Advancing Clean

Technology
Manufacturing
An Energy Technology Perspectives
Special Report
INTERNATIONAL ENERGY
AGENCY
The IEA examines the IEA member IEA association
full spectrum countries: countries:
of energy issues
including oil, gas and Australia Argentina
coal supply and Brazil
Austria
demand, renewable
Belgium China
energy technologies,
electricity markets, Canada Egypt
energy efficiency, Czech Republic India
access to energy, Denmark Indonesia
demand side Estonia Kenya
management and Finland Morocco
much more. Through France Senegal
its work, the IEA Germany Singapore
advocates policies that Greece South Africa
will enhance the Hungary Thailand
reliability, affordability Ireland Ukraine
and sustainability of
Italy
energy in its
31 member countries, Japan
13 association Korea
countries and beyond. Lithuania
Luxembourg
Mexico
Netherlands
New Zealand
Norway
Poland
Portugal
Slovak Republic
Spain
Sweden
Switzerland
Republic of Türkiye
This publication and any United Kingdom
map included herein are United States
without prejudice to the
status of or sovereignty over
any territory, to the
delimitation of international
frontiers and boundaries and
to the name of any territory, The European
city or area. Commission also
participates in the
work of the IEA

Source: IEA.
International Energy Agency
Website: www.iea.org
Advancing Clean Technology Manufacturing Abstract
An Energy Technology Perspectives Special Report

Abstract

Governments and firms around the world are racing to define their place in the
clean energy economy, which is growing quickly as policy makers develop new
industrial strategies that also bolster energy security and address climate change.
This Energy Technology Perspectives Special Report is structured to provide
decision makers with an analytical toolkit to design and evaluate their strategies
for clean technology manufacturing. Acknowledging that there is no “one size fits
all” approach, it lays out guiding principles that can help inform future planning.

This analysis was produced in response to a request from G7 Leaders in 2023. It

benefits from the insights gathered during a High-level Dialogue on Diversifying
Clean Technology Manufacturing held at the IEA headquarters in Paris in
November 2023. It also builds on analysis conducted as part of the latest edition
of the IEA’s flagship technology publication, Energy Technology Perspectives, and
two Special Briefings on the topic of clean technology manufacturing during the
course of 2023.

I EA. CC BY 4.0.

PAGE | 3
Advancing Clean Technology Manufacturing Acknowledgements
An Energy Technology Perspectives Special Report

Acknowledgements

This study was prepared by the Energy Technology Policy (ETP) Division of the
Directorate of Sustainability, Technology and Outlooks (STO) of the International
Energy Agency (IEA). The study was designed and directed by Timur Gül, IEA
Chief Energy Technology Officer. Araceli Fernandez Pales, Head of the
Technology Innovation Unit, provided strategic guidance throughout the
development of the project. Peter Levi co-ordinated the analysis and production
of the report.

The lead authors (in alphabetical order) were: Simon Bennett (innovation), José
Bermudez Menendez (electrolysers), Chiara Delmastro (heat pumps), Mathilde
Fajardy (manufacturing costs), Alexandre Gouy (manufacturing costs), Carl
Greenfield (policy), Mathilde Huismans (wind and data management), Teo
Lombardo (batteries), Rafael Martinez Gordon (heat pumps), Faidon
Papadimoulis (solar PV) and Francesco Pavan (electrolysers), and Chang Tan
(manufacturing costs). Other key contributors were: Caleigh Andrews, Piotr Bojek,
Johannes Hampp, Jean-Baptiste Le Marois and Biqing Yang.

Per-Anders Widell and Anna Kalista provided essential support throughout the
process. Lizzie Sayer edited the manuscript. Thanks also to the IEA
Communications and Digital Office for their help, particularly to Jethro Mullen,
Poeli Bojorquez, Curtis Brainard, Astrid Dumond, Merve Erdil, Grace Gordon,
Julia Horowitz and Clara Vallois.

Valuable comments and feedback were provided by other colleagues within the
IEA, in particular Heymi Bahar, Alessandro Blasi, Laura Cozzi, Trevor Criswell,
Tim Gould, Dennis Hesseling and Thomas Spencer.

The analysis and findings in this report draw on strategic guidance and insights
gathered during a High-level Dialogue on Diversifying Clean Technology Supply
Chains in Paris in November 2023.

The work could not have been achieved without the financial support provided by
the Government of Japan.

Several senior government officials and experts provided essential feedback to

improve the quality of the report. They include: Dries Acke (Solar Power Europe);
Jeremy Avins and Giulia Siccardo (Department of Energy, United States); Ben
Backwell (Global Wind Energy Council); Marco Baresi (Turboden); Edwin Basson
(WorldSteel); Marek Bielewski, Francesco Dolci, Aliki Georgakaki, Arnulf Jaeger-
Waldau, Evdokia Tapoglou and Agne Toleikyte (JRC, European Commission);
I EA. CC BY 4.0.

PAGE | 4
Advancing Clean Technology Manufacturing Acknowledgements
An Energy Technology Perspectives Special Report

Reed Blakemore (Atlantic Council); Roberto Bocca (World Economic Forum);

Rina Bohle Zeller (Vestas); Laura Casuscelli (WindEurope); Sam Cornish (IIGCC);
Leandro de Oliveira Albuquerque (Ministry of Mines and Energy, Brazil); Rebecca
Dell (ClimateWorks Foundation); Miriam D’Onofrio and Sarah Ladislaw (National
Security Council, United States); Daniel Dufour (Natural Resources Canada);
André Eckermann (GIZ); Martin Forsen (NIBE); Marie-Laetitia Gourdin and
Christin Töpfer (Vattenfall); Rishabh Jain and Dhruv Warrior (CEEW); Leif
Christian Kröger (Thyssenkrup nucera); Thomas Kwan and Silvia Madeddu
(Schneider Electric); Jon Lezamiz Cortazar (Siemens Gamesa); Xiao Lin (Botree
Recycling Technologies); Johan Lindahl (ESMC); Michael Lippert (SAFT); Joseph
Majkut (CSIS); Monika Merdekawat (ASEAN Centre for Energy); Yasuko
Nishimura and Atsushi Taketani (Ministry of Foreign Affairs, Japan); Thomas
Nowak (European Heat Pump Association); Jared Ottmann (Tesla); Gaurav
Pundir (Department of Commerce, India); Marta Ramos Fernandez (Airbus);
David Reiner (University of Cambridge); Mark Richards (RioTinto); Agustín
Rodríguez Riccio (Topsoe); Javier Sanz (Innoenergy); Oliver Sartor (Agora);
Christian Schmidt (Federal Chancellery, Germany); Ulrik Stridbæk (Ørsted);
Jacopo Tattini (European Commission); Peter Taylor (Leeds University); Denis
Thomas (Cummins); Fridtjof Unander (Aker Horizons); Noé van Hulst (IPHE);
Anne van Ysendyck (ArcelorMittal); David Victor (University of California San
Diego); Natasha Vidangos (Environmental Defence Fund); Miki Yamanaka (Daikin
Industries).

I EA. CC BY 4.0.

PAGE | 5
Advancing Clean Technology Manufacturing Table of contents
An Energy Technology Perspectives Special Report

Table of contents

Executive summary ........................................................................................................ 7

Part I. Clean technology manufacturing today ........................................................... 12
Chapter 1. An introduction to clean technology manufacturing ............................... 13
Clean technology manufacturing in context ............................................................................. 13
Tracking progress on clean technology manufacturing ........................................................... 18
Chapter 2. The new clean energy economy is emerging ........................................... 27
The project pipeline continues to expand ................................................................................ 27
Rapid - if uneven - progress ..................................................................................................... 34
Part II. Advancing clean technology manufacturing .................................................. 49
Chapter 3. Cost fundamentals of clean technology manufacturing.......................... 50
Levelised cost of manufacturing............................................................................................... 50
Upfront costs ............................................................................................................................ 54
Operational costs ..................................................................................................................... 60
Policy incentives for manufacturing.......................................................................................... 65
Chapter 4. The role of innovation in advancing clean technology manufacturing .. 70
The link between energy innovation and manufacturing has strengthened ............................ 72
Innovation can overcome high cost factors to maintain manufacturing competitiveness ........ 74
Policy missions for innovation to unlock new manufacturing opportunities ............................. 77
The value of technology innovation besides lowering manufacturing costs at home .............. 82
Chapter 5. Policy priorities for advancing clean technology manufacturing ........... 87
Enlarging domestic markets with climate policy....................................................................... 87
Compressing lead times ........................................................................................................... 89
Boosting the availability of skilled workers ............................................................................... 90
Synergies from supply chain integration .................................................................................. 93
Reducing supply chain uncertainty with trade agreements ..................................................... 93
Reducing environmental impacts and addressing social considerations ................................ 95
Part III. Key principles for decision makers ................................................................ 98
Domestic actions to advance clean technology manufacturing ............................................... 98
International co-operation to support domestic investment and global progress .................. 100
Technical annex.......................................................................................................... 106
Abbreviations and acronyms .................................................................................................. 112
Units ....................................................................................................................................... 113
I EA. CC BY 4.0.

PAGE | 6
Advancing Clean Technology Manufacturing Executive summary
An Energy Technology Perspectives Special Report

Executive summary

Clean technologies shine a spotlight on manufacturing

The manufacturing sector – long an engine of economic growth and development
– is increasingly at the forefront of considerations on energy, climate and
economic policy. Countries are racing to capitalise on benefits that clean
technology manufacturing can bring to economic security, employment and the
resilience of clean energy transitions. Following a request by G7 Leaders in 2023,
this Energy Technology Perspectives Special Report is designed to aid policy
makers as they prepare their industrial strategies. It focuses on five key clean
energy technologies – solar PV, wind, batteries, electrolysers and heat pumps.

Investment in clean technology manufacturing is becoming so significant that it is

starting to register in broader macroeconomic data. In 2023, it accounted for
around 0.7% of global investment across all sectors of the economy, driving more
spending than established industries like steel (0.5%). In growth terms, the
contribution is even starker – in 2023, clean technology manufacturing alone
accounted for around 4% of global GDP growth and nearly 10% of global
investment growth.

The recent surge in investment looks set to continue

New, first-of-its-kind analysis in this report shows that investment in clean
technology manufacturing stood at around USD 200 billion in 2023, growing by
more than 70% relative to 2022. Investments in solar PV and battery
manufacturing plants led the way, together accounting for more than 90% of the
total in both years. Investment in solar PV manufacturing more than doubled to
around USD 80 billion in 2023, while investment in battery manufacturing grew by
around 60% to USD 110 billion.

China accounted for three-quarters of global investments in clean technology

manufacturing in 2023, down from 85% in 2022, as investment in the United States
and Europe grew strongly – particularly for battery manufacturing, for which
investments more than tripled in these regions. For solar PV manufacturing,
investments in China more than doubled between 2022 and 2023. Outside these
three major manufacturing hubs, India, Japan, Korea and countries in Southeast
Asia made important contributions in specific areas, while investment in regions
such as Africa, Central America and South America was negligible.

Near-term momentum for clean manufacturing looks strong. Around 40% of

investments in 2023 were in facilities that are due to come online in 2024; for
battery manufacturing facilities, this share is nearly 70%. Committed projects –
I EA. CC BY 4.0.

PAGE | 7
Advancing Clean Technology Manufacturing Executive summary
An Energy Technology Perspectives Special Report

those that are under construction or have reached final investment decisions –
through 2025, together with existing capacity, would exceed by 50% the global
solar PV deployment needs in 2030 based on the IEA’s Net Zero Emissions by
2050 Scenario (NZE Scenario) and meet 55% of battery cell requirements. This
momentum is also spreading to adjacent sectors – nearly half of committed battery
manufacturing announcements in the United States will be via joint ventures with
automakers.

The project pipeline is expanding rapidly, if unevenly

Existing manufacturing capacity for solar PV modules and cells could today
achieve what is necessary to meet demand under the NZE Scenario in 2030 – six
years ahead of schedule, with only modest gaps remaining for the upstream steps
of wafer and polysilicon manufacturing. However, facilities making cells and
modules are currently seeing relatively low average utilisation rates of around 50%
globally. Key factors that explain this are a solar PV module supply glut, together
with the rapid expansion of manufacturing capacity. While the sharp increase in
supply has driven down module prices, supporting wider consumer uptake,
stockpiles of solar PV modules are growing and there are signs of downscaling
and postponements of planned capacity expansions, particularly in China.

Battery manufacturing also had a record year in 2023. Production totalled more
than 800 gigawatt-hours (GWh), rising 45% from 2022. Capacity additions also
surged, with almost 780 GWh of cell manufacturing capacity added – around a
quarter more than in 2022. This raised total installed capacity to around
2.5 terawatt-hours (TWh), or almost three times current demand. Globally, battery
manufacturing capacity could exceed 9 TWh by 2030 if all announcements are
realised. Battery manufacturing deployment needs in 2030 under the NZE
Scenario are within reach: more than 90% could be met by announced expansions
that have reached final investment decisions.

New manufacturing capacity for wind and electrolysers also grew faster in 2023,
although the gains were not as dramatic. Existing capacity for wind could deliver
nearly 50% of NZE Scenario needs in 2030, while announced projects could meet
a further 12%. Meanwhile, capacity additions for heat pump manufacturing slowed
due to stagnation in the majority of leading markets. Existing capacity could deliver
only around one-third of 2030 needs in the NZE Scenario – though this could
change quickly given the short lead times typical of capacity expansions in this
industry.

Geographic concentration in manufacturing looks set to

remain high for most clean energy technologies
China, the United States and the European Union together account for around
80% to 90% of manufacturing capacity for solar PV, wind, battery, electrolyser and
I EA. CC BY 4.0.

PAGE | 8
Advancing Clean Technology Manufacturing Executive summary
An Energy Technology Perspectives Special Report

heat pump manufacturing. Little change to this overall concentration is foreseen

to 2030, even if all announced projects come to fruition. China alone accounts for
more than 80% of global solar PV module manufacturing capacity and 95% for
wafers. This looks unlikely to change significantly this decade, with the country set
to match or exceed the capacity additions planned in other countries like the
United States and India. For battery cell manufacturing, the situation is somewhat
different: Planned capacity additions in Europe and the United States look set to
reduce China’s present share of global capacity, with both regions reaching
around a 15% share by 2030 if all announced projects are realised. In Europe and
the United States, announced battery cell manufacturing capacity is sufficient to
meet the 2030 domestic deployment needs associated with their own climate
goals.

The geographic concentration of manufacturing for wind, electrolysers and heat

pumps also shows little change through 2030. Outside of the main producer
countries, Central and South America account for a small share of global
production of the main wind turbine components (4% to 6% for nacelles, blades
and towers). However, virtually no clean technology manufacturing takes place in
Africa today. Concentration is even more pronounced for upstream solar PV and
battery components, but the prospect of surplus capacity may open up possibilities
for greater diversification of production in this area.

Production cost gaps are significant, but not immutable

New data and analysis, including plant-level assessments of more than 750
facilities, provide insight into key drivers of manufacturing costs and the
differences between regions. Our analysis shows that China is the lowest-cost
producer for all the technologies highlighted in this report, before accounting for
explicit supportive policy measures, though it also points to opportunities for
reducing cost gaps.

The main upfront cost that contributes to overall production costs is the capital
expenditure to set up a clean energy manufacturing plant, and the associated
financing costs. Facilities in the United States and Europe are typically 70% to
130% more expensive per unit of output capacity than those in China for solar PV,
wind and battery manufacturing, before accounting for the difference in the cost of
capital between regions. India’s capital costs are around 20% to 30% higher than
China’s, but significantly lower than those of the United States and Europe.

However, upfront costs make only a modest contribution to the overall levelised
cost of manufacturing. Annualised capital expenditure amounts to just 15% to 25%
of the total cost of producing solar PV modules, with a cost of capital of 8%. The
proportions are similar for batteries (10-20%), wind turbines and heat pumps (2-
10%) and somewhat higher for alkaline electrolyser stacks (15-30%). Operational
costs, including energy, material, component and labour costs, make a far more
I EA. CC BY 4.0.

PAGE | 9
Advancing Clean Technology Manufacturing Executive summary
An Energy Technology Perspectives Special Report

important contribution in aggregate. Using global average commodity prices, and

regional labour and end-user prices for energy inputs, ongoing operational costs
account for 70% to 98% of total manufacturing costs. Reducing the costs of
energy, materials and components is therefore an important lever for reducing
cost gaps.

Cost is not the only factor that influences investment

Many factors besides the cost of manufacturing shape the decisions of firms to
invest: the size of the domestic market, the availability of workers with the
necessary skills, infrastructure readiness, permitting processes and other
regulatory regimes, proximity to customers and synergies with existing industries
are just some examples. Policy interventions can therefore raise the attractiveness
of investing in a given region without directly subsidising the costs of
manufacturing. Training and certification schemes for workers, compressing
project lead times while maintaining environmental standards, enlarging domestic
markets and reducing uncertainty with robust, stable climate policies are some key
“low regret” measures that can increase incentives to invest, irrespective of the
role of direct incentives in industrial strategies.

Innovation is another key focus for industrial strategy design; as the portfolio of
energy technologies shifts towards mass-manufactured equipment, the energy
sector is likely to include more R&D-intensive companies with factories and R&D
hubs in their home countries and elsewhere in the world. Being at the frontier of
innovation is an important opportunity to compete in the market, which is one
reason why countries with relatively high labour and energy costs continue to
manufacture goods in trade-exposed sectors. While private-sector R&D can be
stimulated by policies that promote manufacturing investment and experience,
direct innovation support is also needed. Government measures, including R&D
grants or loans, project finance, support for rapid prototyping, start-ups and
production scale-up, can be targeted towards specific innovation missions to
advance manufacturing.
I EA. CC BY 4.0.

PAGE | 10
Advancing Clean Technology Manufacturing Executive summary
An Energy Technology Perspectives Special Report

Key principles to support industrial strategy design

The purpose of this report is not to prescribe a single approach to industrial
strategy or to make recommendations to a specific country, but rather to support
decision-making. Alongside its analysis of competitiveness, innovation and other
specific areas of policy, the report distils a set of key principles to guide policy
makers.

When considering domestic actions, governments should:

 Prioritise and play to strengths, with clearly defined goals and metrics to gauge
success, and with experimentation and the ability to change course built in.
 Attract and support innovators, including by creating strong links between
manufacturing and each component of the broader innovation system.
 Plug cost gaps strategically and for the long-term, including through measures
to reduce lead times and upskill workforces.
Governments should also collaborate internationally, which in turn enhances
opportunities for domestic investment and global progress. To do so, they should:

 Collect data and track progress, including on the trade and production of clean
technologies and their components.
 Co-ordinate efforts across supply chains to enhance resilience by sharing
experiences and collaborating.
 Identify and build strategic partnerships, backed by clear frameworks for co-
operation.

I EA. CC BY 4.0.

PAGE | 11
Advancing Clean Technology Manufacturing Part I: Clean technology manufacturing today
An Energy Technology Perspectives Special Report

Part I. Clean technology

manufacturing today

Clean technology manufacturing is increasingly in the spotlight. The production of

key technologies to support the transition to clean energy has become the
cornerstone of industrial policies designed to boost employment and economic
development in many countries. Moreover, it is a critical enabler for meeting
climate goals, such as the pledge made at COP 28 to triple the world’s renewable
energy capacity by 2030. The first part of this report puts clean technology
manufacturing in context, considering the role of manufacturing in the global
economy, and the latest progress on ramping up manufacturing capacity in line
with an acceleration of clean energy technology deployment.

Chapter 1 examines how manufacturing contributes to economic development in

different regions worldwide, and recent investment trends in clean technology
manufacturing. Chapter 2 tracks progress being made in expanding
manufacturing capacity for five key clean energy technologies: solar PV, wind,
batteries, electrolysers and heat pumps. It analyses whether manufacturing
capacity is on track to meet deployment needs consistent with a pathway to net
zero emissions by 2050, both within countries and regions and at the global level.
Finally, it considers the geographical distribution of manufacturing, examining
potential to meet domestic demand and opportunities for exports, and analyses
levels of concentration in the supply chain.

I EA. CC BY 4.0.

PAGE | 12
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

Chapter 1. An introduction to clean

technology manufacturing

Virtually all aspects of the modern built environment are products of the
manufacturing sector, or an assembly thereof. Besides those we can see and
touch, there are thousands of manufactured products, materials and chemicals
that most people never come into contact with, existing only as intermediates in a
complex network of industrial processes that make up global supply chains. In a
single year, the manufacturing sector takes in billions of tonnes of minerals and
biomass and transforms them into trillions of dollars’ worth of products, adding
value at each processing step. The manufacturing sector is, therefore, essential
to modern society and an important contributor to economic growth and
development.

Clean technology manufacturing in context

The manufacturing sector is also a critical enabler of the clean energy transition.
To transform the current stock of power generation equipment, vehicles, buildings,
industrial facilities and other capital stocks needed to generate and harness clean
energy, the manufacturing sector will itself need to undergo a transformation. This
shift is already underway, due to increasing consumer demand for clean
technologies, falling costs, and supportive government policies aimed at the
products of the manufacturing sector, and how it produces them. Clean
technologies like solar PV panels, wind turbines and batteries together account
for a small fraction of global manufacturing activity today, but they are emerging
as an important contributor to the sector’s growth.

This chapter introduces the manufacturing sector more broadly and provides an
overview of recent trends, including the role of manufacturing in the economy. It
then moves on to define ‘clean technology manufacturing’ using internationally
recognised industrial classification systems, thereby locating its different elements
within established definitions of the wider sector. The chapter concludes with new
IEA analysis on manufacturing investments, providing a bottom-up, granular view
of trends for individual technologies and components, filling a gap in existing top-
down statistical data collection activities.
I EA. CC BY 4.0.

PAGE | 13
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

Manufacturing plays a crucial role in the global economy

Manufacturing is one of the principal drivers of economic growth, accounting for
nearly one-fifth of GDP and employing around 350 million 1 people globally. The
United States, the European Union, Japan and the People’s Republic of China
(hereafter, “China”) accounted for around 70% of manufacturing value added in
2023, a share that has – in aggregate – remained almost constant over the past
two decades. The United States, Japan and the European Union’s combined
share of output from the global manufacturing sector has declined significantly
over this period, while China has emerged as the world’s manufacturing hub.
China nearly tripled its share of global manufacturing in monetary terms during
2005-2023, with its output increasing fivefold in absolute terms. The increase in
China’s share of global manufacturing output underpins a broader trend of
geographic concentration. The top five manufacturing countries in terms of value
added accounted for around 56% of the global total in 2005, with this share rising
to around 64% in 2023.

Figure 1 Share of global manufacturing value added by geography and by sector

Share by geography Share by sector

100%
Chemicals
United States
80%
Metals
European Union
Electronics
60%
Machinery
Japan
Motor vehicles
40% China

20% Other
Rest of World

0%
2005 2023 2005 2023

IEA. CC BY 4.0.
Note: Chemicals includes pharmaceuticals and fertilisers (ISIC Rev. 4 20-22); Metals includes basic and structural (ISIC
Rev. 4 24-25); Electronics includes computer, electronic and optical products and electrical machinery (ISIC Rev. 4 26-27);
Motor vehicles includes all road vehicles and parts (ISIC Rev. 4 29); Other includes all manufacturing (ISIC Rev. 4 10-33)
less that covered under the other categories disaggregated. European Union corresponds to membership in 2023, held
constant over the period analysed.
Source: IEA analysis based on OECD TiVA database and Oxford Economics Global Industry Service.

The sectoral composition of manufacturing has changed much more slowly than
its geographic distribution. Categorised by a series of aggregated sub-sectors
(see Box 1), five industries – chemicals (incl. pharmaceuticals), electronics,

1
Includes employment in the construction sector.
I EA. CC BY 4.0.

PAGE | 14
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

machinery, metals and motor vehicles – accounted for 61% of manufacturing

value added in 2023, increasing slightly from 55% in 2005. Most of these sub-
sectors have maintained relatively consistent shares of total manufacturing output
over the past two decades globally, even if substantial portions of each have been
relocated geographically. The automotive sector contracted by around 20% in
absolute terms in the aftermath of the 2008 financial crisis, rebounding to its pre-
crisis level of output in 2010. The electronics industry, including the production of
electrical machinery, computers, electronic and optical products, is the only one
of these five sectors to have made significant gains relative to the others,
increasing its share of global manufacturing output from 11% in 2005 to 16% in
2023, in part owing to the widespread adoption of portable electronic devices like
laptop computers and mobile phones.

Box 1 International systems for categorising manufacturing activities

and products
The International Standard Industrial Classification of All Economic Activities (ISIC)
is a system administered by the United Nations Statistics Division for categorising
various products, sectors and activities that contribute to a country’s economy. The
ISIC provides an internationally recognised structure for collecting data and
statistics such that they are comparable between countries. The boundaries defined
in the structure of the IEA’s energy balances are based on the ISIC. The ISIC unifies
national and regional statistical frameworks such as the North American Industry
Classification System and the European Classification of Economic Activities. A
given economic activity can be identified at different levels of specificity using its
Section (letters A-U), Division (2-digit number, subset of Section), Group (3-digit
number, subset of Division) or Class (4-digit number, subset of Group).
The ISIC classifies industrial activities associated with producing and transforming
goods (and services and other activities), rather than the goods themselves. The
Harmonized System (HS), administered by the World Customs Organization, is a
classification system that is used for the classification of goods for trade purposes.
Like the ISIC, the HS provides a unified architecture for use across different
jurisdictions, allowing the mapping of one country’s customs classifications to
another. A given product – or more often a group thereof – can be identified by
Chapter (2-digit number), Heading (4-digit number) or Subheading (6-digit number),
with each providing an increasing level of specificity.

Manufacturing continues to be an important pillar of economic development for

emerging market and developing economies (EMDEs). Few countries have
achieved high and sustained levels of economic prosperity without growth in
manufacturing. The role of the sector in the broader economy varies significantly
I EA. CC BY 4.0.

PAGE | 15
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

by country and over time, but has been shown to rise fastest during the early
stages of economic development as a country industrialises. In China, GDP per
capita increased by a factor of 3.5 between 2005 and 2023, during which time the
contribution of manufacturing to the economy increased from 24% to 30%. Even
in India, where services made an outsized contribution to economic growth
compared to other countries at a similar stage in their economic development, the
contribution of manufacturing to the broader economy increased from 15% to 17%
of total value added over the same period.

Figure 2 The role of manufacturing in the broader economy at different stages of

economic development, 2005-2023
30%
Share of VA by manufacturing

Korea
China 2005 2023

Mexico
Japan
Indonesia
20%

South Africa European Union

India United States

10%

0%
0 20 000 40 000 60 000 80 000 100 000
GDP per capita (USD 2023 PPP)
IEA. CC BY 4.0.
Notes: VA = value added. Manufacturing value added includes the activities corresponding to ISIC Rev. 4 Divisions 10-33.
Grey data series shows annual data for 2005-2023 for 68 countries and regions beyond those explicitly identified.
Sources: IEA analysis based on OECD TiVA database and Oxford Economics Global Industry Service.

