IEA and Oil Track record analysis and assessment of Petter Henke

IEA and Oil Track record analysis and assessment of Petter Henke
UPTEC ES 14014
Examensarbete 30 hp
Maj 2014
IEA and Oil
Track record analysis and assessment of
oil supply scenarios in WEO 2000-2013
Petter Henke
Abstract
IEA and Oil - Track record analysis and assessment of
oil supply scenarios in WEO 2000-2013
Petter Henke
Teknisk- naturvetenskaplig fakultet
UTH-enheten
Besöksadress:
Ångströmlaboratoriet
Lägerhyddsvägen 1
Hus 4, Plan 0
Postadress:
Box 536
751 21 Uppsala
Telefon:
018 – 471 30 03
Telefax:
018 – 471 30 00
Hemsida:
http://www.teknat.uu.se/student
The World Energy Outlook (WEO), an annual publication from the International
Energy Agency (IEA), is often considered to be the most authoritative source of
future energy scenarios for policy decision makers. The demand and supply scenarios
for oil, one of the most irreplaceable resources in the global energy system, are
central in each report. For the last decade, the outlook for oil supply in 2030 in the
main IEA scenario has been reduced by almost 20 million barrels per day.
The aim of this study is to examine the revisions to the oil supply scenarios, both at
global and individual country level, and note if and how the IEA has motivated these
revisions. The accuracy of past WEO scenarios is quantified by track record analysis
and the latest WEO scenario is assessed in detail in relation to current scientific
literature. Finally, implications of the latest WEO scenario for the long term oil supply
are assessed.
It is noted that the IEA generally motivate upward revisions to their scenarios, while
downward revisions are often left unmentioned. Some recent revisions are attributed
to the financial crises of 2008 and the largest revision in absolute terms is the gradual
downward revision of OPEC production motivated by an underestimation of key
producing countries’ will and ability to expand capacity.
The track record analysis indicates that the accuracy of the IEA scenarios has
increased on a five year prediction basis following the extended methodology applied
in the WEO2008. The analysis also shows that the accuracy of scenarios decrease
with time. On a ten year horizon, the mean absolute error for the IEA aggregate
‘World oil supply” was estimated to 8.2%.
The WEO2013 ‘New Policies Scenario’, with a time frame of 2012-2035, was
assessed using decline and depletion rate analysis, and compared to empirically proven
rates. The scenario was found to provide a realistic but optimistic view of the future
of oil supply. An alternative scenario, with depletion rates in line with the fastest
observed regional rates, resulted in somewhat lower production rates throughout the
scenario time frame.
A long term extrapolation to year 2100 of the WEO2013 scenario, based strictly on
resource and production data from the WEO reports, indicated that oil supply will
reach a peak in 2035 and then enter decline for the remainder of the century. A
sensitivity analysis showed that changes to the assumed resource base only moves the
peak by a few years, but has a significant effect on the rate of the following decline.
Handledare: Henrik Wachtmeister
Ämnesgranskare: Mikael Höök
Examinator: Petra Jönsson
ISSN: 1650-8300, UPTEC ES 14014
Sammanfattning
World Energy Outlook (WEO), den årliga publikationen släppt av International Energy Agency (IEA), är
en av de mest auktoritära källorna till framtida energiscenarier för politiska beslutsfattare. Scenarierna
för efterfrågan och tillgång på olja, en av de mest oersättliga resurserna i det globala energisystemet,
har en framträdande roll i rapporterna. Inom det senaste decenniet har utsikten för oljeförsörjningen
i IEAs huvudscenario minskat med nästan 20 mb/d.
Syften med denna studie är att undersöka dessa förändringar, sett både i global skala och för enskilda
länder, och notera huruvida IEA motiverar dessa. Även träffsäkerheten i IEAs tidigare scenarier
undersöks och kvantifieras. Studien syftar dessutom till att bedöma hur det senaste oljescenariot i
WEO står sig mot närmare vetenskaplig granskning. Till sists undersöks även vilka implikationer det
senaste scenariot kan komma att ha för oljeförsörjningen på längre sikt.
Studien finner att IEA mestadels motiverar ökningar i scenarierna, medan minskningar oftast lämnas
oomnämnda. En del förändringar i IEAs scenarier kan härledas till 2008 års globala finanskris, medan
andra inte får några förklaringar.
Resultaten från analysen indikerar att träffsäkerheten i IEAs scenarier ökade i och med att WEO2008
släpptes, baserat på jämförelser mellan 1 och 5 år fram i tiden. Analysen visar även att scenarierna blir
osäkrare med tiden, och det absoluta medelfelet för den globala oljeförsörjningen ligger på 8.2%, mätt
på en tio-års basis.
Huvudscenariot i WEO2013, IEAs senaste oljescenario med en tidsram på 2012-2035, utvärderades
med decline- och depletion rate analys, och jämfördes med empiriskt och historiskt bevisade värden.
Det fanns att scenariot ger en realistisk med likväl optimistisk bild av framtidens oljeförsörjning. Ett
alternativ scenario, med en depletion rate av nya fält begränsad till den snabbast observerade
utvecklade regionen, Nordsjön, resulterade i genomgående lägre produktionsnivåer för hela
scenariotiden.
En extrapolation av WEO2013s huvudscenario till år 2100, baserad på resurs- och produktionsdata från
IEAs rapport, indikerar att oljeförsörjningen når sin peak år 2035, och sedan för att sedan minska
kontinuerligt fram till 2100. En känslighetsanalys visar att förändringar i resursbasen endast flyttar en
peak några få år, men har desto större inverkan på minskningshastigheten efter en peak.
Acknowledgments
This study is the result of a master thesis within the Master of Science program in Energisystem at
Uppsala universitet and Sveriges lantbruksuniversitet. The work has been conducted at the institution
for Global Energy Systems (GES) at Uppsala University, under the oversight of supervisor Henrik
Wachtmeister, Research assistant at GES, and topic examiner Mikael Höök, Associate Professor, also
at GES.
First and foremost I would like to thank Henrik for your support throughout the study. It has been
invaluable to be able to bounce ideas off you, and your input has been greatly appreciated. I would
also like to thank Mikael for interesting insights and rewarding discussions. And to the rest of the Global
Energy Systems group, a big thank you for providing a good and fun working environment.
Finally, I would like to thank our Lord God for creating the dinosaurs1 some 7000 years ago and mashing
them up into our vital petroleum oil. And Satan too of course, for providing the heat and pressure.
Petter Henke
Uppsala 19/5-2014
1
Yes, dinosaurs. In your face evidence based science!
List of abbreviations
BTL – Biomass to liquids
CTL – Coal to liquids
EIA – Energy Information Administration
EOR – Enhanced oil recovery
Gb – Giga-barrels
GTL – Gas to liquids
IEA – International Energy Agency
LTO – Light tight oil
Mb – Million barrels
NGL – Natural gas liquids
OECD – Organization for Economic Co-operation and Development
OPEC – Organization of Petroleum Exporting Countries
RRR – Remaining recoverable resources
RTRR – Remaining technically recoverable resources
SCO – Synthetic crude oil
URR – Ultimately recoverable resources
USGS – United States Geological Surveys
WEM – World Energy Model
WEO – World Energy Outlook
YTD – Fields yet to be developed
YTF – Fields yet to be found
Disposition
This is a brief explanation of the disposition of this report and the focus of each chapter included,
aimed to provide the reader with information so as to decide which parts of the report are best suited
to the reader’s interest. As this report is the result of a master thesis, readers with some experience of
the global oil market and industry may find some sections to be a bit too basic. The report also focuses
on a few different results, which to some extent coincide, but not all chapters are necessary if looking
for specific results.
Chapter 1 features a basic level of introduction to the global oil market, pricing, forecasting and the
IEA. Readers familiar with those subjects may skip those parts and instead focus only on chapter 1.4,
which describes the aim of this study.
Chapter 2 contains the methodology and definitions used within the report and is for most part
recommended for all readers, regardless of previous experience. Chapter 2.1 provides a bullet list of
certain assumptions and limitations that this study is subject to, and chapter 2.2 describes the main
differences between scenarios and forecasts. Chapter 2.3 describes how the different types of oil are
categorized by the IEA. For those familiar with the theory behind depletion and decline rate analysis,
chapter 2.4 may be skipped.
Chapter 3, by far the longest, encompasses most of the data analysis of the Outlook reports. Chapter
3.1 describes some of the modeling methodology used by the IEA, and the subchapters review
resources and economic predictions. Chapter 3.2 features a short review of the primary global oil
supply scenario from each WEO, the changes to which is the main reason behind this project. Chapter
3.3 and associated subchapters review the oil production scenarios for some of the most prominent
producers and exporters in the world. The results of this review is mainly used for the track record
analysis, and does not affect the revised supply model that this project results in. The data used for
the model is instead supplied in chapter 3.4 and associated subchapters, which reviews the scenarios
by type of oil. These chapters also focuses on a comparison between WEO2008 and WEO2013.
Chapter 4 presents the results of the track record analysis, based on the findings in chapter 3. It reviews
and compares past IEA predictions to actual production, and assesses whether the IEA have proven a
reliable source for future oil production scenarios. While of interest for, and the main part of, the
review of IEAs past performance, this chapter holds little significance for the evaluation and
remodeling of the WEO2013 New Policies Scenario.
Chapter 5 contains the aforementioned remodeling of the WEO2013 New Policies Scenario, called the
Alternative scenario, alongside long-term extrapolations of both scenarios based on resource data
used and presented by the IEA. The Alternative scenario represents what is argued to be a more
realistic outlook for global oil supply for the period 2012-2035. The long-term extrapolations are
designed to review some of the implications for global oil supply in the longer term, based mainly on
resource restrictions.
Chapter 6 discusses some of the results, uncertainties and implications that the findings of this project
may hold, and chapter 7 concludes the report.
