Energy Innovation

Why We Need It and How to Get It

BY

B I L L

G AT E S

N OV E M B E R

3 0, 2 0 1 5

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

G E T

I T

P1

Introduction
In 30 years the world will consume much more energy than it does today. This should
be good news. Wherever access to reliable, affordable energy goes up, so does the
quality of life. But today more than 1 billion people lack access to the most basic energy
services. Energy keeps schools and businesses running, city lights shining, tractors
plowing, and cars and trucks moving. Without plentiful energy, the poverty rate could
not have dropped by more than half since 1990, and hundreds of millions of people
would have been denied the opportunity to improve their lives. There would be no
steel, fertilizer, cement, or many of the other materials that make modern life possible.
Of course, the worlds growing appetite for energy is not

By far the biggest source of emissions is the production

unalloyed good news. More than 80 percent of the energy we

of energy, which accounts for around two thirds of all

use today comes from fossil fuels, which produce greenhouse

greenhouse gases produced by human activity.2 Although

gases and drive climate change. Scientists generally agree

using energy more efficiently can help, the only way to

that to limit warming to 2 degrees Celsiusessentially the

dramatically reduce or eliminate global carbon emissions

least bad optionthe worlds biggest carbon emitters need

is to switch to sources that do not emit carbon.3

to reduce emissions by 80 percent by 2050, and all countries

need to essentially eliminate them by the end of the century.

I M PA C T O F R E D U C I N G E M I S S I O N S O N G LO B A L T E M P E R AT U R E R I S E

The blue line shows how dramatically we need to cut emissions from energy production if we want to limit warming to 2 degrees Celsius.

gCO 2 / kWh

600

GLOBAL AVERAGE

500

Historic
6 Degree Scenario

400

4 Degree Scenario
2 Degree Scenario

300

N E W C A PA C I T Y

200

6 Degree Scenario
4 Degree Scenario

100

2 Degree Scenario

0
1990

2000

2010

2020

2030

2040

2050

Source: International Energy Agency

Vaclav Smil, Energy Transitions: History, Requirements, Prospects.
International Energy Agency, Energy Technology Perspectives 2015.
The world will need to reduce emissions in other areas too, including manufacturing. For example, making steel and cement involves chemical reactions that produce carbon dioxide. Cattle produce methane,
another greenhouse gas. Zero-carbon energy sources will not change these facts. But because energy represents such a large share of global emissions, there is no way to solve the climate problem without
carbon-free energy. It is necessary but not sufficient.
1

2
3

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

G E T

I T

P2

Affordable, reliable energy would be a priority even if climate

substantially. As the International Energy Agency (IEA) has

change were not a problem. It would help millions more

written: Recent trends reaffirm the need to accelerate energy

people escape poverty and become more self-sufficient. It

technology innovation.5 To get this acceleration, governments

would ease international tensions, increasing global security

and the private sector should invest far more in clean-energy

and making more countries energy-independent. It would

research, development, and deployment.

unlock new economic opportunities in a growing trillion-dollar

industry. It would lessen the problems caused by the expensive
and often dangerous work of extracting fossil fuels. It would
reduce air pollution, which kills millions of people every year.4
And it would stabilize energy prices, which will have an even
bigger impact on the global economy as more people come to
rely on energy in their daily lives.

The work needs to start immediately. The history of energy

transitions is clear: It takes years to develop new sources of
energy, and decades to make them a significant part of our
energy mix. Today, renewables account for less than 5 percent
of the worlds energy. It took four decades for oil to go from
5 percent of the worlds energy supply to 25 percent. Natural
gas took even longer.6

The opportunity is especially clear for developing countries.

I believe we can make this transition faster, both because

Their most immediate need is to keep their hospitals and

the pace of innovation is accelerating, and because we have

schools running and help their economies grow. If forced to

choose between energy that is clean and energy that is reliable
and affordable, it is completely responsible to prioritize the

never had such an urgent reason to move from one source of

energy to another. Some critics argue that only governments
can manage such a big transition. Others argue that only

health and welfare of their people today over the serious

the private sector can create the innovations we need. Both

implications of an uncertain future with climate change.

sides have a point. Private companies will ultimately develop

We need to resolve this dilemma by making energy reliable,

energy breakthroughs. But their work will rely on the kind of

affordable, and clean.

basic research that only governments can fund. The primary

In other words, moving to new energy sources is both a wise

drivers of innovation are government research and high-risk

response to the threat of climate change, and an opportunity to

private funding.

make the world more secure and more equitable.

