWh	£46/MWh	£45/MWh

Table 13.

Future PV farm LCOE trends.

6. Conclusions

There are multiple factors that can affect the efficiency and cost of a solar farm. Calculations showed that OTM and OTB arrays generated the highest electrical production when facing South, in contrast VTB arrays generated the highest electrical output when facing East/West.

Simulations were carried out at each of the following locations in the South-West of the UK; Camborne, Bridgwater, Bristol, Truro, Torquay, and Weymouth, the first three on the North coast and latter three on the South. The LCOE simulations were carried out in each of the three orientations: OTB, OTM and VTB. Of these locations the results show that the OTB array was 5–7% more cost-effective than the OTM. Conversely, the VTB arrays produced the lowest electrical output and LCOE which ranged between £53–£60/MWh.

Monofacial PV farms are commonplace in the UK and their effectiveness and profitability are supported by data, this does not exist for bifacial PV farms and therefore they can be considered a higher risk to the investor. The project may require an increased level of equity before the remainder of the funding can be provided and they may incur a higher weighted average cost of capital (WACC). Monofacial farms are likely to have a WACC ranging between 4 and 6% which produces an LCOE of £43/MWh-£45 MWh in contrast bifacial farms may have a higher WACC ranging between 6 and 8% which provides an LCOE of £42MWh-£45/MWh. As there is an overlap it could therefore be the case that under certain financial conditions monofacial PV farms are more cost-effective when compared to OTB.

Using the BEIS future projected energy costs, simulations were carried out to determine the cost of PV farms in 2030, 2035 and 2040. In all cases the cost of PV farms reduced, OTB PV arrays continue to be more cost-effective into 2040 showing an LCOE of £34/MWh, with monofacial LCOE slightly higher £37/MWh and vertical LCOE at £45/MWh. By this stage it is expected that OTB arrays will be more commonplace according to ITRPV predictions which state that by 2028 40% of installations will be bifacial, this provides more confidence that funding will be easier to receive, and capital expenditure (CAPEX) costs will reduce.

It can therefore be concluded that in the South-West currently despite bifacial solar farms showing improved LCOE their cost-effectiveness can be influenced by the way in which they are initially financed and in certain circumstances monofacial PV arrays are just as affordable if not more. However, this trend is unlikely to continue and as more data is collated on bifacial PV and they gain improved investor confidence it is predicted that bifacial solar farms will become more cost-effective in the future. Vertical solar farms have shown to be the least cost-effective and should not be used in this type of arrangement, this is not to say vertical PV are not cost-efficient when used under different circumstances and when used in conjunction with bifacial modules to spread the energy production throughout the day.