Net-zero power

Long duration energy storage

for a renewable grid
Published in November 2021 by the LDES Council. Copies of this document are available
upon request or can be downloaded from our website: www.ldescouncil.com.

This report was authored by the LDES Council in collaboration with McKinsey & Company as
knowledge partner. This work is independent, reflects the views of the authors, and has not been
commissioned by any business, government, or other institution. The authors of the report confirm that:

1. There are no recommendations and/or any measures and/or trajectories within the report that
could be interpreted as standards or as any other form of (suggested) coordination between the
participants of the study referred to within the report that would infringe EU competition law; and

2. It is not their intention that any such form of coordination will be adopted.

Whilst the contents of the report and its abstract implications for the industry generally can be
discussed once they have been prepared, individual strategies remain proprietary, confidential and the
responsibility of each participant. Participants are reminded that, as part of the invariable practice of the
LDES Council and the EU competition law obligations to which membership activities are subject, such
strategic and confidential information must not be shared or coordinated – including
as part of this report.
Contents
Acronyms i

About the Long Duration Energy Storage (LDES) Council ii

Preface iii

Executive summary vi

Data collection and benchmarking xi

1. Introduction 1

2. LDES technologies characterization and current status 7

3. Modeling the flexibility needs of future power systems 15

4. Cost analysis 25

5. LDES business cases 35

6. Road to competitiveness and key market enablers 41

Conclusion 46

Appendix A: Methodology 47

Appendix B: Examples of business cases 51

 Acronyms

BoP Balance of plant LDES Long duration energy storage

Capex Capital expenditure MEDC More economically

developed countries
CCS Carbon capture and storage
MPM McKinsey Power Model
CO2 Carbon dioxide
MW Megawatt
CAES Compressed air energy storage
MWh Megawatt-hour
CSP Concentrated solar power
NDC Nationally determined
EV Electric vehicle
contributions
Gt CO2eq Gigatonnes of carbon
NPV Net present value
dioxide equivalent
NMC Nickel, Manganese and Cobalt
GW Gigawatt
O&M Operation and maintenance
GWh Gigawatt-hour
PV Photovoltaic
GHG Greenhouse gas
PPA Power purchasing agreements
IEA International Energy Agency
PSH Pumped storage hydropower
IRR Internal rate of return
RE Renewable energy
IPCC Intergovernmental Panel
on Climate Change R&D Research and development

kW Kilowatt RTE Round-trip efficiency

kWh Kilowatt-hour TW Terawatt

LCOE Levelized cost of electricity TWh Terawatt-hour

LCOS Levelized cost of storage TAM Total addressable market

Li-ion Lithium-ion T&D Transmission and distribution

LAES Liquid air energy storage WACC Weighted average cost of capital

i Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 About the Long Duration
Energy Storage (LDES) Council

The LDES Council is a global, CEO-led class, power and energy systems and the
organization that strives to accelerate broader energy transition. The Council will also
decarbonization of the energy system at proactively engage with other parties on ways
lowest cost to society by driving innovation to accelerate decarbonization of energy systems
and deployment of long duration energy in line with the Paris agreement.
storage. Launched at COP26, the LDES Council
The following organizations have announced
provides fact-based guidance to governments
the intention of forming the Council and are open
and industry, drawing from the experience of
to receive expressions of interest from additional
its members, which include leading energy
founding members ahead of the official launch in
companies, technology providers, investors,
early 2022 (Exhibit 1).
and end-users.
The report has been prepared by the members
With this first report the Council has focused
of the LDES Council in collaboration with
on the role of LDES solutions in electrical power
McKinsey & Company as knowledge partner.
systems. In the future, the LDES Council will
provide further insights into the LDES asset

Exhibit 1
LDES council members
Technology providers Anchors

Industry and services customers Low-carbon energy system

integrators & developers

Capital providers Equipment manufacturers

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company ii
 Preface

As the world considers how to establish a path variability of wind and solar power. This first
towards limiting the rise in global temperatures report from the LDES Council aims to explore
by curbing emissions of greenhouse gases one of the key solutions to this challenge:
(GHG), it is widely recognised that the power long-duration energy storage (LDES).
generation sector has a central role to play.
LDES is defined as any technology that can
Responsible for one third of total emissions, it is
be deployed competitively to store energy for
in fact doubly crucial, since decarbonizing the
prolonged periods and that can be scaled up
rest of the economy depends vitally on growing
economically to sustain electricity provision,
demand for renewable energy, for example in
for multiple hours, days, or even weeks, and
electric vehicles and residential heating. And the
has the potential to significantly contribute to
good news is that the global power industry is
the decarbonization of the economy. Energy
making giant strides towards reducing emissions
storage can be achieved through very different
by switching from fossil-fired generation to wind
approaches, including mechanical, thermal,
and solar power.
electrochemical, or chemical storage (see Box 1).
However, the rising share of renewables in
The provision of flexibility, defined as the ability
the power mix brings with it new challenges.
to absorb and manage fluctuations in demand
Not least of these are the structural strains on
and supply by storing energy at times of surplus
existing power generation infrastructure created
and releasing it when needed, is a critical
by new flows of electricity and by the inherent

Exhibit 2
LDES play a central role in energy system flexibility
LDES use cases Scope of this first report

Power

• H2 peaking plants Power-to-H2 Power-to-heat • CHP for district

• Transport (material H2-to-power Heat-to-power heating/cooling
handling, heavy • Heat pumps/engines
duty vehicles) Desalination
CHP with H2
production and use

• Solid oxide fuel cells /

Hydrogen Heat electrolyzers
H2-to-heat • H2 turbines
• H2 household boilers
• Industrial heat
(furnaces, boilers)

iii Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
enabling factor to decarbonize the economy
in a cost-efficient way. Across the portfolio
of technologies, LDES can provide flexibility
in the energy system as a whole, comprising
power, heat, hydrogen and other forms of
energy (Exhibit 2). For example, some LDES
technologies can discharge both heat and
power (i.e., power-to-heat or heat-to-power) that
can be used to decarbonize industries, or can
use power to produce hydrogen via electrolysis,
which can be reconverted back to power at a
later time. The ability to integrate different sectors
makes some of the technologies unique, and
strengthens the business cases for their use in
decarbonizing industries where the transition is
a challenge.

LDES technologies are attracting unprecedented

interest from governments, utilities, and
transmission operators, and investment in the
sector is rising fast. This report focuses on the
role of novel LDES solutions in electrical power
systems (please refer to Box 1 for more details
on the LDES technologies covered in this
report). It first examines the characteristics of
the technologies and how they may be suited
to help manage structural issues in the power
industry. It then considers LDES costs, how they
may develop as the industry matures, and how
they compare with those of other technologies
that can be used to manage supply and demand
such as Lithium-ion (Li-ion) batteries and
hydrogen. Finally, it proposes some actions
policy makers and industry players can consider
to enable LDES to fulfil its potential as part of the
world’s net-zero solution.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company iv
 What is the How do LDES technologies help?
issue? LDES are a host of different Alongside Li-ion battery technology
technologies that store and and hydrogen, LDES technologies
To avoid catastrophic climate
release energy through mechanical, can play a critical and distinctive role
change, we need to rapidly build thermal, electrochemical, or in delivering flexibility on times
a net-zero power sector chemical means. ranging from hours to weeks.
predominantly powered by
renewable energy.
As the proportion of renewables
grows, we are presented with 3
challenges; balancing electricity
supply and demand; a change
in transmission flow patterns;
and a decrease in system
stability.
LDES can help address these
issues by increasing the
flexibility of the power system.
Where are we today and where
do we need to get to?
Many LDES technologies currently By 2040, LDES need to have scaled
PROPORTION OF exist, but they are at different levels up to ~400x present day levels to
RENEWABLES
of maturity. Some have been 1.5–2.5 TW (85–140 TWh). 10% of
deployed commercially, some are all electricity generated would be
NEED FOR still at the pilot phase. stored in LDES at some point.
FLEXIBILITY

Our projections show that LDES Present-day LDES deployment is

need to be scaled up dramatically low, but momentum in LDES is
over the next 20 years to build a growing exponentially.
Projected installed cost-optimal net-zero energy system.
capacity
1.5–2.5 TW Global deals in the LDES industry, 910 2,640
85–140 TWh USD million

360
260
980 130
0.1–0.4 TW
~0 TW 4–8 TWh
~0 TWh
Today 2030 2040 Pre-2018 2018 2019 2020 2021 Total

How can we make this happen?

For LDES to be cost optimal, costs invested in transmission and
must decrease by 60%. However, distribution networks every 2–4 The value of LDES can be
even greater cost reductions have years. unlocked through
already occurred in other clean regulation change:
technologies like solar and wind. This investment has the potential
to create economic and • Long-term system planning
Between 2022–40, USD 1.5 tr–3.0 tr environmental benefit. The • Support for first deployment
of total investment in LDES will be business cases for LDES can often and scaling up
required. The total investment over be positive if sufficient mechanisms • Market creation
this period is comparable to what is are in place to monetize the value.

v Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Executive summary

The world is not on track to limit the rise in global transmission flow patterns, and the potential for
temperature to 1.5° Celsius. To achieve the greater system instability as the built-in inertia
commitments made in the Paris Agreement, provided by fossil generation is removed. All of
significant efforts must be made to reduce these call for new solutions to create flexibility
emissions across all sectors. The power in electricity supply and demand over different
sector, which accounts for roughly one-third durations — intraday, multiday/multiweek, and
of global emissions, will be central to global seasonal.
decarbonization, with many suggesting that
LDES is one of these solutions, since LDES
it will need to achieve net-zero emissions by
technologies entail low marginal costs for
2040. As a result, innovative solutions will be
storing electricity: they enable decoupling of
essential to meet three critical challenges
the quantity of electricity stored and the speed
for the power sector: tripling the amount of
with which it is taken in or released; they are
electricity produced to meet rising consumption,
widely deployable and scalable; and they have
transforming the power system from fossil-
relatively low lead times compared to upgrading
powered generation to renewables, and meeting
of transmission and distribution (T&D) grids. As
the social and economic cost of the transition.
a result, there is increasing investment interest in
Based on more than 10,000 cost and these technologies, with more than 5 gigawatts
performance data points, this study shows that (GW) and 65 gigawatt-hours (GWh) of LDES
Long Duration Energy Storage technologies announced or already operational.
(LDES) can play a crucial role in helping create
This is only a start: modeling suggests that
the system flexibility and stability required by an
LDES has the potential to deploy 1.5 to 2.5
increasing renewable share in power generation,
terawatts (TW) power capacity—or 8 to 15 times
alongside other technologies such as Lithium-
the total storage capacity deployed today—
ion (Li-ion) batteries and hydrogen turbines.
globally by 2040. Likewise, it could deploy 85 to
LDES encompasses a range of technologies that 140 terawatt-hours (TWh) of energy capacity by
can store electrical energy in various forms for 2040 and store up to 10 percent of all electricity
prolonged periods at a competitive cost and at consumed. This corresponds to a cumulative
scale. These technologies can then discharge investment of USD 1.5 trillion to USD 3 trillion and
electrical energy when needed—over hours, to potential value creation of USD 1.3 trillion by
days, or even weeks—to fulfill long-duration 2040.
system flexibility needs beyond short-duration
The scale of these numbers reflects the multiple
solutions such as Li-ion batteries. The various
use cases for LDES technologies and the
LDES technologies are at different levels of
central role they can play in balancing the power
maturity and market readiness. This report
system and making it more efficient. These
focuses on the relatively nascent mechanical,
include support for system stability, firming
thermal, chemical, and electrochemical storage
corporate power purchase agreements (PPAs)
technologies, instead of Li-ion batteries,
and optimisation of energy for industries with
dispatchable hydrogen assets, and large-scale
remote or unreliable grids. Similarly, there is a lot
aboveground pumped storage hydropower
of potential in using LDES in off-grid systems,
(PSH) (more details about LDES technologies
which have a lower level of flexibility and currently
are provided in Box 1).
rely heavily on fossil fuels. But by far the largest
The rapid integration of large RE capacities with proportion of deployment is expected to be
their inherent variability creates large challenges related to the central tasks of energy shifting,
for the power system, including potential capacity provision, and T&D optimization in bulk
imbalances in supply and demand, changes in power systems.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company vi
 In sum, LDES offers a lower-cost flexibility capacity being installed by 2025 under a fast
solution in many—but not all—situations. decarbonization scenario. A key milestone
A diversified suite of solutions is likely to be for LDES is reached when RE reaches 60 to
deployed in order to achieve a cost-optimal 70 percent market share in bulk power systems,
decarbonization of the grid by 2040. The prize of which countries with high climate ambitions aim
deploying LDES at scale, however, is great. It is to reach between 2025 and 2035. This catalyzes
estimated that by 2040, LDES deployment could widespread deployment of LDES as the
result in the avoidance of 1.5 to 2.3 gigatonnes of lowest-cost flexibility solution.
carbon dioxide equivalent (Gt CO2eq) per year, or
Before these targets are reached, however,
around 10 to 15 percent of today’s power sector
government action will be required to help lower
emissions. In the US alone, LDES could reduce
costs, mobilize the necessary investment and
the overall cost of achieving a fully decarbonized
create market signals enabling investors to
power system by around USD 35 billion annually
make an attractive return on LDES. An enabling
by 2040.
governmental ecosystem would include the
Achieving this order of scale requires significant implementation of (i) long-term system planning,
reductions in the cost of LDES technologies. But (ii) early compensation mechanisms that reduce
projections provided by LDES Council member uncertainty for investors while the market is still
companies show these are achievable and in nascent, and (iii) supportive policies, regulations,
line with learning curves experienced in other and market designs.
nascent energy technologies in the recent past,
Long-term system planning, including clear RE
including solar photovoltaic and wind power.
targets, is critical to creating investor confidence.
In turn, cost reductions will be dependent on
Targeted support for early deployments and
improvements in research and development
scale-up would help kick-start the market and
(R&D), volumes, and scale efficiencies in
trigger the learning curve on costs. Finally,
manufacturing. Similarly, total LDES deployment
supportive market designs such as capacity
is closely tied to the rate of decarbonization of
mechanisms and policies that capture the
the power sector and the deployment of variable
full value of LDES would enable investors to
renewable energy (RE) generation.
monetize their outlays. Together, these measures
While LDES technologies are still nascent, will ultimately help ensure that the energy
deployment could accelerate rapidly in the transition is achieved at the lowest societal cost.
next few years. Modeling projects installation of
30 to 40 GW power capacity and 1 TWh energy

vii Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 Box 1.

LDES technology space of this report

The term LDES is used to encompass a chemical storage. (Exhibit 3)

wide technology family with various levels of
A. Mechanical LDES
technological maturity and market readiness.
While this class does not exclude Lithium-ion The most widespread and mature storage
(Li-ion), hydrogen turbines, or large-scale, technology is PSH, a form of mechanical storage
aboveground pumped storage hydropower that accounts for 95 percent of the total energy
(PSH), this report focuses on novel technologies storage capacity worldwide. New versions of this
that can fulfill the flexibility space beyond Li-ion established technology are emerging to reduce
batteries and other short-duration solutions. its dependence on geographical conditions, for
These technologies are herein referred to as example, geomechanical pumped hydro, which
“LDES”, and do not include hydrogen, Li-ion, or uses the same principles as aboveground PSH
large-scale aboveground PSH. but with subsurface water reservoirs.

Novel LDES can be broadly classified into: Other emerging mechanical energy storage
mechanical, thermal, electrochemical, and solutions include compressed air energy

Exhibit 3
Overview of LDES categories

There are 4 kinds of

Mechanical
novel LDES Mechanical LDES store potential or kinetic
energy in systems for future use.

All LDES allow energy to be stored E.g., raising a weight with surplus energy and
then dropping it when energy is needed.
when there is a generation surplus
and released when there is a shortage. • Novel PSH
• Gravity based
• CAES

Thermal • LAES
• Liquid CO2
Thermal energy storage systems use thermal
energy to store and release electricity and heat.
E.g., heating a solid or liquid medium and then
using this heat to power generators at a later
date.
• Sensible heat
• Latent heat
• Thermochemical heat

Chemical
Electrochemical Chemical energy storage systems store electricity
through the creation of chemical bonds.
Electrochemical LDES refers to batteries of
E.g., using power to create syngases, which can
different chemistries that store energy.
subsequently be used to generate power.
E.g., air-metal batteries or electrochemical
• Power-to-gas-to-power
flow batteries.
• Aqueous flow batteries
• Metal anode batteries
• Hybrid flow batteries

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company viii
 storage (CAES) and gravity-based energy including resistance heaters, heat engines, or
storage. The first stores energy as compressed high temperature heat pumps among others.
air in pressure-regulated structures (either
The most widespread thermal LDES technology
underground or aboveground). In its adiabatic
are molten salts coupled with concentrated solar
form, CAES also includes thermal storage
power (CSP) plants, however, this technology
to store the heat that is generated during
is different from other novel LDES as it presents
compression and reuse it in the discharge
different characteristics (e.g. it cannot be widely
cycle. Gravity-based energy storage is another
deployed as it is not modular, the CSP plant has
promising form of mechanical storage, which
a large footprint and is only effective in regions
stores energy by lifting mass that is released
with high solar radiation). Nonetheless, molten
when energy is needed. This technology is in
salts can effectively be used in novel thermal
an earlier stage of commercial development.
LDES to store electricity independently of CSP
Lastly, mechanical LDES can also take the plants.
form of liquid CO2, which can be stored at high
Thermal LDES technologies can discharge
pressure and ambient temperature and then
both electricity and heat, supporting the
released in a turbine in a closed loop without
decarbonization of the heat sector, which
emissions.
accounts for around 50 percent of the global
Liquid air energy storage (LAES) works similarly final energy consumption (compared to
to CAES by compressing air but uses electricity 20 percent by the electricity sector in 2019). Of
to cool and liquify the medium and store it in the total heat consumption, it is estimated that
cryogenic storage tanks at low pressure. For only around 10 percent is supplied by RE.1 LDES
this reason, LAES is sometimes classified as could support the decarbonization of this sector
mechanical storage and sometimes as thermal through the provision of zero-emissions high-
storage. grade heat to energy-intensive industries—that
rely on fossil fuels and have few decarbonization
B. Thermal LDES
alternatives—and other heat applications (such
Thermal energy storage technologies store as district heating networks).
electricity or heat in the form of thermal energy.
C. Chemical LDES
In the discharge cycle the heat is transferred
to a fluid, which is then used to power a heat Chemical energy storage systems store
engine and discharge the electricity back to electricity through the creation of chemical
the system. Depending on the principle used bonds. The two most popular emerging
to store the heat, thermal energy storage can technologies are based on power-to-gas
be classified into sensible heat (increasing the concepts: power-to-hydrogen-to-power, and
temperature of a solid or liquid medium), latent power-to-syngas (synthetic gas)-to-power.
heat (changing the phase of a material), or
In the first case, electricity is used to power
thermochemical heat (underpinning endothermic
electrolyzers, which produce hydrogen
and exothermic reactions). These technologies
molecules that can be stored in tanks, caverns,
use different mediums to store the heat such
or pipelines. The energy is discharged when the
as molten salts, concrete, aluminum alloy, or
hydrogen is supplied to a hydrogen turbine or
rock material in insulated containers. Likewise,
fuel cell. If the hydrogen is combined with CO2
the charging equipment options are diverse,
in a second step to make methane, the resulting

1
“Renewables 2020,” IEA, 2020.

ix Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
gas—known as syngas—has similar properties There are also hybrid flow batteries with liquid
to natural gas and can be stored and later electrolytes and a metal anode which combine
burned in conventional power plants. Similarly, some of the properties of conventional flow
hydrogen can be converted to ammonia for batteries and metal-anode systems.
direct combustion.
Li-ion, hydrogen turbines, and large-scale
D. Electrochemical LDES aboveground PSH

Different batteries of varying chemistries are This report distinguishes between LDES
emerging to provide long duration flexibility. and Li-ion as the scaling up of costs for a
long-duration flexibility range makes Li-ion
Electrochemical flow batteries store electricity
uncompetitive for a long-duration flexibility
in two chemical solutions that are stored in
range.
external tanks and pushed through a stack
of electrochemical cells, where charge and Hydrogen-based storage and reconversion to
discharge processes take place thanks to a power via turbines (and fuel cells) can serve a
selective membrane. These batteries are suited role for long-duration storage but are called out
for long-duration applications where separately in the report due to dissimilar cost
low chemical and equipment costs are possible. performance at lower storage durations2 and
the specific interest that has evolved around
Emerging metal air batteries rely on low-cost,
hydrogen in the energy community.
abundant earth metals, water, and air – meaning
they have the potential for high scalability and Large-scale aboveground PSH are not included
low installed system costs. Furthermore, many in the considered technology space as the
of these solutions do not suffer from thermal deployment benefits and economics of novel
runaway, making them safe to install and LDES technologies, including novel PSH, are
operate. expected to outcompete these plants and LDES
have fewer geographical limitations.