There are notable exceptions, like Indonesia’s economy, whose GDP increased
rapidly in size over the past two decades, but saw a decline in the share of
manufacturing in total value added. In this case, growth in valued added by mining,
agriculture and services outpaced the growth in manufacturing. Korea’s per capita
economic growth outpaced even China’s in absolute terms during the period 2005-
2023, and the role of manufacturing in the economy continued to grow even at
relatively high levels of per capita income, in part due to supportive industrial
policies. These and other exceptions aside, as countries achieve higher levels of
per capita income, manufacturing tends to make a smaller contribution to the
economy, giving way to services and consumption as the main drivers of growth.
Even among countries with high per capita GDP there is significant variation, with
I EA. CC BY 4.0.

PAGE | 16
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

the United States seeing manufacturing having just an 11% share of in total
economic value added in 2023, compared with 23% in Germany and 21% in Japan.

In EMDEs, the growing importance of manufacturing in total economic output has

been driven by increasing investment. Investments in factories, equipment and
other fixed capital assets for manufacturing grew by nearly 30% during the period
2005-2023 in advanced economies, compared with 260% over the same period
in EMDEs (including China). In 2005 the world’s EMDEs accounted for little under
half of global manufacturing investment. By 2023 this share had risen to over 70%.

More value is added downstream

Downstream manufacturing sub-sectors, i.e. those closer to the final consumer,
tend to account for larger shares of total value added than those upstream, which
produce the main input materials for the downstream industries. The production
of the main large-volume materials, including metals, cement, glass, basic
chemicals, timber, plastic, rubber and paper, accounted for around one-quarter of
total value added from manufacturing in 2023, down from around one-third in
2005. These commodities are often traded in highly competitive markets and have
limited product differentiation, leading to slim margins for their producers.

Figure 3 Global manufacturing investment and value added, 2005-2023

Manufacturing investment Manufacturing value added

20 100%
USD trillion

15 75%

10 50%

5 25%

0 0%
2005 2023 2005 2023

Large-volume materials Other goods EMDE share (right axis)

IEA. CC BY 4.0.
Notes: EMDE = emerging market and developing economies. ‘Large-volume materials’ includes ISIC Rev. 4 Divisions 16-
19, 22-25 and Group 201. ‘Other goods’ includes ISIC Rev. 4 Divisions 10-33, less the large-volume materials. Values
shown in real USD 2023.
Source: IEA analysis based on OECD TiVA database and Oxford Economics Global Industry Service.

Downstream manufacturing sub-sectors have accounted for a larger share of

value added over time, owing to an increase in the variety of products produced,
I EA. CC BY 4.0.

PAGE | 17
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

an increase in the efficiency with which their input materials are used, and
increasing product differentiation in line with consumer demands. For example,
products like mobile phones and tablet computers – for which markets were still
nascent in 2005 – have led to growth in the value of outputs from the ‘computer,
electronic and optical products’ sub-sector that is disproportionate to the growth in
output of key materials from the upstream sub-sectors.

There has also been a marked shift in where these downstream sub-sectors are
located. Value added from downstream sub-sectors in EMDEs increased by a
factor of three between 2005 and 2023, outpacing growth in value added from the
main material-producing sectors, which saw a 240% increase over the same
period. Even in China, which has seen unprecedented growth in the volume of
energy-intensive material production – the country accounted for more than 50%
of global crude steel production and aluminium production in 2023, up from 31%
and 23% respectively in 2005 – the growth in value added from downstream
sectors has risen 40% faster than in upstream material-producing sectors.

Tracking progress on clean technology

manufacturing
The core focus of this report is on five clean technologies and their main
components – solar PV, wind, batteries, electrolysers and heat pumps – as an
indicative sample of key technologies for the clean energy transition, but there are,
of course, many others, such as nuclear reactors; carbon capture, utilisation and
storage (CCUS) technologies; and fuel cells. Clean technology manufacturing is
not presently defined as an individual sub-sector in any systematic data collection
or statistical frameworks. However, products and industrial activities relevant to
clean technology manufacturing can be identified at varying levels of granularity
in established categorisation systems (see Box 2).

Box 2 Defining clean technology manufacturing using established

international classification systems
The manufacturing of these technologies and their primary components is spread
across a handful of manufacturing sub-sectors, most of which are downstream of
the main material-producing sectors. The electrical equipment (ISIC Division 27)
and machinery (ISIC Division 28) sub-sectors account for the majority, including
components of wind turbines, batteries, electrolysers and heat pumps. The
computer, electronic and optical products (ISIC Division 26), chemical and
chemical products (Division 20) and fabricated metal products (ISIC Division 25)
sub-sectors account for the other main components (see table below).
I EA. CC BY 4.0.

PAGE | 18
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

The products themselves, as opposed to the industrial activities that produce them,
can be identified in even more granular terms using the classification systems used
by customs agencies for international trade. The Harmonized System (HS)
administered by the World Customs Organization (see Box 1) identifies several
clean technologies and their components, either individually or as part of wider
groups of products. For example, HS 854143 corresponds to “Photovoltaic cells
assembled in modules or made up into panels”, which matches closely with the
quantities shown for “solar PV modules” in this publication. A perfect match is not
available for all technologies and components within this publication’s core scope,
even with the detailed six-digit HS. Electrolysers, for example, fall within a broader
category of products under code 854330, “machines and apparatus for
electroplating, electrolysis or electrophoresis”.

Mapping clean technologies using a selection of relevant HS and ISIC codes

HS codes ISIC codes
Technology (products) ISIC description of activities
(activities)
Modules 854143, 854190 Manufacture of electronic
2610
Cells 854142, 854190 components and boards
Solar PV Manufacture of other chemical
Wafers 381800 2029
products n.e.c
Polysilicon 280461 2011 Manufacture of basic chemicals
Manufacture of electric motors,
generators, transformers and
Nacelles 850231 2710
electricity distribution and
control apparatus
Wind
Manufacture of fluid power
Blades 841290 2812
equipment
Manufacture of structural metal
Towers 730820 2511
products
850710, 850720
Cells 850730, 850750 Forging, pressing, stamping
850760, 850780 and roll-forming of metal;
2591
powder metallurgy;
Batteries Anodes 850790, 854519 2720
Manufacture of batteries and
850790, 284290 2790
accumulators; Manufacture of
Cathodes 284169, 382499 other electrical equipment
284190, 285390
Manufacture of other electrical
Electrolysers 854330 2790
equipment
Manufacture of other general-
Heat pumps 841581, 841861 2819
purpose machinery
Note: Six-digit HS 2022 and four-digit ISIC Rev 4.

The shortcomings of existing classification schemes for providing a sufficiently

granular description of the industrial activities associated with clean technology
manufacturing mean that more granular data has to be assembled ‘bottom-up’ from
a variety of data sources and research. Company filings are an alternative source of
financial information that can be used as proxies for tracking inputs to, and outputs
from, clean technology manufacturing operations, mostly in economic terms.
I EA. CC BY 4.0.

PAGE | 19
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

Aggregating investment, revenue and earnings data for companies engaged in the
manufacture of clean technologies provides a glimpse at some of the latest trends.

Private sector indicators

Revenue from the top five solar PV 2 manufacturing companies surged from just
USD 10 billion on an annualised basis in Q1 2017 to over USD 100 billion in Q4
2022, before falling slightly during the first three-quarters of 2023. Battery
manufacturing revenue, measured on the same basis for the top five firms, 3
experienced a similar rate of growth, increasing from around USD 26 billion in Q1
2017 to almost USD 200 billion by the third quarter of 2023. Operating income, a
measure of earnings after subtracting operational costs like materials and labour
(but before accounting for taxes, investments and debt interest) grew even faster
than revenues for the top five solar PV firms, and at a similar rate to revenues for
the top five battery firms. These financial metrics indicate highly profitable
operations in absolute terms, with the ten firms’ earnings equating to around 15%
of the global gross operating surpluses (a measure of aggregate profits) from the
sub-sectors in which they are situated (ISIC Divisions 26 and 27). Most of the ten
firms analysed maintained average profit margins (measured here as operating
income as a percentage of revenue) in the range of 5-15-% during 2017-2023.

Figure 4 Aggregated financial indicators for the top five solar PV and battery
manufacturing companies, 2017-2024
Revenue Operating income Capital expenditure
200 20 40
USD billion

150 15 30

100 10 20

50 5 10

0 0 0
2017 2024 2017 2024 2017 2024

Solar PV manufacturers Battery manufacturers

IEA. CC BY 4.0.
Notes: Includes the top five publicly listed firms by installed manufacturing capacity for solar PV modules and batteries.
Gaps in time series are filled with interpolations, and 2023 results are estimates where year-end filings are not available.
Quarterly data shown on an annualised basis in nominal terms.
Source: IEA analysis based on company financials from S&P Capital IQ database.

2 Top five publicly listed solar PV manufacturing companies by module manufacturing capacity operational in 2023 for which complete

time series data from Q1 2017 to Q3 2023 were available.

3 As per solar PV, with the caveat that some of the largest battery manufacturers are also large electric vehicle producers. In these

instances the financial indicators correspond to the overall company totals, including both battery and vehicle manufacturing operations.
I EA. CC BY 4.0.

PAGE | 20
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

Capital expenditure, a measure of overall investment by firms, which for these

companies is mostly in manufacturing plants and equipment, also grew rapidly in
the second half of the period analysed. To put the investment spending in context
with the wider manufacturing sector explored earlier in the chapter, the combined
investments from the ten firms examined – five for batteries and five for solar PV
– totalled around USD 160 billion per year at their peak. When compared with total
manufacturing sector investment in 2023 (around USD 6.4 trillion), the figure
represents a small share, at 2.5% globally. Given that all of the firms analysed are
headquartered in China – and carry out most of their manufacturing operations in
the country – a more relevant comparison is with Chinese manufacturing sector
investment, where the share rises to 6%. A more granular comparator still is the
investment in the manufacturing sub-sectors that account for many of the activities
associated with producing solar PV and battery components. At their peak, the
investments by the ten firms in aggregate account for around 30% of investment
in the electrical equipment and computer, electronic and optical products sub-
sectors (ISIC Divisions 26 and 27) in China, and 60% when stripping out domestic
appliances, consumer electronics and optical products.

While comparisons with samples of company financial data are instructive as to

broad trends in clean technology manufacturing, there are several limitations to
this type of analysis with respect to tracking progress in these industries. First, a
subset of companies is unlikely to be a consistent sample of a broader industry,
as their combined share of sectoral activity is likely to vary over time. Second, it is
difficult to obtain comprehensive data on all manufacturing entities, which can
undergo mergers and acquisitions, and transfer assets to (or acquire them from)
other players. Some companies are private, and not publicly listed, which
generally results in less financial information being made publicly available. Third,
data on manufacturing can include those corresponding to a variety of operations,
and not all companies are “pure play” clean technology manufacturers. This is
particularly the case for battery manufacturing, where, for example, the second
largest battery manufacturer globally (BYD) is also one of the largest electric
vehicle (EV) manufacturers. Tesla is also vertically integrated, with battery
manufacturing operations and investments in lithium refining. Vertical integration
is an increasingly common strategy in the EV supply chain, as it is for the various
steps in solar PV manufacturing. Fourth, little-to-no physical data are required to
be presented in company filings, which means analysis is generally limited to the
use of economic measures of input, output and profitability.

These limitations underscore the need to analyse clean technology manufacturing

operations on a physical basis, tracking where possible, capacity, production and
I EA. CC BY 4.0.

PAGE | 21
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

investments at the facility level. Chapter 2 comprises the latest instalment of our
analysis of progress in clean technology manufacturing, following Special
Briefings on the topic in 2023. 4

Clean technology manufacturing investment

New first-of-its-kind analysis by the IEA of clean technology manufacturing
investment indicates that the sector is booming. Investment in manufacturing of
the five clean energy technologies that this report focuses on reached
USD 200 billion in 2023, up from USD 115 billion in 2022, growing by more than
70%. Investments were dominated by solar PV and battery manufacturing
installations (including those for producing their main components), which together
accounted for 95% of the total in 2023. China accounted for three-quarters of the
investment in 2023, down from 85% in 2022. Both the United States and the
European Union made significant inroads in 2023, with their combined share of
total clean technology manufacturing investment reaching 16% in 2023, up from
11% in 2022. India, Japan, Korea and Southeast Asia made up most of the
remaining share, with virtually no manufacturing investment taking place in either
Africa or Central and South America.

Figure 5 Clean technology manufacturing investment by technology and region,

2022-2023
Clean technologies Solar PV Batteries
200 100 125
USD billion

80 100
150

60 75
100
40 50

50
20 25

0 0 0
2022 2023 2022 2023 2022 2023
Solar PV Batteries Wind Other United States European Union China India RoW

IEA. CC BY 4.0.
Notes: RoW = Rest of world. Solar PV includes facilities producing polysilicon, wafers, cells and modules; Batteries
includes facilities producing packs and cells, anodes and cathodes; Wind includes facilities producing nacelles, blades and
towers; Other includes electrolysers and heat pump manufacturing.
Sources: IEA analysis based on InfoLink, Thomson Reuters, Bloomberg New Energy Finance, Wood Mackenzie, S&P
Global Commodity Insights, EV Volumes, and Benchmark Mineral Intelligence.

4
IEA (2023), The State of Clean Technology Manufacturing, and IEA (2023), The State of Clean Technology Manufacturing
– November 2023 Update.
I EA. CC BY 4.0.

PAGE | 22
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

There are three tell-tale signs of continued momentum in clean technology

manufacturing investment going into the mid-2020s. First, we estimate that around
40% of the global clean technology manufacturing investment in 2023 was for
facilities that will come online in 2024. 5 For battery manufacturing facilities, the
equivalent figure is nearly 70%. Second, when looking in detail at the pipeline of
announced projects for clean technology manufacturing, 85% of committed (i.e.
those having reached final investment decision (FID) or under construction) solar
PV manufacturing projects and around one-third of battery manufacturing facilities
are scheduled to come online by 2025. Just these facilities, combined with those
already installed today, could produce around 150% (solar PV) and 55%
(batteries) of the global deployment levels in 2030 in the IEA’s Net Zero Emissions
by 2050 Scenario (NZE Scenario) (see Chapter 2). Third, a portion of the capacity
that is scheduled to come online by the end of the decade is seeing some degree
of financial commitment now. This momentum is also spreading to adjacent
sectors – nearly half of committed battery manufacturing announcements in the
United States will be via joint ventures with automakers.

This advanced spending commitment takes various forms, including land

purchases, ground works, preparation ahead of manufacturing facility construction
and the front-loaded investment in greenfield projects that later leads to less
capital-intensive brownfield manufacturing expansion. Some of these dynamics
are visible in our bottom-up investment estimates. For example, half of the global
battery manufacturing capacity envisaged for 2030, including both committed and
preliminary projects, is either an existing facility or a planned expansion thereof.
Others only show up in macroeconomic indicators. For example, construction
spending on US manufacturing facilities more than doubled between the
beginning of 2021 and the end of 2023, driven by the clean technology
manufacturing investments incentivised by the Inflation Reduction Act (IRA) and
Chips and Science Act. Some of this construction spend is on “site preparation
and outside construction of fixed structures or facilities such as sidewalks,
highways and streets”, which will take place well in advance of manufacturing
equipment being installed. Granular data on US manufacturing investment
spending from the Clean Investment Monitor shows a continuing upwards trend
for each quarter of 2023, led by battery facilities.

The corollary of the current investment boom is that today’s geographical

concentration is set to persist, in particular for solar PV, with committed capacity
potentially exceeding the 2030 deployment needs in the NZE Scenario. China
continued to account for the lion’s share of investments in solar PV manufacturing,

5
Assuming a 2-year period between FID and the plant becoming operational for all facilities analysed except solar PV module,
cell and wafer manufacturing facilities where 1-year is used. An even spending profile during this period is assumed, meaning
that an investment with a 2-year FID-to-operation period will see 50% of the spending take place in the year the facility
becomes operational and 50% the year before.
I EA. CC BY 4.0.

PAGE | 23
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

with investments growing by more than a factor of two between 2022

(USD 35 billion) and 2023 (USD 77 billion), but its share of the total remained flat
at 91%, due to similar rates of increases in the rest of the world. China’s high share
of manufacturing investments in components was even more pronounced in the
upstream steps of the solar PV supply chain, at 95% of the global total for wafer
production capacity investments, and 96% of investment in polysilicon production
facilities, compared with 83% for modules. This is despite the fact that China is
estimated to have the lowest cost of any country for the installation of these
facilities (see Chapter 3), which therefore require less investment per unit of
capacity. Many of the capacity expansions in 2023 in China were brownfield and
integrated (i.e. multiple process steps) facilities, which are generally lower-cost
per unit of capacity than greenfield and standalone facilities, respectively.

Battery manufacturing investment – including cell, anode and cathode

manufacturing – also showed strong growth in 2023, reaching USD 110 billion, up
from USD 70 billion in 2022. The locations for these investments were more
diverse than for solar PV. The combined investments in the European Union and
the United States more than tripled in absolute terms, and together their share
rose to more than a quarter of the global total, up from 14% in 2022. Given the
assumption that a battery manufacturing facility takes around 2 years to construct
following an FID, and a large swathe of facilities are projected to come online in
2024 and 2025 in these regions, much of this investment is attributable to facilities
that are not yet operational. Looking upstream in the battery supply chain, a similar
pattern can be observed as for the solar PV supply chain. While facilities outside
China accounted for around a quarter of battery cell manufacturing investments in
2022, and nearly half in 2023, China accounted for 98% of investments in facilities
for producing anodes in 2023 and 87% for cathodes.

Investment in other clean energy technologies – wind, heat pumps and

electrolysers – accounted for a much smaller fraction of total investment, at around
7% in 2022 and 4% in 2023. Investment in wind manufacturing, including nacelle,
blade and tower production facilities, fell slightly in absolute terms in 2023. China
accounted for virtually all of the investment in wind manufacturing facilities.
Electrolyser and heat pump manufacturing investments were the two areas where
the European Union and the United States accounted for a larger combined share
of investment than China in 2023, with virtually no investments in manufacturing
for these technologies taking place elsewhere.

The growth in clean technology manufacturing investment in 2023 was so

significant that it is starting to register in broader macroeconomic trends. Just the
direct investments in the facilities described above accounted for around 0.2% of
global GDP in 2023, doubling its share relative to 2022. When considering just the
contribution of investment to GDP – gross fixed capital formation – these shares
rise to 0.4% (2022) and 0.7% (2023). While these figures may seem small, they
I EA. CC BY 4.0.

PAGE | 24
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

are larger than the shares of global investment attributable to the entire aerospace
(0.2%), glass (0.3%) and steel (0.5%) industries; and approaching those of the
basic chemicals (0.9%) and pharmaceuticals (1.1%) industries. Moreover, the
investment in these comparatively mature industries grew only incrementally in
2023, where it grew at all, whereas investment in clean technology manufacturing
grew by more than 70%. Measured as a share of overall global investment growth
in 2023, clean technology manufacturing accounted for nearly 10%, and as a
share of global GDP growth it accounted for around 4%.

Figure 6 Share of clean technology manufacturing in global investment and growth

thereof in comparison to other manufacturing sub-sectors, 2022-2023

Pharmaceuticals

Basic chemicals

Steel

Glass

Aerospace

CTM

0% 2% 4% 6% 8% 10%

2022 2023 2023 growth

IEA. CC BY 4.0.
Notes: CTM = Clean technology manufacturing. Shares of investment calculated as sectoral investment divided by gross
fixed capital formation on a global basis. Sectors correspond to the following ISIC codes: ‘Pharmaceuticals’ = Division 21,
‘Basic chemicals' = Group 201, ‘Steel’ = Groups 241-243, ‘Glass’ = Group 231, ‘Aerospace’ = Group 303.
Source: IEA analysis based on OECD TiVA database and Oxford Economics Global Industry Service.

Box 3 IEA stakeholder engagement on clean technology manufacturing

Manufacturing in clean technology industries is highly concentrated geographically,
and accelerating the clean energy transition will require an expansion of global
manufacturing capacity to other countries and regions. If governments are to make
progress towards establishing secure, resilient and sustainable supply chains for
the critical components and materials for clean energy transitions, they will need
carefully designed industrial strategies that unlock investment, while at the same
time maintaining competitive markets and international trade.
In light of these considerations, on 6 November 2023, the IEA hosted a High-Level
Dialogue that convened experts from government, industry, research, development
I EA. CC BY 4.0.

PAGE | 25
Advancing Clean Technology Manufacturing Part I: Chapter 1
An Energy Technology Perspectives Special Report

institutions and civil society to discuss ways to make clean technology

manufacturing and its supply chains more resilient. The workshop provided a
platform for stakeholders to share experience and priorities for developing and
building out manufacturing bases at the country and regional level. The need for
policy clarity to provide stability to attract private sector investment, and the use of
trade as a tool to support progress towards climate goals, were central themes
throughout the workshop. The outcomes of the discussions provided invaluable
input to the design and considerations of this report.
In addition, the IEA is undertaking an industry survey to gather evidence on the
factors that influence company investment decisions across the supply chain, and
across the world. Initial responses to the survey have been used to inform the
analysis in this report, and will be presented in various IEA publications throughout
the year, notably the forthcoming Energy Technology Perspectives report.

I EA. CC BY 4.0.

PAGE | 26
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Chapter 2. The new clean energy

economy is emerging

Getting on a path to net zero emissions by 2050, in line with global climate goals,
will require a substantial, accelerated expansion of clean energy technology
manufacturing. This has implications for every step of the manufacturing supply
chain, from mining and raw material processing, through to component
manufacturing and final assembly. Against this backdrop, recent years have seen
clean technology manufacturing becoming increasingly dynamic, as successive
major governments have placed it at the centre of new industrial strategies.

This chapter focuses on the latest developments on manufacturing capacity

through 2023 for five key technologies for the clean energy transition: solar PV,
wind, batteries, electrolysers and heat pumps. 6 Together, these technologies
account for almost 40% of the emissions savings that need to be achieved by 2030
in the IEA’s NZE Scenario. We track announcements relating to capacity additions
across different stages of development, and compare existing and announced
capacity to deployment levels envisaged by government targets for 2030 and
under the NZE Scenario. It builds on the analysis presented in two Special
Briefings on clean energy technology manufacturing released in May and
November 2023, with new data on announcements up to the end of 2023.

The project pipeline continues to expand

Global clean technology manufacturing capacity registered strong growth across
several technologies and regions in 2023. Some technologies, like solar PV and
batteries, saw record annual increases on the back of unprecedented
development progress in recent years. Some have now become the most cost-
competitive options available, such as solar PV, for which electricity generation
costs are now lower than fossil fuel-based alternatives in most countries. The
prices of electric cars are falling as competition intensifies, especially in China,
though they remain more expensive than internal combustion engine vehicles in
other markets. In addition, the global energy crisis has contributed to accelerated
deployment of heat pumps, and of electrolysers for producing low-emissions
hydrogen, particularly in Europe. In many instances, electrolysers and heat pumps
remain more expensive than their fossil fuel counterparts. However, their role in
helping to reduce dependency on fossil fuels has made them prominent targets

6
See the Technical annex for an explanation of the analytical boundaries used in this report.
I EA. CC BY 4.0.

PAGE | 27
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

for government industrial strategies and incentive schemes, particularly in the

United States, the European Union and China.

Heat pumps are the only technology among the five covered in depth in this report
for which manufacturing capacity growth slowed in 2023. This was a consequence
of stagnation across the majority of leading heat pump markets, with sales and
installations declining in the European Union, the United States and Japan amid
higher interest rates and inflation. Sales continued to grow in China, which is
currently the largest heat pump market.

Clean technology manufacturing capacity additions in 2023 were also heavily

concentrated in three major markets – the United States, the European Union and
China. While Central and South America account for non-trivial shares of the
production of the main wind turbine components (3-8% of global production for
nacelles, blades and towers), virtually no clean technology manufacturing takes
place in Africa today.

Box 4 Scenarios used in this report

Analysis in this Special Report is underpinned by global projections of clean energy
technologies derived from the IEA’s Global Energy and Climate (GEC) model, a
detailed bottom-up modelling framework composed of several interlinked models
covering energy supply and transformation, and energy use in the buildings,
industry and transport sectors. The modelling framework includes 29 regions or
countries covering the whole world.
The most recent year of complete historical data from the GEC model is 2023, to
which year-end 2022 and 2023 manufacturing installed capacity data have been
added as part of the analysis for this Special Report. For projected values to 2030,
we make use of two IEA scenarios produced using the GEC model that describe
possible energy system pathways:
The Net Zero Emissions by 2050 Scenario (NZE Scenario) is a normative scenario
that sets out a pathway to stabilise global average temperatures at 1.5°C above
pre-industrial levels. The NZE Scenario achieves global net zero energy sector CO2
emissions by 2050 without relying on emissions reductions from outside the energy
sector. In doing so, advanced economies reach net zero emissions before
developing economies do. The NZE Scenario also meets the key energy-related
UN Sustainable Development Goals, achieving universal access to energy by 2030
and securing major improvements in air quality.
The Announced Pledges Scenario (APS) assumes that governments will meet, in
full and on time, all the climate-related commitments they have announced,
including longer-term net zero emissions targets and Nationally Determined
I EA. CC BY 4.0.

PAGE | 28
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Contributions, as well as commitments in related areas such as energy access. It

does so irrespective of whether these commitments are underpinned by specific
policies to secure their implementation. Pledges made in international fora and
initiatives on the part of businesses and other non-governmental organisations are
also taken into account wherever they add to the ambition of governments.
Neither scenario should be considered a prediction or forecast. Rather, they are
intended to offer insights into the impacts and trade-offs of different technology
choices and policy targets, and to provide a quantitative framework to support
decision-making in the energy sector, and strategic guidance on technology choices
for governments and other stakeholders. The scenarios and results are consistent
with those presented in the World Energy Outlook 2023.

Latest announcements present a varied picture of

manufacturing for net zero deployment needs
With existing solar PV module and cell manufacturing capacity alone, the
deployment levels for solar PV envisaged by the IEA’s NZE Scenario in 2030 could
already be achieved today, were it to be run at near-full capacity across all existing
facilities – 6 years ahead of schedule. Ensuring that this existing manufacturing
capacity is used to its full potential would therefore require an acceleration in
deployment. There are, however, still shortages at the upstream end of the solar
PV supply chain: current capacity for producing wafers and polysilicon is not yet
fully sufficient to meet 2030 deployment needs in the NZE Scenario.

For the other technologies considered here, existing manufacturing capacity could
already deliver between 15% (in the case of electrolysers) and close to 50% (in
the case of wind energy) of the NZE Scenario deployment needs by 2030 at the
time of writing (Figure 7). I EA. CC BY 4.0.

PAGE | 29
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Figure 7 Output from existing and announced manufacturing capacity relative to Net
Zero Emissions by 2050 Scenario deployment in 2030

150%

2030 NZE Scenario deployment

100%

50%

0%
Solar PV Wind Batteries Electrolysers Heat Pumps

Output from existing Output from announced

manufacturing capacity: manufacturing capacity:
2022 output 2023 output Increased utilisation Committed Preliminary

IEA. CC BY 4.0.
Notes: NZE Scenario = Net Zero Emissions by 2050 Scenario. 2022 and 2023 output values reflect estimates of actual
utilisation rates. Increased utilisation refers to the gap between 2023 production levels and existing capacity being utilised
at 85%. A utilisation rate of 85% is used for both existing and announced manufacturing capacity in 2030. Refer to the
Technical annex for more details on the analytical boundaries and methodologies used in this analysis.
Sources: IEA analysis based on data from Benchmark Mineral Intelligence, Bloomberg New Energy Finance, EV Volumes,
InfoLink, S&P Global Commodity Insights, UN Comtrade, Wood Mackenzie and announcements by manufacturers and
personal communications.