Table of Contents
1
2
3
Introduction ..................................................................................................................................... 3
1.1
The global oil market ............................................................................................................... 3
1.2
Predicting oil production and price ......................................................................................... 3
1.3
The International Energy Agency and World Energy Outlook ................................................ 4
1.4
Aim of study............................................................................................................................. 5
Methodology and definitions .......................................................................................................... 5
2.1
Assumptions, definitions and limitations ................................................................................ 5
2.2
Scenarios versus forecasts....................................................................................................... 6
2.3
Categorizing oil ........................................................................................................................ 7
2.4
Depletion and decline rates .................................................................................................... 8
Review of scenario revisions ........................................................................................................... 9
3.1
IEA modeling methodology ..................................................................................................... 9
3.1.1
Reserves and resources ................................................................................................. 10
3.1.2
Oil price.......................................................................................................................... 11
3.1.3
Investments ................................................................................................................... 13
3.2
Review of global supply scenarios ......................................................................................... 14
3.3
Review of regional production scenarios .............................................................................. 14
3.3.1
OPEC .............................................................................................................................. 15
3.3.1.1
Saudi Arabia ............................................................................................................... 17
3.3.1.2
Iraq............................................................................................................................. 19
3.3.1.3
Venezuela .................................................................................................................. 20
3.3.2
Russia ............................................................................................................................. 21
3.3.3
North America ............................................................................................................... 22
3.3.4
Brazil .............................................................................................................................. 23
3.4
Review of production scenarios by type of oil ...................................................................... 25
3.4.1
Crude oil ........................................................................................................................ 26
3.4.1.1
Currently producing fields ......................................................................................... 27
3.4.1.2
Yet to be developed .................................................................................................. 29
3.4.1.3
Yet to be found .......................................................................................................... 32
3.4.1.4
Enhanced oil recovery ............................................................................................... 34
3.4.2
Unconventional oil ........................................................................................................ 36
3.4.2.1
Canadian oil sands ..................................................................................................... 37
3.4.2.2
Coal- and gas-to-liquids ............................................................................................. 38
3.4.2.3
Light tight oil .............................................................................................................. 40
3.4.3
Natural gas liquids ......................................................................................................... 41
1
4
Track record analysis ..................................................................................................................... 44
5
Remodeling and extrapolation ...................................................................................................... 45
6
5.1
The Alternative scenario ....................................................................................................... 46
5.2
Long term extrapolation of the scenarios ............................................................................. 48
5.3
Sensitivity analysis ................................................................................................................. 51
Discussion ...................................................................................................................................... 55
6.1
IEA track record ..................................................................................................................... 55
6.2
The Alternative scenario and long term extrapolations........................................................ 56
7
Conclusions.................................................................................................................................... 58
8
References ..................................................................................................................................... 60
9
Appendix I ......................................................................................................................................... i
2
1 Introduction
1.1 The global oil market
Since the first oil was struck in the U.S. in the middle of the 19th century, demand on the fossil energy
source has grown. Initially used as a more humane replacement for whale oil for lighting, petroleum
demand around the world soon surged. While oil had been struck before, the beginning of the oil
industry if often attributed to a discovery in Pennsylvania in 1857 (Campbell, 2013a). Over the last 150
years, demand and production of petroleum oil has continually increased, oil becoming a fundamental
part of the global energy system and a key driver of the global economic development. Now used for
anything from transportation fuel to heating to manufacturing of plastic products, petroleum oil
satisfies more than 30% of the world’s energy needs. As petroleum is formed from organic matter
compressed and heated for millions of years, it is for all human intents and purposes a non-renewable
resource. As such, the production of oil is bound to peak at one point or another, the only question
remaining is when.
History of global oil supply
100
90
80
Mb/d
70
60
50
40
30
20
10
0
1930
1940
1950
1960
1970
1980
1990
2000
2010
Figure 1: History of global oil supply in Mb/d2. Sources: (Campbell, 2013b; Rystad Energy AS, 2014).
Figure 1 presents the increasing global oil supply from 1930 to 2012. Previous oil shortages such as the
1973 oil crisis and the Iranian revolution of 1979 has had great effects on economic growth (Salameh,
2004), especially in countries with high imports such as the members of the Organisation for Economic
Co-operation and Development (OECD), and caused the formation of the International Energy Agency
(IEA). The transport sector is particularly vulnerable to shortages, as no viable replacement fuel for oil
has yet been found or implemented. Insights into the prospects of future production is therefore of
vital importance in order to avoid being caught off guard by another supply crunch.
1.2 Predicting oil production and price
Predicting future oil supply and prices is notoriously difficult. While oil forecasts and scenarios has
been around for many decades, most have proven largely inaccurate (Miller and Sorrell, 2014). The
global oil market is an incredibly complex system, dependent on a vast combination of geological,
socioeconomic and political factors. The most fundamental underlying factors of supply and demand
are themselves incredible intricate systems based on numerous variables which are at best very
2
Mb/d = Megabarrels per day or million barrels per day. 1 barrel equals 159 liters.
3
difficult to predict. Still, while oil production scenarios and forecasts have a bad track record, they are
still essential for efficient political and economic policy decisions and planning.
Historically, many scenarios and forecast have stemmed from the belief that either demand or supply
is the determining factor in oil production. From the economist standpoint, global demand for oil is
the deciding factor and as long as the demand increases, so will supply. Others believe that geological
factors will limit supply, which will the not be able to meet the global demand. Most modern forecasts
and scenarios take both supply and demand in consideration, and view them as interconnected,
although they may still lean towards one factor being more deciding than another.
Aggregate oil demand is driven by the need for petroleum based fuels and products in the
transportation sector as well as in a vast amount of industry sectors ranging from agriculture to
petrochemical plants. Since the services and materials provided by these sectors are fundamental
drivers for economic growth, there exists a strong correlation between oil consumption and gross
domestic product (GDP) (Chu, 2012; Hirsch, 2008). As such, forecasting the demand on oil is generally
based on assumptions on GDP growth, both globally and for individual countries, which in turn is based
on a multitude of macroeconomic variables and future demographics. Demand is also affected by the
availability of substitutes, such as biofuels for the transportation sector in which petroleum products
are by far the most dominating and difficult to replace.
A diverse global oil supply system has developed to meet the demand, and oil production is spread out
over some hundred countries worldwide. The aggregate demand is met not only by the traditional
conventional crude oil but also by natural gas liquids (‘NGL’), biofuels and unconventional sources such
as Canadian oil sands, extra heavy oil and light tight oil (‘LTO’). Still, while diversified, global oil supply
hinges on a few key producers and exporters such as the members of OPEC. The top ten largest
producers in the world provide over 60% of all supply (IEA, 2012a), and around 80% of the aggregate
demand is met by conventional crude oil, with NGL making up most of the rest (Hughes and Rudolph,
2011). Since oil is a non-renewable resource the geological availability and accessibility as well as
technological and economic dynamics of extraction are central to understanding supply.
Besides the economical and geological factors of the oil supply and demand, the strategic importance
of oil from any state or nations perspective adds to the complexity. Oil resources can provide
geopolitical power and a stable oil supply is of great importance for any oil importing country’s energy
security. In summary, forecasting and scenario-design of future oil production is inherently coupled to
high uncertainty and unknown parameters. Still, policy decisions makers need reliable predictions of
future oil production due to the fundamental role of oil in the global energy and economic system.
1.3 The International Energy Agency and World Energy Outlook
The International Energy Agency (‘IEA’) was founded by the OECD as a response to the oil crisis of
1973/74 with the initial purpose of helping its member countries coordinate a collective response to
major disruptions in oil supply. Since its foundation, the organization has expanded on its initial
purpose, and now also works to ensure reliable, affordable and clean energy for its members, hence
moving its scope from oil to all energy. An important part of this is actively analyzing and predicting
trends in the global energy market in terms of supply and demand, and imports and exports of energy.
In the 1990’s3, the IEA published the first version of its now annual report on the current and future
status of the global energy market, the World Energy Outlook (‘WEO’)4. The report was aimed to
provide its members with insights into energy trends and support for energy-related policy decisions,
3
4
The actual year of the first report is somewhat unclear
The World Energy Outlooks are referred to as either ‘WEO’ or ‘Outlook’ for the remainder of the report.
4
and subsequent reports have followed suit. From the turn of the century, the WEOs provide much
more detailed insights and future scenarios for different energy markets such as oil, coal, natural gas
and electricity. It is generally perceived as one of the most authoritative energy publications available
(Kohl, 2010; Van de Graaf, 2012). Perhaps one of the most important scenarios the IEA presents in the
WEO is the outlook for the global oil supply and demand, which has changed drastically since the
beginning of the century.
Current WEOs provide three different scenarios, mainly differing in assumptions on governmental
policy decisions regarding environmental issues. The main scenario, the New Policies Scenario assumes
that governments will continue to enforce existing policies and implement recently announced
commitments and plans in a cautious matter. The Current Policies scenario envision a more ‘businessas-usual’ future, where no new environmental policies or limitations on emissions are adopted, while
the 450 scenario sees the world adopting policies in a manner as to have a 50% chance of limiting the
global increase in average temperature to 2 °C.
1.4 Aim of study
This study aims to review the oil production and supply scenarios published by the IEA in their annual
World Energy Outlook report for the years 2000-2013 and:
-
Identify revisions of production and supply scenarios and their motivations in regard to
underlying assumptions and methodology.
Quantify the accuracy of past IEA scenarios by way of track-record analysis.
Furthermore, the aim of the study is to assess the latest IEA main scenario, the WEO2013 New Policies
Scenario, based on current scientific understanding of the area and utilizing analytical methods such
as decline and depletion rate analysis. Based on this assessment the study will present an alternative
world oil supply scenario, derived from resource estimations, production data and scenarios from
WEO2013.
Finally, the study aims to explore the long term implications of both the WEO2013 and the alternative
oil supply scenarios in regard to resource availability.
2 Methodology and definitions
2.1 Assumptions, definitions and limitations
The following list presents the assumptions and limitations that this project is subject to, due to for
instance the lack of relevant and available data. It also gives a short introduction to some of the
definitions used within the report.
-
-
-
Only the central scenario of each WEO is analyzed, even in Outlooks where the IEA deem all
scenarios as equally probable. This is as the central scenarios are presented in greater detail.
Effectively, this means the Reference scenario for WEO2000-2009, and the New Policies
Scenario for WEO2010-2013.
The production of natural gas liquids (‘NGL’), an increasingly important part of global oil supply,
is heavily coupled to natural gas production (see chapter 3.4.3). However, natural gas is outside
the scope of this project, and therefore the scenarios for natural gas production in the
Outlooks are simply accepted and not subjected to any review.
The track-record analysis and modeling is restricted to figures presented in the WEO reports,
with a few exceptions in the long-term extrapolation in chapter 5.2 and data pertaining to
historical production. As the analyzed categories are not always presented in each separate
Outlook, there exists some inconsistencies in the data presented in this report.
5
-
-
-
-
-
-
Unless otherwise noted, all numbers in the analysis of the scenarios are taken from the text,
tables or figures presented in the Outlooks reports. The numbers from figures have been
extracted using a digitizer program, and are subject to some uncertainties.
Most WEO scenarios are presented in tables using point values for a few key years in the
scenario. In order to aid visual understanding of the scenarios and shed some light on the rate
of change, this report interpolates linearly between the data points. The interpolated values
are in most cases not used in any quantified results, the exception being the track-record
analysis of global production in chapter 4.
This study differentiates between production and supply, and between oil and all liquids. Oil
define all liquid hydrocarbons of fossil origin, while all liquids encompass oil and biofuels.
Production marks all oil extracted from the ground, while supply are the amounts available for
end use and includes the addition of processing gains5.