2
Where a hydrogen technology demonstrates very similar behaviors and cost profiles to other LDES it has been included
(such as solid oxide fuel cells).

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company x
 Data collection and
benchmarking

The data used in the analysis of this report was median figures were processed, yielding eight
collected from the LDES Council members, finalized data sets: first quartile and median,
who submitted a total of more than 10,000 data for 8 to 24 hours and 24 hours or more in
points outlining the cost and performance of central and progressive scenarios. The created
their technology. The data was aggregated and archetypes were used as inputs to model the
processed by an independent third-party clean total addressable market (TAM) and to generate
team. insights on cost competitiveness with alternative
technologies presented in this report. Future
Council members provided cost and
iterations of the analysis aim to incorporate more
performance data for two projected trajectories
data points per technology type, allowing for a
for how these metrics would change from a
disaggregate analysis for each LDES category
“progressive” to a “central” scenario:
(mechanical, thermal, electrochemical, and
• Progressive scenario: council data reflecting chemical) and duration archetype.
ambitious cost-reduction trajectories and
The technology benchmarking in the report
learning rates
builds on the McKinsey Power Model (MPM),
• Central scenario: council data reflecting McKinsey Battery Cost Model, McKinsey Energy
conservative cost-reduction trajectories and Insights modeling of RE costs and capacity
learning rates factors, other proprietary assets, and numerous
benchmarks from external data providers and
The data was grouped into two archetypes
databases. The analytics team also tested
based on their nominal duration: 8 to 24 hours3
the findings from these analyses with experts
and 24 hours or more, with some members
outside the Council and with individual Council
offering products in both ranges. For every
members, who provided industry expertise.
archetype, aggregated data points for each
cost, design, or performance metric created
representative numbers while preserving the
data confidentiality of each individual technology.
After the data point aggregation, top-quartile and

3
The 8-hour threshold does not imply that LDES is not expected to provide services below this duration.

xi Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 1.
Introduction

Chapter summary

LDES can have a role to play in increasing

power system flexibility, which will
be crucial to achieve net-zero
The decarbonization of power systems by 2040 • Intraday flexibility (<12 continuous hours4)
will be essential to achieve net zero economies
• Multiday and multiweek flexibility (12 hours3
and limit the rise in global temperatures to 1.5°
– weeks)
Celsius
• Seasonal flexibility
High renewable penetration will have an impact
on the reliability and stability of the power • Flexibility to respond to extreme weather
system. To fully decarbonize the power sector, events
three key challenges need to be overcome:
Whilst solutions exist today, they are either
• Power supply and demand imbalances carbon emitting (such as gas plants), physically
constrained (such as large-scale aboveground
• Change in transmission flow patterns
pumped storage hydropower, or PSH) or are
• Decrease of system inertia not cost effective for addressing all future
needs of the power system (such as Lithium-ion
These three challenges are solvable by
batteries). To achieve a cost-effective energy
introducing flexibility into the power sector
transition, long duration energy storage (LDES)
across different time spans:
technologies are required.

4
Assumes symmetrical design of the charge and discharge durations, which is not the case for all LDES systems. The optimal
design of LDES systems for the provision of intraday flexibility would be case-specific and can comprehend durations above and
below 12 hours.

1 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
The decarbonization of power To reach a 1.5° Celsius decarbonization pathway,
systems by 2040 will be essential this study assumes that the global power sector
to achieve net-zero economies and
will need to achieve net-zero emissions by
limit the rise in global temperatures
2040 (Exhibit 4). To achieve such a target, it is
to 1.5° Celsius
assumed that more economically developed
Accounting for roughly 80 percent of global GDP, countries (MEDCs) achieve net-zero emissions
123 countries have pledged to achieve net-zero by 2035 and the rest of the world by 2040. This
greenhouse gas (GHG) emissions and/or carbon milestone is consistent with the most recent
neutrality by 2050.5 However, current efforts are net-zero report from the International Energy
insufficient to achieve such targets and move the Agency (IEA).
world onto the 1.5° pathway set out in the Paris
The enabling low-carbon power-generation
Agreement. Human activity has already led to a
technologies are already available at scale.
rise in global temperatures.6 Economies are not
In many instances, they can be deployed at a
on track to reduce their emissions rate, which
lower cost of generation than thermal sources,
is rising again after a brief dip caused by the
allowing the power sector—including large,
Covid-19 pandemic and is expected to rise
interconnected networks, isolated grids, and
further in the coming years. As a result there
mini-grids—to decarbonize ahead of other
is a growing risk of severe climate change
sectors.
in the coming years and decades with
convulsive environmental and socioeconomic
Power systems will have to rapidly
consequences.
accommodate large amounts of
To meet climate targets and limit the impact of renewable energy (RE), which will
climate change, immediate action is required. pose new system challenges
An ambitious combination of solutions is
To limit carbon emissions, power generation will
needed to achieve the necessary GHG
have to accelerate its transition to RE. The falling
emission reductions, including aggressive
LCOEs (levelized cost of electricity) of RE are
decarbonization rates and systemic changes in
already accelerating the adoption of wind and
energy supply across all sectors.7
solar as existing plants retire, and power demand
The power sector is among the largest emitters grows—even in the absence of policy support.
of GHGs, and its decarbonization is crucial If governments adopt strong policies and create
to establishing the pathway towards a net- appropriate market designs, the transition
zero economy by 2050. Electricity generation could be accelerated. While negative carbon
worldwide was responsible for emissions of emissions solutions will be critical to achieving
12.3 gigatonnes of carbon dioxide equivalent full decarbonization of economies, their impact
(Gt CO2eq) in 2020,8 around a third of total by 2040 will be limited and largely influenced by
emissions. Demand for electric power is tailwinds supporting their scale-up, including the
growing, driven by increased electrification availability of appropriate carbon dioxide (CO2)
across multiple end uses, for example, by transportation and storage infrastructure and
electric vehicles (EVs) and residential heating. social acceptability.
New sources of demand are linked to the
The rapid integration of large RE capacities in the
integration of energy-consuming and supply
system—with estimated annual wind and solar
sectors (what is known as “sector coupling”),
photovoltaic (PV) capacity additions of more than
increased population, and higher living
1 terawatts (TW) by 2030 in the electricity sector
standards in emerging markets and developing
alone10—entails some challenges for system
economies. In a deep decarbonization scenario,
planners and market players alike, calling for
widespread electrification could cause power
new solutions that help accommodate increased
consumption to triple by 2050.9

5
Net Zero tracker, accessed on 29 October 2021.
6
NASA Goddard Institute for Space Studies.
7
“Climate change 2021: the physical science basis,” IPCC, 2021.
8
“Net zero by 2050, a roadmap for the global energy sector,” IEA, 2021.
9
“Climate math: what a 1.5-degree pathway would take,” McKinsey & Company, 2020.
10
“Net zero by 2050, a roadmap for the global energy sector,” IEA, 2021.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 2
Exhibit 4
Power sector emission reduction pathways
Global historical emissions of the power Historical emissions
sector and assumed reduction pathways
Net-zero 2035 (MEDCs reference case)1
Net-zero 2040 (Global reference case)2
Gt CO2e

15 Required emission abatement in the

power sector with respect to 2019 levels

12 -35%
-55%
9 -75%

0
1990 95 2000 05 10 15 20 25 2030 35 40 2045
Years
Key assumptions
1. Informed by IEA Net Zero 2050 report on more economically developed countries (MEDCs) needs to get to net zero power
by 2035. Consistent with US President Biden climate ambition.
2. Informed by IEA Net Zero 2050 report on the world’s power sector needs to get to net zero by 2040.

amounts of renewable power. (Exhibit 5) the United Nations Intergovernmental Panel on

Climate Change (IPCC), flooding and extreme
Power supply and demand imbalances
precipitation are projected to increase at global
By definition, the addition of renewables to the
warming levels exceeding 1.5° Celsius in nearly
electricity mix creates imbalances in supply
all regions. Similarly, the frequency, duration,
and demand, since the natural fluctuations
and intensity of hot extremes are very likely
in wind and solar PV power do not match
to increase.11 In this context, power systems
fluctuations in power demand. Increased shares
will need to be resilient to prolonged supply
of geographically concentrated wind and solar
disruptions and ensure sufficient firm capacity
power in the generation mix will thus lead to
to guarantee the security of supply in extreme
more frequent periods of power surplus and
weather events.
shortage. In the case of prolonged periods
without sufficient sun or wind, these imbalance Change in transmission flow patterns
periods may last days or even weeks. As a result, Power systems will also see a shift in the
and as RE becomes more common, the grid will geographical supply pattern, and an alteration
need to become more flexible to develop the of transmission line power flows. These
capacity to maintain the supply-and-demand changes result from the increased deployment
balance while incentivizing RE deployment. of decentralized RE generation driven by
technological developments and accelerated
Compounding the challenge, the higher
cost improvements (for example, in residential
frequency of extreme weather events caused by
PV and behind-the-meter batteries). They will
climate change, such as heat waves and heavy
also reflect the geographical dependency of RE
precipitation, will also create more strain on a
capacity, which will tend to be concentrated in
grid dominated by RE generation. For instance,
areas with abundant supplies of sun and wind.
according to the latest Assessment Report by

11
“Climate change 2021: the physical science basis. Contribution of Working Group I to the sixth assessment report of the
intergovernmental panel on climate change,” IPCC, 2021.

3 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 5
A net-zero power system cannot be built without also
developing different types of system flexibility
Shifting to a power system that predominantly relies
on renewable energy presents 3 key challenges …

Power supply and Change in transmission Decrease in

demand imbalances flow patterns system inertia
The supply of electricity Changes in the distribution of the Removing conventional generators from the
from renewables does not energy system can require costly and system also removes the inertia from rotating
always match the demand lengthy developments to transmission lines masses from the system

… to resolve these challenges, flexibility

on different time scales is needed

Intraday Multiday and Multi-month

flexibility multiweek flexibility flexibility
Flexibility that allows daily variations Flexibility that allows day to week long Flexibility that allows seasonal
in supply and demand to be smoothed fluctuations in supply and demand to mismatches in supply and demand to
out (such as peak energy demand in be balanced (such as taking into be managed (such as energy demand
the evening) account weather anomalies) peaks in winter)

This gives LDES technologies an advantage providing

electricity system flexibility between 8 and 150 hours in length

High-cost
Short duration storage

LDES typically the Fully dispatchable assets

Increasing the Short-duration batteries most cost competitive (eg, hydrogen turbines, CCS)
amount of energy (including Li-ion) typically solution for storage potentially the most cost
stored is… the most cost competitive durations between competitive solution1
solution 6-8 and 150 hours

Very long duration storage

Low-cost
Low-cost High-cost
Increasing the power is …

1. Technologies not mature yet (still in commercial demonstration) requiring cost reductions

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 4
 Changes on the consumer side will switch the which use power electronics to set the correct
traditional one-way design of electricity lines to frequency (“artificial inertia”) and synchronous
a two-way system, where an increased number condensers are current technological solutions.
of end users will generate their own electricity
and inject it into the power grid.12 This will pose A net-zero power system will need
challenges to the conventional distribution flexibility resources at different
systems such as voltage control and stability. An duration levels, where long duration
example of this trend can be seen in California, energy storage (LDES) can play a
where public incentives and governmental crucial role
support have led to the deployment of more than A broad range of flexibility levers and enablers
10 gigawatts (GW) of distributed solar generation already exist to help balance RE generation.
(representing 10 percent of its total generation Existing solutions include dispatchable capacity
capacity mix) in the past ten years.13 (for example, gas peakers, or generation plants
that can be activated at times of peak electricity
Similarly, regions with high RE yield potential
use, and pumped hydro), the expansion of
will likely become new generation centers that
transmission grids, including internal and
impact how networks operate. For example,
cross-market interconnections, feed-in
one study showed that historical transmission
management and RE curtailment, as well as
flow patterns in New York State are likely to be
short-duration batteries.
reversed due to increased solar and offshore
wind power injection. Flow directions will also However, these traditional approaches are not
vary over time as RE yield fluctuates throughout an adequate answer to the evolving needs of
the day and year.14 Long lead times and slow the system. The most widespread solution—
grid adaptation to these system changes could gas peakers—emits carbon and requires
result in more frequent congestion, reducing the deployment of carbon capture and storage
stability of the power system and jeopardizing its (CCS). This increases its capital intensity and
ability to meet decarbonization targets. generally requires it to be installed close to a CO2
storage formation. Grid expansion can reduce
Decrease of system inertia
congestion risk but is costly, has long lead times,
The stability of the system is also challenged as
and is unsuitable in some population centers.
the bulk of power generation transitions from
Furthermore, constructing physical infrastructure
synchronous to asynchronous technologies.
to accommodate peak demands tends to have a
Conventional generators (for example, fossil
low return on investment. Feed-in management
fuels and nuclear) have played a crucial role in
and power curtailment are inherently inefficient,
safeguarding the stability of the electricity system
as they result in lost supply. Lastly, short-
through their provision of inertia: in a system
duration energy storage has technical and
disturbance, the rotating machines connected to
economic limitations that mean it cannot meet
the grid help all generators remain synchronized
the full range of flexibility durations required.
by resisting a change in the frequency of the
grid. If unrectified, stability faults can result in As a result, new low-carbon flexibility sources
blackouts with high economic and societal are starting to emerge, including demand-side
costs. response mechanisms, hydrogen dispatchable
plants, and LDES technologies. A diversified
By contrast, newer technologies like solar
suite of solutions is likely to be deployed in order
PV and wind lack rotating masses directly
to achieve a cost-optimal decarbonization of the
connected to the grid and therefore cannot
grid by 2040 (Exhibit 6).
provide inherent system inertia. As a result,
generation disturbances, frequency, and Intraday flexibility
voltage deviations necessitate the installation This covers the need for flexibility for durations
of new stability sources. Grid-forming inverters, below 12 continuous hours15 and generally

12
“Distributed energy resources for net zero: An asset or a hassle to the electricity grid?,” IEA, 2021.
13
California Distributed Generation Statistics.
14
“The global relevance of New York State’s clean-power targets,” McKinsey & Company, 2019.
15
Assumes symmetrical design of the charge and discharge durations, which is not the case for all LDES systems. The optimal design
of LDES systems for the provision of intraday flexibility would be case-specific and can comprehend durations above and below 12
hours.