This outlook changes significantly if all announcements for manufacturing capacity

expansion are taken into account. For solar PV modules, based only on
announced expansions that are already committed (see Box 5), output could
exceed 2030 requirements of the NZE Scenario by more than 50%. If all
announced expansions are considered, including those that have not yet reached
a final investment decision (FID), output rises to over 60% more than envisaged
demand. This comes with both benefits and drawbacks. On the downside, it points
to a significant level of surplus capacity, which may lead to stranded or under-
utilised assets, and has also led to intense competition among manufacturers,
resulting in a significant module spot price drop over 2023. This has already
resulted in cancellations and downward revisions of planned expansions in solar
PV manufacturing, especially for PV modules. However, the new manufacturing
sites are likely to produce new-generation components with improved
characteristics, as a consequence of growing competition and technology
innovation. New capacity could therefore outcompete existing manufacturing
capacity, reduce the risk of over-capacity and lead to the commercialisation of
products that are cheaper and perform better. This effect could be multiplied if the
prospect of new capacity leads to upgrades in existing manufacturing capacity that
is based on older cell technologies.
I EA. CC BY 4.0.

PAGE | 30
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Box 5 Manufacturing data categories

In this Special Report the manufacturing data for the focus five clean energy
technologies can be categorised as follows:
“Installed manufacturing capacity” refers to the maximum rated output of facilities
for producing a given technology. Capacity is stated on an annual basis for the final
product and does not refer to the capacity for any intermediate products or
components. Where available, manufacturing capacity for key components is
provided separately. Annual manufacturing “output” is a fraction of the installed
manufacturing capacity. Output depends on the utilisation rate of production
capacity, for which 85% is a typical annual average target level under normal
operation. However, utilisation rates for clean technology manufacturing facilities
tend to be much lower on average today, reflecting significant degrees of capacity
surplus globally. The year 2023 – the base year for the analysis in this Special
Report – is the most recent year for which installed manufacturing capacity data has
been collected.
“Announced projects” refers to the aggregate stated capacity – or estimated
output of that capacity (assuming a default utilisation rate of 85%) – of potential
manufacturing facilities that have been announced. This includes projects for
building new facilities or expanding existing ones that are at different stages of
development. “Committed” projects include those that have already reached an
FID, or are under construction, whereas “preliminary” projects include those that
have not yet reached an FID, meaning feasibility studies or earlier steps are
underway. Wherever data is available, we distinguish committed projects from
preliminary announcements across the key technologies in focus, which allows for
more robust projections of future manufacturing capacity. Unless otherwise stated,
the announced projects dataset assembled for this Special Report comprises
announcements dated up to the end of 2023.

In the case of batteries and electrolysers, if all announcements for expansion are
realised, it will be possible to achieve the level of deployment of the NZE Scenario
by 2030, although the maturity of announcements differs. For batteries, the growth
trajectory is quite clear: committed expansions (which account for over 60% of
total announcements) are already sufficient to match more than 90% of the 2030
global deployment needs in the NZE Scenario. When also considering other
announced (though not yet committed) projects, the pipeline of new manufacturing
capacity for batteries comfortably exceeds deployment requirements in the NZE
Scenario by around 30% in 2030.

For electrolysers, the outlook is less certain: all announced expansions would
need to be realised to meet deployment needs in the NZE Scenario, but only 13%
(19 GW) are already committed (i.e. have reached FID). This share reflects
I EA. CC BY 4.0.

PAGE | 31
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

advances made in 2023, having increased from 7% (6 GW) in 2022, but

uncertainty around future demand for low-emissions hydrogen continues to limit
the rate of progress. In contrast to the maturing market for electric vehicles – which
is resulting in more certainty about demand for batteries – the cost of electrolytic
hydrogen production remains high when compared to alternative technologies. In
addition, policies to support project development and stimulate demand are being
implemented slowly, although a growing number of countries are moving into the
implementation phase of their hydrogen strategies, which could trigger further
deployment.

Wind and heat pump manufacturing currently present the least optimistic outlook.
In the case of wind, the majority of announced expansions are already committed,
but the output from these facilities will be able to deliver only around 60% of what
is needed in the NZE Scenario. For heat pumps, announcements have slowed
down in quantity and size over 2023, and even if all announcements for expansion
(including those that are still preliminary) are implemented in full and on time, it
will only be possible to manufacture around one-third of the heat pumps needed
in the NZE Scenario by 2030. However, it should be noted that announced
expansions of heat pump manufacturing capacity are only common in Europe.

Geographical concentration is expected to persist

despite the growing number of expansion
announcements
Level of geographical concentration is an important indicator of the robustness of
a supply chain. Highly concentrated supply chains – or individual steps within the
supply chain – are more vulnerable to disruption in the case of unforeseen events
such as natural disasters, unexpected events or accidents (e.g. the closure of the
Suez Canal due to a shipping collision) or geopolitical conflicts and price distortion
by non-market conditions. In the case of clean energy technology manufacturing,
geographical concentration is also an indicator of the extent to which individual
countries or regions are set to reap a potential economic benefit from clean energy
transitions.

All the technologies under the scope of this report currently present a high level of
geographical concentration in manufacturing, with the three largest producing
countries or regions accounting for around 80% or more of the capacity in all
cases. If all the announced expansions are realised, the situation is expected to
remain the same through 2030, with only minor variations in the relative shares of
the main three producing countries or regions for each technology.

Solar PV manufacturing is the most regionally concentrated of all the key

technologies analysed, with more than 80% of capacity located in China, driven
by the relatively low production costs across the full supply chain. The
I EA. CC BY 4.0.

PAGE | 32
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

United States and India are expected to increase their share of global
manufacturing capacity from now to 2030, but these expansions do not
significantly dent China’s share, which is likely to remain around 80%.

Battery manufacturing is also highly geographically concentrated today, with

China accounting for more than 80% of the manufacturing capacity, followed by
the United States and the European Union, with around 5% each. The capacity
accumulated by these three regions is expected to remain above 90% through
2030, but the share of China could fall to around 60%, as the European Union and
the United States nearly triple their shares, thanks to boost from ambitious policies
such as the Important Projects of Common Interest (IPCEI) and Net-Zero Industry
Act (NZIA) in the European Union and the US IRA. In the case of the
United States, actual expenditure on EV battery manufacturing from 2020 to the
end of the third quarter of 2023 totalled over USD 40 billion. In addition, half of the
committed manufacturing capacity in the United States will be delivered by joint
ventures between battery manufacturers and automotive original equipment
manufacturers (OEMs), which demonstrates that automakers are committed to
electric vehicles over the long term.

With regards to the concentration of manufacturing for electrolysers, wind and heat
pumps, the 2030 outlook is little changed from today. In the case of electrolysers,
despite the significant growth that would be achieved if all announcements are
realised, China, the European Union and the United States will still be home to
around 80% of all capacity. However, this situation could still change significantly:
around 20% of all announced expansions of manufacturing capacities have no
specified location.

For wind and heat pumps, the distribution of manufacturing in 2030 is little different
to today, as a consequence of the very limited number of announced expansions.
The share of wind manufacturing in China looks set to grow, reducing the current
share of the other major manufacturing regions and countries. For heat pumps,
the share of manufacturing in Europe will grow the most on the basis of announced
capacity additions, although this may be a reflection of expansions being
announced more prominently in Europe.
I EA. CC BY 4.0.

PAGE | 33
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Figure 8 Geographical concentration of current and announced manufacturing

capacity, 2023-2030
100%

80%

60%

40%

20%

0%
2023 2030 2023 2030 2023 2030 2023 2030 2023 2030
Solar PV (modules) Wind (nacelles) Batteries (cells) Electrolysers Heat pumps

China United States European Union Viet Nam India Rest of world

IEA. CC BY 4.0.
Notes: 2030 value includes all operational capacity in 2023 together with the capacity of announced manufacturing projects
through to 2030. For electrolysers, the analysis only includes projects for which location data was available. Shares are
based on manufacturing capacity. Refer to the Technical annex for more details on the analytical boundaries and
methodologies used in this analysis.
Sources: IEA analysis based on data from Benchmark Mineral Intelligence, Bloomberg New Energy Finance, EV Volumes,
InfoLink, S&P Global Commodity Insights, UN Comtrade, WoodMac and announcements by manufacturers and personal
communications.

Rapid – if uneven – progress

Announcements on capacity additions paint a varied picture of the potential for
manufacturing to scale up in line with 2030 deployment in the NZE Scenario. We
now turn to consideration of how announcements compare to government
ambitions through 2030, as envisaged by their announced pledges and
commitments.

Steep growth in solar PV manufacturing capacity, though

utilisation rates remain low
Global manufacturing capacity for solar PV modules increased dramatically in
2023, by almost 500 GW, with the vast majority – nearly 440 GW – added in China.
Output also grew to around 560 GW, compared to around 360 GW in 2022.
However, there was a slight decrease in average utilisation rates across PV
module manufacturing facilities, which hovered around 50% in 2023, with facilities
for newer technologies, like Tunnel Oxide Passivated Contact (TOPCon) cells,
having higher utilisation rates than older ones.

The extensive capacity added or announced in recent years has allowed output to
grow substantially, outpacing demand. This has driven down prices, and reduced
utilisation rates, meaning that production costs per unit increase. The result has
I EA. CC BY 4.0.

PAGE | 34
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

been a downscaling of expansion plans, including delays and cancellations, as

well as job cuts. Indicatively, most of the global downscaling appears to come from
China. For example, the solar panel maker Changzhou EGing Photovoltaic
Technology announced it would put on hold the expansion at its TOPCon solar
cell manufacturing base in Anhui province, and companies from other sectors that
planned to diversify into solar have abandoned their plans. Recently, the world’s
largest solar manufacturer, Longi, revealed that it will reduce its workforce.
However this downscaling is not limited to China, as evidenced by other facilities
that have been cancelled or shut down, like the Meyer Burger plant in Germany
or the CubicPV startup in the United States.

Furthermore, low-priced solar PV module imports, primarily from China, and to a

lesser extent from Southeast Asia and destined for the United States, have led to
the accumulation of significant inventories by European and North American
developers, leading to concerns about the future competitiveness of upcoming
production.

Overall, announced capacity additions have been revised downwards across the
supply chain (with the exception of polysilicon, at least for the time being, as further
detailed below). Despite this, the picture varies depending on the region: India will
be able to achieve levels consistent with the APS in 2030 on the basis of
announced capacity, and China already exceeds the APS levels in 2030 today,
indicating significant capacity for exports. As a consequence, China is likely to
remain the lead exporter of solar panels (and their subcomponents) in the near
term. China currently produces twice as much solar PV (modules) as it installs,
supplying regions in which manufacturing capacities are expected to remain well
below their deployment needs in the APS, such as the United States and the
European Union. The projected Chinese surplus output in 2030 (i.e. beyond
output required to meet its domestic needs in the APS) alone could easily
accommodate the global APS demand for installations in the same year. I EA. CC BY 4.0.

PAGE | 35
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Figure 9 Output from existing and announced solar PV manufacturing capacity in

selected regions relative to deployment in the Announced Pledges Scenario
in 2030
1 200 120
GW

2030 APS domestic

deployment

Output from announced

manufacturing capacity:
800 80 Preliminary

Committed

Output from existing

manufacturing capacity:
400 40 Increased utilisation

2023 production

2022 production

China United States European Union India

IEA. CC BY 4.0.
Notes: APS = Announced Pledges Scenario. 2023 production values reflect estimates of actual utilisation rates. Increased
utilisation refers to the gap between 2023 production levels and existing capacity being utilised at 85%. A utilisation rate of
85% is used for both existing and announced manufacturing capacity in 2030. Refer to the Technical annex for more
details on the analytical boundaries and methodologies used in this analysis.
Source: IEA analysis based on data from InfoLink and Bloomberg New Energy Finance.

Solar PV component-level concentration intensified in 2023

Solar PV is a paradigmatic example of a technology that presents a high level of
geographical concentration across the whole supply chain: China accounts for
more than 90% of cell, wafer and polysilicon manufacturing. Concentration
intensified in 2023 across almost all steps, even though it was already high in
2022. Indicatively, more than 90% of existing polysilicon manufacturing capacity
is in China, whereas 5 years ago the share was less than 60%.

The geographical distribution of prospective manufacturing capacity did not

change significantly on the basis of announcements made in 2023, with China
continuing to account for around 80% of planned and existing capacity for
modules, followed by the United States and India with 5%, and Europe with just
1%. However, both India and the United States are mostly expanding in module
and cell manufacturing, and to a lesser extent in the upstream components of the
solar PV supply chain.

The high geographical concentration of the full solar PV supply chain is unlikely to
change significantly on the basis of announced projects, with China’s share of
capacity for modules, cells and wafers decreasing marginally (e.g. to 90% for
wafers) and increasing for polysilicon, to reach close to 95% in 2030. Locally
I EA. CC BY 4.0.

PAGE | 36
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

controlled polysilicon manufacturing can provide a competitive edge, as the

energy-intensive commodity is traded internationally. Companies therefore have
an incentive to expand vertically in order to cover multiple steps of the supply chain
and reduce their exposure to fluctuations in prices for key inputs.

Figure 10 Output from existing and announced solar PV component manufacturing

capacity and 2030 deployment levels in the Announced Pledges Scenario
and Net Zero Emissions by 2050 Scenario

2 000
GW

announcements
Committed

RoW
Existing capacity

1 600 Other Asia Pacific

utilised at 85%

Europe
United States
1 200
China

800 2030 NZE deployment

2030 APS deployment

400

Modules Cells Wafers Polysilicon

IEA. CC BY 4.0.
Notes: APS = Announced Pledges Scenario; NZE = Net Zero Emissions by 2050 Scenario; RoW = Rest of World. A
utilisation factor of 85% is assumed for all regions. Refer to the Technical annex for more details on the analytical
boundaries and methodologies used in this analysis.
Source: IEA analysis based on data from PV InfoLink, Bloomberg New Energy Finance, IEA PVPS, SPV Market Research,
and RTS Corporation.

China accounts for the largest share of announced wind

manufacturing capacity additions
Manufacturing output for the nacelles, towers and blades that make up both
onshore and offshore wind turbines remained largely unchanged through 2023,
with China alone significantly increasing production. Across all components, China
saw an average year-on-year increase of more than 30%, and remains the largest
producer of nacelles, blades and towers. The world’s second largest
manufacturer, the European Union, saw no increase in manufacturing capacity
from 2022 to 2023.

China also accounts for the largest share of announced capacity additions to 2030,
with 42 GW announced for nacelles, 35 GW for blades and nearly 20 GW for
towers. Across all components, nearly all of the announced capacity in China is
already committed – over 90% for nacelles, 85% for blades and 100% for towers.
There were very few announcements of capacity additions for blade and nacelle
I EA. CC BY 4.0.

PAGE | 37
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

manufacturing outside of China in 2023. However, there were notable

announcements for tower manufacturing in the European Union and
United States. Towers are more likely than other components to be manufactured
closer to demand centres, due to the difficulty of transporting such large structures
over long distances.

In the European Union, offshore towers account for the largest share of
manufacturing, with around 10 GW of offshore tower manufacturing capacity
compared to 6 GW and 8 GW for offshore blade and nacelle manufacturing,
respectively. In contrast, in China, offshore nacelle and blade manufacturing
capacity is over 30 GW for each, whereas offshore tower manufacturing capacity
reaches 20 GW. It is likely that many of these tower manufacturing facilities are
local steel companies fabricating these structures on demand, as the number of
facilities dedicated to these components specifically is lower than blades and
nacelles facilities.

China’s manufacturing capacity is tracking ahead of 2030 wind deployments

envisaged by announced policies, by 50 GW for blades and 65 GW for nacelles.
Notably, China’s 2030 target for cumulative solar PV and wind capacity was close
to being reached in the first quarter of 2024, 6 years ahead of schedule. This
opens up potential for exports to other markets, especially given that other
countries and regions largely do not have the manufacturing capacity across
different components to meet their deployment pledges by 2030 on the basis of
existing capacity and announced additions. By 2030, China would be able to
provide 50% of the blades and almost 60% of the nacelles needed to close the
gap between deployment needs in the rest of the world in the APS and the
manufacturing capacity in those regions.

This gap is most prominent in the United States, where existing production and
announced capacity additions for blades and nacelles would result in a shortfall of
more than 30 GW (over 70%) between total output and deployment needs
consistent with announced targets in 2030.

In the European Union, existing and announced capacity for blades and nacelles
is almost 30% lower than would be needed to meet 2030 deployment needs
envisaged in the APS, leading to a gap of 12 GW and 17 GW for nacelles and
blades, respectively.

However, after a challenging 2022, the early signs from 2023 annual financial
reports released to date suggest that several European OEMs may have turned a
corner, with Vestas and Nordex confirming a return to growth. In early 2024,
Vestas broke ground on a new factory for manufacturing offshore nacelles and
hub assembly in Poland, expected to begin operation in 2025. Vestas also
announced plans to establish a new blade factory. Furthermore, in the
United Kingdom, a new Offshore Wind Industrial Growth Plan details actions to
I EA. CC BY 4.0.

PAGE | 38
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

triple current manufacturing capacities, highlighting the opportunities for

expanding tower and blade manufacturing capabilities, while expanding nacelle
assembly is made a lower priority.

While China, the European Union and the United States remained the largest
manufacturers for wind in 2023, India also increased production, and has more
than 60% of the capacity needed to meet 2030 domestic deployment needs in the
APS for nacelles and towers. India is also emerging as an alternative export hub
for blades in the near term, as it is currently oversupplied for the deployment needs
of coming years, although it would come 2 GW short on its domestic targets for
2030. Elsewhere, many Chinese and European OEMs have made investments in
wind manufacturing facilities in Latin America in recent years, principally in
Argentina and Brazil. Attention is turning to capacity for recycling wind turbines at
the end of their lifespan, with six dedicated factories in Europe for recycling blades
announced by Continuum in early 2023.

Figure 11 Output from existing and announced wind manufacturing capacity in

selected regions relative to Announced Pledges Scenario deployment in
2030
150 45
GW

2030 APS domestic

deployment

100 30

50 15

0 0
Blades
Nacelles

Towers

Blades

Blades

Blades
Nacelles

Towers

Nacelles

Towers

Nacelles

Towers

China United States European Union India

Outpu from existing Output from announced
manufacturing capacity: manufacturing capacity:
2023 production Increased utilisation Committed Preliminary

IEA. CC BY 4.0.
Notes: APS = Announced Pledges Scenario. The figure includes data on facilities which are specifically dedicated to wind
manufacturing for blades, nacelles and towers, except for tower manufacturing in China, where an implausible shortfall is
assumed to be met by additional generic fabrication capacity. 2023 production values reflect estimates of actual utilisation
rates. A utilisation rate of 85% is used for both existing and announced dedicated manufacturing capacity in 2030. Refer to
the Technical annex for more details on the analytical boundaries and methodologies used in this analysis.
Source: IEA analysis based on data from S&P Global Commodity Insights.
I EA. CC BY 4.0.

PAGE | 39
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Battery manufacturing capacity reaches new highs,

mostly in major electric vehicle markets
Battery production has ramped up quickly in the past few years to meet increasing
demand resulting from growth in electric car sales. In 2023, battery manufacturing
capacity reached 2.5 TWh, with 780 GWh of new capacity added relative to 2022.
The capacity added in 2023 was 25% higher than that added in 2022.

Global battery manufacturing capacity could exceed 9 TWh by 2030 if all

announcements are completed in full and on time. About 70% of the 2030
projected battery capacity worldwide is already operational or committed, though
announcements differ across regions. Over 40% of future manufacturing capacity
in China relies on the expansion of current plants, indicating the strengthening of
industrial actors that are already part of the Chinese market. In contrast, 80% of
US and EU manufacturing capacity is expected to come from new plants, with a
significant number of new actors entering those markets in the coming years.

Much of the currently announced battery manufacturing capacity remains

concentrated in today’s major EV markets – China, the United States and the
European Union – which are all set to have enough capacity to reach their
announced pledges for 2030. Of course, as EVs and battery storage increasingly
reach global markets, and battery demand diversifies geographically, there will be
new opportunities to be seized around the world to produce batteries near demand
centres. Locating battery manufacturing close to EV manufacturing hubs would
reduce exposure to import/export tariffs, as well as insurance costs associated
with shipping lithium-ion batteries.

Outside of today’s major EV markets, announced manufacturing capacity – of

which 85% is already committed – meets around half of APS needs in 2030 in
those regions. Almost all of this committed manufacturing capacity is divided
among other European countries and Canada (with about 35% each), India (12%),
other Southeast Asian countries (8%), like Malaysia, Viet Nam, and Singapore,
and Japan and Korea (5%). Korea and Japan, however, also account for over 80%
of today’s capacity in these regions.

There is considerable space for growth in South American countries, which today
have no significant announcements for battery manufacturing capacity through
2030, and for countries with manufacturing capacity that falls short of their
pledges, such as India, whose announced capacity would cover only one-quarter
of its 2030 demand in the APS. These gaps could increase the risk of countries
failing to meet their long-term decarbonisation targets, and could have important
implications for future battery trade.
I EA. CC BY 4.0.

PAGE | 40
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Figure 12 Output from existing and announced battery manufacturing capacity in

selected regions relative to Announced Pledges Scenario deployment in 2030
6 000 1 500
GWh

2030 APS domestic

deployment

Output from announced

manufacturing capacity:
4 000 1 000 Preliminary

Committed

Output from existing

manufacturing capacity:
2 000 500 Increased utilisation

2023 production

2022 production

China United States European Union Rest of the World

IEA. CC BY 4.0.
Notes: APS = Announced Pledges Scenario. 2023 production values reflect estimates of actual utilisation rates. Increased
utilisation refers to the gap between 2023 production levels and existing capacity being utilised at 85%. A utilisation rate of 85%
is used for both existing and announced manufacturing capacity in 2030. Demand refers to both EV battery and stationary
storage demand. Battery capacity refers to battery cells. Battery refers to lithium-ion batteries. Refer to the Technical annex for
more details on the analytical boundaries and methodologies used in this analysis.
Source: IEA analysis based on data from Benchmark Mineral Intelligence, Bloomberg New Energy Finance and EV Volumes.

China currently has a leading role in battery production, accounting for over
650 GWh in 2023, or almost 80% of the global total. However, this also comes
with surplus capacity: In 2023, China used less than 40% of its maximum battery
cell production capacity. China’s projected battery manufacturing capacity alone
in 2030 could cover global demand in the APS in the same year. If only its already
committed capacity is considered, China could cover about 85% of 2030 global
demand under the APS. China is currently the world’s largest exporter of EV
batteries, accounting for about 70% of total exports in 2023, but surplus capacity
has also significantly reduced producers’ margins, which may put some at risk if
they do not find enough customers outside of China.

High levels of capacity over the entire battery supply chain, above and beyond
levels of demand, can put downward pressure on prices. This is attractive for end
consumers and can help boost uptake towards a level needed to meet climate
targets, but it also lowers cash flows and provides smaller margins for mining,
refining and manufacturing companies. For example, the drop in battery material
prices in 2023 led to a 14% decrease in average battery pack price, but it also put
I EA. CC BY 4.0.

PAGE | 41
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

at risk several mining companies, with many of them now struggling to stay afloat
and announcing spending and job cuts in 2024. 7

Battery component manufacturing remains heavily concentrated,

but capacity surplus may lead to greater diversification by 2030
While final battery manufacturing becomes less geographically concentrated
through 2030, the project pipeline for battery components shows little sign of
diversification. The manufacturing of lithium-ion batteries requires a stable, high-
quality supply of cathode and anode materials. Their production is heavily
concentrated in China, which currently accounts for nearly 90% of global capacity
for cathode active materials, and over 97% of capacity for anode active materials.
China also accounts for more than 85% of both committed and preliminary
capacity additions announced for cathodes and anode active materials by 2030.

Different supply chains are, however, required for different battery chemistries, with
lithium-iron phosphate (LFP) dominating the Chinese market and lithium nickel
manganese cobalt oxide (NMC) the European and North American markets. China is
home to about 100% of the LFP production capacity, and more than three-quarters
of the installed production capacity for NMC, followed by Korea, with about 20%.

Figure 13 Output from existing and announced battery component manufacturing

capacity in selected regions relative to Announced Pledges Scenario and
Net Zero Emissions by 2050 Scenario deployment in 2030
20
TWh

RoW
Other Asia Pacific
15
Europe
announcements

United States
Committed

China
Existing capacity

10
utilised at 85%

2030 NZE deployment

5
2030 APS deployment

Cells Cathode Anode

IEA. CC BY 4.0.
Notes: APS = Announced Pledges Scenario; NZE = Net Zero Emissions by 2050 Scenario; RoW = Rest of World. A
utilisation factor of 85% is assumed for all years and regions. Battery capacity refers to battery cells. Calculations for
cathode and anode assume a cathode and anode materials energy density of around 670 Wh/kg (NMC and lithium nickel
cobalt aluminium oxide cathode active material), 465 Wh/kg (LFP cathode active material) and 1 500 Wh/kg (graphite
anode active material), respectively. Demand refers to both EV battery and stationary storage demand. Battery and battery
components refer to lithium-ion batteries. Cathode and anode refer to cathode and anode active materials. Refer to the
Technical annex for more details on the analytical boundaries and methodologies used in this analysis.
Source: IEA analysis based on data from Benchmark Mineral Intelligence, Bloomberg New Energy Finance and InfoLink

7
See: IEA (2024), Global EV Outlook 2024.
I EA. CC BY 4.0.

PAGE | 42
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

The 2030 projected manufacturing capacity for cathode active materials is about
two times greater than the projected battery cell manufacturing capacity in the
same year. In the case of anode active materials, this ratio increases to five times
greater, raising doubts about whether all manufacturers will be able to remain
competitive in the face of such a surplus.

The prospect of capacity surplus may also open the door to a more diversified
supply chain: In 2030, cathode and anode active material manufacturing capacity
outside of China can potentially cover up to 70% of the maximum demand for
battery production within those regions. Nonetheless, this cannot be taken for
granted: of the installed capacity and announcements outside of China, over 60%
of the cathode active material capacity, and over 90% of the anode active material
capacity, is still at the announcement stage and has not yet started construction,
underlining the need for close attention to this part of the battery supply chain.

Of course, manufacturing capacity is not the only parameter determining whether

battery manufacturers choose one supplier rather than another. The main
challenge for Chinese manufacturers in the coming years will be finding big
enough export markets to use their massive manufacturing capacity surplus and
increase currently low margins, while manufacturers in regions like the
European Union and the United States will need to demonstrate their cost
competitiveness. The quality, cost and characteristics of the cells and components
provided by different suppliers, together with regulations on local content
requirements, and environmental, social and governance (ESG) standards, will be
key to determining the winners and losers in these markets.

Sustained interest in electrolysers is encouraging,

though the outlook remains uncertain
Current manufacturing capacity for electrolysers increased to about 23 GW per
year at the end of 2023, up from more than 12 GW in 2022 (Table 1). However,
this figure is based on the announced nominal capacity of each facility, which in
some cases may only be reached after a few years of operation. Manufacturing
capacity remains geographically concentrated, with China accounting for 60% of
2023 capacity, followed by Europe with 20% and the United States with 16%.
I EA. CC BY 4.0.