The IEA has called for both deploying immediately available

In this paper I make the case that, although we have made good

solutions and developing more complex solutions needed

progress on this front, the pace of progress needs to increase

for long-term deep decarbonisation.7 Most discussions

E N E R G Y T R A N S I T I O N S TA K E D E C A D E S
% of Total Global Energy Supply

60

50

Gas

40

Oil
Coal

30

Nuclear
Hydropower

20

Traditional Biofuels
Other Renewables

10

0
1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

Source: Vaclav Smil

World Health Organization (http://www.who.int/mediacentre/news/releases/2014/air-pollution/en/).
International Energy Agency, Energy Technology Perspectives 2015.
Vaclav Smil, Energy Transitions: History, Requirements, Prospects. The figure for renewables current share of global energy does not include hydropower.
7
International Energy Agency, Energy Technology Perspectives 2015.
4
5

2000

2010

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

G E T

I T

P3

about energy and climate focus only on the former. As

For example, with todays technology, it is not economical to

important as those solutions are, we also need to focus

store mass amounts of solar or wind energy for use on cloudy,

on the latterthe breakthrough innovations that will

windless days. Depending on the type of battery used, it costs

ultimately solve the problem on a global scale. With the right

between 30 cents and 80 cents to store a kilowatt-hour of

investments, the world can get the tools to meet its growing

electricity.9 Given that the price of electricity averages around

energy needs while also cutting our carbon emissions to

10 cents per kilowatt-hour in the United States,10 these

near zero.

storage costs would at least triple the price of electricity, even

if the electricity itself were generated for free. The European
Union, where prices average 20 cents per kilowatt-hour,
and India, where they range from 2 to 15 cents, would see

Why We Need
Energy Innovation

similarly dramatic increases.11

Todays batteries also have a far lower energy densitythat
is, they store much less by weightthan fossil fuels. Coal
provides 37 times more energy per kilogram than the best
lithium-ion batteries available today. Gasoline provides 60

The past decade has seen remarkable progress in zero-carbon

times more.12 This explains not only why electric cars have a

sources, especially solar photovoltaic (PV) and wind. The cost

shorter range than ones that run on hydrocarbons, but also

of solar PV cells has dropped by nearly a factor of ten. There is

why there are not electric airplanes, freight trucks, or cargo

no question that these technologies can play an important role

ships. Any battery big enough to power these vehicles would

in the worlds zero-carbon energy mix; the IEA has estimated

be prohibitively heavy.

that wind and solar PV could cut the worlds annual emissions
from electricity generation 22 percent by 2050.8

If we are going to take full advantage of the benefits of solar

Why not 100 percent? One reason is that solar PV and wind

factories, skyscrapers, and other large consumers of energy

power are intermittent sources. Customers still need energy

we need innovative ways to stitch them together into a

even when the sun is not shining and the wind is not blowing.

reliable, affordable system.

Today that requires maintaining a parallel system powered by

These technologies could be one path to a zero-carbon

fossil fuels. On bright summer days, the two systems together

future, but they are not the only one. Given the scale of this

generate so much electricity that the price drops below

challenge, we should be exploring all potential avenues.

zeroutilities pay others to take their product. At night, on

Doing so is a job for both the public and private sectors.

PV, wind, and batteriesespecially if they are going to power

the other hand, the only source is electricity from fossil fuels.
Because energy companies have to recover their capital costs
and they are not getting any return during the day, they either
raise the price of energy at night, or they slowly go bankrupt.
It would help solve this problem if we could store lots of
solar or wind energy and then use it on cloudy or still days.
Another option would be to deploy these resources wherever
sun and wind are plentiful and connect them to other places

The Case
for Government
Investment

using a high-voltage grid. These options are being used

today in some places, but we need to make them cheaper and

Government-funded research programs have produced many

improve their performance.

of the innovations that define modern life. For example,

International Energy Agency, Energy Technology Perspectives 2015.

Based on the capital cost of the battery, amortized over its useful life.
U.S. Energy Information Administration (http://www.eia.gov/electricity/monthly/epm_table_grapher.cfm?t=epmt_5_6_a).
11
Eurostat (http://ec.europa.eu/eurostat/tgm/table.do?tab=table&init=1&language=en&pcode=ten00117&plugin=1); Government of India, Department of Economic Affairs (https://data.gov.in/resources/
state-wise-average-rate-electricity-domestic-and-industrial-consumers/download).
12
Based on energy density of 0.2 kilowatt-hours per kg for batteries. Gasoline and coal statistics are drawn from Energies: An illustrated Guide to the Biosphere and Civilization, by Vaclav Smil.
8
9

10

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

G E T

I T

P4

during the Cold War, the U.S. Department of Defense

This time lag between invention and impact helps explain why

developed computer networks that could survive a nuclear

the energy industry tends to invest less in research, development,

attack. Today that same technology helps run the Internet.

and deployment than other industries do. American

Private companies have made the Internet a hub of global

pharmaceutical companies put 20 percent of their revenues

commerce and communication, but it would not exist

into these activities. For IT and semiconductor companies, the

without early government investments in computer chips

number is around 15 percent of revenues. For energy firms, it is

and communications technology. The digital revolution,

0.23 percent.14

which has empowered people around the world, owes a

considerable debt to federally funded research.
Government investments also strengthen the academic
institutions that are the birthplace of so much innovation.

