Li-ion batteries for mobility and

stationary storage applications

Scenarios for costs and
market growth

Tsiropoulos I., Tarvydas. D.,

Lebedeva N.

2018

EUR 29440 EN
This publication is a Science for Policy report by the Joint Research Centre (JRC), the European Commission’s
science and knowledge service. It aims to provide evidence-based scientific support to the European
policymaking process. The scientific output expressed does not imply a policy position of the European
Commission. Neither the European Commission nor any person acting on behalf of the Commission is
responsible for the use that might be made of this publication.

Contact information
Name: Ioannis Tsiropoulos
Address: European Commission, Joint Research Centre, P.O. Box 2, NL-1755 ZG Petten, The Netherlands
Email: [email protected]
Tel.: +31 224 56 51 26

EU Science Hub
https://ec.europa.eu/jrc

JRC113360

EUR 29440 EN

PDF ISBN 978-92-79-97254-6 ISSN 1831-9424 doi:10.2760/87175

Luxembourg: Publications Office of the European Union, 2018

© European Union, 2018

The reuse policy of the European Commission is implemented by Commission Decision 2011/833/EU of 12
December 2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Reuse is authorised,
provided the source of the document is acknowledged and its original meaning or message is not distorted. The
European Commission shall not be liable for any consequence stemming from the reuse. For any use or
reproduction of photos or other material that is not owned by the EU, permission must be sought directly from
the copyright holders.

All content © European Union 2018, except: cover page, © Stillfix. Source: AdobeStock

How to cite this report: Tsiropoulos, I., Tarvydas, D., Lebedeva, N., Li-ion batteries for mobility and stationary
storage applications – Scenarios for costs and market growth, EUR 29440 EN, Publications Office of the
European Union, Luxembourg, 2018, ISBN 978-92-79-97254-6, doi:10.2760/87175, JRC113360
Contents
Abstract ............................................................................................................... 1
Acknowledgements ................................................................................................ 2
Executive summary ............................................................................................... 3
1 Introduction ...................................................................................................... 6
2 Current situation of Li-ion battery markets and costs ............................................. 9
2.1 Li-ion battery markets .................................................................................. 9
2.1.1 Electric vehicles ................................................................................... 9
2.1.2 Stationary storage ............................................................................. 12
2.2 Li-ion battery costs and cost structure .......................................................... 13
2.2.1 EV battery pack costs ......................................................................... 13
2.2.2 Stationary battery system storage costs ............................................... 17
2.2.3 Cost reduction drivers ........................................................................ 22
3 Future growth and costs of Li-ion batteries ......................................................... 24
3.1 Market growth of Li-ion batteries ................................................................. 24
3.1.1 Near-term manufacturing capacity growth globally and in the EU ............. 24
3.1.2 Global long-term manufacturing capacity growth ................................... 26
3.1.3 Global market growth of electric vehicles .............................................. 27
3.1.4 Global market growth of stationary storage ........................................... 28
3.2 Future costs of Li-ion batteries for EVs and stationary storage ......................... 29
4 Scenario-based cost trajectories of Li-ion batteries .............................................. 33
4.1 Method ..................................................................................................... 33
4.1.1 Learning rates ................................................................................... 33
4.1.2 Selected global growth scenarios ......................................................... 33
4.1.3 Implicit assumptions .......................................................................... 35
4.2 Cost trajectories ........................................................................................ 36
4.3 Sensitivity scenarios .................................................................................. 41
5 Conclusions .................................................................................................... 45
References ......................................................................................................... 49
List of abbreviations and definitions ....................................................................... 57
List of boxes ....................................................................................................... 58
List of figures ...................................................................................................... 59
List of tables ....................................................................................................... 61
Annexes ............................................................................................................. 62
Annex 1. Cost components and cost boundaries of Li-ion batteries......................... 62
Annex 2. Cost components and cost boundaries of Li-ion batteries......................... 63
Annex 3. Parameters and detailed input assumptions ........................................... 64
Annex 4. Results ............................................................................................. 67

i
Abstract
Li-ion battery costs could decrease rapidly, by at least 50 % in 2030 and up to 75 % in
2040, due to learning from mass production driven by electric vehicles. Stationary
storage systems may benefit from somewhat slower yet substantial cost reduction of
65 %. Market barriers or inaction on climate goals can affect these trajectories.

1
Acknowledgements
The authors would like to acknowledge the support of Andreas ZUCKER (DG ENER) at
early stages of this work.
The authors would also like to express their gratitude to JRC colleagues for reviewing the
report and offering their feedback: Ignacio HIDALGO GONZALEZ, Konstantinos
KAVVADIAS, Darina BLAGOEVA, Leonidas MANTZOS, Evangelos TZIMAS, Efstathios
PETEVES and Fulvio ARDENTE.
This acknowledgment extends to Brittney ELZAREI and Jean-Michel DURAND of the
European Association for Storage of Energy (EASE) for providing constructive comments.

Authors
TSIROPOULOS, Ioannis
TARVYDAS, Dalius
LEBEDEVA, Natalia

2
Executive summary
Recent cost reduction of Li-ion batteries raise the expectations that electric vehicles and
energy storage at grid and/or household level will become cost-competitive and will
penetrate the respective markets in great numbers. Based on announcements, the global
Li-ion cell manufacturing capacity is expected to quadruple or even increase six times by
2021 – 2022 compared with 2017 levels. By 2040, 150 to 900 million electric vehicles
are projected to be on the road, which is two to three orders of magnitude higher than
today. Over the same period, stationary storage may reach up to 1 300 GWh, compared
with about 3 – 4 GWh installed front-of-the-meter today. These projections point towards
a potentially significant market growth of Li-ion batteries, but also towards a range of
views on the magnitude of these developments. While in the near-term the global
manufacturing capacity is set to increase, in the longer-term the range of projections is
wide. The projections depend on the direction the world will take, for example, in view of
action against climate change by decarbonising road transport. As such, cost trajectories
of Li-ion batteries may be influenced by the total deployment levels of electric vehicles
and stationary storage due to economies of scale, and by the cumulative manufacturing
experience gained globally.
Focusing on Li-ion batteries as the family of batteries for mobility and stationary storage
applications of today and the near future, this report contextualises their potential cost
trajectories in line with global production scale, based on three different scenarios for the
global energy system up to 2040 (high, moderate and low).

EU policy context
The EU is transitioning to a secure, sustainable and competitive energy system as laid
out in the European Commission's Energy Union strategy. Li-ion batteries are often seen
as the technology that can help decarbonise transport, lift the penetration levels of
intermittent renewable energy (wind and solar) and offer a competitive edge to the EU
industry in the Li-ion battery value chain. Batteries, including Li-ion, are recognised as a
key enabling technology for the energy transition of the EU under the Energy Union and
as such, they are specifically mentioned in several policy initiatives that address
transport, raw materials and energy economic sectors, EU industrial policy and EU
Research and Innovation. The strategic importance of batteries for the EU is further
demonstrated by the formation of the European Battery Alliance and the adoption of the
Strategic Action Plan for batteries as an integral part of the third 'Europe on the Move'
package. Li-ion batteries also link with the European Commission's actions on raw
materials, namely the Raw Materials Initiative, the European Innovation Partnership on
Raw Materials and the assessments on Critical Raw Materials.

Main findings and conclusions

By 2040, according to key projections, the global annual sales of Li-ion batteries increase
to around 4 TWh in the high, 2 TWh in the moderate and 0.6 TWh in the low scenario.
This corresponds to about 110, 55 and 15 operational gigafactories in each scenario,
respectively, assuming 35 GWh annual production capacity for each gigafactory. For
comparison, the sales volume in 2017 was about 60 GWh.
These projections entail that by 2030, Li-ion electric vehicle battery packs could come at
least at half the cost of today's production (100 €/kWh instead of about 200 €/kWh
today) due to mass production. By 2040, the cost could drop an additional 50 %,
ultimately reaching 50 €/kWh. Such trajectories are feasible based on costs of new Li-ion
cathode chemistries and other battery pack materials that are estimated at around
30 €/kWh (other additional costs are estimated at around at 10 – 20 €/kWh). Meeting
policy goals such as the Strategic Energy Technology Plan (SET Plan) cost target of
75 €/kWh is feasible in both high and moderate scenarios (which entail fast ramp-up of
Li-ion manufacturing capacity, of about 2 gigafactories globally per year until 2030).

3
Cost reduction of Li-ion battery packs for electric vehicles spills over to stationary storage
systems, but cost reduction in this sector occurs somewhat slower due to the
contribution of other major cost components (e.g. inverters, balance of system hardware,
soft costs such as engineering, procurement and construction). The benchmark costs
of Li-ion stationary storage systems in 2017 were about 500 €/kWh for energy-designed
systems, about 800 €/kWh for power-designed systems, and 750 €/kWh for residential
batteries (1). Ultimately, by 2040, stationary storage system costs will range between
165 and 240 €/kWh for energy-designed utility-scale systems, between 280 and
410 €/kWh for power-designed utility-scale systems and between 250 and 365 €/kWh for
households. Lowering of balance of system and other soft costs can potentially help
further cost reduction of stationary energy storage systems, lifting barriers for their
widespread deployment. The SET Plan target on investment costs for stationary storage
at a system level at 150 €/kWh (2) could be attainable at high global deployment of EVs
and more than 1 TWh of stationary storage. These findings are influenced by the
assumed learning rates, the assumption that cost reduction from Li-ion battery packs for
EVs spills over to stationary storage systems, and on whether aged batteries from
transport are repurposed to storage at a portion of the price of a new battery pack. Other
assumptions, such as the evolution of the battery size or the battery lifetime were not
found to influence significantly the cost projections based on learning.
The main messages to take away from this analysis are the following:
 Wide-spread deployment of electric vehicles will lead to a rapid decrease of Li-ion
battery costs in the near term.
 Investment costs of Li-ion battery stationary storage systems will decrease, yet
improvements should focus also on non-battery pack system components.
 European manufacturing of Li-ion battery cells will increase its share in global
production, provided that announced plans materialise. Supplying domestic
demand may prove challenging if capacity does not ramp up after 2025.
 Re-using and repurposing of Li-ion batteries to energy storage applications after
their end of life in electric vehicles contributes to further cost reduction.

Related and future JRC work

This report complements a series of other JRC publications on key issues related to Li-ion
batteries, namely the competitiveness of the EU in the Li-ion battery sector and
opportunities for Europe in the Li-ion value chain. Future work could relate with economic
assessments to complement modelling activities on the feasibility of energy storage or
other economic metrics such as levelised costs, lifecycle costs and so forth.

Quick guide
This report is structured as follows: Chapter 1 introduces the policy context around Li-ion
batteries, their relevance in the energy transition and the knowledge gaps on deployment
and costs. Chapter 2 continues with an overview of historical developments of Li-ion
batteries markets and costs, focusing on EVs and storage. Chapter 3 presents near-term
expectations and long-term projections for growth and costs of Li-ion batteries based on

(1) Batteries for stationary storage are used for a range of applications with some being more suited to store
energy and others to supply power. In the present report, batteries that can provide energy for more than
1 hour are called energy-designed and batteries that can provide energy for less than 1 hour are called
power-designed. Smaller scale residential batteries provide energy for more than 1 hour and do not require
system integration components. The cost structure of these batteries is different both per kW and kWh.
(2) The SET Plan target for stationary energy storage reads: "For stationary energy storage the SET‐Plan R&I
will aim at developing and demonstrating technology, manufacturing processes, standards and systems,
which have the potential of driving high‐efficiency (>90%) battery based energy storage system cost below
€150/kWh (for a 100kW reference system) and a lifetime of thousands of cycles by 2030 to enable them to
play an important role in smart grids" [23]. Although only the investment cost is cited here and hereafter,
it is imperative to consider it in conjunction with the requirements on efficiency, power rating and lifetime.

4
literature. Chapter 4 provides estimates on scenario-based cost trajectories for Li-ion
batteries for mobility and stationary storage applications based on the learning curve
method. Chapter 5 draws the final conclusions of the analysis.

5
1 Introduction
The EU is transitioning to a secure, sustainable and competitive energy system as laid
out in the European Commission's Energy Union strategy [1]. Energy storage, and in
particular batteries, is frequently addressed as the technology that may unlock the
transition to a decarbonised and clean energy system due to their potentially broad
application in the power sector and in transport [2–4].
In the proposed "Clean Energy for all Europeans" legislative package, the EC sets global
leadership in Renewable Energy Sources (RES) as a priority for the EU (COM(2016) 860
final; [5]). To achieve this goal and deliver on international climate change mitigation
commitments [6], at least 32 % of energy supply must be sourced by renewables as
agreed by the European Parliament and Council [7]. The power sector, however, could
be technically challenged by high supply of electricity from intermittent RES, even more
so as costs of wind and solar decrease, making them more appealing to the market [8–
11]. Wind turbines and photovoltaics generate electricity when the resource is available,
but not necessarily when electricity is needed. At the grid level, batteries offer an
electricity storage option that can moderate the variability of intermittent RES and
increase their share. They do so by providing reliable grid services (e.g. peaking
capacity, frequency and voltage control, peak shaving, congestion management, black
start), which become increasingly important in the face of baseload fossil fuel-based
assets phasing out. In behind-the-meter applications, batteries improve power quality,
reduce demand charge, and increase the reliance on self-generation. In integrated
systems supported by smart market designs, batteries may contribute to decentralisation
and the shift of consumers to prosumers, thereby empowering the participation of the EU
citizens in the energy market as envisaged in "Clean Energy for all Europeans" legislative
package [5].
Transport is a highly contributing sector in the EU that has been witnessing growth in
greenhouse gas emissions [12]. Road transport accounts for about 73 % of all
greenhouse gas emissions in transport [12]. Together with low-carbon options such as
hydrogen or advanced biofuels, the deployment of Electric Vehicles (EVs) at large scale
seems to be a prerequisite in order to transform the sector to a low-emission activity. For
a safe, clean and connected mobility, the EC promotes electro-mobility and
communicates the significance of batteries in the sector's clean energy transformation
and in the competitiveness of the EU's automotive industry with the "Europe on the
Move" and "Delivering on low-emission mobility" packages (COM (2018) 293 final, COM
(2017) 283 final, COM (2017) 675 final); [13–15]). The important role that batteries will
have in a modern and competitive automotive EU industry is acknowledged in the 2017
"A renewed EU Industrial Policy Strategy" communication (COM (2017) 479 final; [16]).
Batteries also stand at the interface of power and transport supporting their sectoral
integration. In the longer term, coupling these sectors may introduce cost efficiencies in
the system and help bring their emissions closer to zero [17,18].
For anticipated boosted future demand for Li-ion batteries the sustainable and secure
supply of raw materials (e.g. lithium, cobalt) is of strategic importance for the EU.
Scenarios on market growth of Li-ion batteries therefore also link with the European
Commission's actions on raw materials, namely the Raw Materials Initiative [19], the
European Innovation Partnership on Raw Materials [20] and the assessments on Critical
Raw Materials [21].
The EU's Research and Innovation agenda on batteries is set in the frame of a dedicated
Key Action of the Integrated SET Plan (C(2015) 6317 final; [22]), where the Declaration
of Intent defines targets for performance, cost, recycling and manufacturing of batteries
and the Implementation Plan outlines the research actions and priorities to meet the
agreed targets [23,24]. The strategic importance of batteries for the EU is further
demonstrated by the formation of the multi-stakeholder group "European Battery
Alliance" [25] and the adoption of the Strategic Action Plan for batteries as an integral
part of the third 'Europe on the Move' package.

6
One of the barriers that delay the large scale deployment of batteries, and especially of
lithium ion (Li-ion) batteries, is their high capital investment costs. Even though Li-ion
battery prices fell almost 80 % since 2010 [26] they remain substantial. In the power
sector investment costs of stationary storage systems are still too high to justify a
business case, partly owing to the market design [27,28]. Similarly, although the uptake
of photovoltaics in households has been increasing, their integration with home battery
systems is limited and so far driven by preference for self-sufficiency rather than
decisions on return on investment [29,30]. In transport, upfront and total costs of
ownership of EVs are still high; more than 50 % of EV costs are attributed to battery
packs [31]. However, with strong government support in some countries, the total cost
of ownership of EVs is already in parity or even lower, compared to internal combustion
engine vehicles [32,33]. Depending on oil prices, taxation policy and use profiles, studies
expect that EVs will reach cost parity with internal combustion engines in most regions
by mid-2020 towards 2030, should battery costs decrease [26,31].
Targeted EC policies recognise that costs need to decline for mass adoption of batteries
in mobility and stationary storage applications. For example, the development of
affordable and integrated energy storage solutions is a priority area stated in the
"Accelerating Clean Energy Innovation" communication [34], and cost reduction has been
the subject of several EU-funded projects [35]. Moreover, the "Declaration of Intent of
SET Plan Key Action 7" sets a target of 75 €/kWh for a battery pack for automotive
applications and 150 €/kWh for stationary storage applications at a system level by 2030
[23] (3). The aim of these policies is straightforward: to help the EU become the global
leader in sustainable battery production and use.
Dynamics of the Li-ion battery sector –currently the fastest growing battery type–
outside the EU, however, exerts a significant influence on technological progress,
innovation and costs. While the EU has strong presence in downstream segments of the
value chain (e.g. battery pack assembly, recycling and re-purposing) [36], slightly less
than half of the battery pack costs lie in cell manufacturing [26,37], where the position of
the EU needs to be strengthened. The existing and announced manufacturing capacity of
Li-ion cells is mainly in Asia (China, Japan and South Korea) [26,36]. Besides
manufacturing capacity, Asian countries now also have the lion's share in EV sales,
reaching about 55 % of global sales in 2017 [38]. Under existing or near-term market
conditions this trend could continue, as more than one-third of the global EV fleet is
projected to be deployed in China and about one-fifth in Europe by 2040 [26]. Moreover,
the size of the EV market strongly depends on the global action required to meet agreed
climate goals. Based on existing policies, the global EV stock could reach 60 million cars
in 2030, yet this projection could more than triple should climate goals below the 2 oC
target be attained [31]. Differences in market share are even more pronounced on
batteries for stationary storage. By 2040, about 30 % of the global capacity may be
installed in China while slightly more than 10 % in Europe [39].
With these possible developments ahead, future costs of batteries may ultimately depend
on the scale and the cumulative manufacturing experience gained globally, a relationship
that has been empirically observed for several other technologies [40,41]. Despite the
plethora of assessments on how Li-ion battery costs may develop, they do not always
capture one or more of the following aspects: a) due to synergies or competition, the
market and production rate for batteries also depend on deployment of other
technologies in the energy system (e.g. residential photovoltaics), RES or climate change
mitigation goals, b) system costs of batteries for stationary storage may benefit from
innovations and production scales of battery packs for EVs, as they use similar or the
same electrochemistry, materials and manufacturing process, c) recent cost reductions in

(3) Taking into account the cost of electrochemical modules, the cost of the inverters and power electronics
and the installation and integration costs. This target refers to high efficiency (>90 %) battery-based
energy storage system (100 kW reference system) and a lifetime of thousands of cycles by 2030. SET Plan
Key Action 7 has a separate cost target for stationary applications per cycle. This target is not assessed in
the present report.