Biofuels are for the most part excluded from this analysis, hence the focus lies on global oil
supply rather than global liquids supply. For individual countries and typed of oil, processing
gains are not included and hence the scenarios are referred to as oil production and not supply.
Some of the earlier WEO reports are known as Insights, meaning that they feature a focus on
a specific area, such as Investments in WEO2003. The earliest Insights (2001 and 2003), feature
scenarios for oil that are almost identical to the 2000 and 2002 Outlooks respectively, and are
thus omitted from the analysis in all cases where they do not differ.
Economics has been, for the most part, left out of this study. Some financial scenarios such as
oil prices and investments are featured and put in relation with oil scenarios, but not critically
analyzed. Some simple economic relationships and simplified assumptions on the demandsupply relation is featured as part of the analysis.
2.2 Scenarios versus forecasts
Forecasts, in their most basic form, are attempts to predict the future, often extrapolating events from
recent or present trends. As such, the accuracy of the forecasts depends on a number of variables that
must change accordingly to the assumed trends. A forecast can be described as the “best” prediction
by a particular individual or party, or based on a particular technique. For policy makers, the credibility
of a forecast depends mainly on the credibility of the forecast technique (MacCracken, 2001).
Scenarios on the other hand present a few possible futures and are based mainly on expert opinions,
combining known facts with assumptions about the future. In practice, no single scenario should be
emphasized as more likely since there should not be any probability distribution between them. Unlike
forecasts, scenarios are very different from predictions in that depend on assumed changes in key
boundary conditions (MacCracken, 2001). Assessing the quality of a scenario necessitates access to the
basic assumptions and prerequisites for the scenario.
The IEA consistently refers to their future Outlooks as scenarios, meaning that they only present range
of futures that are possible if certain assumptions are met. The World Energy Outlook has become one
of the most authoritative sources for future energy scenarios, and often referred to in long term energy
policy decisions. The scenarios are however often quoted by others as predicting or showing the future
of energy, or often subjected to claims that this is the future according to the IEA. Hence the IEA
scenarios are often used as forecasts instead of scenarios, which may present a problem since policy
makers often rely heavily upon the IEA (Aleklett et al., 2010; Monbiot, 2008)
5
Processing gains is the added volume gained in the refinery process due to addition of various chemicals and
changes to the specific gravity.
6
It can be argued that the IEA energy and oil scenarios are a hybrid between forecasts and scenarios,
due to some modeling and extrapolation of trends coupled with assumptions and evaluations. The IEA
has however chosen not to disclose their main model, and many assumptions remain unstated. They
also present one of their three scenarios as the main scenario, implicitly stating this to be the most
probable. This in turn forces policy makers to rely on the good judgment of the IEAs analysts. Parts of
this study consists of a track-record analysis of former WEO scenarios, to assess whether such trust is
well placed.
2.3 Categorizing oil
Oil is actually a collective term for liquid hydrocarbons which comes in many forms with different
properties of the oil itself, and the geological structures in which the hydrocarbons are found are very
diverse. The various properties and geological structures lead to diverse extraction techniques being
used and large differences in the cost of extraction. Because of the differences in costs and properties,
oil is often categorized as either conventional or unconventional oil, with oil defined as unconventional
being more expensive in general. However, the exact definition of what oil is conventional or
unconventional oil is somewhat arbitrary (de Castro et al., 2009). Some authors choose to categorize
oil purely based on its properties, such as API-gravity6, while others define unconventional oil as liquid
hydrocarbons extracted by unconventional means, and some base their definition on the viscosity.
Table 1: IEAs definition of Unconventional oil in the different Outlooks. Green = included, Red = excluded, Yellow =
Unclear/not specified. OS = Oil sands.
Oil Shales
OS
Synthetic
crude oil
OS
Derivative
products
CTL
GTL
BTL
Venezuelan
extra heavy
OS Raw
natural
bitumen
Light tight
oil
WEO2000
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
This report strictly follows the IEAs definition of conventional and unconventional oil, which is based
more on the extraction techniques used than the properties of the oil. While the definition of
unconventional oil has changed between the Outlooks as seen in Table 1, the WEO2013 definition is
6
API gravity is a term constructed by the American Petroleum Institute which measures density of the oil
compared with water, and defined as API gravity = (141.5/specific gravity at 60° Fahrenheit) – 131.5 (Höök et
al., 2014a). An API greater than 10° means that the oil will float on fresh water. Other properties of oil include
low or high sulfur-content (making the oil sweet or sour respectively), and the viscosity of the oil (reported in the
unit centipoise (Chew, 2014)).
7
used for all sections of this report, unless otherwise noted. Biomass-to-liquids (‘BTL’) or biofuels,
previously included as unconventional oil, is now an entirely separate category and not included in the
definition of the World oil supply. When biofuels are included, it will instead be referred to as World
liquids supply. The conventional oil category is also divided into conventional crude oil and natural gas
liquids, or NGL for short (see chapter 3.4.3 for a more detailed description of NGL).
2.4 Depletion and decline rates
This study employs decline and depletion rate studies on conventional crude oil scenarios in the latest
Outlook, the WEO2013, in order to assess whether the scenarios fall in line with empirical evidence.
The following chapter includes a brief explanation of the concept and definition of decline and
depletion rates. A more comprehensive description of the methodological foundations and empirical
evidence for the use of decline and depletion rate analysis can be found in (Höök et al., 2014a) and
(Höök, 2014).
Figure 2: Idealized production profile for an oil field. Source: (Höök et al., 2009b)
Figure 2 shows an idealized production profile for a standardized oil field. The field is first discovered
and the appraised to give a general view of the size and amount of oil expected to be in place. It usually
takes a few years from the discovery to first oil production, the average lead time for giant oil fields
being around 5 years (Höök et al., 2009b). After the first oil there is a build-up phase where an
increasing amount of wells are drilled and production increases. At some point production from the
field will even out, as the field enters a plateau phase7. While new wells may still be drilled in the field,
these will only compensate for decreasing production from older wells. The production profile for an
actual oil field may differ greatly from that presented in Figure 2, as the plateau phase may consist of
many smaller peaks or in some cases, a single peak followed by rapid decline. Regardless of the profile
of the plateau phase, every oil field inevitably enters a state of decline, where production decreases
steadily. At some point or another, the field is abandoned. In most cases, a large amount of the oil
originally in place will still remain in the ground, as the production becomes too expensive.
The rate at which oil can be produced from a field or a region, which can be described as a fraction of
either the ultimately recoverable resources (‘URR’) or the remaining reserves8, is called the depletion
rate. There has been a lack of a standardized use and definitions of depletion rates, leading to
7
A plateau in oil production is a phase where production stays at approximately the same levels for an
extended period of time
8
See chapter 3.1.1 below for definitions of resources and reserves.
8
confusion surrounding the concept. In this study the depletion rate is defined as the rate at which the
remaining recoverable resources are produced at time t, described by equation 1.
𝑞
𝑞
𝑑𝑅𝑅𝑅,𝑡 = 𝑅𝑡 = 𝑈𝑅𝑅 𝑡−𝑄
𝑟
𝑡
𝑡
(1)
where qt is the production at time t, URRt the ultimately recoverable resources and Qt is the cumulative
production at time t (Höök et al., 2014a). There exists no defined theoretical maximum for the
depletion rate of a field or a region, as it depends on both geological and economic factors, where for
example added investments can increase the development speed. The depletion rate of a region can
however never exceed that of the highest depletion rate of a field in that particular region, just as the
depletion rate of the world cannot be higher than the fastest regional depletion rate. The maximum
depletion rates of regions rarely exceed 5%/year (Sorrell et al., 2012), with the fastest developed
region in the world, the North Sea, reaching a maximum of 6.3% (Aleklett et al., 2010). The global
average depletion rate of the remaining recoverable resources is at present approximately 1.2%
(Sorrell et al., 2012).
The use of decline rate analysis suffers less from confusion than depletion rate analysis, and is generally
more accepted and agreed upon. Conceptually, the decline rate refers to the annual rate of change in
production from any field or region, and is defined by equation 2 from (Höök, 2009).
𝐷𝑒𝑐𝑙𝑖𝑛𝑒 𝑟𝑎𝑡𝑒𝑛 =
𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛𝑛 −𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛𝑛−1
𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛𝑛−1
(2)
The decline rate can be positive in for example a still developing individual field, although in general
decline rates are used to described fields that have already peaked or groups of fields where the total
production falls. It has been shown that in general smaller fields decline at greater rates than larger,
and that offshore fields are both depleted faster and declines quicker (Höök et al., 2009a). Decline
rates can also be used to describe the change in all existing conventional crude oil production, i.e. the
rate of change in aggregate global production, where all fields in production, whether still they are
expanding, in plateau or in post-peak decline are included. There exists a broad agreement among
analysts that the decline in existing production is between 4-8% annually (Höök et al., 2014a).
3 Review of scenario revisions
3.1 IEA modeling methodology
The basis for all scenarios in the IEAs World Energy Outlook is the World Energy Model (‘WEM’)9. The
model encompasses a large number of areas such as energy supply and demand, power generation,
emissions, energy prices and investments to name a few. This study focuses mainly on oil supply and
to a smaller extent demand, oil prices and investments. The oil module of the WEM has evolved, from
a more simple demand driven scenario-builder to a supply-demand hybrid, based partially on a
bottom-up approach10. Earlier version of the model featured mainly a top-down approach, with some
bottom-up analysis of mainly the OECD countries (Kato, 2003). In these versions the assumption was
mainly that demand drives supply, and the members of OPEC were assumed to fill the gap between
expected oil demand and supply from other countries. In later versions of the model, OPEC supply is
based on a similar approach as the rest of the world, and demand and supply interact more.
9
A more detailed description of the 2013 World Energy Model can be found in (IEA, 2013a).
Meaning it is based on a field-by-field analysis of each country. The IEA began using this approach in
WEO2008. A more detailed description of what constitutes as bottom-up modeling can be found in (Jakobsson
et al., 2014)
10
9
The model changed most radically with the incorporation of a field-by-field and decline rate analysis
in the WEO2008. The 2013 WEM features a bottom-up approach to such things as historical production
series for countries, a decline rate analysis on fields and country levels, estimations of ultimately
recoverable resources and a methodology that attempts to replicate the industry decision making
process (IEA, 2013a). Currently, the exogenous variables to the WEM are the CO2 prices, future policies,
technological advancements and socioeconomic drivers (such as GDP and population growth). Fossil
fuel prices are perhaps best described as a semi-exogenous variable, as they are based on assumptions
in combination with an iterative process (see chapter 3.1.2).