RESEARCH & DEVELOPMENT

How much do U.S. industries invest in R&D?

% of Revenue (2010)

It is no accident that the worlds top computer science

INDUSTRY

INVESTMENT IN R&D

schoolsStanford, Carnegie Mellon, MITare based in a

PHARMACEUTICAL

20%

country whose government has made big investments in that

INFORMATION TECHNOLOGY

15%

field. Nor is it an accident that the United States, which has

ENERGY

0. 23%

invested heavily in biological research, produces an outsize

Source: International Energy Agency, Global Gaps in Clean Energy R&D (2010)

share of the worlds innovation in health and medicine.

This is why governments play an indispensable role in
Similarly, research funded by the U.S. government has defined

supporting energy research. The field will become only

the state of the art in energy production since World War II.

more important in the years ahead. As the worlds need for

Early advances in wind and solar technology were developed

energy continues to grow, bright people with new ideas will

with federal money. And energy research offers a strong

flock to the places that support efforts to make cheap, reliable,

return on investment. The Department of Energy found that

clean energy. Any country that wants to lead the world in

the $17.5 billion (adjusted for inflation) it spent on research

innovation and energy independence should be doubling

between 1978 and 2000 provided a yield of $41 billion. A

down on this research.

study by the OECD found a strong link between increased

funding for R&D and a rise in high-value patents for cleanenergy technology.13
If investments in energy research are so valuable, why
not simply let the market guide them? The answer is that
the energy industry differs from other industries in some
crucial ways.

How much are governments investing today? The United

States spends more on energy research than any other country,
so its federal budget is a good proxy. In the chart below, the
first two columns show total spending and the U.S. research
budget in three sectors: energy, health care, and defense. The
third column shows the ratio of government research to total
spending in each sector.15
The contrast is striking. The ratio of government R&D to total

Most technology companies know within a couple of years

spending in energy is 0.4 percent. In health, the ratio is more

whether a given innovation will pay off. Pharmaceutical

than twice as high. In defense, it is more than 22 times higher.

companies work on roughly the same time scale. But energy

research can take decades to pay off, because it takes so long
to adopt new technologies. Rudolf Diesel invented his engine

T O TA L
U.S. SECTOR
SPEND (2013)

U.S. GOV R&D

SPEND (2013)

R AT I O O F R & D
T O T O TA L
SECTOR SPEND
(2013)

in the late 19th century, but the diesel engine did not become

ENERGY

$1.4 TRILLION

$5. 3 BILLION

0.4%

commonplace until well after his death in 1913.

HEALTH CARE

$2.9 TRILLION

$31 BILLION

1.1%

DEFENSE

$6 40 BILLION

$69.8 BILLION

11%

13

International Energy Agency, Global Gaps in Clean Energy R&D (2010).

International Energy Agency, Global Gaps in Clean Energy R&D (2010).
Figures in the table were drawn from the U.S. Energy Information Administration (http://www.eia.gov/totalenergy/data/annual/pdf/sec1_13.pdf and http://www.eia.gov/state/seds/data.cfm?incfile=/state/seds/
sep_sum/html/sum_ex_tx.html&sid=US), Centers for Medicare and Medicaid Services (https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/
NationalHealthAccountsHistorical.html), Belfer Center for Science and International Affairs at Harvard University (http://belfercenter.ksg.harvard.edu/publication/24065/doe_budget_authority_for_energy_re
search_development_demonstration_database.html), and U.S. Bureau of Economic Analysis (http://www.bea.gov/industry/xls/io-annual/GDPbyInd_GO_NAICS_1997-2014.xlsx). Sector spend for defense is the
U.S. defense budget.

14
15

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

To make the point another way: In the United States,

G E T

I T

P5

could make reliable zero-carbon energy available to everyone.

consumers spend more on gasoline in a week than the

Some of these companies will fail. We expect the successful

government spends on clean-energy research in a year.