7
batteries have not been included in frequently cited price forecasts, and as shown in the
case of RES technologies, effects could be substantial [42]. Energy system models that
are used for assessments of policies and technologies ( 4) so far rely on cost trajectories
that have paid limited attention to these aspects. As such, their results and the
information provided to policy makers may be influenced to the extent they relate with
battery costs. These are caveats that need to be addressed to inform the EC policy
process.
Focusing on Li-ion batteries as the family of batteries for mobility and stationary storage
applications of today and the near future, this report contextualises their potential cost
trajectories in line with the global production scale based on different technology
developments in the energy system and scenarios up to 2040. This report complements a
series of other JRC publications on key issues related to Li-ion batteries, namely the
competitiveness of the EU in the sector [43] and opportunities for Europe in the Li-ion
value chain [36].

(4) Examples of such models are PRIMES [124], POTEnCIA [125] and JRC-EU-TIMES [126].

8
2 Current situation of Li-ion battery markets and costs

2.1 Li-ion battery markets

Since their introduction into the mass market in 1990, Li-ion batteries have been used in
applications such as electronics, medical devices and power tools. By 2010, their total
market volume increased one order of magnitude (from about 2 to 20 GWh), reaching a
total annual market value of about 6.5 bn € largely owing to portable electronics [44].
From 2010 onwards Li-ion batteries have been growing annually at 26 % in terms of
production output and 20 % in terms of value (5). In 2017, the total market size of Li-ion
batteries was about 120 GWh (24 bn €).
The electronics market gradually slowed down, as indicated by the observed 6 % annual
growth in 2010 – 2017 (Figure 1) and is expected to saturate further as shown by near
term forecasts (section 3). Despite the saturation of the electronics market, the sales of
Li-ion batteries continued to grow primarily due to the staggering demand from EVs, the
niche yet fast growing industrial and stationary storage applications, and the slower but
substantial growth and size of other markets such as electric bikes (Figure 1). Within four
years since their introduction, annual sales of Li-ion batteries for EVs surpassed those for
electronics. They have been witnessing an average year-over-year growth of 67 % and
these trends are expected to continue (section 3). Li-ion batteries for stationary storage
begun to emerge in large market volumes around 2011 and since then their total
installed capacity increased rapidly to reach about 2 GWh in 2017 [26]. With the
penetration of renewable energy increasing, Li-ion battery sales for stationary storage
may grow even faster (section 3). Overall, the market share of Li-ion batteries for EVs
and stationary storage increased from about 5 % early this decade to more than 60 % in
2017.

Figure 1 Global historical annual growth Li-ion batteries in main market segments

140
Other
120 Energy storage & industry
EV
Li-ion battery sales

100 Electronics
[GWh/yr]

80

60

40

20

0
2010 2011 2012 2013 2014 2015 2016 2017

Source: JRC based on Avicenne Energy [44]. Note: Data include sales and stock. Electronics includes mainly
portable electronics, EV include BEV, PHEV and electric buses, Energy storage & industry includes stationary
storage, UPS, telecom, and applications in industry, Other includes medical devices, power tools, electric bikes
and gardening tools.

2.1.1 Electric vehicles

More than 3 million passenger light duty EVs were on the road in 2017 (Figure 2), which
represent about 0.2 % of the total global passenger car fleet [26,45]. Cumulative sales,
of EVs including electric buses surpassed 4 million in mid-2018 [46]. Since their

(5) Compound annual growth rate.

9
introduction to the mass market in 2010, the annual sales of EVs exceeded 1 million cars
for the first time in 2017 (excluding non-plug in hybrid electric vehicles; Figure 3).
Battery Electric Vehicles (BEV) represent two-thirds of total EV sales globally (Figure 3).
In specific markets, however, such as in Japan, Plug-in Hybrid EVs (PHEV) have the lion's
share (two-thirds of new sales). In the EU, BEVs and PHEVs are currently sold annually in
roughly equal amounts [38,47,48]. The EU represents a sizeable market (15 % of new
EV sales in 2017), but more than 50 % of new EVs are nowadays sold in China.

Figure 2 Total global EV fleet (excluding electric buses) in different regions in 2010 – 2017

3 500
Norway
3 000 Other
EV total car fleet
[1 000 vehicles]

2 500 Japan
EU 28
2 000
USA
1 500 China
1 000

500

2010 2011 2012 2013 2014 2015 2016 2017

Source: JRC based on IEA [45]. Note: EU 28 represents the markets of Finland, France, Germany, the
Netherlands, Portugal, Sweden and the UK.

Figure 3 Annual new EV sales per EV type (BEV or PHEV)

1 400
PHEV
Annual new EV sales

1 200
BEV
[1 000 vehicles]

1 000
800
600
400
200

IEA BNEF IEA BNEF IEA BNEF IEA BNEF IEA BNEF IEA BNEF IEA BNEF
2011 2012 2013 2014 2015 2016 2017

Source: JRC based on IEA [45] and BNEF [38].

Electric buses are a market segment of EVs that is quickly gaining market share,
especially in China [26]. A fleet of about 386 000 electric buses in 2017 is reported,
which is roughly 10 % of the global EV fleet [26]. In mid-2018 the cumulative sales of
electric buses reached 421 000 [46]. About 97 % of electric buses and 75 % of their
batteries are currently produced in China [49]. Electric buses have large battery capacity
(60 to 550 kWh [50]), and their demand for Li-ion batteries is sizeable and comparable
with that of passenger light duty EVs (Figure 4).
Based on annual sales, the weighted average battery capacity of BEV is around 39 kWh
and of PHEVs around 11 kWh. The EV performance characteristics vary depending on the

10
producer (Figure 5). On BEV, US producers clearly stand out due Tesla vehicles, which
have large-sized batteries (75 to 100 kWh depending on the model) and consequently
long range. In terms of efficiency (6), however, Chinese and Korean manufacturers are
leading as they focus on producing lighter vehicles. PHEV manufacturers focus on
different properties as shown, for example, by the low range and efficiency of vehicles
produced by EU companies which are more power-oriented (e.g. acceleration of
roadsters, load of sport utility vehicles (7)) compared with Chinese PHEVs.

Figure 4 Li-ion battery demand for electric buses in 2017/2018

50
40
Electric buses
[GWh]

30
20
10
0
BNEF Avicenne IDTechEx

Source: JRC based on different literature sources [47,51,52]. Note: Avicenne Energy data include sales and
stock in 2017 [47]. IDTechEx data represent their forecast for 2018 [52].

Figure 5 Weighted average performance of BEV and PHEV based on sales in 2013 and 2017 per
producer

(6) Defined as km/kWh.

(7) Notably, the efficiency drop of EVs between 2013 and 2017 made by EU producers, is attributed to the
increase in the share of sales of BMW i3 in 2017, which has larger battery capacity but lower range
compared to Renault Zoe ZE sold in 2013. Similarly, the efficiency drop of PHEVs is due to the shift of sales
concentrated to Volvo V60 and Opel Ampera in 2013 to a diverse set of PHEV models in 2017 such as
BMW 330e, xDrive40e, Volkswagen Passat GTW, Audi A3 e-tron, etc.

11
Source: sales data from BNEF [38] and technical specifications from WattEV2Buy [53], High Edge [54] and ITRI
[55]. Note: country grouping based on headquarter location of the producer (see Annex 1 for classification of
producers per country). Sales data include models that sold more than 8 100 since 2011 (i.e. 93 % of total light
duty EV sales).

2.1.2 Stationary storage

In 2017, global capacity of stationary storage exceeded 1.5 TWh (160 GW) [56]. The
vast majority of installed capacity is pumped hydropower (more than 98 %).
Electrochemical system storage is about 1.6 GW (2.8 bn €) and Li-ion batteries are the
main type used (8), accounting for about 1.3 GW (0.85 bn €) or 81 % (in terms of power
capacity) of all electrochemical system storage in 2017 (Figure 6). The total battery
energy storage market value is roughly equally divided between lead-acid, li-ion and
other battery technologies. Since their large-scale introduction in 2011, the sales of Li-
ion batteries for stationary storage have witnessed an overall growth, although sales
have not been following a constant increase, as shown by the drop between 2016 and
2017 (Figure 7). Nonetheless, Li-ion batteries for stationary storage applications are by
far the technology with the strongest growth and market share compared with other
electrochemical energy storage options.

Figure 6 Global cumulative installed capacity of electrochemical system storage

2 000
Other
Total installed capacity

1 600 Capacitor
Sodium
1 200 Flow
[MW]

Nickel
800 Lead acid
Li-ion
400

0
2010 2011 2012 2013 2014 2015 2016 2017

4.0
Other
Total installed capacity

3.5 Capacitor
3.0 Sodium
[GWh]

2.5 Flow
Nickel
2.0
Lead acid
1.5
Li-ion
1.0
0.5
0.0
2010 2011 2012 2013 2014 2015 2016 2017

(8) A third of rated power capacity of sodium-sulphur batteries is built in Japan, including a large-scale facility
of 50 MW/300 MWh built for arbitrage, in 2016. Sodium-sulphur batteries are characterised by high
energy-to-power ratio, thus capacity expressed in energy terms (Figure 6 lower figure) is shown to
represent more than one-third of global electrochemical capacity in 2017.

12
 Source: JRC based on US DOE [56]. Note: including only grid-connected energy storage; the reported power
(kW) and storage duration were used to calculate the total installed capacity of the projects in GWh.

Deployment of residential Li-ion batteries behind-the-meter, was estimated at around

600 – 650 MWh (or about 200 MW) in 2016 [57,58], which is not negligible considering
that it represents almost 20 % of the total Li-ion battery capacity installed for system
storage. BNEF reports additional behind-the-meter storage capacity of 650 MWh in
commercial and industrial sectors [58]. Other sources estimate that global installed
capacity of behind-the-meter storage was 2 300 MWh, without, however, specifying the
battery type or the sectoral application [59] (Figure 8).

Figure 7 Global annual sales of electrochemical storage

500
Other

400 Capacitator
Annual sales [MW]

Sodium
300 Flow
Nickel
200 Lead acid
Li-ion
100

0
2010 2011 2012 2013 2014 2015 2016 2017
Source: JRC based on US DOE [56].

Figure 8 Global installed capacity of energy storage installed behind-the-meter in 2016

2 500
Behind-the-meter storage

Unspecified
2 000 Industrial
Commercial
1 500
[MWh]

Residential

1 000

500

Schmidt et al. 2017 BNEF IEA

Source: JRC based on different literature sources [57–59].

2.2 Li-ion battery costs and cost structure

2.2.1 EV battery pack costs

One of the most widely cited overviews of EV battery pack costs is provided by Nykvist
and Nilsson [60], which is summarised and complemented with recent data in Figure 9.
The figure reveals a wide variation of reported historical costs, which on average range

13
about ± 60 % from observed values reported by market analysts (9). Observed costs of
EV battery packs based on BNEF have decreased from about 870 €/kWh in 2010 to 170 –
215 €/kWh in 2017 (Figure 9) [26]. The range or reported values from the observed
values of BNEF (190 €/kWh in 2017) is about ± 15 % in 2017. The lower end of the cost
range of 2017 coincides with announcements of market leaders, like Tesla, at about
170 €/kWh [61].

Figure 9 Reported Li-ion battery pack costs for EVs

1 600
Reported values
Observed average (BNEF)
EV battery pack costs

1 200 Average of reported values

[€/kWh]

800

400

0
2010 2011 2012 2013 2014 2015 2016 2017

Source: JRC based on Nykvist and Nilsson [60], IEA's World Energy Outlook [62,63], IEA's Global EV Outlook
[45], SET Plan [23], Boston Consulting Group [64], IDTechEx [52], Deutsche Bank [65] and Avicenne Energy
[66]. Note: Observed average prices are based on BNEF's industry survey and are a volume-weighted average
of Li-ion battery pack prices for EVs [26,67]. Reported average is the simple average of the data included in the
graph.

One reason for the spread is that announcements made by market leaders were typically
lower than what the industry reported [60]. Reported metrics are not always consistent
across the different sources, as these may represent either production costs or market
prices of battery packs. For example, the difference between these two metrics was
found to be about 15 % in 2015 [68]. Data published by Avicenne Energy shows low
profitability or operation at loss of Li-ion cell manufacturing business with Earnings
Before Interest & Tax (EBIT) in 2016 of market leaders such as Panasonic, LG Chem and
Samsung SDI being 0 %, -1 % and -13 % [47]. Large size and multi-sectoral business
structures may enable some battery manufacturers to offer their products at a very
competitive price and to absorb large financial losses caused, for example, by product
recall (e.g. the recall of Galaxy Note 7 due to the safety issues with its battery costed
Samsung 5.3 bn $ [69]). A factor that works in the opposite direction, i.e. towards price
increase, is illegal price fixing. Several Li-ion cell manufacturers, including Samsung,
LG Chem, Sony, Sanyo, Panasonic, NEC, Toshiba and Hitachi, have been accused of a
long-term antitrust violating price-fixing conspiracy that kept prices for cylindrical Li-ion
batteries artificially high from 2000 to 2011 [70]. While settlements have been reached
with Sony, LG Chem, NEC and Hitachi, the case is continuing against the remaining Non-
Settling Defendants. The above reported factors and practices underline the important
difference between the terms “price” and “cost” in the context of the global Li-ion sector
and offer additional insights into the reasons behind widely varying numbers reported.

(9) Compared with the volume weighted average of BNEF's price survey [26].

14
Different battery sizes could offer another explanation, as large batteries tend to have
lower specific costs [45] (10). Battery chemistries (11) (Table 1, Figure 10), cell quality, or
cell size and format also affect the cost of a battery pack, with a 18650 cylindrical cell
being about 30 % cheaper than a large prismatic EV cell [45,47]. The material
requirement of different cathode chemistries is shown in Figure 11 (12). Nickel-
Manganese-Cobalt oxide (NMC-111) cathodes are one of the main types used in EV
batteries (e.g. by Nissan Leaf, BMW i3, GM Chevrolet Bolt [71]). Besides the EV market,
behind-the-meter storage applications, such as Tesla's Powerwall, also use cells with
NMC cathodes [72,73]. NMC and Lithium-Iron-Phosphate (LFP) batteries seem to be the
mainstream choice for storage applications from 2017 onward [74]. The industry aims at
reducing cobalt demand in cathodes by developing and bringing into the market nickel-
rich NMC-811 cathodes (e.g. LG Chem, BYD, SK innovation [74–76]). This may reduce
supply risks on critical materials and improve battery pack costs and performance (e.g.
capacity, energy density) [71,77–79]. Some experts and analysts see a wide adoption of
NMC-811 cathodes in the EV market after 2025 [66,77], while others expect it to become
the primary choice by the early 2020s [26] (Figure 12). The second type of cathodes with
widespread commercial use is Nickel-Cobalt-Aluminum oxide (NCA) mainly produced by
Panasonic for Tesla EVs [80]. These cathodes use about 65 % less cobalt than NMC-111,
which could at least partially explain the difference between Tesla's announcements and
other manufacturers on reported costs [72].

Table 1 Main cathode chemistries used in Li-ion battery packs and their application

Main cathode chemistries Main application

Nickel Manganese Cobalt oxide (NMC) EVs, storage, other (e-bikes, medical devises,
industrial)

Nickel Cobalt Aluminum oxide (NCA) EVs, storage, other (medical devices,
industrial)

Lithium Cobalt Oxide (LCO) Portable electronics

Lithium Manganese Oxide (LMO) Power tools, medical devices

Lithium Iron Phosphate (LFP) EVs, electric buses, storage

Based on the weighted average of the global EV car fleet on the road between 2011 and
2017, NMC batteries represent 53 % of the market, NCA batteries 46 % and the
remainder are LFP or other chemistries ( 13). The predominant battery chemistry for
electric buses is LFP and their main market is in China (and to a lesser extent NMC
cathodes) [49]. In 2018, 88 % of the battery chemistries used in electric buses were LFP
[49]. By 2028, it is expected that NMC batteries will gain market share in this segment
(42 % NMC and 58 % LFP chemistries [49]).

(10) IEA mentions that a 70 kWh battery is expected to have a 25 % lower cost per unit of energy stored than a
30 kWh battery, due to the higher cell-to-pack ratio of the former [45].
(11) Based on IEA, the cost of a NMC-111 battery is about 5 % higher than the cost of an NCA battery [45].
(12) When taking into account the process yield of the battery (by cell format and application), the actual
material consumption is higher than the element composition shown in Figure 11. Based on Avicenne
Energy, the cobalt needs for NMC-111 were 0.49 kg/kWh in 2015 decreasing to 0.41 kg/kWh by 2025, and
for NCA from 0.22 kg/kWh to 0.18 kg/kWh over the same period [66]. Back of the envelope estimates
from other sources and for unspecified cathode chemistry, indicate much higher range for cobalt use from
0.36 up to 1.44 kg/kWh [78].
(13) This estimate excludes China, which historically has been relying mainly on LFP batteries and is now
gradually shifting to NMC batteries [127,128].

