Whole life carbon of photovoltaic installations

TECHNICAL REPORT - FEBRUARY 2022

Abstract
This study investigates the embodied and operational Contents
carbon impact of roof mounted solar photovoltaic
(PV) installations. It includes an in-depth assessment
of the embodied carbon of PV systems and frames a
Abstract2
discussion on the most relevant metric for making
Introduction4
decisions on the installation of PV.
Photovoltaic (PV) is an important source of rooftop solar PV should continue to play a vital role
Objectives6
renewable energy generation, and rooftop solar in supporting this transition. Nevertheless, the study
can contribute a significant part of the government found that the embodied carbon to be between 520 –
Scope of the Study 8
roadmap to meet UK climate targets. This paper 780 kgCO2e/ kWp, which is less than the operational
explores the interplay between embodied carbon carbon savings over a 25-year study period based on
Methodology and Assumptions 10
impact and the operational carbon savings for a UK grid.
various roof mounted PV system installations. Results14
The study concludes that rooftop solar PV can
This study concludes that a ‘payback’ approach for PV
installations (i.e. embodied carbon impact compensated represent a valuable investment of embodied Discussion18
carbon, but that built environment professionals
by the operational carbon savings) is not the right
metric while we are in midst of an energy transition should employ detailed embodied carbon Conclusions22
assessments to ensure that impacts are minimised
towards 100% renewables. Embodied carbon will have
through intelligent design and specification
to be invested to achieve full grid decarbonisation and
decisions on a project-by-project basis.
Report Authors

Louise Hamot - Lead Author Doug Drewinak Tom Burgess

Global Lead of Sustainable Innovation Building Performance Manager Mechanical Engineer
Elementa Consulting Willmott Dixon Elementa Consulting

Report Contributors

Alasdair Donn Clara Bagenal George Hugh Dugdale Chris Worboys

Head of Building Performance Associate Associate Principal Senior Sustainability Consultant
Willmott Dixon Elementa Consulting Elementa Consulting Etude

Whole life carbon of photovoltaic installations 2

Introduction
Why does it all matter? There is an initial embodied carbon impact associated with producing and installing the entire system onsite
(modules, mounting system, associated equipment, etc.). This is shown by the tall blue column in Figure 1.
As the latest IPCC report1 reminds us, 2020-2030 is a very decisive decade, our last The PV panels produce renewable electricity and for every kWh generated, it is assumed that the grid does not
chance at preventing climate breakdown and run-away climate change. All parts of the need to produce this kWh of electricity, thus the ‘avoided’ carbon emissions are thought of as an operational
carbon saving. The ‘dirtier’ the electricity grid, the greater the amount of carbon is avoided. This also means
economy will have to transition to net zero with urgency and at pace for us to stabilise that as the electricity grid decarbonises, the avoided emissions are smaller, and thus the energy generated by
the climate and limit global warming below 1.5 degrees C. PV panels has less and less of a carbon-reducing impact (or offset mechanism).
Elementa Consulting, forms part of an international network of engineers and consultants collaborating There is a period of repair and replacement (e.g., the inverter), shown at 12.5 years, which also has an
under a single “deep green” umbrella. We enable every client to protect the health of our planet, by taking a associated embodied carbon impact. At end of life of the PV system - the decommissioning, waste processing
regenerative approach to the design, performance and function of buildings, communities, districts, and cities. and recycling also have an embodied carbon impact.
We provide a full range of building and district systems engineering, analysis, and sustainability consulting
services, delivered by staff widely regarded as innovative leaders in their fields. Our work spans the globe,
CarbonCarbon
balance of a PV
balance ofinstallation in the UK
PV installation
delivered from offices in Australia, Canada, Europe, and the United States. Our projects are located in over
30 countries – with more than 100 net zero energy buildings. Elementa Consulting are proud to be founding

(Tonnes CO2e)
signatories of the World Green Building Council’s Net Zero Carbon Buildings Commitment, signatories of UK

(TonnesCO2e)
Engineers Declare, and primary authors of CIBSE TM65 Embodied carbon of building services: a calculation
methodology and CIBSE TM65.1 Embodied carbon of building services: residential heating.
Willmott Dixon has responded to the climate emergency with a sector-leading sustainable development

GWPGWP
strategy, Now or Never. Our decisive decade, which sets out how the business will commit to becoming net
zero carbon by 2030. Furthermore, by 2030, all of Willmott Dixon’s projects, where they have early-stage
design responsibility, will be delivered to net zero operational carbon standards. As part of Now or Never,

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

2034

2035

2036

2037

2038

2039

2040

2041

2042

2043

2044

2045
the contractor is assembling a suite of technical solutions they intend to test using whole life carbon and
cost modelling approaches, to be able to offer a net zero carbon option on all projects going forward. One of
those solutions is Willmott Dixon’s community solar energy offer, that provides PV infrastructure, capital free years

through a power purchase agreement partnering with a community energy organisation.

What is this research trying to address?