3.1.1 Reserves and resources
In any scenario development of future oil supply it is of utmost importance to know how much oil
remains to be extracted, how much of that is technically feasible and financially acceptable. Since there
exists oil in viable areas that have not yet been explored, the remaining resources of oil are divided
between Discovered and Undiscovered oil. Also, due to technical and economical limits the resources
are again split into categories of Commercial and Sub-commercial. Figure 3 illustrates the division of
resources originally proposed by McKelvey (1972), which is still used today. It is of great importance
to make the distinction between Reserves and Resources. Resources generally account for all oil
thought to remain underground, including that in sub-commercial fields and in those fields that have
not yet been discovered. Meanwhile, Reserves only encompass oil in known fields thought to be
extractable both from a technical and financial standpoint (Sorrell et al., 2009).
Figure 3: McKelvey box. Adapted from Sorrell et al. (2009)
As seen in Figure 3, reserves can also be divided in 1P, 2P and 3P reserves, marking the confidence of
estimations. 1P marks proved reserves, 2P is proved + probable and 3P proved + probable + possible,
and hence the least certain. Most countries only report 1P data as reserves, but in the absence of an
international standard, the definitions of 1P can vary greatly between different countries and different
companies (Sorrell et al., 2009). 1P data also tend to underestimate the amount of extractable
reserves, due to for instance reserve growth, and many authors believe that 2P data would be more
accurate to use (Bentley et al., 2007). Reserve growth marks increasing amounts of for instance 1P
reserves, without any new oil being discovered, and can have many explanations such as initial
underestimations of the OOIP (original oil in place), technological advancements making more oil
extractable or increasing oil prices, rendering formerly sub-commercial fields viable. Reserve growth
has led many to overestimate the rate of new discoveries, as the increases to the reported reserves
are sometimes reported for the current year, and not the year that the oil was originally discovered.
This can be avoided by backdating discoveries, hence reporting the reserve growth as part of the
10
original discovery. A more comprehensive investigation of this and other issues regarding reserves and
resources can be found in (Sorrell et al., 2009). Some authors also suggest that the more advanced
seismic instruments and increased knowledge of the geological structures of newly discovered fields
will decrease future reserve growth (Schindler and Zittel, 2008; Sorrell et al., 2012).
Ultimately recoverable resources (‘URR’) is another important term which denotes the sum of all oil in
a field or a region that is expected to be extracted over all time. It includes estimations of undiscovered
oil and oil that is thought to be recoverable with future technology and oil prices, as well as cumulative
production up to date. While it can be argued that URR is only an estimation of the total volume of
recoverable oil, some argue that URR should be defined in such a way that it takes into account
anticipated technological advancements and changes in market conditions (McGlade, 2012). Removing
the cumulative production from the URR estimate gives the remaining ultimately recoverable
resources (‘RURR’), sometimes simply referred to as ‘RRR’, which is used for the depletion rate analysis
portions of this study. Some organizations, such as the IEA sometimes choose to report a subset of the
RRR, known as remaining technically recoverable resources (‘RTRR’). This is the fraction of oil which is
thought to be extractable by current technologies, not taking economic viability into account.
Table 2: IEA WEO2013 estimation of reserves and resources
Conventional oil
(including NGL + lease condensate)
[Gb]11
Proved
reserves
URR
RRR
NGL
Cumulative
production
1702
3804
2668
465
1136
Unconventional oil
Kerogen
(Oil
Light
EHOB shales) tight oil
Total
1879
3297
1073
345
Table 2 shows the IEAs estimation of both reserves and resources in the latest Outlook report,
WEO2013. Proved reserves has shown a continual increase throughout all Outlooks reporting specific
resource estimations, with the trends for the other categories being less clear (there are only a few of
the more recent Outlooks that report more detailed resource data). While incorporated in the
conventional category here, the proved reserves may include some unconventional oil reserves such
as extra-heavy oil from Venezuela (see chapter 3.3.1.3 for more details). The proved reserves numbers
are sometimes used as an argument against an impending peak in oil production, referred to as ‘Peak
Oil’12. However, the Peak Oil concept is not about running completely out of oil, but about limitations
in the production rates. A peak in oil production is therefore thought to occur long before the world
runs out of oil (Sorrell et al., 2009).
3.1.2 Oil price
The future of oil prices in as previously stated notoriously difficult to predict, with most methods and
models showing little success. It has been shown that the most commonly used forecasting techniques
cannot consistently beat a random-walk scenario out of a sample (Alquist et al., 2013). This is mostly
due to oil prices being extremely volatile, even on a year-average basis as shown in Figure 4 (actual
11
Gb = Gigabarrels = one billion barrels of oil
The concept of Peak Oil was first proposed by Marion King Hubbert in 1956 (Hubbert, 1956), and means
simplified that at one point oil production will reach a maximum and then start to decline. It is sometimes
limited to the peak in conventional oil production (Campbell and Laherrère, 2008; Chapman, 2014; Sorrell et
al., 2009). Although the theory is not always recognized (Bardi, 2009), the date of a peak in production and the
effects on the global economy are much debated and explored (de Almeida and Silva, 2009; Holland, 2013;
Kerschner et al., 2013),.
12
11
prices vary day-by-day), as the price is influenced by a number of factors such as demand, production,
OPEC policies, spare capacity etcetera. Benes et al. (2012) notes that the rapid recovery of oil prices
after the financial market crash in 2008 coincides with strains on spare capacity. The oil price is a
seemingly semi-exogenous variable in IEAs World Energy Model and reflects the IEA view on the pricing
necessary to meet the demand. The process is iterative, if the current assumption on oil price fails, a
price feedback process then recalculates the demand on oil (IEA, 2013a).
Oil price scenarios in year 2012 dollars
160
2012 US Dollars/barrel
140
120
100
80
60
40
20
0
1970
1980
WEO2004
WEO2007
WEO2010
WEO2013
1990
2000
2010
WEO2005
WEO2008
WEO2011
BP statistical review
2020
2030
WEO2006
WEO2009
WEO2012
Figure 4: Oil price scenarios in real terms converted to year 2012 dollars and compared to actual oil prices from BP, also in
year 2012 dollars (BP, 2013)
Figure 4 shows the IEAs assumptions on the changes in oil prices, presented in the Outlook reports in
real terms and converted here to year 2012 dollars for comparison to BPs historical prices. In general,
and somewhat like expected, the IEAs assumptions has been flawed. The 2004-2007 Outlooks in
particular fails to identify the continuing increase of the oil price. Most other WEOs sees oil prices rising
and leveling at approximately today’s prices, somewhere between 100-140 US dollars/barrel, with a
tendency for slightly higher prices in each successive Outlook, WEO2008 excluded. WEO2008, while
(not surprisingly) missing the price drop following the financial crisis, is still successful in anticipating
the current oil price. Most IEA scenarios suggest only a moderate increase, with the oil price staying
close to the current price. It has been shown that as far as accuracy goes, it is better to assume that oil
prices will not change (Benes et al., 2012), and the current IEA standpoint is that the high oil prices are
here to stay. This would suggest an end to the era of cheap oil, in line with what other authors have
suggested (Jakobsson et al., 2012; Tsoskounoglou et al., 2008).
12
3.1.3 Investments
130
125
120
115
110
105
100
95
90
85
80
500
450
400
350
300
250
200
150
100
50
0
Global oil supply by end of scenario (left axis)
Billion USD/y
Mb/d
Average annual total investment (year 2012 USD) and oil
supply by end of scenario
Annual total investment (right axis)
Figure 5: Forecasted investment in year 2012 USD compared to forecasted global oil supply by end of scenario (year 2030 for
WEO2003-2009 and year 2035 for WEO2010-2013)
Figure 5 compares IEAs predictions of the global oil supply by the end of each scenario, compared to
the assumed investments necessary to reach those levels. The investment costs in the Outlooks are
presented in real terms and have subsequently been converted to year 2012 USD for comparative
purposes. Also for comparative purposes, the total investments presented in the Outlooks have been
divided by the scenario length, since the length differs between Outlooks. The figure is limited to oil
supply, hence excluding biofuels, as the investments only cover oil refining, infrastructure and
upstream costs (exploration, drilling etc.). WEO2013 only presents upstream investments costs, to
which refining and infrastructure investments from WEO2012 have been added, both of which have
been fairly constant.
As seen in Figure 5, the annual investments deemed necessary in each Outlook has continually
increased, with the exceptions of WEO2005 and WEO2009, the latter likely owing to the financial crisis.
Meanwhile, the predicted oil supply has decreased, highlighting the assumed increased cost of oil
production and the necessity of heavy investments to reach IEAs targets. Interestingly, the correlation
factor between the global oil supply and the investment needs are approximately -0.9, indicating a
strong inverse relationship between the two. It would appear that the IEAs earlier scenarios have been
far too optimistic, not only in terms of oil supply but also in investments needs. Provided that the latest
Outlooks are more accurate, WEO2003 seems to have underestimated the investment needs the most,
even though it’s an Insight-report specifically dedicated to the subject of investments.
13
3.2 Review of global supply scenarios
Global liquids supply scenarios
130
120
110
Mb/d
100
90
80
70
60
50
40
1970
1980
1990
WEO2000
WEO2005
WEO2008
WEO2011
BP statistical review
2000
WEO2002
WEO2006
WEO2009
WEO2012
2010
2020
2030
WEO2004
WEO2007
WEO2010
WEO2013
Figure 6: Global liquids supply scenarios compared to actual supply data from BP (BP, 2013). Processing gains and biofuels are
included. Scenarios for biofuels have been added separately for WEO2006-2009 to make the scenarios comparable
Figure 6 shows the WEO scenarios for global liquids supply from 2000-2013, and compares it to
historical oil supply data from BP. The BP data excludes both biofuels and coal-to-liquids (CTL) (BP,
2013). While biofuel and CTL supply has not been significant historically, their absence in the BP data
can explain the small differences between the scenarios starting values and the BP supply data.
Analyzing Figure 6 makes it obvious that the IEA has changed its view of the future of oil supply rather
dramatically in the last decade, which is the main reason behind this project.
The 2008 and 2013 scenarios are highlighted in Figure 6, and in all subsequent figures featuring the
Outlook scenarios. The WEO2008 scenario is important in that it was the first Outlook scenario to be
based upon a more scientific approach, with the IEA utilizing tools such as decline rate analysis. It was
also the subject of the analysis performed in the article The Peak of the Oil Age (Aleklett et al., 2010),
sometimes referred to as the ‘Uppsala critique’ (McGlade, 2013). This project draws upon the
methodology used by Aleklett et al. in the analysis of the WEO2013 scenario (which features an
updated decline rate analysis), and compares the results.
From Figure 6 it is evident that the IEAs outlook for global oil supply has been reduced. It does not
however tell us anything about the specific changes to IEAs stance, or whether the changes have been
made on a regional level or on a specific type of oil. In order to review the changes the IEA has made
to their central oil supply scenario, specific portions of the scenarios are analyzed, both on a regional
basis (chapter 3.3) and by type of oil (chapter 3.4). The IEAs view on future oil prices, investments and
changes made to global oil resources are also reviewed.