5 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 6
Summary of existing and emerging flexibility solutions
for different flexibility duration needs
Solution Partial solution

Curtailment
Flexibility Power system Dispatchable Grid rein- or feed-in Li-ion Demand-side
duration challenge generation forcement management batteries LDES response
Intraday Intermittent daily
generation
Reduced grid
stability
Multiday, Multi-day
multiweek imbalances

Grid congestion

Seasonal Seasonal
duration unbalances
Extreme weather
events

involves providing grid stability services and Seasonal flexibility and extreme weather
peak-shifting. Lithium-ion (Li-ion) batteries events
are currently the cheapest zero-emissions The need for seasonal flexibility arises from
option for providing balancing services of less natural variability of solar irradiation, wind
than 4 hours. In the 4- to 8-hour range, other speed, temperature, and rainfall over weeks and
technologies can also accommodate load months, and also from potential exposure to
cycles. These technologies include LDES, extreme weather events. Grid strengthening, RE
demand-side response mechanisms, power oversizing and curtailment, dispatchable assets,
curtailment, and peaking assets. In this range, including hydrogen and biogas with CCS, and
the cost of Li-ion four-hour systems is below natural gas with CCS, could fulfill the/se needs.
USD 400 per kilowatt-hour (kWh) today, and LDES can as well, while also providing resilience
forecasted to decrease to around USD 200 in the face of extreme weather conditions.
per kWh in the next 10 years. With increasing
The set of flexibility needs is likely to evolve
RE shares in the power mix, the need for
following the transition of the power mix. In the
8-to-12-hour flexibility is projected to grow
short term, between now and 2030, as the share
and become a significant market for LDES
of RE remains limited, power systems will mainly
technologies.
require intraday flexibility. Nevertheless, there will
Multiday and multiweek flexibility be local specific applications with high RE shares
This stretches from 12 hours15 to periods and the consequent need for longer durations,
lasting days or weeks. It is needed to address even in the short term. Modeling suggest the
extended periods of imbalanced RE output adoption curve of longer flexibility durations
or potential outages caused by transmission accelerates at levels of RE penetration of 60 to
constraints. Traditionally, the system has relied 70 percent, which will likely be reached in many
on conventional power plants, electricity supply places over the next decade. To achieve global
curtailment, and gradual transmission grid net-zero power by 2040, seasonal flexibility
expansion. LDES technologies are a promising solutions are required to ensure decarbonization
zero-carbon solution for these long-duration in regions with limited potential for a balanced RE
flexibility needs, especially those lasting several portfolio and with limited regional transmission
days. lines.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 6
 2.
LDES technologies
characterization and
current status

Chapter summary

LDES technologies can play a critical

and unique role delivering flexibility on
times ranging from hours to weeks
LDES technologies, like other forms of electricity are modular and do not depend on
storage, allow energy to be stored at times when rare-earth-elements
energy supply exceeds demand and released at
• They have relatively low lead-times compared
times when energy demand exceeds supply
to transmission and distribution (T&D) grid
Novel LDES technologies have distinctive upgrade and expansion
features relative to other forms of electricity
Novel LDES technologies have been deployed
storage:
today:
• The marginal costs of storing additional
• Total investment in major LDES companies
energy are low (i.e., each additional kWh
has reached more than USD 2.5 billion and
of energy stored does not increase cost
has accelerated in the past years
significantly)
• Excluding large-scale aboveground PSH,
• There is decoupling of the quantity of energy
more than 5 GW and 65 gigawatt-hours
an LDES technology can store and the rate
(GWh) of LDES is already operational or has
at which an LDES can uptake and release
been announced. Nevertheless, the majority
energy (i.e., LDES can create a very large
of these deployments are associated with
store of energy with a small stream of energy)
traditional molten salts for concentrated solar
• They are widely deployable and scalable as power (CSP) and compressed air energy
they have few geographical requirements, storage (CAES) technologies

7 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 LDES comprises several technologies, each do not need to stack services to recover the
operating on different storage or physical investment. This results in the low degradation
principles and with different architectures. of their storage capacity, and in the potential to
As a result, it is challenging to provide a reach very long life spans, of around 30 years,
unified perspective of LDES performance before requiring significant upgrades. Some
characteristics. However, some features are LDES technologies also have very low capacity
inherent to LDES technologies and are crucial degradation even at high levels of operation.
for transitioning to a clean grid. (Exhibit 7)
Being as a modular solution, Li-ion batteries
deliver rated power and energy as a bundle
LDES technologies are characterized
precluding the optimal independent scaling
by a low energy storage capacity
capex and by their ability to decouple of power and energy capacities, and limiting
power and energy capacities their ability to provide long-duration services
economically. These technologies can maintain
LDES provide significant benefits in terms of output for prolonged periods by reducing
optimal system sizing and scaling-up costs, discharge rates and derating discharge capacity
including low energy storage capacity capital (i.e., providing less than the rated power), which
expenditure (capex) and decoupling capabilities. is a sub-optimal solution to achieving longer
Bulk energy storage capacity can be scaled storage durations.
up at a low incremental cost while not affecting
Importantly, the charging power of some LDES
the charging and discharging cycle design; in
technologies can be designed independently
other words, systems can be designed for long
of the discharging power, which highlights
durations without the need for additional costly
their versatility and adaptability to ecosystems
power capacity. As a result, these systems can
with different supply and load profiles.16 Some
provide power for long durations and generally
mechanical LDES, for example, are charged

Exhibit 7
LDES key concepts
Power and energy are the key features of LDES

Power capacity of LDES Energy capacity of LDES

The maximum electricity output that can be The maximum amount of electricity the
physically discharged by an LDES system in a LDES system can store (an amount). It
given instant (a flow). It is measured in watts (W) is measured in watts-hour (Wh)

In LDES technologies power and energy capacity is decoupled

Unlike other forms of electricity … which makes it … without significantly

storage, LDES energy capacity cheaper to increase compromising the
can be scaled without scaling the amount of power supply.
up power capacity … electricity stored …

In addition to this, LDES technologies often also have other beneficial features

Projects typically Storage solutions are not Solutions don’t depend

have short lead times geographically limited on rare-earth-materials

16
This is not the case for some electrochemical LDES storage technologies, which have symmetric power charging and
discharging capacities.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 8
 by a compressor and discharged by a turbine, deployment restrictions (Exhibit 8). These
with each process designed independently systems, for example, do not have specific
and with different efficiencies. Moreover, the geographical requirements, such as dams in
asymmetry opens up more possibilities for the case of traditional PSH, and have lower
revenue optimization. For example, technology footprints per installed capacity. Depending on
owners can optimize energy arbitrage by slowly the specific technologies, some can be built
charging overnight when power prices are low underground or very close to populated areas
and discharging energy in a shorter amount of due to their low safety risks.
time when prices are high.
In addition, many technologies have a modular
architecture that allows initial deployment of
Additional operational and
systems at shorter durations or smaller power
deployment benefits of some LDES
technologies can add significant capacities that can be scaled up as the system
value to the system evolves.

LDES can also repower or upsize existing

LDES can offer additional operational and
plants, which will be increasingly relevant as the
deployment advantages, such as shorter lead
presence of RE sites grows. This would optimize
times than grid upgrades and expansion, and
land use and allow RE facilities to leverage grid
fewer large-scale deployment constraints. These
connection permits. Additionally, some LDES
advantages vary by technology, and some must
technologies present opportunities for the
still be demonstrated in pilots and commercial
reutilization of potentially stranded fossil assets.
plants.
For example, gas storage fields can be used for
Shorter lead times than transmission compressed air energy storage (CAES) systems,
and distribution (T&D) grid upgrades and or coal and gas plants can be converted into
expansion thermal storage plants. Thermal LDES solutions
Historically, the connection of new generation can provide additional flexibility by coupling the
plants to constrained grids has been heat and power sectors and supporting the
addressed by upgrading existing lines. Grid decarbonization of end uses that rely on
capacity expansion reduces congestion risk; fossil-based heat.
however, it is a capital-intensive process that
In terms of practicality, several LDES
requires long-term planning. Furthermore, it is
technologies rely on existing supply chains,
becoming increasingly difficult for operators as
most of which use earth-abundant materials
decentralized generation plans proliferate and
available in large quantities globally, both in
as project connections become less certain.
the core technology and the balance of plant
Moreover, the complexity and permitting
(BoP) system. This safeguards against potential
requirements of transmission grid projects cause
future supply chain shortages of certain Li-ion
nearly 20 percent of all projects to be delayed or
technologies, such as nickel, manganese and
canceled.
cobalt (NMC) batteries: more than 65 percent
LDES entails a cost-effective solution for of global cobalt production concentrated in the
transmission optimization, increasing grid Democratic Republic of the Congo.17 However,
utilization and virtual grid capacity while deferring this is not the case for all LDES equipment, as
grid upgrades. LDES technologies have average some use certain scarce metals (for example,
construction times of one year and less onerous vanadium) and electric motors or generators
permitting requirements than grid upgrades. with rare-earth magnetic materials. While these
Similarly, they can be applied to large corridors products do not face supply constraints now,
with multiple sites at capacity, allowing for the there is potential for scarcity in the future.
construction of new RE sites.
Specific characteristics of novel LDES
Widely deployable and scalable technologies can be found in Box 2.
Most emerging LDES technologies have few

17
“Lithium and cobalt – a tale of two commodities,” McKinsey & Company, 2018.

9 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 8
LDES Council technologies benchmarking for different
deployment parameters
Unit footprint Companies with site constraints
m2/MW %

250,000
10,000

75

7,500
1,140 25
500 300 350 500

Average Median Top Traditional Onshore Solar PV CCGT Nuclear Yes No

LDES LDES quartile PSH wind
LDES LDES companies with some
Unit footprint scarce material dependency1
m2/MWh %
95

105
30 25
15
4 7 5
0.1 0.1
Average Median Top Traditional Onshore Solar PV CCGT Nuclear Yes No
LDES LDES quartile PSH wind
LDES
1. Use of vanadium and magnetic materials for electric generators, not experiencing supply constraints now,
but presenting potential scarcity issues.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 10
 Box 2.

Novel LDES technologies present very

different characteristics, making them
suitable for different applications

Exhibit 9
Key LDES storage types and parameters
Energy storage Market Max deployment Max nominal Average RTE1
form Technology readiness size, MW duration, Hours %
Mechanical Novel pumped hydro (PSH) Commercial 10–100 0–15 50–80

Gravity-based Pilot 20–1,000 0–15 70–90

Compressed air (CAES) Commercial 200–500 6–24 40–70

Liquid air (LAES) Pilot (commercial 50–100 10–25 40–70

announced)

Liquid CO2 Pilot 10–500 4–24 70–80

Thermal Sensible heat (eg, molten R&D/pilot 10–500 200 55–90

salts, rock material, concrete)

Latent heat (eg, aluminum Commercial 10–100 25–100 20–50

alloy)

Thermochemical heat (eg, R&D na na na

zeolites, silica gel)

Chemical Power-to-gas-(incl. hydrogen, Pilot (commercial 10–100 500–1,000 40–70

syngas)-to-power announced)

Electrochemical Aqueous electrolyte flow Pilot/commercial 10–100 25–100 50–80

batteries

Metal anode batteries R&D/pilot 10–100 50–200 40–70

Hybrid flow battery, with liquid Commercial >100 25–50 55–75

electrolyte and metal anode

1. Power-to-power only. RTEs of systems discharging other forms of energies such as heat can be significantly higher.

Investor interest in LDES has investment in major LDES companies exceeded

increased in recent years, with
USD 2.5 billion in 2021, having nearly tripled in
more than USD 2.5 billion invested
the last four years (Exhibit 10).
in LDES companies

The potential of LDES technologies to More than 5 GW and 65 GWh of LDES

increase the integration of low-cost wind and is already operational or has been
solar resources while reducing the cost of announced
decarbonized power systems has prompted a
Over 260 LDES projects have been announced
surge of new commercial initiatives and research
worldwide at different commercial stages
and development (R&D) efforts. Cumulative

11 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 10
Investment activity in LDES companies
Global deals1 in the LDES industry xx Number of deals
USD millions

102 22 14 27 18 183

910 2,640

+90% p.a.

360

260
130
980

Pre-2018 18 19 2020 2021 Total

1. Based on public investments, VC, PE, corporate, and debt investments of 25 major LDES companies.

(Exhibit 11).18 These projects total 5 GW and batteries account for the highest number of
65 GWh, with roughly 230 projects and projects (over 100), but their average announced
75 percent of the capacity already contrac- capacity is significantly lower at around 4 MW.
ted, under construction, or operational. This means that, while the potential of other
This capacity does not include large-scale LDES technologies is high, their widespread
aboveground PSH projects, which represent adoption is dependent on their commercial
more than 95 percent of all LDES capacity demonstration and cost developments.
installed globally today (for more information
The US, Spain, and Germany have the
on PSH refer to Box 3).
largest reported capacities and projects in
However, the majority of the capacity is terms of regions. The capacity in the US is
associated with traditional molten salts balanced between mechanical, thermal, and
and CAES technologies, which have some electrochemical projects, accounting for roughly
deployment limitations compared to novel 30 percent of global capacity. Most LDES
LDES (such as their large footprint and limited projects in Spain, which account for 20 percent
modularity). Thermal LDES accounts for the of global announcements, are thermal LDES.
largest share of the total announced capacity Germany also has two CAES projects with more
(60 percent), attributable primarily to a number than 200 MW, accounting for 10 percent of
of molten salt storage facilities for concentrated the total announced capacity globally. In Asia,
solar power (CSP) in the megawatt (MW) scale. Japan and China have announced at least 30
Traditional CAES holds the second-largest electrochemical projects, combining both flow
capacity share (around 30 percent) and the and metal anode batteries.
largest average plant size (80 MW). Flow

18
DoE Global Energy Storage Database. The shown figures exclude PSH

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 12
Exhibit 11
LDES project pipeline (excluding PSH)

Category
Mechanical Thermal Chemical Electrochemical
Rated power, MW
≤1 1–10 10–100 100–200 >200

Source: DOE Global Energy Storage Database Transparency denotes projects that have not been built yet

Box 3.

Large-scale aboveground PSH

PSH is a type of hydroelectric energy storage cost per unit of energy. Currently, around 160
that consists of two different elevation water GW of power capacity is installed globally, with
reservoirs that can generate power as water another 130 GW planned or under construction.
flows down from one to the other, passing Future deployments concentrate in Asia, where
through a turbine. Different configurations China accounts for around 60 percent of
of these systems exist, being the most global capacity announced, planned or under
implemented aboveground open-loop PSH and construction, the US, and India. Of the total
closed-loop PSH. The former are connected to capacity, more than 70 percent is associated
a naturally flowing water stream (i.e., on-stream), with closed-loop projects (Exhibit 12). Existing
whereas the latter are not continuously and announced PSH projects generally have
connected to a river (i.e., off-stream). durations ranging from 10 to 24 hours (but in
some cases reaching multiple days), and project
Large-scale, aboveground PSH is the most used
sizes up to 3 GW.19
energy storage solution globally due to its mature
technology, high efficiency, and low capital Total estimated investment in PSH projects

19
International Hydropower Association.

13 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 over the last 10 years is estimated in and as a primary solution for grid stability due to
USD 100-150 billion, with USD 230-320 billion its fast response times. Its major development
more in thepipeline until 2030. System costs constraints are a lack of available sites, long
vary greatly depending on location, which lead times, high construction costs, and
mainly influences EPC costs, and system environmental concerns. Nevertheless, it has
design (including duration of the system and potential to meet increased electrification needs
technology). Global average capex costs and demand for zero-carbon molecules (such
are above USD 2,000 per kW. However, as hydrogen) to decarbonize hard-to-abate
short-duration standalone designs in regions industries, particularly in emerging economies
with very low EPC costs (like India), values below that hold the majority of the untapped natural
USD 1,000 per kW can be reached. potential and whose electricity demand may
triple in the coming years.
Large-scale aboveground PSH has historically
been used for baseload applications, as it
provides low-cost, dispatchable generation,

Exhibit 12
Annual PSH capacity additions by year
PSH capacity additions by year1
GW
South America Middle East Asia Open-loop Closed-loop
North America Europe
35 35

30 30

25 25

20 20

15 15

10 10

5 5

0 0
1980 85 90 95 2000 05 10 15 20 25 30 85 90 95 2000 05 10 15 20 25 30

1. Includes upgrades to existing plants and construction of new plants

Source: DOE; International Hydropower Association

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 14
 3.
Modeling the flexibility
needs of future
power systems

Chapter summary

LDES technologies need to be scaled

dramatically over the next 20 years to
enable a net-zero power system
Modeling shows that in a net-zero scenario, the value in a range different on-grid and off-grid
total addressable market (TAM) for LDES has the applications not accounted in the modeling
potential to reach between 1.5 and 2.5 TW scale and which could increase the cumulative value
by 2040. creation to around USD 1.3 trillion by 2040

Energy shifting, capacity provision and LDES plays a significant role in all modeled
optimization of T&D applications will account scenarios but the precise uptake is sensitive to
for the vast majority of deployments. This is true cost, the performance of alternative technologies
across markets. and to the pace of decarbonization broadly.
Under alternative assumptions, deployments
The estimated value of this market could reach
could be up to 40 percent lower.
over USD 1 trillion by 2040. LDES can create

15 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 LDES is expected to play a significant role in remote and unreliable grid applications, and
achieving cost-effective decarbonization of bulk corporate RE power purchase agreements
power systems and other specialized power (PPAs). However, as bulk power systems achieve
applications. An overview of the projected total high RE penetration (around 60 to 70 percent
addressable market (TAM) for LDES based globally) from 2030 onward, LDES capacity can
on modeling results is herein provided. The accelerate toward the total value of 1.5 to
TAM values outlined below are an outcome on 2.5 TW in 2040 (Exhibit 13). This represents
the cost-optimal net-zero trajectory for power 8 to 15 times the total energy storage capacity
systems and do not account for announced deployed today.
RE government targets or policy measures
In the next five years, significant investment
(more details on the modeling methodology are
will be required to facilitate the widescale
provided in Appendix A). Data ranges refer to the
deployment of LDES and achieve a lower-
central and progressive scenarios.20
cost decarbonization pathway. It is estimated
that by 2025, around USD 50 billion will have
The total addressable market for
to be deployed to install sufficient pilots and
LDES can reach a 1.5 to 2.5 TW scale
by 2040 to achieve the required commercial plants for early decarbonization
flexibility in net-zero power systems while enabling cost-reduction trajectories.
This funding could come from private sources
Based on the projected cost trajectories, combined with a level of public support. Overall,
modeling results suggest that LDES will play the cumulative investment needed to realize
a leading role in providing flexibility as power deployments through 2040 is expected to reach
systems approach net zero. USD 1.5 trillion to USD 3 trillion globally. While
this is striking, this figure is comparable to what
LDES TAM can see initial deployment at
is invested in T&D networks every 2 to 4 years.
scale from 2025 (30-40 GW, 1 TWh, or 6 to
8 times the current announced capacity), with
LDES can create value in a range
accelerated growth toward 2030 (150 to
of different on-grid and off-grid
400 TW and 5 to 10 TWh) as RE penetration applications
of the energy system continues. In 2025, more
than 95 percent deployment will be driven by LDES’ projected technological and economic
non-bulk grid applications such as island grids, features allow them to serve a wide variety of end

Exhibit 13
LDES total addressable market and cumulative capex investment by year
~1,500–2,500
GW
~ 900–1,700
Cumulative installed
power capacity1
~150–400
~30–40
2025 2030 2035 2040

TWh
Cumulative installed ~1 ~5–10 ~35–70 ~85–140
energy capacity1

Cumulative capex ~1,100– ~1,500–

~50 ~200–500
investment1, USD bn 1,800 3,000

1. Range is LDES central scenario and LDES progressive scenario.

20
Central scenario: assumes first-quartile costs for LDES, conservative learning rates, and new-build nuclear capped at previous
peak, and retired as planned. Progressive scenario: assumes first-quartile costs for LDES, aggressive learning rates, no new-build
nuclear, and retired as planned.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 16
 uses. Five main value-creation segments have are briefly described below, while the following
been identified (Exhibit 14), including: section provides in-depth explanations of
energy shifting, capacity provision and T&D
• Energy shifting, capacity provision, and T&D
optimization.
optimization
Optimization of energy for industries
• Optimization of energy for industries with
with remote or unreliable grids
remote or unreliable grids
LDES can become crucial to enabling onsite
• Isolated island grid optimization RE and ensuring continuous power supply
where it is a requisite (for example, in continuous
• Firming for PPAs
manufacturing lines). Relevant end users that
• Stability services provision may need a clean, reliable, and cost-effective
power supply include large off-grid users (like
Energy shifting, capacity provision, and T&D
mines, agribusinesses and military bases)
optimization in bulk power systems are projected
and industrial users in locations with low grid
to result in the largest proportion of deployment
reliability (like chemical and steel plants in less
(80 to 90 percent in 2040); however, the other
economically developed countries). In these
applications can also add significant value
cases, LDES would have advantages over grid
while ensuring full decarbonization of the power
expansion in terms of shorter lead times and
system (Exhibit 15). Additionally, it is projected
fewer geographical constraints.
that early market development in 2025 will be
driven by supply optimization for industries with In total, cumulative LDES capacity deployed for
remote/off-grid or unreliable grid grids (50 GW), the relevant applications could amount to 60 GW
firming for PPAs (30 GW), and isolated island grid and 1.5 TWh by 2030 and 110 GW and around
optimization (15 GW). The different applications 4 TWh by 2040. The value created by LDES—

Exhibit 14
Overview of LDES applications
Energy shifting, capacity
provision, and T&D optimization ENERGY SHIFTING
ENERGY SHIFTING

PEAK SOLAR
Peak generation from renewables does not align with peak demand GENERATION

for electricity. LDES can play a role in shifting electricity from times
of high supply to times of high demand.
MORNING DEMAND EVENING DEMAND

Alongside this, LDES provides

ND
value in other ways, for example …
LA
IN Supporting island grids
A
M Power systems that are not
connected to a large grid can use
LDES to generate reliable power
(e.g., a power grid on a small island).

