DNV ETO 2021 Executive Summary Single Page Highres

ENERGY
TRANSITION
OUTLOOK
2021
EXECUTIVE
SUMMARY
A global and regional forecast to 2050

DNV Energy Transition Outlook 2021
FOREWORD
This is the fifth annual Energy Transition Outlook issued by DNV. For half a
decade, we have said, consistently, that the most likely future for the world’s
energy system is one that will result in global warming exceeding 2°C by
2100, and that is our conclusion once again this year.
The impact of global warming is becoming alarmingly Not everything can be electrified. That is what makes
apparent, and I believe there is a widening understand- tackling the hard-to-abate sectors of high heat, aviation,
ing of the long-term risks for humanity. However, while shipping and trucking so very urgent. Yet our forecast
perceptions may have changed, reality has not. shows that hydrogen enters the picture at scale only in
Each year, in the foreword to this Outlook, I have the late 2030s. That is far too late: climate science points
stressed the need for governments and companies to to the considerable risks of allowing emissions to
take decisive action on climate change. COVID-19 has accumulate before we act.
more than demonstrated that governments can act
boldly. Yet, from an energy transition perspective, the The verdict is clear: the world needs vastly more green
pandemic has been a lost opportunity. Recovery electricity, both direct and indirect, more biofuel, and
packages have largely focused on protecting rather more carbon capture and storage on a dramatically
than transforming existing industries. There are excep- accelerated timescale.
tions to this, and our forecast incorporates slightly more
clean energy in the mix over the next three decades than In October this year we will publish our first Pathway to
we did a year ago. net zero emissions report – a detailed look at how best
to close the gap between this forecast and one that is
But large-scale action is still needed urgently, and aligned with the Paris Agreement. It is vital that we
our forecast provides clear guidelines on where such mobilize all the forces of the Fourth Industrial Revolution
efforts should be directed. Wind and solar PV will towards a green energy transformation, including
expand 15- and 20-fold respectively in our forecast innovative ways to finance this shift. This year we
period. Twinned with the plunging costs and advancing address that critical subject in our supplementary
technology of battery storage, variable renewables report – Financing the Energy Transition.
are already enabling a phase out of thermal power
generation and the business case will become
overwhelming by 2030.
Electricity demand will more than double by 2050 and

by then over 80% of power will be provided by non-
fossil sources. The accompanying efficiencies are
staggering, both in the avoidance of heat losses in
Remi Eriksen
power generation and in end use – for example, with
EVs and heat pumps. But the problem is this: even if all Group president and CEO
electricity was ‘green’ from this day forward, humanity
DNV
would still fail to achieve net zero emissions by 2050.
2
Foreword
Highlights
CORE INSIGHTS
1. We are not meeting Paris ambitions; there is a very short window to close the gap
− Global energy-related emissions will fall only 9% by 2030, and the 1.5˚C carbon budget is emptied by then
− We estimate the global average temperature increase to reach 2.3˚C by end of the century
2. Electrification is surging ahead, and renewables will outcompete all other energy sources
− Electrification of final energy demand will grow from 19% to a 38% share by 2050, powered
mainly by solar and wind
− 50% of all passenger-vehicle sales will be EVs in 2032
− Heat pump use will triple, providing 32% of heat in 2050 while consuming 9% of energy use for heating
3. Efficiency gains lead to a flattening of energy demand from the 2030s

− Energy efficiency remains our greatest untapped resource against climate change
− Energy intensity (unit of energy per dollar of GDP) improvements at 2.4%/yr outpace GDP growth
during the coming three decades
− Efficiency gains are driven mainly by electrification
4. Fossil fuels are gradually losing position, but retain a 50% share in 2050
− Gas maintains its current position, oil demand halves, and coal falls to a third of current use by 2050
− CCS deployment is too slow, and only 3.6% of fossil CO2 emissions are abated in 2050
NEW INSIGHTS 2021
1. COVID-19 economic recovery spending is a lost opportunity

− Apart from the EU, COVID-19 stimulus packages are largely locking in carbon-intensive systems
2. Variability and low power prices are not roadblocks to a renewable-based power system
− Power-to-X, storage, connectivity, demand response, and carbon pricing will all help solar PV and
wind maintain their competitiveness
− Solar + storage is emerging as a new power plant category which will provide 12% of all
grid-connected electricity by 2050
3. Decarbonizing hard-to-abate sectors requires far greater scaling of hydrogen, e-fuels, and biofuels
− Combined, hydrogen and e-fuels will cover only 5% of global energy demand by 2050
− Aviation, maritime, and heavy industry increase their relative share of emissions and remain heavy
users of unabated fossil fuels
4. Most hydrogen will be produced from dedicated renewables-based electrolysers by 2050

− Green hydrogen will dominate over time, with 18% of hydrogen supply produced via electrolysis from cheap
grid electricity and 43% from electrolysis using dedicated off-grid renewables
− Blue hydrogen will lose its cost advantage, providing only 19% of hydrogen supply for energy purposes
by 2050
3
Highlights – Core insights

We are not meeting Paris ambitions; there is a short Electrification is surging ahead, and renewables will
window of opportunity to close the gap outcompete all other power sources
Global emissions likely peaked in 2019, followed by an Electrification is by far the most dynamic element of the
unprecedented 6% drop in 2020 due to COVID-19. energy transition. The share of electricity in final global
Emissions are now rising sharply again and will grow for energy demand is set to double from 19% to 38% within
the next three years before starting to decline. the next 30 years.
While they are being added at great speed, renewables Solar PV and wind are already the cheapest form of new
currently often supplement rather than fully replace power almost everywhere, and within a decade will also
thermal power generation. By 2030, global energy-related be cheaper than operating existing thermal power in
CO2 emissions are likely to be only 9% lower than 2019 most places. By 2050, solar and wind will represent 69%
emissions, and by 2050 only 45% lower. This is in sharp of grid-connected power generation, and fossil power
contrast to ambitions to halve GHG emissions by 2030 and just 13%. Connectivity, storage and demand-response
to achieve the net zero emissions by 2050 required to limit will be critical assets in the decarbonized power system.
global warming to 1.5˚C. Our forecast is that we are most
likely headed towards global warming of 2.3˚C by 2100. On the demand side, passenger and commercial EV
uptake is rising quickly in Europe, China and to some
As CO2 emissions continue to accumulate, the window of extent the US. Government incentives, cost reductions
opportunity to act narrows every year. Relying on large- and technology improvements in both batteries and
scale net-negative emissions technologies and carbon charging infrastructure will drive a rapid expansion.
removal in the latter half of the century is a dangerous, By 2032, half of all new passenger vehicles sold globally
high-risk approach. With global warming, every fraction will be electric, with some regions lagging owing to
of a degree is important, and all options to reduce infrastructure challenges. In buildings, heat pumps use
emissions need urgent realization. will triple, providing 42% of space heat in 2050 while
consuming only 15% of energy used for space heating.
4
Executive summary
Efficiency gains lead to a flattening of energy demand Fossil fuels are gradually losing position, but retain a
from the 2030s 50% share in 2050
Energy efficiency is the unsung hero of the energy transi- Fossil fuels have held an 80% share of the global energy
tion and should be the number one priority for companies mix for decades. We forecast that, by mid-century, fossil
and governments. Many efficiency measures have fuels will decrease, but still hold a 50% share of the
marginal or even negative costs, but due to split incentives energy mix, a testament to the inertia of fossil energy in
and/or a lack of long-term thinking, industry standards and an era of decarbonization.
regulations are needed to ensure implementation.
Coal use will fall fastest, down 62% by 2050. Oil use stays
Energy intensity (unit of energy per dollar of GDP) improve- relatively flat until 2025 when it starts a steady decline, to
ments will average 2.4%/yr during our forecast period just above half of current levels by mid-century. Gas use
– against the 1.7%/yr average over the last 20 years. Most of will grow over the coming decade, then levels off for a
the accelerated efficiencies are linked to electrification, 15-year period before starting to reduce in the 2040s.
with the remainder coming largely from efficiency Gas will surpass oil as the largest energy source and will
improvements in end uses, such as better insulation. The represent 24% of global energy supply in 2050.
largest efficiency gains happen in the transport sector, but
there are significant gains also in manufacturing and Decarbonized fossil energy is an important aspect of
buildings. reaching the Paris Agreement, but the uptake of carbon
capture and storage (CCS) is forecast to be woefully slow,
Overall efficiency gains will result in a levelling off in global mainly for reasons of cost, with just 3.6% of fossil CO2
energy demand despite a population increase of 22% and emissions abated in 2050.
the global economy growing 111% over the next 30 years.
Global energy demand will grow only 8% from 2019 to
2035, thereafter remaining essentially flat the next 15 years.
5
Highlights – New insights