3.3 Review of regional production scenarios
This section looks at scenarios for a few selected oil producing regions of great importance for world
supply. It analyses IEAs view of the three largest producing countries in the world: Saudi Arabia, Russia
and USA (featured as part of North America), as well as some countries presumed to be of great
14
importance for future oil supply, such as Iraq and Brazil. Venezuela is featured because of its large
reserves of unconventional oil. Due to time constraints in the project, other important producing
countries and regions are not featured here. Due to some differences in the reporting methodology
between the Outlooks, such as not including NGL for separate countries in certain reports, some
scenarios may not be directly comparable between WEOs. Cases where a particular scenario differs
from the rest are noted in the respective chapters.
The regional scenarios are not analyzed by decline and depletion rates as reserve data on a regional
basis is usually either restricted or uncertain (which is the case for members of OPEC in particular).
Rather this section aims to analyze in which way the IEA justifies these scenarios, accounts for
significant revisions between Outlooks and handle uncertainties. Forecasts for countries which have
been featured consistently are used for the track record review in chapter 4. The results of this section
are not used for the reevaluated scenario, as that model is based on oil by type rather than region. As
uncertainties are great for some of the regions however, they are discussed in the sensitivity analysis
(chapter 5.3) in which the impact of some above-ground events such as continuing instability in Iraq is
assessed.
3.3.1 OPEC
OPEC production scenarios
70
60
Mb/d
50
40
30
20
10
0
1970
1980
1990
WEO2000
WEO2005
WEO2008
WEO2011
BP Statistical review
2000
WEO2002
WEO2006
WEO2009
WEO2012
2010
2020
WEO2004
WEO2007
WEO2010
WEO2013
2030
Figure 7: Production scenarios for OPEC compared to actual production from BP (BP, 2013). Unconventional oil is not included
before WEO2006.
The Organization of Petroleum Exporting Countries (OPEC) consists of Algeria, Angola, Ecuador, Iran,
Iraq, Kuwait, Libya, Nigeria, Qatar, Saudi Arabia, United Arab Emirates and Venezuela. While many of
these countries are important oil exporters to the global market, due to time constraints in this project
the only scenarios studied in greater detail are Saudi Arabia, Iraq and Venezuela. Saudi Arabia currently
has the largest oil production of any OPEC member by far, and is as of 2012 the largest producer in
the world (BP, 2013). Iraq is featured as the IEA expects Iraqi production to increase rapidly, and later
Outlooks sees the country contributing more to the growth of global oil supply than any other.
Venezuela’s current production is approximately the same as Iraq’s, but is expected to grow much
more slowly. Instead, Venezuelan scenarios are of interest because of their dependence on
unconventional extra-heavy oil, and the fact that this resource is estimated to be the largest in the
world.
15
As seen in Figure 7, the expectations on OPEC in earlier Outlooks have been great, while more recent
reports suggest that OPEC production will grow at significantly slower rates. Indeed, the difference
between the highest predicted production point in 2030 (WEO2002-2004, both with almost 65 Mb/d)
and the lowest (WEO2013 some 42 Mb/d) is around 23 Mb/d. This more than covers the entire
difference between the global supply in WEO2002 and WEO2013, which is slightly above 20 Mb/d.
Figure 8 illustrates the changes made to global oil production scenarios for the year 2030 compared to
OPECs anticipated production. WEO2002 is used as the baseline for the comparison as it is the first
Outlook that runs the scenario to 2030. The correlation factor between the global and the OPEC
production outlook is 0.96, indicating that the two are heavily related.
Changes in oil production scenarios for 2030 compared to
WEO2002
5
0
Mb/d
-5
-10
-15
-20
-25
Global
OPEC
Figure 8: Changes in oil production scenarios for 2030, with WEO2002 as the baseline.
The large differences between the earlier and later Outlooks likely stems from a combination of the
more simple demand-driven WEM that the IEA previously used, and their assumption that OPEC would
simply fill the gap between oil demand and non-OPEC supply (“call-on-OPEC”) (IEA, 2006, 2004). From
WEO2005 an onwards the IEA regularly discuss the importance of upstream investments, in particular
in the OPEC-Middle east countries. Both limited investments and OPEC production policies (basically
holding back production to increase net profits from the oil) are seen as risks to the rapid expansion of
OPEC oil production predicted in the Outlooks (IEA, 2005). In Figure 7 it is seen that expectations on
OPEC have been drastically reduced from earlier Outlooks and Figure 8 shows that the OPEC scenarios
for 2030 have dropped in roughly the same manner as world oil production. This can perhaps point to
that the IEAs predictions on demand have been severely overestimated, which in turn would reduce
the “call-on-OPEC”. However, on page 253 of WEO2008 IEA also states that
“It now looks much less likely that the key producing countries, in particular, will be willing and able to
expand capacity as much and as quickly as previously assumed.” (IEA, 2008)
From the context in the report, it seems clear that “key producing countries” means mainly the OPEC
members, showing that the IEA now more than previously acknowledges OPECs stated production
policies. It should be noted that the IEAs expectations on OPEC production, at least the 2007 and 2013
16
scenarios, closely matches the view presented in OPECs own World Oil Outlook (WOO) (OPEC, 2013,
2007). It should also be noted that while OPEC production scenarios have been revised downwards,
the organizations importance for the IEA scenarios has stayed roughly the same in terms of total
market share, dropping only slightly in the latest Outlooks.
In general, it appears that the IEA standpoint is that OPECs capacity for increased production is mainly
limited by the organizations production policies. Ample oil reserves in Venezuela and the Middle East
countries, and the cheap nature of Middle Eastern oil, indicates that the OPEC members indeed have
the possibility of expanding production rapidly. However, some critics point to the fact that most OPEC
members suddenly increased their reported oil reserves in the 1980s, in what has become known as
the OPEC ‘quota wars’. These revisions came as OPEC policies changed to allow larger production
quotas for members with larger reserves. Many critics point to these changes as casting doubt on
actual OPEC oil reserves, while others claim the uselessness of 1P reserve data in favor of 2P reserves
(Proven + Probable reserves) (Bentley et al., 2007). The IEA have also been criticized for consistently
underestimating the growth of consumption in the OPEC and Middle East countries (Gately et al., 2013,
2012). A continued underestimation of domestic consumption growth could lead to misconceptions of
the export capabilities of these countries.
In conclusion, OPEC remains a key part of the IEA oil supply scenarios. While predicted OPEC
production has dropped, it has done so at approximately the same rate as the scenario for world oil
supply, and the organizations anticipated market share has remained roughly intact. Hence, it appears
that revisions to the OPEC scenarios explains the main part of the difference between earlier and later
world oil supply scenarios. The following chapters reviews the scenarios for a few key OPEC members,
and examines if the revisions to these match the changes to the total OPEC scenarios.
3.3.1.1
Saudi Arabia
Saudi Arabia production scenarios
20
18
16
14
Mb/d
12
10
8
6
4
2
0
1970
1980
1990
WEO2005
WEO2008
WEO2011
BP Statistical review
2000
WEO2006
WEO2009
WEO2012
2010
2020
2030
WEO2007
WEO2010
WEO2013
Figure 9: Production scenarios for Saudi Arabia compared to actual production from BP (BP, 2013). Note that the WEO2009
scenario does not include NGL.
As stated above, Saudi Arabia is by far the biggest producer of all OPEC members and currently in the
world. It holds the world largest reserves of conventional crude oil, at around 265 Gb, and second only
17
to Venezuela if unconventional oil is included (BP, 2013). Approximately half of Saudi Arabia’s crude
production comes from a single field, the super-giant13 Ghawar, by far the world’s largest field with
initial recoverable resources of 140 Gb. The field is in a very slow decline of 0.3% per year since its
production peak of 5.6 Mb/d in 1980. In 2007, production from Ghawar still amounted to 5.1 Mb/d
(IEA, 2008). Saudi Arabia is one of the few countries in the world with any significant spare capacity,
and as such acts as the main ‘swing producer’ of OPEC. Hence when global supply drops, Saudi Arabia
can step in to meet the rest of the demand, which can explain some of the smaller bumps in Saudi
Arabia’s production history in Figure 9. Recent reports suggest however that even Saudi Arabia is
currently pressed for spare capacity (Benes et al., 2012).
Figure 9 shows the production scenarios from all Outlooks that has featured Saudi Arabia individually.
As with the production scenarios for OPEC as a whole, the outlook for Saudi Arabia drops significantly
somewhere between WEO2008 and 2010 (the 2009 scenarios lacks NGL). The earlier high expectations
can be attributed to the call-on-OPEC assumption, while more recent Outlooks adhere to Saudi stated
production policies of maintaining crude oil production capacity of 12.5 Mb/d with a spare capacity of
some 1.5-2 Mb/d (IEA, 2013b). In WEO2012-2013 the outlook for Saudi production is by far the lowest,
even decreasing within the first half of the scenario. This is based on the IEAs assumption that OPEC
will rein in its production in the light of the growing contribution from North American light tight oil
(IEA, 2013b). Interestingly the Saudi production growth rate is by far the highest in WEO2008, which
featured a field-by-field analysis of world oil production, and significantly lower in subsequent reports.
While part of the revision is likely due to cancelled projects in the aftermath of the financial crisis, the
WEO2008 scenario still raises some questions about IEAs assumptions. Some 2.5 Mb/d of the increase
in the 2008 outlook for Saudi production comes from NGL (page 266 of WEO2008). The 2008 NGL
scenario as a whole was heavily criticized by Aleklett et al. (2010).
In WEO2011, Iraq overtakes Saudi Arabia as the largest contributor to projected growth contribution
for the scenario period, and in WEO2013 Saudi Arabia barely contributes anything to global production
growth. This likely owes in part to the upswing in Saudi production in the years since the great
recession and the tightening of spare capacity14 (Benes et al., 2012), but mostly to the IEAs changing
stance on Saudi Arabia’s stated production policies.
13
A giant oil field is defined by URR exceeding 0.5 Gb or annual production levels over 100 000 b/d (Simmons,
2002).
14
Some researchers have suggested that Saudi spare capacity is best viewed as theoretical due to uncertainties
(Büyükşahin et al., 2013).
18
3.3.1.2
Iraq
Iraq production scenarios
12
10
Mb/d
8
6
1980 IranIraq War
4
1990 First
Gulf War
2003 Second
Gulf War
2
0
1970
1980
1990
WEO2003
WEO2008
WEO2011
BP statistical review
2000
2010
WEO2005
WEO2009
WEO2012
WEO2012 Delayed case
2020
2030
WEO2006
WEO2010
WEO2013
WEO2012 High case
Figure 10: Production scenarios for Iraq compared to actual production from BP (BP, 2013). Note that WEO2006 and 2009
excludes NGL. The WEO2012 High case and delayed case are featured here to illustrate the uncertainty of the scenarios.