Providing stability services

Electricity grids require stability.
LDES can be used to correct
instabilities (e.g., transmission
outages can be rectified by LDES).
Supporting industries with
Firming for RE PPAs remote or unreliable grids
Renewable power-purchase Large power users can use LDES to
agreements can use LDES to ensure ensure reliable power in areas where
that businesses can procure 100% they are isolated from the grid or the
renewable electricity. grid is unreliable (e.g., remote heavy
industry).

17 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 15
Total addressable market and cumulative value creation
by application by 2040
xx T&D optimization value xx Value/spend measures
2040
Cumulative LDES
Installed power Installed energy Cumulative Annual LDES
capacity capacity value creation capex spend
Value created by LDES GW TWh USD bn USD bn

Energy shifting, capacity provision, ~4701

~1,300–2,300 ~80–135 ~175–215
and T&D optimization ~300–6502

Optimization of energy for industry

~110 4 ~120 ~4–5
with remote or unreliable grids

Isolated island grids ~90–100 3 ~30–35 ~10

Firming for PPAs ~40 2 ~5–10 <1

Stability services provision (inertia) 0 0 ~5–103 na4

Total ~1,500–2,500 ~85–140 ~950–1,300 ~190–230

Key assumptions
1. Based on reduction in cumulative system cost vs. “No LDES Case.”
2. Value of transmission and distribution expansion deferral or substitution. Figures only account for infrastructure optimization and
do not quantify the value of reduction in generation curtailment costs and reduction of energy not served.
3. Other services are potential material revenue streams for LDES, but not sized in this report.
4. Inertia provided through assets that are deployed for energy generation and capacity provisions, not through additional build-out.

reducing fossil fuel consumption, increasing based power. Furthermore, communities

operational uptime, and replacing fossil connected to weak power systems could also
generation backup capacity—could total USD benefit from LDES inertia provision and other
20 billion to USD 30 billion by 2030 and around services.
USD 120 billion by 2040.21 The impact of climate
By 2030, the cumulative installed capacity for
change, including increased wildfire risk and
isolated islands could amount to 15 GW and
its effect on grid reliability or corporate targets,
150 GWh; by 2040, this could increase to 90 to
could further accelerate adoption.
100 GW and around 3 TWh of installed capacity.
Isolated island grid optimization The potential value creation of LDES arises
LDES can support the stabilization and security from cost savings on fossil fuels and carbon
of the supply of off-grid or microgrid facilities, emissions, totaling up to USD 30 billion by
including island power systems. For instance, 2040.22 Islands with accelerated decarbonization
these technologies could help decarbonize pathways or higher carbon prices could increase
islands and remote communities by minimizing the deployment of LDES and create more value
their dependency on diesel engines and fossil- for these systems. The value of LDES in inland

21
To estimate the LDES market size, different off-grid and backup LDES value propositions were identified with specific industrial and
geographic scope, with each proposition sized following a tailored analysis. The duration for each application depends strongly on
the specific use case and the geographic characteristics.
22
For the market sizing, the most relevant islands (with a population of 0.1 million to 5 million) were identified and their energy storage
needs estimated based on an in-depth analysis of particular case geographies to assess total LDES deployment. The sizing
assumes a lowest cost pathway to decarbonizing island grids by 2040, implying a modular buildout of both Li-ion batteries and
LDES to fulfill storage needs.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 18
 off-grid or isolated communities could have great of LDES for conventional power plants is
potential as well, especially in developing nations that LDES can provide inertia while ensuring
where electricity needs are still either partially or 100 percent RE supply. Furthermore,
fully unmet or depend on diesel generation. mechanical and thermal LDES technologies can
also offer inertia without grid-forming inverters,
Firming for RE power purchase
which would raise the system’s total cost.
agreements
LDES allows for securing premium PPAs with a Suitable LDES technologies can capture value
particular baseload requirement. Both private from inertia and stack it with other remunerated
and public organizations are increasingly services such as capacity provision. The total
interested in using RE to supply their electricity value created from inertia accessible to LDES
as a means to reduce operational costs, hedge is estimated at USD 0.5 billion by 2030 and
against volatile fossil fuel prices and CO2 costs, USD 5 billion to USD 10 billion globally by 2040,
and achieve corporate environmental targets. considering the costs of the next cheapest
Businesses with ambitious pledges to reduce alternative (that is synchronous condensers
carbon emissions typically rely on RE guarantees combined with flywheels). However, it is unlikely
of origin (GOs)—commonly integrated into that the inertia and stability services will ever
PPAs—to source zero-emission electricity. justify the installation of LDES alone. On a
However, RE PPAs are often insufficient to freestanding basis, synchronous condensers are
decarbonize their total consumption; hence the more cost-effective inertia solution.
businesses frequently offset the remaining
Grid systems with limited interconnections
emissions with carbon credits purchased
are expected to be of particular interest at the
in voluntary carbon markets. LDES enables
beginning of the market, as they have fewer
companies to increase their actual RE supply
alternative sources of grid stability. Pilots for this
to near 100 percent while providing resiliency
service have already commenced: for example,
to operations. In the same way, utilities can use
in the UK, a six-year tender for inertia provision
LDES to offer such 100 percent RE PPAs to their
was contracted in 2020.
customers.

By 2025, the global cumulative deployment of Deep-dive: Energy shifting, capacity

LDES for firming RE PPAs could total 10 GW and provision, and T&D optimization
0.5 TWh, rising to around 40 GW and 2 TWh
LDES are expected to play a unique dual role in
by 2040 and generating up to USD 10 billion
bulk power systems, avoiding the need to use
in cumulative value in cost savings on RE GOs
hydrogen turbines for peaking capacity while
and carbon credits. This application should be
also serving intra- and multiday cycling needs.
primarily viewed as a near-term opportunity,
During summer and winter demand peaks,
as RE penetration in bulk grids will increase
LDES can discharge energy over several days
significantly beyond 2030 to provide 24/7 RE
to provide critical clean energy and capacity
coverage. As a result, companies’ willingness to
reserve; during shoulder seasons, LDES could
pay premiums for storage for firming RE PPAs
primarily perform intraday and multiday energy
will likely decline.
shifting. In the very long duration ranges, at
To ensure near 100 percent RE supply, durations presently projected system costs, a mix of
needed for this application are expected to hydrogen turbines and LDES will likely be cost
be above 24 hours. Nevertheless, required optimal. Nevertheless, more rapidly reducing
durations will be dependent on the existing costs or slower hydrogen cost reductions would
capacity mix of the grid. influence the capacity mix.

Stability services provision (inertia or Regarding the TAM, energy shifting and firm
synthetic inertia) capacity provision in RE-intense power systems
LDES technologies can provide a wide range will be the largest market for LDES, accounting
of ancillary services to maintain grid stability for 80 to 90 percent of deployed volumes
(exact services vary by technology). One of those in 2040. T&D expansion optimization could
services is inertia, which is growing in demand generate an additional cumulative value of
as RE penetration grows. A differentiating feature between USD 300 billion and USD 650 billion by

19 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 2040, primarily through the complementation, and reaching higher shares than these two
deferral, or substitution of the distribution technologies (Exhibit 16).
network, where investments are higher.
A balanced mix of flexibility durations for
Reducing curtailment and energy not served LDES will be necessary through 2040
could increase the value pool further. LDES also Given the lower RE share of the total generation
has the potential to provide distributed capacity mix, the largest share of flexibility needs before
to meet local needs while also providing a cost- the end of this decade are likely to fall on shorter
effective alternative to lengthy T&D lead times. durations below 24 hours, providing intra- and
While not currently accounted in the TAM, interday cycling. Nevertheless, early deployment
distributed thermal LDES applications could be in the 24-hour or more range will also be driven
especially attractive where heating needs are by local conditions and specific applications
also present given the high energy losses of heat (such as backup in low grid reliability regions or
transport. high availability corporate PPAs). Both duration
archetypes are likely to see commercial demand
If cost projections unfold as projected, LDES
in the near future.
could account for a large share of countries’
capacity mix. For instance, in the US, LDES By 2030, the projected deployed capacity of
could store around 10 to 15 percent of total the 8-to-24-hour archetype23 could account
energy consumed by 2040, displacing for more than 80 percent of total LDES power
some Li-ion and hydrogen turbine capacity capacity and more than 60 percent of total LDES

Exhibit 16
Projected capacity mix in the US under a net zero 2040 trajectory and
different cost development scenarios
Capacity mix in the US under different scenarios1
Other Gas Coal
GW
No LDES LDES central scenario LDES progressive scenario
5,000
4,510 LDES 24+
4,500 4,380 600
4,230
440 LDES 8-24
4,000 3,760 3,810 Li-ion battery
3,630
3,500 Wind offshore
3,000 Wind onshore
2,500
2,000 1,860 1,860 1,870

1,500
1,130 1,140 1,130 1,140 1,130 1,250 Solar
1,000 H2 turbines
500 Hydro
Nuclear
0
2019 25 30 35 2040 2019 25 30 35 2040 2019 25 30 35 2040

Key assumptions
1. Two LDES archetypes were designed, one with 8–24 hours duration and one with 24–150 hours duration. The LDES central
scenario is based on 1st-quartile cost data and conservative learning rate trajectories, while the LDES progressive scenario is
based on 1st-quartile cost data and aggressive learning rate trajectories.
̶ New nuclear capacity is only allowed to be built in the No LDES and LDES central scenarios, but capped at 50 GW by 2040.
Existing capacity is assumed to be retired according to schedule.
̶ Gas turbines are allowed to be built in the No LDES and LDES central scenarios, but no biomethane or H2 co-firing is allowed.
̶ H2 turbines are allowed to be built in all scenarios, including gas turbine retrofits and new-build H2 capacity.

23
The 8-hour threshold does not imply that LDES is not expected to provide services below this duration.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 20
 energy capacity. LDES technologies that offer Regions with abundant RE resources and high
more than 24-hour flexibility could see significant solar penetration throughout the year, such as
growth after 2030, owing primarily to an increase Australia and Chile, could mainly require shorter
in RE. Longer duration LDES technologies durations for bulk power services.
could account for roughly 80 percent of total
LDES demand in emerging markets will be driven
cumulative energy capacity by 2040. Required
not only by the replacement of fossil-based
investments for different duration installations
assets with RE, but also by increased electricity
are expected to follow a similar pattern to power
demand, which is expected to rise significantly
capacity deployment due to the greater weight of
in the coming years. LDES’s projected TAM in
this component in the total overall system costs
India is 125 GW to 250 GW and 15 TWh to
(Exhibit 17).
25 TWh by 2040, with average installed durations
RE growth and electrification could lead in the 100-hour scale. However, systems will be
to increased demand for LDES systems providing the full range of flexibility durations,
across all markets including intraday and multiday, with shorter
LDES has the potential to support the cost- durations greatly demanded in the short-term as
optimal decarbonization of bulk power across the RE capacity ramps-up.
all markets (Exhibit 18). The US shows the
Policy measures and government targets
greatest need for LDES systems among the
could influence deployment pace and result in
modeled locations, mainly due to limited
an earlier rollout than projected. For example,
transmission connections across the country.
India’s target of deploying 450 GW of RE by
In this market, LDES would help in reducing
2030 could result in a high demand for energy
curtailment and congestion, while increasing
storage capacity before the end of this decade,
transmission utilization. Demand in Europe and
accelerating LDES deployment. Similarly, the US’
Japan could primarily be driven by peaking
new commitment to zero-emissions electricity
capacity from 2035 to 2040, with longer average
by 2035, as well as China’s target of 1,200 GW of
durations (above 50 hours) being installed.
RE by 2030, could have a positive impact as well.

Exhibit 17
LDES total addressable market for the different archetypes
Duration of system 8–24 hour 24+ hour

~150–400 ~900–1,700 ~1,500–2,500

GW 20% 30% 40%
Cumulative installed
power capacity1
80% 70% 60%

~5–10 ~35–70 ~85–140

TWh
Cumulative installed 40%
energy capacity1 65% 80%
60%
35% 20%

~200–500 ~1,100–1,800 ~1,500–3,000

Cumulative capex 20%
investment1, USD bn 35% 50%

80% 65% 50%

2030 2035 2040

1. Range is LDES central scenario and LDES progressive scenario.

21 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 18
Total addressable market by modeled markets
Before 2030 2030-40
Cumulative LDES installed power Cumulative LDES installed energy Average installed
capacity capacity duration
Modeled regions GW TWh Hours
2030 2040

US 440–600 30–40 15–20 70–75

Europe 140–290 5–20 20–30 50–60

India 125–250 15–25 8–10 95–130

Japan 40–80 1–5 14 35–90

Australia 20–40 0.5–1 15 25

Chile 10–15 0–0.5 10–15 18

Extrapolation to RoW 1–230 490–840 0–5 20–40 14 63

Total 1,300–2,300 ~80–135 14 64

The TAM is most sensitive to cost

hydrogen deployment (120 to 250 GW in the
and performance, alternative
US by 2040). If the round-trip efficiency (RTE)
technologies, and decarbonization
developments of systems falling in the 8-to-24-hour archetype
do not exceed 70 percent, Li-ion could take
The future of power markets is by definition approximately 65 GW of LDES deployed. On
uncertain. Commitments and actions by public the other hand, the impact on longer durations
and private players, new market designs, and would be minimal, with only 15 GW being
technological developments, are all highly displaced by hydrogen.
interconnected and will ultimately determine
Deviations in the cost projections of alternatives
whether climate targets are met. Similar
could significantly impact LDES adoption.
unpredictability surrounds LDES, which in
If hydrogen costs decrease (for example, if
addition carries technology maturity risk.
hydrogen storage in salt caverns increases), an
Projections for LDES deployment are thus highly additional 90 TWh of hydrogen-based energy
sensitive to different assumptions, as shown could be generated, displacing more than
in Exhibit 19 (the figures are for the US market, 170 GW of LDES capacity. Nevertheless, this
but behaviors are representative of the rest of is expected to have its limitations, given that
the world). lower-cost hydrogen requires infrastructure
or geographical conditions that may be highly
The projected TAM is most sensitive to weaker
constrained. A more aggressive Li-ion cost
than projected LDES cost and performance
scenario would replace roughly 40 GW of shorter
developments. If companies only meet average
duration LDES systems (namely the 8-to-24-hour
capex cost-reduction trajectories, the take-up
archetype). On the contrary, slower Li-ion cost
of LDES could be reduced by further Li-ion and

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 22
Exhibit 19
Sensitivities to US bulk power market size to the variation of
different parameters
2040, GW x Bulk power total addressable market variance to central scenario 8–24 hour 24+ hour

Changes in model Variance to central scenario in the US Result observed

Weaker cost and performance for 8–24 +50% +210 GW Li-ion storage; replacing 8–24 hour
hour archetype (median assumed instead -150 -250 100 archetype as Li-ion becomes more
of 1st quartile) -100% competitive in short-durations

Weaker cost and performance for 24+ -65% +50 GW of H 2 turbines; LDES less
hour archetype (median assumed instead -5 -120 115
+45% competitive in firm capacity provision
of 1st quartile)

Weaker cost and performance for both +1% +260 GW Li-ion storage and +70 GW H2
archetypes (median assumed instead of -245 -250 <5 turbines; LDES 24+ hour archetype energy
1st quartile) -100% capacity halved

Additional contingency costs -90% +220 GW of H 2 capacity and 320 GW of

-420 -250 -170
for pre-commercial technology1 -100% Li-ion; LDES not competitive at this cost

Weaker RTE improvement for 8–24 hour 20 +10% +67 GW Li-ion storage; Li-ion more
-45
archetype (frozen at 2025 level, ie, 70%) -65 -25% competitive in short-duration energy shifting

Weaker RTE improvement for 24+ hour 20 -10% Minimal change, as no other competitive
5
archetype (frozen at 2025 level, ie, 50%) -15 +10% alternative in long-duration energy shifting

1. Project contingency spend for technologies that are in early stages of commercialization. Assumes 30% increase for all capex costs
for technologies without a pilot plant, and 15% increase for technologies currently with pilot plants (based on Rubin et al. 2013).

reductions would increase the market size

of LDES below 24 hours by roughly 50 GW
(Exhibit 20).24

Lastly, LDES deployment is closely tied to the

rate of decarbonization and deployment of
variable RE generation. A slower transition to
net zero in power, say by 2050, or a 90 percent
reduction in emissions by 2040, could see only
25 to 40 percent of the 1.5 to 2.5 TW power
capacity and 85 to 140 TWh energy capacity
deployed in 2040; although in this scenario
installation would be likely deferred rather than
eliminated altogether.