Solar photovoltaic (PV) panels play a central role in decarbonising our grid. PV panels are becoming
a ubiquitous solution to increase on-site renewable energy generation, on both new build and major
refurbishment projects, to meet net zero operational carbon goals. Capital costs for PV systems have also Embodied carbon Operational carbon savings Total
decreased significantly in recent years due to the economies of scale manufacturing for large grid-scale solar (A1-A5, B4, C2-C4) (with grid decarbonisation)
installations. The Renewable Energy Hub reports that in 2021 enough energy was generated by PV across the
Figure 1 – Whole life carbon balance of a typical PV installation. The operational carbon factor used include a decarbonisation scenario (see p.11).
globe to power over 30 million households2 . However, like any other product, PV installations come with an
embodied carbon impact: greenhouse gas emissions associated with production; construction; in use and end
of life stages. Moreover, PV panels require accessory equipment such as support, cabling, and inverters, which
also have an embodied carbon impact.
Rooftop solar PV is required to achieve a decarbonised grid, therefore the embodied carbon of PV needs to be Definition of Terms
better understood. However, we often find that we don’t yet have all the data available to make decisions. This
Carbon definitions and lifecycle stages - This study aligns with the WLCN, LETI, RIBA definition of carbon published
study is a first attempt to provide further insight.
in May 20213 and EN 15978 framework around lifecycle stages:
We encourage others to further this work with larger datasets and look to manufacturers of these systems,
• Embodied carbon: greenhouse gas emissions associated with A1-A5 (product & construction), B1-B5 (use), C1-C4
to collect better data and make it more freely available to built environment professionals. We also suggest
(end of life).
similar studies are carried out for other renewable energy generation technologies (e.g., wind, solar thermal).
• Operational carbon: greenhouse gas emissions associated with energy use (B6) and water use (B7) during the
Carbon balance of PV installation over life span building lifetime. This study only explores carbon savings within B6 associated with PV electricity generation.

Figure 1 shows an example of the total carbon balance of a typical PV installation over an assumed life span of • Whole life carbon: greenhouse gas emissions associated with embodied carbon, operational carbon and any
25 years. The carbon impact of a PV installation over its lifetime can be expressed as the cumulative sum of its benefits or loads associated with reuse, recycling and recovery (module D). However, Module D is not considered
associated embodied carbon and its operation carbon savings, taking into account repair/replacement and end within this study.
of life decommissioning.
Environmental Product Declaration (EPD): independently verified and registered document that communicates
1 IPPC Report from Working Group I issued in August 2021 AR6 Synthesis Report: Climate Change 2022 — IPCC transparent and comparable information about the life cycle environmental impact of a product (CIBSE, 2020).

2 https://www.renewableenergyhub.co.uk/main/solar-panels/solar-panels-carbon-analysis/ CIBSE TM65 4: embodied carbon calculation methodology for building services equipment published by CIBSE in
2020, authored by Elementa Consulting, to be used when no EPDs are available. Two levels of calculations are possible
3 https://www.leti.london/_files/ugd/252d09_879cb72cebea4587aa860b05e187a32a.pdf
depending on manufacturer data collected: basic and mid-level.
4 https://www.cibse.org/knowledge/knowledge-items/detail?id=a0q3Y00000IPZOhQAP

Whole life carbon of photovoltaic installations 4

5
Objectives
The use of solar photovoltaic (PV) for
onsite renewable energy generation is now
a significant factor in the vast majority
of new projects to help achieve net zero
operational carbon objectives for new-
build or major refurbishment projects,
in line with LETI guidance and the UKGBC
Advancing Net Zero Framework.
While solar PV is a proven and highly reliable
means of renewable energy generation,
it has, along with all MEP products and
construction materials, an embodied carbon
impact associated with its manufacture,
supply, maintenance and end of life.

This research aims to understand the embodied and operational carbon

impact of rooftop solar PV. It explores different installation types across
4 different scenarios.
As building embodied carbon targets are further refined and developed
(e.g., RIBA 2030 challenge and LETI targets5), the question of whether to
include the embodied impacts of PV installations within the embodied
carbon building target becomes important. PVs are needed therefore we
need to understand their embodied carbon implications and identify any
opportunities to reduce its impact.

5 Embodied Carbon Target Alignment.

Right: Boulder Commons Campus

The large PV[ array for this project is sized to offset all of the energy that the campus uses over the
course of a year, in order to achieve the Zero Net Energy goal. Elementa/Integral Group project.

Whole life carbon of photovoltaic installations 6

7
Scope of the Study
Case studies efficiencies (15-24% compared to 13-20% in polycrystalline cells). The efficiency gain in Monocrystalline systems
can result in generating greater power and annual energy output from the same roof area. Most modules
To investigate this topic, calculations were based on two actual UK projects currently in their preconstruction
employing crystalline silicon wafers require an aluminium frame for structural robustness (or glass backing
phase (both set to complete in 2023). Project specific data on equipment specifications, quantities and
for frameless modules), which add significant embodied carbon. This is in addition to the mounting structure
energy generation estimates were made available via Willmott Dixon’s supply chain partners using industry-
which is often an aluminium frame system to accommodate a variety of installations and roof types.
recognised and accredited design software (see Table 1).
Thin-film PV module: A PV product made up of thin-film semiconductors, often about 20 times thinner than
Building kWh annual PV PV panel area Roof type / PV more traditional crystalline silicon wafers, deposited on glass, plastic or metal. The reduction in thickness
Project Location kWpeak makes thin-film solar panels more flexible and lightweight, when encased in plastic these can become flexible
Type generation m2 orientation
enough to mould to roof shapes. There are three common thin-film divisions: amorphous silicon, cadmium
A School Wiltshire 417,265 kWh 437.6 2,103 Flat - various telluride and copper indium gallium selenide. Efficiencies in thin-film products are typically lower than more
traditional panel systems (<17% for thin-film vs >21% for monocrystalline), so more roof area might be required
Pitched – East/ to achieve the same power and annual energy output. However, the product is much lighter, more adaptable
B Theatre Bristol 114,076 kWh 125.3 634
West and requires less fixings or support. Potential downsides are about recycling the modules, where material
recovery is more difficult due to the bonding process. Furthermore, the technology does not work on some
Table 1 – Case study projects key characteristics roof coverings (e.g., roof tiles).