As evident by the actual production rate shown in Figure 10, Iraq’s oil production history has been
rocky. Several conflicts directly involving Iraq has led to severe drops in production rates, and while
these reductions are usually accompanied by a rapid rebound a few years later, Iraq has yet to recover
to its historical high point of 3,5 Mb/d by 1979 (BP, 2013). Iraq has a reported official oil reserve of
some 150 Gb, although the numbers are debated (see the above discussion of OPEC reserves). Such
uncertainties notwithstanding, Iraqi production potential is still hampered by security concerns and
underdeveloped infrastructure (Miller, 2011). Iraq’s importance is highlighted especially in later
Outlooks where it contributes more to production growth than any other country (earlier Outlooks has
Saudi Arabia as the largest contributor to global growth).
WEO2003, released a few months after the American invasion of Iraq, predicts a quick recovery in the
country’s oil production up to around 3.7 Mb/d by 2010 in the reference case, shown in Figure 10. It
also features both a slow and a rapid expansion scenario for Iraq, reaching 3.5 and 6 Mb/d respectively
by 2010, both well above the actual production of 2.5 Mb/d (BP, 2013). Although the 2003 Outlook
can be seen as overly optimistic of short term Iraqi growth, the later Outlooks matches the current
trend rather well. Every WEO featuring a separate scenario for Iraq predict a large and rather similar
increase in production, the lowest being WEO2010 and the highest WEO2003, with 6.1 and 8 Mb/d
respectively by 2030 (with WEO2010 reaching 7 Mb/d by 2035).
The IEA has in all of its scenarios for Iraq been careful to note the uncertainties stemming from the
political instability that has plagued Iraq for the last decade. A characteristic outlook for Iraq is usually
defined as on pages 274-275 of WEO2008, which claims that “The prospects for oil production in Iraq
remain very uncertain” and “Achieving this will hinge on political stability and providing opportunities
for foreign oil companies on satisfactory and fiscal terms” (IEA, 2008). By WEO2010, several
international oil companies had signed technical contracts with Iraq, which if all were achieved would
boost production capacity up to 12 Mb/d by 2017, the official target set by the Iraqi Ministry of Oil
(IEA, 2010a). Many experts has deemed these numbers to be unrealistic (Miller, 2011) and the target
19
has since then been reduced to 9 Mb/d, with further reductions in the order of millions of barrels per
day currently being negotiated (Kent, 2013).
The IEA have also been skeptical of the previous Iraqi target, and no WEO scenarios have come close
to predicting 12 Mb/d by 2017 (the closest is the Iraq high case in WEO2012 with 9.2 Mb/d by 2020,
see Figure 10). On page 426 of the special section on Iraq in WEO2012 however, IEA claims that
“…the Ministry of Oil has solid grounds for the formal position that, well before the end of the
decade, Iraq should have more than 12 Mb/d of production capacity from fields covered by technical
contracts alone” (IEA, 2012b). IEA also acknowledges that there exists a large number of constraints
on such an expansion, and predict that the actual growth will take on a more moderate pace. The
2013 scenario is slightly lower, “as a result of slow progress on the ground” (page 490 of WEO2013),
but still requires production to almost double by 2020 (from 3 to 5.8 Mb/d).
In conclusion, Iraq’s role in the WEO scenarios is large (in WEO2013 Iraq produces around 8% of all
oil by 2035), but comes with great uncertainties as acknowledged by IEA. The resources are thought
to be there and plentiful enough to sustain a rapid increase in production, although security concerns
and infrastructural challenges limit the pace. The WEO2013 prediction sees Iraqi production
increasing by almost 5 Mb/d during the scenario, but that will require political stability. Should more
unrest break out, it could severely hamper production which could have a great effect on IEAs
scenario. Recent reports on the Al Qaida resurgence in Iraq further suggest that the country might be
a long way off from political stability (Riedel, 2014). These uncertainties will be further investigated in
chapter 5.3.
3.3.1.3
Venezuela
Venezuela production scenarios
5
Mb/d
4
3
2
1
0
1970
1980
WEO2006
WEO2010
WEO2013
WEO2010 (EH)
WEO 2013 (EH)
1990
2000
2010
WEO2008
WEO2011
BP statistical review
WEO2011 (EH)
Extra-heavy actual
2020
2030
WEO2009
WEO2012
WEO2002 (EH)
WEO2012 (EH)
Figure 11: Production scenarios for Venezuela. Scenarios marked EH are for extra-heavy oil only. Note that WEO2006
excludes NGLs. Actual extra-heavy oil production is taken from the IEA scenarios
According to the EIA, Venezuela is the eight largest exporter of oil and its derivatives in the world, and
the largest of all in the western hemisphere. The country also holds the largest reported reserves in
the world, mainly in the form of extra-heavy oil (API<10°) found in the Orinoco belt. Since 1997,
Venezuela’s exports has reportedly dropped by some 50%, likely due to a combination of declining
conventional oil production, limited refining capacity, increased domestic demand and decreasing
20
interest from the U.S. to import (EIA, 2013a). While Venezuela’s somewhat shaky production history
is mainly the effect of above ground events such as OPEC production restrictions during the 1980’s and
worker strikes, most analysts including the IEA, believe Venezuelan conventional oil to be in an
irreversible decline. To mitigate this the IEA foresees the necessity of a rapid expansion of production
capacity of the country’s extra-heavy oil resources (IEA, 2013b).
As seen in Figure 11, the IEAs scenario for Venezuela has not changed greatly between the Outlooks,
although WEO2006 stands out slightly with higher production and faster rate of expansion. In all the
scenarios increased extra-heavy production is more than capable of offsetting the decline in
conventional oil, although the latest Outlooks has Venezuela in a plateau. There are some uncertainties
regarding actual production from Venezuela, which can explain the poor matchup between the
scenario starting values and BPs actual production data. For example, WEO2006 claims actual
production from Venezuela to be 2.2 Mb/d in 2005, in contrast to 3.3 Mb/d reported by BP. Extraheavy oil production is continually expected to rise, although it appears that the expansion is slower
than the IEA has anticipated. While Venezuela’s own production goals are set to reach 6 Mb/d by 2019,
recent developments such as oil companies dropping their contracts (Grant and Chazan, 2013), has
many believing that those goals will not be achieved. The IEA also sees problems in regards to
investments, refinery capacity and stability, and has kept its scenarios lower than the official goals.
Due to aforementioned problems with actual production data, this section is not featured as part of
the track-record analysis.
3.3.2 Russia
Russia production scenarios
12
10
Mb/d
8
6
4
2
0
1985
1995
WEO2000
WEO2005
WEO2008
WEO2011
BP Statistical Review
2005
2015
WEO2002
WEO2006
WEO2009
WEO2012
2025
2035
WEO2004
WEO2007
WEO2010
WEO2013
Figure 12: Oil and NGL production scenarios for Russia compared to historical production data from (BP, 2013). Note that
scenario data for WEO2006 and 2009 lack the inclusion of NGL.
Russia is one of the largest producers of oil in the world along with Saudi Arabia and USA, with a
production of some 10.6 Mb/d as of 2012. Periodically, Russia’s been the de facto largest oil producer
in the world and is currently second only to the U.S. in natural gas production (BP, 2013). The fall of
the Soviet Union (formally on the 26th December 1991) had a large impact on Russia’s economy and
the impact was felt in the country’s oil industry. As Figure 12 shows, Russian production started to
21
rebound around the turn of the century. Analyzing the data shown in Figure 12, the IEA predictions for
Russia can be divided into three stages. The first stage is featured in the early Outlooks from 20002003, in which IEA recognizes the increase in Russian production, but underestimates the speed at
which the recovery occurs. In the second stage of scenarios (WEO2004-2007), IEA successfully
identifies both the speed and size of recovery, and projects a continuous albeit slow rise after reaching
production levels over 10 Mb/d by 2010.
The third stage of scenarios (2008-2013) differs from earlier projections in that IEA predicts a peak in
production, usually around 2015 at slightly over 10 Mb/d. As seen in Figure 12, the projection from
WEO2009 also follow this pattern but are much lower in volume. This is likely because of the fact that
NGL is not reported as part of Russian production in this Outlook. The peak first occurs as a result of
the field-by-field study conducted in WEO2008, and is attributed greatly to the decline in old giant
fields in Western Siberia and the Volga-Urals. WEO2008 and later Outlooks argue that while Russia still
have large reserves of untapped oil, fields yet to be discovered or found are “often located in remote,
northerly areas” (IEA, 2012b), or fields of much smaller size. IEA believes that development of these
fields will slow the overall decline in Russian production, but will not be enough to offset it. It is also
carefully noted in most Outlooks that maintaining production even at current levels require heavy
investment, exemplified by this quote from page 499 of WEO2010
“…in 2009 Russia became the world’s leading oil producer, at 10.2 Mb/d. Many of its largest fields are
in decline, however, and sustained development of Russia’s oil and gas resources depends on very largescale investment.” (IEA, 2010a).
Whether a peak in Russian production will actually occur in the near future - as predicted by IEA in
later Outlooks - remains to be seen. Judging from the track-record of Russian scenarios however, and
with the exception of the earliest Outlooks, IEA seem to be doing well in this category.
3.3.3 North America
North America production scenarios
20
Mb/d
15
10
5
0
1970
1980
WEO2002
WEO2006
WEO2009
WEO2012
Crude oil
1990
2000
WEO2004
WEO2007
WEO2010
WEO2013
+ NGL
2010
2020
2030
WEO2005
WEO2008
WEO2011
BP statistical review
+ Unconventionals
Figure 13: Production scenarios for North America compared to actual production from BP (BP, 2013). Note that some
scenarios include only Crude oil, some Crude + NGL, and other Crude + NGL + Unconventional
22
In this definition, North America comprises the U.S., Canada and Mexico. While each country has
unique conditions in terms of production and reserves, they are most consistently featured as a group
in the Outlook reports, and the scenarios are therefore presented here as such. Scenarios specific for
unconventional oil production from this region, such as Canadian oil sand or LTO (mainly from the U.S.)
are featured as part of the section on scenarios by type of oil (chapter 3.4). As seen in Figure 13, the
IEA practice of reporting the scenarios has changed between the Outlooks, making comparisons
difficult. The earliest Outlooks presents conventional oil (crude + NGL) scenarios, all who peak and then
decline and the WEO2008 scenario exhibits the same pattern if U.S. and Canadian unconventional oil
is excluded. The 2006 and 2009 scenarios only feature conventional crude oil, and show the same
patterns of expected decline. Indeed, in all scenarios for North America, conventional oil, and
especially crude oil, is continuing to decline. In later Outlooks this is remedied by the contribution of
unconventional oil, mainly oil sands from Canada and in the 2012 and 2013 Outlooks, the emergence
of LTO.
