ENERGY TRANSITIONS

ALSO BY VACLAV SMIL

China’s Energy
Energy in the Developing World
Energy Analysis in Agriculture
Biomass Energies
The Bad Earth
Carbon Nitrogen Sulfur
Energy Food Environment
Energy in China’s Modernization
General Energetics
China’s Environmental Crisis
Global Ecology
Energy in World History
Cycles of Life
Energies
Feeding the World
Enriching the Earth
The Earth’s Biosphere
Energy at the Crossroads
Creating the 20th Century
Transforming the 20th Century
Energy: A Beginner’s Guide
Energy in Nature and Society
Oil: A Beginner’s Guide
Global Catastrophes and Trends
Why America is Not a New Rome
Prime Movers of Globalization
ENERGY TRANSITIONS
History, Requirements, Prospects

Vaclav Smil
Copyright 2010 by Vaclav Smil
All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system, or transmitted, in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, except for the inclusion of brief quotations in a
review, without prior permission in writing from the publisher.
Library of Congress Cataloging-in-Publication Data
Smil, Vaclav.
Energy transitions : history, requirements, prospects / Vaclav Smil.
p. cm.
Includes bibliographical references and index.
ISBN 978–0–313–38177–5 (hard copy : alk. paper) — ISBN 978–0–313–38178–2 (ebook)
1. Renewable energy sources. 2. Power resources. I. Title.
TJ808.S64 2010
333.79—dc22 2010002985
ISBN: 978–0–313–38177–5
EISBN: 978–0–313–38178–2
14 13 12 11 10 1 2 3 4 5
This book is also available on the World Wide Web as an eBook.
Visit www.abc-clio.com for details.
Praeger
An Imprint of ABC-CLIO, LLC
ABC-CLIO, LLC
130 Cremona Drive, P.O. Box 1911
Santa Barbara, California 93116-1911
This book is printed on acid-free paper
Manufactured in the United States of America
 C ONTENTS

The Book’s Raison d’être vii

1 Energy Systems: Their Basic Properties 1
Resources and Prime Movers 2
Conversions and Uses 7
Infrastructures and Impacts 12
Energy Transitions 17
2 Global Transitions: Universal Patterns 25
Grand Fuel Sequence: From Biomass to Coal and Hydrocarbons 26
A New Quality: Generation of Electricity 39
History of Prime Movers: From Muscles to Machines 48
Quantifying the Transitions: Uncertainties and Trends 61
3 National Transitions: Commonalities and Particularities 75
European Experience: Britain, France, and the Netherlands 78
U.S. Transitions: The Consuming Superpower 85
Japan and China: The Asian Leaders 92
New Energy Superpowers: Russia and Saudi Arabia 99
4 Coming Transitions: Expectations and Realities 105
Renewable Energies: Resources, Conversions, and Constraints 107
Pace of Transitions: Innovation, Infrastructures, and Inertia 119
National Aspirations: Goals and Realities 129
Cautious Anticipations: Trends and Possibilities 142
Appendix 155
References 157
Index 167
This page intentionally left blank
 T HE B OOK ’ S R AISON D ’ ÊTRE

T
he generic meaning of transitions—as passages from one condition or
action to another—is quite straightforward and hence readily understood,
but adding the energy qualifier complicates the comprehension. Energy, a
concept that in itself is notoriously hard to define in an easy intuitive manner,
encompasses a veritable universe of states and processes, and that is why the term
energy transitions deserves some annotation. The focus should be always on a
process, not just on its initial and concluding stages, but the most revealing
analysis must deal with several key variables and use different measures to trace
their change.
There is no formal or generally accepted hierarchy of meanings, but the term
energy transition is used most often to describe the change in the composition
(structure) of primary energy supply, the gradual shift from a specific pattern of
energy provision to a new state of an energy system. This change can be traced
on scales ranging from local to global, and a universally experienced transition
from biomass to fossil fuels is certainly its best example. Many specific inquiries
are possible within this grand shift: For example, the focus can be on transitions
from wood to charcoal in heating, from coal to oil in households and industries,
from oil to natural gas in electricity generation, or from direct combustion of
fossil fuels to their increasingly indirect use as thermal electricity.
These studies of changing structure of energy supply often focus on the time
elapsed between an introduction of a new primary energy source and its rise to
claiming a substantial share (arbitrarily defined) of the overall market, or even
becoming the single largest contributor or the dominant supplier on a local,
national, or global scale. But given an often impressive growth of energy supply
over time, close attention should be also given to absolute quantities involved in
the transitions as well as to qualitative changes that result in wider availabilities
of energies that are more flexible, more efficient, and more convenient to use even
as they create substantially lower environmental impacts. Combination of all of
these approaches would provide the best understanding of the transition process.
viii THE BOOK’S RAISON D’ÊTRE

But the study of energy transitions should be also concerned with gradual
diffusions of new inanimate prime movers, devices that had replaced animal and
human muscles by converting primary energies into mechanical power. Focus on
the prime movers also brings to the forefront the notion of a transition as a process
of successful technical and organizational innovation, and energy transitions can be
also studied as specific subsets of two more general processes of technical innovation
and resource substitution. I will use all of these approaches in my examination of
global and national energy transitions.
There is only one thing that all large-scale energy transitions have in
common: Because of the requisite technical and infrastructural imperatives and
because of numerous (and often entirely unforeseen) social and economic impli-
cations (limits, feedbacks, adjustments), energy transitions taking place in large
economies and on the global scale are inherently protracted affairs. Usually they
take decades to accomplish, and the greater the degree of reliance on a particular
energy source or a prime mover, the more widespread the prevailing uses and
conversions, the longer their substitutions will take. This conclusion may seem
obvious, but it is commonly ignored: Otherwise we would not have all those
repeatedly failed predictions of imminent triumphs of new sources or new prime
movers.
And an inherently gradual nature of large-scale energy transitions is also the key
reason why—barring some extraordinary and entirely unprecedented financial
commitments and determined actions—none of today’s promises for greatly accel-
erated energy transition from fossil fuels to renewable energies will be realized.
A world without fossil fuel combustion is highly desirable and (to be optimistic)
our collective determination, commitment, and persistence could hasten its
arrival—but getting there will exact not only a high financial and organizational
cost but also persistent dedication and considerable patience. As in the past, the
coming energy transitions will unfold across decades, not years—and a few facts
are as important for appreciating energy prospects of modern civilization as is an
informed appreciation of this reality. This, in just half a dozen paragraphs, is the
book’s raison d’être.
 THE BOOK’S RAISON D’ÊTRE ix

Units and prefixes used in this book

Units
a are area
g gram mass
Hz hertz frequency
J joule energy
K Kelvin temperature
L liter volume
m meter length
m2 square meter area
m3 cubic meter volume
Mtoe million t of oil equivalent energy
N newton force
Pa pascal pressure
ppm part per million concentration
t tonne (metric ton) mass
W watt power
Wh watt-hour energy
Prefixes
h hecto- 102
k kilo- 103
M mega- 106
G giga- 109
T tera- 1012
P peta- 1015
E exa- 1018
Z zetta- 1021
Y yotta- 1024
This page intentionally left blank
 Chapter 1

E NERGY S YSTEMS : T HEIR

BASIC P ROPERTIES

A
ny anthropogenic energy system—that is any arrangement whereby the
humans use the Earth’s resources to improve their chances of survival
and to enhance their quality of life (and, less admirably, also to increase
their individual and collective power and to dominate, and to kill, others)—
has three fundamental components: natural energy sources, their conversions,
and a variety of specific uses of the available energy flows. The simplest systems
in the past tapped only a small number of sources by using just one or two kinds
of inefficient energy conversions for basic, and mostly precarious, subsistence,
while modern systems can draw energy from numerous natural sources, convert
them in many (and increasingly efficient) ways and use them in a myriad of ways
in order to power complex high-energy societies.
Existence of the earliest hominin foragers was not that different from the sur-
vival of scavenging omnivorous animals as their somatic energy (conversion of
food into muscle power) was just a segment of naturally cascading energy degra-
dation beginning with solar radiation and ending with the dissipation of heat
during walking, running, and gathering food. Our hominin ancestors may have
used the first deliberate extrasomatic energy conversion as early as nearly
800,000 years ago by mastering the control of fire (Goren-Inbar et al., 2004).
In contrast, modern high-energy societies tap many natural energy stores and
flows, convert them by using some astonishingly sophisticated devices, and use
them for purposes ranging from intensive food production to rapid long-
distance travel. In today’s world final per capita energy consumption ranges over
two orders of magnitude, from the miseries of the sub-Saharan Africa to the
excesses of the richest urban societies of America, Europe, and Asia. And in the
most affluent societies even the average per capita energy use is now well beyond
the level required for healthy and comfortable living.
All energy systems require infrastructures and their operation consumes con-
siderable amounts of energy. Energy infrastructures comprise not only tangible
components (exemplified by high-voltage transmission lines or pipelines) but—
in order to extract, store, and process fuels and harness energy flows—they also
include intangible organizational and managerial arrangements. Energy cost of
2 ENERGY TRANSITIONS

energy is obviously a critical determinant of the viability of any energy system as

only high-energy returns can create affluent societies with plenty of time left for
leisure. These inescapable costs of energy are not measured in energy terms but
are monetized as capital and operating costs. In the long run, most energy prices
have shown some very impressive declines, particularly when compared in terms
of actually delivered energy services (such as the cost of a lumen of light or a
passenger-kilometer flown).
All anthropogenic energy systems also create environmental impacts, ranging
from locally devastating deforestation to globally worrisome changes of the
atmospheric composition, above all the emissions of CO2, CH4, SO2, NOx,
and volatile organic compounds from fossil fuel combustion that have been
responsible for increasing tropospheric temperatures, acid deposition, photo-
chemical smog, and higher ground ozone levels. Some of the potentially highly
damaging externalities arising from energy conversions have been either com-
pletely eliminated or reduced to acceptable levels by resorting to better produc-
tion techniques and to efficient controls: Surface coal extraction and flue gas
desulfurization are two excellent examples. Others, most notably the greenhouse
gas emissions, are yet to be factored into the real cost of energy.
And, obviously, all energy systems evolve. During the preindustrial era there
were only very slow changes in the composition of the primary energy supply
(dominated by biomass fuels) and in the use of prime movers (dominated by
human and animal muscles)—but the last two centuries have seen a series of
remarkable energy transitions. These changes can be traced (and where statistics
allow, be studied in revealing quantitative details) as shifts in the shares of indi-
vidual fuels and in the origins of electricity generation as well as the adoption
and diffusion rates of new prime movers and as new patterns of final energy uses.
Scores of books could be consulted to get a more detailed understanding of the
matters introduced in this chapter: Fouquet (2008), Smil (2003, 2008), and
WEC (2007) might be convenient places to start.