COVID-19 economic recovery spending is a lost Variability and low power prices are not roadblocks to a
opportunity renewable-based power system
Government interventions, to stop the spread of the The present power system is not set up for variable
virus and then to restart activity, revealed how effective renewables as the dominant source of production.
national and global actions can be. Similar action and Yet plunging costs, government support for renewable
funding have yet to be applied to the unfolding global power buildout, and carbon pricing will ensure that
climate crisis. renewables will eventually dominate power generation.
Over the coming 30 years, USD 12 trillion will be invested
The trillions of dollars pushed into the global economy in both building a larger grid and adapting it to the
over the past 20 months have mainly been directed variability of solar and wind through technical solutions
towards emergency measures like wage supplements such as connectivity, storage, and demand response.
and on building back the existing economic and indus-
trial engine. Yet the opportunity for a green reset of The cost of power from solar and wind will continue to
production, transport and economic activity was unique, reduce but price cannibalization threatens the investment
and as we wrote in ETO 2020, “The post-COVID-19 case for renewable capacity if cheap power is unused at
stimulus packages hold the potential to alter the speed of times of ample supply. However, indirect electrification
the transition.” With some notable exceptions, particularly through power-to-X will require massive renewable
in the EU, governments have not steered recovery electricity production, and along with various storage
spending towards a decarbonized outcome. solutions, will ensure that surplus power will be used,
and capture prices maintained at a satisfactory level.
Global CO2 and GHG emissions fell 6% in 2020 but will rise
again this year. While the emissions trajectory has shifted Solar PV + storage will make solar more directly competi-
down slightly, that is due to lost economic activity, not tive with thermal generation, nuclear and hydropower.
energy-system renewal. The overall pace of the transition We find that one third of all solar production will be built
has not accelerated, and that is a lost opportunity. with direct storage, and by 2050, solar PV + storage will
produce 12% of all grid-connected electricity.
FIGURE 5
Pandemic recovery spending 2020
18%
USD 341bn
USD 1,553bn
82%
Source: UNEP report ‘Are we building back better’, Feb 2021

Note: In a separate report the IEA’s Sustainable Recovery Tracker
estimated that as of Q2 2021, clean energy measures totalled
USD380bn or just 2% of total fiscal support related to COVID-19
6
Executive summary
Decarbonizing hard-to-abate sectors requires far Most hydrogen will be produced from dedicated
greater scaling of hydrogen, e-fuels, and biofuels renewables-based electrolysers by 2050
Hard-to-abate sectors are those that cannot easily be The current production of hydrogen as an energy carrier
decarbonized through electrification, and include is negligible compared with the 75m tonnes of grey/
aviation, maritime, long-haul trucking and large parts of brown hydrogen produced annually for fertilizer and
heavy industry. These sectors are currently responsible chemicals production.
for around 35% of global CO2 emissions, and progress in
reducing these emissions is stubbornly slow. Blue hydrogen, produced by steam methane reforming
(SMR) from gas with CCS, will replace some of the grey
Hydrogen is seen as the main decarbonization alternative and brown hydrogen in the coming decades. In total,
for these sectors, with biofuels in a supporting role, mainly blue hydrogen will also comprise 18% of hydrogen
in aviation. Direct hydrogen use is often not suitable, and supply for energy purposes by 2050.
ships and aircraft require hydrogen derivatives and
e-fuels such as ammonia and synthetic jet fuel. Green hydrogen from electrolysis will be the main
long-term solution for decarbonizing hard-to-abate
Global hydrogen production for energy purposes is sectors, including hydrogen as a basis for other e-fuels.
currently negligible and will only start to scale from the
late 2030s, meeting 5% of global energy demand by Electrolysis powered by grid electricity is disadvantaged
2050. Government incentives, similar to those given to by the limited number of hours of low-priced electricity.
renewables, are needed to stimulate technology develop- Its CO2 footprint will, however, improve as more renewa-
ment and accelerate uptake of hydrogen and e-fuels. bles enter the power mix. The future production of
hydrogen for energy purposes will be dominated by
Aviation, maritime and heavy industry thus retain high electrolysis using dedicated off-grid renewables, such as
unabated fossil-fuel shares towards 2050, slowing the solar and wind farms. By 2050, 18% of hydrogen will be
transition and significantly impeding the achievement of grid-based and 43% will come from dedicated capacity
the Paris Agreement. comprising solar PV (16%), onshore wind (16%) and
fixed offshore wind (9%).
7
PRIMARY ENERGY 2019-2050

Shown here is our forecast for primary energy through to industry’s own use of energy, which is considerable –
2050. Primary-energy supply is the total amount of typically 7% of primary energy consumption. Primary
energy behind the provisioning of energy services. energy, which was 594 EJ before the pandemic, will return
Considerable losses occur in the conversion and transport to 2019 levels in 2022, but will then only increase by a
of energy (currently exceeding 100 EJ annually) and these further 4% percent and peak at 617 EJ in 2030 before
are included in primary energy numbers, as is the energy slowly reducing by 4% to some 590 EJ in 2050.
World Energy Supply Transition 2020–2050

Units: Exajoules (EJ) per year
160
140
120
100
Wind Solar
80 Bioenergy
60
40
Hydropower
20
0
2019 2050 2019 2050 2019 2050 2019 2050
8
Executive summary
Contributions to changes in the primary-energy supply 50% by mid-century. Nuclear will be stable at 5% over the
are shown in the figures below. In the forecast period, entire period, while renewables will triple from 15% today
renewables will constantly be adding to the primary- to 45% by the end of this forecast period.
energy supply, while fossil fuels will be reducing, except
for natural gas, which only decreases in the 2040s. The
fossil share of the energy mix will fall from 80% today to
Oil
Natural gas Coal
Nuclear
Geothermal
2019 2050 2019 2050 2019 2050 2019 2050 2019 2050
9
ENERGY TRANSITION ENERGY TRANSITION ENERGY TRANSITION ENERGY TRANSITION

OUTLOOK 2017 OUTLOOK 2018 OUTLOOK 2019 OUTLOOK 2020
After five years of About this Outlook

This annual Outlook presents the results from our
forecasting… independent model of the world’s energy system. It
covers the period through to 2050 and forecasts the
Our findings this year do not differ fundamentally to energy transition globally and in 10 world regions. Our
those of our first forecast issued four years ago. Over the forecast data may be accessed at eto.dnv.com/data.
years, we have extended and refined our model, for More details on our methodology and model can be
example by delving deeper into the dynamics of the key found on page 34. The changes we forecast hold signifi-
demand sectors of transport, manufacturing and build- cant risks and opportunities across many industries.
ings. We have also added additional energy carriers and Some of these are detailed in our supplements:
sectors such as hydrogen, floating offshore wind and
solar PV+storage, and introduced the modelling of — Maritime forecast
power generation on an hourly basis. But one of our key — Financing the energy transition
findings – that the global energy mix will be split in — Technology progress report
roughly equal shares between fossil and non-fossil — Pathway to net zero emissions
sources by 2050 – has not changed. Neither has our
conclusion that the world will fail to achieve the climate Independent view
goals of the Paris Agreement by an alarming margin. DNV was founded 157 years ago to safeguard life,
property and the environment. We are owned by a
The fact that these findings have remained consistent is a foundation and are trusted by a wide range of customers
major cause of concern: In a half-decade where the costs to advance the safety and sustainability of their busi-
of inaction on climate change have been mounting and nesses. 70% of our business is related to the production,
the evidence of its effects are growing ever more visible, generation, transmission and transport of energy.
it is sobering to reflect on the fact that the pace of the Developing an independent understanding of, and
energy transition has not accelerated beyond our first forecasting, the energy transition is of strategic impor-
forecast. tance to both us and our customers.
10
Executive summary
The impact of COVID-19 workplaces will reduce but that will be counterbalanced by
a rise in residential energy demand. In developed regions
The impact of the pandemic is beginning to recede in
the increase is 4%, half of which arises from bigger dwelling
developed countries where half the adult population will
sizes and half from increased requirements for heating
be fully vaccinated by the end of Q3. Vaccination in the
and cooling.
developing world lags far behind; some countries have
barely begun their rollouts. GDP impact differs strongly
We expect a permanent drop in work-related air travel of
across world regions, but overall, we follow the IMF’s
20% compared with our pre-pandemic forecast. Leisure
expectation (April 2021) of a rebound of 6.1% in 2021 and
travel, which account for two thirds of air miles, is likely to
4.9% in 2022. COVID-19 results in a 2023 global economy
rebound fully by 2024.
that has lost 4.1% of GDP compared with pre-pandemic
projections. This loss is almost permanent, but the post
COVID-19 boost in 2023 will result in some regional
economies growing slightly faster than they otherwise
would have, and in 2050 the loss declines to 3.2%. Very little of the COVID-19 spending which
has a bearing on energy has been steered
From an energy perspective, the pandemic will leave
certain lasting effects in its wake. Working from home will towards decarbonization.
intensify, reducing the demand for workspace by 5%
compared with our pre-pandemic estimates in developed
regions (including China), and by half of that in developing
In the early months of the pandemic many were hopeful
and emerging economy regions. The energy intensity of
that it would provide an opportunity to reset behaviours
and economic activity towards greener outcomes. So
far, that has not proved to the be the case. Of the near
USD 20 trillion in COVID-19 related spending, most has
been allocated to emergency measures that lock in the
dynamics of the pre-pandemic energy and manufacturing
industries. Very little of the COVID-19 spending which
has a bearing on energy has been steered towards
decarbonization, although Europe has proved to be an
exception. While comparatively marginal, this new
spending on clean energy is contributing to slightly
more non-fossil energy into our long-term forecast. It is
nowhere near enough, however, to make the substantial
breakthrough required for achievement of a net zero
energy system by 2050, and thus we regard COVID-19
as a lost opportunity for accelerating the pace of the
energy transition.
11
DEMAND
Buildings Manufacturing
The buildings sector will collectively consume 26% more Manufacturing energy demand, at 131 EJ in 2019, is
energy in 2050 than in 2019, growing its share from 28% to forecast to grow by 8% by 2050. The economic output
one third of global energy use. from production of base materials, construction, manu-
factured goods and mining will grow by 75%, indicating
A growing, more prosperous population will result in a an efficiency improvement of 1.6%/yr.
rapid expansion in floor space (both commercial and
residential) of 62%. Space cooling will quadruple over the The manufacturing sector had the largest share (30%)
next three decades, driven by increases in standards of of final energy demand in 2019. The base materials
living and by climate change. Space heating demand on subsector was responsible for 38% of manufacturing
the other hand will reduce by 17%, owing to considerable energy use. Manufactured goods was responsible for 31%
efficiencies brought by electrification and heat pumps, of energy use, followed by iron and steel at 26%, and by
which provide much more useful energy as heat than they construction and mining at 5%. Substantial energy-
consume as electricity. Energy demand from appliances efficiency gains, including increased recycling, will
and lighting will double – tracking slightly below GDP balance the growth in demand for goods, such that goods
growth, due to moderate efficiency gains from the design manufacturing energy use will grow by 6% to 2033 and
of end use products. Other end uses such as cooking and remain flat to 2050. The base materials sector will see
water heating will stay relatively stable, as efficiency gains, initial growth to 2032, but reduce by a third thereafter to
particularly the shift to modern cooking methods, 2050, due to increased recycling and efficiency gains.
balance out any additional demand.
Manufacturing dominates demand sector emissions at
Buildings are currently responsible for 25% of energy- 12 GtCO2/yr (35% share). These will exactly halve over
related emissions, including indirect emissions from the forecast period owing to reduced coal use and a
electricity and direct heat production. By mid-century progressively greener electricity mix.
these emissions will decline in absolute terms by 44%,
owing to cumulative efficiencies and a greener energy mix Energy use for heating purposes will see the largest
that includes substantially more electricity. efficiency gains towards 2050, due to changes in fuels and
12
Executive summary
the more widespread use of heat pumps. With greater Transport

