Module 212: Solar PV for buildings

This module explores existing solar photovoltaic (PV) options and introduces up-and-coming PV technologies

The International Energy Agency (IEA) recently reported1 that global solar photovoltaic (PV) electricity generation increased by a record 179TWh (up 22% in a year) in 2021 to exceed 1,000TWh – the second-largest absolute generation growth of all renewable technologies in 2021, after wind. The IEA contends that solar PV is becoming the lowest-cost option for new electricity generation in much of the world, and this is likely to drive investment in the coming years. This CPD will provide an overview of the core PV technology choices for today and in the near future.

To meet net zero emissions by 2050, the IEA predicts that in the period 2022-2030 global average annual generation growth of 25% is needed. The IEA report1 highlights the impact of smaller PV installations, with residential, commercial, and industrial segments providing almost half of global PV generation, and installations on investors’ own buildings and premises being responsible for almost 30% of total installed PV capacity as of 2021.

Solar PV is typically referred to in terms of cells, modules, panels, and arrays, as illustrated in Figure 4. PV panels and arrays are increasingly seen used as ground-mounted (as in the example of Figure 1), attached to building roofs (and walls) – typically referred to as building attached PV (BAPV) – as in the example of Figure 2, and as building integrated PV (BIPV), as in the example of Figure 3. At the end of September 2022, 55% of UK capacity (7,739MW) came from ground-mounted or standalone solar installations.2

Solar cells contain a material that conducts electricity only when energy is provided by photons from the sun, as direct, diffuse and reflected irradiance, and are specified using terminology as summarised in the panel on page 53.

Modern commercial solar cells used in building applications are typically made of crystalline silicon cells. Silicon makes up3 26% of the Earth’s crust, and using a reduction process, in which the silica is heated with a carbon material, oxygen is removed from molten silicon-bearing quartz aggregates, leaving behind purer, metallurgical-grade silicon. Higher-purity electronic-grade silicon requires further refining, and the resulting molten silicon is then cast or drawn into cylindrical ingots that are sliced into 0.2-0.5mm thick discs or wafers.

Impurities are added during the silicon production process (known as ‘doping’) so that the otherwise non-conducting disc of silicon, which is now termed a ‘crystalline polysilicon’, acts as a semi-conductor (as well as a diode). Phosphorus or arsenic are typically used for n-type doping (introduced on one side of the wafer) and boron or gallium are mixed into the silicon (typically when molten) as p-type doping forming ‘holes’ that allow an electric current to flow through the silicon. Solar irradiance will ionise the atoms in the semiconductor so that electrons jump from their atomic bond and are then free to move through ‘holes’ in the material.

If the top, solar facing, n-type layer is connected by electrical conductors (typically via external circuit) to the rear p-type layer the electrons will travel from the n-type to the p-type layers and then pass through their junction to create an electrical current.

The circular wafers are typically trimmed to a rectangle (with major dimensions approximately 155mm or larger), with the corners cropped off to create an irregular octagon shape. The resulting monocrystalline cells (as shown in Figure 5) are black (or very dark blue) and offer a higher efficiency (recently as high as 23%) compared with the cheaper irregular bluish polycrystalline cells (composed of multiple crystals – currently edging over 20% efficient) that are normally cast into a rectangular block and sliced into rectangular wafers (without further trimming). Wafers are fragile and if handled inappropriately can suffer from microcracks that are invisible to the eye, but which can significantly degrade a cell’s performance.

The application of monocrystalline silicon accounts for the majority of installations, having overtaken the slightly cheaper polycrystalline silicon in recent years. The maximum theoretical efficiency for a silicon solar cell is around 32%.

Traditionally, multiple thin, metallic lines, known as ‘fingers’, are laid down on the front surface of the cell to transfer electrons from the solar cell to the busbars. Wider busbars, typically made of aluminium or silver-plated copper, run perpendicular to the fingers. The rear of a cell semi-conductor also has a printed grid of conductors.

Low resistance wires connect multiple cells to the PV junction box. Optimisations to the collecting conductors – which in some cases remove the need for visible fingers and busbars – are employed to reduce electrical resistance and increase absorption. Anti-reflection coatings and textured cell surfaces are employed to reduce wasted reflection.