16

Some countries are stepping up to this challenge and

ones to attract large amounts of traditional capital investments

as they are demonstrated and deployed.

investing more in clean-energy research. Through

Over the next year, we will be analyzing potential investments,

Mission Innovation (www.mission-innovation.net),

creating vehicles to support this effort, and expanding the pool

more than 10 governments have committed to significant

of investors who join us. In the meantime, we have agreed on

increases. I am confident that the money can be put to

five principles that will guide our work. We will:

good use quickly, thanks to the outstanding university and

laboratory systems that already exist in the United States
and other countries. There is research capacity to spare
it has lacked only the necessary funding.

INV EST EA R LY

We will take a flexible approach to early-stage work,

providing seed, angel, and Series A investments, with the
expectation that once these investments are de-risked,

The Role of Markets

Private-sector efforts are the flip side of government
research. Energy is already a trillion-dollar market, and clean
energy could one day be a multi-trillion-dollar market. But
private investors are reluctant to get into the field, for the
same reason that energy companies tend to under-invest in
R&D: Breakthroughs can take decades to play out and their
inventors see relatively little reward. Although the profit
motive is not entirely absent, it is muted because investors
have little incentive to take risks. Promising concepts and
viable products are separated by a Valley of Death that
neither government funding nor conventional investors can
bridge completely.
A key part of the solution is to attract investors who can
afford to be patient, and whose goal is as much to accelerate
innovation as it is to turn a profit. I am joining with a
number of other investors who are fortunate enough to be
in this position.
We are creating the Breakthrough Energy Coalition
(www.breakthroughenergycoalition.com), a global,
private investment group which intends to help promising
approaches cross the Valley of Deathto take the risks that
allow companies to get innovation out of the lab and into the
marketplace. We will focus on early-stage companies that

traditional commercial capital will invest in the later stages.

2

INV EST B ROA D LY

Because it is too early to tell which paths to a clean-energy

future will succeed, we will support a wide range of
approaches in a number of sectors: electricity generation and
storage, transportation, industrial use, agriculture, and energy
system efficiency.
3

INV EST B O LD LY

We will look for novel technologies as well as ways to make

existing technologies dramatically cheaper, more efficient, or
more scalable. We will require a credible pathway to scaling
up rapidlyproviding affordable energy to the greatest
number of people without overburdening essential resources.
4

INV EST W ISELY

This effort requires a deep understanding of complicated

science and technology as well as public policy. We will
work with leading public and private institutions and energy
experts to guide our decisions.
5

INV EST TO G ETH ER

Because so many breakthroughs rely on government research,

we will focus our investments on countries that commit to
expanding their funding for research on clean, affordable,
reliable energy. The countries that have come together in the
Mission Innovation initiative are showing great leadership on
this issue.

According to the U.S. Energy Information Administration (http://www.eia.gov/tools/faqs/faq.cfm?id=23&t=10), Americans used more than 374 million gallons of gasoline per day in 2014. At an average of $2.50 /
gallon, this works out to more than $6.5 billion per week.
16

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

Promising Research

G E T

I T

P6

PHOTOANODE

2H 2 0

4e CATALYST

0 2 + 4H +

As an investor, I have been involved with many companies

ION EXCHANGE

working on promising energy solutions. I wish there were

MEMBRANE

many more to choose from. There should be hundreds if

ELECTR ICAL
MEMBRANE

not thousands of companies around the world exploring

H+

different approaches.

CATALYST
4H +

Below I give three examples of promising technologies I

have learned about. Researchers are exploring other exciting
ideas too; I am mentioning these only to illustrate a few types

4e

2H 2

PHOTOCATHODE

But they show how ingenious researchers throughout the

Solar chemical cells would use sunlight to make fuel.

They work much the same way a plant uses photosynthesis to convert sunlight
into carbohydrates. The simplest application involves cells that use sunlight
to generate enough energy to split water, producing oxygen and hydrogen
gas. The hydrogen can be used directly as fuel or in commercial processes

United States and around the world are developing ideas that

like making fertilizer.

could solve the energy problem. The faster governments and

Source: California Institute Of Technology

of ideas that deserve support. None of them is likely to be

ready to deploy for at least a decade, if they even get that far.

investors accelerate research, the faster we can make the

transition to clean, affordable, reliable energy.

The Challenge
The technology is still in its first generation. One challenge
is that researchers need to develop a light-absorberthe top
piece of bread in the club sandwichthat will work under

SOLAR CHEMICAL

the same conditions as all the other pieces.

The Opportunity
Solar photovoltaic technology uses photons to generate
electricity from sunlight. Solar thermal uses mirrors to
channel heat from the sun. Solar chemical takes a different
approach, using solar energy to create fuel.