automation and digitalization, the machines, motors & Efficiency gains and fuel switches – mainly to electricity
appliances end-use will see its share of manufacturing and hydrogen – will result in transport energy demand
energy demand increase to 32%, up from 24% in 2019. falling from 125 EJ in 2019 to 111 EJ in 2050. Oil will supply
just above half of this demand by mid-century, with
Feedstock electricity supplying a quarter, hydrogen 10%, and
In 2019, roughly 8% of global primary fossil-fuel supply biofuels and natural gas 7% each. That is a very different
was used for non-energy purposes, including 13% of oil. picture from today, with transport at 29% of global final
Petrochemicals are the largest consumer of feedstock energy demand, almost entirely in the form of fossil fuels.
and, of the consumption in this sector, about 45% was 92% of road transport energy use is oil, with biofuels and
used to produce plastics in 2019, with the rest going to the natural gas taking 3% and 4% shares respectively and
manufacture of cosmetics, fertilizers, paints, and other electricity 1%. The energy mix in aviation and maritime is
chemicals. We expect that in 2050 the plastic proportion fairly similar; only rail is supplied mainly by electricity.
will have grown to about 61% of petrochemical feedstock
demand. Transport services will grow significantly in the next 30
years: the passenger vehicle fleet will expand by two thirds
While plastic demand continues to grow to 2050, reuse, reaching 2 billion units by 2050; aviation demand will
substitution and especially recycling grows more rapidly recover from the pandemic – although not without effects
– encouraged by taxes levied on unrecycled plastic. We on job travel – with global passenger flights growing 130%
estimate the global rate of plastic recycling will improve from 4.4 billion flights in 2019 to reach 10.2 billion flights in
from around 13% in 2018 to 47% in 2050 as it is bolstered 2050; in the maritime industry, cargo tonne-miles will rise
by more efficient (and potentially circular) chemical by nearly a third. Despite all this additional activity,
recycling, which supplements or replaces traditional, transport energy demand will fall by 13% by 2050.
mechanical recycling. That is a major reason for non-
energy feedstock use peaking in 2032 and then declining This counterintuitive development is mainly due to
sharply towards 2050. efficiency improvement associated with the electrification
of road transport, supported by efficiency gains in
13
DEMAND
aviation and maritime. By 2042, half of all passenger Commercial vehicles require much larger batteries, and
vehicles on the road will be electric. we expect significantly higher and more-prolonged
subsidy levels per vehicle in OECD regions and Greater
Road transport China, where ICEVs will also be made less attractive
President Biden’s recent fuel-efficiency plan includes a through higher carbon prices.
target that every other car sold in the US by 2030 will be an
EV. Our forecast indicates that the goal is not farfetched The rate of EV uptake will differ regionally, but in all
and will be achieved by 2031 in the North America region. regions the 50% of new sales figure will be reached before
Indeed, Europe and Greater China will have already mid-century, and globally half of the passenger vehicle
reached that milestone in 2027 and 2028 respectively. fleet will be electric by 2042. By 2050, they will consume
just 36% of the road subsector’s energy demand.
Driving the surge in EV sales in those regions is a combi- Conversely, oil will still account for 60% of road transport
nation of factors, including a range of subsidies and other energy, fueling just 30% of passenger and 59% of
preferential treatments for owners. On the technology commercial vehicles, owing to the much lower efficiency
side, the average range of EVs is increasing, charging of the combustion engine.
speeds are rising and charging infrastructure is expanding.
Some 16% of the commercial EV fleet in the OECD region
With better chemistries, manufacturing processes and and Greater China by 2050 will be fuel-cell electric
pack designs, battery costs will keep plunging. By the vehicles (FCEVs). The combination of electricity, hydrogen
middle of this decade, EVs will have demonstrably and biofuel will chase 21 million barrels per day of oil out
reached total cost of ownership (TCO) parity with internal of the road transport energy mix by 2050 from its peak in
combustion engine vehicles (ICEVs), and sales will rise 2019. In sum, our forecast indicates a rapid and significant
dramatically but not uniformly across regions: uptake will electrification of all parts of road transport, in all regions.
be hampered by availability of electricity and infra-
structure challenges, particularly in developing regions.
14
Executive summary
Maritime Aviation
Nearly 3% of the world’s final energy demand, and 7% of Aviation consumes over 3% of the world’s energy, almost
the world’s oil, is presently consumed by ships, mainly by entirely in the form of oil. By 2050, oil still accounts for 53%
international cargo shipping. By 2050, the dominance of of aviation fuel, but in absolute terms oil use will be 26%
oil in the fuel mix will have been displaced (42%) by low- lower than today. Following recovery from the pandemic
and/or zero-carbon fuels like ammonia, hydrogen, and demand slump, the number of flights will rise steadily in
other electro-based fuels such as e-methanol. Natural gas line with GDP growth to a level 140% higher in 2050
– mostly LNG – will take a 39% share. compared with 2019. Fuel use, however, increases by just
40% owing to efficiency gains in aircraft and engine
Driving this fundamental fuel switch are the IMO targets technology and in logistics.
for a 50% absolute reduction in GHG emissions from 2008
to 2050. Our forecast assumes that a mixture of improved The three main options to replace oil-based fuel in
utilization and energy efficiencies, combined with a aviation are: electricity, hydrogen and sustainable aviation
massive fuel decarbonization, will see this goal being met. fuel (SAF). All three are and will remain more expensive
than oil-based fuels, and fuel and related technological
The digitalization of logistics and supply-chains will changes are expected to be driven by regulatory and
increase fleet efficiencies. However, in a world with GDP consumer-supported forces. Electricity, limited to the
doubling, cargo transportation needs will outweigh short-haul segment, takes just a 2% share of the aviation
efficiency improvements, and tonne-miles will increase by fuel mix in 2050. Cost, technology and regulatory chal-
32%. The exception is coal and oil transport, reducing by lenges will see hydrogen-powered airplanes in use only
more than 50% and 20%, respectively, by 2050. after 2040 in the first few regions.
The energy density of batteries both today and in the SAFs represent the most feasible decarbonization route,
future is likely to remain too low to play a larger role in with biofuels reaching a 26% share of the aviation mix by
deep-sea shipping. For further details on the fuel mix and 2050, and e-fuels (plus pure hydrogen) 19% - a pace of
use, refer to our Maritime Forecast companion report. decarbonization running ahead of the CORSIA goals.
15
ENERGY CARRIERS
Electricity Towards 2030, demand growth and an expansion of VRES
Electrification is pivotal to the ongoing energy transition. will see a steady increase in grid investments, rising by
Electricity demand will more than double between now between USD 150-200bn/yr from pre-pandemic levels. In
and 2050, with the share of variable renewable energy terms of circuit-km, transmission lines will double and
sources (VRES) in the power mix growing from 8% today to distribution lines more than double by 2050. Some 15% of
reach 69% of power generation in 2050. grid investments will be steered towards digital infrastruc-
ture, to address the complexity of a more decentralized
Growing at almost 3% per year to reach some 60,000 power system.
GWh in 2050, electricity demand will outpace economic
growth despite continuous efficiency improvements. This With rapid retirements from 2026 onwards, coal will fall to
is due to vast new categories of demand totaling 35,400 almost 75% of present capacity by 2050. With lower
TWh/yr by 2050. Of this new demand, the electrification emissions and higher flexibility, gas remains competitive,
of road transport (2.8 billion EVs by 2050) is responsible with capacity reducing just 22% from current levels.
for one fifth. Electrolysers producing green hydrogen will Nuclear-power capacity will stay flat throughout the
take a 23% share, new space cooling requirements 11%, forecast horizon, with new capacity additions, largely in
and a similar share goes to the growing manufacturing Greater China, compensating for retirements in Europe
subcategory of machines, motors & appliances. and North America. In relative terms, nuclear more than
halves its share, dropping from 10% in 2019 to 4.3% in
Historically, prices have been set by the variable cost of 2050. Hydropower will be limited by resource constraints,
the most-expensive generation technology, providing reducing its share in the global electricity mix from 16% in
revenue for all generators. With the growing dominance 2019 to 12% in mid-century.
of new technologies, including solar, wind, storage, and
power-to-X, new rules will emerge. For example, in the In 2019, only 26% of electricity was supplied from renewa-
2050 power system, the maximum price arises when wind ble sources. By 2050 that share will have risen to 82%,
and solar supply are at their lowest, unlike the current along with major changes in flexibility and storage, which
power system where peak price would typically be at the we address in the sections that follow.
time of maximum load.
16
Executive summary
Hydrogen existing gas networks, pure hydrogen requires expensive