PAGE | 43
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Table 1 Selected electrolyser manufacturing facilities commissioned in 2023

Announced
Company Location Country Technology
capacity
Cummins Fridley United States 500 MW PEM

Plug Power Rochester United States 1.2 GW PEM

Siemens Energy Berlin Germany 1 GW PEM

HydrogenPro Tianjin China 500 MW ALK

Sunfire Solingen Germany 500 MW ALK

E-Gen Energy Shanghai China 100 MW SOEC

Notes: PEM = proton exchange membrane electrolyser; ALK = alkaline electrolyser; SOEC = solid oxide electrolyser cell.

Based on announcements made in 2023, almost 170 GW of cumulative installed

manufacturing capacity could be reached by 2030, an increase on the 102 GW
that had been announced at the end of 2022. However, close to 90% of the
announced capacity is at a preliminary stage of development, and more than 40%
of this capacity has been announced without a target year of commissioning. Only
13% of the announced capacity has reached FID or is under construction, half of
which is in China. Today, capacity that is committed accounts for 19 GW,
compared with 6 GW at the end of 2022, and 17 GW at the time we published the
November 2023 Special Briefing.

This progress is encouraging, though there are reasons to remain cautious about
the expansion of electrolyser manufacturing capacity. Firstly, the manufacturing
output in 2023 has nearly doubled compared to 2022, but utilisation rates remain
very low. Output of manufacturing capacity in 2023 is estimated at 2.5 GW, mainly
from projects under construction in China, where the majority of the electrolyser
deployment is taking place. 8

Secondly, many of the announced factories are assembly facilities that will require
a supply of components (membranes, cathodes, anodes, bi-polar plates, power
electronics, etc.) to produce the electrolyser stacks and the final electrolyser
system. Many of these components are also used in other technologies with more
mature markets (such as power electronics for batteries). There is currently limited
visibility on the expansion plans for these components, and on whether
manufacturers will be able to serve the competing needs of both markets.
Reaching the capacity expansion envisaged by announced projects will depend
on the scale-up of manufacturing capacity of all these components in parallel, to
prevent bottlenecks occurring in certain parts of the supply chain.

8
This estimation includes manufacturing of electrolysers for the chlor-alkali industry, which has been traditionally the core
market for electrolysers, as well as electrolysers manufactured for dedicated production of hydrogen, which is now the largest
market for electrolysers.
I EA. CC BY 4.0.

PAGE | 44
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

China, with 56 GW, today accounts for one-third of the total manufacturing
capacity that could be operational by 2030. This level of deployment is significantly
above the level of deployment needed in the APS. Moreover, with only the
capacity that the country has available today (13 GW, about 60% of the global
capacity), if fully utilised, China could already meet the deployment needed to
reach its climate objectives by 2030. This could lead to a situation similar to that
of solar PV, where capacity that is surplus to the needs of the domestic market
can enable exports to other regions. In fact, almost 45% of global APS needs by
2030 could be met just by the surplus of manufacturing output from China.
However, for this to happen, Chinese manufacturers will need to modify their
current designs to comply with the standards required in other regions, and to
respond to doubts about equipment reliability that have arisen from the operational
challenges experienced with the largest project to date.

The European Union accounts for 14% of the total capacity that could become
operational by 2030. If only the committed capacity is considered, this share rises
to about 25%. If fully utilised, the committed capacity could be enough to meet the
level of deployment needed in the APS, but this would require faster
implementation of support schemes for low-emissions hydrogen production
projects, as well as policies for demand creation and the development of hydrogen
infrastructure to link producers and users. Sluggish implementation of the
announced programmes has led to delayed FIDs, which is translating into slower
growth in demand for electrolysers.

The United States accounts for about 16% of the total capacity that could be
operational by 2030, or 15% of all committed capacity. After the IRA was signed
into law in 2022, there were high expectations about the United States becoming
a particularly attractive location for capacity additions, but delays in providing the
final guidelines on provisions are resulting in subsequent delays in manufacturers
reaching FID. Consequently, less than 3 GW of manufacturing capacity (11% of
the announced manufacturing expansions) in the United States is at least at the
FID stage, which, added to the existing capacity, accounts for around half of what
would be needed in the APS in 2030.

Another 20% of the announced manufacturing capacity currently has no specified

location, and the final decision on project siting could be influenced by policies and
subsidies. In addition, a large part of the announced manufacturing capacity –
more than one-third of the 145 GW – has been announced without a specific target
year of deployment. The realisation of this capacity will depend on the demand for
electrolysers, and therefore on the deployment of announced electrolytic hydrogen
production projects. Under 4% of electrolytic hydrogen production projects around
the world have reached FID.
I EA. CC BY 4.0.

PAGE | 45
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Figure 14 Output from existing and announced electrolyser manufacturing capacity

relative to Announced Pledges Scenario deployment in 2030
60 30
GW

2030 APS domestic

deployment

Output from announced

manufacturing capacity:
40 20 Preliminary

Committed

Output from existing

manufacturing capacity:
20 10 Increased utilisation

2023 production

2022 production

China United States European Union India

IEA. CC BY 4.0.
Notes: APS = Announced Pledges Scenario. 2023 production values reflect estimates of actual utilisation rates. Increased
utilisation refers to the gap between 2023 production levels and existing capacity being utilised at 85%. A utilisation rate of
85% is used for both existing and announced manufacturing capacity in 2030. Refer to the Technical annex for more
details on the analytical boundaries and methodologies used in this analysis.
Source: IEA analysis based on announcements by manufacturers and personal communications.

Announced capacity additions for heat pumps have

slowed, but could see a quick turnaround
New announcements of manufacturing projects for heat pumps slowed in 2023
relative to 2022. Global heat pump sales declined by 3% in 2023, after two
consecutive years of double-digit growth fuelled by the energy crisis. Most major
markets showed negative trends in sales, with the exception of China, and heat
pump markets in general were hit by rising interest rates and inflation. This global
trend increased uncertainty among manufacturers, undermining potential
investment decisions in the short term in some regions.

In line with the decline in sales, manufacturing output in the European Union,
United States and Japan also fell, with an average reduction of 10% in the
utilisation rates of existing manufacturing facilities. In contrast, China’s
manufacturing capacity slightly increased to accommodate growing domestic
demand (12% increase relative to 2022), which compensated for a 20% decrease
in heat pump exports.

In the United States, heat pump sales fell by 15%, but sales of fossil fuel-based
heating systems plummeted even further, by a record 20%, indicating a slowdown
in the national heating market. However, these trends could be reversed in the
short term thanks to recent policy developments, such as the allocation of
I EA. CC BY 4.0.

PAGE | 46
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

USD 250 million under the IRA to support the expansion of domestic heat pump
manufacturing. Further, nine states (accounting for almost a quarter of residential
energy use) set a target for heat pumps to account for around two-thirds of heating
and cooling equipment sales by 2030.

In the European Union, sales fell by 5% after a decade of steady growth. The EU
market was particularly affected by a slowdown in the construction of new
buildings, which accounts for a large share of heat pump installations. In addition,
the fall in natural gas prices from their peak in 2022 has favoured the operation of
natural gas boilers, and there is continued uncertainty about policy support
schemes and regulations in some countries. The European Union remains the
only global region where manufacturers tend to announce their expansion
ambitions on a large scale, with over 30 GW of manufacturing capacity expected
to come online during this decade.

China was the only major market where sales increased, driven by demand for
air-source heat pumps for space heating, while heat pumps for domestic water
heating, a segment where China is the world leader, stagnated. In Japan, one of
the most mature markets for heat pumps, sales were down 10% due to low
consumer spending.

Sales of heat pumps in the European Union, Japan and the United States fell by
5%, 10% and 15%, respectively. In line with this decline, manufacturing output in
these regions also fell, with an average reduction of 10% in the utilisation rates of
existing manufacturing facilities. In contrast, China’s manufacturing capacity
slightly increased to accommodate growing domestic demand (13% increase
relative to 2022), which compensated for a 20% decrease in heat pump exports.

I EA. CC BY 4.0.

PAGE | 47
Advancing Clean Technology Manufacturing Part I: Chapter 2
An Energy Technology Perspectives Special Report

Figure 15 Output from existing and announced heat pump manufacturing capacity in
selected regions relative to Announced Pledges Scenario deployment in
2030
90 120
GW

2030 APS domestic

deployment

Output from announced

manufacturing capacity:
60 80 Preliminary

Committed

Output from existing

manufacturing capacity:
30 40 Increased utilisation

2023 production

2022 production

China United States European Union India

IEA. CC BY 4.0.
Notes: APS = Announced Pledges Scenario. 2023 production values reflect estimates of actual utilisation rates. A
utilisation rate of 85% is used for both existing and announced manufacturing capacity in 2030. Refer to the Technical
annex for more details on the analytical boundaries and methodologies used in this analysis.
Source: IEA analysis based on trade data from UN Comtrade and announcements from manufacturers.

Overall, announced manufacturing projects for heat pumps currently meet around
40% of deployment needs in the APS in 2030. However, manufacturing capacity
for heat pumps can typically be adjusted or expanded quickly in response to
growing demand, by either increasing the utilisation of existing lines, adding new
production lines, or building entirely new manufacturing sites. Any policies
designed to support an expansion in heat pump manufacturing should therefore
prioritise action to stimulate sustained market demand. Moreover, manufacturing
expansion plans for heat pumps are not announced as prominently as those for
other technologies, so the slowdown in capacity additions may be less significant
than it appears.

The European Union is today the only region with sufficient announced
manufacturing capacity to come anywhere close to meeting the 2030 deployment
needs of the APS, if announcements are realised in full and on time. However, the
decline in sales in 2023 has created uncertainties for manufacturing investment
decisions, as domestic manufacturing capacity by 2025 would be 50% greater
than sales in 2023 if all announced expansion plans are completed in full and on
time. Manufacturers in Europe are responding to this slowdown by downsizing
some production lines and, according to the European Heat Pump Association,
nearly 3 000 employees in Europe were affected by either job cuts or significant
shift reductions between September 2023 and February 2024.
I EA. CC BY 4.0.

PAGE | 48
Advancing Clean Technology Manufacturing Part II
An Energy Technology Perspectives Special Report

Part II. Advancing clean

technology manufacturing

Part I of this report illustrates some of the ways manufacturing contributes to

countries’ economic development, and how clean technology manufacturing – and
the surging levels of investment it is attracting – is today making an important
contribution to the global economy. Part I also shows that rapid – if uneven –
progress is being made on the deployment of manufacturing facilities. More looks
set to come if governments follow through on their climate pledges.

The emergence of clean technology manufacturing as a pillar of the new energy

economy presents clear opportunities in the form of expanding markets and
sources of employment, but also some important challenges. As with any
structural change to the economy, the growth of new sub-sectors disrupts the
status quo for incumbents, posing risks for workers and economic security. Part I
also shows that the high levels of geographic concentration in clean technology
supply chains identified in early 2023 in Energy Technology Perspectives, and in
follow-up Special Briefings in May and November, persist in virtually all
manufacturing steps analysed. This poses further risks to the security and
resilience of clean technology supply chains.

Countries across the world are already responding to these risks. When designing
their industrial strategies, governments have several options at their disposal:
boosting domestic production, forming strategic partnerships, stockpiling,
resource efficiency and input substitutions are some of the oft-cited examples.
Some of these options come with trade-offs for industrial competitiveness, while
others necessitate international collaboration. The purpose of this report is not to
prescribe a single approach, or make recommendations to a specific country, but
rather to provide a ‘toolbox’ for governments when examining some of the key
considerations of their industrial strategies.

Part II explores three categories of these considerations in turn. Chapter 3

examines some of the fundamental cost drivers for clean technology
manufacturing. Chapter 4 focuses on innovation, specifically the links between
energy and manufacturing innovation, the opportunities innovation can unlock and
the value of innovation as an exportable good. Chapter 5 explores other key areas
of government policy, identifying ‘low regret’ actions governments can take to
advance clean technology manufacturing.
I EA. CC BY 4.0.

PAGE | 49
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Chapter 3. Cost fundamentals of

clean technology manufacturing

The clean energy transition offers many opportunities for growth and employment
in new and expanding industries, including in the manufacturing of clean energy
technologies. As countries strive to meet their climate goals at the same time as
maintaining energy security and affordability through designing resilient clean
energy technology supply chains, they are also inherently competing to capture
some of this economic opportunity. Understanding the key determinants of
manufacturing costs can help to inform the development of fit-for-purpose
industrial policies to achieve these goals while maintaining a competitive edge and
creating value domestically.

Manufacturing costs for clean technologies are highly company- and facility-
specific. Individual contracting arrangements, overheads to fund R&D and
corporate expenses, financing terms, tariffs, taxes and levels of profitability across
the supply chain all have an impact on actual realised costs, as do government
subsidies and incentives. Prices for clean technologies will be influenced by all the
factors affecting costs, and more; in particular the extent to which supply and
demand are in equilibrium, which is very challenging to predict accurately.
However, an assessment of the main components of manufacturing cost – and
the principal factors contributing to differences across technologies and regions –
is an important tool for policy makers designing industrial strategies, in order to
gauge the impact of any proposed action.

Levelised cost of manufacturing

Levelised cost is a good proxy for overall manufacturing cost per unit of output
(e.g. the cost of producing 1 kW of solar PV modules), capturing both upfront and
operational costs, together with some of the main sources of regional variation in
each of its components. This section presents an overview of levelised cost
estimates for manufacturing key clean energy technologies in different regions,
based on the best data currently available. The Technical annex provides more
details on the analytical boundaries and methodologies used in the underlying
analysis.

Explicit policy incentives for manufacturing are intentionally excluded from this
analysis in order to provide policy makers with a baseline for comparison.
However, it is important to note that this does not mean the costs presented are
exclusive of all subsidies (e.g. subsidies for electricity generators), which can be
I EA. CC BY 4.0.

PAGE | 50
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

embedded – intentionally or otherwise – in all of the main components of levelised

cost explored here. Subsidies and manufacturing incentives are explored
separately in a dedicated section below.

Figure 16 Breakdown of total levelised costs of manufacturing for key clean

technologies and components in 2023

100%
Wafers

Towers
Polysilicon

Cathodes
80%
Cells

60%
Blades
Anodes
40%
Nacelles

20%

0%
Wafers

Blades

Heat pumps
Modules

Towers
Nacelles

Cathodes
Cells

Turbines

Cells

Electrolysers
Polysilicon

Anodes
Solar PV Wind Batteries
Component directly upstream Material and other component costs
Energy costs Labour costs
Fixed operational costs Capital costs

IEA. CC BY 4.0.
Notes: ‘Electrolysers’ refers to the stack of an alkaline system, and ‘Heat pumps’ refers to the final assembly step. Cost
shares presented here are calculated using energy prices, capital costs and other region-specific factors for China, and so
can differ for other countries. Values exclude any explicit policy incentives for manufacturing, transportation, profit margins,
taxes and tariffs, and therefore may not match market prices for these units. A depreciation period of 25 years, a weighted
average cost of capital (WACC) of 8%, a utilisation rate of 85% and an annual fixed operational cost set at 5% of initial
capital cost are used for all technologies and all manufacturing steps. Refer to the Technical annex for more details on the
analytical boundaries and methodologies used in this analysis.

Solar PV module manufacturing costs are estimated to be around 35-65% lower

in China than in the United States and Europe when considering region-specific
values for capital, energy and labour costs. Absolute values of cost are highly
sensitive to material and energy prices, which together account for around three-
quarters of the total levelised cost of production, with annualised capital costs and
labour costs making up less than 15% and 5% respectively. Energy costs for solar
PV manufacturing are mainly incurred at the polysilicon production step, at the
upstream end of the value chain, given the high temperatures required. In contrast,
in the downstream steps of cell and module production, material costs have a
proportionately much greater impact on manufacturing costs.
I EA. CC BY 4.0.

PAGE | 51
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Using illustrative values for the weighted average cost of capital (WACC) (8%) and
capital costs (USD 190/kW), average industrial end-user prices for electricity in
China (USD 90/MWh) and global average material prices, our bottom-up
estimates indicate a manufacturing cost of around USD 160/kW in 2023. However,
lower production costs are achievable for facilities with lower financing costs, and
for those that have access to electricity and materials at lower prices than the
national and global average values considered in our indicative figures. A recent
estimate 9 for best-in-class facilities in China puts total production costs at around
USD 125/kW in March 2024. These production cost figures are significantly lower
than global weighted average selling prices for modules during 2023 of around
USD 250/kW, which include the impact of costs and margins associated with
several intermediate transactions, and a variety of different contracting terms.

The cost of manufacturing the main components (nacelles, blades and towers) of
onshore wind turbines is estimated at around USD 385/kW in China, compared to
between USD 485/kW and USD 525/kW in Europe and the United States.
Offshore units are around 20% more costly per kW to produce on average, mostly
as a result of the more complex and material-intensive towers required. Based on
the indicative costs for onshore turbine components, material costs can account
for up to 60% and labour between 5% and 10%. At the component level, labour
costs make up a larger share of the costs of manufacturing blades than for any
other technology, whereas tower manufacturing cost is governed largely by
material costs (mostly high-strength steel). The manufacturing costs for nacelles
are largely determined by the costs for the other components (the generator and
gearbox), which are assumed in this analysis to be manufactured by external
facilities rather than in the nacelle assembly facility.

Anode and cathode active materials account for around 10% and 50% of the
manufacturing cost for battery cells, respectively, and therefore make up the
majority of overall battery manufacturing costs, which also include the electrolyte
and cell casing as additional components in this analysis. A total manufacturing
cost of over USD 100/kWh is estimated for the United States and Europe, if the
battery cell and its components are all produced locally, with costs being between
20% and 35% lower in China. These calculations assume that the prices of the
key input materials do not vary regionally, given that they are globally traded
commodities. In reality, manufacturers procure materials under different contract
terms, and integrated producers – which are more common in China – are likely
to benefit from lower prices for key inputs, which could mean the manufacturing
cost gap is even wider (see section “Synergies from supply chain integration” in
Chapter 5). Moreover, Chinese battery manufacturers currently mainly produce

9
‘Solar Supply Chain Index March 2024: Pitched Battle’ accessible via BNEF subscription. USD 125/kW is an average of
values estimated for TOPCon and PERC technologies, including overhead expenses (sales, general administration and
R&D) but excluding profit margins.
I EA. CC BY 4.0.

PAGE | 52
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

LFP cells, which are up to 30% cheaper to produce than high-nickel chemistries
that are more common in the United States and Europe.

Our bottom-up estimates suggest that the costs of manufacturing alkaline

electrolyser stacks could range between USD 45-65/kW, with the upper end of the
range corresponding to costs in the United States and Europe, and the lower end
in China. However, these illustrative figures are more representative of the levels
that could be achieved once the industry is mature, and do not account for several
factors that lead to much higher costs for manufacturers in what is currently a
nascent industry. First, global average utilisation rates today are around 10%;
much lower than the value of 85% used in this analysis to obtain comparable
figures between technologies. Just accounting for this difference in manufacturing
facility utilisation would lead to a three to fourfold increase, to a range of
USD 130-260/kW. Second, these costs exclude manufacturers’ recuperation of
R&D and other overhead costs that could equate to as much as 100% of the total
stack cost, when distributed among the small volumes of units currently produced.
Furthermore, comparisons with electrolyser system costs would not include other
components that comprise the balance of plant (e.g. rectifier, gas treating
equipment etc.).

Heat pump manufacturing is estimated to cost around USD 200-250/kW in Europe

and the United States today, which is around twice the cost estimated for China.
Manufacturing is assessed in this analysis at the final assembly step, so
components (e.g. compressors) and their materials make up the bulk of
manufacturing costs. In aggregate these inputs account for around three-quarters
of the total in the European Union and the United States, and over 80% in China.
Energy, labour and capital costs make up relatively small shares of the total
compared with other technologies such as solar PV modules. Heat pump
manufacturers that are vertically integrated tend to have a competitive advantage
over those more dependent on purchasing their components from other firms.
Similarly, those that specialise in air-to-air heat pump units can benefit from
synergies with the manufacture of air conditioners. Chinese manufacturers tend
to benefit from both of these advantages, together with lower labour and capital
costs relative to those seen in the other main manufacturing centres (the
United States, Europe and Japan).

The cost gap between the manufacturing of clean energy technologies in China
and in other countries is not set in stone. The real cost differences seen across
different regions today – which our estimates provide an indication of – are a
function of many factors, including surging energy prices in the aftermath of the
Russian Federation’s (hereafter, “Russia”) invasion of Ukraine, lingering supply
chain disruptions and currency inflation following the Covid-19 pandemic,
increased interest rates, fierce competition in a race for market share for nascent
I EA. CC BY 4.0.

PAGE | 53
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

technologies, a variety of subsidy regimes supporting manufacturing, and

uncertainty about future demand, all of which are subject to change in the future.

Upfront costs
The main upfront cost that contributes to overall production cost for clean energy
technologies is the capital expenditure on the manufacturing facility. We have
carried out a detailed analysis of data on the cost of manufacturing facilities for
this report, focusing on regions that account for the majority of global
manufacturing output today. In addition to the cost of the manufacturing facility
itself are the financing costs and, in particular, the cost of capital. While the cost
of capital is highly project-specific, there are significant variations between
regions, in particular between advanced and developing economies.

Box 6 New IEA analysis of capital costs for manufacturing facilities

Data on the cost of manufacturing facilities are scarce. When data are collected or
reported, it is quite often not clear what is included (e.g. equipment costs,
construction costs, land purchases, financing costs). Information on certain
attributes of facilities is not always available to identify comparable costs, for
example whether the facility was a greenfield (i.e. no existing facility at that site) or
brownfield investment (i.e. a sizeable expansion of an existing facility), or whether
all process steps in a supply chain are included, or just a subset.
Our analysis focuses as much as possible on the overnight facility costs, including
the core equipment and construction costs, but excluding land purchases and
financing costs (financing costs are revisited separately below). 320 facilities were
analysed for solar PV, 340 for batteries and 90 for wind, with China accounting for
the largest share of plants where both cost and capacity data were available.
Significant variation is present in these underlying data, which have been distilled
into country/region averages for the United States, Europe, China and India.
The data on capital costs presented in this section are used both in the levelised
cost and the manufacturing investment calculations. See the Technical annex for
more details on the methodologies and data sources used.

Capital costs of manufacturing facilities

Wind turbine manufacturing, including nacelle, blade and tower production
facilities, is the most capital-intensive among the five clean technologies we focus
on in this report, with costs for manufacturing facilities in the range of USD 300-
540/kW for the countries and regions we examine (average of facilities for
manufacturing onshore and offshore components). These facilities require large
buildings to house the huge components, and heavy-duty machinery for
I EA. CC BY 4.0.

PAGE | 54
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

manoeuvring them around the site. The steadily increasing size of wind turbines
in recent years has also limited the ability to standardise – an issue that impacts
not only the cost of the finished components and their installation, but also to some
extent the manufacturing facilities, as amortisation of specialised equipment is
spread over fewer units. Recent wind manufacturing facility data for Europe and
the United States are limited, due to the small numbers of manufacturing capacity
additions in recent years, hence the same average capital cost shown for both.

Figure 17 Estimated overnight unit capital costs for clean technology manufacturing
facilities in selected countries, 2023
600 150
USD/kW

USD/KWh
400 100

200 50

0 0
Solar PV Wind Electrolysers Heat pumps Batteries
United States Europe India China

IEA. CC BY 4.0.
Notes: Capital costs are shown per unit of annual rated capacity. Solar PV includes polysilicon, wafer, cell and module
production facilities; Batteries includes cell, anode and cathode production facilities; wind includes nacelle, tower and blade
facilities. Electrolysers and heat pumps include only the final assembly step. Costs refer to greenfield, non-integrated
facilities where these attributes could be isolated in the data and constitute averages across plants of different sizes today.
Data gaps filled using regional multipliers based on differentials in cost for constructing other facilities where more data are
available. No explicit policy incentives (e.g. investment tax credits) are applied in this assessment. Refer to the Technical
annex for more details on the analytical boundaries and methodologies used in this analysis.
Sources: IEA analysis based on data from Clean Investment Monitor, InfoLink, Ofweek, Black Hawk Solar, InnoEnergy,
ITDCW, IN-EN, Benchmark Mineral Intelligence, IPCEI, S&P Global Commodity Insights, GWEC and BNEF.

Facilities for solar PV manufacturing are similarly capital-intensive to wind

(USD 190-440/kW), owing to the multiple processing steps and complex nature of
the processing equipment needed, particularly for manufacturing cells and
polysilicon. Module assembly, cell and wafer production are more amenable to
economies of scale and short cycles of innovation, given the modular nature of
the technology. Much the same is true for battery cell manufacturing facilities,
which, however, cannot be compared directly in capital intensity terms given the
difference in output unit (kWh vs. kW).

For electrolysers and heat pump facilities, capital costs lie in the ranges of
USD 65-150/kW and USD 60-175/kW, respectively. These estimates are
based on far fewer data points than those for solar PV, wind and batteries, due
to the nascent (electrolysers) and slower-growing (heat pumps) nature of these
I EA. CC BY 4.0.

PAGE | 55
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

supply chains and consequently more limited data availability (see the
Technical annex for details on this analysis).

Figure 18 Estimated overnight unit capital costs for solar PV, battery and wind
technology manufacturing facilities in selected countries, 2023

Solar PV, USD/kW

Polysilicon
China
Wafers
Cells
India
Modules

Europe

United States

0 100 200 300 400 500

Batteries, USD/kWh
Anodes
China
Cathodes
Cells
India

Europe

United States

0 50 100 150
Wind onshore, USD/kW Wind offshore, USD/kW

China

India

Europe/
United States

0 150 300 450 0 200 400 600

Nacelles Blades Towers

IEA. CC BY 4.0.
Notes: Capital costs are shown per unit of annual rated capacity. Costs refer to greenfield, non-integrated facilities where
these attributes could be isolated in the source data. Investments in metallurgical grade silicon manufacturing for solar PV
are not included, nor are those associated with electrolyte, separator, or foil manufacturing for batteries. Data gaps are
filled using regional multipliers based on differentials in cost for constructing other facilities where more data are available.
No explicit policy incentives (e.g. investment tax credits) are applied in this assessment. Refer to the Technical annex for
more details on the analytical boundaries and methodologies used in this analysis.
Sources: IEA analysis based on data from Clean Investment Monitor, InfoLink, Ofweek, Black Hawk Solar, InnoEnergy,
ITDCW, IN-EN, Benchmark Mineral Intelligence, IPCEI, S&P Global Commodity Insights, GWEC and BNEF.

China is the lowest-cost region for manufacturing facility capital investment for all
technologies and for all manufacturing steps. Costs of clean technology
manufacturing facilities in the United States and Europe are between 70% and
195% more expensive per unit of output capacity. India’s capital costs are around
20-90% more than China’s for the five technologies analysed, but still significantly
I EA. CC BY 4.0.

PAGE | 56
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

lower than in the United States and Europe. These cost differentials are likely due
to differences in underlying labour, material and construction costs. China also
benefits from the experience gained in building its large stock of existing facilities,
as well as the economies of scale from larger facilities, industrial clusters covering
the full value chain, lower interest rates and a deflationary environment. A facility
that can be built more quickly at a larger scale and with less uncertainty will yield
cost reductions throughout the construction and procurement process.