F L O W B AT T E R I E S

The Opportunity
The current gold-standard in storing electricity is the

It works much the same way a plant uses photosynthesis to

lithium-ion battery. Some researchers are focused on

convert sunlight into carbohydrates. The simplest application

optimizing this technology, which has important applications

involves a kind of club sandwich of cells: a series of catalysts

including in cars. Another promising storage technology that

separated by a membrane, and surrounded by light-absorbing

deserves more investment is called a flow battery. Rather

material on the top and bottom. These cells use sunlight to

than using self-contained cells, as lithium-ion batteries do,

generate enough energy to split water, producing oxygen

flow batteries use a rechargeable liquid electrolyte inside two

and hydrogen gas; the hydrogen can be used directly as fuel

pairs of tankstwo receiving tanks and two holding tanks.

or in commercial processes like making fertilizer. Another

While the battery is being charged, the liquid flows from

approach uses water, carbon dioxide, and sunlight to create

the receiving tanks to the holding tanks, through a radiator-

hydrocarbonsmaking it possible to produce and burn fuels

like set of fins that charge the liquid with energy. Energy is

with no net gain or loss of carbon dioxide in the atmosphere.

released when the liquid flows out of the receiving tanks,

Solar chemical would put us on a path to decarbonizing

both the electricity and transportation sectors. It would also
help a lot with the storage problem, because the world is

back through the fins and into the holding tanks. The tanks
could be as small as a fish tank or as large as a swimming pool;
the larger the tank, the more energy it would hold.

already very good at storing fuels and moving them around in

While flow batteries would be impractical for personal

pipelines, oil tankers, and other infrastructure.

devices (imagine a cell phone the size of a fish tank), their

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

G E T

I T

P7

scale could prove very useful for large industry. Another

advantage is that, unlike lithium-ion batteries (whose
capacity falls by half after 1,500 charges), a flow battery
could last for decades and the rechargeable electrolyte
liquid could last indefinitely.

Solar paint would transform almost any surface into a cheap solar panel.
It could work on rooftops, walls, cars, cell phones, and more. It involves applying
a conductive layer, then a white base layer, and finally a light-sensitive dye on
topthis is what generates the electricityand applying heat to cure the paint.
In theory, it would be as straightforward as painting a wall in your house.
Source: NDnano, University of Notre Dame
TANK 1

TANK 2

PUMP 1

PUMP 2

MEMBRANE

The Challenge
Researchers are trying to refine the light-sensitive dyes that
generate electricity. The most efficient dyes come from a
family of chemicals called perovskites, and they convert about

Flow batteries would last longer and store more energy than
todays batteries.
They use a rechargeable liquid electrolyte inside two pairs of tankstwo
receiving tanks and two holding tanks. Energy is stored and discharged by
moving the liquid between the tanks. These batteries can be as small as a
fish tank or as big as a swimming pool, making them more attractive for
industrial use than todays batteries.

20 percent of solar energy into electricitya higher rate than

all but the most efficient solar panels currently on the market.
Unfortunately, these dyes contain lead, which is poisonous.
(The United States has banned lead-based paints from houses
and public buildings.) Researchers will need to develop a
non-toxic dye that is efficient, stable, and readily applied by

The Challenge
Many working prototypes use vanadium, a relatively rare
element, as the active electrolyte. Future systems will need
to use a more easily available electrolyte before they are
commercially feasible.

S O L A R PA I N T

The Opportunity
Although the cost of solar panels is coming down quickly,
installing and maintaining them remains expensive. The idea
behind solar paint is to make solar power much cheaper and
easier to install. Almost any surface could be transformed
into a cheap solar panel: rooftops, walls, cars, cell phones, and
more. It involves applying a conductive layer, then a white
base layer, and finally a light-sensitive dye on topthis is
what generates the electricityand applying heat to cure the
paint. In theory, anyone could do this; it would be almost as
straightforward as painting a wall in a house.

anyone, anywhere there is sunlight.

E N E R GY

I N N OVAT I O N

W H Y

W E

N E E D

I T

A N D

H OW

TO

G E T

I T

P8

Conclusion
It is hard to overstate the impact that clean, affordable, reliable energy will have. It
will make most countries energy-independent, stabilize prices, and provide low- and
middle-income countries the resources they need to develop their economies and help
more people escape povertyall while keeping global temperatures from rising more
than 2 degrees. I am optimistic that the next 15 years can bring the big breakthroughs
we need to accomplish all of these things.
This is a fantastic opportunity. It is also an unmistakable challenge. Humans have
changed their energy diets before, but never as rapidly as we need to today. Moving
this fast is unprecedented, which is all the more reason to start now.