The WEO2008 scenario for North America stands out as by far the largest prediction before the LTO
boom in the U.S. This is likely explained by the fact that WEO2008 features the largest production
scenarios for both Canadian oil sands and NGL of any Outlook (see chapters 3.4.2.1 and 3.4.3
respectively). The WEO2012-2013 scenarios are by far the largest, owing to the rapidly increased
production from LTO and the expectation that the trend will continue. In conclusion, the scenarios for
North America depend mainly on unconventional oils and NGL, the prospects of which are all discussed
in separate subchapters in the scenarios by type of oil chapter (3.4).
3.3.4 Brazil
Brazil production scenarios
7
6
Mb/d
5
4
3
2
1
0
1980
1990
2000
WEO2000
WEO2005
WEO2008
WEO2011
BP statistical review
2010
WEO2002
WEO2006
WEO2009
WEO2012
2020
WEO2004
WEO2007
WEO2010
WEO2013
2030
Figure 14: Production scenarios for Brazil compared to historical production data from (BP, 2013). WEO2008-2009 does not
include unconventional oil.
Forecasts for Brazilian oil production have been featured in every Outlook analyzed in this project, all
of them predicting increasing production levels throughout the scenario. The one exception is the
WEO2008 scenario which seems to peak somewhere around 2015 (the report only features values for
2015 and 2030). This may be explained by the exclusion of unconventional oil the 2008 and 2009
scenarios for Brazil, although WEO2008 predicts a faster rate of growth than WEO2009.
23
As seen in Figure 14, actual oil production in Brazil has grown steadily since 1980, the pace increasing
somewhere in the mid-90s according to BP (2013). In 2006 Brazil became self-sufficient in terms of oil
on a net-basis (IEA, 2006). Every Outlook so far has slightly overestimated the rate of growth for Brazil,
although the difference is rather small. The largest differences occurs in WEO2004 and WEO2005, both
predicting 2.5 Mb/d in 2010 with actual production landing at 2.1 Mb/d. WEO2008 shows a greater
difference, although all points before 2015 are interpolated from which no conclusions are drawn.
More recent WEOs feature a much greater increase in Brazilian production, owing in large to increased
prospects of pre-salt oil (named so because they predate the formation of thick salt layers (IEA,
2012b)). EIA characterizes pre-salt oil as “oil reserves situated exceptionally deep under thick layers of
rock and salt and requiring substantial investment to extract” (EIA, 2013b). The first pre-salt find, the
Tupi field (renamed Lula in 2010 (Riveras and Luna, 2010)) in the Santos basin, was made in 2007 and
first discussed in WEO2008. With Lula located in 2000 meters of water and over 4000 meters below
the sea bed, IEA saw considerable technological challenges with production (IEA, 2008). By WEO2010,
several other pre-salt fields had been found and the IEA increased its scenarios by several millions
barrels, in line with the Brazilian national oil company Petrobras’15 stated production goals (IEA, 2011).
WEO2013 features the largest scenario for Brazil as of yet, reaching 6 Mb/d by 2035, with almost all of
the growth coming from the Santos basin. Petrobras has set even higher ambitions, aiming to increase
production to 5.3 Mb/d by 2020, but the IEA justifies its lower scenario by claiming that Petrobras’
plans leaves no room for any slippage (IEA, 2013b). Recent reports suggest that Petrobras is finding
success in some pre-salt fields but are encountering difficulties in others (Bloomberg, 2013), which
suggests that IEAs approach is reasonable. Saraiva et al. (2014) uses a multi-Hubbert model to predict
future Brazilian oil production, based on different estimations of the URR (P5, P50 and P95). If the presalt finds are included, the study found that the P50 multi-Hubbert peaks at 5.4 Mb/d by 2034,
approximately in the same magnitude as the WEO2013 prediction. If pre-salt finds were ignored the
P50 peak occurred at 3.3 Mb/d by 2022, well in line with the earlier Outlooks. However, the results are
noted to be uncertain (Saraiva et al., 2014).
In summary, the IEAs predictions for Brazil has shown a tendency to be slightly optimistic, although
differences between actual and predicted production has been rather small. The current scenario puts
Brazilian oil production in rapid expansion, although a bit lower than Petrobras stated goals. There are
some uncertainties whether Brazil can achieve the production levels set by either Petrobras or the IEA,
especially as the goals require current production to almost triple and the pre-salt fields are still new
and untested at large scale.
15
In early 2011, Petrobras was responsible for over 92% of Brazil’s oil and gas production (de Oliveira, 2011)
24
3.4 Review of production scenarios by type of oil
WEO2008
WEO2004
100
100
80
80
Mb/d
120
Mb/d
120
60
60
40
40
20
20
0
2000
2005
2010
Currently producing
2015
YTD
2020
YTF
EOR
2025
0
2000
2030
Unconventional
2005
2010
Currently producing
100
100
80
80
60
40
20
20
2010
Currently producing
2015
YTD
2020
YTF
NGL
2025
2030
0
2000
2035
Unconventional
2005
2010
Currently producing
100
100
80
80
60
40
20
20
Currently Producing
2010
YTD
2015
YTF
2020
NGL
2025
2030
Unconventional
2015
YTD
2020
YTF
NGL
2025
2030
2035
Unconventional
60
40
2005
NGL
WEO2013
120
Mb/d
Mb/d
WEO2012
120
0
2000
2020
EOR
60
40
2005
YTF
WEO2011
120
Mb/d
Mb/d
WEO2010
120
0
2000
2015
YTD
2025
Other Unconv
2030
0
2000
2035
Light tight
2005
Currently Producing
2010
YTD
YTF
2015
EOR
NGL
2020
2025
Other Unconventionals
2030
2035
Light Tight Oil
Figure 15: IEA scenarios by type of oil. WEO2004-2011 has been digitized. WEO2012 is digitized and then interpolated.
WEO2013 is interpolated from table 14.4 in the 2013 Outlook.
Figure 15 shows a more detailed version of the world oil production scenarios, with a division for the
different types of oil into Conventional crude oil, NGLs and Unconventional oil, and dividing the crude
oil into four subcategories:
-
Fields currently producing
Fields yet to be developed (YTD)
Fields yet to be found (YTF)
Enhanced oil recovery (EOR), only a separate category in 2004, 2008 and 2013 and otherwise
likely divided between the other conventional crude categories.
25
Unconventional oil can also be further divided into subcategories, as described in chapter 3.4.2. The
more detailed description of the scenarios is mainly a result of the special focus on oil and the fieldby-field decline rate analysis in WEO2008. This analysis is used for conventional crude oil in WEO20102012, while WEO2013 featured a new decline rate analysis. The 2004 Outlook features a similar
division, although the specifics are hardly mentioned at all in the report. NGL is likely included as part
of the crude oil categories, in part explaining the differing looks compared to the other Outlooks. With
the exception of the EOR chapter, the 2004 projection is not analyzed in any great detail. The other
scenarios by type of oil are more similar, although WEO2008 stands out as it predicts a much larger
production by 2030 (the WEO2010-2013 scenarios run to 2035 but still sees lower production). EOR is
a separate category only in three of the Outlooks while the effect of enhanced oil recovery in the other
reports is presumed to be included in the other crude oil categories.
The following chapters examine all the oil categories presented in Figure 15, as well as some of the
more important subcategories of unconventional oil. The results of this analysis is used in the design
of the Alternative scenario (chapter 5.1) and the long term extrapolation of the scenarios (chapter 5.2).
Some of the categories are also featured as part of the track-record review (chapter 4), although such
an analysis is not applicable on the conventional crude oil subcategories, as historical data on
production from fields coming on stream in a specific year is seldom openly available.
3.4.1 Crude oil
Conventional crude oil scenarios
95
90
85
Mb/d
80
75
70
65
60
55
50
1990
2000
WEO2006
WEO2011
2010
WEO2008
WEO2012
2020
WEO2009
WEO2013
2030
WEO2010
IEA actual
Figure 16: Forecasts for crude oil and condensate. Note that extra heavy oil from Venezuela is included in WEO2006-2009
These chapters on conventional crude oil draws upon the methodology used by Aleklett et al. (2010)
and will serve as a comparison mainly between WEO2008 and WEO2013. The reason for this being
that these are the only Outlooks to feature separate decline rate analysis (scenarios by crude oil type
in WEO2009-2012 are based upon the study done in 2008 while WEO2013 is based on an update). The
2008 and 2013 Outlooks are also among the few to present viable URR data for fields YTD and YTF
which is a necessity for depletion rate analysis.
Figure 16 shows the scenarios for conventional crude oil for all Outlooks which separate conventional
crude and NGL. WEO2007 and the reports preceding WEO2006 only report scenarios for total
conventional oil (crude + NGL), which is why they are not featured here. Some condensate production
26
is included in the scenarios as some countries report it as part of the conventional crude oil production
instead of NGLs (see chapter 3.4.3 for more details). Figure 16 shows that previous expectations on
conventional crude has been unrealistically high, with WEO2006 predicting an increase from some 70
Mb/d to almost 90 Mb/d. Both the 2008 and 2009 Outlooks see a slower increase in production, while
the most recent reports have conventional crude oil in a slowly declining plateau phase until the end
of the scenario. A small part of the differences for the 2006-2009 Outlooks is the categorization of
Venezuelan extra heavy oil as conventional oil, although as seen in chapter 3.3.1.3 scenarios for
Venezuelan extra heavy has never surpassed 3 Mb/d. In the more recent Outlooks it is often stated by
the IEA that conventional crude oil peaked around 2005-2006, and has been in a plateau ever since.
Current IEA scenarios predict that the plateau phase will last throughout the scenario time frame, with
a slight decline (IEA, 2013b).
3.4.1.1
Currently producing fields
Scenarios for fields currently producing
80
70
60
Mb/d
50
40
30
20
10
0
1
6
11
16
21
26
Year from scenario start
WEO2004
WEO2011
WEO2008
WEO2012
WEO2010
WEO2013
Figure 17: Production scenario for fields currently producing. Note that all data has been digitized, with the exception of
WEO2013 which applies decline rates from page 468 of the 2013 Outlook.
Figure 17 shows the scenarios for oil from fields that were in production during each Outlook’s release
year, all moved to the same starting year for comparative purposes. There are some slight differences
in the shapes of the production curves, but with the exception of WEO2004 all scenarios are mostly
similar. Some small differences may be caused by the digitization process as well, which is more
evident in the decline rate analysis (Figure 18). WEO2004 likely includes NGL in the currently producing
category, explaining the larger historical production at the start of the scenario. The category includes
all fields in production, whether they are in a build-up phase, plateau or in post-peak decline. As
evident by Figure 17, the fields in the pre-plateau phase are not enough to outweigh the decline in
fields that have already peaked.