Equipment included
Most existing studies6 looking at the embodied carbon impact of PV typically only include the modules
themselves. However, PV panels require accessory equipment: support, wiring, inverters and quite often
optimisers (to maximise the efficiency of electricity generation). This study investigates the entire installation,
and thus the scope (see Table 2):

Category Equipment Project A Project B

PV PV panels  
Optimiser Optimisers  
Inverter Inverters   Figure 2 - Monocrystalline PV installation on flat roof Figure 3 - Thin-film PV installation on standing seam metal roof
Image source
Mounting rails   Image source

Supports Module Clamps   Scenarios considered

Supports   PV optimisers can offer benefits in terms of increased output for partially shaded systems, and improved
system control/data acquisition, but are not always included within PV installations. We decided to investigate
Ballast Concrete tiles  N/A
two further system options (with and without optimisers) in this study to understand the effects on the overall
Generation meter   carbon balance.
The thin-film technology was only tested on project B, due to the roof type and covering on project A not being
Ballast   suitable for this system.
Other Electronics
DC wiring   Table 3 below summarises the four scenarios analysed in this study.
Local AC isolator   Scenario Project PV type Optimisers Support type Notes
Table 2 – Equipment scope considered as part of the study
Ballast + mounting
Scenario 1 A Monocrystalline Included
Ancillary equipment related to roof access and maintenance were not included in the calculations scope, as this frames
was required for general roof construction/maintenance.
Roof clamps + mounting
Scenario 2 B Monocrystalline Included
Different types of PV panels (known also as modules) exist. It was decided to explore the research question rails
with two different module types, as quoted for by our supply chain partners on the projects:
Different higher efficiency
Roof clamps + mounting
Monocrystalline PV module: Monocrystalline cells are solar cells made from silicon crystallised into a Scenario 3 B Monocrystalline Not included
rails
inverter type used than in
single (i.e. mono) crystal. The manufacturing process is more complicated than other technologies (e.g., scenario 2
polycrystalline silicon cells) but achieves fewer defects or impurities which generally results in higher
Reduced PV electricity
Scenario 4 B Thin-film Not included Bonded to existing roof generation because of lower PV
6 For instance, Hilson Moran’s study: (3) To PV or not to PV? | LinkedIn (payback in 6 years), or this study Solar Panels Carbon Payback | The panel efficiency
Renewable Energy Hub (pay back in 2.5 years) or this study Energy payback time and carbon footprint of commercial photovoltaic systems -
ScienceDirect (pay back in 0.68 to 1.96 years)
Table 3 – The four scenarios evaluated in this study

Whole life carbon of photovoltaic installations 8

9
Methodology and Assumptions
Study period Operational carbon calculations
The study was carried out for a lifecycle period of 25 years, which is normally considered the guaranteed The annual kWh associated with PV generation was assessed by Willmott Dixon’s supply chain partner
service life of a PV panel. Most manufacturers provide both product and power warranties for their PV panels. Solarsense using industry standard software PV*Sol premium 2021 (Valentin Software GmbH), which is based
Product warranties are typically shorter (perhaps 10-12 years) and cover manufacturing defects, environmental on UK climate data and project geometry/orientation.
issues and premature wear and tear. Power warranties are over a longer time span and guarantee a certain
Power warranties (at 25 years) allow for year-on-year power degradation estimates, typically <2% in the first
level of performance at year 25, taking module degradation into account (typically a minimum of 80% power
year of operation and 0.55% from year 2 to 25 for monocrystalline modules and slightly higher for the thin-film
at 25 years). PV modules can last a lot longer than 25 years, but power drop off becomes harder to predict,
modules. For the monocrystalline panel types, we have assumed a year-by-year degradation of 0.6%, and for
so warranties typically only extend to 25 years. For simplicity and accuracy, we have based the service life on
the thin-film modules: 0.8% (Table 5 below).
these power warranties in line with most Environmental Product Declarations (EPDs).

PV types Thin-film Monocrystalline

Embodied carbon calculations
Power warranty at year 25 80% 85%
Bill of materials
Material quantities of the systems tested in this study are based on supply chain partner data and can be Assumed year on year degradation 0.8% 0.6%
found in the following Table 4:
Table 5 - Assumed power degradations from 25-year power warranty

Scenario Scenario 1 Scenario 2 Scenario 3 Scenario 4

Total weight of PV panels (kg) 24,186 7,290 7,290 3,531 The conversion of generated energy data into consequent savings in operational carbon from displaced grid
electricity was carried out in two different ways, as shown in the results section below:
Total weight of optimisers (kg) 1,258 365 N/A N/A
• With a variable carbon emission factor over time, based on the UK Green book 2020 predictions for
Total weight of inverter (kg) 330 48 84 48 the period 2021 to 2046. These emission factors are forecast to drop considerably over time as the grid
decarbonises due to increased renewable energy generation from large scale wind, solar etc.
Total weight of mounting system (kg) 3,118 1,087 1,087 N/A
• With a fixed carbon emission factor for 2021 of 0.291 kgCO2e/kWh as per DUKES (Digest of UK Energy
Total weight of ballast (kg) 21,884 N/A N/A N/A Statistics); this approach is consistent with what is adopted for most building energy compliance modelling
Total weight of cables (kg) 90 18 18 18 software such as SAP, IES etc. This ensures a consistent approach as embodied carbon impact associated
with repair in the future does not currently take into account grid decarbonisation.
Total weight of electricity meters (kg) 6 2 2 2
Table 4 – Bill of quantities by scenario

Data
Unfortunately, EPDs are not yet available for all the equipment considered within the study. As the aim was to
do a generic study, the data hierarchy was set up as follow:
• 1st: create average of relevant & comparable EPDs (minimum of 3)
• 2nd: if not possible, find one plausible EPD (ideally industry average EPD)
• 3rd: if nothing available, using CIBSE TM65 calculations to estimate embodied carbon based on
manufacturer information about material composition breakdown.
Table 6 on page 12 summarises the data sources. In the case of CIBSE TM65 calculations, the basic calculation
method was used due to limited manufacturer information available. The basic calculation method relies on
material composition breakdown, the weight of the product, as well as the complexity of the product. CIBSE
TM65 defines 3 levels of complexity, level 1 designates products with a short supply chain, and level 3 long
supply chains. More information concerning the selection of PV panel EPDs can be found in the results section.