24
All cases use central scenario assumptions, and test sensitivity to one cost axis at a time.

23 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 20
Capacity mix by flexibility technology under different cost sensitivities
LDES1 Li-ion2 H2 turbine3 Total installed amount of
flexibility capacity in the US
2040, GW

Low-cost LDES 68% 9% 23% 790

High-cost Li-ion 65% 7% 28% 745

H2 High-cost H2 56% 17% 27% 785

Central-cost LDES 56% 17% 27% 785

Low-cost Li-ion 50% 24% 26% 815

H2 Low-cost H2 34% 19% 47% 800

High-cost LDES 23% 45% 32% 870

Key assumptions
1. LDES: low cost represents 1st quartile cost data and fast learning rate cost-reduction scenario; central cost represents 1st
quartile cost data and slow learning rate cost-reduction scenario; high cost presents median cost data and slow learning
rate cost-reduction scenario.
2. Li-ion: high cost is based on McKinsey Battery Cost Model Reference Case; low cost assumes 10% decrease in capex in
all years from McKinsey Battery Cost Model Aggressive Case.
3. Hydrogen: high cost assumes +$1/kg to H2 price due to lower than expected investments; low cost assumes H2 storage in
salt caverns rather than in above ground steel tanks.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 24
 4.
Cost analysis

Chapter summary

Achieving the scales outlined in this report

requires learning rates comparable to other
emerging clean technologies to occur
Novel LDES are nascent technologies that will a limited but growing number of applications.
reduce in cost as they are scaled. The Council The levelized cost of storage (LCOS) analysis
have identified that a large portion of the costs shows that if these learning curves are achieved,
will have learning curves LDES is cost-competitive for durations above
6 hours and below 150 hours
Projected capex learning rates are between
12 to 18 percent, consistent with other similar • In 2030, LDES can be LCOS-competitive
breakthrough energy technologies such as against Li-ion for durations above 6 hours,
offshore wind and batteries. Technology with a distinctive advantage above 9 hours
developments and gaining operational scale will
• In 2030, LDES can be LCOS-competitive
be the largest drivers of cost improvements
against hydrogen turbines with the same
The competitiveness of LDES is driven largely operational profiles for durations below
by energy storage capacity costs, which are 150 hours
expected to decline by 60 percent. The
To overcome the current cost gap and
round-trip efficiency (RTE) of these technologies
technological uncertainties of this nascent
is also projected to improve by 10 to 15 percent
market, the right ecosystem that accelerates
Some technologies are competitive today for investments should be in place

25 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 Technology costs roadmap Current system costs and
performance are comparable to
As with any new technology, competitive other nascent technologies on
costs and performance are critical to ensuring the verge of commercialization
widespread adoption and providing societal
LDES show potential for cost savings as a
benefits versus alternatives. For LDES, the key
result of technological learning rates. Both
parameters to consider are energy capacity
archetypes are sensitive to learning rates, with
cost25 (USD per kWh) or energy capex, power
75 to 90 percent of their capex being influenced
capacity cost26 (USD per kW) or power capex,
to some extent (Exhibit 21). In the 8-to-24-hour
operation and maintenance (O&M) cost (USD
archetype, 35 percent of capex is susceptible to
per kW-year), and round-trip efficiency (RTE).27
learning rates, rising to more than 50 percent in
Because the cost breakdown changes the 24-hour or more archetype as the impact of
significantly with duration, two LDES archetypes procurement costs decreases.
(8 to 24 hours28 and 24 hours or more) have been
Cost reductions are likely to be dictated by two
created based on more than 10,000 data points
factors: 1) cost improvements from increased
from the LDES Council. The analysis shows top
industrywide deployment, supplier development,
performers’ projections for both archetypes.29
and supply chain learnings; and 2) improved cost
Only the most competitive LDES technologies
reductions linked to manufacturing advances
are expected to receive the capital to scale up
and increased production volumes (namely
over the next decade and therefore constitute
learning at a manufacturer level).
the dominant portion of the mix by 2030. The
energy capex has been chosen as the defining The LDES cost-reduction rate compared to other
metric of top performers since the total cost low-carbon flexibility systems, such as Li-ion
of decarbonization (like the system cost) is and hydrogen turbines, will determine the level
especially sensitive to this metric in deeply of uptake of these technologies. However, in
decarbonized scenarios. specific applications, the distinguishing factors

Exhibit 21
Capex breakdown by sensitivity to learning rates (2025)
No sensitivity Low sensitivity Medium sensitivity High sensitivity

8–24 hour archetype 24+ hour archetype

6%
13% 14%
23% Eg, commercial &
Eg, logistics, commissioning training costs, consu-
materials, storage medium mables, material costs,
standard equipment
28% 27%
53%

36% Eg, engineering, Eg, tailored equipment for

commissioning labor, civil storage technology, project
works, permitting costs structure

Source: LDES Council member technology benchmarking

25
Capex associated with the energy storage equipment, representing the investment required to store energy.
26
Capex associated with charge and discharge equipment and BoP. The BoP includes auxiliary components such as inverters, circuit
breakers, or transformers.
27
The ratio of the total energy discharged over the total energy charged. It is calculated as an average value in standard temperature
and pressure conditions. It accounts for the electricity lost in the inverter for those storage technologies which need one and does
not include ancillary consumptions.
28
The 8-hour threshold does not imply that LDES is not expected to provide services below this duration.
29
Based on top-quartile data.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 26
 of LDES such as modularity, short time to absolute power capex, lower duration systems
market, and the ability to provide a diverse set present lower values, as they are usually
of services, will be critical in unlocking business optimized to be competitive at shorter durations
cases in the short term. and higher cycling profiles. This advantage tends
to be reduced for longer storage durations as the
Energy and power capex could energy capex becomes the main cost driver.
decline by 60 percent in the next
15 years, while RTE could grow The energy capex differs more significantly
by 10 to 15 percent as the across archetypes and scenarios. The energy
commercialization of systems capex of the 24-hour or more archetype can
accelerates reach considerably lower values than the
8-to-24-hour archetype (around three times
In 2040, the power capex is likely to be between
lower), enabling the design of these systems
USD 380 and USD 960 per kW and the energy
for longer durations due to the lower cycling
capex between USD 4 and USD 17 per kWh.
requirements to generate profits (Exhibit 22).
This compares to USD 60 to 110 per kW and
For more information on the energy capex of
USD 70 to 80 per kWh for Li-ion batteries,
median performers please refer to Box 4.
and to USD 800 to 900 per kW for single cycle
gas turbines in 2040. The power capex, which The O&M costs can experience a significant
includes charging and discharging equipment decrease between 2025 and 2040, down to
and BoP costs, is expected to show a USD 1.5 to USD 10 per kW annually, thanks to
comparable overall decline of around the deployment of larger facilities. The 24-hour
60 percent across both archetypes, or more archetype is likely to achieve O&M per
experiencing the steepest drop in the next ten kW around ten times lower than the 8-to-24-hour
years. Power-only-related costs are likely to archetype, a benefit mainly due to scale effects.
decrease faster than BoP costs as they mainly Longer duration systems present a lower RTE of
comprise standard equipment. In terms of the about 55 percent compared to above 75 percent

Exhibit 22
LDES power and energy capex trajectories
Central (conservative learning rate) Progressive (ambitious learning rate) 8–24 hour archetype 24+ hour archetype

Power and BoP capex Energy capex

USD/kW USD/kWh
2,800 24
22
2,400
20 25–40%
18
2,000
16
~60%
1,600 14
12
1,200 10
8
800 35–50%
~60% 6
4
400
2
0
2025 2030 2035 2040 2025 2030 2035 2040

Source: LDES Council member technology benchmarking

27 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 for shorter duration systems by 2040 (Exhibit industry anticipates learning rates of 12 to
23). Most of the RTE increase could be achieved 18 percent for the benchmarked period, based
before 2035 and is largely attributable to material on technology providers’ forecast deployments
science breakthroughs and adjustments in the calculated on a per-technology basis. LDES
system design. For more information on the RTE technologies’ learning rates align with similar
of median performers please refer to Box 4. energy technologies’ historical data, as seen
in Exhibit 25. However, these learning rates are
Projected capex learning rates ambiguous—as are any forecasts of nascent
for LDES systems are consistent technologies—as they have little historical
with similar breakthrough energy information to draw on.
technologies such as wind, PV, and
electrolyzers The potential learning rates for different LDES
technologies also vary as they are influenced
The total equipment capex is of primary
by the equipment used, bill of material,
importance in driving the total cost of
and sensitivity to capex improvements.
ownership to competitiveness. Learning rates
Generally, more mature technologies, such as
are a measure of how costs decrease as
electrochemical batteries, have lower-than-
output increases. For example, doubling the
average learning rates (four to five percentage
installed capacity of PV and wind technology
points below average), while novel LDES
is associated with an 18 to 24 percent cost
technologies, such as mechanical or thermal
reduction.
energy storage, may enjoy higher-than-average
Emerging LDES technologies have a significant learning rates (up to three and five percentage
potential to achieve economies of scale and points respectively).
further decrease costs through R&D. The

Exhibit 23
LDES’s yearly O&M and RTE benchmark capex reduction
Central (conservative learning rate) Progressive (ambitious learning rate) 8–24 hour archetype 24+ hour archetype

Yearly O&M RTE

USD/kW %
22 80

20
75
18
~10%
16 ~50–55% 70
14

12 65

10
60
8

6
55
4 ~40–45%
~15%
2 50

0
2025 2030 2035 2040 2025 2030 2035 2040

Source: LDES Council member technology benchmarking

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 28
 Box 4.

Top performance data compared to

median performance data
To achieve greater societal benefits and be to overachieve on today’s projections, which
competitive with other low-carbon storage has already proven possible for other energy
technologies, the broader LDES industry technologies when supported by policies and
must achieve the objectives set by the most industrial objectives. Exhibit 24 presents the
competitive market players. gap between the median and top quartile.
The current aggregation methodology does
The gap between median and top-quartile
not create artificial best-in-class players as
performance data must also be covered for
demonstrated by the fact that top-quartile
the LDES industry to achieve the results of this
players in terms of energy capex also present
study. This will require the industry as a whole
a slightly lower RTE than the median.

Exhibit 24
LDES’s benchmark capex reduction for top-quartile and
median performance data
Central (conservative learning rate) Progressive (ambitious learning rate) 8–24 hour archetype 24+ hour archetype

RTE Energy capex

% USD/kWh
76 32
74 30 33%
72 28
70 26
24 45%
68
22
66
20
64
8% 20% 18
62 16
60 5% 14
58 12
56 10
54 8
52 6 40%
50 4
70%
2
10%
0
2025 2030 2035 2040 2025 2030 2035 2040

Source: LDES Council member technology benchmarking

29 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 Projected capex learning rates The potential learning rates for different LDES
for LDES systems are consistent technologies also vary as they are influenced
with similar breakthrough energy
by the equipment used, bill of material,
technologies such as wind, PV, and
and sensitivity to capex improvements.
electrolyzers
Generally, more mature technologies, such as
The total equipment capex is of primary electrochemical batteries, have lower-than-
importance in driving the total cost of average learning rates (four to five percentage
ownership to competitiveness. Learning rates points below average), while novel LDES
are a measure of how costs decrease as technologies, such as mechanical or thermal
output increases. For example, doubling the energy storage, may enjoy higher-than-average
installed capacity of PV and wind technology learning rates (up to three and five percentage
is associated with an 18 to 24 percent cost points respectively).
reduction.
Technology developments and
Emerging LDES technologies have a significant
gaining operational scale will be the
potential to achieve economies of scale and largest drivers of cost improvements
further decrease costs through R&D. The
industry anticipates learning rates of 12 to R&D and volume will be key levers to realize
18 percent for the benchmarked period, based aspirational cost trajectories and will require
on technology providers’ forecast deployments attention from the industry to be competitive.
calculated on a per-technology basis. LDES The 45 percent reduction for the 8-to-24-hour
technologies’ learning rates align with similar archetype and the 50 percent reduction for the
energy technologies’ historical data, as seen 24-hour or more archetype until 2035 will mainly
in Exhibit 25. However, these learning rates are be driven by increased efficiencies—as a result
ambiguous—as are any forecasts of nascent of R&D—and scale, depending on the maturity
technologies—as they have little historical level of each technology. Manufacturing and
information to draw on. supply chain improvements will have a slightly

Exhibit 25
Historical learning rates for selected clean technologies and
LDES technology families
Historical learning rates for selected technologies
%
24
23

12–18 18
LDES average 0–3
% pts 0–2
14

4–5

Electrolyzers Global LDES Wind EV Li-ion Solar Electro- Mechanical Thermal

average batteries chemical

Source: LDES Council member technology benchmarking

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 30
 lower impact on the overall cost projections that each technology shows distinct
(Exhibit 26). However, they will still play a cost-reduction drivers. These diverge from
fundamental role in reaching the levers shown for LDES.
cost-competitiveness.
Li-ion’s future trajectory will be set by the
demand in EVs (more than 85 percent of future
The expected LDES cost-reduction
total demand from 2021 to 2040), with storage
trajectory is comparable with Li-ion
battery and hydrogen energy storage potentially accounting for up to 10 percent of
cost projections in the next 20 years this demand. As such, learning rates of Li-ion
batteries will be linked more closely to the EVs
The LDES system capex reduction forecast demand than the output of Li-ion stationary
(55 to 60 percent by 2040) is comparable storage. Similar chemistries still allow similar
to cost-reduction expectations reported for learning-by-doing cost reductions, procurement
utility-scale Li-ion systems (around 70 percent) scale benefits, and cell assembly benefits from
and LCOE for hydrogen turbines (around manufacturing scale that apply across EVs and
50 percent).30 Moreover, the pace of reduction stationary batteries. The largest cost component
is similar across the technology groups, with of Li-ion stationary systems is the battery pack
the fastest learning phase occurring in the next (50 percent in 2021), which is often common to
decade. This implies that the relative competitive both EVs and stationary applications and will
positioning and economic trade-offs between account for 32 percentage points of the cost
the technologies will likely remain similar over reduction due to greater value chain integration,
this period. manufacturing scale, and improvements in raw
material refinement. The remaining capex will be
Li-ion and hydrogen face a level of future
reduced by refining and specialization in other
uncertainty comparable with LDES but can rely
hardware systems, engineering, procurement
on a higher pledged level of capital investment
and construction fees, and soft costs.
and attention at the moment. There are many
detailed perspectives on the cost-reduction The hydrogen-to-power cost trajectory is most
trajectory of Li-ion and hydrogen,31 underlining sensitive to fuel costs, which currently contribute

Exhibit 26
Projected impact of different cost reduction levers on total system cost
Breakdown of cost-reduction levers, 2025–40
% of total reduction

Total capex reduction

Increased cost efficiency, eg, due to design

R&D improvements 35–60%
optimizations of major components and efficiency
of materials used

Cost reduction due to Learnings from volumes, eg, more efficient project
35–60%
scale production (volume) management and scaling up of logistics

Increased manufacturing efficiency, eg, leaner

Manufacturing and supply
15–30% production processes, cost-efficient sourcing,
chain improvements
automated assembly

Source: LDES Council member technology benchmarking

30
“Hydrogen economy outlook,” BNEF, 2020.
31
NREL; AEMO ISP; BNEF; Hydrogen Council.

31 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
to around 70 to 80 percent of the LCOE, with constraints in densely populated areas and
hydrogen turbine (or fuel cells) and transport the availability of waste heat supply).
costs being the other key components. The cost
The LCOS can be the first effective proxy to
of renewable hydrogen is expected to decline
evaluate the cost competitiveness of LDES
by an average of between 67 and 74 percent
solutions at different storage durations. With
globally by 2040, becoming widely competitive
consistent global assumptions and utilization
in the early 2030s.22 This hydrogen fuel cost
rates, LDES can be compared to Li-ion in
will be set by the development of hydrogen
shorter durations and hydrogen turbines in
for industrial, commercial, and transport
longer durations through the LCOS metric.
decarbonization applications. In turn, the
Acknowledging that storage duration is a
extensive use of hydrogen in these sectors will
continuum and that partial charge or discharge
drive developments in electrolyzer technologies,
often plays a significant role in achieving the
hydrogen transport (including infrastructure,
flexibility requirements of a project, this static
pipelines, and shipping), and fuel cell technology
analysis is helpful to understand the range of
improvements that underpin the cost reduction
durations where the cost and performance
of hydrogen for power.
parameters of LDES could allow for the most
One other aspect to consider with hydrogen- competitive applications. For more details on
to-power systems is the synergy with hydrogen the methodology and assumptions please refer
used in other energy systems. In a future with to Appendix A.
high levels of hydrogen used in nonpower
decarbonization and transported by pipeline, LDES can be LCOS-competitive
the interplay could significantly influence power compared to Li-ion batteries for
system economics. For example, in times of durations above 6 hours, with a
reduced global economic growth or recession, distinct advantage above 9 hours
there could be an oversupply of hydrogen power
Assuming a constant yearly utilization of
as cyclical industries such as steel and cement
45 percent (average real storage utilization
reduce demand.
reflected by the modeling), by 2030 LDES will
have a lower LCOS than Li-ion batteries in
Levelized cost of storage (LCOS) applications requiring more than 9 hours of
competitive benchmarking storage, with USD 80 to USD 95 per megawatt-
hour (MWh) (Exhibit 27). Competitiveness
Analysis of LCOS in static conditions against Li-ion batteries is more challenging
and comparable operations helps in applications with storage durations of less
define durations where LDES can than 6 hours, as Li-ion’s low power capex
compete costs drive low prices at shorter durations.
Due to comparable learning rates between
The LCOS provides a discounted unit value of
Li-ion and LDES technologies, the relative cost
all technical and economic factors that influence
competitiveness of LDES technologies to Li-ion
the lifetime cost of storing electricity by taking
is unlikely to change significantly before 2035.
a technology cost perspective rather than a
system one. However, when considering the
In peaking capacity applications,
potential for LDES to replace other technologies,
LDES are likely to be LCOS-
such as gas turbines or transmission, or its competitive against hydrogen
contribution to the overall system value, LCOS turbines for consecutive discharge
alone is insufficient. In these cases, it is also durations of less than 150 hours
critical to consider the operational profiles,
Some LDES already match the operational
duration requirements, commodity prices,
profile of gas peakers32 when providing grid
and other system conditions. Aside from the
reliability. For similar use cases, LDES is
cost, several other application and instance-
expected to show a cost-competitive advantage
specific properties will influence the choice of
against hydrogen turbines in durations below
a technology (such as, presence and safety

32
“Solving the clean energy and climate justice puzzle,” Form Energy, 2020.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 32
Exhibit 27
Energy storage LCOS competitiveness by duration for Li-ion and
LDES, 2030
Central (conservative learning rate) Progressive (ambitious learning rate) 8–24 hour archetype Li-ion