RESOURCES AND PRIME MOVERS

Energies used by human societies can be classified dichotomously according
to their origins either as renewable and nonrenewable or primary and secondary.
Renewable energies include solar radiation (radiant or electromagnetic energy)
and all of its biospheric transformations: plant mass (phytomass) formed by
photosynthetic conversion of solar radiation into chemical energy of plant tis-
sues; wind, arising from pressure gradients created by differential heating of
the ground; moving water originating in radiation-driven evaporation and pre-
cipitation (stream flows) or as wind-driven waves and ocean currents; and the
temperature difference between the surface of tropical oceans and dark cold
waters below the thermocline (water layer, usually about 200 m thick, whose
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 3

temperature fluctuates in contrast to deeper layers that stay at about 4°C). There
is yet another renewable flux, the Earth’s heat (geothermal energy) generated by
the decay of heat-producing isotopes in the planet’s crust and by heat rising from
its core.
The spectrum of solar radiation contains the shortest gamma rays and x-rays,
ultraviolet light (<400 nm), visible wavelengths (400–700 nm), and infrared
(>700 nm). Nearly all of the UV wavelengths are screened by the stratospheric
ozone layer, almost exactly 30% of the incoming radiation is reflected back to
space and 20% is absorbed by the atmosphere; as a result, solar energy reaching
the ground is only half of the solar flux in space. Active use of solar energy to
generate electricity or to produce hot water is still rather limited, but all build-
ings have always benefited from passive solar heating, and architectural design
can enhance this reality by optimizing the orientation of buildings, ingress of
winter rays into rooms, and blocking of summer rays.
Photosynthesis uses only a small part of available wavelengths (principally
blue and red light amounting to less than half of the energy in the incoming
spectrum) and its overall conversion efficiency is no more than 0.3% when mea-
sured on the planetary scale and only about 1.5% for the most productive terres-
trial (forest) ecosystems. Phytomass produced by photosynthesis is dominated by
carbohydrates and absolutely dry phytomass has a fairly uniform energy density
of about 18 MJ/kg; air-dry wood, the most important fuel for household heating
and cooking and small-scale manufacturing in all preindustrial societies, con-
tains about 15 MJ/kg, as do various cereal and legume straws and stalks that have
been burned by households in arid and deforested regions.
Only a very small part of insolation (no more than 2%) energizes the global
atmospheric circulation but the total power of winds generated by this differen-
tial heating is a meaningless aggregate when assessing resources that could be
harnessed for commercial consumption because the Earth’s most powerful winds
are in the jet stream at altitude around 11 km above the surface, and in the
northern hemisphere their location shifts with seasons between 30° and 70° N.
Even at altitudes reached by the hubs of modern large wind turbines
(70–100 m above ground) only less than 15% of winds have speeds suitable
for large-scale commercial electricity generation. Moreover, their distribution is
uneven, with the Atlantic Europe and the Great Plains of North America being
the premiere wind-power regions and with large parts of Europe, Asia, and
Africa having relatively unfavorable conditions.
Similarly, the total potential energy of the Earth’s runoff (nearly 370 EJ, or
roughly 80% of the global commercial energy use in 2010) is just a grand sum
of theoretical interest: Most of that power can be never tapped for generating
hydroelectricity because of the limited number of sites suitable for large dams,
seasonal fluctuations of water flows, and the necessity to leave free-flowing
sections of streams and to store water for drinking, irrigation, fisheries, flood
control, and recreation uses. As a result, the aggregate of technically exploitable
4 ENERGY TRANSITIONS

capacity is only about 15% of the theoretical power of river runoff (WEC,
2007), and the capacity that could be eventually economically exploited is obvi-
ously even lower.
There are four other water-based energy resources: tidal power and, as already
noted, wind-driven waves, ocean currents, and the difference in temperature
between the warm ocean surface and cold deeper waters. Each of them has a sig-
nificant overall global potential but none of them is easy to harness. Large-scale
tidal projects have remained in the conception/proposal stage for decades,
wave-harnessing devices are in their early development stage, there have been
no serious attempts to capture the power of major ocean currents, and even in
the warmest tropical seas (where the difference between the surface and deep
water surpass 20°C) the ocean thermal differences can be tapped for electricity
generation only with a very low efficiency and none of a few isolated experiments
with such generation had progressed to commercial projects.
Fossil fuels are by far the most important nonrenewable energies: All coals
and most hydrocarbons (crude oils and natural gases) are transformations of
ancient biomass, buried in sediments and processed by high pressures and tem-
peratures (for millions to hundreds of millions of years), but a significant share
of natural gases may be of abiogenic origin. All fossil fuels share the dominant
presence of carbon, whose content ranges from nearly 100% in the best anthra-
cite coals to 75% in methane; most common bituminous coals used in electricity
generation, as well as most hydrocarbons, contain sulfur (a mere trace in some
gases, up to 4% in some coals, with 2% being a common mean). Coals also con-
tain varying shares of incombustible ash and moisture, as well as traces of heavy
metals that are also present in many crude oils, and natural gases often contain
dissolved nitrogen, water, and hydrogen sulfide.
Energy density of coals ranges from just 8 MJ/kg for low-quality lignites to
about 30 MJ/kg for the best anthracites, with most bituminous (steam) coals
between 20 and 25 MJ/kg. Crude oils are much more uniform (40–42 MJ/kg),
as are the natural gases (mostly between 35 and 40 MJ/m3). Resources of fossil
fuels (their total mass present in the Earth’s crust) are not known with a high
degree of certainty, and their reserves (that part of resources that is recoverable
with existing technical means and at profitable costs) keep changing as new tech-
niques (such as horizontal drilling or steam-assisted recovery of oil from oil
sands) lower their extraction cost to the point that previously uneconomical
deposits become profitable sources of energies.
Resource recovery and depletion has engendered passionate debates about an
imminent peak of global crude oil production, about the eventual magnitude of
natural gas resources, and about the durability of coal deposits. What is not in
doubt is that a large share of fossil fuel resources will be never exploited because
their extraction and conversion would be technically forbidding or exceedingly
costly: This is true about thin seams of poor-quality coal located at great depths
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 5

as well as about many tiny hydrocarbon reservoirs or very heavy oils or deeply
buried oil sands and oil shales. The same conclusion applies to fissionable
materials abundant in very low concentrations in many rocks as well as in
seawater.
Nuclear energy can be released either by fission of the isotopes of the heaviest
natural elements (a process exploited in all nuclear electricity-generating plants)
or by fusion of the lightest ones (a process whose commercial realization has been
a frustratingly receding mirage). Since the late 1950s uranium fission has been used
in commercial nuclear stations to generate electricity by the same means as in
fossil-fueled stations (i.e., expanding the pressurized steam in a turbine that rotates
a generator). In contrast, there are no fusion-based plants; none are even on a
distant horizon and fusion may remain nothing but an ever-receding promise.
Division of energies into primary and secondary categories is based on the
method of their production. Primary fuels (stores of chemical energy) are harvested
(wood, crop residues) or extracted from the uppermost strata of the Earth’s crust
(all fossil fuels, including peats, coals, crude oils, and natural gases). Their combus-
tion provides heat (thermal energy) or light (electromagnetic or radiant energy).
Their processing to yield secondary fuels may change only their physical state
(making solid briquettes by compressing coal dust, with or without binders), but
it usually involves chemical transformation.
The only secondary fuel in preindustrial societies was charcoal made by
pyrolysis (thermal decomposition in the absence of oxygen) of woody phyto-
mass. With all volatile components driven out, the fuel is virtually pure carbon,
nearly smokeless (its well-oxidized combustion produces only CO2), and with
high energy density of almost 30 MJ/kg (see Figure 1.1). Coke, made by high-
temperature pyrolysis of coal, was first used in England during the 1640s in malt
roasting, but only when its cost declined sufficiently did it begin to replace char-
coal as a fuel in blast furnaces by the middle of the eighteenth century, and it has
remained the fuel of choice for all primary iron production ever since. During
the nineteenth century another secondary fuel—coal gas (town gas or manufac-
tured gas)—became a common urban illuminant (in- and outdoors) as well as a
fuel for cooking; it was eventually displaced by electric lights and natural gas, but
in some cities its use lingered until after World War II.
Today’s most important, as well as by far the most common, secondary fuels
are various liquids produced by refining crude oils. Refining was done initially
by simple thermal distillation (fractions separated by temperature); now the
crude oils are transformed with the help of catalytic cracking used to produce
higher shares of gasoline and jet fuel (kerosene), lighter and more valuable fuels
that power passenger cars and airliners. Heavier diesel oil is also used to fuel cars
but its principal consumer is truck and railways transport, while the heaviest
residual oil powers the marine transportation. Diesel oil and residual fuel oil
are also used in stationary generation of electricity.
6 ENERGY TRANSITIONS

Figure 1.1 Steps in preparing wood piles for charcoaling illustrated in Diderot and
D’Alembert’s L’Encyclopédie (1769–1772).

Commercial electricity generation and transmission added a new dimension

to human energy use and, as in the case of fuels, electricity’s origin is classified
as either primary or secondary. Primary electricity involves all conversions of
natural, renewable energy flows including those of water and wind, the Earth’s
heat, and solar radiation. Primary electricity could also be generated by harnessing
ocean waves and the temperature differences between the surface layer of the
warmest ocean and constantly cold waters underneath. Nuclear electricity is yet
another form of primary energy, with steam for large turbogenerators derived from
controlled splitting of uranium. Secondary electricity uses heat released from the
combustion of fossil fuels, mainly coal for steam turbogenerators and natural gas
for gas turbines.
Prime movers are energy converters able to produce kinetic (mechanical)
energy in forms suitable for human uses. Human muscles (somatic energy) were
the only prime movers (converting chemical energy in food to kinetic energy
of walking, running, and countless manual tasks) until the domestication of
animals provided more powerful animate prime movers used in fieldwork, trans-
portation, and for some industrial tasks. Animate prime movers continued to
dominate energy use long after the introduction of first mechanical prime mov-
ers, beginning with simple sails, followed, millennia later, by small water wheels,
and roughly another millennium afterwards by small windmills.
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 7

During the eighteenth century the steam engine became the first mechanical
prime mover powered by the combustion of fuels. Steam turbine and two key
types of internal combustion engines (sparking gasoline-fueled machine and
non-sparking engine fueled by heavier fuels or by residual oils) were invented
before the end of the nineteenth century, and gas turbine became practical
during the 1930s. Electric motors present a classification dilemma: They are,
obviously, prime movers in the sense of the definition I offered at the outset of
the preceding paragraph, but they are powered by electricity that has been pro-
duced by prima facie prime movers, be it steam turbogenerators or gas, water,
and wind turbines.
Major criteria used to classify energy uses, as well as the deployment of prime
movers, are the location of the conversion process, temperature of the final use,
and principal economic sectors. Stationary combustion provides space heating
for households, public institutions, and industries, as well as hot air and steam
for industrial processes. Stationary prime movers (dominated by steam turbo-
generators and water turbines) produce the world’s electricity and electric motors
and internal combustion engines power most of the modern industrial processes.
Heavy horses were the most powerful commonly used mobile prime movers in
preindustrial societies. Mobile steam engines, introduced between 1805 and
1835, revolutionized both land and water transportation and dominated the
two sectors until the middle of the twentieth century.
Mobile steam turbines were first used in ship propulsion at the beginning of
the twentieth century, but marine transport became eventually dominated by
diesel engines. Diesels also power heavy road transport and a variety of off-
road vehicles, while the automotive gasoline-fueled internal combustion engines
emerged as the world’s most numerous mobile prime movers. Commercializa-
tion of gas turbines began during the late 1930s but their widespread adoption
had to wait until the 1960s. Larger stationary machines are used mostly in elec-
tricity generation and, starting in the 1950s, lighter and increasingly powerful
gas turbines rapidly displaced reciprocating internal combustion engines in
long-distance air travel. During the 1980s modified jet engines began to be used
also for stationary applications as aeroderivative turbines for peak demand or
decentralized electricity generation.