Capital cost makes a modest contribution to the overall levelised cost of

manufacturing clean technologies, accounting for – once annualised – 15-25% of
the cost of producing solar PV modules, 5-10% for heat pumps, and 10-20% for
batteries, wind turbines and electrolysers. However, the variation in capital cost
between regions accounts for a significant share of regional variation in total
manufacturing cost; for example, around half of the difference in cost between
producing solar PV in China and the United States.

For battery and wind manufacturing facilities, similar differences in unit capital
costs between regions can be observed at each step of manufacturing. Cell
production is the most capital-intensive step in solar PV manufacturing for the
boundary considered, accounting for 35-45% of total capital costs of
manufacturing. The same is true for batteries, but more so, with cell production
accounting for around 80% of total facility costs for cells, anodes and cathodes.
Wind manufacturing facilities are more expensive per unit of capacity for the larger
offshore turbines, particularly for blades and towers, owing to the greater
complexity and scale of equipment needed to assemble individual components
over 100m in length.

Regional average cost figures mask plant-specific variation. For instance, all
capital costs presented here are estimates for greenfield facilities – to aid
comparability – but some significant differences in the cost of greenfield and
brownfield installations can be observed for certain components of clean
technology supply chains. Polysilicon manufacturing is a case in point, where
greenfield facilities in China – the only region where the distinction can be made
based on the data available – cost around two-thirds more per unit of output than
brownfield facilities. A recently announced JinkoSolar manufacturing facility
illustrates the lower costs that are achievable for a specific plant relative to the
national average values we have compiled. At 56 GW, it is larger than any plant
that exists today and is fully integrated (i.e. polysilicon to modules in a single
facility), factors which lead to economies of scale and lower running costs. While
estimates are not outturn costs, the facility is projected to come in at
USD 7.8 billion, or USD 140/kW for full-chain solar PV manufacturing, compared
with our national average figure of USD 185/kW for China.
I EA. CC BY 4.0.

PAGE | 57
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

The regional average figures are also static and aimed at capturing costs of the
most recently constructed facilities, thereby concealing any variation in costs over
time. Most of the data in this assessment are for facilities that have begun
construction over the past 5 years, a period in which costs and prices have
declined significantly for technologies like solar PV modules and batteries. It does
not necessarily follow that manufacturing facilities have undergone the same
capital cost declines, as more expensive facilities could lead to production cost
decreases for other cost components. However, some cost declines can be
observed directly in the plant-level data. In China, for example, capital costs for
solar PV cell and module manufacturing capacity both declined on a weighted
average per unit basis by around 35% over the period 2020-2023, whereas costs
for greenfield polysilicon and wafer production facilities appear to be broadly flat
over the same period.

Cost of capital
The cost of capital is the minimum return that a company requires to justify a
decision to invest. As such, it is also a measure of real and perceived risk: the
riskier the project, the higher the rate of return that would be required to justify
investing. Today, the cost of capital for clean energy projects is considerably
higher in developing economies than in advanced economies and in China. This
explains to a significant degree the variations in capital flows to clean energy seen
across these regions. Mobilising more capital to manufacturing projects in
developing economies will largely depend on reducing risks that push up the cost
of capital. The IEA’s report on Reducing the Cost of Capital explores ways in which
these risks can be reduced, and the Cost of Capital Observatory provides an on-
going analysis of the current cost of capital in EMDEs, together with tools and
analysis to help governments understand the main underlying risks.

The weighted average cost of capital (WACC) is a function of the cost of debt
capital, cost of equity capital, and the proportions of debt and equity used for
financing. Together with the length of the period over which an asset is
depreciated, the WACC directly influences the calculation of the annual
contribution of capital costs in the levelised costs presented above. Small
percentage point increases in WACC make a big difference in annualised capital
costs. For example, a billion-dollar investment with a WACC of 5% and a 25-year
depreciation period results in an annualised capital cost of around USD 70 million
per year. If the WACC is increased to 15%, this increases the annualised capital
cost by more than double, to around USD 155 million. This spread of values for
WACC (5-15%) is illustrative of the gap seen between advanced economies and
emerging economies today for renewables deployment projects in the power
sector.
I EA. CC BY 4.0.

PAGE | 58
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Figure 19 Cost of capital for solar PV and battery power plants in selected regions,
2022

Solar PV Battery
20%
Brazil South Africa
18%
Mexico
16%
India
14%

12%

10%

8%

6%
Indonesia
4% Advanced
2% economies

0%
Range Median

IEA. CC BY 4.0.
Notes: Solar PV corresponds to a 100 MW plant and Battery to a 40 MW plant.
Source: IEA (2023), Cost of capital observatory.

Weighted average cost of capital (WACC) can also be very sector dependent. A
manufacturing facility is likely to have a different financing structure to that of the
power plant in which its finished units may be installed. For example, the WACC of a
solar power plant in the United States can be around 5-6%, whereas a WACC of 9%
is more typical of PV module manufacturers, and latest reported company-levels of
WACC for US module manufacturers lie between 8-10%. In polysilicon production,
WACC values of 15% and above have been observed in the United States. Similar
trends are observed in China, with polysilicon producers reporting higher company-
level WACC ranges (7-17%) than PV module manufacturers (2-13%).

Today, virtually all clean technology manufacturing capacity is located in advanced

economies and China, with solar PV manufacturing installations in India and
Southeast Asia being notable exceptions. Therefore the cost of capital is likely to be
more project- and company-dependent than region-dependent, for regions currently
active in clean technology manufacturing, with variations driven by differences in debt-
to-equity ratios, costs of debt, and cost of equity across projects. Nonetheless, given
that governments may identify strategic partnerships or foreign direct investment as
features of their industrial strategies, the WACC for manufacturing facilities in
developing economies – and ways to reduce it – are important factors to consider.
I EA. CC BY 4.0.

PAGE | 59
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Operational costs
Important differentials in total manufacturing cost can arise from differences in
operational costs, i.e. those incurred after upfront costs are met and once a facility
is up and running. Three important categories of operational costs are energy,
materials and labour, which together account for 70-98% of total manufacturing
cost for the key technologies examined in this report. There are many other on-
going expenses that accrue or may be allocated during the operation of a
manufacturing facility, such as company overheads, infrastructure charges,
transport of finished goods, quality control and assurance, wastage and warranty
processing, to name but a few. Here we focus on energy, materials and labour as
three major contributors that would need to be examined as part of a wider
assessment of the viability and competitiveness of manufacturing operations at a
given site or in a given jurisdiction.

Energy costs
Natural gas and electricity are the two main energy inputs for clean technology
manufacturing today, the prices of which, including taxes and excises for industrial
users, vary significantly between countries. The prices of coal and oil products,
which are more easily traded internationally (as transport costs represent a smaller
share of the price), vary much less between regions, excluding the impact of taxes
and duties like CO2 pricing. None of the major manufacturing regions today have
an industry CO2 pricing system that covers clean technology manufacturing
directly, but several have policies that cover electricity generation, including China,
the European Union and certain states in the United States. The indirect impact of
these policies will mostly be captured in electricity prices.

Thermal energy needs are much lower for clean technology manufacturing than
for heavy industries like steel and cement, and natural gas – together with coal,
particularly in China – tends to be the energy source used today to generate the
direct process heat required. Countries with abundant domestic natural gas
reserves (e.g. the United States) tend to have much lower industry end-user prices
for natural gas, whereas regions that are dependent on imports tend to see higher
prices. Higher prices still are seen for countries that rely on liquified natural gas
imports as opposed to pipeline gas. Over the period 2013-2023, industry users in
the United States saw the lowest natural gas prices among major clean technology
manufacturing regions, with prices fluctuating within the range of
USD 15-30/MWh. Industrial gas prices in China and India tended to be two to four
times higher during this period. In the aftermath of Russia’s invasion of Ukraine,
which sparked a global energy crisis, prices in Western Europe shot up to levels
three to five times higher than the United States during the period 2022-2023.

Industrial electricity prices also vary significantly between regions, and some clean
technology manufacturing operations are electricity-intensive. The United States
I EA. CC BY 4.0.

PAGE | 60
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

has some of the lowest industrial end-user prices for electricity, and Europe some
of the highest, particularly since Russia’s invasion of Ukraine. Price differentials
between regions are similar to those for natural gas, which is intuitive, given that
natural gas is often the price-setting mode of generation in liberalised electricity
markets. In China, electricity prices are strongly regulated for many users, which –
together with the fact that a large share of electricity is generated from domestic
coal supplies – has led to falling prices for industry customers even during a time
when wholesale energy prices have been rising. Industrial end-user prices for
electricity in China reached levels similar to those of the United States in 2022-2023.

Figure 20 Industry end-user prices for natural gas and electricity in selected regions

Natural gas Electricity

90 300
USD/MWh

60 200

30 100

0 0
2010 2023 2010 2023

United States Western Europe India China Japan

IEA. CC BY 4.0.
Notes: Prices are shown in 2023 US dollars using market exchange rates on an annual average basis. End-user prices
include taxes (such as VAT), subsidies and tariffs.
Sources: IEA analysis based on the IEA Energy Prices database.

Energy can be an important component of operational costs for clean technology

manufacturing, accounting for 1-30% of the total manufacturing cost of
components excluding polysilicon and up to 75% for polysilicon, depending on the
region, and the step in the supply chain. Energy costs tend to account for a larger
share of production costs for upstream supply chain steps, like anode and
polysilicon production. For example, in solar PV manufacturing, energy costs
remain an important driver of the differences in module cost between countries,
particularly for the energy-intensive polysilicon and wafer production steps, which
consume around two- to three-times more energy per unit of production than cells
and modules. For wafers, electricity is estimated to account for around 25% of
overall production costs in China today, and for polysilicon, more than 60%. For
heat pumps and wind turbines, which are typically less energy-intensive to
manufacture, energy prices – and regional variations thereof – have less of an
impact on the total manufacturing cost in one region relative to another.
I EA. CC BY 4.0.

PAGE | 61
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Labour costs
Manufacturing wages, one of the primary determinants of labour costs, along with
labour intensity, are a significant source of regional variation in manufacturing costs.
In the United States and Western Europe, manufacturing earnings can be 6 times
higher than in China, and around 30 times higher than in Southeast Asia and the
Middle East. Important variations can also be observed within regions, with earnings
in Eastern Europe on average four times lower than those in Western Europe.
Beyond differences in economy-wide earnings, variations are also driven by
differences in the distribution of skills in the manufacturing workforce. Labour
productivity levels vary by region, driven largely by greater uptake of mechanisation
and automation in advanced economies or manufacturing centres. This in turn
reduces the magnitude of the workforce and the share of lower-skilled labour as
simpler manual tasks are automated. For example, while economy-wide earnings
are on average higher in the United States than in Germany, the opposite is true for
manufacturing earnings, as Germany has a lower share of lower-skilled (ISCO-08
Level 1) employees 10 than the United States.

Figure 21 Current average wage expenses for key clean energy technologies and
average annual manufacturing earnings and skill distribution by region
Labour cost Average annual earnings
400 80
0 1 2 3 4 5 6 7 8
USD thousand
Index (100 = China)

Economy wide: higher skill

Manufacturing: average
300 60

Economy wide: lower skill

200 40

100 20

0 0
United States

Slovakia

China

Türkiye

India

Indonesia

Egypt
Germany

Solar PV Wind Batteries

China Europe
North America Other Asia Pacific

IEA. CC BY 4.0.
Notes: Average earnings represent the gross remuneration in cash for employees, excluding employers’ contribution to
social security or pension schemes. All values are for 2022, except the average earnings in China, which is for 2021.
‘Lower skill’ refers to the ILOSTAT ISCO-08 occupations classification levels 1-2; ‘Higher skill’ referring to levels 3-4.
Information on wage per skill level is not available for China.
Sources: IEA analysis based on ILOSTAT.

10
Skill levels are defined according to ILOSTAT ISCO-08 occupations classification.
I EA. CC BY 4.0.

PAGE | 62
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Similar differences can be observed at the sectoral level, with factors including
industry maturity, degree of labour representation, local skilled labour availability,
degree of automation and mechanisation, and role of government incentives all
contributing. China, the leading clean technology manufacturing country globally,
typically has the lowest average manufacturing wages, but there are some important
differences at the level of individual technologies. For solar PV manufacturing, for
example, wages in Europe and North America can be as much as four times higher
than those in China, while the same gap decreases to a factor of two for wind
turbines and EV battery manufacturing. On a per unit output basis, wage premiums
in advanced economies or regions with more mature industries can also be offset
to some degree by greater labour productivity. This effect can be clearly observed
with heat pump manufacturing: In Japan, the average wage in this sector is about
60% greater than in China, but fewer workers are required for each unit produced,
leading to relatively comparable total wage expenses per GW.

The availability of skilled workers must also be considered alongside costs. In

theory, the levelised cost of battery production is similar in India and in China, but
the lack of qualified workers (see Chapter 5) constitutes a barrier to scale-up of
manufacturing facilities.

Materials costs
Materials make up around 25-80% of the total manufacturing cost for the key clean
technologies examined in this report. For several technologies and supply chain
steps thereof, materials are the single largest contributor to overall costs. Materials
tend to be priced in international markets, with small variations between regions
compared to energy (natural gas and electricity) prices. However, prices over time
can be highly volatile, and/or cyclical, like many commodity markets. Prices for
large-volume metals used in the manufacture of clean technologies – steel,
aluminium and copper – have shown significant variation over the last 5 years,
with prices increasing by as much as 100% for steel relative to their levels at the
start of 2018. At the start of 2024, aluminium prices were up 2% relative to their 5-
year average, whereas steel prices were up 5%.

Prices for critical minerals such as lithium, cobalt and nickel have shown
substantially more volatility than those of the larger-volume metals over the past
5 years. Prices are directly impacted by the manufacturing demand for clean
technologies, which make up 10-45% of the global total for these metals. Prices
of cobalt have fallen by around 70% since 2018, returning nearly to their 2018
level in 2022, but then falling sharply again thereafter. Lithium prices fell by 75%
between the beginning of 2018 and late 2020, before rocketing by a factor of 10
during 2021-2022. Prices then crashed during 2023, leaving the commodity priced
at around half its early 2018 level. Nickel prices showed similar – albeit more
muted – swings.
I EA. CC BY 4.0.

PAGE | 63
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Figure 22 Commodity prices for key inputs to clean technology manufacturing, 2018-
2024

Large volume metals Critical minerals Average price

Index (January 2018 = 100)

100 000

USD/t
200 300

150
200 10 000

100

100 1 000
50

0 0 100
2018 2024 2018 2024 2018-2024
Steel Aluminium Copper Lithium Cobalt Nickel

IEA. CC BY 4.0.
Notes: All figures based on contracts and price markers in China.
Sources: IEA analysis based on price information from Bloomberg Terminal.

The impacts of material and critical mineral costs are felt very differently between
clean technologies. On a price-weighted basis using the average prices for
materials over the period 2018-2024, metals produced in large volumes, like
copper, aluminium, steel and glass, together account for 70-90% of the material
costs of solar PV modules, wind turbines and heat pumps. The remainder of the
material costs are composed of silver (solar PV), nickel (wind and heat pumps)
and a variety of other materials including composites. Material costs for batteries
and electrolysers depend to a great degree on the prices of specific critical
minerals.

Taking a price-weighted average of the material requirements for the battery

chemistries currently in use today, lithium is by far the largest contributor to
material costs, followed by copper, nickel and cobalt. The remainder of battery
material costs is composed of much smaller contributions from aluminium and
graphite, the latter being the main input to anode manufacture. Electrolyser
material costs depend almost entirely on the cost of nickel (for alkaline models),
or iridium and platinum (proton exchange membrane models). Based on the
differing levels of volatility seen in recent years for large-volume metals (steel,
aluminium, copper) and critical minerals (lithium, cobalt, nickel), and the
proportions in which they are used in each technology, batteries and electrolysers
are significantly more exposed to material cost volatility than wind and solar PV
technologies.
I EA. CC BY 4.0.

PAGE | 64
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Figure 23 Share of individual material costs in total material cost for key clean
technologies

100% Other
Composite
80% Glass
Iridium
60% Platinum
Cobalt

40% Nickel
Lithium

20% Copper
Steel
Aluminium
0%
Alkaline PEM
Solar PV Wind Batteries Electrolyser Heat pumps
IEA. CC BY 4.0.
Notes: PEM = proton exchange membrane. Composite includes carbon fibre, fibreglass and other composite materials.
Solar PV covers the finished module, including the interim steps for producing cells, wafers and polysilicon. Wind covers
the nacelle, tower and blades on a deployment weighted average basis for onshore and offshore turbines. The figures for
batteries are calculated using the weighted average of battery chemistries based on their current market share in the
electric vehicle market. For Alkaline and PEM, the material costs refer to the stack and exclude balance of plant
components. For Alkaline, a nickel intensity of 800 kg/MW is assumed; for PEM, an iridium intensity of 0.4 kg/MW is
assumed. None of the materials required for installation of these technologies in power plants or vehicles are included.
Refer to the Technical annex for more details on the analytical boundaries and methodologies used in this analysis.
Sources: IEA analysis based on price information from Bloomberg Terminal.

Policy incentives for manufacturing

Financial and other incentives for manufacturing increasingly feature as
components of governments’ industrial strategies. The aim of these policies is to
reduce the cost of production – and thereby increase the attractiveness to invest
– for firms, usually by transferring aspects of cost to governments’ balance sheets.
Explicit measures include direct grants to projects or firms, government loans,
government equity purchases, loan guarantees, investment and production tax
credits, among others. These measures can be deployed on an ad-hoc basis (e.g.
one-off government support for a specific facility or project) or on a systematic
basis (e.g. codified in policy documents).

Beyond explicit measures, there are many ways embedded financial support can
influence manufacturing costs, whether intentionally or not. A vast literature has
identified and measured various subsidy mechanisms across a range of
industries, sectors and activities including studies published by the World Bank,
the International Monetary Fund, the Kiel Institute, the Centre for Prospective
Studies and International Information, Center for Strategic & International Studies
and the OECD. Such an examination is beyond the scope of this report, but it is
an important consideration for policy makers that any notion of “pure” cost for
I EA. CC BY 4.0.

PAGE | 65
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

industries like those producing clean technologies is challenging to interrogate in

practice. The measures used to deliver these embedded financial support
schemes can be similar to those comprised by explicit incentive regimes, but can
be deployed on upstream inputs like labour, capital and energy costs, and
therefore may be harder to observe and/or quantify.

As noted, the levelised cost estimates and their components examined above
should be considered as a guide to the observable costs for firms, including
embedded financial support where they exist. These figures may well be the most
appropriate tool for policy makers looking to examine the potential impact of
explicit policy support domestically, as they are a better starting point for assessing
cost gaps than theoretical estimates of unsubsidised costs. Explicit manufacturing
incentive regimes should be considered on top of these levelised cost estimates.
Some highlights of the latest developments in this area are provided below.

United States
In the United States, the Inflation Reduction Act (IRA) provides production tax
credits for the manufacturing of components across solar PV, wind and battery
supply chains through to 2032. An investment tax credit for manufacturing facilities
is also available for a wider range of technologies, including electrolysers and heat
pumps. In April 2024, USD 1.14 billion was made available for clean energy
technology manufacturing projects in a first round for this tax credit, with around
60% of the total allocated during that round, which also included funding for grid
modernisation, critical minerals production, and industrial decarbonisation. The
Defense Production Act was also extended through the IRA to provide grants and
loans to strategic domestic clean energy manufacturing projects.

Further provisions under the IRA to stimulate demand for clean energy
technologies could also indirectly incentivise domestic technology and component
manufacturing. For example, the Clean Vehicle tax credit, which provides up to
USD 7 500 for the purchase of new electric vehicles, does not apply to vehicles
containing battery components or critical minerals that come from select “Foreign
Entities of Concern” (FEOC). 11 In December 2023, the US Department of Treasury
published further guidance around FEOC-related restrictions, with an impact for
some Chinese battery and component manufacturers. The Clean Hydrogen
production tax credit is an incentive to increase demand and potentially boost
electrolyser manufacturing.

11
The United States defines a FEOC as an actor that potentially poses economic or security threats. This includes businesses
that are significantly influenced by selected governments (China, Russia, North Korea and Iran).
I EA. CC BY 4.0.

PAGE | 66
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

Table 2 Examples of the funding available for clean energy technology

manufacturing through the Inflation Reduction Act

Technology Value Description

Modules USD 0.07/W

Cells USD 0.04/W

Solar PV
Wafers USD 12/m2 • Advanced Manufacturing Production Tax
Credit (45X MPTC).
Polysilicon USD 3/kg • Available through to 2032 for clean
energy components manufactured in the
Nacelles USD 0.05/W United States.
• The value of the tax credit starts
Wind Blades USD 0.02/W decreasing from 2030 (75% of credit in
2030, 50% in 2031, 25% in 2032).
Towers USD 0.03/W • Cannot be combined with Advanced
Battery USD 10/kWh (cells) Manufacturing Investment tax credit (48C
modules USD 45/kWh (no cells) ITC).
Batteries Cells USD 35/kWh
Cathodes
10% of costs
Anodes
• Advanced Energy Project credit or
Advanced Manufacturing Investment tax
Up to 6% of investment
credit (48C ITC).
(30% if apprenticeship
• USD 10 billion available.
and wage requirements
are met) • Covers manufacturing of solar PV, wind,
battery, electrolyser and heat pump
Cross-cutting components.
• Enhanced use of the Defense Production
Act to correct domestic manufacturing
shortfalls.
• USD 500 million in grants and loans
available until September 2024.
• Covers manufacturing of solar PV,
electrolyser and heat pump components.

European Union
Policy support schemes for clean technology manufacturing have also moved
forward in the European Union. In April 2024, the members of the European
Parliament adopted the Net-Zero Industry Act (NZIA). Unlike the IRA, the NZIA
does not provide financial support to specific projects, but rather aims to boost
investment in clean energy technologies by simplifying permitting procedures,
enhancing the skills of the European workforce, and creating favourable
frameworks to boost innovation. The NZIA is part of the broader Green Deal
Industrial Plan which aims to improve regulation, access to funding, skills building,
and establish trade partnerships to boost net zero industry in the European Union.
The plan is complemented by the Critical Raw Materials Act, adopted in March
2024, which aims to support recycling and environmentally friendly supply of
critical minerals to the European Union. Other EU-level programmes are in place
to provide support to manufacturing more directly, such as the IPCEI programme
that facilitates grants to EU electrolyser and battery manufacturing.
I EA. CC BY 4.0.

PAGE | 67
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

The European Union is in the process of implementing its carbon border

adjustment mechanism (CBAM), which may indirectly support domestic
manufacturing and prevent “carbon leakage” by shielding producers of low-
emissions materials, components and technologies from competition with more
emissions-intensive goods imported from other jurisdictions. The scope of the
CBAM, which was passed in May 2023 and will enter into force in 2026, is currently
limited to a selection of goods and precursors including cement, iron and steel,
aluminium, fertilisers, electricity and hydrogen, but could be extended to all sectors
covered under the EU Emissions Trading Scheme by 2030.

Individual member states are also taking action at the national level to support
clean technology manufacturing. Country-level subsidy schemes such as direct
grants may be approved by the European Union if they are in line with the
Temporary Crisis and Transition framework. This can, for example, include aid
towards relevant equipment for the transition to a net zero industry. In the
Netherlands, the government just closed a public consultation on the new
Manufacturing Industry Investment Subsidy Climate Neutral Economy
programme, which targets the production of solar panels, batteries and
electrolysers. In Spain, public consultation opened in March 2024 on a
EUR 750 million grant scheme to support manufacturing of clean energy
technologies, including solar panels, batteries, heat pumps, wind turbines and
electrolysers. In October 2023, the European Commission approved a
EUR 100 million scheme in Italy for grants for electrolyser manufacturing.

In addition to grants and loans, subsidy schemes based on lifecycle performance

of technologies can also favour domestic manufacturing, as is the case for a
French EV purchase subsidy available to consumers, which is linked to vehicle
lifecycle analysis rather than emissions from use.

In parallel to these support schemes to encourage domestic manufacturing, the

European Union is also pursuing an investigation to assess potential unfair
subsidy practices for EV manufacturers. In October 2023, the European
Commission launched an anti-subsidy investigation to determine whether electric
vehicle manufacturers in China benefit from unfair subsidisation, and to assess
the effects for EU manufacturers. Based on the findings, the European Union will
decide whether to impose tariffs above the standard 10% EU rate for cars. A
similar investigation was launched in April 2024, targeting two companies which
have responded to a public bid for the design, construction and operation of a
solar PV park in Romania, and who might have benefited from unfair foreign
subsidies.

Other regions
In India, the Production Linked Incentive (PLI) scheme provides financial support
to reduce investment costs in new integrated PV module manufacturing plants
I EA. CC BY 4.0.

PAGE | 68
Advancing Clean Technology Manufacturing Part II: Chapter 3
An Energy Technology Perspectives Special Report

through payments linked with sales volumes achieved. Following the approval in
March 2023 of incentives for solar PV module manufacturing under Tranche II of
the PLI scheme, concerns have been raised about the likelihood of manufacturers
being able to compete across all key components quickly, and meet efficiency
targets. In addition, following the introduction of the Advanced Chemistry Cell PLI
scheme in 2021, the government launched a new consultation in July 2023 to re-
open bids for 20 GWh of unutilised battery cell manufacturing capacity. More
recently, in October 2023 the Minister of Power, New & Renewable Energy
announced that the government will launch another PLI scheme for batteries. In
June 2023 India also announced the implementation of tenders to support 15 GW
of electrolysis manufacturing capacity in the country. A first call (for 1.5 GW),
which launched in June 2023, received 21 bids, with a combined capacity twice
as high as the call itself.

In Indonesia, the first electric vehicle battery cell manufacturing facility in the
country could start operating in April 2024 with a 10 GWh capacity. The plant
received support from the Indonesian government as part of an IDR 142 trillion
(Indonesian rupiah) (USD 9.8 billion) support package for the battery industry.

In December 2023, Korea announced a KRW 38 trillion (Korean won)

(USD 29 billion) envelope in the form of tax incentives, loans, and insurance to
support Korean firms in the battery industry, including in the processing of critical
minerals. This funding also aims to support Korean firms making investments
abroad.

In Türkiye, minimum import prices on solar cells were implemented in January

2023 to protect domestic manufacturers, and since May 2023, solar PV plants
using domestic components have benefited from higher feed-in tariffs than ones
operating with imported systems.

In its 2023 budget, Canada proposed a series of investment tax credits which
could directly and indirectly support domestic clean energy technology
manufacturing. These include the Clean Technology Manufacturing investment
tax credit, which could support 30% of costs of equipment for the manufacture of
clean technologies, as well as investments in critical materials and minerals
processing, extraction and refining.

In April 2024, Australia announced plans to introduce a “Future Made in Australia

Act” to support domestic clean technology manufacturing. The legislation is
expected to be introduced in 2024.
I EA. CC BY 4.0.