A single cell will have an open-circuit voltage – the maximum voltage available at zero current – in the order of 0.45-0.6V, and commonly 36, 60, 72 or 144 cells are connected together in a module to provide a nominal output voltage – for example 12, 24 or 40 volts. Single or multiple modules can then be arranged in parallel or series to form PV panels that are designed to work at a specific range of voltage and current.

The panel is sealed to protect the cells, with the front face covered with a non-reflective transparent material, the back of the panel being sealed to prevent damage and short circuiting, as shown in Figure 6, and the panel mounted into a rigid metallic frame. Panels are connected in arrays (in ‘strings’), and can be series connected to provide higher voltages or parallel connected to increase current – higher currents require larger conductors to reduce losses. (The connections are often ‘intelligently’ controlled to optimise the system performance and reduce the impact of adverse conditions, such as shading of PV panels.)

Passivated emitter and rear contact (PERC) cells have become popular as a low-cost development to monocrystalline cells that increases efficiency by a few per cent, where a ‘passivation’ layer is added to the rear of the cell to reflect photons back into the cell. PERC cells currently have a basic wholesale cost of around £300 per kWp.

Trailing behind crystalline silicon panels in building applications are thin-film solar cells, which can be easier to fabricate but, until recently, are significantly less efficient –thin films make up4 3% to 5% of the global market. Thin-film solar cells are made by coating a thin layer of a highly-absorptive semiconductor material on a sheet of glass, plastic, or metal foil, rather than creating a crystalline wafer.

Flexible materials can be used to apply PV on curved or irregularly shaped surfaces. Thin films are typically dark or partially transparent (at lower efficiencies, as in Figure 3), so the modules look uniform and can replace traditional glazing elements. Some thin-film technologies, such as cadmium telluride (CdTe) copper indium gallium selenide (CIGS) with a maximum theoretical efficiency of 33%, are already being manufactured at efficiencies comparable to traditional crystalline silicon cells.

Higher temperatures cause the semiconductor properties to shift, resulting in a slight increase in current, but a much larger decrease in voltage. BSRIA5 notes that the temperature coefficient of a panel is a percentage of how much more or less energy is produced by the panel per degree above or below 25°C.

For example, if the temperature coefficient of a panel is -0.5%, on a hot day when the panel’s temperature may reach 35°C, the efficiency of the panel will reduce by up to 5%. Both monocrystalline and polycrystalline have temperature coefficients of -0.4% to -0.5%. Thin-film solar panels have temperature coefficients between -0.2% to -0.25%, so thin-film solar panels can be more suitable for locations that experience higher temperatures.

PV panel performance specification

A crystalline panel standardised specification will be dependent on the cell type, cell size, number of cells, layout geometry and resistance of connections. The electrical performance is typically quoted in terms of the following:

  • The peak power output rating, Wp, of a solar panel, is the output under standard test conditions (STC) – that is, cell temperature of 25°C, solar irradiance of 1,000W.m-2.
  • Efficiency – the maximum percentage of sunlight energy that the panel converts into electricity
  • Voltage (Vmp) and current (Imp) at maximum power point
  • Open circuit voltage – the maximum voltage that the panel can provide (Voc)
  • Short circuit current – the maximum current that can be delivered by a panel (Isc)
  • Maximum input voltage
  • Temperature coefficient of power.

Solar panels are typically supplied with a product warranty for basic manufacturing defects of 10+ years, and a power output warranty of 25-30 years. A solar panel degrades over time, the potential electricity production decreasing slowly – typically 2%-3% degradation in year one, and then 0.50% or less per year.6

Output will be impacted by angle and orientation of the panel relative to the sun, shading from nearby objects, and soiling or dust accumulation on the panel surface.

There are several emerging techniques in PV that have the potential to significantly improve the efficiency and performance of solar panels, as described by the US Department of Energy (DoE).4 For example, bifacial solar cells are double-sided to capture light on both sides of a silicon solar module – they capture light reflected off the ground or roof where the panels are installed.