CONVERSIONS AND USES

Modern societies use many forms of energy in order to satisfy many final uses.
While there is no single binding classification of the uses that provide individ-
uals, households, cities, and economies with essential energy services, the princi-
pal categories include heat, light, industrial (overwhelmingly stationary) power,
and freight and passenger transport. All energy conversions involve some loss
of the capacity to perform useful work. This is the essence of the second law
8 ENERGY TRANSITIONS

of thermodynamics: in any closed system (i.e., one without any external supply of
energy), availability of useful energy can only decline. Energy remains conserved
(the first law of thermodynamics) but its practical utility is diminished because
disordered, dissipated low-temperature heat (the final product of all energy
conversions) can be never reconstituted as the original, highly organized fuel or
electricity. This is an irreversible process, as no action can reconstitute a tank full
of gasoline or a truckload of coal from the diffuse heat in the atmosphere.
While such considerations as comfort and convenience are hardly unimpor-
tant, the quest for higher conversion efficiencies underlies the evolution of
modern energy systems. The simplest definition of energy conversion is as the
ratio of output or transfer of the desired energy kind achieved by a converter
to the initial energy input (be it to an organism, a mechanical device or a com-
plex system). This rate does not capture the efficiency limitations due to the sec-
ond law. The second-law (or exergy) efficiency is expressed as the ratio of the
least available work that could have performed the task to the available work that
has been actually used in performing it. This measure provides a direct insight
into the quality of performance relative to the ideal process, and it is concerned
with a task to be performed, not with a device or a system used for that end.
As a result, all conversions using high-temperature combustion (flame in
excess of 1200°C) to supply low-temperature heat (to pasteurize food at 72°C,
to heat bath water to no more than 49°C in order to avoid third-degree burns)
will be particularly wasteful when judged in terms of the second-law efficiency.
But, as the following examples show, applying that efficiency to many human
actions may be actually irrelevant or inappropriate. One of the most efficient
ways to produce animal protein is carp aquaculture (as those cold-blooded her-
bivorous species have inherently low metabolic needs) while the most inefficient
way to produce animal protein is beef from cattle fed a mixture of corn and
soybeans in a giant feedlot. But most people with good incomes prefer to buy
beef, not carp. Similarly, corn is the most efficient staple grain crop—but unlike
gluten-rich hard wheat, its flour cannot be used to bake leavened breads. And a
periodic bleeding of cattle by Kenya’s Maasai is a vastly more efficient means
of converting grasses to food than slaughtering cattle for meat—but how many
societies would be ready to make such a switch?
Combustion, that is, rapid oxidation of carbon and hydrogen in biomass and
fossil fuels, has been the dominant energy conversion since the early stages of
human evolution. For hundreds of thousands of years of hominin evolution it
was limited to wood burning in open fires, and combustion of biomass fuels
remained the principal means of securing heat and light until the advent of
industrialization—and even the most advanced of today’s postindustrial societies
derive most of their useful energies from the burning of fossil fuels. What has
changed, particularly rapidly during the past 150 years, are the typical efficien-
cies of the process. In open fires less than 5% of wood’s energy ended up as useful
heat that cooked the food; simple household stoves with proper chimneys
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 9

(a surprisingly late innovation) raised the performance to 15–20%, while today’s

most efficient household furnaces used for space heating convert 94–97% of
energy in natural gas to heat.
The earliest commercial steam engines (Newcomen’s machines at the begin-
ning of the eighteenth century) transferred less than 1% of coal’s energy into
useful reciprocating motion—while the best compound steam engines of the late
nineteenth century had efficiencies on the order of 20% and steam locomotives
never surpassed 10% (see Figure 1.2). The first internal combustion engines
(stationary machines powered by coal gas) had lower efficiencies than the best
contemporary steam engines, and even today’s best-performing gasoline-fueled
engines do not usually surpass 25% efficiency in routine operation.
But the very first working prototype of Rudolf Diesel’s non-sparking engine
(officially tested in 1897) surpassed that rate and the world’s largest marine diesel
engines are now the only internal combustion machines whose efficiency can
reach, and even slightly surpass, 50%. For comparison, today’s best gas turbines
(used in aviation and electricity generation) are about 40% efficient (Figure 1.2).
When the hot gas ejected by large stationary gas turbines is used to heat water for a
steam turbine, this combined cycle gas turbine can reach overall efficiency of about
60%. In contrast, the maximum efficiency of coal-fired electricity-generating

Figure 1.2 Maximum efficiency of prime movers, 1700–2000. There has been an
order of magnitude gain during the last two centuries, from about 6% for steam
engines to about 60% for the combined-cycle gas turbines.
10 ENERGY TRANSITIONS

plants using the standard configuration of a boiler and a steam turbogenerator is

just over 40%.
Rising efficiency of individual conversions has been reflected in the improving
performance of entire economies. As a result, the difference between average per
capita energy use in modern and traditional societies is significantly greater when
compared in useful terms rather than as the rates of gross energy consumption.
For example, thanks to a relatively easy access to extensive and rich forests, the
average U.S. wood and charcoal consumption was very high: about 100 GJ/capita
in 1860, compared to about 350 GJ/capita for all fossil and biomass fuel at the
beginning of the twenty-first century. But as the typical 1860 combustion efficien-
cies were only around 10%, the useful energy reached only about 10 GJ/capita.
Weighted efficiency of modern household, industrial, and transportation conver-
sions is about 40% and hence the useful energy serving an average American is
now roughly 150 GJ/year, nearly 15-fold higher than during the height of the
biomass era.
Energy uses have undergone some significant changes even during the prein-
dustrial period when most fuels were used by households and in small-scale arti-
sanal manufactures, and when most prime movers were deployed in subsistence
agriculture. Expansion of manufactures and metallurgy led to a spreading use of
water power (most efficiently by larger vertical water wheels) and iron metal-
lurgy and the preference for smokeless fuel in richer urban homes had also cre-
ated higher demand for charcoal. Crop rotations including leguminous food
and cover crops enabled farmers to divert a greater share of harvests to animal
feeding and made it possible to deploy larger numbers of more powerful animals
in agriculture.
Industrialization brought a radical change in the composition of national
energy use as coal mining, metallurgy, and heavy machinery sectors became
eventually the leading consumers of energy, followed by light manufactures
(textiles and various consumer items) and a rapidly expanding land and sea
transportation. In Europe and North America this shift was accomplished
already before 1900. Households claimed a relatively small share of overall
energy use during the early phases of industrialization, first only as coal (or coal
briquettes) for household stoves, later also as low-energy coal (town) gas, and
(starting during the 1880s) as electricity for low-power light bulbs, and soon
afterwards also for numerous household appliances.
Subsequently, modern energy use has seen a steady decline of industrial and
agricultural consumption and increasing claims of transportation and household
sectors. For example, in 1950 industries consumed more than half of the world’s
primary commercial energy, at the time of the first oil crisis (1973) their share
was about one-third, and by 2010 it declined to about 25%. Major appliances
(refrigerators, electric stoves, washing machines) became common in the United
States after World War I, in Europe only after World War II, and private car
ownership followed the same trend. As a result by the 1960s households became
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 11

a leading energy-using sector in all affluent countries. There are substantial

differences in sectoral energy use among the industrializing low-income nations
and postindustrial high-income economies. Even after excluding all transporta-
tion energy, U.S. households have been recently claiming more than 20% of
the country’s primary energy supply in 2006, while in China the share was only
about 11%.
But the boundaries of standard sectoral classification can be redrawn to yield a
different breakdown. Perhaps most notably, modern agriculture consumes
directly only a few percent of the total energy supply as fuels and electricity to
operate field machinery (tractors, combines, irrigation pumps) and mostly as
electricity for heating, cooling, and machinery used in large-scale animal
husbandry. But the indirect energy cost of agricultural production (to produce
agricultural machinery, and to synthesize energy-intensive fertilizers, pesticides,
and herbicides) and, even more so, energy costs of modern industrial food
processing (including excessive packaging), food storage (the category domi-
nated by refrigeration), retailing, cooking, and waste management raise the
aggregate cost of the entire food production/distribution/preparation/disposal
system to around 15% of total energy supply.
Inevitably, changing sectoral requirements have affected the final uses. Before
the advent of extensive steam-driven electricity generation (during the 1890s),
coal had four major final uses: as the leading household fuel, as the principal
source of both process heat and steam and mechanical power in industries, as
the prime energizer of land and water transport, and as the feedstock to produce
metallurgical coke needed to smelt pig iron. A century later, coal ceased to be an
important transportation fuel, only in a few countries was it still used for house-
hold heating and cooking, and its rising use was confined largely to only two
markets, the dominant one for electricity generation and a smaller one for coke
production.
Similarly, refined oil products were used first as illuminants and lubricants
and only the mass ownership of cars (the era that began in the United States with
Ford’s Model T in 1908) required mass production of gasoline. After World
War I diffusion of Diesel’s efficient engine in trucking and shipping claimed
the heaviest fuel oils, and the post–WWII commercialization of jet engines
made kerosene the third most important refined product. And natural gas
became the world’s premiere source of household heat only after 1950. There
were also some notable shifts in non-energy uses of fuels: During the late nine-
teenth century coal became an important feedstock for chemical industries, but
its use was soon displaced by crude oil and natural gas. Currently on the order
of 10% of all extracted oil and slightly more than 5% of all natural gas are used
as chemical feedstocks, above all for syntheses of ammonia and various plastics.
Another revealing classification that ignores the traditional sectoral divisions
is according to the prevailing temperature of final uses. Most energy needs
are for low-temperature heat, dominated by space heating (up to about 25°C),
12 ENERGY TRANSITIONS

hot water for bathing and clothes washing (maxima of, respectively, about 40°C
and 60°C), and cooking (obviously 100°C for boiling, up to about 250°C for
baking). As already noted, ubiquitous heat waste is due to the fact that most of
these needs are supplied by high-temperature combustion of fossil fuels. Steam
and hot water produced by high-temperature combustion also account for
30–50% of energy needs in food processing, pulp and paper, chemical and
petrochemical industries. High-temperature heat dominates metallurgy, produc-
tion of glass and ceramics, steam-driven generation of electricity, and operation
of all internal combustion engines.

INFRASTRUCTURES AND IMPACTS

Only the simplest harnessing and conversion of energies (gathering of woody
debris and its burning in primitive stoves) does not require special infrastruc-
tures whose existence must either precede a particular form of energy use or must
accompany its expansion. Some early infrastructures could be relatively simple.
For example, in the eighteenth century an unpaved road leading to a coal seam
outcropping in a previously uninhabited valley would make it possible to bring
in the material necessary for opening a small mine and to haul the mined coal
in horse-drawn wagons to the nearest settlement. But a large nineteenth century
mine would have to be connected to its markets by a railroad, or its coal would
be shipped by barges, and the mining of deeper seams could not be accom-
plished without first installing adequate steam-powered water pumping and
ventilation facilities.
Infrastructural needs reached an entirely new level with the exploitation of
hydrocarbons whose large-scale extraction requires complex and expensive infra-
structures. Pipelines are needed to carry the crude oil and natural gas to markets
(or to the nearest coast for overseas exports) and a pretreatment (separation of
water, brine, petroleum gases, or hydrogen sulfide) may be required before send-
ing such fuels by a pipeline. When natural gas is used for household heating it is
necessary to have voluminous storages to meet high winter peak demand. Crude
oil is too valuable a resource to be burned as is and it needs expensive refining
that converts it into gasoline, kerosene, diesel oil, residual oil, and non-energy
products (lubricants, paving materials).
Electricity generation presents an even more demanding case of infrastruc-
tural prerequisites. Not only it is necessary to have extensive networks of trans-
mission and distribution lines in place before any large-scale generation can
take place, it is also necessary to have large numbers of converters (lights, appli-
ances, electric motors, electric furnaces, electrochemical processes) ready to use
the delivered electricity. Consequently, size of electricity-generating stations has
been driven by rising demand—and it has been also constrained by the existing
(and anticipated) load. For example, the maximum size of turbogenerators in
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 13

the U.S. thermal stations stopped growing (and the average size had actually
declined) as the demand weakened during the 1970s. Perhaps the most exacting
infrastructural challenge has been presented by the exports of liquefied natural
gas (LNG). High costs of liquefaction plants, LNG tankers, and regasification
facilities mean that the economies of scale dictate the construction of a system
capable to deliver at least a million tonnes of gas a year.
Energy systems have also become more interdependent and their integration
has been steadily expanding. Preindustrial energy systems were just patchworks
of independent entities. Their spatial extent could have been as small as a village
that relied on nearby forests and on crop residues for all of its fuel and feed needs
and that produced virtually all of its food by growing a variety of crops in rota-
tions. Modernization began to enlarge the boundaries of energy systems, first
with railway and shipborne transport of coal, then with increasingly large-scale
production of industrial manufactures that were traded not only nationwide
but even overseas and with adoption of simple agricultural machines.
Today’s energy system is truly global, with nearly 50 countries exporting and
almost 150 nations importing crude oil (and with nearly as many trading refined
oil products), with more than 20 states involved in natural gas sales (either by
cross-border pipelines or by using tankers carrying liquefied gas), and with
nearly a dozen major coal importers and a similar number of countries with
substantial coal imports. Electricity is traded relatively less than coal, but even
so at least two dozen countries have interconnections of sufficient capacity to
carry on exchanges on a GW scale. Moreover, there are no national autarkies as
far as extraction, transportation, and processing of energy is concerned: Mining
machinery, oil and gas drilling rigs, pipelines, tankers, and coal-carrying vessels
and refineries are designed and made by a relatively small number of producers
in about a score of countries and used worldwide.
And design and production of the most powerful prime movers have seen an
even greater degree of concentration, with as few as two or three companies
dominating the global market. All of the world’s largest marine diesel engines
that power virtually all large commercial vessels (oil and gas tankers, bulk car-
riers, container ships) come from the duopoly of MAN Diesel and Wärtsilä
(and the companies license their engines to a small number of makers in Europe
and Asia) and all of the world’s most powerful jet engines are designed and made
by America’s General Electric and Pratt & Whitney and Britain’s Rolls-Royce, or
by alliances of these companies.
Because of energy’s central place in nature and in human affairs it is inevitable
that the massive burning of fossil fuels, fissioning of uranium, and capture of
renewable energy flows have many profound consequences for the performance
of economies and for the state of the environment, and hence for the overall
quality of life. Consequently, it is incredible that energy has never been a pri-
mary, not even a major, concern of modern economic inquiry. This also helps
to explain why modern societies began to deal with widespread environmental
14 ENERGY TRANSITIONS