Life cycle stages

All embodied carbon life cycle stages were included except: B1 (use), B2 (maintenance), B3 (repair), B5
(refurbishment), C1 (deconstruction). This was due to the lack of available data and the small impact associated
with those lifecycle stages for PV modules. For A5 (construction), only impact associated with material wastage
was accounted. In summary, the following lifecycle stages: A1-A5 (raw material extraction and processing,
transport, manufacturing, installation), B4 (replacement), C2-C4 (transport to waste facility, waste processing,
and disposal) were estimated. Module D is not included in this study.

Whole life carbon of photovoltaic installations 10

11
Table 6

Waste % Recycling %
Service life
Equipment Product Embodied Carbon Data source Material composition breakdown Weight (kg) during at End of Life
(years)
construction (EOL)

Source Project Specs Data availability Product specs Assumption Specs

Average between 3 EPDs: Trina Solar (TSM-

See section about embodied carbon & PV panels
Monocrystalline -20% efficiency DEG15M.20(II)), Jinko (KMXXXM-72H-TV (Swan)), Sun 11.5 kg per m2 88%
below
Power Energy Solutions (MAXEON 3)
PV modules 25 1%
CIBSE TM65 Mid-level calculation,
complexity level 3
PV thin-film Unknown – used ICE database datapoint 3.81 kg per m2 15%
used ICE average datapoint for A1-A3

Solar Edge P801 Worldwide(v1) 25% aluminium, 60% electronics, 2.1 kg per unit 0%
Optimiser CIBSE TM65 Basic calculation, complexity level 2 15% plastics (Global values) 25
Solar Edge P850 Worldwide(v1) 2.3 kg per unit 0%

SE25K 0.3% stainless steel, 13.4% steel, 4.6% zinc, 12.2% 138 kg per unit 0%
copper, 7.7% aluminium, 14.6% electronics, 3.1%
Inverter SE82.8 CIBSE TM65 Basic calculation complexity level 2 12 48 kg per unit 0%
ceramic, 44.1% epoxide resin
Solis-100K-5G (v3) (Global values) kg per unit 0%
n/a*
K2 SolidRail medium 42 1.34 kg per m 5%
Mounting Rails CIBSE TM65 Basic calculation complexity level 1 100% aluminium (European) 25
K2 D-Dome Rails 1.305 kg per m 5%

K2 OneMid, middle clamp 0.06 kg per unit 0%

Module Clamps CIBSE TM65 Basic calculation complexity level 1 100% aluminium (European) 25
K2 D-Dome clamps 0.06 kg per unit 0%

Roof Clamp / Rail Fixings K2 S-5! Z-Mini-FL Round SeamClamp CIBSE TM65 Basic calculation complexity level 1 100% aluminium (European) 25 0.06 kg per unit 0%

Concrete Tile Ballast Slabs - small Manufacturer EPD from HBF manufacturer for results 4.5 kg per m3 0%
Ballast 100% concrete 25 0%
Concrete Tile Ballast Slabs - bigger A1-A3, remaining assumptions using CIBSE TM65 0%
9 kg per m3

Manufacturer PEP from 12% copper, 0.2% steel, 20% plastic, 4%

Generation meter Legrand meter 25 0.491 kg per unit 0% 80%
Legrand (meter 412010) electronics

18.3% copper wire, 2.6% copper, tin 0.3%, 2.6%

Manufacturer PEP from Hager
Comms cabling cat6 cable UTP polyethylene, 3.4% polypropylene, other plastics 25 0.05 kg per m 5% 17%
(Cable Wan blue RJ45 1m)
60.6%

CIBSE TM65 Basic calculation

DC wiring DC wire 6m2 80% copper, 20% PE 25 0.04 kg per m 5% n/a*
complexity level 1

Manufacturer EPD from Schneider (Compact 1.6% stainless steel, 42.2% UP polyester, 7.7%
Local AC isolator panel & switch - 20 1.15 kg unit 0% 37%
ins100 t0 ins 160) Polycarbonate, 9.5% copper, 28.3% steel

Table 6 – Summary of the embodied carbon data sources

*: n/a because CIBSE TM65 basic calculations were carried out,
therefore it is included within the scale up factor.

13
 Results
Embodied carbon results by scenarios Embodied carbon
Embodied (A1-A5,
carbon B4, C2-C4)ofof
(A1-A4,C1-C4) PVPV panels
Panels (kgCO2e/kWp)
(kgCO2e/kWp)
The following graph shows the embodied carbon impact of a whole PV installation across 25 years (assumed to 1,600
SYSTOVI Monocrystalline,
be a PV service life) for different scenarios. 300W, with micro-inverter
Embodied carbon (A1-A5, B4, C2-C4) over 25 years in kgCO2e/kWp 1,400
JA Solar Monocrystalline, 60
SYSTOVI Monocrystalline, 60
Embodied carbon (A1-A5, B4, C2-C4) over 25 years in kgCO2e/kWp cells, 300Wc
1,000 cells, 300W, silicone

900 1,200

800
1,000
700

kgCO2e/kWp
kgCO2e/kWp
JA Solar Monocrystalline
panel, 300Wc, 60 cells
kgCO2e/kWp
/ kWp