Hydrogen is a major contender for hard-to-abate sectors network retrofitting and a total upgrade of appliances.
where electrification is either infeasible due to the low
energy density of batteries or very costly. The production The leading method of hydrogen production is currently
of hydrogen is, however, expensive and involves significant steam methane reforming (SMR) . However, fugitive
energy losses. Absent some extraordinary policy shift to emissions of methane in the production of the natural gas
subvent its production and use, we do not foresee feedstock are controversial, and CO2 emissions also have
hydrogen supplying more than 5% of global energy to be captured for the hydrogen to be considered ‘blue’.
demand by 2050. Over time, blue hydrogen will steadily lose market share to
green hydrogen produced via electrolysis, as the latter
Synthetic fuels, such as e-methanol, e-ammonia, or benefits from scale economies and the downward cost
sustainable aviation fuels, are all hydrogen-derivatives trajectory of VRES power.
and are included in our total hydrogen demand forecast.
By 2050, synthetic fuels will account for 70% of aviation’s In our forecast, we distinguish further between grid-
and virtually all of maritime’s ‘hydrogen’ demand. powered and off-grid electrolysis. In the former, opera-
tors will have to compete with many other takers for
We forecast that, by 2050, hydrogen will have replaced low-cost VRES electricity such as demand response,
fossil fuels in many industrial heat applications. Manufac- pumped hydro, EVs (storage), and utility-scale batteries.
turing will thus account for 32% of hydrogen demand. The This limits annual operating hours to the extent that
next-heaviest user (17% share) of hydrogen (or rather off-grid dedicated renewable generation for hydrogen
hydrogen derivates) will be maritime, where few battery- production will become increasingly attractive.
electric options exist. One fifth of aviation fuel will be
hydrogen based by 2050, equating to 11% of total hydro- By 2050, 61% of the the world’s 281 Mt/yr hydrogen supply
gen demand. Hydrogen will struggle to replace natural will come from electrolysis, split roughly equally between
gas in buildings for space and water heating and cooking. solar PV (16%), onshore wind (16%), offshore wind (11%),
and grid-electricity (18%). Total installed electrolysis
Although up to 10% of hydrogen can be blended into capacity will reach 3 TW by 2050.
17
LEVELIZED COST OF HYDROGEN PRODUCTION

LEVELIZED COST OF HYDROGEN PRODUCTION
NAM LAM EUR SSA MEA NEE CHN IND SEA OPA Weighted Average
Fossil fuel cost range
for heating without
carbon price Production cost of hydrogen varies
significantly between regions
mostly due to variations in fuel cost
and conditions for renewable
SMR with CCS energy generation.
Electrolysis with grid electricity

2019
Electrolysis with onshore wind
Electrolysis with offshore wind
Electrolysis with solar PV
SMR with CCS
Electrolysis with grid electricity

By 2030, the average
2030
cost of renewable
Electrolysis with onshore wind hydrogen will catch
up with the cost of
SMR+CCS.
Electrolysis with offshore wind
Distribution of electricity price in 2050
SMR with CCS

Hours per year
Variable cost of
Electrolysis with grid electricity SMR+CCS
2050
Electrolysis with onshore wind 3000 hr/yr
Electrolysis with offshore wind 0 20 40 60 80 100 120 140 160 180 200
Wholesale electricity price (USD/MWh)
18 USD/kg 0 1 2 3 4 5 6 7 8
Executive summary
COST COMPONENTS OF HYDROGEN

COST COMPONENTS OF HYDROGEN
Hydrogen Production CAPEX Renewable Electricity CAPEX OPEX Electricity Cost
Transport and Share in final

Electricity Grid Charges and Tax Gas Cost CCS Cost Carbon Cost Storage Cost energy supply
Assuming 90% emissions 0%

SMR with CCS are captured.
Electrolysis with grid electricity 0%
Electrolysis with onshore wind 0%
Electrolysis with offshore wind 0%
Electrolysis with solar PV 0%
Renewable CAPEX can be as low as 50% of the world

average in locations where solar output is higher.
SMR with CCS 0.03%
Electrolysis with grid electricity The 2050s hydrogen 0.07%

market will be shaped by
competition. Electrolysers
will not be able run for
Electrolysis with onshore wind 7000-8000 hours/year as
0.01%
they do now. Instead they
will operate 3000
Electrolysis with offshore wind hours/year when electricity 0.01%
is cheap enough to make
their product cheaper than
Electrolysis with solar PV the competing SMR+CCS.
0.06%
0.90%
SMR with CCS
Due to its high cost,
Electrolysis with grid electricity hydrogen will remain a 1.23%
marginal energy carrier even
in 2050, except for a few
sectors where hydrogen is
Electrolysis with onshore wind the only feasible alternative
1.08%
to decarbonize.
Only 5% of the world's final
Electrolysis with offshore wind energy consumption will 0.72%
be in the form of hydrogen.
Electrolysis with solar PV 1.08%
USD/kg 0 1 2 3 4 5 6 7 8 19
WE ANALYSE 10 GLOBAL REGIONS
EUROPE
Transition policies, European

Green Deal and ‘Fit for 55’ aim
at a net zero greenhouse gas
emission economy by 2050,
which is forecast not to be met
Nearly decarbonized electricity

grid by 2050; < 5% generation
by coal and natural gas
The high carbon price could

spur CCS uptake, but
lukewarm support and a
decarbonized power system
ensure low implementation
KEY
Share of non-fossil Share of fossil

primary energy primary energy
MIDDLE EAST
sources 2050 sources 2050 AND NORTH AFRICA
Natural gas and oil dominate

the primary energy mix and
will do so until 2050
While declining in absolute

terms, this region’s cheap oil
NORTH AMERICA will increasingly dominate
global oil production.
Decarbonizing electricity by
2035, part of the ‘Build Back The region will start to
Better’ plan, is ambitious but realize its vast potential for
unlikely to be met renewable energy, reaching
a 25% share in primary
A massive buildout of grids energy mix in 2050
is forecast to bring aboard
renewables; 55% of generation
in 2030 and 75% by 2050
The switch to passenger