impacts of energy use only after World War II. Modern studies of energy–
economy links have uncovered some broad commonalities that have marked
the path from traditional to industrial to postindustrial societies—but they are
perhaps no less notable for revealing many singularities and peculiarities. Envi-
ronmental impacts of energy use are often so difficult to appraise because there
can be no generally acceptable metric for valuing their consequences for biota,
climate, and human health.
Global growth of primary energy consumption has corresponded fairly
closely to the expansion of the world’s economic product: During the twentieth
century a roughly 17-fold expansion of annual commercial energy use (from
about 22 to approximately 380 EJ) produced a 16-fold increase of annual
economic output, from about $2 to $32 trillion in constant 1990 dollars
(Maddison, 1995; World Bank, 2001). Similarly close relationship is revealed
by studying historical statistics of many individual countries—but comparisons
among the countries clearly indicate that a given level of economic development
does not require an identical, or not even very similar, level of the total primary
energy consumption. This is true among low-income economies as well as
among affluent nations: France has certainly a much higher standard of living
than Russia even though the two countries consume primary energy at a very
similar per capita rate.
Fewer exceptions are found as far as the secular decline of average energy inten-
sity (energy use per unit of GDP) is concerned. That rate’s rise during the early
stages of industrialization (reflecting energy needs for new industrial and transpor-
tation infrastructures) is usually followed by a prolonged decline. The British peak
came early in the nineteenth century, the U.S. and Canadian peaks followed six to
seven decades later—but Japan reached its highest energy intensity only in 1970,
and China’s energy use per unit of the country’s GDP continued to rise until the
late 1970s but since that time the Chinese rate has fallen faster than in any previous
case: By 1990 it was 40% below the 1980 level, and by 2005 the decline reached
just over 70% (Fridley et al., 2008). But comparisons of national energy intensities
and their secular trends require careful interpretation because their differences are
caused by factors ranging from climate to consumer preferences, with the compo-
sition of primary energy consumption and the structure and efficiency of final
conversions as key factors.
Countries with harsh climate, generously sized houses, large territories, and
numerous energy-intensive industries will have relatively high national energy
intensities even if their specific energy conversions are highly efficient, while
countries undergoing modernization will have much higher intensities than
postindustrial economies. These realities help to explain why, for example,
Canada’s energy intensity is more than twice as high as that of Italy, and China’s
intensity is still more than twice that of Japan. Another long-term trend has been
the decarbonization of the global energy supply: The relative shift away
from coal (usually more than 30 kg of carbon/GJ) to liquid hydrocarbons
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 15

(averaging about 20 kg C/GJ) and natural gas (less than 15 kg C/GJ) and rising
generation of carbon-free primary electricity had lowered the carbon content of
the world’s primary energy supply by about 25% during the twentieth century, and
the slowly increasing share of renewable conversion will continue to lower that rate.
Technical innovation, economies of scale, and competitive markets have com-
bined to bring some impressive long-term declines of energy prices, particularly
when compared to rising disposable incomes or when expressed in terms of value
for delivered service. None of these declines has been more impressive than the
cost of electricity for lighting traced as constant monies per lumen: Fouquet
(2008) found that rising incomes, higher conversion efficiencies, and lower gen-
eration costs made the household lighting in the United Kingdom in 2000 about
160 times more affordable than in 1900. In contrast, inflation-adjusted prices of
coal and oil do not show a general declining trend but a great deal of fluctuation
and a remarkable constancy in the long run. When expressed in constant monies
crude oil prices were very low and very stable between the beginning of the twen-
tieth century and the early 1970s, they retreated rapidly after two OPEC-driven
price rises of 1973–1974 and 1979–1981, but their recent fluctuations offer no
safe foundation for looking ahead.
But energy prices have not been usually determined by free-market operation,
as energy industries in general (and oil industry and nuclear electricity genera-
tion in particular) have been among the greatest beneficiaries of government sub-
sidies, tax breaks, and special regulation. Some prices have been subject to cartel
control: In the United States the Texas Railroad Commission fixed prices by
allocating production quotas until March 1971, and since 1973 OPEC has used
its production quota to manipulate the global crude oil supply. Even more
importantly, no energy price expresses the real cost of the delivered service, as
the costs of often significant environmental and health externalities are not
included in the prices of fuels or electricity.
Internalization of these costs has been done adequately in some cases (electricity
cost is higher due to highly efficient capture of particulate matter by electrostatic
precipitators and removal of SO2 by flue gas desulfurization; all modern passenger
cars have three-way catalytic converters to reduce NOx, CO, and volatile organic
hydrocarbon emissions) but it remains a challenge in most instances, above all
because health effects account for most of the cost but are notoriously difficult to
monetize—as are the long-term ecosystemic effects of such complex processes as
photochemical smog, acid deposition, nitrogen enrichment, or climate change.
Strategic considerations further complicate the quest for the real price of energy:
Perhaps most notably, the Pentagon had been devoting a significant share of
its budget to the Middle East even before the 1991 Gulf War or before the Iraqi
invasion of 2003.
As for the environmental impacts of energy industries and uses, it is clear
that the anthropogenic emissions of CO 2 from the combustion of fossil
fuels have become one of the most prominent concerns of modern civilization.
16 ENERGY TRANSITIONS

Their global total rose from just over 0.5 Gt C in 1900 to nearly 8.5 Gt C by
2007, and they have been the major reason (deforestation, mainly in the tropics,
was the second most important contribution) for the rise of tropospheric
CO2 concentration (since 1957 continuously monitored at the Mauna Loa
observatory in Hawaii) from about 295 ppm in 1900 to 386 ppm by 2008
(Figure 1.3). This is a truly global phenomenon, as the average tropospheric con-
centrations rise no matter where the emissions take place. After being the leading
emitter for more than a century, the United States was surpassed in 2007 by
China (but in per capita terms there is a nearly four-fold difference).

Figure 1.3 Global emissions of CO2, 1850–2008 (in Gt C/year) and tropospheric
CO2 concentrations, 1958–2008. Plotted from emissions data in Rotty and Marland
(2009) and from Mauna Loa concentrations data in NOAA (2009).
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 17

Extraction and conversion of energy has many other environmental conse-

quences. Deforestation in the Mediterranean and in North China was the first
environmental manifestation of the growing human use of energy as emerging
cities and expanding metal smelting (first copper, then iron) needed more wood
and charcoal. Underground coal mining created aboveground disturbances
(subsidence, mountains of mine spoils) and localized water pollution (acid
runoff), but emissions of particulate matter and SO2 were the most important envi-
ronmental consequences of coal combustion as the two pollutants often reached
very high concentrations in large cities. After 1950 electrostatic precipitators virtu-
ally eliminated particulate pollution from large sources but long-distance transport
of SO2 (and also NOx) created serious regional to semi-continental problems with
acid deposition.
Extraction and transportation of crude oil created local water pollution and
accidental oil spills, and combustion of refined oil products provided the key
starting ingredients (NOx, CO, and volatile organic compounds) for photo-
chemical smog. Beginning in 1956 generation of electricity by fissioning ura-
nium introduced an entirely new set of environmental problems, ranging from
possibilities of accidental contamination to challenges of long-term storage of
high-level radioactivity waste. And renewable energy flows have a multitude of
their own environmental consequences, ranging from the alterations of water
quality and age caused by large dams (lower temperature, water aging behind
dams) to problems with esthetic acceptability of large wind turbine farms and
with their noise and effect on birds.

ENERGY TRANSITIONS
As this brief review of energy system fundamentals makes clear, there are
many components whose importance and performance evolve and hence there
are many energy transitions whose origins, progress, and accomplishments can
be studied on levels ranging from local to global. Not surprisingly, transitions to
new energy sources (be they gradual diffusions of new fuels or new modes of elec-
tricity generation) have attracted a great deal of attention and I will quantify the
key shifts—from wood and charcoal to coal and then to hydrocarbons, followed
by transitions to a higher share of primary energies consumed in a secondary form
as electricity—from the global perspective as well as by focusing on some notable
national trajectories.
Perhaps even more attention has been paid by the historians of technical advances
to the diffusion of new fuel and electricity converters ranging from better stoves and
lights to more efficient furnaces and boilers, with particular interest in the evolution
and diffusion of new engines and turbines and new electricity-powered motors and
appliances. Technical innovation, emergence of new mass energy markets, and a
steadily rising demand for more efficient, more affordable, and more flexibly
18 ENERGY TRANSITIONS

delivered energy services were both the driving factors behind these changes and,
thanks to numerous reinforcing feedbacks, also their beneficiaries.
In addition to tracing the transitions to new energy sources and new energy
converters it is also revealing to look at the changing uses of individual fuels (most
notable, coal losing all of its transportation markets but becoming the leading
fuel for electricity generation, and the principal use of refined oil products shift-
ing from illuminants and lubricants to transportation fuels) and at changing
patterns of sectoral consumption. The latter shifts are actually an excellent means
of tracing a nation’s trajectory of modernization and its rise to affluence: Diversi-
fication of final commercial energy uses proceeds from the initial pattern domi-
nated by industrial consumption to a combination characterized by the absence
of any dominant use, where each of the four key sectors (households, industries,
commerce, and transportation) claims a major share of the final demand.
As this book’s principal aim is a comprehensive appraisal of energy transitions—
on levels ranging from global to national and looking at trends ranging from aggre-
gate provision of primary energies to specific supplies of individual fuels and
progress of important conversion techniques—I will use this introductory section
only in order to make several general observations by resorting to analogies. When
appropriately understood—that is, in an illuminating, suggestive manner and not
as rigid templates—analogies are a useful tool to emphasize important features of
a complex process. I think that two of them, of widely differing provenience, are
particularly relevant to the understanding of energy transitions.
The first one draws on Tolstoy’s famous observation (in Anna Karenina) regard-
ing families: ‘‘Happy families are all alike; every unhappy family is unhappy in its
own way.’’ Analogically, notable similarities can be seen when looking at all rapid
and apparently easily accomplished energy transitions—while the reasons for pro-
longed, complicated, and delayed transitions are usually very specific, bound with
unique environmental, social, economic, and technical circumstances. Rapidity of
energy transitions is most evident when looking at small countries with compact
territories that have either relatively few people or a high density of population.
No matter if they are affluent economies or still essentially premodern societies with
very low per capita economic product, once they discover a new rich source of pri-
mary energy they can develop it rapidly and end up with completely transformed
energy foundations in less than a single generation.
The Netherlands—thanks to the discovery of a giant Groningen natural gas
field in the municipality of Slochteren in the northern part of the country on
July 22, 1959 (Whaley, 2009)—is perhaps the most apposite example of an afflu-
ent economy following this path (for more detail, see chapter 3), while Kuwait’s
rapid development of its giant oilfields is an iconic example in the second category.
Kuwaiti oil development began only in 1934 with the concession given to the
Kuwait Oil Company, a joint undertaking of the APOC (Anglo-Persian Oil
Company, later BP) and Gulf Oil. The concessionary agreement was signed after
the APOC was assured by an expert it hired to evaluate the country’s oil prospects
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 19