15
 Figure 10 Breakdown of the total cost of Li-ion EV battery in key components (upper figure) and
between cell components and other costs across different chemistries (lower figure)

100%
EV battery pack cost

75%
breakdown

50%

25%

0%
BNEF IDTechEx

Materials Capital Operating Labour

100%
EV battery pack cost

75%
breakdown

50%

25%

0%
NMC-111 LFP NMC-111 NCA NMC-622 NMC-811 NMC
BNEF IEA Avicenne
Energy

Cathode Anode Separator Electrolyte Other costs

Source: JRC based on BNEF, IDTechEx (upper figure; [26,52]) and BNEF, IEA and Avicenne Energy (lower
figure; [26,45,66]).

Based on the average of raw material prices over the last 8 years, the majority of the
materials for the cell and pack of NMC-111 batteries cost about 37 €/kWh, with cobalt
and copper representing about two-thirds of that cost. The majority of the materials in
NMC-811 batteries cost about 30 €/kWh, with copper and nickel covering almost 70 % of
the cost. In the case of NMC-811, lithium and cobalt each represent about 10 % of the
total material costs (see assumptions in Annex 2) (14).

(14) Based on historically high prices of raw materials of the last 8 years, NMC-111 material costs are estimated
at about 80 €/kWh and NMC-811 are estimated at 50 €/kWh.

16
 Figure 11 Element requirement for Li-ion battery cathodes

NMC-811

NMC-622

NMC-111

NCA

LCO

0 0.5 1 1.5
kg/kWh

Lithium Cobalt Nickel Manganese Aluminum

Source: JRC based on Olivetti et al. [77].

Figure 12 Future NMC chemistry mix in cathodes

3%
12% 17%
19%
23%
36%
35%
88%

56%
37%
43%

9% 10% 12%
Avicenne Energy BNEF Avicenne Energy BNEF
2020 2030

NMC-111 NMC-532 NMC-622 NMC-811

Source: JRC based on different experts and analysts [26,81].

2.2.2 Stationary battery system storage costs

Li-ion battery system costs for stationary storage have been witnessing a downward
trend, from 1 800 – 1 900 €/kWh in 2010 to 1 100 – 1 700 €/kWh in 2015 [57,65]. In
2017, the reported figures average at much lower costs at around 570 €/kWh, due to the
dive of battery pack prices and balance of system costs (BOS) [82]. Factors such as
boundaries (e.g. grid connection; see Figure 13), application (e.g. utility, behind-the-
meter), type of service (e.g. peak replacement, frequency control), size (e.g. kW, MW),
chemistry (e.g. NMC, LFP) characterise the technology and its costs. These vary across
the different studies and are not always reported, which could explain the wide spread,
shown in Figure 14. For example, IRENA mentions that already in 2016, under optimal
conditions, energy installation costs of Li-ion batteries for stationary storage may be as
low as 220 €/kWh, comparable with EV battery packs (section 2.2.1), possibly because
they include only electrochemical and power electronic components in the technology
boundaries. This estimate is almost a factor 10 lower than the approximately
1 400 €/kWh required for a behind-the-meter system based on Schmidt et al. [57]. As

17
such, a direct comparison of Li-ion battery stationary storage system based on multiple
sources on system costs or more detailed cost structures is not always feasible.
Costs of other battery types suitable for stationary storage systems range between 100
and 400 €/kWh for lead acid technologies at utility scale [4,83] and 1 250 €/kWh at
residential scale [57], between 220 and 640 €/kWh for sodium-sulphur batteries [4,83],
and from 450 to 1 450 €/kWh for zinc-bromide flow batteries [83].

Figure 13 Illustrative system cost and price structure of stationary battery storage

Source: JRC based on Lazard and BNEF [84,85].

18
 Figure 14 Li-ion battery stationary system costs in 2016 and 2017

1 600

1 400
Stationary battery storage

1 200

1 000
[€/kWh]
costs

800

600 Behind-the-meter
400 Utility
200 Reported average

0
2016 2017

Source: JRC based on Schmidt et al. [57], BNEF [85], IRENA [83], IEA [63], Deutsche Bank [65], Lazard [84]
McKinsey [86] and Navigant [87].

Another aspect to consider is that Li-ion batteries may be used to cover a range of
services in the power sector. The battery configuration (power output and energy
capacity ratio) depends on the system requirements and the desired service. Specific
costs expressed per unit of power range from being similar to a factor 8 higher when
compared with specific costs expressed per unit of energy (Figure 15). Moreover, the
spread of reported costs widens when expressed per unit of power compared with costs
expressed per unit of energy (Figure 15).

Figure 15 Specific costs of Li-ion battery for stationary system storage expressed per kW and kWh

6 000
2015
Stationary battery storage

5 000 2016
2017
4 000
costs

3 000

2 000

1 000

0
€/kWh €/kW

Source: JRC based on Schmidt et al. [57], BNEF [85] and Lazard [84].

Batteries for stationary storage are used for a range of applications with some being
more suited to store energy and others to supply power. It is observed that with
increasing power-to-energy ratio the specific costs tend to decline when expressed per
unit of energy and tend to increase when expressed per unit of power (Figure 16).
Generally, it is preferable to express specific costs of battery systems tailored to deliver
energy for a long period of time (energy-designed system) per kWh and for a system
tailored to deliver more power but for a less amount of time (power-designed system)

19
per kW [85]. In the present report, batteries that can provide energy for more than
1 hour are called energy-designed and batteries that can provide energy for less than
1 hour are called power-designed. Smaller scale residential batteries provide energy for
more than 1 hour and do not require system integration components.
Figure 16 Specific system costs of Li-ion battery for stationary storage expressed per kWh (upper
figure) and kW (lower figure) and power-to-energy ratio

2000
Lazard
1600 Schmidt et al. 2017
BNEF
€/kWh

1200

800

400

0
0 1 2 3 4 5 6 7 8 9
Power-to-Energy ratio

6000
Lazard
5000
Schmidt et al. 2017
4000
BNEF
€/kW

3000

2000

1000

0
0 1 2 3 4 5 6 7 8 9
Power-to-Energy ratio

Source: JRC based on Schmidt et al. [57], BNEF [88] and Lazard [84].

The distribution of power-to-energy ratios of projects with similar size is comparable

across different databases (Figure 17). Utility-scale projects tend to be more energy-
designed (power-to-energy ratio lower than 1), whilst commercial- or residential-scale
projects tend to be power-designed (power-to-energy ratio > 1). The spread of values in
Figure 17 indicates that the design of the system is project specific.
The cost structure of battery storage systems in 2017 is shown in Figure 18. Regardless
of the system service, the battery pack is a key cost component that represents 42 to
50 % of the total cost (15). The Power Conversion System (PCS), comprised mainly by
the inverter, is a major cost component in power-designed storage systems, while in
energy-designed storage systems the Energy Management System (EMS) becomes more
important (a description of the cost components of stationary battery storage systems is
presented in Figure 13).

(15) The cost of battery packs for EV and stationary storage applications are similar. However, different
requirements between applications (e.g. technical, long term functionality and external constraints [119])
could partially explain differences at a pack or at a systems level.

20
 Figure 17 Distribution of power-to-energy ratio based on project size

Utility scale (≥ 1 MWh)

50 %
BNEF
40 % US DOE

30 %

20 %

10 %

0%
C-rate ≥ 2 1 ≥ C-rate < 2 0.5 ≥ C-rate < 1 0.25 ≥ C-rate < 0.1 ≥ C-rate < C-rate < 0.1
0.5 0.25

Commercial or residential scale (≤ 100 kWh)

60 %
BNEF
50 %
US DOE
40 %

30 %

20 %

10 %

0%
C-rate ≥ 2 1 ≥ C-rate < 2 0.5 ≥ C-rate < 1 0.25 ≥ C-rate < 0.1 ≥ C-rate < C-rate < 0.1
0.5 0.25

Source: JRC based on BNEF [88] (214 projects > 1 MWh, 19 projects < 100 kWh) and US DOE [56]
(136 projects > 1 MWh, 119 projects < 100 kWh). Residential applications are not included in the datasets.
Note: Lazard's industry survey data [84], indicate for utility-scale projects (4 to 400 MWh) a ratio between 0.15
and 0.25 (energy-designed systems). For commercial or residential applications (10 to 250 kWh), Lazard's
reported ratio is 0.5 (energy-designed). Note: The discharge rate of a battery is expressed by its C-rate. The
capacity of a battery rated at 1C means that a fully charged battery will be completely discharged in 1 hour. 2C
rate means that the battery can be fully discharged in half an hour. ½C rate means that the battery can be fully
discharged in 2 hours.

According to most studies, the division of Li-ion battery storage systems into components
(Figure 19) shows that materials contribute 65 to 80 % (similar with the contribution of
materials in EV battery packs; section 2.2.1) and the remainder is mainly labour,
overhead, margins and other non-material costs. IRENA estimates the cost contribution
of materials to be less, at 30 % of which two-thirds are cell costs, and attributes the
remainder to other system costs [83]. The disparity in the cost-structure between the
sources is possibly due to different system boundaries (IRENA assesses stationary
system storage, while the other sources assess battery packs).

21
Figure 18 Cost breakdown of power-designed (C-rate 2) and energy-designed (C-rate 0.25) grid-
scale stationary storage system

820 €/kWh or 410 €/kW 500 €/kWh or 2 000 €/kW

100%

75%

50%

25%

0%
Power-designed Energy-designed
Battery pack BOS PCS EMS EPC Grid connection

Source: JRC based on average costs in BNEF's survey [85]. Note: Balance of System (BOS), Power Conversion
System (PCS), Energy Management System (EMS), Engineering, Procurement and Construction (EPC).

Figure 19 Cost breakdown of Li-ion battery storage system between cell components and other
costs

100 %
Other costs
80 %
Rest materials

60 % Separator
Electrolyte
40 %
Cathode
20 % Anode

0%
IRENA 2017 Pillot 2015 Sakti et al. Roland Berger Lowe et al.
2015 2012 2010

Source: IRENA and sources therein [83]. Note: sources in IRENA [83] represent cost breakdown of Li-ion
battery packs.

2.2.3 Cost reduction drivers

The fast growing pace of EV sales over the last years and the slower growth of other Li-
ion battery markets (Figure 1) have led to volume-driven cost reduction, owing to
technical progress (e.g. battery chemistries), improvements in manufacturing and
economies of scale. The growing research and innovation activity as measured by patent
statistics (Figure 20) may have also been contributing at an accelerating rate. Kittner et
al. [89] estimate a reduction in the order of 2 % per 100 patents (Patent Cooperation
Treaty; PCT), while the rest may be attributed to increasing production scale. Beyond
technical factors, the U.S. Department of Energy identifies three reasons that can explain
the price drop, based on market conditions and business strategy [68]. These are:
production overcapacity that may lead to supply-demand imbalances, competitive long-
term supply contract offerings that are based on anticipated cost reduction and finally,
strategic corporate behaviour.

22
 Figure 20 Annual patent activity on batteries filed globally and by the EU

1 600
Global
Number of patent families
on batteries annually

1 200 EU

800

400

0
2008 2009 2010 2011 2012 2013 2014

Source: JRC method [90] based on PatStat data (Autumn 2017 edition). Note: Global numbers include the EU.

The expected market surge, primarily of EVs [91], entails growth in production and
manufacturing scales, in view of the announced megafactory and gigafactory capacities
around the world (section 3.1). This may act as a driver for further cost reduction due to
economies in production scale. Accumulated experience may also lead to improved and
optimised manufacturing processes, bringing the costs of Li-ion further down. Moreover,
by means of vertical integration of production steps across the value chain (e.g. cell and
pack production), transportation expenses and turnaround time could decrease leading to
lower costs [43].
In addition, whilst several chemistries of Li-ion batteries already exist in the market (e.g.
NMC, NCA, LFP, LCO [92]), there are continuous efforts on improving Li-ion battery
cathode chemistry and material composition aiming to deliver better performance (e.g.
higher energy density), lower specific costs and removing other bottlenecks such as the
dependence on cobalt. It is not known which the dominating cathode chemistry will be or
when it will emerge at commercial scale as different expectations are expressed by
analysts (Figure 12).
Ongoing research on further improving Li-ion battery chemistries (e.g. high-voltage
electrolytes, durable lithium manganese oxide cells), combining conversion cathodes with
silicon-containing anodes, or moving beyond Li-ion (e.g. lithium-metal, solid state,
lithium-sulphur, lithium-air) signals to greater cost reduction potential [36,78,83].
Forecasted battery technology evolution, originally presented by the German National
Platform for Electromobility [93] and adopted in the preceding JRC publications on key
issues related to Li-ion batteries [36,43], gives an indication of the timeline for the
commercialisation of future battery technologies for mobility applications. The situation is
less clear with technologies for energy storage because of the large diversity of services.
Finally, the multiple applications in which Li-ion batteries are used, may offer additional
synergies for learning. Stationary storage system costs may benefit from large
photovoltaic inverter manufacturers entering the market thus reducing PCS costs. Scale
effects based on project size could affect other components such as BOS and
Engineering, Procurement and Construction (EPC), as well as standardisation of system
design, design improvements, engineering and competitive markets [82,86].

23
3 Future growth and costs of Li-ion batteries

3.1 Market growth of Li-ion batteries

3.1.1 Near-term manufacturing capacity growth globally and in the EU

Global manufacturing capacity of Li-ion cells for EVs and stationary storage is around
150 GWh, with two-thirds of the capacity being located in China [94]. In recent years,
production was characterised by low utilisation rates of about 50 % [66] and
overcapacity. Based on company announcements, substantial growth is expected in the
near future, ranging from additional 240 to 450 GWh [59,94,95]. That is, by 2022, global
Li-ion cell production capacity will be 2.5 to 4 times higher than today. With both EV
sales and the global manufacturing capacity for Li-ion battery cells rapidly increasing, the
market situation is expected to come into better balance in the near future [36].
Noticeable is that while the future market share of Li-ion cell manufacturing belongs to
Asian players (Figure 21), steepest growth is expected in Europe, owing to the limited
domestic capacity today. By 2022, the global share of European Li-ion battery cell
manufacturing capacity is expected to increase from about 3 % today to 8 % (Figure 21).
By 2028, due to additional capacity and plant expansions in Europe total Li-ion cell
manufacturing capacity may reach about 105 GWh, if all current plans and
announcements materialise. In the coming decade, depending on the year that the new
production lines become operational and near-term market projections, Europe may
serve between 7 and 25 % of global demand (Table 2, Figure 24). Most of the capacity
will be located in Sweden, Germany and Poland (Table 2, Figure 22).

Figure 21 Expected near-term growth in global Li-ion cell manufacturing capacity for applications
such as EVs and stationary storage

700
600
500
+86%
GWh

400
Rest of the world
300 +39%
Asia Pacific
200
100 Europe
0 North America
Today Under By 2021 By 2022 China
construction
BNEF (world) & other sources (Europe) IEA (world) &
other sources
(Europe)

Source: JRC based on BNEF [94] (world) and other sources for Europe [36,95–97]. IEA data used for 2022
[59].

24
Table 2 Producers of Li-ion cells for mobility and stationary storage applications in the EU
(including the UK) today and in the near term [36,94–98] (1)

Manufacturing Company name Capacity, GWh Headquarters

Currently operating facilities
Hungary Samsung SDI Co Ltd 2 South Korea
UK Nissan 1.4-1.5 Japan (2)
France Bollore SA 0.5 France
Germany Leclanche GmbH 0.1 Switzerland
France SAFT 0.06 (3) France
Finland European Batteries Oy 0.03 Finland
Germany Custom Cells 0.02 Germany
Facilities under construction, expected to begin production by 2021
Poland LG Chem Ltd 4 (2019), up to 9-12 South Korea
Italy SERI (FAAM) 0.2 (2019) Italy
Announced facilities, expected to begin production after 2021
Sweden NorthVolt AB 8 (2020), 32 (2023) Sweden
Hungary SK Innovation 7.5 (2022) South Korea
Germany CATL 14 (2022) China
4
Germany TerraE ( ) 0.9 (initially), 34 (2028) Germany
1
( ) Between 2020 – 2028, other announcements include a 35 GWh plant by Tesla at undisclosed location [95].
In addition, according to Reuters, SAFT (FR) and partners plan to produce advanced Li-ion batteries in the EU
from 2020 [97], Varta Microbattery Systems (DE) and Ford plan to establish Li-ion battery cell production in
Germany [99] and BYD (CN) at undisclosed location in Europe [100]. (2) Nissan entered into a definitive
agreement with Envision Group (headquarters in China) for the sale of Nissan's battery operation and
production facilities. The deal, pending regulatory approvals, is expected to be completed by March 29, 2019
[101]. (3) Data on the exact manufacturing capacity of SAFT in France for cells for stationary energy storage
application is unavailable; hence full manufacturing capacity of SAFT is France is cited based on data from
Lebedeva et al. [36]. (4) The dissolution of the TerraE consortium has been announced in October 2018 [102].

Figure 22 Expected evolution of Li-ion cell manufacturing capacity for mobility and stationary
storage applications in the EU (incl. the UK), in GWh

Today By 2021 By 2023

Source: JRC based on various sources [36,94–98]. Note: Excluding the announcement made by TESLA as the
location was not disclosed.

25
Figure 23 summarises the development of specific annual investments of capacity added
and announced battery plant costs based on IEA [59]. While some difference is noticed
between announced capital costs and investments, the trends converge to about
100 million €/GWh in 2017, or 70 % lower compared with the beginning of the
decade (16).

Figure 23 Annual investments in new Li-ion cell manufacturing capacity and announced capital
costs per unit of battery manufacturing capacity

400
Annual investments
350
Announced capital costs
300
Million €/GWh

250
200
150
100
50
0
2011 2012 2013 2014 2015 2016 2017 2018 2019

Source: JRC based on IEA [59].

3.1.2 Global long-term manufacturing capacity growth

Globally and in the longer term (towards 2040), annual sales of Li-ion battery cells are
expected to grow exponentially (Figure 24). Their market value increases from 24 bn € in
2017 to about 65 bn € in 2025 [44]. Across all market segments, global production will
see an unprecedented growth due to EVs, as the main consumer of Li-ion batteries. EVs
are projected to demand 65 – 80 %, 80 – 95 % and 90 – 95% of total production of Li-
ion batteries by 2020, 2025 and 2030, respectively, compared with about 55 % today.