600 800 VOLTEC SOLAR

Monocrystalline, 60 cells, EPDs used in this study
500 300Wc
kgCO2e

400 600
Sun Power Energy Solutions
300 Monocrystalline, 400W
400
200 Trina solar 390 W

100 Jinko 390W

200

-
Scenario 1 Scenario 2 Scenario 3 Scenario 4
0
PV Optimiser Inverter Support Ballast Other electronics 2017 2018 2019 2020 2021 2022
panel
Year EPD Produced
Figure 4 - Embodied carbon over 25 Years
Scenario 1: Project A, Flat roof, PV monocrystalline, Optimisers Figure 5: Embodied carbon of monocrystalline module over time. The size of the dot represent the
Scenario 2: Project B, Pitched roof, PV monocrystalline, Optimisers market share, and the green dot represent the products used in the study.

Scenario 3: Project B, Pitched roof, PV monocrystalline, No optimisers Embodied carbon (A1-A3) in kgCO2e / kWp
Scenario 4: Project B, Pitched roof, PV thin-film, No optimisers Upfront embodied carbon (A1-A3) of PV panels (kgCO2e/kWp)
1,600

Embodied carbon and PV panels 1,400

The study shows that PV panels themselves have the most embodied carbon impact of a PV system: around
50% or more of the total PV installation. It should be noted that this proportion could rise quite a bit more if 1,200
less attention was given to PV panel specification.
The embodied carbon data for monocrystalline PV modules used in this study, represent most recent data 1,000

kgCO2e/kWp
kgC02e / kWp
available (2020-2021) and also most representative of the market: Trina Solar, Sun Power Energy Solutions and value used in this study
Jinko together represent about 30% of the global market share.
800
When exploring embodied carbon data available for this study, we found an important trend in the results, EPDs used in this study
which seemed to be correlated to time – meaning the embodied carbon content of PV panels is decreasing
over time. Figure 5 shows embodied carbon results by kWp for different products for which EPDs were found. 600
Figure 5 (on the following page) is aligned with Etude’s study7, which shows that the embodied carbon of PV
panels is reducing over time. This is likely due to the fact that electricity grids of the countries that manufacture
the panels are decarbonising. 400

Figure 6 (on the following page) shows the embodied carbon impact associated with A1-A3 (raw material
extraction, transport to factory, manufacturing processes) for different types of PV panels per kWp. The dark 200
green bars show the data used in the study to create the average data point used (the line represents the data
used in this study).
0
VOLTEC SOLAR SYSTOVI JA Solar JA Solar SYSTOVI Jinko 390W Trina solar 390 W Sun Power Energy
Monocrystalline, Monocrystalline, Monocrystalline Monocrystalline, Monocrystalline, Monocrystalline Monocrystalline Solutions
60 cells, 300Wc 60 cells, 300Wc panel, 300Wc, 60 60 cells, 300W, 300W, with micro- Monocrystalline,
cells silicone inverter 400W

7 (4) The rapid fall of solar’s embodied carbon | LinkedIn

Figure 6 – Embodied carbon impact associated with lifecycle stages A1-A3 from various EPDs

Whole life carbon of photovoltaic installations 14

 Operational carbon results
Figure 7 shows the ‘avoided’ carbon emissions, due to the fact that the PV panels are generating electricity
rather that the UK grid. These are thought of as operational carbon savings. The first bar (dark yellow) shows
the operational carbon savings assuming grid decarbonisation based on information from the UK Green Guide Embodied carbon (A1-A5, B4, C2-C4) & Operational carbon (B6 savings)
2020. The second bar (light yellow) shows the carbon savings assuming no decarbonisation of the UK electricity with decarbonisation (1) and without (2) over 25 years in kgCO2e / kWp
1,000
grid.
Operational carbon
Operational (B6)(B6)
carbon savings
savings from PVgeneration
from PV generation
over over 25 years
25 years
Scenario 1 without decarbonisation
Scenario 2 (1) and withScenario
(2) 3 Scenario 4 500
-
Decarbonised grid

Non decarbonised grid

Non decarbonised grid

Decarbonised grid

Non decarbonised grid

Decarbonised grid
Decarbonised grid

Non decarbonised grid

-1,000 -
1 2 1 2 1 2 1 2
-2,000

kgCO2e / kWp
-500 Scenario 1 Scenario 2 Scenario 3 Scenario 4
kgCO2e/kWp
/ kWp

-3,000
kgCO2e

-4,000 -1,000

-5,000
-1,500
-6,000

-7,000 -2,000

Figure 7 - Operational carbon (B6) savings for all scenarios with and without decarbonisation predictions
-2,500

Operational carbon savings (due to displacing grid electricity) Other electronics Ballast Support Inverter Optimiser PV
Embodied carbon and operational carbon combined
Figure 8 brings Figure 4 and 7 together to show both embodied carbon impact and operational carbon savings
in kgCO2e/kWp for a period of 25 years. In all scenarios, even with grid decarbonisation, the operational carbon
saving still outweighs the embodied carbon impact over the 25-year life span of each PV system installation.