that ‘‘the absence of geological structure suitable for the accumulation of oil in
commercial quantity shows that there is no justification for drilling anywhere in
Kuwait’’ (Howard, 2008, p. 152).
At that time that small country (with an area less than half that of the
Netherlands) was an impoverished British protectorate with fewer than
100,000 people, a single town, and mostly empty interior with a small number
of desert nomads; export of pearls, harvested by diving, was declining and tradi-
tional maritime trading (horses, spices, coffee) was the only notable economic
activity. The concession was signed on December 23, 1934, and the supergiant
al-Burqān oilfield (a Cretaceous sandstone trapped above a massive swell of
about 750 km2 of salt) was discovered on February 23, 1938. The field was later
proved to be the world’s second largest accumulation of oil, following the Saudi
al-Ghawār (Stegner, 2007; Howard, 2008). In 1946, when it began its oil
exports, Kuwait produced about 800,000 t of oil, a year later 2.25 Mt, annual
output surpassed 50 Mt by 1955 and 100 Mt by 1965 when the country was,
ahead of Saudi Arabia, the world’s fourth largest producer of oil (behind the
United States, USSR, and Venezuela). In energy terms Kuwait thus moved from
a premodern society dependent on imports of wood, charcoal, and kerosene to
an oil superpower in a single generation.
In contrast, large economies, particularly those with relatively high per capita
demand and with extensive infrastructures serving an established fuel, cannot
accomplish the substitutions so rapidly. Comparing the Dutch and the British
experience is particularly revealing in this respect, as both of these countries
benefited from major natural gas discoveries. The first discoveries of natural gas
in the British sector of the North Sea were made by BP in 1965 but despite an
aggressive development of those rich and relatively near-shore deposits, Britain
could not accomplish even in 30 years what the Netherlands did in a decade: Its
share of natural gas stood at a bit less than 5% of the primary energy supply in
1970 and it peaked only 30 years later at about 39%.
Principal reasons for the difference include a much higher total of the absolute
supply needed to provide an identical share of the primary energy (by 1970 the
UK’s primary energy supply was nearly 220 Mtoe/year compared to 60 Mtoe/
year in the Netherlands), UK’s traditionally high dependence on coal-fired elec-
tricity generation, the country’s pioneering role in nuclear generation (it would
have been very costly to shut down those stations and replace them with gas-
fired plants), a higher cost and longer lead times to develop offshore resources
rather than hydrocarbons fields on land (particularly in such an inhospitable
environment as the North Sea), and also the much larger size of the country
(about 244,000 km2) necessitating longer trunk and distribution lines.
And the Japanese progress shows that when the gas has to be imported from
overseas then the pace of substitution must be even slower—regardless of the fact
that the country was one of the pioneers of LNG imports (starting in 1969 with
Polar Alaska and Arctic Tokyo, each with capacity of 71,500 m3 to carry gas
20 ENERGY TRANSITIONS

from Alaska) and that when it commenced its LNG imports it was not only one
of the world’s leading economies but one with an enormous experience in ship-
building. At the same time, a slow pace of substitution comes as no surprise
given the size of Japan’s economy and its nearly total dependence on fossil fuel
imports: This means that despite its relatively high efficiency the country now
requires annually more than 500 Mtoe (nearly 22 EJ) of primary energy. Given
these circumstances Japan’s LNG progress could be actually seen as rather
impressive, as the country had increased the share of natural gas in its energy
supply from 5% in 1979 to about 16% by 2008.
The second analogy illuminating the process of energy transitions is their com-
parison with aircraft accidents. Careful studies of those events show that they are
nearly always due to a number of factors and that the final outcome is a result of
a specific sequence of errors (be they actions or inactions) taken by crews in
response to a sudden change, be it a faulty indicator light, erroneous instrument
reading, or (an increasingly rare occurrence with modern gas turbines) mechanical
failure of one or more of the airplane’s engines. And so it is with energy transitions:
They are never brought about by a single factor, and in the second chapter I will
show that this was the case even with perhaps the most commonly cited claim, por-
traying English wood shortages as the decisive factor forcing the country’s early
transition to coal.
And, as with the aircraft accidents, a careful investigation of energy transitions
always reveals that their progress requires a specific sequence of scientific advan-
ces, technical innovations, organizational actions, and economic and political
and strategic circumstances. Missing a single component in such a sequence, or
delaying its introduction or effects because of some unforeseen events, results
in very different outcomes and in lengthier transition periods. Once again, an
excellent example illustrating this necessity of a specific sequence, and of assorted
events delaying its progress, is provided by the recent emergence of LNG as a
globally available fuel traded competitively on an intercontinental basis.
A long road toward this accomplishment had to include the invention and
commercialization of gas liquefaction, establishment of LNG supply chain
(liquefaction, tanker-borne transport, regasification), increase of typical liquefac-
tion and LNG tankers’ capacities in order to lower unit costs of the delivered gas,
a greater number of importing countries in order to justify the construction and
expansion of larger terminals, and extensive trunk and distribution pipelines in
those importing countries that had previously no natural gas supply. And the
process needed to create this new global industry was delayed by factors ranging
from predictable (high capital costs of the first generation of LNG systems) to
unforeseeable (OPEC-driven energy price increases, the Shah’s fall and Khomeini’s
assumption of power in Iran, hydrocarbon price deregulation in the United States,
concerns about early peak of oil extraction).
The road toward global LNG industry began in 1852 when the pioneering
work done by James Prescott Joule and William Thomson (Lord Kelvin) on
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 21

liquefaction of gases demonstrated that as a highly compressed air flows through

a porous plug (a nozzle) it expands to the pressure of the ambient air and cools
slightly (Almqvist, 2003). Repetition of this sequence creates a cooling cascade,
the temperature of the gas expanded at the nozzle gradually declines and it even-
tually liquefies. Practical designs for commercial liquefaction of oxygen and
nitrogen followed during the last three decades of the nineteenth century, with
the most important contribution made by Carl von Linde (1842–1934), whose
patented process (in 1895) combined the Thomson–Joule effect with what
Linde termed countercurrent cooling, with compressed air expanded through a
nozzle at the bottom of an insulated chamber used to pre-cool the incoming
compressed air in a countercurrent cooler (Linde, 1916).
Because the United States was the only notable user of natural gas before
World War II there was no commercial need for LNG: That is why Godfrey
Cabot’s patented handling and transporting liquid natural gas (Cabot, 1915)
did not have any practical consequences. The first small LNG storage was built
in West Virginia in 1939 and a larger one in Cleveland in 1941 to provide fuel
for the periods of peak demand; in 1944 one of its tanks failed and the ignited
vaporized gas killed 128 people in the plant’s neighborhood. This accident used
to be cited by those who wanted to portray LNG industry as very risky—but the
investigation report concluded that the accident was caused by a poor tank
design and that properly done the gas liquefaction and storage are not exception-
ally dangerous (USBM, 1946).
Post–WWII surfeit of cheap crude oil and rapid expansion of North American
gas extraction had postponed the beginning of the LNG era for another generation:
The first demonstration shipment of LNG (from Lake Charles, LA to Canvey
Island on the Thames) took place in 1959 with a tanker of just 5,000 m3 (Methane
Pioneer, a converted WWII Liberty class freighter). The first methane liquefaction
plant was completed in Arzew, Algeria in 1964 and LNG exports to the United
Kingdom began in the same year with two specifically designed tankers (Methane
Princess and Methane Progress) of 27,400 m3 each (Corkhill, 1975). They were
followed by the Japanese imports from Alaska in 1969 and the French imports from
Libya in 1970. But then the Groningen and the North Sea gas made the LNG
imports uneconomical and when the Arzew–Canvey contract expired in 1979 it
was not renewed.
Similarly, during the 1970s the United States built four regasification terminals
for the import of Algerian gas (the first one in Everett, MA, in 1971) only to reduce
their operation or to shut two of them down as the availability of domestic natural
gas increased with the post-1993 wellhead price deregulation. This left Japan (with
no domestic gas resources) as the world’s leading importer of LNG, adding new
long-term contracts for the gas from Abu Dhabi and Indonesia (in 1977), Malaysia
(1983), and Australia (1989): By 1984 Japanese imports accounted for 75% of all
LNG trade; by 1999 they were still 66% of the total. And while Taiwan (in
1990) and South Korea (in 1991) joined Japan as the other major Asian importers,
22 ENERGY TRANSITIONS

the LNG trade remained confined by uncompetitive long-term contracts served by

dedicated plants and ships along inflexible routes.
These realities were not conducive to any bold technical advances. For more
than a generation, between the mid-1960s and the late 1990s, typical capacities
of LNG trains (liquefaction units) remained at just 1–2 Mt/year, while the
aggregate outputs of entire plants increased only gradually, from the pioneer
Arzew’s rate of 0.45 Mt/year in 1964 to 1 Mt/year in 1970, 1.5 Mt/year in
1980, 2.2 Mt/year in 1990, and 3.5 Mt/year in 2000. Some of these large-
scale liquefiers have used the classic cascade cycle but most of them have relied
on a mixed refrigerant cycle (using such gases as butane, propane, ethane, and
nitrogen) devised by A. P. Kleemenko in 1960. And although the largest ship
capacities increased fairly rapidly during the first decade of LNG trade—from
27,400 m3 for the two pioneering ships in 1964 to 71,500 m3 in 1969 and
126,227 m3 in 1975—three decades later the dominant sizes (largely due to
the Japanese restrictions on the maximum tonnage of LNG tankers) were still
between 125,000 and 130,000 m3.
Given a limited number of exporting countries (1 in 1964, 6 by 1980, 12 by
2000) and LNG tankers (fewer than 60 vessels until 1984, 100 by 1997), this
slow capacity growth meant that the total LNG trade surpassed 50 Mt/year only
by 1991 and that only in 1999 did it carry more than 5% of all exported gas
(Castle, 2007). The industry began to change rapidly at the century’s turn. Qatar
joined the ranks of LNG exporters in 1997, in 1999 a new LNG plant in
Trinidad and Tobago led to the reactivation of the two closed U.S. regasification
plants (Elba Island in 2001, Cove Point in 2003), Nigeria and Oman began ship-
ping LNG in 2000, followed by Egypt in 2005, Equatorial Guinea in 2007, and
Russia (from Sakhalin) in 2009.
Increasing train size (maxima of 5 Mt/year by 2005, more than 8 Mt/year by
2008) and decreasing costs of train and tanker construction resulted in rapid
capacity increases and bold plans for further expansion. Total export capacity rose
from 100 Mt/year in 2000 to about 220 Mt/year by 2009. For three decades the
standard LNG tanker design used large aluminum spheres (Kvaerner-Moss shells
introduced in 1971) covered with insulation inside steel tanks and bolted to the
vessel’s hull. This design wastes storage space and steel spheres increase the ship’s
draft, making voluminous vessels impractical. In contrast, membrane design has
insulated tanks of thin stainless steel shaped to fit the inner hull. As a result, average
size of ships ordered in 2007 was about 180,000 m3 and Qatargas has taken
delivery of the first tankers belonging to new Q-Flex (210,000 m3) and Q-Max
(266,000 m3) classes of ships. The company will eventually have 45 of these large
vessels (Qatargas, 2009).
By 2008 there were 250 LNG tankers with the total capacity of 183 Mt/year and
the global LNG trade carried about 25% of all internationally traded natural gas
(BP, 2009). LNG was imported by 17 countries on four continents, and before
the economic downturn of 2008 plans envisaged more than 300 LNG vessels by
 ENERGY SYSTEMS: THEIR BASIC PROPERTIES 23

2010 with the total capacity of about 250 Mt/year as the global LNG trade has
moved toward a competitive market. LNG trade has been finally elevated from a
marginal endeavor to an important component of global energy supply, and this
has become true in terms of total exports (approaching 30% of all natural gas sold
abroad) and number of countries involved (now more than 30 exporters and
importers) as well as the flexibility of transactions (with a true market emerging).
This brief recounting of LNG history is an excellent illustration of the
decades-long spans that are often required to convert theoretical concepts into
technical possibilities and then to adapt these technical advances and diffuse
them to create new energy industries (Figure 1.4). Theoretical foundations of
the liquefaction of gases were laid down more than a century before the first
commercial application; the key patent that turned the idea of liquefaction into
a commonly used industrial process was granted in 1895, but at that time natu-
ral gas was a marginal fuel even in the United States (in 1900 it provided about
3.5% of the country’s fossil fuel energy), and in global terms it had remained one
until the 1960s, when its cleanliness and flexibility began to justify high price of
its shipborne imports. Even then the first long-term contracts delivered gas only
to affluent countries that could afford the price and that used most of the gas for
shore-based electricity generation (Japan) or had preexisting trunk and distribu-
tion pipelines carrying domestically produced gas in place (United Kingdom,
France, United States) that could be used to sell the imported gas to households
and enterprises.