Figure 24 Future global sales of main Li-ion battery market segments

2 000
Li-ion battery sales

Electronics
1 600 Stationary storage
EV
[GWh]

1 200
800
400
0
BNEF

BNEF

BNEF

BNEF
Avicenne

Avicenne

IDTechEx

Avicenne

IDTechEx

IDTechEx
Energy

Energy

Energy

2015 2020 2025 2028

Source: JRC based on BNEF [26], Avicenne Energy [44] and IDTechEx [52]. Note: EV includes passenger light
duty vehicles and buses; Electronics includes consumer electronics for all studies, power tools and medical in
Avicenne Energy, and wearables in IDTechEx; Stationary storage includes industrial applications in Avicenne
Energy.

(16) The difference between annual investments and announcements was about 400 million €/GWh in 2010.

26
3.1.3 Global market growth of electric vehicles
Figure 25 shows a steep growth of EVs on the road after 2025. Based on a range of
global projections, by 2030, the lowest value is about 50 million EVs and the highest
estimate is 4 times higher (225 million EVs). The lowest projection reflects the potential
deployment of EVs under reference technology development assumptions, captured by
the IEA's Energy Technology Perspectives RTS scenario [103]. The highest projection,
reflects the ambition that 30 % of the global market share of all vehicles will be captured
by EVs [45].
By 2040, the volume increases from about 150 million EVs under reference technology
assumptions [103] up to 900 million EVs in ambitious scenarios with respect to
decarbonisation, improved energy access and air quality (captured by IEA's World Energy
Outlook SDS scenario [63]). The 4-fold and 6-fold difference across scenarios, in 2030
and 2040, respectively, clearly shows the varying perspectives on future EV growth. A
scenario review, conducted by the Center on Global Energy Policy (CGEP), concludes that
different forecasts have widely disparate views on key underlying drivers of oil demand
such as population and economic growth, which could also affect the total number of EVs
[104]. In addition, CGEP mentions that forecasts consider the adoption of EVs is
encouraged by government policies and technology change, which is reflected in battery
costs and is one of the most uncertain factors [104]. Even so, the analysis conducted in
the present report points out that even under the most pessimistic projections the size of
the EV fleet will multiply by 50 in 2040 compared with today. Moreover, the more
ambitious the scenarios in terms of climate targets, the higher the deployment of EVs
would need to be, as shown by the number of total EVs in decarbonisation scenarios, due
to higher electrification rates. Notably, similar growth is also seen by market scenarios
such as the one considered by BNEF [26,105], in which climate targets are not met.

Figure 25 Projections of the global EV fleet over time

1 000
XOM IHS
900 BP ET OPEC
800 IEA WEO SDS (2017) Wood Mackenzie
IEA WEO 450 (2016) Deutsche Bank
Global EV fleet

700
[Million EVs]

IEA WEO NP (2016) IEA WEO NP (2017)

600 IEA ETP B2DS (2017) IEA ETP 2DS (2017)
500 IEA ETP RTS (2017) Avicenne Energy
400 BNEF EVO IRENA
IEA GEVO NPS IEA GEVO EV30@30
300
200
100

2017 2020 2025 2030 2035 2040

Source: JRC based on different studies and scenarios [26,45,47,63,65,103,106–109]. Abbreviations of the
studies can be found in the references. Note: the fleet includes passenger light duty vehicles (BEV and PHEV)
and electric buses.

Another difference across scenarios lies in the composition of the EV fleet. Figure 26
shows that the more ambitious the scenario is in terms of decarbonisation (e.g. IEA ETP
B2DS), EV push in the market (e.g. IEA GEVO EV30@30), or favourable market
assumptions (e.g. BNEF), the higher the share of BEVs. As the battery capacity of BEVs is
higher than that of PHEVs, the associated global manufacturing capacity will be
influenced by the composition of the EV fleet. Heavier duty vehicles, and specifically
electric buses, could also have an effect on the demand for Li-ion batteries, as they
typically require much larger batteries. IEA forecasts the total number of buses to reach
1.5 and 4.5 million units by 2030 in their New Policies and EV30@30 scenarios,

27
respectively [45]. BNEF expects the global electric bus fleet to be somewhere between
these values, at 2.3 million buses in 2040 [26].

Figure 26 Share of BEV and PHEV in the global EV fleet according to different scenarios in 2030
and 2040

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
IEA ETP B2DS
IEA ETP 2DS
2040

IEA ETP RTS

BNEF
IEA GEVO EV30@30
IEA GEVO NPS
IEA ETP B2DS
2030

IEA ETP 2DS

IEA ETP RTS
BNEF

BEV PHEV

Source: JRC based on BNEF and IEA [26,45,103].

3.1.4 Global market growth of stationary storage

From a niche application today, Li-ion batteries for stationary storage are projected to
increase rapidly over time. In the near-term, most projections see an increase by up to
an order of magnitude, from about 3 – 4 GWh today to 100 GWh in 2025. Beyond 2025,
the steep growth continues: the lowest estimates range from 8 to 100 GWh and the
highest estimates reach 400 GWh in 2030, ultimately leading to 1 300 GWh in 2040
(Figure 27). Compared with EV outlooks, fewer studies report on stationary storage (see
e.g. Figure 25). Also most forecasts extend up to 2025 – 2030. Overall, the slowest
growth is projected by IEA [103], which is possibly due to the fact that their projections
do not include residential applications. Highest growth is foreseen by BNEF, which in its
latest New Energy Outlook, revised its projections upwards [105].

Figure 27 Projections of total stationary storage installed front- and behind-the-meter globally

1 400
IRENA Ref min IRENA Ref max
1 200 IRENA Doubling min IRENA Doubling max
BNEF GESF (2017) BNEF NEO (2017)
Stationary storage

1 000 BNEF NEO (2018) IEA ETP 2DS (2017)

IEA ETP B2DS (2017) Deutsche Bank
[GWh]

800 JRC low JRC high

600 IDTechEx

400

200

2017 2020 2025 2030 2035 2040

Source: JRC based on various literature sources [31,39,43,52,58,65,83,105]. Abbreviations of the studies can
be found in the references. Note: IEA [103] is not clear as to whether they include behind-the-meter
applications in their projections for storage.

28
3.2 Future costs of Li-ion batteries for EVs and stationary storage
The average value across cost forecasts of battery packs for EVs taken from literature
show a decline from about 250 €/kWh in 2020 to 110 €/kWh in 2040, a trend that is
more conservative than recently announced prices at about 200 €/kWh (Figure 28). The
range of estimates is rather wide, with values reported from above 400 €/kWh in the
short term to as low as 40 €/kWh in the long term. One explanation is that studies
published between 2010 and 2015 anticipated the costs to decline but possibly owing to
the unforeseen rapid drop of prices after 2015 (see Figure 9) their estimates turn out to
be conservative. Another aspect is that some studies refrain from offering a single cost
trajectory but provide a set of values based on different assumptions (e.g. slow,
moderate and rapid advancement considered in the study of NREL [110], or low, high
and the global average value from the IEA World Energy Outlook [63]). As a result, a
wide range of values is available in literature when it comes to long-term battery pack
costs for EVs.

Figure 28 Cost forecasts of Li-ion battery packs for EVs

500
Studies conducted after 2015
400
EV battery pack costs

Studies conducted in 2010-2015

300
[€/kWh]

200

100

0
2020 2025 2030 2035 2040

Source: JRC based on literature sources (NREL [110], IEA [63,103], Schmidt et al. [57], Avicenne Energy [44],
Deutsche Bank [65], BNEF [26], and several sources in Nykvist and Nilsson [60] and Berckmans et al. [111]).
Note: the figure includes the cost target based on SET Plan for 2030 (filled symbol) [23].

Furthermore, there is no consensus as to when batteries may reach price which will bring
EVs at parity with internal combustion engine vehicles (e.g. at a set price of 90 €/kWh),
as forecasts range from 2020 to 2033 (Figure 29).

Figure 29 Year when EVs are at price parity with internal combustion engine vehicles based on
different scenarios

Source: JRC adapted from CGEP [104]. EV battery pack at 90 €/kWh. Note: Each symbol indicates a different
study. Differentiation between shape and colour is used to characterise the type of the organisation that
conducted each study.

29
Box 1 Approaches used to estimate costs of Li-ion batteries

1. Detailed engineering studies

Bottom-up calculations based on each cost component required to produce specific
battery types according to manufacturing process flows. This approach takes into
account fixed and variable expenses such as material and energy costs, direct
labour, plant investment costs, research and development, general sales and
administration, overhead, which are parameterised based on the manufacturing
scale and the plant's annual output. The estimated costs are representative for a
specific battery (e.g. power, energy, type, size, hence also performance) and
chemistry (e.g. NMC-111, NMC-811, NCA). The production costs are related with
specific assumptions on raw material and energy prices. In this method, costs and
performance parameters are correlated. Future costs can be estimated with
detailed engineering studies when potential improvements, for example in
production processes or material efficiency, are considered. The ANL model [112],
Berckmans et al. [111] and today's costs of Li-ion batteries from BNEF [26] are
few studies that apply this approach. Other studies simplify their estimation by
adding components at a higher aggregation level (e.g. Abdon et al. [113], IRENA
[114]).
2. Learning curves
Applying the empirical observation that production costs of a technology decline at
a specific rate (learning rate) every time the manufactured capacity doubles, this
method estimates top-down the evolution of battery costs over time (e.g.
[57,60,89,111]). One challenge is to establish a representative learning rate, as it
requires robust historical timeseries on production capacity and costs. The three
most prominent recent examples that apply the learning curve method are:
̶ Nykvist and Nilsson [60], who conducted a systematic review of historical Li-ion
EV battery pack costs to determine learning rates.
̶ Schmidt et al. [57], in which cost trajectories of Li-ion EV battery packs,
residential and utility-scale storage systems were assessed, in line with
deployment that was derived from energy storage diffusion curves.
̶ Kittner et al. [89] who analysed the deployment and innovation of batteries
using a two-factor learning curve model.
3. Surveys, expectations and announcements
Several consultancies and market research organisations (e.g. Avicenne Energy,
Bloomberg New Energy Finance, IDTechEx, Boston Consulting Group) initiate
surveys and collect information directly from the producers on the evolution of
battery costs. In addition, several companies (e.g. car manufacturers), in support
of their strategy, proceed with own announcements on how they expect costs to
evolve. These are often near-term forecasts without technical details as they are
typically proprietary to the manufacturer. These costs are often quoted in
literature, yet, in the absence of technical details and more evidence, they could be
highly uncertain or speculative.
4. Levelised cost of storage
Whilst a different metric than capital investment costs and primarily relevant for
stationary storage applications, levelised costs of storage are reported by several
organisations, taking into account the service provided by the battery (e.g. IRENA
[114], Lazard [84]). Results based on this metric are usually provided on a kWh
basis, but are not comparable with the capital investment cost metric that is
assessed in this report. Unlike levelised costs of energy from conventional
technologies, the metrics on storage are not yet established and results vary as
new metrics are also being proposed [115].

30
Cost forecasts of Li-ion battery systems for stationary storage systems show a similar
declining trend and wide ranges. Battery packs make up large of part of the total cost.
The range is somewhat wider for stationary storage, firstly, because additional
components are included (e.g. inverters, BOS) and secondly, because diverse
applications call for different battery and system design (e.g. power shaving is a power-
designed system compared to self-consumption which is energy-designed). As such, on a
kWh basis, power-designed systems (Utility in Figure 30) are shown to cost more than
energy-designed systems (Res / Com, in Figure 30). In addition, studies provide different
cost estimates either based on different chemistry (e.g. IRENA [114]), or based on
different modelling assumptions (e.g. Schmidt et al. [57]). Another observation is that
some studies tend to be more optimistic (e.g. IRENA [114] in Figure 30) than others
(e.g. Schmidt et al. [57] in Figure 30). Besides the different method used to estimate
costs (Box 1), different system boundaries could also be an explanation.

Figure 30 Cost forecasts of Li-ion battery stationary system storage based on literature

1 600
IRENA Schmidt et al. 2017
Stationary battery storage

BNEF Deutsche Bank

1 200 McKinsey Navigant
[€/kWh]

800
costs

400

Res / Com
Res / Com

Utility

Res / Com

Utility

Res / Com

Utility

Res / Com

Utility

Utility
2020 2025 2030 2035 2040

Source: JRC based on IRENA [114], Schmidt et al. [57], BNEF [85], Deutsche Bank [65], McKinsey [86] and
Navigant [87]. Note: IRENA estimates are based on ranges for all Li-ion chemistries and selected applications
(self-consumption as residential and peak-shaving as utility). The size of the inverter is selected accordingly
(small-scale for residential and large-scale for utility).

31
Box 2 Techno-economic performance of EV batteries

Between 2011 and 2017, the techno-economic performance of Li-ion batteries has seen
significant improvements (Figure 31). On average, costs of packs have decreased by
65 %. While the capacity of the batteries in EVs more than doubled, the total cost
dropped by 10 %. Larger battery size also entails longer ranges, which have more than
doubled since 2011.

Figure 31 Historical and near-term Li-ion battery performance improvement of BEVs in western
markets (i.e. excluding China)

Source: Historical estimates based on weighted average sales from BNEF [38] and technical specifications
from WattEV2Buy [53], High Edge [54] and ITRI [55], future estimates based on technical performance in
IRENA [108] and BNEF [26].

Historically, this step change in weighted average costs and performance can be largely
attributed to the production of Tesla vehicles, which, compared with other BEVs, have
larger battery size and range. Since 2013, Tesla cars represent from about one-fifth to
one-quarter of all new BEV sales. Anticipated improvements of new batteries of the
remainder of the fleet show that the historical improvements will continue in the near-
term.

32
4 Scenario-based cost trajectories of Li-ion batteries

4.1 Method

4.1.1 Learning rates

The costs of energy technologies have been correlated with their cumulative installed
capacity as an indication of the experience gained through increased global
manufacturing of the technology and its components [40,41]. Researchers and analysts
extend this historical correlation to assess top-down how costs of energy technologies
may develop in the future. This method is known as learning or experience curve. It
combines the historical rate of cost reduction achieved for every doubling of installed
capacity of a technology (learning rate) with projections on its deployment over a period
of time (for a description of the method see, for example, Tsiropoulos et al. [116]). A
technology learning rate is derived from historical data on cost development over time
and on manufacturing output or installed capacity. Figure 32 shows different learning
rates that have been reported for Li-ion batteries. The period of analysis, the technology
boundaries, the metrics used (e.g. cost or price, annual or cumulative production) offer
possible explanations as to why the values range. For Li-ion batteries, Schmidt et al. [57]
note that learning rates tend to decrease with increasing technology scope.
Learning rates of inverters, a key component of stationary storage systems (section
2.2.2), are reported at 19 % (± 1 %) [8,9,57].

Figure 32 Learning rates of Li-ion batteries for different applications

40%
Learning rate

30%

20%

10%
Electronics EVs Stationary storage
0%
BNEF EVO 2017

Sandalow 2015

Kittner et al. 2017

Schmidt et al. 2017

Gondelach et al. 2012

Gondelach et al. 2012

Schmidt et al. 2017

Schmidt et al. 2017

Schmidt et al. 2017

Nykvist and Nilsson 2015

Nykvist and Nilsson 2015

Nykvist and Nilsson 2015

Nykvist and Nilsson 2015

(residential)
(market leaders)

(utility)
(all industry)

(historical)

(future)

Source: JRC based on literature [57,60,67,89,117,118].

4.1.2 Selected global growth scenarios

Available global growth scenarios of EVs and stationary storage are shown in Figure 25
and Figure 27, respectively. Long-term projections on future deployment of a technology
are inherently uncertain, as they are based on different approaches (e.g. market
penetration models, energy system models) and critical assumptions therein (e.g. CO2
emission reduction target, energy efficiency goals, technology costs). Furthermore, the
deployment of Li-ion batteries for mobility or stationary storage applications does not
occur in isolation from the rest of the energy system. It depends on a number of factors
such as the synergies with residential photovoltaics that may boost the deployment of
behind-the-meter applications or the competition with other technologies such as fuel

33
cells in the transport sector. In the long-term, dynamics from synergies and competition
may prove important. Using deployment projections of EVs and stationary battery
storage from distinctly different scenarios provides a range of growth trajectories, which
has the advantage of taking system dynamics into account and is consistent across all
technologies within each scenario.
To address different possible long-term futures of Li-ion batteries, three global
deployment scenarios are selected from literature, which outline growth trajectories of
EVs (Figure 25) and stationary storage (Figure 27). The selected scenarios cover a wide
range of projections and they are adequately different in terms of technology portfolio
and deployment levels of all technologies (Figure 33). The deployment projections for
EVs and stationary storage in the high scenario (Box 3) are based on the 2018 Electric
Vehicle Outlook and the 2018 New Energy Outlook of Bloomberg New Energy Finance
(BNEF NEO 2018) [94,105]. Projections in the middle of the range are covered by the
moderate scenario (Box 4), which is based on the deployment figures of the International
Energy Agency's 2DS scenario of the 2017 Energy Technology Perspectives (IEA ETP 2DS
2017) [103]. Finally, the lower end of the range is covered by the low scenario (Box 5),
which relies on deployment on the International Energy Agency's Reference Technology
Scenario of the 2017 Energy Technology Perspectives (IEA ETP RTS 2017) [103].

Figure 33 Selected global growth scenarios for EVs (upper figure) and stationary storage (lower
figure)

40
high
35
moderate
30 low
25
EV fleet
[TWh]

20
15
10
5
0
2017 2020 2025 2030 2035 2040

1.4
high
1.2 moderate
Stationary storage

1.0 low

0.8
[TWh]

0.6

0.4

0.2

0.0
2017 2020 2025 2030 2035 2040

Note: Global EV fleet battery capacity estimated based on million EVs on the road (Figure 25) and battery
capacity of 40 kWh/EV.

34
 Box 3 Description of the high scenario (BNEF NEO 2018)

Among the selected scenarios, the high scenario sees the highest deployment of EVs
and stationary storage. According to this scenario, market forces drive decisions based
on technology costs. EVs gain a significant market share as they gradually come at
cost parity with internal combustion engines (EVs are 33 % of the total fleet in 2040).
Renewables become increasingly competitive and wind and solar supply almost all new
electricity demand globally. Growth in intermittent renewables goes hand in hand with
growth in stationary battery storage, which allows more than 1/3 of total demand to be
supplied by wind and solar in 2040. The growth in renewables and decoupling of
emissions in passenger road transport allows for significant changes in emission
profiles of large economies, however, global CO2 emission reduction goals in line with
the Paris Agreement are not met. The deployment projections for EVs and stationary
storage are based on the 2018 Electric Vehicle Outlook and the 2018 New Energy
Outlook of Bloomberg New Energy Finance [94,105].