1,000
Embodied carbon (A1-A5, B4,C2-C4) & Operational carbon (B6) savings over 25 years

-1,000

-2,000
Scenario 1 Scenario 2 Scenario 3 Scenario 4
/ kWp
e/kWp

-3,000
kgCO2e
kgCO2

-4,000

-5,000

-6,000

panel
-7,000

Operational carbon savings (due to displacing grid electricity) Other electronics Ballast Support Inverter Optimiser PV
Hackbridge Primary School, Elementa Consulting
Figure 8 - Operational carbon savings and embodied carbon scenarios for all scenarios
designed the UK’s first net zero carbon school at the
Scenario 1: Project A, Flat roof, PV monocrystalline, Optimisers. Scenario 2: Project B, Pitched roof, PV monocrystalline, Optimisers time and the first to be ‘Passivhaus plus’ certified.
Scenario 3: Project B, Pitched roof, PV monocrystalline, No optimisers. Scenario 4: Project B, Pitched roof, PV thin-film, No optimisers

Whole life carbon of photovoltaic installations 16

17
Discussion
Can we really compare embodied carbon and operational carbon? Technical discussion points Flat vs Pitched roof
Flat vs Pitched roof
In the results section, we combined operational carbon savings and embodied carbon values even though they
Difference between flat and pitched roof:

(B6) 25 years in kgCO2e/kWp

1,000
follow different calculation methodologies: operational carbon savings are based on predicted data which the
industry has a good track record at estimating, whereas the embodied carbon values rely on different types of The results (expressed in kgCO2e per kWp) indicate
data (both EPDs and CIBSE TM65 manufacturer forms), a number of assumptions and don’t take into account that the flat roof option has a slightly higher embodied
decarbonisation scenarios. carbon impact, as it needs more support and requires 500

concrete tiles. Surprisingly the PV yield is also higher for

Moreover, it is debatable whether embodied and operational carbon should be compared at all in this case. If

years
the flat roof option, which is a result of the suboptimal
we account the embodied carbon impact of on-site renewable generation, we would need to account for the

Carbon
roof orientation of our pitch roof case study. This has
embodied carbon impact associated with energy grid in case of off-site renewable or non renewable generation.

over 25
resulted in a better net whole life carbon balance for the -

& Operational
Scenario 1 Scenario 2
The whole life carbon study should not be done in isolation and all the relative benefits and impacts of flat roof, but also highlighted the importance of project
solar PV generation for net zero carbon buildings should be considered. specific factors.

e/kWp

Decarbonaised grid

Decarbonised grid
B4, C2-C4)
For the case that both systems were set up to have a -500

2
kgCO
similar orientation, it might reasonably be expected
PV embodied carbon compared to building embodied carbon targets for the flat roof (with lower pitch) to generate less solar

Embodied Carbon (A1-A5,

Proportion of energy yield, thereby creating fewer carbon savings. A further
Building embodied carbon targets have been 1,000 -1,000
consumption generated factor was that project B (the pitched roof installation)
developed by LETI and RIBA8 which are setting ‘best 87% by PV panels also suffers from shading issues from neighbouring
900
practice’ benchmark levels for the embodied carbon
buildings. So, while the embodied carbon impact is a
of typical building types. 800 relative constant number (generally higher for flat roof -1,500

Figure 9 shows the embodied carbon impact of the PV 111 systems), the operational offsets vary by many factors PV Optimiser
700
installation of Project A compared to the rest of the such as the geography, orientation, and pitch. This Inverter Support
Ballast Other electronics
embodied carbon in the building, assuming that the 600 shows that we cannot make general conclusions and
e/m2

675 Carbon savings from PV generation

kgCO2e/m2

embodied carbon of the rest of the school building is specific project-based calculations need to be carried
2

675 kgCO2e/m2 – aligned with a LETI C rating. 500 (Building out to make an informed design decision.
kgCO

Figure 9 - Embodied carbon (A1-A5,B4,C2-C4) & operational carbon (B6)

Embodied Scenario 1: Project A, Flat roof, PV monocrystalline, Optimisers
If we assume that the project A meets the LETI energy 400
carbon Scenario 2: Project B, Pitched roof, PV monocrystalline, Optimisers
use intensity (EUI) target of 65 kWh/m2 /year, then the
target)
PV system will generate 87% of the annual energy
300
The impact of PV optimisers:
consumption of the building. 200 Optimiser vs No optimiser
PV optimisers offer a range of benefits such as safety, Optimisers vs No optimisers
wih better inverter
Appendix 1 explores, how the proportion of 100
control, monitoring and more flexible system design 1,000

embodied carbon of a PV system differs with schools (different panels and orientations can be utilised) but
with a different number of storeys. 0 the most relevant to yield are the removal of mismatch
Project A losses (from manufacturing tolerances and shading).
PV Embodied carbon (kgCO2e/m2) 500
Figure 9 – Embodied carbon of a PV installation for project A (2 Building embodied carbon target (675 kgCO2e/m2)
Optimisers are most useful for difficult roof
story) compared to total building embodied carbon target configurations, e.g. a roof with more than one

kgCO2e/kWp over 25 years

orientation and localised shading issues. A simple roof
design without any shading issues will benefit very little -
Other points to consider about PVs: Scenario 2 Scenario 3
(in terms of additional yield) but will take an embodied
carbon penalty for the additional components. This is

Decarbonaised grid

Decarbonised grid
the case for our case study project B (with optimisers)
The focus of this study is about carbon, however there are other issues to consider when procuring PV panels: -500
where very small yield increases are offset with a
• Ethics: the supply chain should follow anti-slavery legislation to ensure it is free from slavery, forced labour and rather large increase of embodied carbon (due to
other human rights violations *. additional components). Optimisers are often applied
for panel pairs, so result in a large number of additional
• Other environmental impacts: the product should ensure for instance low environmental toxicity, high recycling -1,000
equipment use.
rates. EN 15804 compliant EPDs disclose other environmental impacts than climate change and should be also
taken into consideration. In our case study, the no optimiser option (Scenario
4) was further improved by a more efficient inverter
• Human health impacts: the supply chain should ensure an extraction and manufacturing process which is not -1,500
(as specified by the supplier), resulting in higher
harmful with red-list free materials. PV Optimiser
conversion efficiency (i.e. increased yield). Again, Inverter Support
*/ some studies show it is not always the case: https://www.shu.ac.uk/helena-kennedy-centre-international-justice/research-and-projects/all- while the embodied carbon figures are locked through Ballast Other electronics

projects/in-broad-daylight; https://www.antislavery.org/solar-panel-industry-uyghur-forced-labour/ specification, the project specifics will govern whether Carbon savings from PV generation

optimisers can provide operational carbon offset

advantages through better yield. This shows that once Figure 10 - Embodied carbon (A1-A5,B4,C2-C4) & operational carbon (B6)
8 https://www.leti.london/carbonalignment
again project specific calculations need to be carried out Scenario 2: Project B, Pitched roof, PV monocrystalline, Optimisers
to make informed design decisions. Scenario 3: Project B, Pitched roof, PV monocrystalline, No optimisers