EVs is slow but gaining
momentum, and contributes LATIN AMERICA
to declining oil use SUB-SAHARAN AFRICA
Hydrogen from dedicated,
low LCOE solar PV may Least-developed and least-electrified world region;
change the region's energy only 42% of its people currently have access to
landscape electricity
Power generation mix will Soaring energy demand from a growing population
switch from hydropower/ and economy will be counteracted by efficiencies,
natural gas/fuel oil to e.g. traditional biomass cooking replacement by
hydropower/solar/wind gas and electricity
Renewables > 50% of 2050 Off-grid solar PV plays a significant role in energy
primary energy mix; access, and with grid-connected solar, accounts for
bioenergy above 20% almost 40% of power generation in 2050
20
NORTH EAST EURASIA
The region’s dependence on

oil and gas export revenues
will remain strong, disincen-
tivizing change
Two thirds of the region’s

primary energy will still be
met by natural gas in 2050
This region lags and remains

a laggard on decarbonization,
but there is considerable
focus on energy efficiencies
SOUTH EAST ASIA OECD PACIFIC
GREATER CHINA Energy demand, especially Falling population and

from space-cooling and improved efficiencies will
INDIAN SUBCONTINENT Ambitious net zero emissions is appliances, grows significantly almost halve energy use by
a 2060 goal; not evident from but levels off towards 2050 2050
500 million more people and shorter-term energy policies,
GDP growing four-fold will see but pressure is mounting Increasing use of natural gas 2050 electricity mix
energy demand doubling and renewables to supply dominated by wind; at 45% of
Electricity in final energy domestic demand for electricity, final energy demand, this is
Despite the rapid growth of demand grows from 23% in will result in lower importance of the most electrified region in
renewables, fossil-energy 2019 to 55% in 2050 – highest coal and oil 2050 after Greater China
sources will be 60% of the of all regions; >90% from
energy mix in 2050 renewable sources Manufactured goods Hydrogen will gain a foothold
production more than doubles (9% of energy use), sourced
Electricity’s share in building Peak coal already achieved. by 2050, driving demand for initially from Australia through
energy demand will triple by Share of coal in power mix gas and transforming this SMR processes, but later
2050, enabling universal (currently 60%) reduces to 5% region into a net-importer of mainly via renewably
access to electricity over the forecast period LNG powered electrolysis
21
SUPPLY
Hydrocarbon decline Natural gas – overtakes oil as the largest source of primary
Coal – coal demand peaked in 2014 at 7.9 billion tons. energy in 2032 and holds the top spot throughout our
Since then, it has been losing ground to natural gas and forecast period. Natural gas use will grow slowly this
renewables in Europe and North America. Coal use in decade, have a flat development in the 2030s, and
Greater China will decline significantly after 2030. Strong thereafter taper off by some 10% to 2050. Demand
growth in the Indian Subcontinent and South East Asia will changes will differ regionally: in OECD countries, gas
level off by then, and with coal exiting the energy mix consumption will gradually decline; in Greater China, it
elsewhere, except for high-heat processes in industry, its will peak in the early 2030s; in the Indian Subcontinent,
use will decline to just above a third of current levels by demand will almost triple by mid-century. Almost half the
2050. demand for gas derives from its final use in buildings,
transport and manufacturing. The other half involves
Oil – global oil demand may have peaked in 2019, but transformation into other uses: electricity, petrochemicals
post-pandemic recovery could see a new all-time high. and hydrogen production.
Our forecast demand for 2019 and 2025 differs by just 1%.
From that point oil use reduces slowly towards 2030, but The LNG share of total gas export will grow throughout
then the decline steepens by an average -2.8%/yr over the the forecast period. The big producers are also the big
following two decades – much faster than the average exporters, and North America – which is distant from its
growth of 1% per year we have seen in the past. Oil falls gas customers – will see the largest growth in liquefaction,
most sharply in the transport sector, halving over the next accounting for 38% of global capacity by 2050. By 2050,
30 years due to electrification of road transport and the just 15% of gas will be carbon free, led by regions with
rising use of low- and zero-emission fuels in aviation and higher decarbonization ambitions and higher carbon
maritime after 2030. Oil use in petrochemicals will reduce prices: Europe, Greater China, North America, and OECD
after 2035 due to recycling and bio-derived feedstock. Pacific. Hydrogen will grow to supply 3.5% of gas demand,
Globally, oil use declines 45% to 2050 compared with with CCS in power and industry and biomethane making
2019, with production concentrated ever more strongly in up the balance of ‘carbon free’ gas.
Middle East and North Africa.
22
Executive summary
Wind Cost reduction in onshore wind will total 42% over the
Wind power provided 5% of the world’s electricity output period 2020 to 2050, driven by rising capacity factors and
in 2019, almost exclusively in the form of onshore wind. By cheaper turbines. Cost reductions for the less-mature
mid-century that share will rise to 33% as electricity segments will be even steeper. For fixed and floating
generation from wind increases from 1,420 TWh/yr in offshore wind levelized costs will fall 44% and 80%
2019 to 19,000 TWh/yr in 2050. respectively, boosted by improvements in operating and
maintenance expenses as experience in installing and
Onshore wind installation will increase 8-fold by 2050 as it operating offshore wind turbines builds.
outcompetes fossil sources on cost from the current
decade onwards. By the 2040s, it also gains advantage Global wind capacity additions will increase from
over solar because wind turbines generate electricity 60 GW/yr in 2019 towards 340 GW/yr by mid-century,
when prices are high more often than PV. For offshore with marked differences in regional developments as
wind, we expect strengthened support in developed illustrated in Figure 20 below.
countries to bypass community opposition to onshore
turbines. The share of offshore wind in total wind electricity Installed capacity reached 709 GW at the beginning of
generation will increase steadily, rising globally from 6% in 2020. We forecast 1 TW in 2022, 2 TW in 2029, 4 TW in
2019 to 40% in 2050. However, by 2050, only Europe and 2043, and 5.9 TW in 2050, of which 1.7 TW will be
OECD Pacific will have more offshore than onshore wind. offshore. These developments are linked to larger
turbines, mega-sized projects, and a more dedicated
With new turbine types and continued increases in offshore supply chain. In addition, the 2020s will see
turbine, blade, and tower sizes, capacity factors will floating wind progress to full-scale demonstration
improve, raising the world average for onshore wind projects and on to commercial-scale deployments.
turbines from 21.5% in 2019 to 31% by 2050. For offshore We predict that floating offshore wind projects will have
wind turbines, where wind conditions are more favourable, 264 GW of installed capacity by 2050.
the average capacity factor is already 34%. We expect
this to rise to 50% by 2050.
23
Solar PV progressively lower prices for daytime production.

Grid-connected solar PV electricity will grow from 3.2%
of global grid electricity generation in 2019 to 36% by Lower received prices will not, however, be a showstopper
2050. Installed capacity increases 20-fold over the next for the strong growth of PV generation. Increasingly, PV
30 years to reach 11.5 TW in 2050. Greater China will hold and storage systems are designed as a ‘package’ that can
a 35% share of this capacity, followed by the Indian produce energy on demand, just like hydropower,
Subcontinent at 20%. nuclear or combustion power plants. Solar PV + storage
is thus a distinct power station category.
In 2020, despite the supply-chain disruptions caused by
the pandemic, new solar PV installations again set a The levelized cost of solar + storage is currently more
record at 129 GW. From 2030 onwards, we expect annual than double that of solar PV without storage. However,
additions of between 300 and 500 GW. with a continued drop in battery prices, the gap between
the two will narrow to below 65% by 2030. By then, the
A high cost-learning rate for solar panels (26%, declining to capture price advantage of solar + storage over regular
17% by 2050) for every doubling of installed capacity and solar PV plants will surpass the cost disadvantage on a
rising capacity factors (improving from 19% today to 26% globally averaged basis. Within a decade, about a
in 2050) will see solar PV costs continue to fall, as illustrated quarter of all PV installed will be with dedicated storage,
below (Figure 21). Currently, the global weighted average and by mid-century this share will have risen to half. In
levelized cost of energy (LCOE) for solar PV is breaking the 2050, total installed capacity will be 7.6 TW for solar PV
USD 50/MWh barrier, with individual project costs well and 3.9 TW for PV + storage.
below USD 20/MWh in locations like the Middle East and
North Africa and Latin America. Total installed off-grid solar PV capacity for hydrogen
production will be around 800 GW by 2050. By then, a
Cost leadership is a necessary but not sufficient condition further 130 GW of off-grid solar PV, coupled with inex-
for the expansion of solar PV. That is because, from a value pensive battery storage, will be providing hundreds of
perspective, solar can become a victim of its own success. millions of less-affluent people in the Indian Subcontinent
Typically, as the share of solar PV in a grid grows, it captures and Sub-Saharan Africa regions with access to energy.
24
Executive summary
Storage and flexibility Storage technologies will increasingly allow power

As we move towards a decarbonized electricity system, generation to be decoupled timewise from power
there is both opportunity and need for flexibility. With demand. Storage in today’s power system is mostly in the
high shares of solar and wind, traditional sources of form of pumped hydro. Although it is a mature technol-
flexibility will need to be accompanied by a large amount ogy and limited by geography, pumped hydro is set to
of storage. Over the next 30 years, utility-scale storage grow by 20% over the next three decades.
capacity will grow 160% to reach 7.3 TWh (Figure 23).
High penetration of wind and solar raises price variability
With the value of flexibility increasing, many conventional and strengthens the business case for storage, as does
generation technologies, like gas-fired power stations, the plunging cost of battery technology (Figure 22). We
will seek ways to accelerate their ramp rates and reduce forecast widespread expansion of battery storage,
their start times. There will be a growing emphasis on dominated by Li-ion batteries with 2-4 hours capacity.
shifting electricity usage from peak periods to times of From 2040 onwards, throughput of vehicle-to-grid
lower demand. Better prediction of renewable power systems in the world will be almost as large as that of
generation – and also consumption – levels will evolve, dedicated Li-ion batteries and pumped hydro, reaching
and new technologies and market mechanisms will allow 240 TWh/yr globally by mid-century.
more consumers to provide flexibility in the form of
demand response. In larger markets for utility-scale battery storage (e.g.,
China, South Korea, Japan, US), demand for longer-
Converting cheap electricity from VRES to other energy duration storage (> 5 hours) is already developing and
carriers, such as hydrogen, will add more flexibility. will intensify. This trend will boost new technologies like
Adoption of smart meters and smart grids, continued vanadium redox flow batteries, zinc-based chemistries,
investment in the interconnectors between physical or compressed air. Long-duration storage capacity is
transmission systems, and in the links between genera- likely to be nearing 1 TWh by 2050.
tion and load centres, will also contribute towards better
utilization of excess renewable supply.
25
COMPARISON OF ENERGY FLOWS: 2019 AND 2050