Box 4 Description of the moderate scenario (IEA ETP 2 DS 2017)

The world moves towards decarbonising its energy system up to a 2 oC average global
temperature increase, with 50 % likelihood, by the end of this century. This
challenging transformation reduces emissions from energy use and other sectors. EVs
and energy storage participate in a broad technology portfolio, which among other
options includes, renewable electricity generation, biofuels in transport, hydrogen,
carbon capture and storage (CCS) and demand-side measures (e.g. energy efficiency).
Intermittent renewables provide somewhat less than 1/3 of total electricity demand,
while baseload and flexible generation from solid fossil fuels, gas and biomass still have
a strong presence in the energy mix in 2040, as they can be combined with CCS.
Growth of EVs (23 % of the global fleet in 2040) and system storage is based on the
International Energy Agency's 2DS scenario of the 2017 Energy Technology
Perspectives [103].

Box 5 Description of the low scenario (IEA ETP RTS 2017)

The world moves from earlier "business-as-usual" paradigms to more ambitious

pathways with respect to emission reduction, yet no major transformation takes place
before mid-century. Nationally Determined Contributions on CO2 emission reduction,
pledged by the parties under the Paris Agreement are met, but they fall short in
reaching the level of the internationally agreed climate change mitigation targets. In
this context, electrification of passenger road transport is limited (EVs are 7 % of the
total fleet in 2040), penetration of intermittent renewables increases only up to 15 %
and fossil fuels still dominate the global electricity mix. This scenario is used as a
baseline against which the impacts of growth of EVs and stationary storage can be
measured. It is based on the International Energy Agency's Reference Technology
Scenario of the 2017 Energy Technology Perspectives [103].

4.1.3 Implicit assumptions

Combining learning rates with global deployment projections of EVs and stationary
storage, implicitly assumes that Li-ion batteries are the winning technology that take up
the entire share in these markets, across all scenarios (high, moderate, low). Even if lead
acid batteries dominate the broader electrochemical storage market today (including
energy storage, telecommunications, industrial and other applications), Li-ion is expected

35
to be the leading technology by 2025 [119]. Li-ion technologies represent 81 % of
electrochemical energy system storage (in terms of power capacity; section 2.1.2). In
the longer term, Li-ion batteries are suitable for most services in the energy system;
other battery storage technologies will ultimately compete on costs.
Besides applications in mobility, more explicit assumptions on stationary storage are
that: a) Li-ion stationary storage benefits from the learning, hence the cost reduction,
achieved in battery packs driven by the market growth of EVs (spill-over learning), b)
costs of PCS components reduce based on learning in production of inverters, in line with
the growth of photovoltaics in each of the selected scenarios, c) costs of the remaining
components (e.g. BOS, EPC, EMS) reduce based on growth and learning rate of
stationary storage applications only. As benchmark power-designed, energy-designed
systems and residential batteries the cost structures of BNEF are assumed [85,105] (see
also Figure 18 and Annex 3).
The lifetime of Li-ion EV batteries is assumed to be 10 years, which is reasonable taking
into account that many manufacturers provide warranty for 8 years and that significant
effort is made by both cell manufacturers and automotive producers to make robust and
long-lasting batteries. The lifetime of stationary storage is assumed to be 20 years. The
majority of Li-ion storage projects in US DOE's database have a lifetime between 15 and
20 years [56].
The influence of several of these assumptions on the cost trajectories is assessed in
detail in section 4.3.
All data and parameters used to estimate cost trajectories of Li-ion batteries for mobility
and stationary storage applications can be found in Annex 3.

4.2 Cost trajectories

The cost trajectories of Li-ion batteries estimated with the learning curve method are
presented as a function of cumulative deployment (Figure 34) and time (Figure 35,
Figure 37-Figure 38, Figure 40).

Figure 34 Learning curves for Li-ion batteries for EVs and stationary storage systems

Note: cost reduction of stationary storage is estimated based on total deployment of Li-ion cells (i.e. for
stationary storage and EVs), while the figure presents the deployment range associated with stationary storage
only.

36
The main remarks on cost trajectories of Li-ion batteries for EVs and stationary
storage as a function of cumulative deployment (Figure 34) are the following:
 The costs of power-designed and energy-designed stationary storage systems
converge with increasing installed capacity, due to the changing cost structure
over time (lower contribution of battery packs thus increasing contribution of
other components; see also Figure 39).
 The learning rate of EV packs is 16 % and is based on literature [57]. The
combined learning rate of energy-designed stationary storage systems ranges
between 14 and 16 %, while for power-designed systems it ranges between 14
and 15 %.
 The effect of spill-over learning is shown by the steeper decline of stationary
storage costs compared with EV costs, albeit the lower learning rate and range of
cumulative deployment of the former.
 The SET Plan target for stationary storage system costs at 150 €/kWh [23] could
be attainable at high global deployment of EVs and more than 1 TWh of stationary
storage deployed globally. It could be beneficial to define separate targets for
power-designed systems, as a significant cost difference is noticed when
compared with energy-designed systems.
As regards the cost trajectories of Li-ion battery packs for EVs as a function of time
(Figure 35) the main remarks are the following:
 Costs are reduced by more than 50 % by 2030 and by 63 to 75 % by 2040,
compared with today, depending on the storyline. That is a 4 to 6 % reduction in
cost when the fleet size increases between 16 and 25 % on an annual basis.
 Early in the time horizon, costs decrease faster in the moderate scenario, possibly
due to the push for EVs in order to achieve greenhouse gas emission mitigation
targets. After 2030, EVs penetrate much faster in the high scenario, as they
become competitive in more regions, and costs decline faster than in the
moderate scenario. Overall, the cost trajectories in these scenarios are similar.
 Costs decrease relatively quickly, yet more conservatively in the low scenario.
This indicates that EV penetration is the main cost reduction driver. In this
scenario costs remain above 100 €/kWh until 2030. This may impede EV
deployment at large scales and delay any efforts for transport decarbonisation.
 The costs estimates fall consistently within the lower-end of literature results.
 The SET Plan target for Li-ion EV battery pack cost (75 €/kWh by 2030 [23]) is
feasible in both high and moderate scenarios. This entails fast ramp-up of Li-ion
manufacturing capacity, of about 2 gigafactories globally per year until 2030
(Figure 36) (17). Should unfavourable conditions prevail (e.g. low scenario),
meeting the SET Plan cost target seems more likely around 2040.
 Based on these cost trajectories and annual sales, the global Li-ion battery market
for EVs could reach 40 – 55 bn €/yr in 2025. For European production this could
entail a growth from about 450 M€/yr in 2017 to 3 – 14 bn €/yr in 2025. Globally,
the market size may exceed 200 bn €/yr by 2040.

(17) Assuming 35 GWh/yr per Gigafactory. After 2030 the gigafactory additions increase significantly up to 8
gigafactories per year to meet the demand in 2040 in line with the high scenario.

37
Figure 35 Cost-development of Li-ion battery packs for EVs over time based on three different
deployment scenarios

500
high
Li-ion EV battery pack cost

moderate
400
low
[€/kWh]

300

200

100

0
2017 2020 2025 2030 2035 2040

Note: literature estimates are represented by the boxplots and include values presented in section 3.2.

Figure 36 Global ramp-up of manufacturing capacity of Li-ion cells and annual gigafactory
additions in each scenario to 2030 (upper figure) and from 2030 to 2040 (lower figure)

1 400
Annual
+ 2.4
1 200
Global Li-ion battery

gigafactory
1 000 additions:
+ 1.8
[GWh/yr]
demand

800
600
400 + 0.5
200
0
low moderate high
2017 2030

4 000 +8
Annual
3 500
Global Li-ion battery

gigafactory
3 000 additions:
[GWh/yr]

2 500
demand

2 000 +3

1 500
1 000
+1
500
0
2030 2040 2030 2040 2030 2040
low moderate high

Note: assuming 35 GWh per gigafactory.

38
On cost trajectories of Li-ion stationary storage systems as a function of time (Figure
37 and Figure 38) the main findings are:
 By 2030, costs are lower between 30 and 55 % compared with today. By 2040
they are lower by up to 66 % compared with today, depending on the storyline.
That is an annual cost decline of 3 to 5 % when installed capacity increases 13 to
24 % per year.
 In the short-term (2020-2025) there is no substantial difference between the high
and the moderate scenario. Thereafter the decline in the high scenario outpaces
that of moderate.
 The difference in costs between scenarios reaches 35 – 40 €/kWh for energy-
designed and 55 – 70 €/kWh for power-designed systems in 2040. Compared with
system costs of stationary storage today this may not seem significant, but it
becomes substantial considering that it could represent 15 to 25 % of total
system costs in 2040.
 Costs below 200 €/kWh can be reached for energy-designed systems after 2030
under the scenario with the highest deployment of batteries for EVs and storage
due to spill-over effects (high scenario). For comparison, the cost target of SET
Plan is at 150 €/kWh at a system level for 100 kW [23].
 For energy-designed systems, the costs fall within the lower range of values found
in literature. After 2035, the cost estimates of this report are lower than other
literature estimates. For power-designed systems, estimates fall within the range
of costs in reported in literature. There is limited publicly available data beyond
2030 to draw more detailed conclusions.
 Over time, all major cost components of stationary system storage decline (Figure
39), with major reduction noticed in battery packs (around 75 % by 2040) and by
a similar rate in PCS and other components (about 60 % by 2040). These
downward trends change the cost structure of stationary storage systems over
time and, by 2040, BOS, EPC and other soft costs become the major cost
component.
 Li-ion battery pack costs are set to decrease based on deployment of EVs, and
PCS costs also steeply drop based on growth of inverters of photovoltaics. The
next frontier to reduce costs of stationary storage further are BOS hardware, EPC
costs, and soft cost components (e.g. land acquisition).

Figure 37 Cost-development of Li-ion battery for utility scale energy-designed stationary storage
systems (C-rate 0.25) over time based on three different deployment scenarios

1 500
Li-ion stationary storage cost

high
Energy-designed [€/kWh]

1 250 moderate
low
1 000

750

500

250

0
2017 2020 2025 2030 2035 2040

Note: literature estimates are represented by the boxplots and include values presented in section 3.2.

39
Figure 38 Cost-development of Li-ion battery for utility scale power-designed stationary storage
systems (C-rate 2) over time based on three different deployment scenarios

1 500
high
Li-ion stationary storage cost
Power-designed [€/kWh]

1 250 moderate
low
1 000

750

500

250

0
2017 2020 2025 2030 2035 2040

Note: literature estimates are represented by the boxplots and include values presented in section 3.2.

Figure 39 Cost structure and cost structure development of utility scale stationary storage
systems over time in the high scenario

600
Other
stationary storage [€/kWh]

500
Cost of energy-designed

500 PCS
Battery pack
400 225

300
25 223
200 164
116 - 48%
250 91 - 22 %
100 14 - 44%
11 - 21 %
93 - 63 % 62 - 33 %
0
2017 2030 2040

1 000
Other
stationary storage [€/kWh]

820
Cost of power-designed

800 PCS
Battery pack
305
600

170 380
400
- 48% 282
157
200 - 43% 124 - 21 %
344 97
72 - 26 %
129 - 63 % 85 - 34 %
0
2017 2030 2040

40
 Figure 40 Cost development of Li-ion residential energy storage (C-rate 0.35)

800
high
700
Li-ion residential energy
storage system [€/kWh]
moderate
600 low
500
400
300
200
100
0
2017 2020 2025 2030 2035 2040

The reduction rate of residential storage is assumed to be the same with that of utility-
scale systems (18). Apart from the different absolute cost levels (determined largely by
the C-rate; see Note in Figure 18), the contribution of each major component also differs
in the three systems (energy-designed and power-designed utility-scale systems and
residential batteries). Ultimately, however, the cost reduction over time in relative terms
is found to be similar for the stationary storage systems and steeper for battery packs. In
absolute terms, the savings are more substantial for residential and power-designed
systems (on a kWh basis).

4.3 Sensitivity scenarios

Anticipated developments on Li-ion battery performance or on system characteristics
may influence the main assumptions that were used to estimate the reference cost
trajectories (section 4.2). To provide a holistic view, Box 6 presents six additional
scenarios that capture the sensitivity of the reference costs.

Results show that the assumed learning rate is by far and foremost the most influential
assumption across all systems assessed in this report, as it may lead to a cost difference
of up to 45 % for EV packs and up to 38 % for stationary storage systems, compared
with the reference costs. This highlights the need of using well-established learning rates
when applying the method on Li-ion batteries, and updating them as necessary, because
combined with high growth projections of EVs the ultimate costs may diverge
significantly. This observation is further supported by the scenario on learning only at cell
level, which is a major cost component of Li-ion batteries. The scenario assumes the high
historical learning rates observed for Li-ion cells for electronics (i.e. 30 % [57]) and
applies them on cells for EVs and stationary storage. It shows that, decomposing this
technology to a more granular level but using learning rates of similar yet not identical
applications, could drastically alter the findings (more than 30 % lower costs were found
in reference results in 2040).
Another key assumption of the method is that the experience gained due to growth in
manufacturing of Li-ion cells and packs for EVs spills over to stationary storage. This
assumption proves influential as costs of energy-designed systems may be higher by
about 20 to 30 % and of power-designed systems by about 18 to 25 %. Influential as

(18) Based on the methodology applied in the present report, each cost component (battery pack, inverter,
other) reduces at the same rate across the different stationary storage systems. However, the overall
reduction rate differs between the storage technologies due to the different contribution of these
components. The cost contribution and the cost levels are based on BNEF [85,105].

41
this assumption may be, it is unlikely that Li-ion battery manufacturing will be isolated to
their respective markets, as even today, they are produced by the same companies.

Box 6 Description of what-if and sensitivity scenarios

1. Learning rates
The learning rate is a parameter that directly affects the results as it is the rate used to
estimate the cost reduction over time. In this set of sensitivity scenarios, cost
trajectories are assessed under the same growth assumptions (section 4.1.2) with the
reference results but for high and low learning rates of Li-ion battery packs, inverters
and stationary storage systems (Table 13, Annex 3).
2. Effect of battery size in manufacturing capacities
The total annual manufacturing volume of Li-ion batteries is based on the assumption
of an average size of 60 kWh for BEVs and 20 kWh for PHEVs in 2040 (Table 12, Annex
3). As manufacturers are moving towards larger battery sizes, this assumption may
prove important and additional capacity may be required if battery sizes increase more
than what is assumed. This scenario assesses two widely different situations: a) the
battery size increases to 90 kWh for BEVs and 30 kWh for PHEVs (high scenario) and
b) the battery sizes remain as today (low scenario).
3. Battery lifetime
The lifetime of batteries is directly related to the rate at which EVs, residential or utility
scale batteries are replaced by new capacity. At a high lifetime scenario, fewer
batteries are replaced thus cumulative production is lower compared with a low lifetime
scenario where replacements occur more frequently and cumulative production will
need to be higher (Table 10, Annex 3).
4. Second life of batteries
According to this scenario, all Li-ion EV battery packs can be used in stationary storage
applications after the end of their life in an EV (1st life with duration of 10 years). As
such, the demand for Li-ion batteries from energy storage applications is largely
covered by the primary EV market. As a result, the total production of Li-ion batteries
is lower compared with the reference scenarios. In this scenario it is assumed that
used EV battery packs are re-sold at 50 % of the cost to Li-ion battery pack
manufacturers for energy storage (this cost is assumed to account for complete
revamping of the used EV battery pack).
5. Spill-over learning
In this scenario it is assumed that manufacturing of Li-ion batteries for stationary
storage is fully independent from the production of EV batteries. Similarly, learning
from inverters of photovoltaics does not transfer to PCS components of storage
systems. Learning effects on cost of batteries for stationary storage depend only on the
deployment of energy storage applications and not on EVs.
6. Cell level learning
This scenario aims to capture that learning is primarily relevant at a cell level, where
production scale, material substitution, improved performance and synergies due to
multiple applications may take place. Li-ion cell manufacturing for EVs and storage
benefits from experience also from electronics production. In this scenario learning
rates for electronics apply for Li-ion cells in all segments. For the remaining
components (i.e. pack, inverter, rest for stationary storage) reference learning rates
apply (Table 13, Annex 3).

42
Re-using Li-ion EV battery packs for stationary storage applications, lowers the costs of
the latter due to lower purchasing costs of battery packs by stationary battery storage
producers (19). As such, the results for this scenario are entirely driven by the cost of the
used EV battery pack (here assumed that refurbished batteries come at a 50 % cost of a
new battery pack). This reveals the need for estimating cost effects and life cycle cost of
EV batteries when used for other applications more precisely. Moreover, second life of
electric vehicle batteries contributes towards improving their sustainability performance
and facilitating circular economy [120].
Other assumptions, such as the evolution of the battery size or the lifetime of Li-ion
batteries are not found to affect the results on costs as the overall difference is less
± 10 %.
The absolute range of results across the sensitivity scenarios is shown in Annex 4. Figure
41 summarises the results.

(19) The issue of second life of EV batteries is acknowledged to be very complex. There are indications that
refurbished battery packs may be suitable for a limited number of energy storage applications (such as
UPS, frequency regulation and voltage support) and may just as well be re-directed to mobility
applications; the development of Vehicle-to-Grid applications may have an impact on the EV battery life
and could as such influence the second life potential [129]. Nevertheless, the overall effect of lower
required manufacturing capacity thanks to re-deployment of the EV batteries in various second life
applications would remain and results of this analysis are not expected to vary significantly.