Whole life carbon of photovoltaic installations 18

19
 PV installation with
Difference between monocrystalline and 1,000
Monocrystalline vs Thin
Monocrystalline vs film
Thin-film

thin-film technology:

Savings) 25 years in kgCO2e/kWp

Although thin-film technology has a lower efficiency, it
doesn’t require any additional roof mounting systems/
500
supports, and the embodied impact of the panel itself
is similar to that of an equivalent monocrystalline panel
- this could be interesting from a whole life carbon

(B6years
perspective. -

25
Scenario 2 Scenario 4
However, thin-film technology is currently only available

e/kWp over Carbon

at significantly higher total capital cost, compared to

Decarbonaised grid

Decarbonised grid
& 2Operational
conventional roof mounted PV systems. Moreover, the
availability of embodied carbon data is scarce. Thin- -500

film manufacturers were contacted to carry out CIBSE

kgCO
TM65 calculations, but unfortunately data was not

Embodied Carbon (A1-A5, B4, C2-C4)

made available, so instead ICE A1 to A3 data was used to
estimate total embodied impact. -1,000

Although it seems from these initial calculations that

there is a carbon benefit in using thin-film rather than
more standard monocrystalline PV, it should be noted -1,500
that it is said that the recycling rate is quite low for
PV Optimiser
thin-film and it is also known for its toxicity issues,
Inverter Support
especially regarding the use of cadmium-telluride (See
Ballast Other electronics
article Review on Life Cycle Assessment of Solar
Carbon savings from PV generation
Photovoltaic Panels for further information).
Figure 10 - Embodied carbon (A1-A5,B4,C2-C4) & operational carbon (B6)
Scenario 2: Project B, Pitched roof, PV monocrystalline, Optimisers
Scenario 4: Project B, Pitched roof, PV thin-film, No optimisers

PV as an offset mechanism - ‘why it is not about payback’

We are used to thinking about the payback of measures that reduce carbon emissions – for example a
financial payback (a result of energy savings) when installing additional insulation.
Recently, the industry has started looking at both the embodied carbon impact and operational carbon
savings to evaluate the net effect of carbon reduction measures. This can inform decision making based on
whether the embodied carbon outlay is worth the operational carbon reductions. It is tempting to take this
same approach when considering whether to install PVs or not, However, doing so might have unintended
consequences and could ignore other important global factors. For example, when carrying out these
calculations, the future decarbonisation of the grid is taken into account based on the assumption that
significantly more renewable generation will be added to the grid in coming years. In order to meet our
climate targets, we need to shift progressively to 100% renewables, so new installations of PV and other
renewable energy systems are required to decarbonise the grid further
• other project picture
This means we need to ‘invest’ embodied carbon into installing renewable energy infrastructure. Without
that initial ‘embodied carbon’ investment the grid will not decarbonise further. As the grid decarbonises, local
supply chains also benefit from accessing renewable energy, reducing the upfront embodied carbon content
of their products.
Intuitively we can understand that PV installations are required to decarbonise our electricity grids
and to move away from fossil fuels such as coal and gas. The UK grid needs to substantially increase
capacity to deal with the likely increased demand of the energy in the future (e.g. heat pumps and
electric cars) and rooftop solar PV represents a significant opportunity to support this renewable
energy generation push.
University of Warwick’s Interdisciplinary Biomedical
Even though our results suggest that PV as a pure carbon offset mechanism will be less useful going
Research Building (IBRB): project from Willmott
forward (as operational offsets diminish in line with decarbonisation), the additional renewable Dixon which features 390 vertical PV solar panels.
capacity to help balance supply and demand will be far more important in its contribution to the
energy transition.