TITLE
Bioenergy Coal Direct heat Electricity Geothermal Hydrogen Hydropower
Natural gas Nuclear Oil Solar Wind
2019
Primary energy supply
Direct use & transformations Final energy demand
Direct use Transport

Oil
Road
Maritime
Aviation
Manufacturing
Rail
Coal
Manufactured
goods
Base materials
Natural gas
Iron and steel
Buildings
Construction
and mining
Feedstock
Power generation
Bioenergy
Space heating
Other
Water
Solar heating
Energy sector
Geothermal Losses own use
Hydropower Cooking
Appliances
Nuclear
and lighting
Space cooling
Wind
Electricity
These concentric pie charts illustrate the generation
losses associated with thermal (fossil,
biomass and nuclear) and non-thermal 27
(renewable) power generation. PWh/yr
The inner two circles show the input source.
The third circle shows the electricity/losses
associated with each source, while the outer
circle shows the total output in the form of Thermal
electricity, direct heat, and losses.
Non-thermal
Lost
2050
Direct use & transformations
Primary energy supply Hydrogen production
Coal
Direct use Final energy Road
demand
Transport
Oil Manufactured
goods
Maritime
Aviation
Natural gas
Base
materials
Rail
Manufacturing
Hydropower Iron and steel
Nuclear
Construction
and mining
Buildings Feedstock
Wind
Space heating
Bioenergy Power
generation
Water heating
Cooking
Solar Losses
Energy Appliances
sector and lighting
own use
Geothermal
Other Space cooling
Biomethane production
Electricity
generation
58
PWh/yr
EFFICIENCIES AND EXPENDITURE

Energy efficiency to be 100% efficient by the conventions of the energy-
Accelerating efficiencies in the production and use of accounting method we use. Therefore, conversion losses
energy are pivotal to the energy transition. as a percentage of input energy in power generation
reduce from 51% in 2019 to 19% in 2050.
Globally, energy intensity (unit of energy per dollar of
GDP) has been reducing by 1.7%/yr on average for the Most end-use efficiencies are also linked to electrifica-
last two decades. This decline has not been smooth, and tion. The obvious example is the electrification of road
the COVID-19 pandemic introduced further short-term transport, where EVs are more than three times more
spikes with varying fluctuations in both energy consump- efficient than their fossil counterparts.
tion and GDP.
In calculating sectoral efficiencies, we take a range of
Irrespective of short-term impacts of the pandemic, activity, technological and structural changes into
energy intensity will continue to decline faster than in account in addition to electrification. For example,
previous decades, by 2.4%/yr on average over the next technology improvements resulting in better engine
30 years (Figure 24). The cumulative effect is that energy performance, hull hydrodynamics or insulation. We find
intensity will drop from 4.3 MJ/ USD in 2019 to 2.0 MJ/ that without expected efficiency gains in transport,
USD in 2050. In other words, by mid-century we will use buildings, and manufacturing, energy demand would be
less than half of the energy used today to produce one 65% higher than we forecast for 2050.
dollar of GDP.
There is considerable potential to accelerate efficiencies
From an energy production perspective, rapid electrifi- beyond our forecast. That, however, will require new
cation powered by renewables is the core driver of mandates from governments and international bodies
accelerating energy efficiency in the next three decades. that we are as yet not able to forecast, along with unprece-
The typical thermal efficiency for utility-scale electrical dented co-operation within industries on new standards
generators is some 30 to 40% for coal and oil-fired plants, and recommended practices.
and up to 60% for combined-cycle gas-fired plants. In
comparison, solar PV and wind generation are assumed
28
Executive summary
Energy expenditure Upstream oil and gas expenditures will decline by 46%
The energy transition that we forecast is not only affordable through to 2050. Oil investments will drop, while gas
but leads to considerable savings at the global level. investment will remain constant before slightly decreasing
in the 2040s. On the other hand, given the long lifetimes
Renewables, storage and grid buildouts, will indeed of installed capacity, operating expenses will remain quite
require very large upfront investment; this is why some high, only decreasing by 35% for oil and remaining flat
believe that the transition is ‘unaffordable’. Our results for gas.
suggest the exact opposite: whereas global GDP will
more than double by 2050, global energy expenditures Fossil fuel-fired power investments will decline spectacu-
do not increase very much from today’s levels. This is larly by 96% over this same period, to a mere USD 5bn.
thanks to improvements in energy efficiency and in Operating and maintenance of fossil power stations
renewable-energy technologies. (excluding fuel) will remain at around USD 350bn due to
the inertia of long-lifespan installations.
World energy expenditures will shift from fossil to non-
fossil sources, and the annual sum expended will increase The increase in electricity demand will lead to an almost
by only 4%, from USD 4.5trn in 2019 to USD 4.7trn in 2050 doubling of non-fossil power expenditures by 2050. The rise
The fossil-energy share will decline by more than a third of is particularly visible for solar PV and wind power. Together,
today’s 76%, dropping to 42% by mid-century. they will represent a fifth of global energy expenditures in
2050, an almost four-fold increase compared with 2019.
There are various definitions of ‘energy expenditure’, and
we use a strictly stipulated terminology. However, we have The bottom line however is that the share of GDP devoted
excluded investments in energy-efficiency measures, as to energy worldwide will halve within the space of a single
well as in downstream carbon-mitigation costs. Nor do we generation. Conceptually, these ‘savings’, amounting to
incorporate costs related to end-use spending (in manu- tens of trillions of dollars per decade, could be applied to
facturing, transport, etc.). accelerating the transition.
29
NOT FAST ENOUGH

The latest science is clear on the need to achieve net zero Regions
emissions by mid-century if we are to reach the ambitions Absolute emissions will increase in the Indian Subconti-
of the Paris Agreement. For that to occur, we must drive nent and Sub-Saharan Africa to 2050. The highest-
down emissions from human activities to as close to zero emitting region, Greater China, will reach peak emissions
as possible by 2050. The energy-related emissions we before 2030; emissions will then decline by 75% from
forecast are very far from zero by 2050, and, together today’s levels. All other regions will reduce their emissions,
with other emissions from human activity, suggest a with OECD Pacific, together with Europe, experiencing
warming of 2.3°C by 2100 – a level considered dangerous the biggest relative change. North East Eurasia will have
by the scientific community. the highest emissions per capita at 7.5 tonnes/person in
2050, followed by North America and Middle East &
Emissions North Africa at 3.5 tonnes/person.
Energy production and use represents 70% of global
GHG emissions, of which most is CO2. We forecast that Carbon capture
annual energy-related CO2 emissions in 2050 will be at CCS uptake will be very limited in the near- to medium-
19 Gt, a 45% reduction compared to the present level. term, and effectively too late and minimal in the longer
term. It is only in the 2040s, when carbon prices start to
50% of energy-related emissions have been added to the approach the cost of CCS, that deployment begins at
atmosphere in the last 50 years. After staying virtually flat scale. By 2050, total carbon capture will amount to just
between 2014-2016, global energy-related CO2 emis- 6% of all annual energy-related emissions.
sions grew to reach a peak of 34.3 Gt CO2 in 2019.
Overshooting the carbon budget
The effects of COVID-19 resulted in emissions dropping To limit global warming to below 1.5°C, the IPCC
by approximately 6% in 2020. But as economic activity is concludes that we have to limit cumulative emissions to
now picking up, energy use and emissions are rising 400 Gt CO2 from the start of 2020 and into the future,
again. Energy-related emissions will climb back up 3% and 1150 Gt CO2 to stay below 2.0°C.
during 2021 and grow to 33 Gt CO2 by 2023 before
declining gradually to 31 Gt CO2 in 2030, a level only 9% Using the IPCC carbon budgets and the aggregated CO2
lower than 2019. emissions from our forecast, we find that the 1.5°C
budget will be exhausted in 2029. To exhaust the budget
Combustion emissions currently come mainly from coal associated with the 2.0°C threshold it takes a further 24
and oil use, but within the next three decades emissions years until 2053.
will be increasingly dominated by natural gas. We
forecast that emissions from coal will fall 62% by 2050, In arriving at these estimates, we add emissions from
emissions from oil will halve, whereas emissions from non-energy sources (e.g. agriculture and industrial
natural gas will grow towards 2030 and then drop back to processes) to give a full picture of CO2 emissions from
a level only 15% below today’s emissions. human activity.
The main demand sectors of manufacturing, buildings To estimate the CO2 emissions and global warming by
and transport will all see emissions fall between 40 to the end of the century, we extrapolate the development
50% by 2050 due to the rise of renewably powered of emissions and their capture towards 2100. The
electricity and ongoing efficiencies described on updated climate response from IPCC AR6 (IPCC, 2021)
page 28. suggests that, with a likely budget overshoot of 370 Gt
CO2, the world will reach a level of warming of 2.3°C
above pre-industrial levels by 2100.
30
Executive summary
IPCC Sixth Assessment The report also describes significant effects of the
climate changing in areas such as rainfall patterns, the
Report on climate science melting of icesheets and glaciers, permafrost and
seasonal snow cover, and rising sea levels.
The United Nations’ Intergovernmental Panel on
“ "
Climate Change (IPCC) published the first part of its
Sixth Assessment Report (AR6) on 9th August 2021. Today's IPCC Working
The report concludes that almost all emissions scenarios
Group 1 Report is a
expect to result in 1.5°C of warming, “in the early 2030s”. code red for humanity.
Without reaching “net zero” CO2 emissions, as well as an
immediate, sharp reduction in emissions of both CO2 and
other greenhouse gases – the climate system will
continue to warm.
A key enhancement in AR6 is that its warming projections António Guterres