43
 Figure 41 Effects of what-if and sensitivity scenarios on the reference cost estimates

Cell level learning

Battery pack for electric

Spill-over
vehicles

Second life

Battery lifetime

Battery size

Learning rate

-50% -40% -30% -20% -10% 0% 10% 20% 30% 40% 50%
Cost difference compared with reference results

high moderate low

Energy-designed stationary

Cell level learning

Spill-over
storage system

Second life

Battery lifetime

Battery size

Learning rate

-50% -40% -30% -20% -10% 0% 10% 20% 30% 40% 50%
Cost difference compared with reference results

high moderate low

Power-designed stationary

Cell level learning

Spill-over
storage system

Second life

Battery lifetime

Battery size

Learning rate

-50% -40% -30% -20% -10% 0% 10% 20% 30% 40% 50%
Cost difference compared with reference results

high moderate low

44
5 Conclusions
Li-ion batteries are seen as the technology that can help decarbonise transport, lift the
penetration levels of intermittent renewable energy and offer a competitive edge to the
EU's industry in the Li-ion battery value chain. Batteries, including Li-ion, are recognised
as a key enabling technology for the energy transition of the EU under the Energy Union
and as such, they are they are specifically mentioned in several policy initiatives
addressing transport [2,13–15], raw materials [19–21] and energy [5] economic sectors,
EU industrial policy [16] and Research and Innovation in the EU [22–24]. The strategic
importance of batteries for the EU is further demonstrated by the formation of the
European Battery Alliance [25]. Until recently, costs of Li-ion batteries for mobility and
stationary storage applications have been one of their barriers to their large scale
deployment. The cost reduction over the last two years, largely owing to experience from
electronics and surge of EV sales, raises expectations that it will continue into the future,
ultimately leading to penetration of electric vehicles and stationary storage in great
numbers. Based on announcements, the global Li-ion cell manufacturing capacity is
expected to quadruple by 2021 or even increase six-fold by 2022 compared with 2017.
In the longer term, studies estimate that 150 to 900 million EVs may be on the road by
2040, which is about two to three orders of magnitude higher compared with today. Over
the same period, stationary storage may reach up to 1 300 GWh, compared with 3 –
4 GWh today. These projections point towards a potentially significant market growth but
also to diverging views about the future. These projections depend on the direction the
world will take, for example, on action against climate change, or on when and how
steeply costs will decline. Despite the near-term announcements, global manufacturing
capacity is subject to uncertainty in the longer-term. Future costs of Li-ion batteries are
directly influenced by this uncertainty to the extent that their production costs depend on
economies of scale and the cumulative manufacturing experience gained globally. In
turn, technology costs and cost projections are a fundamental parameter that feeds in
the policy process.
Focusing on Li-ion batteries as the family of batteries for mobility and stationary storage
applications of today and the near future, this report contextualises their potential cost
trajectories based on the annual production output which is derived from global energy
scenarios up to 2040. To do so, this report operationalises the learning curve method,
which combines the historical rate of cost reduction achieved for every doubling of
installed capacity of a technology (learning rate) with projections on its deployment over
a period of time.
The demand for EVs and stationary storage is based on three scenarios, namely high,
moderate and low that cover a range of possible futures. The high scenario sees the
highest deployment of EVs and stationary storage, assuming that in the longer-term
market forces drive decisions based on technology costs. The storyline of the moderate
scenario is focused on CO2 emission reduction. It sees a strong growth of EVs to
decarbonise the transport sector, yet lower than in the high scenario, possibly due to the
role of biofuels in reducing road transport emissions. In the power sector, several
technologies compete as low carbon energy solutions (e.g. CCS, biomass, intermittent
renewables) and the role of stationary storage is less pronounced compared with the high
scenario. Finally, based on the low scenario, the world is at a standstill when it comes to
further action against climate change. As a result, sales of EVs and stationary storage are
limited. In each of these scenarios, the global manufacturing capacity required to meet
the annual demand is enormous. By 2040, almost 4 TWh would be sold annually in the
high scenario, around 2 TWh in the moderate scenario, and 600 GWh would be needed in
the low scenario, compared with about 60 to 70 GWh of annual sales in 2017. This
translates to about 110, 55 and 15 operational Li-ion battery gigafactories in each
scenario, respectively, by 2040 (assuming that the annual production capacity of a
gigafactory is 35 GWh). At these scales, next to anticipated improvements in Li-ion cell
chemistries, optimisation of manufacturing processes, standardisation of design and

45
possibly vertical integration of plants, the costs of Li-ion battery pack for EVs and
stationary storage could fall drastically, as also implied by recent literature.
The present report shows that by 2030, Li-ion EV battery packs could come at least at
half their cost of today. By 2040 the cost could drop another 50 %, ultimately reaching
50 €/kWh. These cost trajectories are in line with most recent estimates of other studies.
The SET Plan target for Li-ion EV battery pack cost (75 €/kWh by 2030 [23]) is feasible
in both the high and moderate scenarios. This entails fast ramp-up of Li-ion
manufacturing capacity, of about 2 gigafactories globally per year until 2030. Should
unfavourable conditions prevail (e.g. low scenario), meeting the SET Plan cost target
seems more likely around 2040. If the demand of cells increases rapidly in the following
years, there is the caveat that existing and under construction manufacturing capacities
may not be enough to cover global demand. Shortage of batteries may result in
temporary increase their price above the costs projected in the present report. Similar
situation may result from shortage of mining capacity for the critical materials (e.g.
cobalt).
As EVs drive the demand for Li-ion battery packs, cost reduction spills over to stationary
storage systems, but somewhat slower due to the contribution of other major cost
components (e.g. inverters, BOS hardware, soft costs). Overall in 2040, in the high
scenario, costs of stationary storage systems may be lower by two-thirds compared with
today. Costs of inverters may decrease by more than half (20) benefiting from the strong
deployment of photovoltaics. BOS hardware, soft costs and other cost components may
reduce up to 60 % by 2040 but only in the high scenario, in which stationary storage
sees strong growth. In the moderate and low scenarios these costs reduce less than
50 %. As such, BOS and other soft costs become an area to look further into for cost
reduction potential to lift barriers that may occur in slower growth futures. Ultimately,
stationary system storage costs range between 165 and 240 €/kWh for energy-designed
utility-scale systems, between 280 and 410 €/kWh for power-designed utility-scale
systems and between 250 and 365 €/kWh for households. Results show that costs below
200 €/kWh can be reached after 2030 for energy-designed systems in scenarios with
high deployment of EVs and stationary storage, due to spill-over effects (high scenario).
For comparison, the SET Plan cost target for stationary storage is 150 €/kWh at a system
level and could be attainable at high global deployment of EVs and more than 1 TWh of
stationary storage installed.
Based on inventories and the historical average of commodity prices, materials for Li-ion
battery packs may cost around 30 €/kWh. In today's cost structures, materials represent
between 60 and 75 % of the total cost, which entails that Li-ion battery packs may
ultimately cost 40 – 50 €/kWh. As such, the lower bound of the cost estimated in the
present report (i.e. 50 €/kWh by 2040) seems feasible. Economies of scale could reduce
capital cost, standardisation and automation could limit operating and labour costs, and
improved Li-ion chemistries could further reduce the demand for materials, hence overall
costs.
The learning curve approach that is used in the present report has, however, its
limitations [43]. For instance, it is not suitable to foresee future step changes induced by
disruptions, spill-over from other sectors or commodity price changes. Furthermore, it
does not capture market distortions, which possibly characterise even today's Li-ion
market, such as lower prices due to overcapacity, predatory pricing, or dumping (as was
the case with solar panels). Moreover, the learning curve method generally estimates a
monotonous cost reduction, which increases or slows down based on the demand
projections in the scenarios.

(20) On a kW basis.

46
Box 7 Main take-away messages

Wide-spread deployment of electric vehicles will lead to a rapid decrease of

Li-ion battery costs in the near term
 Cost levels between 75 and 100 €/kWh can be reached in 2025 – 2030,
indicating the feasibility of the cost targets set by the SET Plan.
 At these cost levels, massive deployment of electric vehicles can be expected
globally. By 2030, the numbers of electric vehicles on the road will be 10 to 50
times higher than today. Comparable growth levels are expected in Europe: 12
- 19 % of passenger car fleet are expected to be electric.
 These expected developments offer signals to the EU industry, energy system
providers (e.g. infrastructure) and other stakeholders (e.g. market design) to
anticipate the disruption.
Investment costs of Li-ion battery stationary storage systems will decrease,
yet improvements should focus also on non-battery pack system components
 Li-ion battery stationary storage system costs will decrease significantly
following the development of batteries for electric vehicles due to spill-over
effects.
 By 2040, stationary storage system costs are expected to range between 165
and 410 €/kWh, depending on system configuration.
 High cost share of system storage lies in components such as power electronics,
energy management system, balance of plant, system integration and soft
costs.
 A significant cost reduction for the mentioned system components would be
required, for example, through intensified R&D and increase in manufacturing
capacity.
European manufacturing of Li-ion battery cells will increase its share in global
production, provided that all announced plans materialise. Supplying domestic
demand may prove challenging if capacity does not ramp up after 2025
 In the coming years, the global share of European Li-ion battery cell
manufacturing capacity is expected to increase from about 3 % today to 7 –
25 %. Slightly more than half of this capacity will be deployed by well-
established Asian Li-ion battery cell producers.
 By 2025, the global market for Li-ion batteries for electric vehicles could reach
40 – 55 bn €/yr. For European production this could entail a growth from about
450 M€/yr (2017) to 3 – 14 bn €/yr (2025).
 Albeit relying on imports today, Europe could cover almost completely its annual
demand for Li-ion battery cells by 2025. Thereafter, due to rapid increase of
electric vehicle sales, domestic capacity would need a further significant ramp
up for Europe to remain self-sufficient.
 The global demand for Li-ion batteries is foreseen to grow also in the long term
and may exceed 200 bn €/yr by 2040. To ensure the competitiveness of the EU
on the global scene, efforts on increasing the European manufacturing base
should continue after 2025.
Re-using and repurposing of Li-ion batteries to energy storage applications
after their end of life in electric vehicles contributes to further cost reduction
 In the longer term the volume of batteries from electric vehicles will be
sufficient to supply the needs for stationary storage if batteries are re-used.
 Second life of batteries could reduce costs of stationary storage further by
about 20 % or 30 to 45 €/kWh.

47
The learning rates that were used were found to be the single most influential parameter
in the assessment. The uncertainty assessment on spill-over effects reveals that they
may also prove important, but are unlikely to occur given that Li-ion batteries for
mobility and stationary storage applications are produced by the same manufacturers.
Finally, if re-purposed Li-ion batteries from the transport sector to energy storage are
sold at a fraction of the cost of new batteries, then the effects in stationary storage
systems may be more significant than learning. Other assumptions, such as the evolution
of the battery size or the battery lifetime, while technically highly relevant, were not
found to influence cost projections based on learning.
The top-down cost estimates presented in this report can be complemented by a detailed
bottom-up engineering assessment on the influence of material prices, different
chemistries and innovations. Nonetheless, the upfront costs of Li-ion batteries for
mobility and stationary storage applications presented in this report can be used in
further analysis, for example, to determine the total investment costs of different
technological options to meet peak demand (e.g. open cycle gas turbines against
photovoltaics and different configurations of storage systems), to assess levelised costs
of storage and compare different storage technologies or even to assess the
competitiveness of the EU industry and value chain on Li-ion batteries taking global
developments into account.

48
References
[1] European Commission, 2015, Energy Union Package - A Framework Strategy for a
Resilient Energy Union with a Forward-Looking Climate Change Policy COM(2015)
80 final, European Commission, Brussels, pp: 1-21.
[2] European Commission, 2017, Energy Storage - The role of electricity SWD(2017)
61 final, European Commission, Brussels, pp: 1-25.
[3] IRENA, 2015, Battery Storage for Renewables : Market Status and Technology
Outlook, International Renewable Energy Agency (IRENA), Bonn and Abu Dhabi,
pp: 1-60.
[4] A. Darmani, A. Saleem, B. Normark, L. Krosse, C. Julienne, F. Gardumi, G.
Avgerinopoulos, B. Zakeri, 2017, REEEM Innovation and technology roadmap:
Energy Storage Application, InnoEnergy, KTH, AALTO, pp: 1-92.
http://www.reeem.org/wp-content/uploads/2017/09/REEEM-D2.1a.pdf.
[5] European Commission, 2016, Clean Energy For All Europeans COM(2016) 860
final, European Commission, Brussels, pp: 1-13.
[6] European Union, 2015, Intended Nationally Determined Contribution of the EU and
its Member States, European Union, pp: 1-5.
[7] EC, 2018, Europe leads the global clean energy transition : Commission ambitious
agreement on further renewable energy deployment in the EU, European
Commission (EC), Strasbourg, pp: 1-2. http://europa.eu/rapid/press-
release_STATEMENT-18-4155_en.htm.
[8] IRENA, 2016, The Power To Change: Solar and Wind Cost Reduction Potential to
2025, International Renewable Energy Agency (IRENA), ISBN: 9789295111974,
pp: 1-112.
[9] Fraunhofer ISE, 2015, Current and Future Cost of Photovoltaics, Freiburg, pp: 1-
82.
[10] APS, 2010, Integrating Renewable Electricity on the Grid, American Physical
Society (APS), Washington, DC, pp: 1-40.
[11] H.-W. Sinn, 2017, Buffering Volatility : A Study on the Limits of Germany ’ s
Energy Revolution, Centre for Economic Studies & Ifo Institute, Munich, pp: 1-42.
[12] EEA, 2018, Greenhouse gas emissions from transport, European Environment
Agency (EEA). https://www.eea.europa.eu/data-and-maps/indicators/transport-
emissions-of-greenhouse-gases/transport-emissions-of-greenhouse-gases-10
(accessed July 24, 2018).
[13] European Commission, 2018, Europe on the Move - Sustainable Mobility for
Europe: safe, connected, and clean COM(2018) 293 final, European Commission,
Brussels, pp: 1-14.
[14] European Commission, 2017, Europe on the move - An agenda for a socially fair
transition towards clean, competitive and connected mobility for all COM(2017)
283 final, European Commission, Brussels, pp: 1-18.
[15] European Commission, 2018, Europe on the Move - Sustainable Mobility for
Europe: safe, connected, and clean COM(2018) 293 final Annex 2, European
Commission, Brussels, pp: 1-11.
[16] European Commission, 2017, Investing in a smart, innovative and sustainable
Industry A renewed EU Industrial Policy Strategy COM(2017) 479 final, European
Commission, Brussels, pp: 1-18.
[17] EEA, 2016, Electric vehicles and the energy sector - impacts on Europe’s future
emissions, pp: 1–5. https://www.eea.europa.eu/themes/transport/electric-
vehicles/electric-vehicles-and-energy (accessed August 10, 2018).
[18] T. Brown, D. Schlachtberger, A. Kies, S. Schramm, M. Greiner, 2018, Synergies of
sector coupling and transmission reinforcement in a cost-optimised, highly

49
 renewable European energy system, Energy, 160: 720–739,
doi:10.1016/j.energy.2018.06.222.
[19] European Commission, 2008, The raw materials initiative — meeting our critical
needs for growth and jobs in Europe COM(2008) 699 final, European Commission,
Brussels, pp: 1-14.
[20] European Commission, 2018, The European Innovation Partnership (EIP) on Raw
Materials, Growth - Internal Market, Industry, Enterpreneurship and SMEs.
https://ec.europa.eu/growth/tools-databases/eip-raw-materials/ (accessed
November 6, 2018).
[21] European Commission, 2017, on the 2017 list of Critical Raw Materials for the EU
COM(2017) 490 final, European Commission, Brussels, pp: 1-8.
[22] European Commission, 2015, Towards an Integrated Strategic Energy Technology
(SET) Plan: Accelerating the European Energy System Transformation C(2015)
6317 final, European Commission, Brussels, pp: 1-17.
[23] European Commission, 2016, SET‐Plan ACTION n°7 – Declaration of Intent “
Become competitive in the global battery sector to drive e ‐ mobility forward”
European Commission, pp: 1-9.
[24] SET Plan Temporary Working Group, 2018, Integrated SET-Plan Action 7 -
Implementation Plan “Become competitive in the global battery sector to drive e-
mobility and stationary storage forward,” European Commission, pp: 1-70.
[25] EC, 2018, European Battery Alliance, European Commission (EC).
https://ec.europa.eu/growth/industry/policy/european-battery-alliance_en
(accessed July 24, 2018).
[26] BNEF, 2018, Long-Term Electric Vehicle Outlook 2018 (EVO 2018), Bloomberg New
Energy Finance (BNEF), pp: 1-100.
[27] C. Pieper, H. Rubel, 2011, Revisiting Energy Storage: There Is a Business Case,
Boston Consulting Group (BCG), pp: 1-23.
http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Revisiting+Ener
gy+Storage:+There+Is+a+Business+Case#0.
[28] EASAC, 2017, Valuing dedicated storage in electricity grids, European Academies
Science Advisory Council (EASAC), Halle, ISBN: 978-3-8047-3729-7, pp: 1-51.
[29] K. Ardani, E.O.’ Shaughnessy, R. Fu, C. Mcclurg, J. Huneycutt, R. Margolis, 2017,
Installed Cost Benchmarks and Deployment Barriers for Residential Solar
Photovoltaics with Energy Storage: Q1 2016, National Renewable Energy
Laboratory (NREL), Denver, ISBN: December 2013, pp: 1-41.
http://www.nrel.gov/docs/fy17osti/67474.pdf.
[30] A. Dietrich, C. Weber, 2018, What drives profitability of grid-connected residential
PV storage systems? A closer look with focus on Germany, Energy Economics, 74:
399–416, doi:10.1016/J.ENECO.2018.06.014.
[31] IEA, 2017, Global EV Outlook 2017: Two million and counting, Organisation for
Economic Co-operation and Development (OECD), International Energy Agency
(IEA), France, ISBN: 9789264278882, pp: 1-71, doi:10.1787/9789264278882-en.
[32] K. Palmer, J.E. Tate, Z. Wadud, J. Nellthorp, 2018, Total cost of ownership and
market share for hybrid and electric vehicles in the UK, US and Japan, Applied
Energy, 209: 108–119, doi:10.1016/j.apenergy.2017.10.089.
[33] D. Karoline, K.J. Steinbacher, Minke Goes, 2018, Incentives for Electric Vehicles in
Norway, Ecofys und adelphi, pp: 25. https://www.euki.de/wp-
content/uploads/2018/09/fact-sheet-incentives-for-electric-vehicles-no.pdf.
[34] European Commission, 2016, Accelerating Clean Energy Innovation COM(2016)
0763 final, European Commission, Brussels.
[35] European Commission, 2017, Batteries - A major opportunity for a sustainable