Whole life carbon of photovoltaic installations 20

21
Conclusions
Main takeaways of this study Our call to action
The whole life carbon performance of PV installations is strongly affected by design While the whole life carbon impact of a PV installation is a key factor in net zero building decision making, we
decisions (e.g. roof type, pitch, orientation, etc.) and as such, we strongly recommend to should acknowledge that the availability and quality of embodied carbon data at product level is still quite
model and test different scenarios to find the optimum project specific solutions. limited, and that encouraging further work in this area is important.
While embodied carbon impacts generally are more fixed than operational savings (which To improve data and measurements, the key action we believe is needed, is to bring the PV manufacturing
depend on project specific parameters), it can be valuable to test different PV module industry on this journey with us to work on lower embodied carbon components and provide better EPD data.
types against their whole life carbon performance. In some cases, lower efficiency
Embodied carbon can be further reduced through intelligent design, specification and procurement decisions,
components can also reduce embodied carbon impacts (e.g. thin-film modules).
but we need to validate scenarios through models, and we need to feed these models with the best data
A ‘payback’ approach for PV installations, i.e. embodied carbon impact compensated by the available.
operational carbon savings, is not the right metric during an energy transition towards 100%
We need to work with the PV manufacturing industry to enhance the service life of PV installations to reduce
renewables. Embodied carbon will have to be invested to achieve full grid decarbonisation
replacement frequency (therefore less material extraction and manufacturing) and ensure components can be
and rooftop solar PV should continue to play a vital role in supporting this transition.
readily recycled and recovered at the end of their useful life. Also needed is full disclosure of impacts through
the supply chain, not only on climate change but also on other environmental impacts while ensuring ethical
practices.
Our findings We need supply chains to take this feedback on board and work with consultants and contractors to improve
data and decarbonise their own production facilities.
• While the embodied carbon impact of a whole PV installation (i.e., PV modules and all supporting
infrastructure) appear significant in all options explored as part of this study, the carbon savings We need customers to specify low embodied carbon components and drive the market in the right direction.
from PV generation in a UK context provided a significant net benefit (i.e. operational savings Further research is required into how PV embodied carbon impacts compare to the embodied carbon of other
outweighed the impact) in all tested scenarios (with and without decarbonisation scenario). renewable energy generation systems, to fossil fuel energy generation systems, and importantly, how such
embodied carbon impacts are reflected in grid emission factors from generation systems and the distribution
infrastructure.
• In the future when operational savings reduce further (in line with grid decarbonisation), the embodied
carbon content of PV components will become even more important, and this is something that supply
chains will have to respond to by decarbonising their own operations, responsibly sourcing components
and better data.

• When looking at PV installations at a component level, PV modules remain the single largest embodied
carbon impact. However, contribution from the supporting infrastructure (mounting systems, ballast,
inverters, optimisers etc.) is significant and should be accounted for and minimised in design and
embodied carbon calculations.

• While embodied carbon impacts for equipment are relatively fixed, the operational savings from generated
renewable energy depend heavily on the project specific parameters: roof orientation, pitch, geography,
roof and panel design. Although it is not possible to draw many general and universal conclusions from our
small study size, there are strong cost and value incentives for designers to maximise generation efficiency
and yield, and therefore maximise operational savings.

• Our study found that a thin-film PV option might be able to deliver a measurable embodied carbon saving
against more conventional PV system configurations. Questions around recycling of thin-film systems
(components are bonded), limited roof system compatibility, and current higher capital costs mean that
this likely to remain a specialist option for now.

• The pros and cons of PV optimisers (from a whole life carbon standpoint) are not straight forward and
should always be tested on a project-by-project basis and evaluated together with other criteria (e.g.
shutdown control benefits, monitoring requirements). In some instances, more efficient inverters might
give equivalent yield benefits while decreasing the embodied carbon impact, while projects with complex
roof geometry and shading issues would probably still benefit from optimisers.

• One of the main conclusions of this initial study is the need for the solar PV manufacturing industry to
provide lower embodied carbon PV panels and further robust EPDs to investigate the embodied carbon
impact associated with PV installations.

Whole life carbon of photovoltaic installations 22

23
Appendix 1
What is the embodied carbon of PV compared to the
rest of the building - for various building heights?
Industry embodied carbon total building targets are now appearing, such as LETI/RIBA targets, which are
setting ‘best practice’ benchmark levels for the embodied carbon of typical building types.
These targets do not typically include the embodied carbon of PV, as it is thought that this would disincentivise
the installation of PV panels. PVs in effect become part of the wider energy grid infrastructure, helping to
decarbonise electricity; if they are not installed on the building they (or other renewable energy) will need to
be installed elsewhere in the UK. However, as PVs are paid for and installed by a specific building contract,
excluding them potentially means the associated embodied carbon is not counted anywhere. Not including
them also raises questions about where the boundary lies for building integrated PV; for example, when used
as a facade. An additional benefit would be the incentive for manufacturers to develop products with reduced
embodied carbon and increased energy efficiency.
To help understand the relative impact of the embodied carbon of PV systems compared to the rest of the
building a study was undertaken to understand the embodied carbon of a PV system compared with the total
building. The relative PV system size compared to total building floor area relates to the number of storeys
of the building. Hence the study uses the embodied carbon data from the PV installation of Project A from
the main body of this report, to understand the relationship between embodied carbon of the PV system,
embodied carbon of the building, and the proportion of energy consumption that is generated onsite for a
single storey school, as well as a school with 2,5 and 10 storeys. In this study it is assumed that the schools
meet the LETI EUI target of 65 kWh/m2 /yr and the embodied carbon of the building aligns with a LETI C rating.
If the school had had just one storey, the embodied carbon impact of the PV system installation (221 kgCO2e/
m2) would have represented an additional 33% of embodied carbon, however it would generate 1.7 times the
annual energy consumption of the school. If the building had been a 10-storey building, the PV system would
have only represented an additional 3% of embodied carbon, however the PV system would only be capable of
generating 17% of the annual energy consumption.

Embodied Carbon of PV Installation

in relation to total building target of 675 kgCO2e/m2 (A1-A5, B4, C2-C4) over 60 years
Proportion of energy
consumption generated by
Embodied Carbon of PV Installation in relation to total building target of 675 kgCO2e/m2
PV panels:
(A1-A5, B4, C2-C4) over 60 years
1000 170% 87% 35% 17%
900
embodied carbon kgCO2e/m2 over 60 years

800
221
111
700 44 22

600
kgCO2e/m2

500

400

300

200

100

0
1 Storey 2 Storeys 5 Storeys 10 Storeys

Building embodied carbon target (675 kgCO2e/m2) PV Embodied carbon (kgCO2e/m2)

Figure 11 - Embodied carbon of the PV installation of Project A (Scenario 1) in relation to total UK building embodied carbon target at various number
of storeys hight.

Whole life carbon of photovoltaic installations 24

25