are based “for the first time” on multiple lines of evidence, United Nations
including observations of historical and recent warming Secretary-General
trends. This is a major shift, as earlier IPCC projections
9 August 2021
were based entirely on climate models.
31
MODEL INPUT
Population Economic growth
The number of people in the world is a central input to any World GDP is expected to grow from USD 138 trn/yr in
energy forecast. The UN’s World Population Prospects is 2019 to USD 292 trn/yr in 2050. This doubling over the
the most widely used resource by energy forecasters. 31-year period is the result of a 22% increase in population
and a 74% increase in average GDP per capita, with large
However, the UN has been criticized for not paying regional differences.
enough attention to country-specific education levels –
data that are relevant for future fertility trends. The pandemic slowed economic growth to a 1.2% CAGR
over the 2019-2021 period, and will leave a permanent
Consequently, this Outlook follows the approach used by impact on the world economy. The post COVID-19 boost
the International Institute for Applied Systems Analysis in 2023 will result in some regional economies growing
(IIASA), which specifically considers how urbanization and slightly faster than they otherwise would have, and in 2050
rising education levels are linked to demographic trends. the loss declines to 3.2%.
Following the latest (2019) update from IIASA, we arrive at By mid-century, today’s fast-growing emerging econo-
a global population estimate of 9.4 billion by 2050 – some mies will experience significantly slower growth as they
4% lower than the most recent (2017) UN median popula- move into the tertiary (service) economy. The combination
tion forecast. Energy use per person varies considerably of slower population growth and decelerating productivity
and this is reflected in our model through a weighting means global GDP growth rates will also slacken.
process in calculating aggregate energy consumption
first at a regional and then global level. The fastest growth in GDP per capita, between 2021 and
2030, will be in Asia. The Indian Subcontinent (IND) will
The pandemic has affected both fertility levels and have the highest growth rate, at an average of 6.2%/yr,
education, however the impacts on long-term population followed by South East Asia (SEA) at 4.9%/yr and Greater
trends have yet to be definitely determined. China (CHN) at 4.6%/yr.
32
Executive summary
Learning curves effect Policy

The cost of a technology decreases by a fraction with every Energy policy has never been under so bright a spotlight
doubling of capacity. Ongoing market deployment brings as at present – in the wake of the publication of the IPCC’s
greater experience, industrial efficiencies, and further R&D. alarming AR6 report, in the build-up to the delayed
COP26, and amid analyses of the disappointing impacts
The cost of solar panels has been declining by 26% for of COVID-19 recovery spending on the energy transition.
every doubling of global cumulative capacity additions.
We expect this to reduce to 17% by 2050. The cost learning Through our work with governments, academia, and
rate (CLR) is 16% for wind turbines and 19% for Li-ion international bodies, DNV keeps close tabs on policies
batteries and will continue at these rates. Apart from these that impact the energy transition at national, regional, and
core technologies, total investment costs include the costs global levels. We constantly analyse a wide range of
of supporting infrastructure, installation kits, labour, legal topics – such as climate goals, air quality, health, job
fees, etc, which all have a lower CLR. Including these other creation, and energy security.
cost components, we forecast global average of the total
investment cost of relatively mature renewable technolo- An overview of the policy factors explicitly factored into
gies like solar PV and onshore wind to decline 11% for our forecast can be found on page 35.
every doubling of global cumulative capacity additions
until 2030. For newer technologies like solar PV + storage In our model, country-level data are translated into
and floating offshore wind, the learning rates are higher. expected policy impacts, then weighted and aggregated
Towards 2050, CLRs will decline as non-technology costs to produce regional figures for inclusion in our analysis.
start to constitute a higher share of the total investment cost. Examples include explicit regional carbon prices, support
for renewable energy and EVs, and fossil fuel taxation in
The decline in investment cost only partially explains why relation to air pollution and climate concerns. Several
technologies like solar and wind experience a massive countries – Chile being an outstanding example – are
uptake. Improvements in design and operation of power gearing up programmes to support the acceleration of
plants allow higher capacity factors, ensuring more energy green hydrogen.
is produced over the year per kilowatt of capacity.
33
MODEL DESCRIPTION
Figure 29 below presents the ETO model framework. influences all aspects of the energy system. Energy-
The arrows in the diagram show information flows, efficiency improvements in extraction, conversion
starting with population and GDP per person, while and end-use are a cornerstone of the transition.
physical flows are in the opposite direction. Policy
FIGURE 29
ETO model framework
POLICY
POPULATION
Source of Final energy Energy Primary

demand demand transformation energy supply
GDP PER
PERSON
TRANSPORT TRANSPORT POWER GENERATION Solar

Measured in Road Electricity Wind
tonne-miles,
Maritime Direct heat Hydropower
passenger-
kilometres and Aviation Nuclear
vehicles Rail
Pipelines
BUILDINGS HYDROGEN Bioenergy

Space heating BUILDINGS Geothermal
& cooling,
water heating, Residential OIL REFINERIES
cooking, and Commercial
appliances & FOSSIL FUEL
lighting EXTRACTION
MANUFACTURING Crude oil

Manufactured DIRECT USE Natural gas
MANUFACTURING Coal
goods
Production output
Base materials
measured as
Manufacturing Iron and steel
Value Added Construction
and mining
NON-ENERGY
Feedstock
OTHER
ENERGY SECTOR’S
OWN USE
ENERGY EFFICIENCY
34
Executive summary
Our approach turing, and so on) and all sources of supply over time.
In contrast to scenario-based outlooks, we present a It encompasses demand and supply of energy globally,
single ‘best estimate’ forecast of the energy future, and the use and exchange of energy between and within
with sensitivities in relation to our main conclusions. ten world regions. The analysis covers the period
1980-2050, with changes unfolding on a multi-year
Our model simulates the interactions over time of the scale that in some cases is fine-tuned to reflect hourly
consumers of energy (transport, buildings, manufac- dynamics.
Our best estimate, Long term dynamics,

not the future we want A single forecast, not scenarios not short-term imbalances
Continued development Main policy trends included; Behavioural changes: some

of proven technology, not caution on untested assumptions made, e.g. linked
uncertain breakthroughs commitments, e.g. NDCs, etc. to a changing environment
FIGURE 30
Policy
Policy factors included in our Outlook
Policy influences all aspects of the energy system,
and Figure 30 gives a snapshot of the policy factors
incorporated into our forecast. Policy considerations
influence our forecast in three main areas:
1 Renewable
power support 2 Energy
support
storage
3 Zero emission
vehicle support
a.) Supporting technology developments and activat-
ing markets that close the profitability gap for
renewable-energy technologies competing with H
existing technologies
b.) Restricting the use of inefficient or polluting 4 Hydrogen

support 5 CCS
support 6 Energy efficiency
standards
products/technologies by means of technology

requirements or standards, or;
TAX
c.) Providing economic signals, for example a price
incentive, to reduce carbon-intensive behaviours 7 Bans and
phase-out plans 8 Carbon-pricing
schemes 9 Fuel-, energy- and
carbon taxation
Country-level data are translated into expected policy

impacts, which are weighted and aggregated to
produce regional figures, and ultimately a global 10 Air-pollution
interventions 11Plastic pollution
interventions 12 Sustainable aviation
fuels support
impact, for inclusion in our analysis.
35
ENERGY TRANSITION OUTLOOK 2021 REPORTS OVERVIEW
Energy transition outlook Technology progress report

Our main publication details our model-based forecast of We explore how key energy transition technologies will
the world’s energy system through to 2050. It gives our develop, compete, and interact in the coming 5 years.
independent view of the most likely trajectory of the The ten technologies are:
coming energy transition, and covers:
— Energy production: floating wind, solar PV, and waste
— The global energy demand for transport, buildings, to fuel and feedstock
and manufacturing — Energy transport, storage, and distribution: pipelines
— The changing energy supply mix, energy efficiency, for low-carbon gas; meshed HVDC grids, new battery
and expenditures technology
— Detailed energy outlooks for 10 world regions — Energy conversion and use: novel shipping technolo-
— The climate implications of our forecast. gies, EVs and grid integration, green hydrogen
production, CCS.
We also provide details of our model and main assump-
tions (i.e., population, GDP, technology costs and govern- We attempt to strike a balance between technical details
ment policy). Our 2021 Outlook explores, inter alia, the and issues of safety, efficiency, cost, and competitiveness.
impact of COVID-19 and the growing importance of The interdependencies and linkages between the
hydrogen as an energy carrier. technologies are a particular area of focus.
36
Executive summary
Financing the energy transition Maritime forecast