50
 society - Research & Innovation Projects for Policy, European Commission,
Luxembourg, ISBN: 978-92-79-68686-3, pp: 1-40, doi:10.2777/864893.
[36] N. Lebedeva, F. Di Persio, L. Boon-Brett, 2017, Lithium ion battery value chain and
related opportunities for Europe, Publications Office of the European Union,
Luxembourg, ISBN: 978-92-79-66948-4, pp: 1-81, doi:10.2760/6060.
[37] ANL, 2017, BatPac Version 3.1, Excel Spredsheet. http://www.cse.anl.gov/batpac/
(accessed March 6, 2018).
[38] BNEF, 2018, Electric Vehicles Data Hub, Bloomberg New Energy Finance (BNEF).
www.bnef.com (accessed September 10, 2018).
[39] BNEF, 2017, New Energy Outlook 2017 Global Overview, Bloomberg New Energy
Finance (BNEF), pp: 1-78.
[40] M. Junginger, W. van Sark, A. Faaij, 2010, Technological learning in the energy
sector: lessons for policy, industry and science., 2012th ed., Edward Elgar
Publishing, Cheltenham and Northampton, ISBN: 9781848448346, pp: 1-332.
[41] E.S. Rubin, I.M.L. Azevedo, P. Jaramillo, S. Yeh, 2015, A review of learning rates
for electricity supply technologies, Energy Policy, 86: 198–218,
doi:10.1016/j.enpol.2015.06.011.
[42] A. Graf, M. Buck, 2017, The cost of renewable energy: A critical assessment of the
Impact Assessments underlying the Clean Energy for All Europeans-Package,
Agora Energiewende, pp: 1-19.
[43] M. Steen, N. Lebedeva, F. Di Persio, L. Boon-Brett, 2017, EU Competitiveness in
Advanced Li-ion Batteries for E-Mobility and Stationary Storage Applications –
Opportunities and Actions, Publications Office of the European Union, Luxembourg,
ISBN: 978-92-79-74292-7, pp: 1-49, doi:10.2760/75757.
[44] Avicenne Energy, 2018, The Rechargeable Battery Market and Main Trends 2017-
2025: 100 pages update, Avicenne Energy, ISBN: 2008200920102, pp: 1-111.
[45] IEA, 2018, Global EV Outlook 2018 (Towards cross-modal electrification),
Organisation for Economic Co-operation and Development (OECD), International
Energy Agency (IEA), pp: 1-141.
https://webstore.iea.org/download/direct/1045?filename=globalevoutlook2018.pdf
.
[46] BNEF, 2018, Cumulative Global EV Sales Hit 4 Million.
https://about.bnef.com/blog/cumulative-global-ev-sales-hit-4-million/ (accessed
September 14, 2018)
[47] Avicenne Energy, 2018, Battery market for Hybrid, Plug-in & Electric Vehicles,
Avicenne Energy, pp: 1-64.
[48] EAFO, 2018, Vehicle Stats. https://www.eafo.eu/vehicle-statistics (accessed
October 3, 2018)
[49] V. Adesanya-Aworinde, 2016, Electric bus sector is game changer for battery
market. https://www.idtechex.com/research/articles/electric-bus-sector-is-game-
changer-for-battery-market-00009175.asp (accessed August 9, 2018).
[50] Z. Gao, Z. Lin, T.J. LaClair, C. Liu, J.-M. Li, A.K. Birky, J. Ward, 2017, Battery
capacity and recharging needs for electric buses in city transit service, Energy,
122: 588–600, doi:10.1016/J.ENERGY.2017.01.101.
[51] BNEF, 2018, Long-Term Electric Vehicle Outlook 2018, Bloomberg New Energy
Finance (BNEF), pp: 1-100.
[52] L. Grande, 2018, Li-ion Batteries 2018-2028 From Raw Materials to New Materials,
through Gigafactories and Emerging Markets, IDTechEx, pp: 1-496.
[53] WattEV2Buy, 2018, Electric Cars Models List. https://wattev2buy.com/ (accessed
July 19, 2018).
[54] High Edge, 2017, Next Generation Electric Vehicle Market 2017 - Automotive

51
 Market Reporting in the Electric Era 2017 Edition, High Edge Corporation, pp: 1-
113. http://hiedge.co.jp/dm/2017ランキングサンプル.pdf.
[55] L. Hsuesh-lung, 2016, An update on the Chinese xEV market and a Technical
Comparison of its Batteries, in: World Mobility Summit, Industrial Technology
Research Institute (ITRI): p. 41. https://www.emove360.com/wp-
content/uploads/2016/11/20161019-An-Update-on-the-Chinese-xEV-Market-and-
a-Technical-Comparison-of-its-Batteries.pdf.
[56] US DOE, 2018, DOE Global Energy Storage Database.
http://www.energystorageexchange.org/projects (accessed March 30, 2018).
[57] O. Schmidt, A. Hawkes, A. Gambhir, I. Staffell, 2017, The future cost of electrical
energy storage based on experience rates, Nature Energy, 6: 17110,
doi:10.1038/nenergy.2017.110.
[58] BNEF, 2017, 2017 Global Energy Storage Forecast, Bloomberg New Energy Finance
(BNEF), pp: 1-54.
[59] IEA, 2017, Who wants to be in charge?, International Energy Agency (IEA).
https://www.iea.org/newsroom/news/2017/november/commentary-battery-
production---who-wants-to-be-in-charge.html.
[60] B. Nykvist, M. Nilsson, 2015, Rapidly falling costs of battery packs for electric
vehicles, Nature Climate Change, 5: 329–332, doi:10.1038/nclimate2564.
[61] UCS, 2018, Electric Vehicle Battery: Materials, Cost, Lifespan, Uion of Concerned
Scientists (UCS). https://www.ucsusa.org/clean-vehicles/electric-vehicles/electric-
cars-battery-life-materials-cost (accessed August 9, 2018).
[62] IEA, 2016, World Energy Outlook 2016, Organisation for Economic Co-operation
and Development (OECD), International Energy Agency (IEA), Paris, ISBN: 978-
92-64-26495-3, pp: 1-684.
[63] IEA, 2017, World Energy Outlook 2017, Organisation for Economic Co-operation
and Development (OECD), International Energy Agency (IEA), Paris, ISBN: 978-
92-64-28230-8, pp: 1-766.
[64] BCG, 2017, The Electric Car Tipping Point - Research highlights, Boston Consulting
Group (BCG), pp: 1-23. https://www.bcg.com/Images/Highlights_BCG Studie_The
Electric Car Tipping Point_tcm108-175822.pdf.
[65] M. Hocking, J. Kan, P. Young, C. Terry, D. Begleiter, 2016, Lithium 101 F.I.T.T. for
investors Welcome to the Lithium-ion Age, Deutsche Bank Market Research, ISBN:
1055100898643, pp: 1-179.
http://www.belmontresources.com/LithiumReport.pdf.
[66] Avicenne Energy, 2017, The Worldwide Rechargeable Battery Market 2016-2025,
26th edition, Avicenne Energy, pp: 1-205.
[67] BNEF, 2017, Long-Term Electric Vehicle Outlook 2017 (EVO 2017), Bloomberg New
Energy Finance (BNEF), pp: 1-58.
[68] U.S. DoE, 2017, Cost and Price Metrics for Automotive Lithium-Ion Batteries, U.S.
Department of Energy (US DOE), Energy Efficiency & Renewable Energy (EERE),
pp: 1-4.
[69] S. De Leon, 2017, Galaxy Note 7 lessons learned, in: Batteries 2017, Nice.
[70] ReverseTheCharge.com, 2018, Reverse The Charge.
https://www.reversethecharge.com/ (accessed November 5, 2018)
[71] Research Interfaces, 2018, What do we know about next-generation NMC 811
cathode?. https://researchinterfaces.com/know-next-generation-nmc-811-
cathode/ (accessed August 10, 2018).
[72] F. Lambert, 2017, Tesla is starting Model 3 battery cell production at Gigafactory 1
“right now,” Electrek. https://electrek.co/2017/06/19/tesla-model-3-battery-cell-
production/ (accessed August 9, 2018).

52
[73] Cobalt27, 2018, Corporate Presentation, pp: 1-41.
https://www.cobalt27.com/_resources/presentations/KBLT-corporate-
presentation.pdf
[74] T. Zhao, 2018, Sustainable Development Strategy for EV Battery, BYD, pp: 1-31.
https://www.iea.org/media/Workshops/2018/Session2ZhaoTongBYD.pdf.
[75] P. Lima, 2017, LG Chem will introduce NCM 811 battery cells for EVs next year.
https://pushevs.com/2017/09/08/lg-chem-will-introduce-ncm-811-battery-cells-
evs-next-year/ (accessed August 10, 2018).
[76] P. Lima, 2017, SK innovation to start producing NCM 811 battery cells soon.
https://pushevs.com/2017/09/02/sk-innovation-start-producing-ncm-811-battery-
cells-soon/ (accessed August 10, 2018).
[77] E.A. Olivetti, G. Ceder, G.G. Gaustad, X. Fu, 2017, Lithium-Ion Battery Supply
Chain Considerations: Analysis of Potential Bottlenecks in Critical Metals, Joule, 1:
229–243, doi:https://doi.org/10.1016/j.joule.2017.08.019.
[78] K. Turcheniuk, D. Bondarev, V. Singhal, G. Yushin, 2018, Ten years left to redesign
lithium-ion batteries, Nature, 559: 467–470, doi:10.1038/d41586-018-05752-3.
[79] D. Bresser, K. Hosoi, D. Howell, H. Li, H. Zeisel, K. Amine, S. Passerini, 2018,
Perspectives of automotive battery R&amp;D in China, Germany, Japan, and the
USA, Journal of Power Sources, 382: 176–178,
doi:10.1016/J.JPOWSOUR.2018.02.039.
[80] N. Nitta, F. Wu, J.T. Lee, G. Yushin, 2015, Li-ion battery materials: present and
future, Materials Today, 18: 252–264,
doi:https://doi.org/10.1016/j.mattod.2014.10.040.
[81] AVICENNE ENERGY, 2017, The Rechargeable Battery Market and Main Trends
2015-2025, in: Advanced Automotive Battery Conference, Avicenne Energy: p. 98.
http://cii-resource.com/cet/FBC-TUT8/Presentations/Pillot_Christophe.pdf.
[82] M. Gupta, 2018, U.S. Front-of-the-Meter Energy Storage System Prices 2018-
2022, GMT Research. https://www.greentechmedia.com/research/report/us-front-
of-the-meter-energy-storage-system-prices-2018-2022#gs.LXJbdqE.
[83] IRENA, 2017, Electricity Storage and Renewables - Costs and Markets to 2030,
International Renewable Energy Agency (IRENA), ISBN: 978-92-9260-038-9, pp:
1-132.
[84] Lazard, 2017, Lazard’s Leveized Cost of Storage Analysis - Version 3.0, Lazard. pp:
1-49
[85] BNEF, 2017, Storage System Costs More than Just a Battery, Bloomberg New
Energy Finance (BNEF).
[86] B.D. Frankel, S. Kane, C. Tryggestad, 2018, The new rules of competition in
energy storage, McKinsey & Company, pp: 1-9.
https://www.mckinsey.com/industries/electric-power-and-natural-gas/our-
insights/The-new-rules-of-competition-in-energy-storage?cid=other-eml-alt-mip-
mck-oth-
1806&hlkid=344ddedf47264589bf915495716bbf03&hctky=10298008&hdpid=9d54
27a6-34aa-4b8c-92d2-18fa621ab.
[87] A. Eller, D. Gauntlett, 2017, Energy storage trends and opportunities in emerging
markets, Navigant Consulting, Inc., pp: 1-52.
[88] BNEF, 2017, Energy Storage Project Database, Bloomberg New Energy Finance
(BNEF). www.bnef.com (accessed April 30, 2018).
[89] N. Kittner, F. Lill, D.M. Kammen, 2017, Energy storage deployment and innovation
for the clean energy transition, Nature Energy, 2: 1–6,
doi:10.1038/nenergy.2017.125.
[90] A. Fiorini, A. Georgakaki, F. Pasimeni, E. Tzimas, 2017, Monitoring R&I in Low-
Carbon Energy Technologies, Publications Office of the European Union,

53
 Luxembourg, ISBN: 9789279655920, doi:10.2760/447418.
[91] B. Eckhouse, D. Pogkas, 2018, EVs Take Over Driving Seat for Global Battery-Pack
Sales:, Bloomberg News, p: 1.
[92] G. Zubi, R. Dufo-López, M. Carvalho, G. Pasaoglu, 2018, The lithium-ion battery:
State of the art and future perspectives, Renewable and Sustainable Energy
Reviews, 89: 292–308, doi:10.1016/j.rser.2018.03.002.
[93] Nationale Plattform Elektromobilität (NPE), 2016, Roadmap integrierte Zell- und
Batterieproduktion in Deutschland, Nationale Plattform Elektromobilität (NPE),
Berlin, pp: 1-63.
[94] BNEF, 2018, Equipment Manufacturers - Battery cells. www.bnef.com (accessed
June 29, 2018).
[95] N. Lutsey, M. Grant, S. Wappelhorst, H. Zhou, B., 2018, Power play: How
governments are spurring the electric vehicle industry, The International Council
on Clean Transportation (ICCT), Washington, pp: 1-41.
https://www.theicct.org/sites/default/files/publications/EV_Government_WhitePap
er_20180514.pdf.
[96] I. Presinger, V. Bryan, 2018, China’s CATL to build its first European EV battery
factory in Germany, Reuters. https://www.reuters.com/article/us-bmw-catl-
batteries/chinas-catl-to-build-its-first-european-ev-battery-factory-in-germany-
idUSKBN1JZ11Y (accessed August 8, 2018).
[97] B. Felix, B. Mallet, G. Guillaume, 2018, France’s Saft targets new generation
battery production from 2020. https://www.reuters.com/article/autos-batteries-
saft/frances-saft-targets-new-generation-battery-production-from-2020-
idUSL5N1VX4TZ (accessed November 16, 2018).
[98] P. Lima, 2018, LG Chem to triple EV battery production in Poland.
https://pushevs.com/2018/03/12/lg-chem-to-triple-ev-battery-production-in-
poland/ (accessed November 13, 2018).
[99] Der Tagesspiegel, 2018, Der Minister und die Batteriezelle.
https://www.tagesspiegel.de/wirtschaft/elektromobilitaet-der-minister-und-die-
batteriezelle/23166586.html (accessed November 14, 2018).
[100] M. Kane, 2018, BYD Looks To Enter European Battery Market With New Factory.
https://insideevs.com/byd-looks-to-enter-european-battery-market-with-new-
factory/ (accessed November 14, 2018).
[101] Nissan Global, 2018, Nissan to Sell Electric Battery Business to Envision Group.
https://newsroom.nissan-global.com/releases/release-
ed7b0014763a42e1693c5c954e0607c2-180803-01-e (accessed November 14,
2018).
[102] electrive.com, 2018, Battery cell production in Germany hung out to dry.
https://www.electrive.com/2018/10/10/battery-cell-production-in-germany-hung-
out-to-dry/ (access November 14, 2018)
[103] IEA, 2017, Energy Technology Perspectives 2017 Catalysing Energy Technology
Transformations, Organisation for Economic Co-operation and Development
(OECD), International Energy Agency (IEA), Paris, ISBN: 978-92-64-27597-3, pp:
1-443.
[104] M. Kah, 2018, Electric vehicles and their impact on oil demand: Why forecasts
differ, Center on Global Energy Policy (CGEP), Columbia University, pp: 1-14.
[105] BNEF, 2018, New Energy Outlook 2018, Bloomberg New Energy Finance (BNEF),
pp: 1-181.
[106] BP, 2018, 2018 BP Energy Outlook, British Petroleum (BP), pp: 1-125.
[107] OPEC, 2016, 2016 World Oil Outlook, Vienna, ISBN: 9783950272222, pp: 428,
doi:10.1190/1.1439163.