Figure 1.4 History of LNG shipments illustrates often very long time spans
required for the maturation and diffusion of innovations in energy extraction,
transport, and conversion.
24 ENERGY TRANSITIONS

Industry’s subsequent growth was affected by a combination of events that

could not have been predicted during the 1960s: by the two oil price crises of
the 1970s, by the collapse of the Iranian monarchy in 1979, by the deregulation
of U.S. natural gas prices (and the consequent boost of the domestic extraction),
and by the collapse of the world oil price in 1985. As a result, many plans were
postponed or cancelled. In 1975 it was expected that by 1981 Nigeria would
begin its LNG exports to Europe, and Iran to Europe, the United States, and
Japan (Faridany, 1975), but Nigerian exports began only nearly two decades later
(in 1999) and Iranian shipments have yet to begin. The industry that began in
1964 moved only about 2% of all traded gas by 1980 and 5% of all natural
gas exports only in 1999. At that time it was clearly an important earner for a
few major exporters (Algeria, Indonesia, Brunei) and a significant source of fuel
for the three leading importers (Japan, South Korea, Taiwan), but it still could
not qualify as a key ingredient of the global primary energy supply.
If we take the years between 1999 (when worldwide LNG exports surpassed 5%
of all natural gas sales) and 2007 (when the number of countries exporting and
importing LNG surpassed 30, or more than 15% of all nations) as the onset of
LNG’s global importance, then it had taken about four decades to reach that point
from the time of the first commercial shipment (1964), about five decades from
the time that natural gas began to provide more than 10% of all fossil energies
(during the early 1950s), more than a century since we acquired the technical
means to liquefy large volumes of gases (by the mid-1890s)—and about 150 years
since the discovery of the principle of gas liquefaction.
By 2007 it appeared that nothing could stop an emergence of a very substan-
tial global LNG market. But then a sudden supply overhang that was created in
2008—and that was due to the combination of rapid capacity increases, lower
demand caused by the global financial crisis, and the retreat of U.S. imports
due to increased domestic output of unconventional gas—has, once again,
slowed down global LNG prospects, and it may take years before the future
course will become clear. In any case, the history of LNG remains a perfect
example of the complexities and vagaries inherent in major energy transitions.
 Chapter 2

G LOBAL T RANSITIONS :
U NIVERSAL PATTERNS

T
he most obvious reality that emerges from the study of energy transitions
done from the global perspective and across the entire historical time span
is a highly skewed division of their progress: Stasis, stagnation, marginal
adjustments, and slowly proceeding innovations marked the entire preindustrial
era—while the process of industrialization and the evolution of postindustrial soci-
eties have been marked (indeed formed) by rapid, often truly precipitous diffusion
of new inventions and widespread adoption of technical and organizational inno-
vations. As a result, nearly five millennia of preindustrial history were almost
completely dominated by reliance on inefficiently burned biomass fuels as the
source of heat for households, metallurgy, and artisanal manufactures, and by
exertions of human and animal muscles to provide nearly all requirements for
mechanical energy (sails being the only early exception).
This situation did not change fundamentally even during the early modern
era when some Western European societies began a small-scale extraction of coal
(or peat) and when they adopted increasingly more efficient and more powerful
water wheels and windmills. The two fundamental transitions, from biomass to
fossil fuels and from animate to inanimate prime movers, have taken place only
during the last few centuries (roughly three in the case of some European societies)
or just a few recent decades (six in China’s, four in India’s case), and the emergence
of electricity as the energy form of the highest quality began only during the 1880s.
Inevitably, these transitions began on small local scales, evolved into nationwide
developments, and eventually became truly global phenomena. Only the earliest
innovators were able to maintain their advantage for a period of time, while the
more recent advances have been diffusing with only a minimum lag (a phenome-
non perhaps best illustrated by China’s rapid post-1980 modernization).
I will trace all of these developments by following first the grand fuel sequence
of millennia-long dependence on biomass energies that was replaced by now
virtually universal dependence on fossil fuels. In the next section I will emphasize
importance of electricity in modern societies and review the development
of thermal, hydro, and nuclear generation. Then I will offer a brief history of a
critical transition from animate to mechanical prime movers, and the chapter
26 ENERGY TRANSITIONS

will conclude with the best possible quantitative appraisal of these trends on the
global scale—and with inevitable caveats regarding the quality of various
historical data used for these analyses.

GRAND FUEL SEQUENCE: FROM BIOMASS TO COAL AND HYDROCARBONS

All preindustrial societies had a rather simple and persistent pattern of primary
fuel use as they derived all of their limited heat requirements from burning biomass
fuels. Fuelwood (firewood) was the dominant source of primary energy, but woody
phytomass would be a better term: the earliest users did not have any requisite saws
and axes to cut and split tree trunks, and those tools remained beyond the reach
of the poorest peasants even during the early modern era. Any woody phytomass
was used, including branches fallen to the ground or broken off small trees, twigs,
and small shrubs. In large parts of the sub-Saharan Africa and in many regions of
Asia and Latin America this woody phytomass, collected mostly by women and
children, continues to be the only accessible and affordable form of fuel for cooking
and water and house heating for the poorest rural families.
Moreover, in some environments large shares of all woody matter were always
gathered by families outside forests from small tree clumps and bushes, from the
litter fall under plantation tree crops (rubber, coconut) or from roadside, back-
yard, or living fence trees and shrubs. This reliance on non-forest phytomass also
continues today in many tropical and subtropical countries: Rural surveys con-
ducted during the late 1990s in Bangladesh, Pakistan, and Sri Lanka found that
this non-forest fuelwood accounted for more than 80% of all wood by house-
holds (RWEDP, 1997). And in less hospitable, arid or deforested, environments,
children and women collected any available non-woody cellulosic phytomass,
fallen leaves (commonly raked in North China’s groves, leaving the ground
barren), dry grasses, and plant roots. For hundreds of millions of people the
grand energy transition traced in this chapter is yet to unfold: They continue to
live in the wooden era, perpetuating the fuel usage that began in prehistory.
Another usage that has been around for millennia is the burning of crop
residues (mostly cereal and leguminous straws, but also corn or cotton stalks
and even some plant roots) and sundry food-processing wastes (ranging from
almond shells to date kernels) in many desert, deforested, or heavily cultivated
regions. And on the lowest rung of the reliance on biomass fuels was (and is)
dry dung, gathered by those with no access to other fuels (be it the westward-
moving settlers of the United States during the nineteenth century collecting
buffalo dung or the poorest segments of rural population in today’s India) or
whose environment (grasslands or high mountain regions) provides no suitable
phytomass to collect (Tibetan and Andean plateaus and subtropical deserts of
the Old World where, respectively, yak, llama, and camel dung can be collected).
But besides constancy there have been also important changes in wood use
and production. Charcoal, produced from a variety of wood species (hardwoods
 GLOBAL TRANSITIONS: UNIVERSAL PATTERNS 27

as well as conifers; see Figure 1.1), was widely used in antiquity and it eventually
became a preferred source of heat for those who could afford its higher price and
valued its smokeless combustion. Remarkably, the British House of Commons
was heated by charcoal until 1791, long after (as I will soon explain) everybody
in cities, including the royal family, switched to coal. Charcoal was also the best
choice for many small manufactures, particularly for metalworking. But this
cleaner fuel was produced so inefficiently that in mass terms up to 15 units of
wood were need for a unit of charcoal, and even with a lower, typical preindus-
trial mean of 5:1 this conversion entailed about 60% loss of initially charged
energy content (air-dry wood has about 15 GJ/t and density of 0.65 t/m3 while
charcoal’s energy density is 29.7 GJ/t).
And in households the universal reliance on phytomass combustion had
evolved from using inefficient open hearths to burning wood in stoves with
proper chimneys. These stoves were a surprisingly late innovation, beginning
with tiled Kachelofen (common in Central Europe by the sixteenth century),
and with various iron stove designs introduced during the eighteenth century
(including the famous but misleadingly named Franklin stove in 1742: it was
actually just an iron-lined fireplace). At the same time industries introduced
more efficient, larger furnaces and steam-generating boilers, and iron makers
began to convert wood to charcoal on a massive scale needed to feed larger blast
furnaces. As a result, growing cities, expanding manufactures, and increasing
iron production led to the demise of surrounding forests, and affordable avail-
ability of nearby wood or charcoal supplies (as long-distance land transport
using pack or draft animals was usually prohibitively expensive) became a key
factor limiting the size of preindustrial cities and the level of iron output.
Unfortunately, we have no reliable records of ancient or medieval household
biomass fuel consumption, and even for the early modern era and the eighteenth
and nineteenth centuries there is only some highly fragmentary information
from a few countries. The United States, with some fuelwood data going back to
the seventeenth century, is perhaps the most important exception. And, of course,
in some regions household use was easily equalled or surpassed by industrial
demand of iron- or glassmaking or for salt production. German medieval glass-
making was particularly wood-intensive, with as much as 2.4 t of wood (97% of
it burned to obtain potassium rather than energy) used per kg of glass, an equiva-
lent of 90 MJ/kg (Sieferle, 2001). Salt works using large heated pans to evaporate
brines produced (depending on salt concentration) 15–100 kg of salt/m3 wood,
demanding as much as 500–600 MJ/kg.
And reliable information on English iron smelting indicates that up to 20 kg
(almost 600 MJ) of charcoal were used to produce 1 kg of hot metal during the
Middle Ages and about 8 kg (240 MJ) of charcoal were needed for 1 kg of iron
by the end of the eighteenth century (Smil, 1994). This demand led to extensive
deforestation, a transformation that undercut not only the viability of charcoal-
using establishments but also the very existence of nearby villages and cities that
28 ENERGY TRANSITIONS

needed wood as timber for their houses and as raw material for making nearly
all of their machines, devices, and utensils of daily life. At the same time, a more
predictable supply of wood was secured in some regions by deliberate planting of
trees in backyards, on roadsides, on otherwise infertile slope land, or in fuelwood
groves to supply nearby farms or villages.
Classical cases of energy transition from biofuels to fossil fuels (both in the
sense of being the most consequential for the subsequent economic advancement
and the best studied from historical and technical perspectives) involve gradual
shifts from total reliance on fuelwood and charcoal to increasing uses of coal,
in both domestic and industrial settings. But that sequence, best known from devel-
opments in England, Germany, or the United States, has not been a universal phe-
nomenon. As I will describe in the next chapter when detailing several prominent
energy transitions on the national level, there was an interesting early exception
when the Golden Age of the Dutch Republic (1608–1672) was energized by a
unique shift from biofuels to peat, the youngest of all fossil fuels, aided consider-
ably by a widespread use of wind power (de Zeeuw, 1978; Unger, 1984).
And during the twentieth century many Asian and African countries with
abundant hydrocarbon resources but with no domestic coal deposits moved
from the biofuel era directly to the use of refined oil products and natural gas,
with some desert countries moving from very low per capita consumption rates
of biomass fuels to some of the world’s highest per capita rates of hydrocarbon
use in just two generations. But, without exception, all of the world’s major
economies—the United States, the United Kingdom, Germany, France, Russia,
Japan, China, and India—had followed the classical sequence from biofuels to
coal, and that is why this transition should receive the closest attention.
During the earliest stage of small-scale local coal extraction there was no need
to discover the fuel and to develop elaborate mines: The first seams to be tapped
were those outcropping to the surface or those under only a shallow overburden
and accessible by open pits or short shafts. In some places and at different times—
ranging from the Roman Britain of the first two centuries CE to Arizona Hopis
of the thirteenth century—coal was used locally for heating, and its first metallur-
gical uses were during the Han dynasty, where the fuel was packed around iron
ore-filled crucibles (Needham, 1964). The oldest European extraction is docu-
mented in Belgium in 1113 and London received its first coal deliveries in 1228
but, as Nef (1932) noted, until the sixteenth century the fuel was regularly burned
only by poor households who could not afford to buy wood and lived close to
coal outcrops.
The genesis of the growing British reliance on coal offers some valuable
generic lessons. Thanks to Nef ’s (1932) influential work a national wood crisis
has been commonly seen as the key reason for the expansion of coal mining
between 1550 and 1680—but other historians could not support this claim,
pointing to the persistence of large wooded areas in the country, seeing such
shortages as largely local and criticizing unwarranted generalization based on the
 GLOBAL TRANSITIONS: UNIVERSAL PATTERNS 29