USD/MWh
240

220

200

180
Li-ion: lower power capex
160 but energy capex increasing
linearly with duration LDES: higher power capex
140
but low energy capex,
making duration scalable
120

100

80

0
4 6 8 10 12 14 16 18 20 22 24
Hours

100 hours when able to match the turbines’ applications—that will shape the technology’s
operational profile (Exhibit 28). In this analysis, competitiveness (Exhibit 29).
a capacity factor of 15 percent, corresponding
Combined, electricity prices and storage
to the maximum utilization associated with a
utilization have the most substantial impact on
peaking capacity asset, is assumed for the
the LCOS. For example, a charging electricity
hydrogen turbine. As the assumed capacity
price of USD 30 per MWh and a 70 percent
utilisation grows, the extent to which an LDES
utilization rate results in an LCOS of USD 70
system can be a potential substitute for turbines
per MWh. The same LCOS is obtained if the
decreases.
LDES has an utilization rate of 45 percent and
A multi-technology portfolio approach, a charging electricity price of USD 15 per MWh
including hydrogen turbines, LDES, and other in the 8 to 24 hour archetype (in line with RE
long-duration solutions, is likely to be the most LCOE in the world’s most competitive regions)
economic path to full decarbonization. Even and USD 120 per MWh in the 24 hour or more
though durations longer than 6 days cover most archetype..
renewable generation “dips”, new dispatchable
The RTE is an influential variable in the LCOS
generation will still need to be part of the
calculation (with a one-on-one correlation)
capacity mix to ensure reliability in case of longer
because it influences charging and discharging
extreme weather events (for example, weeks
requirements; however, its impact on LDES
with little sunshine and wind).
competitiveness and value is limited when
compared to the energy capex. From the
Asset utilization and lifetime
standpoint of LCOS sensitivity, the energy
average charging costs will be major
operational breakeven components capex will have a direct impact on the design
energy storage capacity of the system and on
The LCOS is highly dependent on boundary its utilization. RTE’s improvement is frequently
conditions—including specific market compromised by technological limitations.
conditions, geographical location, and end

33 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 28
Energy storage LCOS comparison by duration for hydrogen and
LDES, 2030
Central (conservative learning rate) Progressive (ambitious learning rate) Hydrogen LDES 24+ hour archetype

USD/MWh

280
USD 2/kg of H2
260

240

220 Hydrogen turbines:

high fuel cost but
200 fully dispatchable

180 USD 1/kg of H2

160
LDES: higher power capex but low
energy capex, making duration scalable
0
50 100 150 200

Hours

Exhibit 29
Impacts on LCOS by ranging different input metrics, 2030
8–24 hour archetype, 16 hours duration 24+ hour archetype, 100 hours duration
Sensitivity to Sensitivity to
LCOS impact 5% change LCOS impact 5% change
Parameter Range 2030, USD/MWh % Range 2030, USD/MWh %

Prior to change 87 151

Per energy capex

8–25 81 92 3 3–9 135 167 1
USD/kWh
RTE
90–40 71 140 6 75–25 110 268 4
%
Per power capex & BoP
400–960 76 97 2 690–1,510 130 173 2
USD/kW
Opex
5–20 80 90 3 0.9–2.6 150 152 0
USD/kWh
Electricity price
16–46 70 112 3 16–46 123 183 2
USD/MWh
Utilization
70–20 71 138 6 70–20 119 257 4
%

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 34
 5.
LDES business cases

Chapter summary

LDES can create significantly economic and

environmental benefit in the energy system
if opportunities are created to pursue it

LDES assets are being commercially installed business cases illustrate the potential for
today, having returns on investment of more than value creation in the near future for some of
10 percent the applications. Integrated utilities with future
transmission bottlenecks benefit from LDES but
To enable wider commercial deployment,
face uncertainty on monetization
LDES must achieve optimal cost-decrease and
performance trajectories, as well as technical Market support mechanisms and regulatory
maturity incentives are required to in the short term to
unlock the competitiveness of certain business
LDES value creation could benefit a broad
cases and attract the necessary private capital
range of customer archetypes. Four customer

35 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 LDES projects will have comparable For the majority of modeled business cases,
investment returns to other energy the implementation of market mechanisms
technologies by 2025, although
is required to bring the IRRs well above
market support will be critical in
10 percent before 2025 (Exhibit 30). Here the
the short term
example of earlier renewable development is
In 2025, all the modeled LDES applications instructive. IRRs of LDES projects deploying in
have internal rates of return (IRRs) well above 2025 that purely rely on existing regulations and
the minimum investor attractiveness threshold revenue streams are comparable with those
and comparable with benchmarked IRRs of PV and wind at the very beginning of their
of current mature energy projects. The high commercialization. Their growth was assisted
profitability is mainly attributable to the cost by dedicated public policy support schemes. In
trajectory assumptions, which will require early a similar fashion, LDES competitiveness could
deployments and investments supported by be unlocked by policy actions in line with the net-
an appropriate market ecosystem (explored in zero goals that capture their value. In general,
Chapter 6). beyond the business cases studies in this report,

Exhibit 30
Unlevered LDES IRRs for 2025 compared to other technologies
Unlevered IRRs of 2025 LDES applications vs Typical project IRRs Potential improvement2 IRR sensitivity to CO 2e price
selected energy projects1, %
24 CO2 price3
Non risk-
Fuel cost adjusted
20 increase5 expected
Ancillary IRR range
services
16 Avoided RE Competitive pressure effects
curtailment lowering IRRs are accounted only
for the selected energy projects
12 Capacity Low maturity
provision4 technologies

8 High maturity
technologies

0
Broader set Decarbonizing RE developer Reliability of Decarbonizing Onshore Offshore Utility Green Integrated LNG
of LDES grid (US) providing isolated grids copper mine wind wind scale PV H2 for RE and project
applications6 PPAs (US) (Chile7) (China ) (North (California)
8 industry H2 (global)
(2025) (Australia) sea) offtaker 9 (Middle
(EU) East)

2025 IRRs of specific modelled LDES projects 2025 IRRs of selected energy projects

Key assumptions
1. Projects IRR are based on country level or competitive landscape benchmarks and thus are not exhaustive of the overall market expected returns
2. Potential improvement in IRR enabled by market mechanisms and regulatory improvements
3. CO2 prices modelled for three scenarios: base scenario (60 USD/tCO2e in 2030), medium scenario (75 USD/tCO2e in 2030) and high scenario
(100 USD/tCO2e in 2030)
4. Maximum modeled weighted average capacity cost of 16.7 USD/kW per month (from LA Basin, California)
5. Islanded grid from 150 USD/MWh to 200 USD/MWh, off-grid mine from 250 USD/MWh to 300 USD/MWh
6. Includes the broader list of LDES projects, not explicitly modelled in this report (
7. Particularly favorable project and not necessarily representative of all mining applications
8. Benchmarked wind onshore projects IRRs in Shandong and Jiangsu regions
9. Assuming subsidies commensurate with track record in currently deployed projects, project IRR is scale-dependent.
Source: Grant Thornton; Renewable energy discount rate survey 2018

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 36
 LDES presents a wider spectrum of applications, T&D system operators, corporates with
related IRRs and sensitivities, described under environmental, social, and corporate governance
the “broader set of LDES applications” category. commitments, industrials with high uptime
requirements, isolated island communities,
LDES assets are being commercially military bases, and public and healthcare
installed today, but there are key services with backup requirements. While the
challenges to be overcome for wider common theme is the need for power resilience,
commercial deployment each use case is driven by local regulation.
For example, public utilities with grid-system
Different use cases of deploying LDES in the
operating functions are likely catered for by
near term (2025 to 2030) are explored through
LDES applications for energy shifting, capacity
four case examples. The selected cases have a
provision, T&D optimization, and stability
positive net present value (NPV), driven by an IRR
service provision. Whereas corporate players,
above 10 percent; this improves as commercial
such as RE developers or owners or industrial
operation dates are shifted to 2030 and beyond,
customers, are more likely to be interested in
benefitting from system cost reductions and
LDES’ ability to firm PPAs or optimize energy
lower capital investment requirements. Where
sourcing for offtake in a remote or unreliable grid.
LDES deployment is not yet economical, several
potential mechanisms could unlock financial The business case of an integrated US-based
viability (explored in Chapter 6). utility is described below. The other case
examples are summarized in Exhibit 31 and
The different LDES use cases could benefit a
described in Appendix B.
broad range of customers, including integrated
utilities, independent power producers,

Exhibit 31
Assessment of LDES-driven business cases
IRR(potential Value drivers
Application Case example Customer example improvement) for LDES Key unlocks
US-based utility Integrated utilities with ~3% T&D Market mechanisms enable
Energy shifting, significant RE build-out optimization remuneration of CO2e
capacity provision, (+11%)
and transmission bene-fits for LDES asset
and T&D bottlenecks between Capacity owners
optimization generation and demand provision
Regulatory options or
CO2e cost incentives ensure WACC
savings commensurate with RE
Stability services development
provision (eg, inertia) RE
curtailment Sustained carbon price in
reduction line with NDCs1
Firming for PPAs RE developer RE developers or owners ~7% Firmed Increase certainty for LDES
in Australia looking to serve corporate capacity RE developers through long-
RE PPAs with firmed PPA term contracting
capacity premiums
Isolated island grid Isolated island Integrated utilities serving ~7% Production Regulatory options or
optimization integrated isolated island power (+5%) cost savings incentives ensure WACC
utility in the US systems with commensurate with RES
decarbonization ambitions CO2e cost development
but limited interconnectivity savings
Optimization of Diesel-powered Industrial customers ~15% Production Market mechanisms enable
energy for copper mine in looking to reduce the costs (+4%) cost savings remuneration of CO2e
industries with Chile of energy supply and benefits for LDES asset
remote/ unreliable reduce carbon footprint of CO2e cost owners
grids products savings
On-demand RE peak RE developer in RE developer in India ~10% Peak and off- Increased electricity pricing
power India providing morning and (+2%) peak power spread and higher need for
evening peak supply as supply clean dispatchable peaking
well as off-peak generation power
1. Nationally determined contributions.

37 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 A US case study shows how will be connected, critical choices about T&D
integrated utilities can benefit from investments for the next decades need to be
multiple LDES applications but face
made.
uncertainty on monetization
For these customers, an LDES system could
In the US, several integrated utilities are
provide multiple solutions. The installation of an
responsible for their local grid’s energy
LDES system displaces a share of the demand
generation and operation. They typically rely
for electricity from gas-peaking power plants
on carbon-intensive gas peaking plants to
and reduces the production and emission costs.
support local load and reliability. In cases where
In addition, LDES storage capacity can absorb
there is a geographic divide between generation
a large volume of electricity currently being
and demand, transmission bottlenecks arise.
curtailed during peak production times. LDES
Limited transfer capabilities and challenges
also provides reserve capacity to replace the gas
when building new transmissions drive the need
power plants that currently deliver this service.
to improve the utilization of T&D networks and
As LDES increases, the utilization of the T&D
maintain reliability in the low-carbon future.
network and costly capacity expansion can be
Given the vast amounts of RE capacity that
optimized. (Exhibit 32)

Exhibit 32
Integrated US utility case example
Main LDES application(s) Value Net value Cost Accessible value with market mechanisms in place

NPV for an integrated utility customer

Customer profile USD millions

Integrated utility in the US that depends

on gas-peaking plants for reliability  Production cost savings ~0

Geographic divide between generation and

demand with transmission bottlenecks  CO2e cost savings 210–220

Challenges building new transmission Market

 RE curtailment reduction 60–70 mechanisms
LDES assets to displace gas-peaker plants in place
and improve RE utilization
 Capacity provision 290
Potential LDES system: 200 MW/2,000 MWh
(10 hours); systems of longer durations are  T&D optimization 230
also seeing demand driven by utilities’ long- Value from
term needs transmission
 Stability services provision savings currently
10–20
inaccessible to
non-grid owners
Total value creation 300–830

 Invested capital 340

 Total fixed O&M 70

IRR
NPV -110–420 3–14%

Assumptions Final investment decision date Commercial operation date CO2e price scenario WACC
2023 2025 Base case 6%

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 38
 The case shows that the NPV ranges between buildout for different scenarios to determine the
-USD 110 million and +USD 420 million. To relative storage investment required to offset
access the higher end of this range, market transmission spend. CO2e cost savings originate
mechanisms would have to be fully in place to from the opportunity of replacing a gas peaking
ensure the benefits can be captured, especially if plant with LDES.
the customer is not an integrated utility that also
The economics of this case are sensitive to CO2e
operates the grid.
prices and project start dates. The IRR increases
Transmission optimization, capacity provision, significantly to between 9 and 25 percent when
and CO2e cost savings (USD 230 million, USD the commercial operation date moves to 2030,
290 million, and USD 210 million to USD 220 with the construction of the system taking place
million, respectively) are the most significant in the two years prior. Furthermore, accelerated
contributors to the overall value creation of USD CO2e price increases could result in IRRs of up to
300 million to USD 830 million. Transmission 16 and 29 percent with operation dates by 2025
savings compare transmission costs to storage and 2030, respectively. (Exhibit 33)

Exhibit 33
Integrated utility case example – US, IRR sensitivity
IRR sensitivity to carbon price and project start date1 Base case <5% 5–10% 10–15% 15–20% >20%

High
3–16% 9–29%
100

CO2e price
Medium
scenario 3–15% 9–26%
75
2030 USD/tCO2e

Base
3–14% 9–25%
60

2025 2030

Commercial operation date

1. Lower end of range for value capture in markets with appropriate mechanisms; higher end of range for full value potential.

39 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
To ensure the financial viability of LDES for
integrated utilities, several key conditions have
been identified. First, the LDES asset owner
would be able to monetize the benefit created
by CO2e emission reductions through adequate
market mechanisms. Second, regulatory
options or incentives would be in place to bring
the owner’s cost of capital in line with that of
other decarbonization efforts. Third, CO2e
prices would be rising in line with the increasing
ambitions of national emission reduction plans,
such as nationally determined contributions
(NDCs). Fourth, RE owners would use LDES
charging as much as possible, especially since
net-zero grids are not fully deployed yet; this
could be ensured by schemes that facilitate the
traceability of energy generation and certify its
origin.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 40
 6.
Road to competitiveness
and key market enablers

Chapter summary

Three potential actions could help unlock

LDES value by changing the way storage
is regulated and remunerated

Driving the economic and technical maturity • Dedicated support programs to reach
of LDES technologies should be aligned with cost cutting potential and test new market
the large-scale deployment of RE to achieve mechanisms
maximum societal cost reductions
• Targeted support schemes such as RE
A supportive ecosystem with concrete actions and LDES tenders to incentivize take-up by
would be beneficial for the prompt development sector players
of the market. In particular, 3 key areas for action
• Support for manufacturing and supply
have been identified:
chain improvement to increase scale and
1. Long-term system planning could help attract reduce capex
adequate levels of private investment:
3. Market creation to ensure financial returns
• National upfront planning to optimize during the lifetime of the assets
the capacity mix, grid infrastructure, and
• Market mechanisms and designs to ensure
storage
compensation for flexibility provision
• Clear RE targets to create demand for
• Enabling regulation to facilitate LDES
energy storage and provide visibility to
uptake (e.g. safety standards, market rules
investors
that capture LDES value)
• International coordination to enhance
A lack of supportive market could significantly
efforts across markets and regions
delay the deployment of LDES technologies
2. Support for first deployments and
scaling-up capabilities to lower
investors’ barrier of entry and risk

41 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 The timely maturity of LDES that mature quickly enough to meet market
technologies is essential to enable demands are likely to make it into the portfolio
the optimal integration of RE in power
of solutions that support the power system
systems
transition.
LDES can play a vital role in decarbonizing
Potential accelerators for the adoption of
the world’s power sector. By 2040, LDES
LDES may emerge. They could include a
deployment could result in the avoidance of
faster trajectory to net-zero power systems
1.5 to 2.3 Gt CO2eq per year33 (around 10 to
than the one assumed in this study, either at a
15 percent of today’s power sector emissions)
local level or in regions or countries with high
by enabling dispatchable RE and the
decarbonization targets. These would naturally
replacement of emitting plants.
demand solutions to de-risk and balance the
Most importantly, they can do so at no additional integration of large amounts of RE. In addition,
societal cost. The overall cost of power systems sustained and incremental CO2 pricing will
could be reduced by around USD 35 billion enhance the value of specific business cases by
annually by 2040 in the US alone (Exhibit 34) creating new revenue streams and reinforcing
under a 100 percent decarbonization scenario existing ones. Last but not least, the evolution
and top quartile performance, of which of alternative solutions and their deployment
USD 5 billion would come from the application constraints (such as supply chain shortages for
of LDES in T&D expansion projects. Li-ion and high demand in the EVs segment)
could heavily influence the demand for LDES
In order to realize their full potential, LDES
technologies.
technologies need to reach technical and
economic maturity alongside the widespread
Achieving the optimal LDES capacity
deployment of RE. Whether LDES developers
deployment through 2040 will require
achieve the cost-reduction trajectories outlined significant investments
in this study or not depends on improved
technological designs, the streamlining and Overall, using LDES to upgrade electric power
optimization of manufacturing capacities, and systems in the most cost-effective manner
scale factors. Furthermore, rapid technological will necessitate significant private investment.
progress will be essential to ensure their Cumulative capex investments of USD 50 billion
adoption at a fast pace. Only the technologies are likely to be required to deploy the projected

Exhibit 34
Total in-year societal cost of bulk energy shifting
XX Societal costs per unit of generated electricity T&D investment Capacity investment