PAGE | 69
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

Chapter 4. The role of innovation in

advancing clean technology
manufacturing

Technology innovation is an engine of economic growth. Investment flows to new

technologies that can outcompete incumbents or offer new value to consumers.
Throughout the history of energy systems, dramatic shifts in fuel sources and end-
uses have been triggered by the emergence of new technology ideas, often led
by governments via research funding, national laboratories or the provision of
related infrastructure. These dynamics have played a central role in the
development of the clean energy technologies featured throughout this report.
Innovation – including R&D, demonstration projects and continued optimisation –
has opened opportunities for market uptake of these products and helped by
government support, has led them to a scale of manufacturing at which heads of
state are now concerned with their contributions to national trade balances.

Government spending on energy R&D is on the increase and boosting

competitiveness is a key reason. Despite challenging economic conditions,
spending rose in all major regions between 2019 and 2023. Globally, four-fifths of
the USD 50 billion that governments spent on energy R&D in 2023 was dedicated
to clean energy topics. The impact of this funding should be to reduce fossil fuel
emissions more quickly than is possible with the current technology portfolio.

From the perspective of an individual government seeking to secure investment in

clean technology manufacturing for policy goals including job creation and supply
chain security, innovation spending can be a means of minimising the costs of the
overall policy package in the medium to long term. For example, if a country is not
the lowest-cost producer today, it may need to use subsidies to help a
manufacturer bridge the cost gap in relation to imports, or to use regulation to raise
the cost of imports. These policy approaches imply a cost for taxpayers or
consumers that is likely to persist after the manufacturing base is established.
Technology improvements to products or production processes can reduce costs
or increase the value to consumers relative to the imported good, thereby
narrowing the cost gap to be addressed by subsidy or regulation.

Another major aim of using public funds in this way is to lock in an early
comparative advantage in an emerging technology area. Innovation can be
nurtured so that recipient firms can be among the first to commercialise new
technologies or improvements to them in order to gain market share or enter new
I EA. CC BY 4.0.

PAGE | 70
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

markets. Firms that are innovating at the technological frontier increase the
likelihood of investment, employment and tax revenues in the country in which
they are headquartered.

Figure 24 Government spending on energy R&D and demonstration, 2015-2023

Billion USD (2023)

60
Rest of world

50
Japan, Korea,
Australia, New
40 Zealand

Europe
30

20 North America

10
China

2015 2016 2017 2018 2019 2020 2021 2022 2023

IEA. CC BY 4.0.
Source: IEA (2024), World Energy Investment 2024, forthcoming.

For hardware innovation in particular, the innovator’s goal is often for their product
to be manufactured and/or used in factories, for domestic markets, or for export.
Whether this is “incremental” innovation (an improvement to an existing
technology that improves performance or lowers costs) or “radical” innovation (a
new type of product that changes the fundamental nature of how a service is
delivered), it will contribute to a country’s manufacturing competitiveness if
successful. There is, of course, a risk that companies will relocate abroad after the
technology is commercialised, but experience shows that most innovative
companies stay close to where they were established and keep a share of their
R&D and manufacturing there: international moves are rare.

There are four main ways in which technology innovation can advance clean
technology manufacturing:

 By capturing a larger share of an existing market or increasing the size of the

market, typically via lowering costs. For example, by commercialising a cheaper
and more efficient heat pump that, in turn, raises demand for heat pumps.
 By raising the value of a certain subset of the overall market. A technical advance
can lead to changes in consumer preferences or regulatory requirements if it
shows the feasibility of meeting demand with lower emissions, more positive social
impact, greater durability or higher safety and security. For example, the
I EA. CC BY 4.0.

PAGE | 71
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

development of a more recyclable battery, or one that eliminates minerals with

poorly governed supply chains, could command a price premium or drive
regulatory change to increase the value of that market segment.12
 By creating intellectual property – whether formalised in patents or not – that
generates wealth from the sale of licences, levies and services. For example, a
designer of wind turbine components that licenses its design production overseas
by a third party and sells its services as an engineering contractor to assemblers
and installers of wind turbines, with the revenue returning to the headquarters of
the company, where its R&D centre is based.
 By inspiring “spillovers” of knowledge into adjacent technology areas that become
more competitive through this “free” innovation resource. For example, electrolysis
techniques or components that have value for mineral processing or CO2 capture
as well as hydrogen production.

The link between energy innovation and

manufacturing has strengthened
A large share of the technologies that can contribute to reaching net zero
emissions have unit sizes that are smaller in scale than those of past energy
systems.

Some of these technologies are small because they are electronic or digital. Some
are small because they operate best when distributed across many small
installations. Some are small because they are end-user technologies that can be
tailored to user needs and sold to millions of separate users. Some could have
cheaper unit costs if they were bigger, but their developers are adopting a modular
approach to control risks related to the budget overruns of past mega projects.
Regardless of the underlying reason, these new breeds of small energy
technologies can be mass-manufactured and traded more easily across borders.

12
In a small number of cases, such as the direct capture of CO2 from the air, entirely new markets could be created this way.
I EA. CC BY 4.0.

PAGE | 72
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

Figure 25 Clean energy technology types mapped according to their general

attributes of size and modularity versus barriers to market entry

IEA. CC BY 4.0.
Notes: SMR = small modular reactor; NH3 = ammonia.
Source: IEA (2022), How Governments Support Clean Energy Start-ups.

This evolution of the scale of energy technologies has implications for the scope of
energy innovation. First, it implies a much bigger role in the energy sector for
innovation in technologies used in manufacturing that reduce the costs or increase
reliability of factory output. Technologies such as multi-wire saws for silicon wafers
have played as much of a role in delivering cost reductions for solar PV as
innovation in the design of the solar cells and modules themselves. Second, it can
reduce the reliance of energy systems on “flows” of fuels that are consumed during
use, and increase the importance of “stocks” of materials that go into manufactured
hardware, such as batteries. New technologies that can enhance the resilience of
highly dispersed supply chains for these inputs, such as those that enable the
I EA. CC BY 4.0.

PAGE | 73
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

extraction of critical minerals from diffuse sources or avoid their use in batteries
altogether, are now an accepted part of the energy innovation landscape.

A further effect of the changing technology portfolio is the greater ease with which
smaller innovators can break into the market for energy products. If the market
demands thousands to billions of individual units each year, then there is scope
for more suppliers to enter it and to differentiate their products. Barriers to
entrepreneurs that relate to working capital, access to regulated infrastructure and
economies of scale tend to be lower for technologies with smaller unit sizes. This
phenomenon is further embedded by the deregulation of markets for electricity
and gas supplies in many countries. As a result, there is now more opportunity for
a clean energy technology spin-off from a university to successfully commercialise
a new technology and steadily grow its market share from almost anywhere in the
world.

Whereas the energy system of the 20th Century was dominated by a limited
number of major engineering firms – often state-owned – selling costly and
customised installations in the oil, gas, coal, nuclear and power grid sectors, today
there is no shortage of smaller players based in China, India, Scandinavia or
South Africa with stated ambitions to become leading exporters of EVs, batteries,
steel or energy management services.

Innovation can overcome high cost factors to

maintain manufacturing competitiveness
The purpose of R&D spending and innovation in the private sector is typically to
capture a larger share of the market or increase the size of the market. Corporate
R&D on energy technologies is estimated at around USD 130 billion per year,
indicating a continuing faith in the ability of technology improvements to keep
prices attractive while meeting customers’ needs and satisfying regulatory
requirements. These large sums spent on innovation each year, in a wide range
of countries, show that industrial competitiveness can only be understood as a
function of technology, labour costs, energy costs and other input costs. For
example, input costs and economies of scale cannot explain all of the cost decline
of solar PV. It is estimated that around 60% of solar PV cost reductions between
1980 and 2001 arose from R&D. Indeed, the role of continued R&D in reducing
costs was roughly equal to that of economies of scale even after wide
commercialisation in 2001.

Improving the technological basis of a company’s products and production

processes can keep costs competitive even in situations where the input costs are
higher than those of competitors. If successful, a manufacturer’s investments in
developing new technologies will yield benefits relating to lower energy and
materials demand via smaller products or higher production efficiency; lower
I EA. CC BY 4.0.

PAGE | 74
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

capital requirements through miniaturisation of processes; or an ability to charge

higher prices than competitors by more closely matching consumer preferences.
In this way, companies can maintain manufacturing at locations where they
already have a knowledgeable workforce, established government relations and
a proven supply chain, rather than uprooting to relocate to a lower-cost region.
Governments often have an interest in helping companies to bolster domestic
employment and revenue, and so provide R&D grants, tax incentives and
networking opportunities. A significant co-benefit of this strategy is that R&D often
generates spillovers of innovation in co-located companies that are in adjacent
parts of the supply chain or related sectors.

Innovation is therefore an important reason why countries that are known for
relatively high labour and energy costs continue to have factories that manufacture
goods in trade-exposed sectors, including automotive parts, engines, heating
equipment and robotics. Firms in these sectors that have successfully operated
manufacturing facilities in advanced economies over several decades generally
produce goods that are at the top end of the markets in which they operate. Their
pursuit of quality helps them access higher profits per unit of output. As they have
globalised, such companies may have also opened manufacturing plants in other
regions – especially where these are close to new markets – but usually keep R&D
facilities in the country of their headquarters and founding. The spending at these
R&D centres is significant, representing 2% to 10% of the firm’s revenue (Table
3). For a company such as Valeo, an automotive parts supplier, annual R&D
spending was over USD 2 billion, or 9% of revenue, in 2023.

Table 3 Eleven selected companies in trade-exposed sectors that have

maintained competitiveness through innovation
R&D Manufacturing Notable
Company R&D locations
intensity locations innovation area
Compact heat
Sweden*, Brazil, China,
Alfa Laval 2.5% Sweden*, Denmark plate for efficient
India, Italy, Poland
heat transfer
Germany*, Brazil,
Germany*, China, China, Czechia, France,
Electronic
Bosch Auto Czechia, India, Italy, Hungary, India, Italy,
8.2% stability control;
Parts Thailand, United Poland, Spain,
EV powertrains
States Thailand, Türkiye,
United States
Denmark*, Germany,
India, United Arab
Emirates, Poland, Low-pressure
Danfoss
Denmark*, China, China, Türkiye, refrigerant
Climate 4.6%
India, United States Slovenia, Mexico, compressors for
Solutions
United States, Czechia, heat pumps
Brazil, Bulgaria,
Romania, Slovakia
United States*, Advanced thin-
First Solar 4.6% United States*
Malaysia, Viet Nam film solar PV
I EA. CC BY 4.0.

PAGE | 75
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

R&D Manufacturing Notable

Company R&D locations
intensity locations innovation area
Switzerland*, Argentina,
Australia, Canada,
Switzerland*, China, Egypt, Mexico,
High-voltage
Canada, China, Colombia, India,
Hitachi direct current
2.9% Germany, Poland, Indonesia, Japan,
Energy components and
Sweden, Korea, Saudi Arabia,
systems
United States Thailand, United States,
Viet Nam, 13 European
countries
Korea*, China, India, Modular platform
Korea*, China,
Mexico, Pakistan, for mass-market
Kia Motors 2.6% Germany,
Slovakia, United States, EVs; waste heat
United States
Uzbekistan, Viet Nam recovery
Canada*, Austria,
Brazil, China,
Czechia, France, Canada*, Argentina,
Composite
Germany, India, Italy, Brazil, China, India,
materials for
Magna Japan, Korea, Korea, Mexico,
2.0% lightweight
International Morocco, Romania, Morocco, Thailand,
vehicles; EV
Spain, Sweden, United States, 17
powertrains
Thailand, United European countries
Kingdom,
United States
Lithium-ion car
Japan*, Brazil, China,
battery coatings
Panasonic Costa Rica, India,
2.6% Japan* to raise stability
Energy Indonesia, Mexico,
and energy
Thailand, United States
density
High efficiency,
United Kingdom*,
Rolls Royce flexible engines
Germany, Hungary, China, Germany, India,
Power 4.7% for marine or
India, Norway, Türkiye, United States
Systems back-up
United States
applications

France*, Brazil, China,

France*, Brazil,
Czechia, Germany, Integrated EV
China, Czechia,
Italy, India, Japan, vehicle
Egypt, Germany,
Valeo 9.2% Korea, Mexico, Poland, powertrain for
Italy, India, Japan,
Romania, Slovakia, low- and high-
Korea, Mexico,
Spain, Tunisia, Türkiye, voltage EVs
Poland, United States
United States

World’s most
efficient 4-stroke
Finland*, Italy,
Finland*, Brazil, China, diesel engine;
Netherlands, Norway,
Wärtsilä 4.3% France, India, Italy, methanol and
Singapore, United
Japan, Spain, Sweden ammonia
Kingdom
engines and
systems
* Country of headquarters (domicile).
Note: R&D intensity = research and development expenditure divided by revenue for the latest year for which data is
available for the company or its parent company.

As clean energy expands, and healthy competition between equipment suppliers

intensifies, it is reasonable to expect that clean energy innovation efforts will
increase. This comes with additional benefits as more innovative companies have
I EA. CC BY 4.0.

PAGE | 76
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

more capacity to weather macroeconomic storms, or adjust to new competitive

landscapes, than those that are reliant on resource rents or cheap factor inputs.

Policy missions for innovation to unlock new

manufacturing opportunities
Policy support for clean technology manufacturing is mounting in many countries
as governments respond to the energy crisis and seek more resilient clean energy
supply chains. Some of these policies provide a boost to the drivers of clean
energy innovation by creating demand for manufactured products, or by helping
companies to advance their plans to build factories and industrial facilities. To
enable more rapid progress towards more competitive manufactured clean energy
technologies, more direct innovation policies can be used, and specific technology
areas are worthy of consideration.

Significant policy focus has been given recently to investments that could diversify
value chains or raise the level of recycled content in manufactured energy
equipment. The US Bipartisan Infrastructure Law and IRA, which include grants
and production and investment tax credits for a range of manufactured products,
are examples of such policies. In Europe, the EU Innovation Fund backs
successful applicants for clean tech manufacturing grants, and there is scope for
the EU IPCEI mechanism to also provide this type of support, and the European
Solar PV Industry Alliance and UK Green Industry Growth Accelerator could
facilitate similar measures in future. Climate Transition Bonds could play a
comparable role in Japan. India’s PLI instrument provides payments to selected
manufacturers of solar PV modules and batteries per unit of output.

These types of policies indirectly support innovation by amplifying the incentives

for established firms and newcomers to gain a competitive edge, for example by
improving the quality of their product or reducing the price. In the case of India’s
PLI, the subsidy is tied to certain criteria for efficiency or chemical composition, an
approach that can send a stronger signal to innovators if set at an ambitious level.
Indirect inducement of innovation can be highly effective.

In cases where it is not clear that the private sector can respond to these signals
by bringing new technologies online quickly and in line with policy goals,
complementary policies are needed. Direct innovation programmes typically
include targeted government grants for underfunded but promising R&D or
demonstration projects. Most countries have budgets for targeted grants, including
China’s so-called “bounty system” that covers research challenges for electric
vehicles, energy storage and hydrogen. The European Investment Bank’s loans
to companies for electric vehicle research are an example of the use of debt as a
targeted R&D instrument, while the US loan guarantees for demonstration projects
illustrate a complementary type of finance tool. Direct support can also include in-
I EA. CC BY 4.0.

PAGE | 77
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

kind assistance, as provided by Canada’s public laboratories in its Critical Minerals

Research, Development and Demonstration Program, and via the research
networking elements of the EU IPCEIs for batteries and hydrogen.

Figure 26 Venture capital investment in energy start-ups, by technology area, for

early-stage and growth-stage deals, 2010-2023

45
Billion USD (2023)

40
35
30
25
20
15
10
5

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023

Renewables Energy efficiency Industry

Energy storage and batteries Mobility Hydrogen and fuel cells
Other power and grids Fossil fuels Other
IEA. CC BY 4.0.
Notes: Industry includes start-ups developing alternative routes to materials such as building materials, steel and
chemicals; mobility includes technologies specific to alternative powertrains, their infrastructure and vehicles, but not
generic shared mobility, logistics or autonomous vehicle technology; “Other” includes CCUS, nuclear, critical minerals and
heat generation; fossil fuels cover start-ups whose businesses aim to make fossil fuel use cheaper or otherwise more
attractive, including fossil fuel extraction and fuel economy of hydrocarbon combustion vehicles.
Source: IEA analysis based on Cleantech Group (2024).

In recognition of the potential for new, manufactured clean energy technologies to

be developed and scaled up into products by smaller companies, government
support for start-ups in this area is rising. While they represent relatively risky bets,
start-ups can be a conduit for the most disruptive ideas that can accelerate energy
transitions. Governments also appreciate their potential to seed new, large
manufacturing businesses in regions that may not currently be leaders in a given
technology area, or regions that risk losing their competitive edge without
innovation. However, establishing the conditions that could foster the next
“breakout” clean energy firm – think of Tesla, BYD, Ola Electric, Northvolt,
Enphase Energy or EVBox – requires attention to a range of policy issues. These
include availability of venture capital (VC) funds, facilities for technology testing,
the preferences and behaviours of potential customers, and barriers to market
access for new entrants. The rapid global growth in VC funding for clean energy
start-ups in the past 5 years in part reflects the improvements to start-up
ecosystems due to targeted government efforts in China, Europe, India and North
America. The dip in funding in 2023 is a clear reminder that governments must
remain alert to the impacts of macroeconomic pressures on clean energy
I EA. CC BY 4.0.

PAGE | 78
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

innovation progress, but they can be partly reassured that the 2023 drop in VC
funding was greater in sectors other than clean energy.

Several technology innovation challenges present themselves as potential means

of advancing clean energy technology manufacturing from the end of this decade.
Four are listed in Table 4 to illustrate possible target areas for direct government
innovation policy in the near term that could support longer-term goals.

They each build from existing trends that create new market opportunities or could
hinder scale-up. For example, energy efficiency efforts and the deployment of
renewable electricity and nuclear power underpin different environmental impacts
of solar PV production among countries – the emissions intensity of Chinese solar
PV has been halved through efficiency and other measures since 2011, for
example – creating opportunities for further product differentiation on this basis.
However, the reliance on silicon-based PV cells has not changed in recent years,
and there is renewed interest in alternative designs that use different raw materials
and could reach higher conversion rates. When used in tandem designs that use
less silicon, halide perovskites have already been demonstrated to reach higher
efficiencies than crystalline silicon alone. Attempts for wider commercialisation
of perovskites gain momentum by overcoming innovation challenges related
to durability and stability, which are focus areas for companies such as
Sekisui Chemical in Japan, and the US Department of Energy Solar Energy
Technology Office.

With respect to adaptation of technology to consumer needs in different locations,

innovators can be encouraged by the precedent of reduced noise and
visual impacts of wind turbines in the past two decades. However, the wind
sector has also experienced misalignment of technology and manufacturing in its
value chain: As economies of scale and land use pressures drove up the size
of the largest wind turbines by nearly 70% in around 10 years, the development
cycles for new components became shorter, increasing the risk of failures, which
is exacerbated by reliance on a small number of third-party factories for
outsourcing production of new designs. Strategies to manage this type of risk
will be required if clean technologies continue to expand rapidly.
I EA. CC BY 4.0.

PAGE | 79
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

Table 4 Technology innovation challenges to advance clean technology

manufacturing

Type of challenge Examples Possible approaches

Battery chemistries that Digital technologies (including

favour more abundant machine learning and digital twins)
minerals over cobalt or that can radically speed up the hunt
lithium. for new chemical combinations that
Electrolyser catalysts that could have desirable properties.
Design technologies that reduce the need for platinum
preserve quality while group metals (see Box 7).
avoiding inputs of scarce
minerals or those at Cut wastage in the solar PV Techniques such as 3D-printing to
greatest risk of supply supply chain to minimise reduce material inputs per unit of
chain disruption requirements for critical power generation.
minerals and energy- Optimising for durability and recycling
intensive inputs. to reduce demand for energy-intensive
polysilicon.
Non-silicon designs, introducing
perovskites.

Integrate low-emissions Digital tools that can schedule and

electricity or waste heat into modulate manufacturing process steps
manufacturing facilities. according to the availability of
renewable electricity.
Connection of multiple facilities with a
Innovate for consumers geothermal or waste heat source.
that value low emissions
intensity Reduce emissions from New fuel and propulsion technologies
shipping products, such as for ships, including large roll-on-roll-off
EVs, internationally. vessels.
Standardised designs for efficiently
shipping large offshore wind turbine
components.

Overcome obstacles to heat Scroll compressors are a first step

pump deployment towards very quiet heat pumps and air
associated with certain conditioners. Thermoacoustic devices
locations, such as the need or responsiveness to background
Tailor product
to reduce noise and/or noise may help further.
specifications to meet the
improve visual impact.
next waves of consumer
demand, which will be
Adapt the specifications, cost Innovation in open access platforms
required for technologies to
and durability of clean for designing reliable, affordable, small
reach their full potentials
energy products to the electric vehicles.
budgets of rapidly-growing
consumer segments in
EMDEs.

Innovation in component and Enhance dialogue among players in

Help align R&D
material supply must lead, the wind energy value chain to guide
programmes throughout the
not lag, turbine design. basic R&D and standardisation of key
value chain to ensure that
components, such as fixation
suppliers can provide
dimensions between wind turbine
components for next
hubs and blades, and share
generation designs
experiences.
I EA. CC BY 4.0.

PAGE | 80
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

Box 7 The value of technology diversity to electrolyser supply chains

For a century, one technology – liquid alkaline – captured the entire market for water
electrolysers, largely due to its dominance in the related and, until recently, larger
market for producing chlorine. However, the number of competing technologies has
proliferated in the past 5 years. Polymer electrolyte membrane (PEM), solid oxide
electrolyser cell (SOEC), anion exchange membrane (AEM), microbial electrolysis
cell (MEC) and decoupled water electrolysis (DWE) are among the technologies at
different stages of maturity for different applications. Depending on the mix of
electrolysers ultimately deployed to meet future demand for low-emissions
hydrogen, the impact on demand for critical mineral inputs could vary considerably.
This has implications for strategic planning of mineral supplies in the public and
private sectors, and for expectations for how prices of these manufacturing inputs
might evolve. It also points towards the importance of early consideration of
recycling infrastructure and requirements.
Unlike for some battery chemistries, electrolysers are not expected to dominate
demand for many minerals, with the possible exception of iridium. However, their
influence on supplies of nickel and platinum could be felt strongly if there is a
“winner-takes-all” outcome to the competition between technologies (see table
below). In the case of alkaline electrolysers dominating the market, demand for
nickel for electrolysers in 2030 could equal 4% of total nickel demand today, a
modest increase in the context of an overall nickel market that would grow 75% by
2030 in the NZE Scenario due to battery requirements. Nevertheless, sourcing
nickel for electrolysers affordably and securely could be more challenging if
suppliers struggle to keep ahead of battery demand. In the case of PEM dominating
the market, there might be a sizeable increase in demand for iridium and platinum,
two elements that are co-produced in a small number of locations today. However,
this would be moderated by a more balanced mix of electrolyser technologies and
further research into new PEM designs that aim to substitute iridium, for example
with ruthenium.

Possible mineral demand from electrolysers in 2030 depending on technology

choices in the Net Zero Emissions by 2050 Scenario
Mineral demand in 2030 as a share of
current demand
Electrolyser technology mix
Iridium Nickel Platinum

If all installations are alkaline - 4% -

If all installations are PEM 420% - 6%
If all installations are SOEC - - -
60% ALK, 30% PEM, 10% SOEC 125% 2% 2%
I EA. CC BY 4.0.

PAGE | 81
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

The value of technology innovation besides

lowering manufacturing costs at home
Countries that host companies and researchers who are innovating at the
technology frontier can generate multiple value streams. While the prospect of
competitive mass-market manufacturing may be a key reason to support such
R&D, not all of the value streams require the country to be a major manufacturer,
and they can lead to additional returns to innovation.

As discussed above, the most common goal for technology innovators is to supply
a large share of the market. Countries that host manufacturing facilities accrue the
advantages of employment, spending of income and tax receipts. They may also
come to host the factories of related suppliers due to so-called “network
externalities” that encompass the benefits of locally aggregated demand,
preferential access to specialised inputs and knowledge exchange. However, not
all successful innovation results in a competitive edge that can corner a market
through mass manufacturing. There are two other mechanisms that can help
create wealth for cutting-edge innovators:

 Production of high-performance products that have the highest value per unit.
 Trade in intellectual property and intangible goods.
In most markets, there are customers that are willing to pay more for a high-quality
product that has specific attributes that are unavailable in mass-market offerings.
In the area of clean energy, higher “willingness to pay” may be exhibited by:

 First-movers who can afford to buy a product that more closely matches their
customer preferences before it is affordable for other consumers. For example,
early EV adopters.
 Buyers in a country, company or sector that is required by regulation or
shareholders to pay for more expensive products with higher environmental
performance. For example, EU carmakers that must buy batteries with CO2
intensity below a specified level and a minimum level of recycled content.
 Customers that require a higher level of reliability and performance than the
market average. For example, military procurement of fuel cells or users of
electrolysers to supply processes with very low tolerance for downtime.
These cases can add up to a significant economic opportunity that is only
accessible to those operating at the technological frontier and staying there over
time. Furthermore, technology developers that seek to supply customers of this
kind often generate inventions that subsequently trickle down to the mass
market. 13 Therefore, companies producing the highest-performing products are

13
When regulation raises the “willingness to pay” in a given jurisdiction, this is part of a phenomena sometimes referred to as the
Porter Hypothesis, for which evidence for its strongest interpretation is mixed. However, the assertion that regulation or other market
incentives can create valuable market differentiation for higher-performing products and processes is not contested.
I EA. CC BY 4.0.

PAGE | 82
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

often also the companies that can profit from owning intellectual property that is
used around the world. Patents for clean energy technologies, which represent
the most formal measure of intellectual property, are growing faster in number
than patents for all inventions globally.

Figure 27 Global growth of patents in low-carbon energy technologies versus all

technologies, 2000-2021
700
Index (2000 = 100)

Clean energy
600

500
Fossil energy
400

300 All technologies

200

100

IEA. CC BY 4.0.
Notes: Shows a count of international patent families, each of which represents a unique invention and includes patent
applications targeting at least two countries
Source: IEA analysis based on data from the European Patent Office (EPO), and EPO and OECD/IEA (2021), Patents and
the Energy Transition.

While data on trade in intellectual property, including patent licences, are scarce,
there is evidence that trade in intangible capital related to manufacturing is
significant to economic prosperity. For several G7 countries, the trade balance
exceeds 1% of GDP. Across countries, intangible capital has been found to
account for around 50% more of the income in global value chains than returns to
tangible capital, such as investments in factories, and a share that is half that of
labour income. Most intangible capital value relates to R&D, including intellectual
property, computer software and databases. A smaller share derives from
organisational capital and brand value.

Globally, income from charges to overseas users of intellectual property grew 50%
in the 10 years to 2022 to a level that is equivalent to 0.5% of world GDP. Just
three countries – the United States, Germany and Japan – were responsible for
half the global total in 2022. In the United States, a country for which data is
available, half of this income relates to licences for the use of outcomes of R&D,
indicating that the trade is not dominated by copyrights, trademarks and non-
I EA. CC BY 4.0.

PAGE | 83
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

innovation intellectual property. 14 However, since 2019, international trade in

intellectual property has stagnated while revenue from total exports has grown. At
present, there are no strong reasons to believe that this phenomenon will persist
in the long term if recent inflationary and supply chain obstacles recede.