54
[108] IRENA, 2017, Electric Vehicles: Technology Brief, International Renewable Energy
Agency (IRENA), Abu Dhabi, ISBN: 9789295111004, pp: 1-52.
http://www.irena.org/-
/media/Files/IRENA/Agency/Publication/2017/IRENA_Electric_Vehicles_2017.pdf.
[109] Wood Mackenzie, 2017, How fast will EV market share grow, Wood Mackenzie.
https://www.woodmac.com/news/editorial/ev-market-share/ (accessed July 24,
2018).
[110] J. Paige, C. McMillan, D. Stenberg, M. Muratori, L. Vimmerstedt, T. Mai, 2017,
Electrification Futures Study : End-Use Electric Technology Cost and, National
Renewable Energy Laboratory (NREL), Golden, Co, pp: 1-109.
https://www.nrel.gov/docs/fy18osti/70485.pdf.
[111] G. Berckmans, M. Messagie, J. Smekens, N. Omar, L. Vanhaverbeke, J. Van Mierlo,
2017, Cost projection of state of the art lithium-ion batteries for electric vehicles
up to 2030, Energies, 10:, doi:10.3390/en10091314.
[112] P. A. Nelson, K.G. Gallagher, I. Bloom, D.W. Dees, 2012, Modeling the
Performance and Cost of Lithium-Ion Batteries for Electric-Drive Vehicles, 2nd ed.,
Argonne National Laboratory (ANL), Chemical Sciences and Engineering Division,
Chicago, pp: 116. http://www.cse.anl.gov/batpac/files/BatPaC ANL-12_55.pdf.
[113] A. Abdon, X. Zhang, D. Parra, M.K. Patel, C. Bauer, J. Worlitschek, 2017, Techno-
economic and environmental assessment of stationary electricity storage
technologies for different time scales, Energy, 139: 1173–1187,
doi:10.1016/j.energy.2017.07.097.
[114] IRENA, 2018, Cost of Service Tool v1-0, IRENA.
http://www.irena.org/publications/2017/Oct/Electricity-storage-and-renewables-
costs-and-markets#RestrictedModal
[115] A. Belderbos, E. Delarue, K. Kessels, W. D’haeseleer, 2017, Levelized cost of
storage — Introducing novel metrics, Energy Economics, 67: 287–299,
doi:10.1016/J.ENECO.2017.08.022.
[116] I. Tsiropoulos, D. Tarvydas, A. Zucker, 2018, Cost development of low carbon
energy technologies - Scenario-based cost trajectories to 2050, 2017 Edition,
European Union, Luxembourg, ISBN: 978-92-79-77479-9,
doi:doi:10.2760/490059.
[117] D. Sandalow, C. McCormick, T. Rowlands-Rees, A. Izadi-Najafabadi, I. Orlandi,
2015, Distributed solar and storage - ICEF Roadmap 1.0, Innovation for Cool Earth
Forum (ICEF), Bloomberg New Energy Finance (BNEF), pp: 1-67.
[118] S.J. Gerssen-Gondelach, A.P.C. Faaij, 2012, Performance of batteries for electric
vehicles on short and longer term, Journal of Power Sources, 212: 111–129,
doi:10.1016/j.jpowsour.2012.03.085.
[119] Avicenne Energy, 2017, Energy Storage Systems - Grid to Behind the Meter -
Market & Main Trends, v03 ed., Avicenne Energy, pp: 1-131.
[120] S. Bobba, A. Podias, F. Di Persio, M. Messagie, P. Tecchio, M.A. Cusenza, U.
Eynard, F. Mathieux, A. Pfrang, 2018, Sustainability Assessment of Second Life
Application of Automotive Batteries (SASLAB): JRC Exploratory Research (2016-
2017): Final technical report: August 2018, Publications office of the European
Union, Luxembourg, ISBN: 978-92-79-92835-2, pp: 140, doi:10.2760/53624.
[121] Metalary, 2018, Metal prices. https://www.metalary.com/ (accesss September 26,
2018)
[122] Northern Graphite, 2018, Graphite pricing. http://northerngraphite.com/graphite-
pricing/ (accesss September 26, 2018)
[123] W. Tahil, 2010, How Much Lithium does a LiIon EV battery really need ?, Meridian
International Research, pp: 1-11.
[124] E3Mlab, 2014, PRIMES Model 2013-2014 Detailed model description,

55
 E3MLab/ICCS, National Technical University of Athens, Athens, pp: 1-155.
[125] L. Mantzos, T. Wiesenthal, I. Kourti, N. Matei, E. Navajas Cawood, A. Papafragkou,
M. Rozai, P.R. Hans, A. Soira Ramirez, 2016, POTEnCIA Model Description -
Version 0.9, European Commission (EC), Joint Research Centre (JRC), ISBN: 978-
92-79-56945-6, pp: 150, doi:10.2791/416465.
[126] S. Simoes, W. Nijs, P. Ruiz, A. Sgobbi, D. Radu, P. Bolat, C. Thiel, S. Peteves,
2013, The JRC-EU-TIMES model. Assessing the long-term role of the SET Plan
Energy technologies, European Commission, Luxembourg, ISBN: 978-92-79-
34506-7, pp: 1-382, doi:10.2790/97596.
[127] S. Gandon, 2017, China’s Green Vehicle Revolution To Reshuffle The Cards For
Cobalt.
[128] InsideEVs, 2017, China Approves NMC Battery Technology For Green Car
Subsidies.
[129] J.M. Durand, 2018, Personal communication, European Association for Storage of
Energy (EASE).

56
List of abbreviations and definitions

BEV Battery Electric Vehicle (or fully electric vehicle)

BOS Balance Of System
CCS Carbon Capture and Storage
CGEP Center on Global Energy Policy
EBIT Earnings Before Interest & Tax
EMS Energy Management System
EPC Engineering, Procurement and Construction
EV Electric Vehicle
LCO Lithium Cobalt Oxide
LFP Lithium Iron Phosphate
Li-ion Lithium ion
NCA Nickel Cobalt Aluminium oxide
NMC Nickel Cobalt Manganese oxide
PCS Power Conversion System
PCT Patent Cooperation Treaty
PHEV Plug-in Hybrid Electric Vehicle
RES Renewable Energy Sources
SET Plan Strategic Energy Technology Plan
US DOE United States Department of Energy

57
List of boxes
Box 1 Approaches used to estimate costs of Li-ion batteries.......................................30
Box 2 Techno-economic performance of EV batteries ................................................32
Box 3 Description of the high scenario (BNEF NEO 2018) ...........................................35
Box 4 Description of the moderate scenario (IEA ETP 2 DS 2017) ...............................35
Box 5 Description of the low scenario (IEA ETP RTS 2017) .........................................35
Box 6 Description of what-if and sensitivity scenarios ................................................42
Box 7 Main take-away messages ............................................................................47

58
List of figures
Figure 1 Global historical annual growth Li-ion batteries in main market segments ........ 9
Figure 2 Total global EV fleet (excluding electric buses) in different regions in 2010 –
2017 ...................................................................................................................10
Figure 3 Annual new EV sales per EV type (BEV or PHEV) ..........................................10
Figure 4 Li-ion battery demand for electric buses in 2017/2018 ..................................11
Figure 5 Weighted average performance of BEV and PHEV based on sales in 2013 and
2017 per producer ................................................................................................11
Figure 6 Global cumulative installed capacity of electrochemical system storage ...........12
Figure 7 Global annual sales of electrochemical storage .............................................13
Figure 8 Global installed capacity of energy storage installed behind-the-meter in 2016 13
Figure 9 Reported Li-ion battery pack costs for EVs ..................................................14
Figure 10 Breakdown of the total cost of Li-ion EV battery in key components (upper
figure) and between cell components and other costs across different chemistries (lower
figure) ................................................................................................................16
Figure 11 Element requirement for Li-ion battery cathodes ........................................17
Figure 12 Future NMC chemistry mix in cathodes ......................................................17
Figure 13 Illustrative system cost and price structure of stationary battery storage ......18
Figure 14 Li-ion battery stationary system costs in 2016 and 2017 .............................19
Figure 15 Specific costs of Li-ion battery for stationary system storage expressed per kW
and kWh..............................................................................................................19
Figure 16 Specific system costs of Li-ion battery for stationary storage expressed per
kWh (upper figure) and kW (lower figure) and power-to-energy ratio..........................20
Figure 17 Distribution of power-to-energy ratio based on project size .........................21
Figure 18 Cost breakdown of power-designed (C-rate 2) and energy-designed (C-rate
0.25) grid-scale stationary storage system ..............................................................22
Figure 19 Cost breakdown of Li-ion battery storage system between cell components and
other costs ..........................................................................................................22
Figure 20 Annual patent activity on batteries filed globally and by the EU ....................23
Figure 21 Expected near-term growth in global Li-ion cell manufacturing capacity for
applications such as EVs and stationary storage .......................................................24
Figure 22 Expected evolution of Li-ion cell manufacturing capacity for mobility and
stationary storage applications in the EU (incl. the UK), in GWh .................................25
Figure 23 Annual investments in new Li-ion cell manufacturing capacity and announced
capital costs per unit of battery manufacturing capacity ............................................26
Figure 24 Future global sales of main Li-ion battery market segments .........................26
Figure 25 Projections of the global EV fleet over time ................................................27
Figure 26 Share of BEV and PHEV in the global EV fleet according to different scenarios in
2030 and 2040 .....................................................................................................28
Figure 27 Projections of total stationary storage installed front- and behind-the-meter
globally ...............................................................................................................28
Figure 28 Cost forecasts of Li-ion battery packs for EVs .............................................29

59
Figure 29 Year when EVs are at price parity with internal combustion engine vehicles
based on different scenarios ..................................................................................29
Figure 30 Cost forecasts of Li-ion battery stationary system storage based on literature 31
Figure 31 Historical and near-term Li-ion battery performance improvement of BEVs in
western markets (i.e. excluding China) ...................................................................32
Figure 32 Learning rates of Li-ion batteries for different applications ...........................33
Figure 33 Selected global growth scenarios for EVs (upper figure) and stationary storage
(lower figure) .......................................................................................................34
Figure 34 Learning curves for Li-ion batteries for EVs and stationary storage systems ...36
Figure 35 Cost-development of Li-ion battery packs for EVs over time based on three
different deployment scenarios ...............................................................................38
Figure 36 Global ramp-up of manufacturing capacity of Li-ion cells and annual gigafactory
additions in each scenario to 2030 (upper figure) and from 2030 to 2040 (lower figure) 38
Figure 37 Cost-development of Li-ion battery for utility scale energy-designed stationary
storage systems (C-rate 0.25) over time based on three different deployment scenarios
..........................................................................................................................39
Figure 38 Cost-development of Li-ion battery for utility scale power-designed stationary
storage systems (C-rate 2) over time based on three different deployment scenarios ...40
Figure 39 Cost structure and cost structure development of utility scale stationary
storage systems over time in the high scenario ........................................................40
Figure 40 Cost development of Li-ion residential energy storage (C-rate 0.35) .............41
Figure 41 Effects of what-if and sensitivity scenarios on the reference cost estimates ...44

60
List of tables
Table 1 Main cathode chemistries used in Li-ion battery packs and their application ......15
Table 2 Producers of Li-ion cells for mobility and stationary storage applications in the EU
(including the UK) today and in the near term ..........................................................25
Table 3 Grouping of EV producers per country ..........................................................62
Table 4 Main raw materials in NMC battery packs .....................................................63
Table 5 Historic prices of main raw materials of NMC battery packs ...........................63
Table 6 Deployment scenarios for EVs .....................................................................64
Table 7 Deployment scenarios for stationary storage ................................................64
Table 8 Deployment scenarios for photovoltaics .......................................................64
Table 9 EV fleet composition over time in the selected scenarios ................................65
Table 10 Capital investment costs of Li-ion batteries in 2017 and assumed technical
lifetime ...............................................................................................................65
Table 11 Assumed cost-structure of stationary storage in 2017 ..................................66
Table 12 Development of Li-ion EV battery size over time ..........................................66
Table 13 Learning rates of Li-ion battery packs for EVs and of other stationary storage
components .........................................................................................................66
Table 14 Li-ion battery pack cost trajectories ...........................................................67
Table 15 Li-ion stationary storage system cost trajectories (energy-designed, C-rate
0.25) ..................................................................................................................67
Table 16 Li-ion stationary storage system cost trajectories (power-designed, C-rate 2) .67
Table 17 Li-ion residential stationary storage cost trajectories (C-rate 0.35) ................67
Table 18 Range of Li-ion battery costs across all sensitivity and deployment scenarios ..67

61
Annexes
Annex 1. Cost components and cost boundaries of Li-ion batteries

Table 3 Grouping of EV producers per country

China Europe USA Japan Korea

BAIC Audi Chevrolet Mitsubishi Hyundai

BYD BMW Ford Nissan Kia

BYD-Daimler Citroen Tesla Toyota

Changan Fiat

Chery Mercedes

DongFeng Opel

Geely Peugeot

Hawtai Porsche

JAC Renault

Jiangling Volkswagen

Kandi Volvo

SAIC

Zotye

62
Annex 2. Cost components and cost boundaries of Li-ion batteries

Table 4 Main raw materials in NMC battery packs based on Olivetti et al. [77] and BNEF [26] in
kg/kWh (1)

NMC-111 NMC-811 Main use

Nickel 0.392 0.75 Cathode

Manganese 0.367 0.088 Cathode

Cobalt 0.394 0.094 Cathode

Lithium 0.139 0.111 Cathode, electrolyte

Graphite 1.2 1.2 Anode

Copper 1.4 1.4 Anode, electronics

Aluminium 1.2 1.2 Casing, cooling system

(1) Excluding mainly the separator and plastic components.

Table 5 Historic prices of main raw materials of NMC battery packs in €/t [121,122]

8 year average Historical high

(2010-2017) (2010-2017)

Nickel 15 660 23 288

Manganese 2 400 3 537

Cobalt 32 380 49 285

Lithium carbonate (1) 5 448 8 387

Graphite 1 294 2 700

Copper 6 718 8 969

Aluminium 1 882 2 440

(1) Lithium carbonate prices were converted to lithium based on 5.3 kg lithium carbonate/kg lithium [123].

Based on the material requirement in Table 4 and the prices in Table 5 it is estimated
that NMC-111 materials cost from 47 €/kWh (historical average) to 83 €/kWh (historical
high) and NMC-811 materials cost from 31 €/kWh (historical average) to 48 €/kWh
(historical high).

63
Annex 3. Parameters and detailed input assumptions

Scenario assumptions

Table 6 Deployment scenarios for EVs, in million EVs

2020 2025 2030 2035 2040 Source

High 10 42 134 322 562 BNEF [26]

Moderate 23 74 160 274 410 IEA [103]

Low 9 27 58 101 146 IEA [103]

Table 7 Deployment scenarios for stationary storage, in GWh

2020 2025 2030 2035 2040 Source

High 25 103 378 777 1 326 BNEF [105]

Moderate 4 6 8 30 51 IEA [103]

Low 2 3 4 15 25 (1)

(1) Assumed as half of the deployment of stationary storage in the moderate scenario.

Table 8 Deployment scenarios for photovoltaics, in GWh

2020 2025 2030 2035 2040 Source

High 751 1 346 2 137 3 323 4 524 BNEF [105]

Moderate 564 885 1 321 1 846 2 354 IEA [103]

Low 498 551 736 919 1 115 IEA [103]

64
Table 9 EV fleet composition over time in the selected scenarios (based on the total fleet)

2020 2025 2030 2035 2040 Source

Passenger BEV 61 % 70 % 75 % 80 % 85 %
BNEF
High Passenger PHEV 31 % 27 % 23 % 19 % 14 %
[26]
Electric buses 7% 3% 1% 1% 0.5 %

Passenger BEV 55.5 % 52 % 48 % 44 % 43 %

IEA
Moderate Passenger PHEV 43.5 % 47 % 51 % 54 % 56 %
[103]
Electric buses 1% 1% 1% 1% 1%

Passenger BEV 51 % 40 % 33% 31 % 32 %

IEA
Low Passenger PHEV 48.5 % 59 % 66 % 68 % 67 %
[103]
Electric buses 0.5 % 1% 1% 1% 1%

Techno-economics assumptions

Table 10 Capital investment costs of Li-ion batteries in 2017 and assumed technical lifetime

Unit Value Source

CAPEX

Li-ion battery pack cost €/kWh 200 Selected from a

range of recent
values (Figure 9)

Li-ion stationary storage €/kWh 500 BNEF [85]

system, energy-designed

Li-ion stationary storage €/kWh 820 BNEF [85]

system, power-designed

Li-ion stationary storage, €/kWh 745 BNEF [105]

residential

Lifetime Sensitivity

High Low

EV battery pack years 10 12 8 Own assumption

Stationary storage years 20 25 15 Own assumption

Photovoltaics years 20 - - Own assumption

65
Table 11 Assumed cost-structure of stationary storage in 2017 (estimated on a kWh-basis) based
on BNEF [85]

Power-designed (1) Energy-designed (1)

Battery pack 42 % 50 %

Inverter 21 % 5%

Other components 37 % 45 %

(1) Power-designed battery of C-rate 2 and energy-designed battery of C-rate 0.25.

Table 12 Development of Li-ion EV battery size over time, in kWh

2017 2025 2030 2035 2040 Source

Passenger BEV 40 52 60 60 60 IRENA [108] (1)

Passenger PHEV 18 19 20 20 20 Own assumption

Mini electric busses BEV 80 80 80 80 80 Own assumption

Mini electric busses 30 30 30 30 30 Own assumption

PHEV

Own assumption
Electric busses BEV 150 150 200 200 200
based on [50]

Own assumption
Electric busses PHEV 40 40 40 40 40
based on [50]

(1) IRENA assumes an average of 60 kWh/EV in 2030, here it is assumed constant thereafter.

Learning rates

Table 13 Learning rates of Li-ion battery packs for EVs and of other stationary storage
components

Reference High Low Source

Battery pack 16 % 20 % 12 % Schmidt et al. [57]

Inverters 19 % 20 % 18 % IRENA [8]

Other components 12 % 16 % 8% Schmidt et al. [57] (1)

Cell 30 % Schmidt et al. [57] (1)

(1) Assuming learning rate for storage systems.

66
Annex 4. Results

Table 14 Li-ion battery pack cost trajectories, in €/kWh

2017 2020 2025 2030 2035 2040

High 200 154 104 75 59 50

Moderate 200 134 96 75 64 56

Low 200 171 127 101 86 75

Table 15 Li-ion stationary storage system cost trajectories (energy-designed, C-rate 0.25), in
€/kWh

2017 2020 2025 2030 2035 2040

High 500 404 294 223 187 164

Moderate 500 393 323 284 233 206

Low 500 443 371 327 273 242

Table 16 Li-ion stationary storage system cost trajectories (power-designed, C-rate 2), in €/kWh

2017 2020 2025 2030 2035 2040

High 820 663 494 382 323 282

Moderate 820 658 546 481 402 355

Low 820 732 617 546 466 411

Table 17 Li-ion residential stationary storage cost trajectories (C-rate 0.35), in €/kWh

2017 2020 2025 2030 2035 2040

High 744 601 442 339 285 249

Moderate 744 590 486 427 354 313

Low 744 662 555 489 414 365

Table 18 Range of Li-ion battery costs across all sensitivity and deployment scenarios

2020 2030 2040

EV battery pack 113 – 178 53 – 121 32 – 98

Energy-designed storage 367 – 459 175 – 406 118 – 279

Power-designed storage 662 – 754 315 – 666 216 – 488

67
GETTING IN TOUCH WITH THE EU
In person
All over the European Union there are hundreds of Europe Direct information centres. You can find the
address of the centre nearest you at: https://europa.eu/european-union/contact_en
On the phone or by email
Europe Direct is a service that answers your questions about the European Union. You can contact this
service:
- by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls),
- at the following standard number: +32 22999696, or
- by electronic mail via: https://europa.eu/european-union/contact_en
FINDING INFORMATION ABOUT THE EU
Online
Information about the European Union in all the official languages of the EU is available on the Europa
website at: https://europa.eu/european-union/index_en
EU publications
You can download or order free and priced EU publications from EU Bookshop at:
https://publications.europa.eu/en/publications. Multiple copies of free publications may be obtained by
contacting Europe Direct or your local information centre (see https://europa.eu/european-
union/contact_en).
 KJ-NA-29440-EN-N

doi:10.2760/87175

ISBN 978-92-79-97254-6
2