worst-case urban situations (Coleman, 1977). This was undoubtedly true, but not
entirely relevant, as transportation constraints would not allow the emergence of a
national fuelwood market, and local and regional wood scarcities were real.
At the same time, the best available reconstruction of energy prices in the
southeast of England shows that (in real terms) fuelwood prices were actually
fairly stable between 1550 and 1650 and that it was a steady decline in coal
prices that primed the expanding fuel extraction (Fouquet, 2008). By 1600 coal
prices were about half that of wood prices when compared per unit of gross
energy, and by the time wood prices began to rise during the latter half of the
seventeenth century (driven not only by growing industries but also by expanded
shipbuilding), coal was the dominant source of energy for nearly all British
industries (with a major exception of iron smelting dependent on charcoal) as
well as for household heating (Hatcher, 1993). Coal production was greatly
boosted with the invention of steam engine and with the spreading replacement
of charcoal by coke in iron smelting.
Again, neither of these innovations was immediately successful. Thomas
Newcomen’s highly inefficient steam engine was introduced by 1712, but even
its improved version had a limited impact (it was widely used only by coal
mines), and it was only the machine’s radical redesign patented in 1769 by James
Watt that gained wider commercial acceptance during the last three decades of
the eighteenth century (Figure 2.1; for more details, see the third section of this
chapter). Expanded coal extraction created its own positive production feedbacks,
as deeper shafts sank far below local water tables and required frequent or constant
pumping, and as deeper mines also needed more energy for ventilation and for
hoisting of the fuel.
Britain was circumventing its early eighteenth-century local charcoal short-
ages by iron imports from Sweden and by lowering the rate of charcoal use in
smelting. Coke was successfully used in iron smelting by Abraham Darby in
1709, but it was too costly (because of its inefficient production) and its wide-
spread acceptance came only after 1750 (Harris, 1988). Coke, much less friable
than charcoal, made it possible to build taller, more voluminous blast furnaces
and due to its higher temperature of combustion it also produced better iron.
By 1800 Britain was extracting about 9 Mt of coal a year while U.S. production
was only about 100,000 t, most of it coming from Pennsylvania (both bituminous
coal and anthracite) with smaller contributions from Virginia, West Virginia, and
eastern Kentucky (Milici, 2003).
Before 1800, major coal-mining regions were also emerging in continental
Europe in northern France, around Liège in Belgium, in Bohemia, and in Silesia,
Saxony, and the Ruhr region of Germany. German transition to coal is a notable
example of the fact that the shift from wood to coal did not have to be primed
by increasing shortages of fuelwood (Sieferle, 2001). Large parts of the country
were always well forested, and even by the end of the eighteenth century it was
possible to secure enough fuelwood at acceptable prices. While the German wood
30 ENERGY TRANSITIONS

Figure 2.1 Cross-section of Watt’s engine. Reproduced from Farey (1827).

crisis was above all a timber crisis rather than a matter of energy shortages, it is also
clear that the country’s wood supply could not have supported economic growth
associated with industrialization. Wood shortages did precipitate the transition to
coal, but only coal could sustain the country’s famously rapid post-1870
industrialization.
Timber shortage could be best addressed by improved forestry practices and by
reserving more wood harvests for timber. These measures led to state-promoted,
even state-subsidized switch to coal, first in some state-owned industries and
later in households. And it was only because of this switch that, as Sieferle (2001,
p. 180) put it, ‘‘the limits of the agrarian solar energy system were burst in the
 GLOBAL TRANSITIONS: UNIVERSAL PATTERNS 31

transition to fossil energy system. Precisely this constituted the solution that
formed the energy basis for the industrial transformation and marked it as a unique
epochal discontinuity.’’ This conclusion has, of course, a universal validity.
Because bituminous coals of acceptable quality, and even more so the poorer
lignites, are widely distributed, it was not long before scores of countries—including
Sweden, Greece, and Spain in Europe, China, India, Japan, and Turkey in Asia,
and Mexico and Peru in Latin America—began producing the fuel on larger
scale, but Britain maintained its coal-mining lead almost until the century’s
end. In 1800 the country produced more than 80% of the global output and
by 1870 its share was still 50%. But by 1899 it was surpassed by the U.S. extrac-
tion, and Germany’s production was not far behind the British extraction (but
more than a quarter of that country’s coal was low-energy lignite). France, the
Austro-Hungarian Empire, Belgium, Russia, and China were the other major
producers at the beginning of the twentieth century when coal became firmly
established as the single most important primary fuel, having surpassed fuelwood
and charcoal sometime during the 1890s.
Three grand trends marked the global coal production of the twentieth century:
continuous decline of its relative importance, continuous growth of its absolute
contribution to the worldwide total of primary energies, and the transformation
from a highly labor-intensive to a highly mechanized industry. In 1900 global
extraction of hydrocarbons was only marginally important; by 2000 they far
surpassed coal’s contribution. In 1900 the worldwide extraction of bituminous
coals and lignites added up to about 800 Mt; a century later it was about 4.5 Gt,
a roughly 5.6-fold increase in mass terms and (because of the declining energy den-
sity of extracted coal) almost exactly four-fold increase in energy terms. Much as
fuelwood energized the early stages of eighteenth- and nineteenth-century industri-
alization, coal energized the building of modern industries and infrastructures in all
European countries as well as in North America, Australia, and Japan during the
first half of the twentieth century, and the fuel continues to play a critical role not
only in China and India but also in the United States and Europe because of its
contribution to electricity generation.
After 1950 high levels of mechanized underground mining were the norm
(except in China’s small mines), easily doubling or tripling typical pre–WWI labor
productivities. Even greater productivities (and much higher safety) have been
achieved by extracting an increasing share of the fuel by surface (open-cast)
mining. And the worldwide coal industry of 2000 differed from its predecessor
of 1900 also because of its profoundly changed spatial distribution. Perhaps the
most notable national trends include the near demise of British extraction, ascent
of the Soviet and then the retreat of the Russian output, continuing high levels of
U.S. production, huge increase of China’s extraction, and the emergence of new
large coal exporters, Australia and Indonesia.
The British output peaked in 1913 (287 Mt from more than 3,000 mines), and
it was still above 200 Mt during the 1950s (Hicks & Allen, 1999; Figure 2.2).
32 ENERGY TRANSITIONS

Figure 2.2 Coal production, 1810–2010. Plotted from data in UNO (1956 and
1976), Etemad et al. (1991), and BP (2009).

But the subsequent oil imports and then the domestic production of crude oil
and natural gas from the North Sea reduced it to less than 100 Mt during the
1980s and by 2000 the United Kingdom was extracting less than 20 Mt/year and
it became an importer of coal, joining such larger buyers as Japan, South Korea,
Taiwan, and India. Most of those imports came from the new coal-exporting
powers, Australia, Indonesia, South Africa, and Colombia and from Russia and
the United States.
The United State’s 1900 extraction doubled by 1913 and it reached the pre–
WWII peak of nearly 600 Mt in 1923; that total was surpassed again only
during 1944 in the midst of wartime effort, and after another post-war pullback
the industry became the prime energizer of rapidly expanding electricity genera-
tion, a period that lasted until the early 1970s. By that time the USSR was the
world’s second-largest producer (having surpassed the United Kingdom in terms
of total energy content already by 1950), with China rising fast. But the United
States yielded to China only in 1991, and by 2005 the gap between the two largest
coal producers was more than twofold in mass terms and about 1.8-fold in terms
of coal’s total energy content (Figure 2.2).
As in coal’s case, it is impossible to date the earliest instance of human famili-
arity with crude oil because in some locales the fuel was known for millennia.
As any liquid under pressure, crude oil has the propensity to seep to the surface
 GLOBAL TRANSITIONS: UNIVERSAL PATTERNS 33

along fracture zones and to form black lakes of tar or bitumen pools and even,
when under high pressure and mixed with natural gas, to create periodically reap-
pearing burning pillars. And as in coal’s case, the fuel was used in small amounts,
often for non-energy applications, since antiquity. In ancient Mesopotamia
asphalts and bitumens were used in floor and wall mosaics and as protective coat-
ings and lighter oils were burned in fire pans for illumination; such uses could be
subsequently copied by the Greeks and the Romans and later they were perpetu-
ated by the inhabitants of the medieval Middle East (Forbes, 1964).
Oil from natural seeps in western Pennsylvania was collected during the late
eighteenth century and bottled to be sold as a medicinal ‘‘Seneca oil,’’ and crude
oil, although not in its liquid form, was also known in preindustrial Europe in
the form of oil sands in Merkwiller-Pechelbronn in Alsace where the first shallow
(9.75-m) pit was dug in 1745, the first refinery was built in 1857, water injection
began in 1879, and small-scale production continued until 1970 (Walther,
2007). There was only one locality in the preindustrial world where active steps
were taken to collect crude oil, the Absheron peninsula of the Baku region on the
Caspian Sea in Azerbaijan. Baku’s oil pools and wells were described by medieval
Arabic travelers and historians, and in 1593 an inscription was affixed near a
35-m deep well that was dug manually in Balakhani (Mir-Babaev, 2004).
By the time Czarist Russia took over Baku (in 1806) the Absheron region had
many shallow wells from which lighter oil was collected in order to produce
kerosene (by thermal distillation) and use it for local lighting as well as for export
by camels (carried in skins) and in wooden barrels on small ships. In 1837 Russians
built the first commercial oil-distilling factory in Balakhani and nine years later
they sank the world’s first (21-m deep) exploratory oil well in Bibi-Heybat and thus
opened up what was later classified as the world’s first giant oilfield (i.e., one having
at least 500 million barrels of recoverable crude oil). Baku was thus the place where
the modern oil era began, and 1846 was its beginning.
North American developments followed soon afterwards, spurred by the
search for an alternative source of lighting to replace whale oil (Brantly, 1971).
In 1858 Charles Tripp and James Miller Williams financed America’s first (man-
ually dug) oil well near Black Creek (Lambton County in southwestern Ontario)
and a year later, amidst the world’s first oil boom, the hamlet was renamed Oil
Springs. And 1859 was also the year of the first commercial U.S. oil discovery
as Edwin Drake (employed by George Bissell, who started the Pennsylvania
Rock Oil Company) supervised the drilling of a shallow well at an oil seep site at
Oil Creek near Titusville, Pennsylvania. The well (whose drilling used an ancient
Chinese percussion method, but powered by steam engine) struck oil at the depth
of 21 m on August 27, 1859, the date the Americans use as the beginning of the
modern oil era.
During the early 1860s crude oil thus began to figure on the balance sheets of
primary energy consumption in Russia, Canada, and the United States. Canada
soon fell out of the new oil league: Another Oil Springs boom began in 1862 with
34 ENERGY TRANSITIONS

the world’s first gusher, and in 1865 oil was discovered also in nearby Petrolea, but
the pressure in Ontario’s small reservoirs soon declined, steam pumps were used to
produce diminishing volumes, and by the 1890s a shrinking industry could not
compete with much cheaper American oil. Canada’s second oil era began only
with the post–WWII discoveries in Alberta.
Russian oil extraction also progressed swiftly thanks to substantial foreign
investment (above all by Ludwig and Robert Nobel, who launched Nobel Brothers
Petroleum Company in 1875, and the Rothschild brothers who established the
Caspian and Black Sea Oil Industry and Trade Society in 1883) and to new major
discoveries at the giant Bibi-Heybat field in 1878 (Figure 2.3). The only other
notable pre-1900 crude oil developments took place in Romania, Indonesia, and
Burma. In Romania pools and shallow wells were known for centuries and the first
commercial refinery was opened in Ploieşti (60 km north of Bucharest) in 1857,
and the country’s only giant oilfield was discovered in 1900.
Oil was discovered in northern Sumatra in 1883 and the Burmese production
began in 1887. Most of this Asian oil was shipped to Europe, where coal was
generally at the peak of its dominance and crude oil amounted to only a minuscule
addition to the total primary energy supply. Before World War I also came the first
major oil discoveries in the Middle East (on May 26, 1908, Masjid-e-Soleiman in
Iran), Mexico, and Venezuela (the giant Mene Grande field on Lake Maracaibo’s
coast in 1914). With the exception of pre–WWI discoveries in Iran, all major
finds in the Persian Gulf region came only between the late 1920s and the early
1960s. Iraqi Kirkuk was first (discovered in 1927, producing since 1934), followed