Figure 28 Income from charges to overseas customers for the use of intellectual
property, and as a share of total exports and GDP, 2010-2022
600 2.0% Other
USD billion (2023)

China
500 France
1.6%
Ireland
400 United Kingdom
1.2%
Switzerland
300 Netherlands
0.8% Japan
200 Germany
United States
0.4%
100 Share of exports
Share of GDP
0 0.0%
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022

IEA. CC BY 4.0.
Source: IEA analysis based on WTO (2024) Trade in commercial services database.

One trend that supports the increase in national income from intangible capital is
the separation of the location of design and production in industries such as
smartphones. Digital technologies have made it possible for firms to unbundle the
value in ideas from the physical manufacturing of the product in question. It has
been calculated that 35-50% of the value of a smartphone accrues to the
companies that own the design and the intellectual property, and not to the
manufacturers of the handset or the components.

The development of Extreme Ultraviolet Lithography (EUV) for semiconductor

manufacture is an example of how R&D can generate outsize value for the developers
of the intellectual property. It also echoes the critical link between innovation and the
continuation of the "learning curve” trends that show declines in prices for solar PV
and batteries over time. Learning curves are not laws of nature and cannot usually be
delivered through economies of scale alone. They will grind to a halt without R&D and
innovation in both products and processes, and sometimes require step changes in
technology approach in order to stay on track in the long term.

14
The link between the location of the R&D and the country receiving the income remains somewhat uncertain as variable
tax rates between countries and other factors can encourage firms to register intellectual property in ways that relocate
ownership to maximise financial returns.
I EA. CC BY 4.0.

PAGE | 84
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

Box 8 Innovation by customers to help semiconductor manufacturers

keep Moore’s Law on track
The long-standing trend towards miniaturisation of semiconductors and increase of
computing power since the 1960s – colloquially known as “Moore’s Law” – is widely
reported. However, in the 1990s, semiconductor production had largely relocated
to cheaper manufacturers in Korea and China Taipei who did not have the in-house
capacity for the innovation necessary to follow the miniaturisation trend. At the time,
increased computing power was enabling unprecedented changes in economic
productivity and entertainment, and it was conceivable for the buyers of silicon chips
to be content to limit risk and accept these products - the cheapest and best chips
ever available – for their future electronic devices. Had they done so, Moore’s Law
would not have been maintained. Instead, the major North American, Japanese and
European designers of semiconductors and computers worked with governments
on a high-risk approach to lithography (the process of engraving functionality in
silicon chips) that would allow it to operate at scales closer to 10 nanometres than
100 nanometres.
Following a programme of R&D investment at US national laboratories and testing
by firms such as Intel and ASML, EUV technology was commercialised by 2010 –
in time to keep the long-term trend on track, and to meet demands from the new
smartphone market. It was the product of international collaboration along the value
chain. Despite the vast majority of semiconductor chips being produced outside the
countries that funded the R&D, the key innovator countries have reaped huge
benefits: they are home to the production of the machines and components for EUV-
based manufacture. In addition, their companies continue to be the leading
designers and creators of cutting-edge products that would not be possible if such
advanced EUV-based processors were not available.

It is sometimes tempting to dismiss the opportunities for countries or companies

to invest in manufacturing R&D in sectors with mature process technology and
high input costs. In these cases, outsourcing or offshoring production may be
accepted as the only options for reducing costs. But this oversimplifies the ways
in which innovation can generate new sources of value for established players and
existing manufacturing regions. One such way is to focus on market segments
that will pay the most for quality or environmental attributes. Well-designed
industrial policies and public procurement can guide incumbent and start-up
companies towards this outcome. Another way is to focus on potential game-
changing process technologies that can raise the quality of manufactured
components worldwide, with the aim of integrating them into higher value clean
energy equipment for end consumers. Hi-tech process technologies are
particularly well-suited to these types of strategies because proprietary
manufacturing processes are typically harder to reverse-engineer than to work out
I EA. CC BY 4.0.

PAGE | 85
Advancing Clean Technology Manufacturing Part II: Chapter 4
An Energy Technology Perspectives Special Report

the details of a product design. A third way is to maximise the comparative

advantage of co-location for producing intangible capital in networks of interacting
experts and for raising efficiency and reliability through physical integration of
processes and suppliers.

The size of the future market for clean technologies is expected to be large enough
to accommodate examples of advanced economies stepping back down value
chains and also EMDEs moving progressively up them. In advanced economies,
where the energy R&D focus has, until recently, been on product and system
design, especially to make final goods that add value to intermediate commodities,
innovation can help industrial processes and mass manufacturing stay
competitive. In particular, investments in innovation today can reduce the costs of
meeting employment and other social policy goals related to clean technology
manufacturing during energy transitions. In EMDEs, stronger innovation policies
can create new sources of value in much the same way, including by enhancing
the local capacity to absorb and adapt technologies originating abroad. In addition,
EMDEs have the potential to seed entirely new manufacturing bases based on
geographic or cost-based comparative advantages that are specific to clean
technologies – such as renewable resources, mineral resources, proximity to
export markets or skilled labour.

I EA. CC BY 4.0.

PAGE | 86
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

Chapter 5. Policy priorities for

advancing clean technology
manufacturing

Governments have a variety of tools at their disposal to advance clean technology

manufacturing in their jurisdictions; cost competitiveness of domestic output is an
important driver for companies to invest, but by no means the only one. As
described in the previous chapter, supporting innovation is one possible means of
advancing clean technology manufacturing by promoting cutting-edge research to
reduce costs and raise product quality. However, as innovation can take time to
be translated into project investment, a range of other complementary policies can
foster investment in the near term. They include, but are not limited to, overarching
climate policies, clear environmental and social standards, accelerated permitting,
workforce training and international co-operation.

How attractive a given location is to potential investors in new manufacturing

capacity is determined by many more factors than cost alone, such as robust local
demand for the technologies being produced. The same is true for the
competitiveness of continued output from an existing facility in that location. In
both cases, million-dollar decisions are routinely taken by investors or owners in
ways that cannot be explained solely by cost, even if cost is always a major factor.

While many of the policies that are currently in place to advance clean technology
manufacturing in the near term are cost-based (see Chapter 3), in this chapter we
consider policy interventions that can improve the attractiveness of investment or
production in a region without subsidising the costs of manufacturing. Cost
disadvantages can also be offset in a range of ways that are not covered in this
chapter, including the stability of the political and economic outlook, the fiscal
regime, the availability of dependable infrastructure, the ease of exchanging
knowledge and skills and the absence of corruption, or reputational factors,
including a track record for quality output.

Enlarging domestic markets with climate

policy
Strong and stable energy and climate policies can help catalyse markets for clean
technologies. Proximity to sizeable domestic demand can help manufacturers
achieve economies of scale and partially offset cost differences with producers in
I EA. CC BY 4.0.

PAGE | 87
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

other regions. In addition, local demand makes domestic manufacturing less

dependent on uncertainties in export markets, thereby reducing project risks.

The prominence of China, the United States and the European Union in global
investment in clean technology manufacturing has been partly driven by strong
domestic demand for technologies, and the sheer size of these markets. China
alone currently accounts for 30-75% of global demand for the technologies we
focus on in this report.

Looking forward, demand for clean energy technologies will expand rapidly if all
countries follow through on their climate pledges. In the APS in 2030, solar PV
capacity additions grow from 420 GW today to 640 GW, and wind capacity more
than doubles, up from 115 GW to 240 GW. Battery capacity increases more than
fivefold, from 865 GWh to 4380 GWh, and electrolyser capacity from 1 GW to
60 GW. Heat pump capacity more than triples, from 110 GW to 355 GW. However,
demand remains concentrated in China, the United States and the
European Union, reflecting rapid growth in the sizes of their domestic markets for
clean technologies. These regions account for 45-75% of global demand for clean
energy technologies in 2030 in the APS, with China alone making up 10-45%.

However, a lack of near-term demand in emerging economies could further

increase the investment gap between advanced and emerging economies.
Emerging economies are projected to account for 0-10% of announced
manufacturing capacity for 2030 on the basis of announced projects, and only
5-25% of deployment in 2030 in the APS. Given the low levels of demand in these
countries today, it will be challenging – though by no means impossible – for many
EMDEs to create sufficient domestic demand on their own in the near term to
attract investment in clean energy technology manufacturing.

Two key pillars can contribute to success towards this aim: first, policy support is
needed to build up demand in order to spark meaningful investment in
manufacturing capacity. Second, co-ordination between countries can play a
crucial role in expanding the size of the market for clean energy technologies. In
Association of Southeast Asian Nations (ASEAN) countries, for example, in order
to achieve each country’s own climate goals as reflected in the APS, the size of
the electric light-duty vehicle market across the ASEAN region needs to grow by
a factor of eight to 2030, to around 1.1 million vehicles being sold. This is
equivalent to the size of the US EV market in 2023. On its own, Indonesia, as one
of the largest markets in the region, reaches around half of the size of this market
in 2030. In May 2023, ASEAN leaders signed a declaration to close this gap by
working together to develop a regional EV ecosystem, from building regional EV
production to improving charging infrastructure.
I EA. CC BY 4.0.

PAGE | 88
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

Figure 29 Share of global deployment for selected clean energy technologies by

region in 2023 and in the Announced Pledges Scenario in 2030

100%

80%

60%

40%

20%

0%
2023 2030 2023 2030 2023 2030 2023 2030 2023 2030
Solar PV Wind Batteries Electrolysers Heat pumps

China United States European Union India Rest of the world

IEA. CC BY 4.0.

Compressing lead times

A regulatory environment that helps to accelerate lead times can be a source of
competitive advantage. Commissioning new manufacturing plants is not typically
considered a bottleneck in clean energy technology deployment, compared to the
time it takes to commission new mining projects, power plants or transmission and
distribution infrastructure. Commissioning manufacturing facilities can take
anything in the range of 3-4 months (e.g. for solar PV module and cell facilities in
China) to more than 3 years (e.g. for polysilicon plants in Europe and the
United States) from FID to operation. It is also typical for manufacturing plants to
operate at a level of output significantly lower than their maximum rated capacity
for the first 1-2 years, while operations are honed. Lead times have a significant
indirect impact on the cost of manufacturing, as capital costs still need to be met
during the period in which facilities are not generating any revenue (see Chapter
3). Regions with shorter lead times to ramp up manufacturing also have the
potential to capture a larger share of the global market in the near term.

Policies to shorten lead times, such as streamlined permitting and clear regulatory
frameworks, when combined with adequate resourcing for regulatory agencies,
can help provide certainty for contractors, suppliers and investors. This should be
balanced against the need to ensure that environmental and social safeguards are
part of the process. Lead times for the downstream installations like power plants
and storage facilities are also relevant, as any project delays can give rise to
I EA. CC BY 4.0.

PAGE | 89
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

uncertainty in manufacturers’ order books. Wind power projects are rightly an area
of focus for policy makers in this regard. In the European Union, the adoption of
the revised Renewable Energy Directive aims to shorten permitting times for
certain wind energy installations (to 1 year for onshore projects and 2 years for
offshore, with an extension of up to 6 months) and limit the grounds of legal
objections to new installations. Moreover, the European Wind Power Action Plan
was proposed in October 2023 to support European competitiveness in the wind
industry. A key pillar of the plan is improved auction design, as well as improved
access to finance, monitoring of unfair trade practices, and skills development.

Grid expansion and modernisation projects can also create uncertainty for clean
technology manufacturers. In the United States, insufficient grid capacity to
integrate new renewable electricity projects is stifling investment. Average queue
lead times there rose from 3 years in 2015 to 5 years in 2022. In the
United Kingdom, 120 GW of projects awaiting connection have been offered it
only in 2030 at the earliest. Meanwhile, France’s backlog of projects has led to
connection delays of 22 months. In Brazil, increased development of solar PV and
onshore wind has increased grid connection queues and project lead times.
Where permitting and connection delays create a lack of visibility on future
demand, the resulting delays in investment in domestic manufacturing facilities
can, in turn, create insecurity in component supply, further delaying installation
projects. In India, higher turbine prices due to supply chain challenges have
reduced the bankability of projects that had already concluded their auctions,
resulting in delays.

Boosting the availability of skilled workers

The clean energy transition requires a rapid expansion of the energy workforce,
with manufacturing clean energy technologies presenting some of the greatest
demand for new workers. In the NZE Scenario, manufacturing jobs in electric
vehicles, solar PV, heat pumps and wind increase 220% between 2022 and 2030.
This is driven primarily by a boom in the workforce for manufacturing electric
vehicles and their batteries, with assembly jobs increasing by more than 400% as
electric cars come to represent 65% of new car sales by 2030. Wind
manufacturing jobs, which accounted for nearly one-third of all jobs in the wind
sector in 2022, more than double by the end of the decade. Expansion is slightly
more limited for solar PV, where installation represents a greater share of
workforce growth, with manufacturing employment increasing by 40% over the
same period. Employment in residential heat pump manufacturing grows by
approximately 180% to 2030 in the NZE Scenario as deployment accelerates.

Manufacturing facilities in some regions already struggle to attract and maintain

adequate staff, and regions with large existing manufacturing workforces have
greater potential to retrain workers from other manufacturing sectors in order to
I EA. CC BY 4.0.

PAGE | 90
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

more rapidly ramp up clean energy manufacturing to meet future demand. Today,
China is home to the majority of manufacturing jobs for all technologies considered
in this report, accounting for around 80% of the workforce in solar PV
manufacturing, about 60% for wind, electric vehicle and battery manufacturing,
and around half for heat pump manufacturing. While Europe boasts a relatively
sizeable manufacturing workforce in electric vehicle assembly, and to a lesser
extent in wind, workforce numbers in solar PV and battery manufacturing are well
behind those of China. Many battery manufacturing companies in Europe already
struggle to hire qualified employees locally, often recruiting personnel from Asia
to build out their workforces. However, well-established industry players in Asia
are also facing difficulties staffing new facilities and hiring sufficient high-skilled
specialists such as engineers. To remedy this problem, major firms often rely on
transferring existing workers to recently built plants to help with training and
upskilling new recruits, resulting in up to 30% of staff in a new plant coming from
existing manufacturing facilities.

Availability of skilled workers can constitute a significant bottleneck. In India, only

around 3% of the population has undergone formal vocational training (though this
figure rises to 17% if informal training is included). This is far from the OECD
average of 44% of upper secondary students being in vocational training. As a
result, even if the levelised cost of battery production would in theory be lower in
India than in China (see Chapter 3), the lack of qualified workers could be a barrier
to expansion of manufacturing.

Figure 30 Number of workers in clean energy manufacturing by technology, 2022, and

global employment needs in manufacturing by region and technology in the
Net Zero Emissions by 2050 Scenario, 2030
Workers (100,000)

% manufacturing workforce

100 100%

80 80%

60 60%

40 40%

20 20%

0 0%
2022 NZE 2030 EV Solar PV Wind Heat pumps

EV Solar PV China Other Asia Pacific

Europe North America
Heat pumps Wind Rest of the world

IEA. CC BY 4.0.
Notes: EV includes both vehicle assembly and battery manufacturing.
I EA. CC BY 4.0.

PAGE | 91
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

Building up a trained workforce to manufacture a new technology takes time, so

proactive and strategic labour planning is needed to prevent shortages. Training
programmes that target clean energy technologies, such as those proposed as
part of the European Commission NZIA, could help build up a skilled workforce.
Stakeholders in industry, education and labour relations should collaborate to
ensure these training programmes are designed with a view to priority skills and
requirements at each supply chain stage, as well as potential synergies across
technologies. For example, electrode and polysilicon manufacturing might require
more chemical and mechanical engineering skills, while automation and
electrochemical engineering skills might be more relevant for battery and PV cell
manufacturing. Taking advantage of existing training structures, for example
through partnerships between industry and educational institutions, can decrease
uncertainties and lead times when building a pipeline of skilled clean energy
workers.

Experience with similar technologies already in use can also be leveraged to

reduce the time and resources needed to train clean energy manufacturers. For
example, a significant portion of the workforce currently manufacturing internal
combustion engine (ICE) vehicles may switch to assembling EVs, which are often
produced by the same companies. However, these “transfers” are not always
straightforward. In EV manufacturing, growth in battery manufacturing jobs has
offset declines in other parts of the automotive manufacturing chain. But firms that
manufacture components for ICE vehicles are not the same as those that make
EV batteries, nor are their production facilities necessarily located in the same
regions. For that reason, some regions with large workforces focused on ICE
upstream manufacturing are set to lose jobs on a net basis without large-scale
investments in EV supply chains.

Even when there is geographic overlap between fossil fuel and clean energy
sectors, wage differentials can present another barrier to skills transfer. While
wages in the energy sector are generally higher than for comparable occupations
in the broader economy, wages in clean energy sectors are often lower than those
in fossil fuel sectors. In the United States, for example, ICE powertrain plants may
pay as much as USD 10-15 more per hour than vehicle battery plants, despite
high labour demand from the latter. To address this obstacle, government funding
or incentives for clean energy manufacturing can include contingencies. The
Department of Energy’s USD 2 billion Domestic Manufacturing Conversion Grants
programme is therefore preferencing applications that transfer workers from ICE
to EV manufacturing at comparable wages, or that maintain collective bargaining
agreements.
I EA. CC BY 4.0.

PAGE | 92
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

Synergies from supply chain integration

Regions that rely on imported components as inputs in their manufacturing processes
are more susceptible to shocks and disruptions in component markets and supply
than regions with an integrated supply chain. Recent surges in international shipping
costs have also led to significant inflation in imported components.

More than in any other region, companies in China have increasingly consolidated
manufacturing in each segment of the supply chain over the last decade,
particularly in solar PV. China now produces the vast majority of the world’s solar
PV modules, with large and medium-sized integrated solar PV manufacturers
producing three out of every four supply chain products on the market. These firms
account for over 90% of global polysilicon, wafer and cell manufacturing capacity,
and around 80% of module manufacturing capacity. The cost efficiencies resulting
from integration, and the consequent ability to absorb price shocks, allow these
firms to produce the lowest-cost solar PV equipment while also introducing labour
and manufacturing efficiencies to reduce variable costs. In contrast, PV module
manufacturing in the United States, Europe and India depends on imported cells,
the cost of which can account for 60-70% of overall module costs.

Similarly, electric vehicle and battery supply chains are highly integrated in China,
more so than in the rest of the world, which leads to cost advantages. Chinese
manufacturers boast a surplus of electrolyte, anode and cathode manufacturing
capacity relative to their domestic production of batteries, opening up potential for
exports, while manufacturers in Europe and the United States largely rely on
imported components throughout the supply chain (see Chapter 2). These
integrated supply chains are part of what enabled Chinese automakers to produce
more than half of the electric cars sold worldwide in 2023.

Owing to the significant cost of transporting large components for wind, particularly
towers and blades, manufacturing tends to be located closer to demand across the
value chain, and supply chains tend to be more integrated at the domestic level.

Reducing supply chain uncertainty with trade

agreements
Trade of clean energy technologies is an opportunity to help strengthen
emerging markets and create new ones, and to facilitate greater co-operation
throughout supply chains. In some cases, relationships that cover the trade of
clean technologies may continue under the umbrella of larger, existing
agreements. In others, the trade of clean technologies may open the door to new
trade partners and create opportunities for a wider group of countries. These
existing and new trade relationships must be carefully evaluated in the broader
context of a country’s industrial strategy.
I EA. CC BY 4.0.

PAGE | 93
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

Recent trade agreements relating to clean technologies have tended to be

bilateral, have relied on existing relationships, and have largely focused on critical
minerals supply and processing. There has been a concerted effort in different
countries worldwide to establish agreements with mineral-rich countries to secure
upstream supply chains.

Table 5 Recent agreements on clean energy technology supply chains

Agreement Description Type Signatories Coverage

Zambia and the Democratic Republic of

the Congo signed a co-operation
agreement in April 2022 to facilitate
development of the battery supply
Zambia- chain for EVs. The two countries, both Zambia, EV battery
Democratic major producers of key critical minerals Democratic supply chain
Bilateral
Republic of the for EV batteries (cobalt and copper), Republic of the (critical
Congo established a Battery Council to oversee Congo minerals)
the new agreement. The United States
and European Union have also both
signed MoUs with each of the two
countries.

Australia,
Pillar III of the IPEF covers a range of Brunei Daruss
issues critical to transitions to clean alam, Fiji,
economies, including an agreement to India,
Indo-Pacific strengthen clean energy supply Indonesia,
Clean
Economic chains across markets by building a Japan, Korea,
energy
Framework for better understanding of the challenges Multilateral Malaysia, New
supply
Prosperity and vulnerabilities of the region’s supply Zealand,
chains
(IPEF) chains and securing more diversified and Philippines,
sustainable sources of critical inputs, Singapore,
including critical minerals or materials for Thailand,
clean energy technologies. United States,
Viet Nam

Australia and the Netherlands signed an

MoU in January 2023 to support the
Australia- Australia, the Hydrogen
development of a renewable hydrogen Bilateral
Netherlands Netherlands supply chain
supply chain from Australia to Europe.

Brunei
Darussalam,
Myanmar,
Cambodia,
In 2023, ASEAN member states and
ASEAN- Canada, Clean
Canada agreed to work together to build
Canada Indonesia, energy
new clean energy supply chains under Multilateral
Strategic Laos, supply
their existing strategic partnership
Partnership Malaysia, chains
framework established in 2022.
Philippines,
Singapore,
Thailand,
Viet Nam
I EA. CC BY 4.0.

PAGE | 94
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

Agreement Description Type Signatories Coverage

In 2023, the United States and Australia

signed a compact to accelerate the
expansion and diversification of end-to-
Climate, end clean energy supply chains; promote Clean
Critical responsible, sustainable, and stable energy
Minerals and supply of critical minerals; drive the United States, supply
Bilateral
Clean Energy development of emerging battery Australia chains;
Transformation technologies; and support the critical
Compact development of emerging markets for minerals
clean hydrogen and its derivatives in the
respective countries and across the Indo-
Pacific.

Notes: EV = Electric Vehicle; MoU = Memorandum of Understanding.

Reducing environmental impacts and

addressing social considerations
The environmental impacts of clean technology manufacturing and its supply
chains could influence the partners that countries and companies choose to work
with. Environmental regulation relating to the lifecycle emissions of clean energy
technologies, including from mining, material production and transport, can
encourage investment in new locations with access to low-cost, low-emissions
electricity, or incentivise investment that serves to reduce the emissions intensity
of existing operations. Regulation may include caps on lifecycle emissions,
emissions intensity-based technical border adjustments, and incentives indexed
to lifecycle emissions performance.

Upstream steps of clean technology supply chains today are generally more
emissions-intensive than those downstream. The production of materials (e.g.
steel, aluminium, copper, nickel) typically generate the largest share of CO2
emissions across the supply chain for key clean technologies – typically upwards
of 60% when including indirect emissions from electricity generation. Technology
manufacturing tends not to be as energy- or emissions-intensive as material
production, with electricity being the main energy input. For example, the primary
factor influencing the carbon intensity of solar PV manufacturing is the share of
fossil fuels in a country’s electricity generation mix. In addition, water needs for
mining and processing the ores and minerals required for clean technologies are
often much higher than those required for technology manufacturing.
I EA. CC BY 4.0.

PAGE | 95
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

Figure 31 Share of CO2 emissions by supply chain segment for key clean technologies
100%

Technology
manufacturing
75%

Material production

50%

Mining

25%

Material production
share
0%
Solar PV Wind Batteries Heat pumps
IEA. CC BY 4.0.
Notes: Includes indirect emissions from electricity generation and the production of chemicals used for mining and material
production. Shares derived based on estimates of global average emission and material intensities. An emission factor of 460 g
CO2/kWh (approximately the global average in 2022) is used to calculate emissions from electricity generation. Shares for
batteries based on NMC 811 battery chemistry. For batteries, material production refers to material refining, while technology
manufacturing refers to cell and pack production, and active material synthesis. Shares for wind based on onshore wind turbine
components. Refer to the Technical annex for more details on the analytical boundaries and methodologies used in this analysis.

Regulations that directly target materials production may play a role, as countries
aim to source near-zero emissions materials for their clean technology
manufacturing base. For instance, regulations that consider emissions intensity
improvement can help incentivise material savings all along supply chains. An
example is the European Union’s Energy Performance of Buildings Directive,
which will cap buildings’ embodied carbon emissions.

Some countries, including France and Korea, have begun to include the embodied
carbon footprint of solar PV panels as a criterion in their competitive tender
evaluations for new power plants. Countries with ambitious climate targets are
also considering policies for imported renewable energy goods, including solar PV
(for example in the European Union and the United Kingdom). In France, since
October 2023, EV purchase subsidies available to consumers have been linked
to vehicle lifecycle analysis, rather than solely considering the emissions
generated during the use-phase of the vehicle. Other countries around the world,
including Canada, are considering similar approaches. Such policies favour
vehicles manufactured in jurisdictions with access to clean energy – particularly
low-emissions electricity – to power their facilities.

As countries look to increase transparency along supply chains, social

considerations are also increasingly factored into decision-making on
procurement. Public and private sector investors are increasingly demanding
I EA. CC BY 4.0.

PAGE | 96
Advancing Clean Technology Manufacturing Part II: Chapter 5
An Energy Technology Perspectives Special Report

greater transparency on critical mineral supply chains, given that mines,

processing facilities and refineries carry risks of harm to the environment, workers,
communities and societies.

Social and human rights considerations, while generally most widely discussed
with reference to critical mineral production, could also play a role in clean
technology manufacturing. In March 2024, the European Union reached an
agreement on a regulation to ban products from the EU market that are made with
forced labour. In June 2021, the US Customs and Border Protection agency
issued a Withhold Release Order on shipments containing polysilicon from several
producers in Xinjiang, China, due to concerns over human rights and international
labour standards.

The use of trade policy to serve non-trade policy objectives, such as with regards
to addressing corruption, can also help create a more stable environment for
businesses and investors over the longer term.

I EA. CC BY 4.0.

PAGE | 97
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

Part III. Key principles for decision

makers

Part I of this report (Chapters 1 and 2) examines the current state of play for clean
technology manufacturing. On the one hand, the analysis reveals that
encouraging progress is being made, both in monetary terms (with around
USD 200 billion invested globally in 2023) and in relation to climate goals, with the
global project pipeline of solar PV, battery and electrolyser manufacturing facilities
on track to serve NZE Scenario deployment levels in 2030 if all announced
projects materialise. On the other hand, the analysis highlights important gaps in
the manufacturing project pipeline for wind and heat pump technologies, and
persistent levels of geographic concentration that pose risks to security and
resilience in clean technology supply chains.

Whichever combination of measures is favoured in countries’ industrial strategies,

maintaining competitiveness and improving resilience will be central themes. Part II
of this report (Chapters 3 to 5) provides an analytical toolkit for examining trade-offs
and complementarities for resilience and competitiveness in three key areas – cost
fundamentals, innovation and other non-cost measures. Part III builds on the
analysis in each of these areas, providing principles for designing key aspects of
industrial strategies. The analysis deliberately stops short of recommending the
specific measures these strategies should comprise, as they are highly context and
country specific. The principles cover two main areas: The first concerns measures
that can be taken domestically, whereas the second pertains to measures that
inherently involve collaborating with international partners.