490
460 450
USD bn -7%

310

90% decarbonization No LDES LDES central LDES progressive

scenario scenario

USD / MWh 34 46 44 43

100% decarbonization

33
Assuming that the total electricity discharged by LDES globally in 2040 would be emissions-free and substitute traditional gas
peaking capacity.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 42
 capacity until 2025, with USD 1.5 trillion to frequently result in the acquisition of equipment
USD 3 trillion needed globally until 2040 to unsuitable for long-term system needs.
realize progressive cost projections.
Academic and industry progress in building
LDES technologies would also benefit from new capacity expansion models has led to an
government support to kick-start the market emerging set of best practices about how to
as quickly as the net-zero transition demands. plan low carbon grids that rely substantially
Short-term funding for these technologies can on renewables and storage. Where possible,
be viewed as a long-term investment that will pay capacity expansion models and the investment
off in the form of a lower-cost power system and decisions they require should be based on: at
a de-risked transition. least one full year of grid operations at hourly
resolution, including weather and load variability
A supportive ecosystem with that reflects day-to-day, week-to-week, and
concrete actions would be beneficial season-to-season fluctuations; multiple weather
for the prompt development of the years and key future system conditions, such
market as technological availability, commodity prices,
or other variables. This would lower consumer
To overcome the current cost gap and
costs as well as the risk of unanticipated power
technological uncertainties of this nascent
outages and supply chain constraints.
market, the LDES Council believes that
governments and business leaders can catalyze Power system planning that includes LDES is
the development of the market by creating the already taking place in some advanced regions.
right ecosystem that accelerates investments. For example, California has already procured
MW-scale LDES, and New South Wales,
Three key dimensions where support actions
Australia, announced last year the procurement
with the highest impact have been identified:
of 2 GW of LDES in its Electricity Infrastructure
4. Long-term system planning to create the right Roadmap.
investment signals
Clear RE targets and strategies to accelerate
5. Supporting the first deployments and permitting would also create early demand
scaling-up capabilities to kick-start the for energy storage to balance the variability in
market renewable generation. RE generation, T&D grids,
and energy storage are highly interconnected.
6. Creating the market to capture LDES value
As such, clear strategies on RE integration,
and allow monetization
storage, and grid upgrades would provide
1. Long-term system planning to create visibility to investors and incentivize uptake
the right investment signals by RE developers.
Clear commitments to net-zero emissions and
Lastly, international coordination is also
comprehensive decarbonization road maps
essential to establishing the world’s path to
from governments and industry are essential
net-zero power. Coordination of efforts yields
to meet climate targets. Long-term system
principles and lessons that can be replicated
planning could attract adequate levels of private
across markets. Applications or regions with
investment in both technological advancements
similar needs could join forces to monitor the
and early system deployment, ensuring the
technologies that best support the energy
timely development of enabling solutions such
transition, establishing industrial or regional
as LDES.
networks capable of formulating needs and
LDES can significantly improve the reliability and providing knowledge. All players need to
resilience of power systems. Net-zero power continue to expand the knowledge base on
systems could benefit from upfront planning LDES technology capabilities, value, and
(similar to Publicly Owned Utility Integrated development trajectories.
Resource Plans in the US) to optimize the
2. Supporting the first deployments and
capacity mix, grid infrastructure, and storage
scaling-up capabilities to kick-start the
deployment. Upfront planning would minimize
market
the number of emergency procurements, which
The majority of novel LDES technologies have

43 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
not reached full commercial maturity yet, which RE players, system operators or off-takers) and
presents a barrier for raising large amounts of collaboration with LDES technology providers
private capital. Without an established market can also accelerate early LDES deployment.
for LDES and a track-record of the performance Early movers (like mining companies or data
of these systems, investor perceptions of high centers) willing to cover the green premium will
risks will limit funding and constrain the ability be essential to kick-start the market and develop
of developers to continue testing and improving learning curves. Tenders for RE co-located with
their technologies. storage with a minimum discharge duration,
public-private partnerships or early market
As a result, dedicated support programs for
mechanisms could accelerate uptake. In this
large-scale demonstration plants would be
way, the maturity of the technologies would
essential to ensure that these technologies can
benefit from early positive cash flows that could
reach their full technological and cost-reduction
be reinvested in further improvements while
potential and that new market mechanisms can
refining market creation.
be tested. Such support could take many forms,
some of which are listed below, and should be Similarly, manufacturing and supply chain
implemented in the short term to accelerate improvements could reduce the total capex
deployment. by 15-30 percent. This support could include
increasing manufacturing efficiency, automated
For example, for the deployment of utility-
assembly, and cost-effective sourcing.
scale, grid-connected demonstration plants,
government-funded grants and financial To achieve full decarbonization, cleantech
instruments would be critical. Grants would solutions would be needed in harder-to-
accelerate design improvements (for example, decarbonize applications. These include the
in the RTE), reduce costs through R&D, and provision of backup power for critical loads, such
decrease uncertainties around operational as hospitals or telecommunication towers, where
performance, which would de-risk such projects decarbonization by any technology will entail
for investors. higher costs than the use of fossil-fuel-based
installations. LDES are suitable for these backup
Initiatives are already underway in countries such
applications; however, the low utilization implied
as the UK, where the Department for Business,
with backup usage may lead to unfavorable
Energy & Industrial Strategy launched a USD 100
economics.
million LDES demonstration competition in early
2021 to accelerate project commercialization. 3. Creating the market to capture
Similarly, the US Department of Energy has LDES value and allow monetization
launched a program to reduce costs of LDES of Even at the commercial-readiness stage, risks
more than 10 hours of duration by 90 percent surrounding the future-cost trajectories and
in one decade. The program has requested a the revenues assets can capture during their
budget of more than USD 1 billion. In the EU, lifetimes will remain. Current markets generally
the Innovation Fund also provides grants to do not capture the full value of LDES since:
energy storage projects based on innovative
• Power markets are mostly short-term
technologies. Not only grants, but financial
(such as day-ahead, intraday markets)
instruments (such as blending financing
and generally do not provide long-term
instruments, thematic growth instruments, or
agreements that could de-risk capital;
credit enhancement mechanisms) would help
to catalyze private funding and de-risk early • Multiday and multiweek market signals are
projects. weak compared to intraday, and therefore
storage technologies are incentivized to cycle
Society could also learn from successful
multiple times per day;
support schemes offered to other clean
technologies such as solar or wind (e.g. tenders • Carbon-reduction compensation schemes
that reward best-performing technologies either do not exist or are insufficient to
against determined criteria), and implement compensate investors for the additional
similar measures on a technology-neutral basis. funding.
Incentivizing sector players’ uptake (such as

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 44
 Therefore, in the short- to mid-term, it is essential in maturity, safety features, and use cases,
to devise markets which allow the benefits resulting in a lack of shared understanding and
created by LDES to result in financial returns valuation. In addition, the role of storage in the
and attractive IRRs. Several market design energy system is complex, both within the power
and regulatory actions could help minimize the system and in conjunction with industrial end
operational risk of commercial plants, providing uses. Therefore, the appropriate valuation of
visibility on revenues during the lifetime of the LDES through the establishment of clear market
assets. rules is necessary. In addition, a clear definition
of LDES, including minimum technical and safety
Optimal market designs that create the right
requirements, would facilitate its development
incentive signals for long-duration services
and implementation in the marketplace. Finally,
will vary by location depending on the local
carbon pricing mechanisms would need to be
resources and infrastructure. For example, it
designed so that low-carbon technologies are
may become increasingly difficult for generators
not outcompeted for similar flexibility services by
and storage owners to generate an income
emitting assets.
solely from energy payments in markets with
limited price volatility. These markets may need Some frontrunner countries or regions have
to be redesigned to compensate for flexibility, produced examples of legislation explicitly
which could be accomplished through long- designed to meet the needs of LDES. For
term capacity payments or new imbalance instance, the California bill AB 2255 (2020)
compensation markets. Alternatively, other proposes the adoption of a new regulatory
markets may need to open to LDES as firm approach as it aims to develop a process to
capacity and balancing providers. This could procure and deploy GWs of LDES across the
contribute to various benefits, including state. In addition, Arizona has launched an
increased competition, increased innovation in incentive program structured to encourage
the electric power industry, and increased grid longer durations, by offering incentives for
flexibility and resilience. storage technologies with more than 5 hours
of discharge. In the UK, there are ongoing
Lastly, a set of requirements and drivers for LDES
conversations on different routes to increase
uptake would be desirable from a regulatory
profitability for LDES, including 15-year capacity
perspective. LDES technologies vary widely
markets or balancing mechanisms34.

34
“Facilitating the deployment of large-scale and long-duration electricity storage: call for evidence,” UK Department for Business,
Energy & Industrial Strategy, 2021.

45 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Conclusion
This report has shown that Long Duration development of a market in LDES, and stimulate
Energy Storage can play a crucial role in fully early investment. Large deployment is required
decarbonizing the power sector and thus in the next few years in order to build scale
enabling a pathway to limit the rise in global and realize the cost projections set out here.
temperatures to 1.5 degrees as set out in the Governments need to establish a supportive
Paris agreement. It can provide the power ecosystem including long-term planning,
system flexibility and stability required to economic incentives and appropriate market
integrate an increasing renewable share in power designs.
generation with its inherent variability, and it can
To be clear, this is not a proposal for an ongoing
do so at a manageable cost.
subsidy regime at the public expense: the
Data from LDES providers shows it has proposed recommendations are designed to
significant potential to become the most cost- kick-start a functioning market that can support
competitive solution for energy storage beyond a society’s objective of rapid decarbonization.
duration of six to eight hours: the social benefits All the evidence suggests that this could be a
of large-scale deployment as solar PV and wind highly attractive market for investors: a sizeable
become the dominant sources of power are new industry providing 1.5 to 2.5 TW of storage
obvious. capacity, requiring an investment that could
reach USD 1 to 3 trillion by 2040 with potential
These projections, however, come with an
competitive returns. The prize is within reach,
important caveat. They will only come to pass
and the time to seize it is now.
if action is taken in the short to medium term
to create the right framework conditions for

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 46
 Appendix A: Methodology

Total addressable market (TAM) and market size restrictions, such as biomethane
modeling
blending, nuclear new build restrictions, and
This report estimates the LDES future transmission expansion restrictions, were also
deployment and TAM by leveraging the modeled.
McKinsey Power Model (MPM), a long-run
The model contains bulk-transmission-
capacity expansion model which includes
level grid connections (i.e., no mid-voltage
elements of production cost modeling, to
transmission or distribution grid), and within
size deployment requirements in bulk grid
the smallest modeling region, transmission is
applications. Additionally, the analysis augments
not represented, i.e., intra-region transmission
the bulk potential with other different on-grid and
effects are not included, corresponding to a
off-grid applications for that are not captured by
‘copper plate’. This modeling limitation will
a large-scale capacity expansion model.
necessarily underestimate the market size of
TAM results are sensitive to assumptions on LDES since transmission constraints, which
LDES and alternatives, hence multiple sensitivity LDES can provide a strong value proposition to
analysis have been carried out. mitigating, are not fully considered. In addition,
the model only covers the power sector as well
McKinsey Power Model as fuel creation related to supply predefined
demand from the power sector, e.g., such as
The MPM is a techno-economic optimization
hydrogen production. No co-optimizations on
that simulates large-scale power systems
other sectors such as dual fuel boilers, space
concurrently on an hourly and multi-decadal
heating co-optimization or global clean fuels
time resolutions. It was used to determine the
flows were considered. These aspects could
cost-optimal pathway to net-zero emissions
be considered in future analysis to further
across a set of real-world systems. The result is
understand the potential of LDES.
a portfolio of technologies and fuel consumption
that minimize the societal cost of the transition in
Non-MPM TAM estimates
the modeling horizon.
As a parallel effort, additional sizing outside
A wide set of technologies ranging from
the MPM was performed, estimating the LDES
traditional thermal generators such as
market size in off-grid applications, and the value
gas turbines and nuclear power plants to
created by LDES in use cases not considered
technologies with increasing potential in the
by MPM. Five additional value streams have
energy transition, such as renewables, CCS,
been defined and assessed: optimization
energy storage, and power-to-fuel were included
of transmission and distribution investment,
in the model. The modeling effort specifically
stability services provision, firming for PPAs,
focused on the role of LDES in the net-zero
isolated island grid optimization, and energy for
emissions transition. The result provides an
industries with remote or unreliable grid.
outlook for the LDES market size and a possible
operational profile. In the optimization of transmission and
distribution case, the generation capacity
Various sensitivities for technologies were
support of LDES is already accounted for in the
defined to study the impact on the technology
MPM results while optimal geospatial placement
portfolio, and specifically, the LDES market
of deployed LDES would provide additional value
size. The capital cost reductions of LDES
creation, which has been separately sized. The
technologies were defined based on the learning
value refers to the savings in transmission and
rate and technology commercial readiness
distribution infrastructure investments, sized
gathered from data submissions of LDES council
by assessing the potential of LDES to increase
members. Different technology build decisions
the grid utilization, therefore reducing buildout

47 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
requirements, while not impacting system understanding what proportion would require
reliability. The methodology was followed in near 100 percent 24/7 RE coverage and
detail for two countries, the US and Germany, assuming a level of LDES penetration (versus
before extrapolating the figures to a global value. Li-ion or other firming options). To calculate the
value of these deployments, the cost of covering
Several different approaches were used to size
any non-RE power consumption with the
the deployment of LDES in remote and unreliable
purchase of RE guarantees of origin and carbon
grids. For remote mines a hybrid model was
credits was calculated. New deployments and
deployed in a method analogous to islanded
value from RE PPAs were assumed to reach a
power grids in a Chilean copper mine, with the
peak in 2030 before being phased out as general
McKinsey MineSpans database used to identify
grids reach high proportions of firmed RE.
the electrical energy requirements of suitable
mines globally. For unreliable grids, sectoral The opportunity for stability services was also
energy demand with requirements for high modeled focusing on inertia provision (as likely
uptime (e.g., chemicals, manufacturing, metal the most significant service). It was reasoned
processing) were identified in countries with that LDES would not be economically installed
high historical records of blackouts. The total purely for inertia provision, so modeling focused
storage requirement to bridge blackout periods on identifying the additional value that could be
was calculated and an LDES penetration was achieved should these services be monetized.
assumed to estimate the deployed systems. To do this, a method for the next lowest-cost
A value equivalent was attributed to the alternative was implemented using the cost of
productivity of avoided downtime. For critical installing synchronous condensers from various
on-grid assets using RE (e.g., military bases real quotations, and scaling for the installed
and hospitals), only the LDES energy capacity quantity of LDES in a region from the MPM
deployment was included (the expectation is results.
that LDES will be used for more than backup
purposes); however, additional value was LCOS modeling
estimated with the removal of backup diesel
The LCOS represents the discounted total
generation.
unit cost of ownership of storage technology
LDES deployment and value of island grids over the project lifetime. This metric accounts
was based on in-depth hybrid energy system for all technical and economic parameters
modeling using real hourly supply/demand load impacting the lifetime cost of discharging stored
profiles of the O’ahu island system in Hawaii to electricity. It is directly comparable to the LCOE
determine the optimum decarbonized energy for generation technologies and represents
setup across RE options, Li-ion, and LDES an appropriate tool for cost comparison of
overtime. This produced a deployment of LDES electricity storage technologies. However, LCOS
per GWh consumed annually and value savings for storage, much like LCOE for dispatchable
based on reduced production costs, stability generators, is not an intrinsic property of the
services, and CO2e reductions. This result was installed technology, but depends heavily on the
then scaled to cover global isolated island grids. operations of the system. As such it is extremely
This was done by identifying all islands with useful in comparing costs, but the user should
0.1 million to 5.0 million inhabitants, then filtering understand input parameters and limitations of
those based on mainland grid connections and the comparison. Modeling deeply decarbonized
other common-sense checks (e.g., removing systems is a complex task, and metrics such
highly populated Indonesian islands that as LCOE and LCOS provide a baseline to orient
would be double counted with the main MPM market participants to the relevant technologies.
modeling) before a moderation of storage need Exhibit 35 reports the LCOS formula, showing its
was conducted based on RE potential in each components.
county. Finally, the result was scaled using the
The LCOS must be handled carefully to create
annual energy consumption of each island.
meaningful results. Successful LCOS use
Sizing of LDES requirements for corporate cases appear in well-defined storage demand
RE PPAs was taken using forecasts for RE and well-understood technology behavior,
PPAs globally based on historical trends, like daily energy arbitrage markets. These

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 48
Exhibit 35
LCOS formula
Computation of the levelized cost of storage (LCOS)
 Round-trip efficiency (RTE)
 Capex of LDES solution,  O&M costs  Ancillary consumption, self-discharge
construction, balance of system  Replacement intervals and costs  Cost of charging energy

Installation cost + lifetime discounted O&M cost + lifetime discounted charging cost
LCOS =
Total lifetime discounted electricity discharged

Used throughout
 Discount rate
 Nameplate capacity
 Depth of discharge
 Duration, unit energy costs
 Annual degradation rate
 Annual and lifetime cycle count

factors imply consistent utilization profiles that utilization values, a capacity factor of
produce sensible LCOS comparisons. Poor use 15 percent for the turbine was assumed.
cases are defined by occasional or sporadic
• LDES energy and power capacity, and
storage demand, like utility reliability markets
charging rate: the energy and power capacity
or integrating multiple revenue streams and
values are system nameplate capacities.
storage uses. For these sporadic use cases
100 MW was the chosen nameplate power
with sparse cycle counts, LDES solutions are
capacity for the different systems compared.
better compared to capex or annuitized capacity
The charge and discharge limitations are
costs and not LCOS since efficiencies and
accounted in the depth of discharge metric,
replacement costs are less important.
which is defined as effective dischargeable
The main assumptions are: energy capacity over nominal energy
capacity (considering both charging and
• Annual LDES asset utilization: 45 percent
discharging limitations). Only nominal
in all durations. This value represents the
charging rates have been considered (i.e.,
portion of time the storage is either charging
1C charging rate for Li-ion).
or discharging. It was inferred as an outcome
of the MPM and chosen as the base case. • Li-ion costs: the progressive and central
scenarios are based on the McKinsey
• Average lifetime charging cost: 30 USD/MWh
Battery Cost Model. The central scenario
• Hydrogen turbine costs: they are based implies a cost improvement learning curve
on findings from the Hydrogen Council35 projections without considering disruptive
and latest academic literature figures. Li-ion technology breakthroughs, while
Assumptions on the cost of hydrogen in the progressive scenario anticipates
2030 are key in determining the two different aggressive component cost improvements.
scenarios (2 USD per kg in the central The assumed Li-Ion and hydrogen cost and
scenario and 1 USD per kg in the progressive performance trajectories are amongst the
scenario). Relying on the academic and most progressive from their sources.
industry consensus on peaking plants
• WACC: 6 percent

35
Hydrogen Insights, 2021

49 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Carbon cost trajectories
CO2 costs differ heavily by region and will have
different development trajectories influenced
by policies and regulation at national and
international level. The MPM sets specific
emission reduction targets rather than assumed
CO2 cost trajectories.

CO2 costs outlooks have been accounted in the

modeled business cases, where three scenarios
have been defined: base (60 USD/tCO2e in
2030), medium (75 USD/tCO2e in 2030) and high
scenario (100 USD/tCO2e in 2030). All scenarios
assume a 8 percent compound annual growth
rate over the period from 2030 to 2040.