Figure 2.3 Baku oil wells. Reproduced from The Illustrated London News,
June 19, 1886, 9, 671.
 GLOBAL TRANSITIONS: UNIVERSAL PATTERNS 35

by Iranian Gachsaran and Haft Kel in 1928, Naft-i-Said in 1935, Pazaran in 1937,
and Agha Jari in 1938.
In that year came also the first large discovery in Kuwait, and in Saudi Arabia
(Dammam on the western shore of the Persian Gulf near Dhahrān), followed by
Abqaiq and Abu Hadrı̄ya in 1940, Qatı̄f in 1945, and in 1948 al-Ghawār (south-
west of Dhahrān), that was confirmed by 1956 to be by far the world’s largest res-
ervoir of crude oil. Canada also rejoined the ranks of major oil countries with the
discoveries of giant oilfields in Alberta (Leduc-Woodland in 1947 and Redwater
in 1948) and the Soviet center of oil production shifted from Baku to the
Volga-Ural region, where the first strike in 1937 (giant Tuymazy) was followed
by two more giants in 1945 and 1948 (Mukhanovo and Romashkino).
Postwar recovery in Europe, the USSR, and Japan and the U.S. baby boom–
driven economy stimulated demand for oil, as did the twin trend of rising car
ownership and shift to suburbs, soon joined by the jet-powered air travel.
Increasing networks of large-diameter pipelines and construction of massive
crude oil supertankers made it possible to export the fuel at a very low cost,
and general adoption of efficient catalytic cracking enabled production of larger
volumes of the most valuable transportation fuels, gasoline, kerosene, and diesel
oil. Low oil prices also accelerated the transition from biofuels and coals in a
number of Asian and Latin American countries.
Future of this increased supply seemed to be secure, as the 1950s and 1960s
were the two record decades for the discovery of giant oilfields. These finds
included giants in Saudi Arabia (Safānı̄ya-Khafjı̄, Manı̄fa, Berri, Shayba), Iraq
(Rumaila), Iran (Ahwaz, Marun, Fereidūn), and Abu Dhabi (Bū Hasa, Zākūm,
Asab), as well as in Canada (Pembina, Weyburn-Midale, Swan Hills) and the
United States (the Prudhoe Bay on the North Slope of Alaska in 1968) and the
largest Soviet supergiant in Western Siberia (Samotlor in 1965). Discoveries in
Algeria, Libya, and Nigeria made Africa into a major new supplier and a super-
giant Daqing oilfield in Heilongjiang (discovered in 1959) finally changed China’s
meager oil fortunes.
Size of the Middle Eastern oilfields and ready availability of Western (and later
also Japanese) investments brought rapid extraction increases: For example, the
Saudi output (all of it managed by the Arabian American Oil Company) tripled
between 1960 and 1970 (from 62 to 192 Mt/year), while the Kuwaiti output
went from less than 1 Mt in 1945 to more than 80 Mt by 1960. Dissatisfaction
with low oil prices led to the establishment of the Organization of Petroleum
Exporting Countries (OPEC) in 1960. As the oil demand continued to rise while
the U.S. production began to fall in 1971 (it remained the world’s largest until
1975), OPEC began raising its prices. Its first round of large increases in 1973–
1974 was followed by the second round in 1979–1981 that was precipitated by
the overthrow of the Iranian monarchy.
At the same, discoveries of giant oilfields were declining, but they included
Mexico’s Cantarell Complex in 1976 and the three largest fields in the North
36 ENERGY TRANSITIONS

Sea’s Norwegian waters (Ekofisk in 1971, Statfjord in 1974, and Gullfaks

in 1978). Concerns arose about the adequacy of the future supply but OPEC
overplayed its hand and as the oil price rose to nearly $40/bbl, the global
demand (and with it the worries about an imminent peak oil production, as well
as new drilling) receded. Global oil production peaked at just over 3.2 Gt in
1979 and it did not surpass that level until 1994 (Figure 2.4). Prices remained
fairly stable until the century’s end, and new major discoveries of the 1990s came
from Mexico, Iran, Brazil, and from the U.S. offshore waters in the Gulf of
Mexico (Ursa in 1991, Auger in 1996, and Alpine and Thunder Horse in 1999).
Meanwhile another major change took place, as the USSR, the world’s largest
oil producer since 1975, dissolved, and the aggregate oil extraction of its former
states declined by nearly a third between 1991 and 1996, making Saudi Arabia a
new leader starting in 1993. Prices remained low and fairly steady and total
extraction increased modestly during the first five years of the new millennium.
Subsequent combination of a weaker U.S. dollar (all international oil trade is
denominated in US$), speculation in commodity futures, and rising demand
in China and India pushed them to new nominal (and demand-destroying)
highs of nearly $150/bbl in July 2008 before they receded, and later somewhat
rebounded. As a result, crude oil’s share of the global primary fuel supply has
been declining while the shares of coal and natural gas have been rising.
Natural gas is actually a mixture of light combustible hydrocarbons, with
methane dominant but with up to a fifth of the volume made up of ethane,

Figure 2.4 Crude oil production, 1850–2010. Plotted from data in UNO (1956
and 1976), Etemad et al. (1991), and BP (2009).
 GLOBAL TRANSITIONS: UNIVERSAL PATTERNS 37

propane, and butane; other gases commonly present include CO2, H2S, N2, and
water vapor. Because of its natural seeps the gas was also known in some regions
in antiquity and its first well-documented commercial use dates to the Han
dynasty (200 BCE), when wells were drilled with percussion tools and the gas
was led through bamboo tubing to burn under large iron pans and evaporate
brines to produce salt in the landlocked Sichuan province (Needham, 1964).
Americans were the industry’s modern pioneers, with the first shallow natural
gas well dug in 1821 in Fredonia, New York, by William Hart. Rising volumes
of the gas became available with the expanding crude oil production (associated
gas, mixed with or dissolved in the liquid fuel) but in the absence of long-distance
pipelines some of the fuel was used for local lighting but most of it was wasted
(set alight and flared).
Moreover, in the cities natural gas faced a long-established competition
by the gas made from coal (town gas, first produced in London in 1812; in the
United States, in Baltimore, in 1816), and by the 1880s gaslights began to be
replaced by new electric lights. Three innovations had to take place before natu-
ral gas could become a major household and industrial fuel: commercialization
of a safe burner mixing the gas and air in correct proportion to produce a safe
flame for cooking and heating; introduction of large-diameter, high-pressure
pipelines that could carry greater volumes of the gas over longer distances; and
efficient compressors to propel the gas through pipes. The first advance began
with Robert Bunsen’s burner in 1885 and it was perfected by temperature-
regulating thermostats that could be used to monitor and regulate the flame
as needed.
Better pipes and pipeline construction method began to diffuse during the
1930s, but it was only after World War II when metallurgical advances and better
welding and pipe-laying techniques brought a pipeline construction boom (with
trunk lines having diameters up to 120 cm), first in the United States, and by the
1960s also in Europe and parts of Asia. Another important post–WWII advance
was the replacement of reciprocating engines or electric motors used in centrifu-
gal compressors that pressurize the transported gas (and spaced in intervals of
60–160 km along the pipeline) by inherently more efficient gas turbines which
could be powered by consuming a small amount of the transported gas. These
technical constraints were the main reason why—with the exception of the
United States, where natural gas extraction rose to more than 10% of all fuel pro-
duction by 1940—the gas industry became an important global presence only
after World War II.
Clean combustion and flexible use made natural gas a much sought-after fuel,
used not only for space heating and cooking but also in numerous industrial
processes as well as to generate electricity (either by producing steam in thermal
plant boilers or by powering gas turbines). Moreover, natural gas has also out-
standing non-energy roles as an excellent feedstock for production of ammonia
and a wide variety of synthetic materials. Not surprisingly, its global extraction
38 ENERGY TRANSITIONS

expanded rapidly, from about 200 Gm 3 in 1950 to 1.2 Tm 3 by 1975 and

2.4 Tm3 in 2000, a 12-fold rise in 50 years. The United States remained the
world’s largest natural gas producer until 1982, when it was surpassed by the
USSR: Ever since, the flow from the supergiant fields in Western Siberia has kept
Russia as the world’s largest producer and exporter of natural gas (Figure 2.5).
Canada has been the world’s third-largest producer since the late 1950s and
discoveries in the North Sea made the United Kingdom the fourth-largest pro-
ducer. Algeria, Indonesia, Iran, the Netherlands, Uzbekistan, Kazakhstan, and
Norway made up the rest of the global top ten in 2000. Growing LNG trade—
by 2009 15 countries, led by Qatar and followed by Malaysia, Indonesia, Algeria,
Nigeria, Australia, and Trinidad and Tobago were exporting LNG and Japan,
South Korea, Spain, the United States, Taiwan, and France were the main
buyers—means that much less gas is now wasted, but in 2005 the share of flared
gas, at about 4% of the global output, was still equal to nearly half of Russia’s
large exports (Mouton, 2005). Gas-flaring sites in Russia, Iraq, Iran, and Nigeria
can be seen on nighttime satellite images as the brightest spots on the Earth’s
surface.

Figure 2.5 Natural gas production, 1880–2010. Plotted from data in UNO (1956
and 1976) and BP (2009).
 GLOBAL TRANSITIONS: UNIVERSAL PATTERNS 39

A NEW QUALITY: GENERATION OF ELECTRICITY

There are many reasons why electricity has become the preferred form of
energy and why it is, in many ways, absolutely essential for the normal function-
ing of modern civilization. Electricity’s use results in economic benefits unsur-
passed by any fuel, as it offers superior final conversion efficiencies, unmatched
productivity, and unequaled flexibility, with uses ranging from lighting to space
heating, from metallurgical to food industries, and from any stationary to all but
one mobile use (commercial flight). Its other much-appreciated advantages
include precise control of delivery (ranging from less than one watt for the most
efficient microchips to multigigawatt flows in large national or regional grids),
focused applications on any conceivable scale (from micromachining to power-
ing the world’s largest excavators and the world’s fastest trains), and, of course,
no need for storage and the ease of using (flipping the switch) energy that is
noiseless and, at the point of conversion, absolutely clean.
And with the now universal reliance on electronic monitoring and auto-
mation (be they in incubators or nuclear reactors, rapid trains or large banks)
electricity’s role as the controller, regulator, and enabler of materials and infor-
mation flows became even more fundamental: Only a small share of its genera-
tion energizes these controls—for example, the best calculations show that
servers, modems, and routers consume about 3% of all U.S. electricity (Koomey,
2007) and that the Internet claims about 5% of the global electricity use (Sarokin,
2007)—but cessation of that supply would have profound effects on modern
societies.
Most of the world’s electricity is generated by the burning of fossil fuels:
In 2005 that share was almost exactly two-thirds, with coal accounting for more
than 60% of the latter fraction, or just over 40% of the global total (WCI,
2009). Hydroelectricity came next with about 17%, followed by nuclear elec-
tricity (just over 15%); despite their recent rapid expansion, all forms of new
renewable electricity generation supplied only about 2% of the 2005 total, with
most of it coming from wind turbines: In 2008, after doubling the 2005 installed
capacity, their share was 1.5% of the global total (WWEA, 2009). Geothermal
plants were the second-largest renewable contributor and photovoltaic (PV)
conversion generated a minuscule share of less than 0.1%. Moving away from
the reliance on fossil fuels in general (and from coal combustion in particular)
would necessitate a massive reshaping of the industry.
Anything beyond a marginal (less than 10% share) shift to generation based
on renewable energy flows would affect the key infrastructural arrangements of
the industry (location, size, and load factors of its generating capacities and high
voltage (HV) transmission links among the primary concentrations of new sup-
ply and major load centers). Alternatively, any bold moves toward a system based
largely on nuclear generation would have to overcome a number of recalcitrant
socioeconomic and technical obstacles (with the post-1970s record giving little