Domestic actions to advance clean

technology manufacturing
Prioritise and play to strengths
It is self-evident that governments cannot prioritise everything at once. For most
countries, it is simply not realistic to effectively compete in all supply chain steps,
or even in parts of all clean technology supply chains. Not only would the costs of
the various support measures be prohibitive for countries with higher-cost
production or no legacy of leadership in these sectors, but the sizes of many
individual economies would be unlikely to accommodate all the requisite
investment (see Chapter 1). Understanding relative strengths and weaknesses,
and where it might be better to build complementary strategic partnerships with
I EA. CC BY 4.0.

PAGE | 98
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

other countries (see below), should be key considerations of industrial strategies

for clean technology manufacturing.

Beyond deciding where to focus, it is also important for governments to define

what they wish to achieve. Clearly articulating thresholds for success before
making any financial commitments gives a government room to cease or redirect
support when things do not turn out as hoped. In this domain, more so than in
many other energy policy areas, precise outcomes are often highly uncertain, and
unintended consequences may arise. Therefore, the ability to experiment and
course-correct should be built in from the start. As industrial policy inevitably
involves picking winners at a sectoral level, and often at a company level too,
governments need to create the political and economic headroom to identify,
monitor and manage any “losers” resulting from the policy measures. Many of the
oft-cited examples of industrial policy backfire stem from a commitment to a
specific measure, firm or project without a clear means of determining or
monitoring success. In such cases, precise objectives (e.g. “limit single-supplier
dependencies for a given product to less than 50% of annual demand”) are
preferable to broadly defined ones (e.g. “improve national security”).

Attract and support innovators

Hosting researchers and other innovators can have multiple benefits for a country.
Chief among these is the strong link between innovation and domestic
manufacturing that is at the cutting edge of technical advances, able to withstand
competition from regions with low-cost inputs and able to command a premium
price. Other benefits include the ability to attract highly educated workers and the
spillovers that accrue from co-location of innovative firms, raising the combined
and individual productivity of clustered firms more quickly. Governments can use
a combination of direct and indirect measures to support clean technology
manufacturing innovation. Indirect measures, which can be very effective, involve
making the market for successful clean technologies larger, more differentiated or
more dependable.

Direct support for innovation – including R&D grants, tax incentives, start-up
incubation, knowledge sharing and demonstration project finance – enables
support to be allocated to important challenges or high-potential domestic
capabilities. In Chapter 4 we identify several potential “missions” for direct
innovation policy to advance clean technology manufacturing: high-quality designs
that avoid scarce or unreliable mineral supplies; products with low emissions
intensity; products with attributes that target the next waves of consumer demand;
and alignment of R&D advances across value chains.
I EA. CC BY 4.0.

PAGE | 99
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

Plug cost gaps strategically and for the long term

Governments may deem it appropriate to subsidise or otherwise provide direct
financial support to manufacturing operations or investments. The intention is
typically to lower the costs faced by manufacturers and thereby raise their
competitiveness, usually by redistributing costs from private to public balance
sheets. This may well be warranted in certain circumstances, but governments
have limited balance sheets just as private sector actors do, and every subsidy
comes with opportunity costs.

Aside from innovation, which is not usually a short-term option, there are some
measures that can help fundamentally reduce the total cost for all stakeholders.
Chapter 5 explores several such “low regret” options that come without significant
production or investment subsidies. Reducing lead times through enhanced
permitting procedures reduces transaction costs, project risks and consequently
the interest paid on monies committed during the early stages of a project. While
some aspects of upskilling or reskilling workforces are costly outlays, targeted
training programmes and certification schemes can increase productivity and
alleviate costly skilled labour shortages.

International co-operation to support

domestic investment and global progress
Collect data and track progress
It is difficult to manage what is not measured, and the current state of data
availability on clean technology manufacturing makes accurate measurement
challenging. Much of the data used in Chapters 1-3 of this report are from
proprietary sources, which can have gaps in their coverage and require significant
post-processing and analysis. Policy makers may often be left with two or more
conflicting data points or trends. Efforts to improve data collection can be
advanced to a certain extent by national statistical offices and agencies. But to
obtain effective, granular comparisons between technologies and along supply
chains, governments could benefit from co-operating on data collection efforts
internationally.

One specific area that deserves prompt attention are the international systems for
collecting production and trade data for clean technologies and their components.
As described in Chapter 1, internationally adopted frameworks for collecting
statistics on industrial activity currently lack the detail to be able to isolate
individual clean technologies and their components. Individual countries’ customs
authorities and other national bodies already collect data at higher levels of
granularity, but often not in a harmonised manner. Tried and tested frameworks
like the ISIC and HS already exist and should be adapted to incorporate sufficient
I EA. CC BY 4.0.

PAGE | 100
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

detail on production activity and products, but further data on clean technology
manufacturing (e.g. energy use, physical production quantities, emissions foot-
printing, investments, costs, employment) should also be sought and harmonised
by governments internationally.

Co-ordinate efforts across supply chains

Much attention is now paid – quite rightly – to the security of supply of critical
minerals, but any supply chain is only as strong as its weakest link. Governments
should co-ordinate the work they are doing at each stage in the supply chain to
increase overall resilience and avoid unwanted duplication, examining remaining
gaps that may lead to bottlenecks. Wherever possible, governments should co-
ordinate efforts to enhance the resilience of supply chains.

This collaboration can cover many specific areas and take many forms. Sharing
best practice in the appropriate fora and at an appropriate level of detail is an
important vehicle for collaboration. This could include domestic experience with
creating favourable investment conditions at home or abroad, accelerating
permitting, designing effective and efficient environmental regulation and
stockpiling of input materials and components. “How-to-guides” for developing
industrial strategies could be a method of disseminating such efforts and
experiences among countries. Beyond sharing experience, governments can also
collaborate on the ground. Efforts to reduce the costs of financing for capital-
intensive components of supply chains in developing economies, by, for example,
pooling investments, is an area where many hands can make for lighter work.

Identify and build strategic partnerships

Strategic partnerships are a way for countries to increase resilience in areas of
manufacturing supply chains where domestic production may otherwise be
uncompetitive. At the same time, such partnerships can facilitate investment in
EMDEs. An appropriate balance should be sought between export opportunities
and support for in-country clean energy transitions and socioeconomic
development. Risks can be mitigated by developing a systematic framework for
identifying and evaluating potential partnerships, rather than proceeding ad-hoc.

Establish a framework for co-operation

It is not always easy for countries to co-operate on interventions that need to
balance the domestic interests of two or more parties with the larger goal of
accelerating the clean energy transition. Having a general framework for co-
operation that takes into account competitiveness factors, trade relationships, and
social, environmental and developmental considerations can help countries
identify the right strategic partner. At a minimum, this framework could include:
I EA. CC BY 4.0.

PAGE | 101
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

 Definition of objective: Identifying strategic partners starts with defining the

objective of the partnership. For some countries the objective may be to increase
the number of suppliers at a certain step in the supply chain; for others it may be
to fill a domestic competitiveness gap, or to find new export markets.
 Competitiveness and resilience considerations: The next step is to gauge
factors that will have a bearing on domestic competitiveness and resilience. Part
II of this report provides an analytical toolkit for examining some cost and non-cost
factors.
 Mutual understanding: It is important that countries come to an agreement on
the definitions and methodologies that underpin the deployment of clean energy
technologies, particularly their upstream inputs like low-emission fuels and
materials.
 Avenues for co-operation: Existing investment and economic co-operation
frameworks, as well as trade agreements, are good “off the shelf” starting points
for strategic partnership discussions. Memoranda of Understanding (MoUs) and
Foreign Direct Investment (FDI) are potentially more expedient routes to
collaboration, where no existing trade agreements are in place.

Figure 32 Framework for establishing strategic partnerships

IEA. CC BY 4.0.
I EA. CC BY 4.0.

PAGE | 102
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

Potential processes and outcomes

There are several types of outcomes and processes that countries may choose to
pursue; each one determined by the objective that a country is hoping to achieve
in the context of its larger industrial strategy. Different outcomes could occur
depending on the stage of the above framework. For instance, as countries assess
competitiveness and resilience considerations, they may choose to partner with
other countries through offtake agreements or co-investment in a project. As
countries work to establish a mutual understanding, pursuing discussion through
multilateral dialogues could be an option. Below is a sample of the types of
outcomes and processes.

Offtake agreements across the supply chain

In general, this type of partnership is based on where individual countries may sit
along clean technology supply chains. It involves finding a partner country that is
at another step in the chain to purchase or offtake a product. This type of
partnership can provide investment security for the exporting partner to develop
capital-intensive clean energy projects, and provide supply security for the
importing partner that may not be able to produce that product competitively.

Such trade agreements are common between companies, but in some cases can
also involve government institutions that are looking beyond their borders to fill in
supply chain gaps. For hydrogen, one of the most developed support mechanisms
of this kind is the H2Global double-auction programme initiated by Germany. A
market intermediary conducts an auction to purchase hydrogen from suppliers
outside of the European Union through fixed price contracts, then a separate
auction is conducted to sell the hydrogen to interested buyers. Using public funds,
the market intermediary covers the price differential. The programme is not limited
to Germany – in fact the Netherlands announced it will dedicate EUR 300 million
in subsidies to use H2Global.

For EVs, this could mean that a country that manufactures EVs enters into a
strategic partnership with another country that can provide anode and cathode
materials needed for EV batteries. For example, Zambia and the Democratic
Republic of Congo signed a co-operation agreement in April 2022 to facilitate
development of the battery supply chain for EVs. The agreement is expected to
provide a framework for bilateral co-operation between these two countries, who
are major producers of cobalt and copper, key critical minerals used in EV
batteries. The United States and European Union have each signed MoUs with
these countries in an effort to secure EV battery supply chains.

On a more detailed level, taking into consideration social and environmental

factors, that same EV manufacturing country may place additional demands on its
suppliers to provide increased transparency on the mineral and material inputs.
I EA. CC BY 4.0.

PAGE | 103
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

For instance, requiring certain labour standards are met for critical mineral
extraction, or opting for near-zero emission steel.

Co-investment and development of a project

This type of partnership involves a shared financial commitment between the two
countries in order to reach a certain level of scale. In this case, a country may
choose to invest in a facility or project located in another country, typically through
direct foreign investment mechanisms, allowing risk to be shared across the
countries. In some cases, this type of partnership could be driven by industry,
where a company identifies its specific needs along the supply chain; in other
cases this could be expanded under a larger government-led strategy to include
a set of projects or targeting a specific sector.

This type of partnership in the energy sector is not new; the initial scale-up of the
liquified natural gas (LNG) market showed that importers seeking access to new
supplies shared much of the risk with the operators of export infrastructure. Japan
was a frontrunner in the development of the LNG market, and it held a 75% share
of global LNG trade through to the late 1980s due to its active development of
contracting and co-investment.

Partnerships based on co-investment and development could help more nascent

supply chains scale up, and simultaneously aid the development of supporting
infrastructure that underpin those supply chains. One example of this type of
partnership is the direct investment in a project in exchange for part of that
project’s output. At the country level this is typically done through state-controlled
enterprises. For example, over the years China has directly invested in cobalt
mines and supporting infrastructure in the Democratic Republic of the Congo
(DRC) to feed its metal refining capabilities. Consequently, the DRC supplies
China with nearly all of its mined cobalt needs and it is estimated that one-third of
China’s imported intermediate products are from mines or smelters in which it has
a stake.

Another example is to leverage development funding as a means to support a

nascent industry in a partner country. This is the case with the proposed HYPHEN
Hydrogen Energy project in Namibia, a massive complex that will use solar and
wind power to produce electrolytic hydrogen. The proposed project, which aims to
export the hydrogen in the form of liquid ammonia, has received, or is in the
process of receiving, funding from a variety of international sources, including the
European Commission, the Netherlands and the United States.

Multilateral dialogues
Establishing a multilateral dialogue on certain clean technology manufacturing
topics may be a potential option to ensure consistency and standardisation across
supply chains. International organisations and fora are natural avenues for
I EA. CC BY 4.0.

PAGE | 104
Advancing Clean Technology Manufacturing Part III
An Energy Technology Perspectives Special Report

countries to discuss supply chain questions, share best practices and

experiences, and co-ordinate assessments.

Multilateral dialogues can be used as a way to remove barriers to the trade of

clean energy technologies. For example, a lack of clarity surrounding what
constitutes low-emissions materials, such as steel and cement, can prevent
countries from procuring these materials if there is commonly agreed standard.
The IEA’s Working Party on Industrial Decarbonisation, established in 2023, is
working to address this barrier by providing a platform for policy officials to discuss
common measurement methodologies. Multilateral dialogues can also provide
alignment on standards for other technologies and fuels, such as the production
of low-emissions hydrogen, or monitoring, reporting and verification requirements
for carbon removal technologies. Without internationally agreed-upon definitions
and approaches, there is a risk that fragmentation in the global market may hinder
global trade. Equally important is the recognition of such standards and definitions
in domestic procurement decisions.

In addition, as some countries and regions finalise subsidies for domestic clean
technology industries, discussions at the multilateral level can provide a platform
to discuss trade concerns and enhance regulatory co-operation. For example, the
WTO Technical Barriers to Trade (TBT) Committee works to clarify proposed trade
measures and enhance alignment with international standards, and could act as
a valuable forum for technical discussions on trade-related aspects of carbon
measurement methodologies and verification procedures.

I EA. CC BY 4.0.

PAGE | 105
Advancing Clean Technology Manufacturing Technical annex
An Energy Technology Perspectives Special Report

Technical annex

Manufacturing capacity, output and demand

Unless otherwise specified, manufacturing capacity and output figures are stated
on an annual (i.e. GW or GWh per year) basis, and in direct current (DC) terms
where relevant. Capacity refers to the maximum rated capacity of the facility.
Default utilisation rates of 85% are used to derive output from capacity, or vice
versa, for 2022-2023, where data for one of these quantities are lacking.

Utilisation rates of 85% are also used to derive forward-looking quantities of output
from existing and announced projects such that a comparison can be made with
deployment levels in IEA scenarios (see Box 4 and Box 5 in Chapter 2 for a
description of the scenarios and manufacturing data categories used in this
report). This level of utilisation is to be interpreted as a practical maximum, and
not as a level that is reflective of all current operations.

In instances where no specific component in a multi-step supply chain is specified,

manufacturing capacity and output figures are stated for the final step. For
example, “solar PV manufacturing capacity”, stated without reference to cells,
wafers or polysilicon, refers to the module assembly (final) step. For batteries, the
final step considered in this analysis is cells. For wind, where the components are
in parallel, nacelles are used as the default capacity/output figure.

Demand numbers for all technologies are approximated using deployment figures
from IEA scenarios in the same year, unless otherwise specified. This is a
simplification for comparative purposes, as not all manufactured units are installed
the same year they are produced.

Beyond these cross-cutting factors, there are a number of technology-specific

considerations for the analytical boundaries used in the analysis.

Solar PV
 Capacity and output figures are stated for each step in the supply chain in series
– polysilicon, wafers, cells and modules.
 Ingot production is included within the wafer production step, with all capacity and
output figures stated for wafers.
 A polysilicon material intensity of 2.9 g/W is used for stating polysilicon in power
units.
 Metallurgical grade silicon production is not considered within the boundary of our
analysis.
I EA. CC BY 4.0.

PAGE | 106
Advancing Clean Technology Manufacturing Technical annex
An Energy Technology Perspectives Special Report

 Demand for solar PV corresponds to total module installations (rooftop and utility
scale) in the same year.
 In this report there has been an update to the method for splitting the committed
and preliminary announced capacity since the last iteration of this analysis.
Specifically, the announced capacity that is due to come online within the next 1-
2 years (i.e. up to end 2025) for modules, cells and wafers, and 2-3 years for
polysilicon (i.e. up to end 2026) is considered “committed” and the rest as
“preliminary”.
 All PV-related capacity numbers are expressed in DC.

Wind
 Capacity and output figures correspond to the final manufacturing step for the
three components considered in parallel: nacelles, blades and towers, including
both onshore and offshore units.
 Upstream sub-components such as generators and gearboxes are not included
within the scope of the analysis, and the figures therefore do not reflect the
capacity or output at these points in the supply chain. Foundations are not included
within the scope.
 Only dedicated facilities for manufacturing towers are included, with the exception
of China, where we assume that an apparent capacity shortfall is being addressed
using generic steel fabrication facilities, which are included within the capacity
figures.
 Demand for wind corresponds to total wind power installations (onshore and
offshore) in the same year.

Batteries
 The supply chain steps considered in the analysis are cells, anodes and cathodes.
 Other components like electrolytes, foils, separators and casings are not included
in the analysis of capacity, output and demand.
 The capacity stated for anodes and cathodes corresponds to the facilities for
making the active material in these components.
 All tiers of battery manufacturing facilities (I-III), indicating both battery
manufacturers certified to serve the EV and stationary market, and manufacturers
currently certified to serve the stationary storage market only, are included in the
capacity and output figures.
 Only lithium-ion batteries are included within the scope of this analysis.
 Demand for batteries includes both electric vehicle and stationary storage
applications, but excludes other segments like consumer electronics, which in
2023 accounted for less than 10% of demand, and which is projected to represent
a minor (<5%) share of battery demand in 2030 and beyond.

Electrolysers
 Only the final assembly step is considered for capacity and output figures.
I EA. CC BY 4.0.

PAGE | 107
Advancing Clean Technology Manufacturing Technical annex
An Energy Technology Perspectives Special Report

 Capacity figures for 2022-23 include facilities used to produce brine electrolyser
units for the chlor-alkali industry.
 The capacity and output figures do not speak to the capacity of the upstream
components such as electrodes.
 All major electrolyser technologies (including alkaline, proton exchange
membrane, solid oxide electrolysis and others) are included in the capacity and
output figures.
 Demand for electrolysers corresponds to electrolysis plant installations in the
same year.

Heat pumps
 Only the final assembly step is considered for capacity and output figures.
 These figures do not speak to the capacity of the upstream components such as
compressors.
 Only heat pumps for residential and commercial buildings for space heating and/or
hot water provision are included in the analysis.
 Reversible air conditioners are included where they are used as primary heating
equipment, i.e. they are not complementary to other equipment such as a boiler.
 Industrial heat pumps are excluded.

Levelised costs of manufacturing

The analytical boundaries used to assess levelised costs of manufacturing (see
Figure 33) are broadly similar to those used in the assessments of capacity, output
and demand (see above). There are some important exceptions. Only the
manufacturing facility for an electrolyser stack assembly is included. All costs
associated with manufacturing the other components that form the balance of
plant for an electrolysis system (e.g. rectifiers, inverters etc.) are excluded. For
batteries, the focus remains on cells, but foils, separators, electrolytes, binders
and casings are also included in the “Other components” cost category for cells.

In instances where the technology category (e.g. wind) comprises multiple designs
or characteristics (e.g. onshore and offshore), the general approach to reflecting
this in our indicative manufacturing cost estimates is to use a deployment-
weighted average. This also applies to battery chemistries, which are summarised
in the latest edition of the Global Electric Vehicle Outlook. For electrolysers, the
levelised cost calculations are based on a plant producing alkaline units.

Beyond these specific considerations for individual technologies, the common

elements of the levelised cost calculations are as follows:

 Unit capital expenditure is derived on an overnight basis using a sample of outturn

costs and capacities for more than 750 plants (see Box 6 in Chapter 3 for a
description of the analysis of capital costs) and is differentiated regionally.
I EA. CC BY 4.0.

PAGE | 108
Advancing Clean Technology Manufacturing Technical annex
An Energy Technology Perspectives Special Report

 Annual fixed operational expenditure is assumed to be 5% of the overnight capital

expenditure.
 Material intensities and global average material prices in 2023 are used to
calculate material costs.
 Energy intensities and regionally differentiated end-user prices in 2023 for
electricity and natural gas are used to calculate energy costs.
 Labour intensities and regionally differentiated manufacturing wages for the most
recent year available are used to calculate labour costs.
 Other components are accounted for as a subset of material costs, and global
average prices are used where relevant, except where a component directly
upstream falls within the boundary of our analysis (e.g. polysilicon for wafer
production).
 To aid comparability, an 85% utilisation rate, a 25-year depreciation period and an
8% weighted adjusted cost of capital are used to compute all levelised costs.
Levelised costs exclude profit margins, expenses associated with company R&D,
and other overheads, and do not reflect supply and demand dynamics. Levelised
manufacturing cost estimates are therefore not intended to align with market
prices for finished units.

Figure 33 Scope of analytical assessment for manufacturing cost and investment

calculations
Capital costs and Fixed operational Other component
Energy costs Labour costs Material costs costs
cost of capital costs

Analytical boundary for costs and investments

e.g. Metallurgical
grade silicon
Polysilicon Wafers Cells Modules Solar PV system

Solar PV
Nacelles

Wind e.g. Generator Blades Wind turbine

Towers

Legend:
Anodes
Examples of other e.g. Battery
Batteries Cells Battery pack
components metals
Cathodes
Core analytical
component
Electrolyser Electrolysis
Electrolysers e.g. Electrodes
stacks system
Global cost factor

Regional cost Heat pump

Heat pumps e.g. Compressors Heat pumps
factor system

Notes: PV = Photovoltaic.
I EA. CC BY 4.0.

PAGE | 109
Advancing Clean Technology Manufacturing Technical annex
An Energy Technology Perspectives Special Report

Manufacturing investment spending

Manufacturing investment spending is calculated using the same analytical
boundary for overnight unit capital expenditure (USD/kW or USD/kWh of annual
capacity) as that for calculating levelised costs of manufacturing (see above).
Overnight capital costs are multiplied by capacity additions (GW or GWh of annual
capacity) for each technology to obtain overnight investments (USD). Investment
spending (USD) is derived from overnight investments using the assumption of an
even distribution of expenditure over the period between final investment decision
(FID) and the start of operations. This period is assumed to be 2 years for all
technologies and components considered in the analysis, apart from solar PV
modules and cell facilities, for which we assume a period of 1 year.

Data sources
The table below summarises the main external data sources used in this report,
which are supplemented by desk research and personal communications with
manufacturers and other technology experts. IEA scenario and modelling data
from the Global Energy and Climate Model are used in conjunction with the data
below.

Table 6 Description of the main data sources used in this report

Technology Data sources Description

InfoLink, BNEF, IEA

PVPS, SPV Marker InfoLink data is the primary source for capacity and output
Solar PV Research, RTS data, supplemented by BNEF for cross-checking.
Corporation

S&P Global
Commodity S&P Global Commodity Insights is the primary data source for
Insights, BNEF, capacity and output data, which are supplement with data from
Wind GWEC, WindEurope, BNEF and GWEC. The Wind Supply Chain
WindEurope, Wood series from Wood Mackenzie and NREL studies were used to
Mackenzie and inform the assessment of levelised costs.
NREL

BMI is the primary data source for current and projected

battery cell manufacturing capacity and for classifying
BMI, EV-Volumes,
Batteries BNEF
announcements as committed or preliminary. BNEF is used as
a supplementary source. EV-Volumes is the primary data
source used to assess demand for EV batteries.

Manufacturing capacity data are based on announcements by

Electrolysers Primary research manufacturers and personal communications, gathered by the
IEA.
Manufacturing capacities are derived combining heat pump
UN Comtrade, sales in different regions and trade flows based on Oxford
Heat pumps Oxford Economics Economics and UN Comtrade. Manufacturing capacity
Trade Prism additions and expansion plans are based on public
announcements by manufacturers.
I EA. CC BY 4.0.

PAGE | 110
Advancing Clean Technology Manufacturing Technical annex
An Energy Technology Perspectives Special Report

Technology Data sources Description

Bloomberg is the primary source for information on material

Bloomberg, IEA prices; IEA data are used for end-user prices for energy and
Other World Energy ILOSTAT data are used to calculate labour costs. Academic
Prices, ILOSTAT literature is consulted to derive material, labour and energy
intensities and to benchmark results for levelised cost.

Note: Many of the data sources are only accessible via subscription – in these instances a link to the data provider’s
website is provided.

I EA. CC BY 4.0.

PAGE | 111
Advancing Clean Technology Manufacturing Abbreviations and acronyms
An Energy Technology Perspectives Special Report

Abbreviations and acronyms

AEM anion exchange membrane

ALK alkaline electrolyser
APS Announced Pledges Scenario
ASEAN Association of Southeast Asian Nations
CBAM Carbon Border Adjustment Mechanism
CCUS carbon capture, utilisation and storage
COP Conference of the Parties
CO2 carbon dioxide
CTM clean technology manufacturing
DRC Democratic Republic of Congo
DWE decoupled water electrolysis
EMDE emerging markets and developing economies
EPO European Patent Office
ESG environmental, social and governance
EUR Euro
EUV extreme ultraviolet lithography
EV electric vehicle
FDI Foreign Direct Investment
FEOC Foreign Entities of Concern
FID final investment decision
GDP gross domestic product
GEC Global Energy and Climate
HS Harmonized System
ICE internal combustion engine
IDR Indonesian rupiah
ILOSTAT International Labour Organization Department of Statistics
IPCEI Important Projects of Common Interest
IPEF Indo-Pacific Economic Framework for Prosperity
IRA Inflation Reduction Act
ISCO International Standard Classification of Occupations
ISIC International Standard Industrial Classification of All Economic Activities
ITC Investment Tax Credit
KRW Korean won
LFP lithium-iron phosphate
LNG Liquified Natural Gas
MEC microbial electrolysis cell
MoU Memorandum of Understanding
NH3 ammonia
NMC lithium nickel manganese cobalt oxide
NZE Net Zero Emissions by 2050 Scenario
I EA. CC BY 4.0.

PAGE | 112
Advancing Clean Technology Manufacturing Abbreviations and acronyms
An Energy Technology Perspectives Special Report

NZIA Net-Zero Industry Act

OECD Organisation for Economic Co-operation and Development
PEM proton exchange membrane
PLI Production Linked Incentives
R&D Research & Development
SMR small modular reactor
SOEC solid oxide electrolyser cell
TBT Technical Barriers to Trade
TOPCon Tunnel Oxide Passivated Contact
UN United Nations
USD United States dollar
VA value added
VAT Value Added Tax
VC venture capital
WACC Weighted Average Cost of Capital
WTO World Trade Organization

Units
g CO2/kWh grammes of carbon dioxide per kilowatt-hour
GW gigawatt
GWh gigawatt hour
kg/MW kilogramme per megawatt
kW kilowatt
kWh kilowatt-hour
m metre
m2 square metre
MW megawatt
MWh megawatt-hour
t tonne
TWh terawatt-hour
W watt
Wh/kg Watt-hour per kilogramme
I EA. CC BY 4.0.

PAGE | 113
International Energy Agency (IEA)

This work reflects the views of the IEA Secretariat but does not
necessarily reflect those of the IEA’s individual member countries or of
any particular funder or collaborator. The work does not constitute
professional advice on any specific issue or situation. The IEA makes
no representation or warranty, express or implied, in respect of the
work’s contents (including its completeness or accuracy) and shall not
be responsible for any use of, or reliance on, the work.

Subject to the IEA’s Notice for CC-licenced Content, this work

is licenced under a Creative Commons Attribution 4.0
International Licence.

This document and any map included herein are without prejudice to the
status of or sovereignty over any territory, to the delimitation of
international frontiers and boundaries and to the name of any territory,
city or area.

Unless otherwise indicated, all material presented in figures and tables is

derived from IEA data and analysis.

IEA Publications
International Energy Agency
Website: www.iea.org
Contact information: www.iea.org/contact

Typeset in France by IEA - May 2024

Cover design: IEA
Photo credits: © Shutterstock