Currency
All financial figures are in 2020 US dollars (USD)
and refer to global averages unless otherwise
indicated.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 50
 Appendix B: Examples
of business cases

1. RE developer in Australia For example, on average, 5 percent of solar

and wind power was curtailed in Queensland,
The business case for RE developers Australia, since 2019, putting pressure on RE
or owners looking to increase developers’ financials. These curtailment levels
the firming of RE PPAs with LDES are likely to increase to double-digit percentages
systems could be attractive in the in the next decade, as seen in countries with
short term higher RE penetration. In addition, stability issues
According to the Australian Clean Energy will become more pronounced as conventional
Council, tenders for RE corporate PPAs with generation plants (such as coal and eventually
firmed capacity are expected to grow in natural gas) are phased out.
Australia. They enable customers to achieve
Given this context, a deployed LDES system can
environmental, social, and corporate governance
participate in multiple value streams. Here the
targets and hedge market volatilities. In addition
case of a RE developer deploying LDES for RE
to upcoming RE needs from end consumers,
PPA firming with a front-of-the-meter contract,
grid systems are experiencing increasing
whilst providing services to system operators is
challenges in bringing RE capacity online.
explored. (Exhibit 36)

Exhibit 36
Australia renewables developer case example, net present value
Main LDES application(s) Value Net value Cost

NPV for an Australian RE developer

Customer profile USD millions

RE developer in Australia providing RE for

corporate PPAs  RE curtailment reduction 25–40

Increasing corporate demand for near 100%

RE to meet decarbonization targets  Stability services provision 10-50 Value of discounted
premiums for RE
LDES asset enables firming of the RE PPA; • Enabling firm capacity PPA firming
additionally, system can provide ancillary 280
RE PPA
services
Total value creation ~300–360
Potential LDES system: 150 MW/1200 MWh
(8 hours); systems of longer durations would
be required in locations with lower annual RE  Capital invested 250
yield

 Fixed O&M 50

IRR
NPV 15–70 6.5–8%

Assumptions Final investment decision date Commercial operation date CO2e price scenario WACC
2023 2025 Base case 6%

51 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 The case example shows an IRR of years prior. However, it is more common to have
approximately 6.5 to 8 percent, indicating shorter rather than longer contract durations,
short-term financial attractiveness. In this case, with shorter RE PPA contract terms resulting in
the most significant value stream is created 1 to 2 percentage point lower, making the IRR
by enabling firm capacity for a RE PPA and fall below the weighted average cost of capital
has a present value of about USD 280 million, (WACC).
indicating regional PPA price indexes have a
Several steps could help unlock additional
considerable influence on the case profitability.
value potential—for example, creating market
The reduction of RE curtailment and provision mechanisms that enable RE developers to
of stability services contribute to a lesser access different value streams outside the
extent, with current values of approximately market for energy shifting such as inertia
USD 25 million and USD 10 million, which could provision. Another option would be increasing
potentially increase to around USD 40 million certainty on accessible value through regulatory
and USD 50 million respectively due to increase schemes that make it more attractive for
curtailed RE volumes and regulation of stability corporate customers to engage in long-term
services such as inertia provision. (Exhibit 37) contracting. The NPV could also increase if
the WACC is lowered through different existing
In addition to the project starting date, three
and innovative financial instruments, including
sensitivities influence the project IRRs. The
insurance for energy storage and public–private
IRR increases significantly to approximately
regulatory options.
14 percent for the base case when the
commercial operation date moves to 2030, with
the construction of the system taking place two

Exhibit 37
Australia RE developer case example, IRRs
IRR sensitivity to contract durations and project start date1 Base case <5% 5–10% 10–15% 15–20% >20%

Curtailment 20 7.5% 15%

volumes 15 7% 14.5%
% of energy
output curtailed 10 6.5% 14%

Ancillary services 3.75 8% 16%

revenues
2.25 7.5% 15%
USD millions
annually
0.75 6.5% 14%

20 6.5% 14%
PPA contract
duration2 15 6% 13.5%
Years
10 4.5% 12.5%
2025 2030
Commercial operation date
1. Lower end of range for value capture in markets with appropriate mechanisms; higher end of range for full value potential.
2. After contract end, value of service ~50% of in-contract value.

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 52
 2. Isolated island integrated could install a hybrid of additional solar and
utility in the US wind with Li-ion batteries and LDES. Detailed
modeling of an isolated island system indicates
The near-term financial viability that the lowest cost pathway to 100 percent
of LDES for integrated utilities on fulfillment of energy demand by RE employs a
isolated power systems with limited combination of Li-ion and LDES. The LCOE of
interconnectivity depends on local this configuration is 15 percent lower than a pure
fuel costs and RE potential Li-ion battery system—caused by the significant
Off the mainland, the US has multiple islands RE overbuild—and 5 percent lower than a pure
that have no connection to neighboring islands LDES system. (Exhibit 38)
or the mainland grid. Hence, they are mainly
The deployment of such a RE hybrid storage
dependent on coal and fuel-oil-generated
system could take place in a multi-phase
power. The electricity cost for consumers
buildout. It is here assumed that 40 percent
on these islands is among the highest in the
of energy demand is fulfilled by the existing
US.36 At the same time, there is considerable
build-out of RE without storage. The costs and
potential for low-cost RE generation. These
benefits of this pre-deployment phase are not
conditions have already led to a buildout of
included in the assessment of the business case.
solar and wind capacity, increasing the share
In the first phase of the deployment additional RE
of RE in the generation mix. With the growing
capacity and Li-ion battery capacity to achieve
decarbonization of the island’s power system,
70 percent RE fulfilment. In the second phase,
thermal generation will be decommissioned and
additional RE capacity, Li-ion battery capacity,
stability services reduced.
and the LDES system would be constructed
To achieve full power decarbonization on such to achieve 100 percent RE fulfilment by 2030.
an island, the incumbent integrated utility (Exhibit 39)

Exhibit 38
Cost comparison of different storage options for decarbonizing electricity
on an isolated island and a remote mine
LCOS in 2040 for different storage mixes in a RE hybrid system
cents of USD per kWh

100% RE fulfilment for an isolated island 100% RE fulfilment for a remote mine
9.0

8.0
7.5
7.0
6.5
6.0

Li-ion LDES Li-ion + LDES Li-ion LDES Li-ion + LDES

Key assumptions: Top-quartile LDES 24+ hour archetype cost figures, conservative learning rates.

36
“Electric power annual”, Energy Information Administration, 2019.

53 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
Exhibit 39
Integrated utility on an isolated US island, net present value
Main LDES application(s) Value Net value Cost Accessible value with market mechanisms in place

NPV for an integrated utility on an isolated US island

Customer profile USD millions

Integrated utility on a US island without

interconnection to the mainland  Stability services provision ~100

Relies on carbon-intensive electricity

generation sources  Production cost savings 6,600 Market
mechanism
Step-wise build-out of RE and storage, incl in place
 CO2e cost savings 3,300
Li-ion and LDES, through a realistic low-cost
scenario
Total value creation ~6,700–10,000
Potential LDES system: 1.3 GW/104 GWh
(80 hours)
 Capital invested 5,400

 Fixed O&M 700

IRR
NPV ~600–3,900 7–12%

Assumptions Final investment decision date Commercial operation date CO2e price scenario WACC
2023 (RE and Li-ion) 2025 (RE and Li-ion) Base case 6%
2028 (RE, Li-ion and LDES) 2030 (RE, Li-ion and LDES)

The resulting IRR of this case example is 7 to The main value streams from the RE hybrid
12 percent, with an NPV of USD 500 million to storage system are production cost savings
USD 3.9 billion. The results imply that for some and CO2e cost savings, which have projected
situations, the business case will be “in the present values of approximately USD 6.6 billion
money” and potentially attractive for investors, and USD 3.3 billion, respectively.
strongly driven by fuel costs and RE potential.
Multiple sensitivities materially influence the
However, for other instances with a gap to
financial viability of this case, namely shifted
viability, it is vital to ensure the value capture of
operation dates, CO2e price, fuel costs, and
CO2e cost savings.
RE capex. The IRR increases significantly to
The resulting IRR of this case example is 7 to between 11 and 17 percent when the LDES
12 percent, with an NPV of USD 500 million to commercial operation date moves to 2035, with
USD 3.9 billion. The results imply that for some the construction of the system taking place in the
situations, the business case will be “in the two years prior. Furthermore, accelerated CO2e
money” and potentially attractive for investors, price increases could result in IRRs of 15 and
strongly driven by fuel costs and RE potential. 20 percent, with LDES operation dates by 2030
However, for other instances with a gap to and 2035, respectively. However, the IRR would
viability, it is vital to ensure the value capture of significantly drop in islands with lower fuel costs
CO2e cost savings. or less advantaged RE potential. A fuel cost USD
50 per MWh lower than the base would reduce

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 54
 the IRR by approximately 3 percent. A similar
effect occurs when RE capex increases by
50 percent. (Exhibit 40)

To ensure the financial viability of LDES for

integrated utilities on isolated islands, multiple
actions could be considered. These actions
include creating market mechanisms for CO2e
benefit remuneration, providing options to
lower WACC, ensuring CO2e price stability, and
facilitating the traceability of energy for LDES
charging.

Exhibit 40
Integrated utility on an isolated US island, IRRs
IRR sensitivity to contract durations and project start date1 Base case <5% 5–10% 10–15% 15–20% >20%

High
100 7–15% 11–20%
CO2e price
Medium
scenario 75 7–13% 11–19%
2030 USD/ tCO2e
Base
60 7–12% 11–17%

200 11–16% 17–22%

Fossil fuel cost
150 7–12% 11–17%
USD/ MWh
100 3–9% 5–13%

0 7–12% 11–17%
RE capex
sensitivity 50 5–10% 8–14%
% increase
100 4–9% 6–11%
2030 2035
Commercial operation date
1. Lower end of range for value capture in markets with appropriate mechanisms; higher end of range for full value potential.

55 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 3. Remote copper mine in Chile allows the mining company to reduce production
costs—relative to their onsite diesel generation—
For an industrial customer looking to and related CO2e emission costs.
reduce electricity production costs
and decrease their carbon footprint, Detailed modeling of a RE hybrid storage system
deploying LDES could be financially for the mine indicates that the lowest cost
attractive in the near term pathway to 100 percent fulfillment of energy
demand via RE employs a combination of Li-ion
A mining company operating a remote Chilean
and LDES. The LCOE of this configuration is 10
copper mine currently relies on onsite diesel
to 15 percent lower than a pure Li-ion battery
generators for a stable electricity supply. The
system and 5 percent lower than a pure LDES
resulting electricity costs represent 10 percent
system.
of the total mining and processing costs.
Reliability of the onsite electricity system is also The deployment of the RE hybrid storage system
critical because of the high opportunity costs could take place in a multiphase buildout, similar
of power outages (12 hours of power outages a to the isolated island system.
year would translate into USD 1.5 million in lost
revenue). (Exhibit 41) The resulting IRR of this example
is 15 to 19 percent, with an NPV
Increasing cost pressure, together with of USD 1.5 billion to 2.7 billion,
decarbonization ambitions and reducing indicating that this type of LDES
renewable LCOEs, pushes the company to application as part of RE hybrid
consider switching to RE and, subsequently, storage systems can be “in the
storage to ensure fully RE generation. money” in the near term.
Like the isolated island integrated utility, the Further acceleration of investments could be
mining company could consider installing a driven by rising ambition levels of corporate
hybrid of additional solar and wind with Li-ion decarbonization targets.
batteries and LDES. This hybrid configuration

Exhibit 41
Remote Chilean copper mine case example, net present value
Main LDES application(s) Value Net value Cost Accessible value with market mechanisms in place

NPV for a remote copper mine in Chile

Customer profile USD millions

Mining company operating a remote copper

mine in Chile  Production cost savings 3,900

High electricity costs with current power

supply from diesel generators  CO2e cost savings 1,200
Market
mechanism
Step-wise build-out of RE and storage, incl 3,900– in place
Total value creation
Li-ion and LDES, through a realistic low-cost 5,100
scenario
 Capital invested 2,300
Potential LDES system: 0.7 GW/56 GWh (84
hours)
 Fixed O&M 100–200

IRR
NPV 1,400–2,700 15–19%

Assumptions Final investment decision date Commercial operation date CO2e price scenario WACC
2023 (RE and Li-ion) 2025 (RE and Li-ion) Base case 6%
2028 (RE, Li-ion and LDES) 2030 (RE, Li-ion and LDES)

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 56
 The main value streams from the RE hybrid considerably increases the IRR between 19 and
storage system are production cost savings 23 percent. In addition, accelerated CO2e price
and CO2e cost savings, which have projected increases could result in IRRs of up to 21 and 26
values of approximately USD 3,900 million percent, with LDES operation dates by 2030 and
and USD 1,200 million, respectively. Also here, 2035, respectively. For mines with simpler fuel
monetization of CO2e cost savings requires an logistics and lower RE potential, the IRR could
adequate market mechanism. (Exhibit 42) drop by 3 percent, for example, due to a USD 50
per MWh lower fuel cost than the base case or a
Like the isolated island situation, a range of
RE capex increase of 50 percent. However, the
sensitivities on IRRs have been evaluated,
IRR would still be significantly above the WACC
particularly a shifted operation date, CO2e price,
of 8 percent, indicating a more robust business
fuel cost, and RE capex. A shift in the LDES
case for this application.
component commercial operation date to 2035

Exhibit 42
Remote Chilean copper mine case example, IRRs
IRR sensitivity to contract durations and project start date1 Base case <5% 5–10% 10–15% 15–20% >20%

High
100 15–21% 19–26%
CO2e price
Medium
scenario 75 15–20% 19–24%
2030 USD/tCO2e
Base
60 15–19% 19–23%

300 19–22% 23–27%

Fossil fuel cost
250 15–19% 19–23%
USD/MWh
200 12–16% 14–19%

0 15–19% 19–23%
RE capex
sensitivity 50 14–17% 16–21%
% increase
100 12–16% 15–19%
2030 2035

Commercial operation date

1. Lower end of range for value capture in markets with appropriate mechanisms; higher end of range for full value potential.

57 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company
 4. On-demand RE peak increase over current RE levels. Hence, meeting
power in India this target while supplying the increased demand
and managing the grid stability is expected to
The business case for a RE developer require the deployment of innovative solutions
in India providing peak and off-peak and flexibility resources like LDES.
supply with a combination of RE and
PSH could show attractive IRRs and Given this context, a deployed LDES system can
more competitive PPA tariffs than provide different services. This case explores
fossil generation procurement a RE developer deploying LDES to enable
dispatchable RE with a front-of-the-meter
Decarbonizing India’s power supply requires
contract. Specifically, a business case of LDES
the widespread deployment of flexibility
to enable dispatchable peaking capacity within
solutions such as LDES. Electricity demand
specific contracted hours of the day, is modelled.
is expected to rise sharply as more end-uses,
A 6-hour system is considered given its suitability
such as heating and transportation, electrify,
to a solar generation profile (solar PV is expected
renewable hydrogen production expands, and
to be increasingly deployed in India in the near
living standards increase. In 2020, nearly 70
term). The 300 MW and 1,800 MWh novel PSH
percent of India’s power generation mix was
system is assessed in combination with 600 MW
thermal, with coal accounting for 85 percent.
of hybrid solar PV and wind capacity. (Exhibit 43)
With more than 75 GW of installed capacity, RE
accounted for nearly 20 percent of the mix, with The case example shows an IRR of
solar seeing the fastest growth. At the Glasgow approximately 10 to 12 percent, indicating
climate change conference, India committed to short-term financial attractiveness for potential
reach 500 GW of non-fossil generation capacity infrastructure investors. Peak power supply
by 2030, representing a nearly 500 percent shows a present value of about USD 700–800

Exhibit 43
India RE developer case example
Main LDES application(s) Value Net value Cost Accessible value with market mechanisms in place

NPV for a renewable developer

Customer profile USD millions

RE developer in India providing morning and Different peak

evening peak supply as well as off-peak  Peak power supply ~700–800 and off-peak
generation with a combination of RE and PSH tariffs assumed
 Off-peak power supply ~650–750
Providing lower cost PPA tariffs compared to
the case of procurement from thermal power1
• Stability services provision2 ~20–40
LDES asset (ie, PSH) enables dispatchable
peaking capacity Total value creation ~1,300–1,500
Modeled LDES configuration: 300 MW/1800
MWh (6 hours); in combination with 600 MW  Capital invested ~1,000
of contracted hybrid RE1 (solar and/or wind)
• Fixed O&M ~250

IRR
NPV ~100–300 10–12%

Assumptions Final investment decision date Commercial operation date WACC RE cost outlook PPA contract duration
2021 2023 10% Progressive3 20 years4

1. Corresponding to ~870 MW of DC installed capacity

2. Inertia provided by the PSH, remuneration based on NGESO’s Stability Pathfinder mechanism
3. Capex figure for PSH takes an industry estimate for off-stream closed-loop systems in India and NREL-ATB Advanced costs for RE
4. After contract end, value of service ~50% of in-contract value

Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company 58
 million, while off-peak power generation two years prior. This behavior is explained by
contributes to a comparable extent, with NPV the high value of storage solution in the next
values approximately between USD 650 million decade, able to mitigate higher price peaks and
and USD 750 million. The revenues streams larger electricity spreads caused by a higher RE
are similar as the higher PPA tariff during peak penetration. The value of green dispatchable
periods is balanced by a smaller yearly energy PPA contracts will increase in value in the
supply (around one third of the total generation). upcoming decade, heavily contributing to the
profitability of this business case. Furthermore,
Market remuneration of inertia for power system
novel PSH systems costs in India could see
stability is not a currently existing value stream
a more rapid cost down trajectory, enabling
in India, but could be worth about USD 20-40
significantly higher returns than in other regions.
million alone with similar mechanisms under
implementation in other countries. (Exhibit 44) Several factors could help unlock additional
value potential—for example, an increased
Both the project starting date and the LDES
spread of electricity prices leading to more
storage CAPEX influence the project IRR. The
valuable PPA contracts, increased demand for
IRR increases significantly to approximately
clean peaking dispatchable power, or sustained
22 to 25 percent for the base case when the
long contract durations as penetration of
commercial operation date moves to 2030,
renewable increases.
with the construction of the system taking place

Exhibit 44
India RE developer case example, IRRs
IRR sensitivity to contract duration and project start date Base case <5% 5–10% 10–15% 15–20% >20%

Progressive 10–12% 14–16% 22–25%

LDES storage
Central 9–11% 13–15% 20–22%
CAPEX1

Conservative 6–8% 9–11% 13–15%

2023 2027 2030

Commercial operation date

1. India industry costs perspective

59 Net-zero power: Long duration energy storage for a renewable grid | LDES Council, McKinsey & Company