Focuses on the financial opportunities and challenges The Maritime Forecast to 2050 offers shipowners
for financiers, policymakers, developers, and energy practical advice and solutions as shipping’s carbon
companies: reduction trajectories rapidly head towards zero.
— An affordable transition – considering whether a — DNV’s new carbon risk framework allows detailed
Paris-compliant transition is affordable, and what may assessments of fuel flexibility and Fuel Ready solutions,
be needed to mobilize and redirect capital the economic robustness of fuel and energy efficiency
— Accelerating the transition – examining the role of strategies, and their impact on vessel design.
financial markets, policy, and regulation, and how to — Decarbonization is leading to increased regulatory
get capital to flow to where it can have the most impact requirements, new cargo-owner and consumer
on emissions expectations, and more rigorous demands from
— Ensuring a just transition – exploring the importance investors and institutions.
of balancing sustainability priorities, ensuring co- — Investments in energy and fuel production will be
benefits, and building climate resilience. essential to shipping’s efforts to decarbonize.
The report combines DNV’s independent energy This is the grand challenge for the maritime industry. But
forecast to 2050 with views from a diverse set of leaders by working together as an industry, embracing fuel
in the energy and finance sectors. flexibility, and consulting with expert partners, shipping
can reach its destination.
37
PATHWAY TO NET ZERO EMISSIONS
This year, ahead of COP 26, we are releasing

a new companion report to our main
Energy Transition Outlook 2021. As outlined
in the Paris Agreement, and confirmed in
the IPCC AR6 WG1 report released in
August 2021, there is a dire need for
urgent, prioritized action tackling energy-
related emissions.
Our new report plots a pathway for how to

close the gap between our forecast and net
zero CO2 emissions by 2050 – i.e. actions
that are likely to limit global temperature
increase to 1.5˚C by end of this century.
The report covers all energy sectors –
including hard-to-abate sectors like aviation,
maritime and cement – and each of the ten
global regions in our Energy Transition
Outlook. We look at which technologies will
contribute to the required change and the
policies needed to achieve that.
Download our forecast data

All the forecast data in DNV’s suite of
Energy Transition Outlook reports, and
further detail from our model, is accessible
on Veracity – DNV’s secure industry data
platform.
eto.dnv.com/forecast-data
38
Executive
The Project
summary
team
ETO TEAM AND CONTACT
This report has been prepared by DNV as a cross- our Energy Transition research programme, part of the
disciplinary exercise between the DNV Group and two Group Development and Research unit, based in Oslo,
of our business areas – Energy Systems and Maritime Norway. In addition, we have been greatly assisted by
– across 15 countries. The core model development and the external Energy Transition Outlook Collaboration
research has been conducted by a dedicated team in Network.
DNV core team Our external collaboration network for our 2021
Outlook includes:
Steering committee Harald Magnus Andreassen (Sparebank 1 Markets),
Remi Eriksen, Ditlev Engel, Ulrike Haugen, Valentin Batteiger (Bauhaus Luftfahrt), Kingsmill Bond
Trond Hodne, Liv Hovem (Carbon Tracker), Sunil Gupta (Vena Energy), Gørild
Heggelund (Fridtjof Nansen Institute), Robert Hornung
Project director
(Canadian Renewable Energy Association), Steffen
Kallbekken (CICERO), Francisco S. Laverón (Iberdrola),
Yang Lei (Institute of Energy, Peking University), Wolfgang
Lutz (Wittgenstein Centre for Demography and Global
Human Capital), Jinglong Ma (APEC Sustainability
Center), Tom Moultrie (University of Cape Town),
Susanne Nordbakke (Transportøkonomisk Institutt, TØI),
Glen Peters (CICERO), Sergei P. Popov (Melentiev Energy
Systems Institute), Thina Margrethe Saltvedt (Nordea),
Sverre Alvik, sverre.alvik@dnv.com Jon Birger Skjærseth (Fridtjof Nansen Institute), Marco
Tagliabue (Oslo Met University), Mena Testa (Enel Global
Modelling responsible
Infrastructure and Networks), Kevin Tu (Agora Energie-
Onur Özgün
wende China), Jørgen Wettestad (Fridtjof Nansen Insti-
Core modelling- and research team and contributing tute), Yongping Zhai (Tencent / ADB)
authors
Bent Erik Bakken, Gudmund Bartnes, Thomas Horschig, Historical data
Anne Louise Koefoed, Erica McConnell, Mats Rinaldo, This work is partly based on the World Energy Balances
Sujeetha Selvakkumaran, Adrien Zambon, database developed by the International Energy Agency
Roel Jouke Zwart © OECD/IEA 2020, but the resulting work has been
prepared by DNV and does not necessarily reflect the
Communication responsible and editor
views of the International Energy Agency.
Mark Irvine, mark.irvine@dnv.com
Energy Systems project manager For energy-related charts, historical (up to and including
Jeremy Parkes 2018) numerical data is mainly based on IEA data from
World Energy Balances © OECD/ IEA 2020, www.iea.org/
Maritime project manager
statistics, License: www.iea. org/t&c; as modified by DNV.
Linda Sigrid Hammer
Published by DNV AS. Design SDG/McCann Oslo/Infogr8. Print 07 Media AS. Paper Arctic Volume White 130/250.
Images Cover image: Shutterstock, page 31: UN/Mark Garten, page 38 and 39: DNV
39
About DNV
DNV is an independent assurance and risk management provider, operating in

more than 100 countries, with the purpose of safeguarding life, property, and the
environment. Whether assessing a new ship design, qualifying technology for a
floating wind farm, analysing sensor data from a gas pipeline or certifying a food
company's supply chain, DNV enables its customers and their stakeholders to
manage technological and regulatory complexity with confidence.  As a trusted
voice for many of the world’s most successful organizations, we use our broad
experience and deep expertise to advance safety and sustainable performance,
set industry standards, and inspire and invent solutions. 
dnv.com/eto
Headquarters:
DNV AS
NO-1322 Høvik, Norway
Tel: +47 67 57 99 00
www.dnv.com
The trademarks DNV® and Det Norske Veritas® are the properties of companies
in the Det Norske Veritas group. All rights reserved.

DNV ETO 2021 Executive Summary Single Page Highres

Uploaded by

Copyright:

Available Formats

DNV ETO 2021 Executive Summary Single Page Highres

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

DNV ETO 2021 Executive Summary Single Page Highres

Uploaded by

Copyright:

Available Formats

ENERGY

A global and regional forecast to 2050

Electricity demand will more than double by 2050 and

3. Efficiency gains lead to a flattening of energy demand from the 2030s

NEW INSIGHTS 2021

1. COVID-19 economic recovery spending is a lost opportunity

4. Most hydrogen will be produced from dedicated renewables-based electrolysers by 2050

Highlights – Core insights

Highlights – New insights

Pandemic recovery spending 2020

Source: UNEP report ‘Are we building back better’, Feb 2021

PRIMARY ENERGY 2019-2050

World Energy Supply Transition 2020–2050

ENERGY TRANSITION ENERGY TRANSITION ENERGY TRANSITION ENERGY TRANSITION

After five years of About this Outlook

the more widespread use of heat pumps. With greater Transport

Hydrogen existing gas networks, pure hydrogen requires expensive

LEVELIZED COST OF HYDROGEN PRODUCTION

Electrolysis with grid electricity

Electrolysis with onshore wind

Electrolysis with offshore wind

Electrolysis with solar PV

SMR with CCS

Electrolysis with grid electricity

Electrolysis with solar PV

Distribution of electricity price in 2050

SMR with CCS

Electrolysis with onshore wind 3000 hr/yr

Electrolysis with solar PV

COST COMPONENTS OF HYDROGEN

Transport and Share in final

Assuming 90% emissions 0%

Electrolysis with grid electricity 0%

Electrolysis with onshore wind 0%

Electrolysis with offshore wind 0%

Electrolysis with solar PV 0%

Renewable CAPEX can be as low as 50% of the world

SMR with CCS 0.03%

Electrolysis with grid electricity The 2050s hydrogen 0.07%

Electrolysis with solar PV 1.08%

WE ANALYSE 10 GLOBAL REGIONS

Transition policies, European

Nearly decarbonized electricity

The high carbon price could

Share of non-fossil Share of fossil

Natural gas and oil dominate

While declining in absolute

The switch to passenger

The region’s dependence on

Two thirds of the region’s

This region lags and remains

SOUTH EAST ASIA OECD PACIFIC

GREATER CHINA Energy demand, especially Falling population and

Solar PV progressively lower prices for daytime production.

Storage and flexibility Storage technologies will increasingly allow power

COMPARISON OF ENERGY FLOWS: 2019 AND 2050

Natural gas Nuclear Oil Solar Wind

Primary energy supply

Direct use & transformations Final energy demand

Direct use Transport

DNV is an independent assurance and risk management provider, operating in