Perovskite solar cells are a relatively new type of thin-film solar cell that have rapidly increasing conversion efficiencies, with some cells already achieving more than 25% conversion efficiency; however, to be commercially viable, perovskite cells have to become more stable and durable enough to survive 20 years outdoors.

Multijunction solar cells have multiple layers, each absorbing a different part of the solar spectrum, making greater use of sunlight than single-junction cells. Light that is not absorbed by the first semiconductor layer is captured by a layer beneath it (and so on, through successive layers). Multijunction solar cells have demonstrated efficiencies higher than 45%, but they are costly and difficult to manufacture.

Organic PV are lightweight solar cells made with carbon compounds that use organic polymers and molecules that conduct and generate electricity in a similar way to those in organic light-emitting diode display technologies. They can be different colours or transparent. Work is continuing to improve lifetime and efficiency and mitigate the visual effects of ageing.

Example costs of commercial systems provided by a UK-based installer7 indicate that commercial installations of rooftop solar panels usually cost approximately £1,500 per kWp (including control equipment). With an approximate solar energy cost of £0.06 per kWh across the life of the installation, compared with grid-supplied electricity (at approximately £0.34 per kWh in the UK), this will provide a simple payback of between eight and 12 years. Since the panels are expected to last approximately 25 years (with one inverter replacement), the return over 20 years is usually 9%-18% (internal rate of return (IRR)%).

A recent report8 by Elementa and Wilmott Dixon estimated the combined embodied carbon and operational carbon for two example UK commercial applications under various scenarios for a period of 25 years, considering complete installations employing monocrystalline or thin-film PV. In all scenarios, even with grid decarbonisation, the operational carbon saving was shown as outweighing the embodied carbon impact over the 25-year life span.

In 2021, the US National Renewable Energy Laboratory (NREL)9 collected together the output of numerous studies and estimated that the life-cycle carbon impact of PV is in the order of 43gCO2e .kWh-1. This compares with 13gCO2e.kWh-1 for wind and nuclear, and 486gCO2e.kWh-1 for gas-powered and 1,001gCO2e.kWh-1 for coal-powered electricity generation.

As noted in the Elementa/Wilmott Dixon report, to head towards net zero embodied carbon, investment will, in any case, be required to achieve grid decarbonisation.

So, across the life of an installation, both the financial and the carbon accounting would appear to stack up (in these UK examples) in support of PV. However, these may not be the only, or even the most pressing, factors in many areas around the globe, particularly in developing countries with limited infrastructure, and in the ‘developed’ world where electricity distribution networks might strain to meet future loads,10 and security and safety of supply is key to short- and medium-term life.

Both BAPV and BIPV installations can play an important part in the diversification and decarbonisation of the world’s electrical supply as a local contribution to the renewable technologies that can contribute to the ambitions of a net zero future.

Further reading:
CIBSE TM25 Understanding building integrated photovoltaics (2000) – a little old, but still useful.
BRE National Solar Centre Solar PV on commercial buildings. A guide for owners and developers (2016).
BSRIA BG 34/2021 The illustrated guide to renewable technologies 2nd Ed, (2021).
Smets, A et al, Solar Energy: The physics and engineering of photovoltaic conversion, technologies and systems, UIT Cambridge, 2016 – freely downloadable Kindle Edition from Amazon.


  1. Solar PV, IEA, Paris 2022, – accessed 9 February 2023.  
  2. Solar photovoltaics deployment in the UK – December 2022 – BEIS, – accessed 9 February 2023.
  3. – accessed 9 February 2023.
  4. – accessed 9 February 2023.
  5. Agha-Hossein, M and Bleicher, D, BG 34/2021 The illustrated guide to renewable technologies 2nd Ed, BSRIA 2021.
  6. – accessed 9 February 2023. 
  7. – accessed 9 February 2023.
  8. Hamot, L et al, Whole life carbon of photovoltaic installations, Elementa/Wilmott Dixon 2022.
  9. Life Cycle Greenhouse Gas Emissions from Electricity Generation: Update, – accessed 9 February 2023.
  10.– accessed 9 February 2023.