27
Decline rates for fields currently producing
9
8
7
6
%
5
4
3
2
1
0
1
6
11
16
21
26
Year from scenario start
WEO2004
WEO2008
WEO2010
WEO2011
WEO2012
WEO2013
Figure 18: Decline rates for fields currently in production for each Outlook. Production data for WEO2012-2013 is linearized,
explaining the smoother decline rates
Figure 18 shows the calculated decline rates for each Outlook’s scenario. The smoother nature of the
WEO2012-2013 decline rates owes to linear interpolation between given production values. The shaky
look of the other Outlooks owes in part to the digitization process used to attain the production values,
as calculating the decline rates amplifies these errors. The sharp drop in the decline rate for the two
last years of WEO2008 may be due to increased reliance on giant fields with slower decline, which
would be an reasonable assumption (Jakobsson et al., 2014). The general trend in most outlooks is
towards increasing decline rates by the end of the scenarios, the main exception being the more stable
decline rate in WEO2013. Table 3 presents the average decline rates for the entire scenario period, for
a better comparison.
Table 3: Calculated average aggregate decline rates of conventional crude oil fields for the Outlooks
WEO2004
Average aggregate
Decline rate [%]
-6.1
WEO2008
WEO2010
-4.1
-5.3
WEO2011
-4.5
WEO2012
-3.9
WEO2013
-4
Most of the scenarios has an average decline rate of 4-5% per year, again with the exception of
WEO2004 which likely includes NGL, making the comparison difficult. Sorrell et al. (2012) found the
average aggregate decline rate of all fields currently in production to be at least 4%, with an upward
trend. This is well in line with most Outlooks, and also with the results from the analysis by Aleklett et
al. (2010), that found the WEO2008 decline rate and the currently producing category to be
reasonable. Jackson and Smith (2014) estimates a slightly lower aggregate decline rate of 3.7% until
2040, which still does not differ greatly from that of the two latest Outlooks. The decline rate in
WEO2013 increases early in the scenario as more fields enters a plateau or post-peak decline, and then
slowly decreases again to a 4% annual decline. The lower decline in the later parts of the scenario is
explained by the IEA as owing to production in 2035 coming mainly from onshore giant fields, known
to decline slower (IEA, 2013b). This is in line with the findings of (Höök et al., 2009a). The average
28
decline rate in WEO2013 is only slightly lower than that of WEO2008, and the scenario can be
considered to be fully reasonable.
3.4.1.2
Yet to be developed
Production scenarios for fields YTD
60
50
Mb/d
40
30
20
10
0
1
6
11
16
21
26
Years from scenario start
WEO2004
WEO2011
WEO2008
WEO2012
WEO2010
WEO2013
Figure 19: Production scenarios for fields YTD. Note that all graphs with the exception of WEO2013 have been digitized, and
that WEO2012-2013 are based on point values and linear interpolation
Fields yet to be developed (YTD) is a term that includes oil fields that are either planned, assumed or
currently being developed, and have already been discovered by the beginning of the scenario. Figure
19 shows IEAs scenarios for YTD production. For comparative purposes all scenarios start at the same
year, since it is the rate of growth that is most interesting to compare. As previously mentioned, the
WEO2004 scenario probably includes NGL as part of the conventional category, which can explain part,
but certainly not all, of the large difference compared to the other Outlooks. The shape of the
WEO2008 scenario also differs from the more recent Outlooks, owing to rapid expansion and limited
URR. WEO2010-2012 are all rather similar, while WEO2013 sees YTD production grow slightly faster in
the beginning of its scenario and later slowing down. WEO2008 and especially WEO2004 should likely
be viewed as somewhat extreme cases.
To assess whether the YTD scenarios are realistic, this report utilizes depletion rate analysis (see
chapter 2.4 or Höök (2014) for detailed terminology). For such an analysis to be possible, knowledge
of the total URR of the YTD category is needed. Of the Outlook reports that feature this division of
conventional crude oil, only WEO2008 and WEO2013 present estimations of the YTD URR. Also, since
the IEA rarely discuss this category in any detail, it is unclear exactly which fields are included in the
YTD URR, and when they have been discovered. This is of essence since techniques for the estimation
of field URR has grown more advanced and accurate in in more recent years, perhaps indicating that
the reserve growth of more recently discovered fields could be smaller than historically observed
growth (Robelius, 2007; Schindler and Zittel, 2008). There are also many undeveloped fields discovered
in the 1970s, where the URR is likely highly uncertain and the oil in such ‘fallow fields’ have at least
previously been regarded as uneconomical (Miller, 2011).
29
On page 257 of WEO2008 the IEA presents world conventional oil reserves in fields YTD as 257 Gb, and
then further divides these reserves as belonging to OPEC/non-OPEC countries and onshore/offshore.
The YTD scenario is then presented in three pages. In WEO2013, YTD is rarely discussed at all, and only
presented in table 14.4 on page 471. Since there is no discussion about YTD production in the 2013
Outlook, the IEA do not directly state what URR is used for the scenario. However, figure 13.5 on page
428 presents estimated conventional crude oil resources by field size, divided into fields producing,
under development, discovered and undiscovered. Digitizing the figure gives 39.5 Gb of URR for fields
under development and 197.3 Gb for discovered fields, which put together results in a URR of 236.8
Gb for the YTD category. While these numbers are not specifically said to be used by the IEA, they
closely match the size of the YTD URR from WEO2008, and will be used for the depletion rate analysis
in Figure 20.
YTD depletion rates compared to the North Sea
12
10
%
8
6
4
2
0
1
6
11
16
21
Year from scenario start
Global (WEO2008)
Global (WEO2013)
North Sea
Linear (North Sea)
Figure 20: RRR depletion rates for fields yet-to-be-developed compared to the depletion of the North Sea region. URR: 257 Gb
for WEO2008, 236.8 Gb for WEO2013. The URR of the North Sea region is approximately 75 Gb (Aleklett et al., 2010)
Figure 20 shows the depletion rate of WEO2008 and 2013 based on the aforementioned URRs,
compared to that of the North Sea which is the fastest developed region in history. While the North
Sea depletion rate is by no means a theoretical maximum, it should be argued that any scenario whose
development pace greatly exceeds that of the fastest developed region in history should be properly
justified (Sorrell et al., 2009). Due to the more detailed data presented by the IEA in the 2008 Outlook,
Aleklett et al. (2010) could analyze the depletion rates of OPEC and non-OPEC fields and
onshore/offshore separately, and found the YTD scenario to be unrealistic and dependent on an
unreasonably high depletion rate. The WEO2008 YTD production of 22.5 Mb/d by 2030 was reduced
to 13.5 Mb/d, considered to be more realistic by Aleklett et al. (2010). As WEO2013 features no such
division of the URR data, the global depletion rate for both the 2008 and 2013 Outlook has been
calculated for comparative purposes. From Figure 20 it is evident that the WEO2008 scenario
significantly exceeds the depletion rate of the North Sea, while the WEO2013 features a depletion rate
only slightly faster than that of the North Sea. As previously mentioned, the North Sea depletion is not
a theoretical maximum and the WEO2013 scenario is therefore viewed as realistic, although both
rather optimistic and heavily dependent on rapid development and investments.
30
YTD production from WEO2013 and adjusted to North Sea
depletion
25
20
Mb/d
15
10
5
0
2012
2017
WEO2013
2022
Adjusted exact
2027
2032
Adjusted trend
Figure 21: Forecasted production from fields YTD in WEO2013, compared to production adjusted to the North Sea's exact
depletion rate and the North Sea trend depletion rate
Figure 21 compares the WEO2013 scenario to the production that comes from applying the North Sea
depletion rate to the YTD URR of 236.8 Gb. Recalculating the production to match the North Sea
depletion gives a slower expansion than the WEO2013 scenario, but reaches the same production
levels by the end of the scenario, further showing that the IEA scenario is at least somewhat realistic.
There are also some uncertainties regarding the URR, and whether or not it allows for any reserve
growth (if not, actual URR could be higher, which is analyzed in the sensitivity analysis in chapter 5.3).
If instead the North Sea depletion rate trend is applied the result is a slower development pace and a
smaller production of 16 Mb/d by 2035 (compared to 19.8 Mb/d from WEO2013). This result is used
in the Alternative scenario in chapter 5.1.
Despite a larger global YTD URR in WEO2008, Aleklett et al. (2010) arrives at a lower production by the
end of the scenario than this study (13.5 Mb/d compared to 16 Mb/d). This is in part due to the more
detailed description of the URR in WEO2008, which allowed for depletion rate analysis of OPEC and
non-OPEC onshore and offshore fields respectively. The analysis of WEO2013 is limited to a global
depletion rate, which encompasses all fields, and compares that an offshore region, which are
generally developed and depleted much faster than onshore regions (Höök et al., 2009b).
In conclusion, it is observed that earlier WEOs have had unrealistic expectations on the development
pace of fields YTD. WEO2013 lies closer to empirical depletion rates seen in history, and the YTD
scenario is accepted, although still perceived as optimistic.
31
3.4.1.3
Yet to be found
YTF production scenarios
25
Mb/d
20
15
10
5
0
1
6
11
16
21
26
Years from start of scenario
WEO2004
WEO2011
WEO2008
WEO2012
WEO2010
WEO2013
Figure 22: Production scenarios for field YTF. Note that all graphs with the exception of WEO2013 have been digitized, and
that WEO2012-2013 are based on point values and linear interpolation
The term yet to be found (YTF) corresponds to oil fields that have as of yet not been found, but are
expected to be discovered within the scenario time frame. Figure 22 presents projected YTF production
for the Outlooks, again adjusted to the same starting year for comparison. Unlike the other crude oil
categories, the WEO2004 scenario does not stand out. Instead, most scenarios are rather similar with
WEO2012 predicting the lowest production levels. This is compensated by a larger contribution from
fields currently producing (see Figure 17). In WEO2008 the production from fields YTF grows slowly in
the beginning of the scenario but expands much faster towards the end.
As with the YTD category, depletion rate analysis is used to determine the plausibility of the scenarios.
As the YTF category consists of fields that still undiscovered, the YTF URR corresponds to the expected
discoveries within the scenario. There of course exists large uncertainties as to how much oil can be
found and where it would be located, but estimations of remaining undiscovered resources and
investments into exploration can give a hint on future discoveries. Some authors promote curve-fitting
to historical discoveries, although others criticize such techniques for being unreliable (Sorrell and
Speirs, 2014).
The YTF production scenario in WEO2008 is based on projected worldwide discoveries of 114 Gb, which
was found to be realistic by Aleklett et al. (2010). WEO2013 predicts larger discoveries within the
scenario time frame of 170 Gb, citing a stabilized discovery rate for the last decade and that the USGS
assessment of undiscovered resources has P95 values of 250 Gb. No issue is taken with the predicted
170 Gb of oil discoveries. While a similar curve-fitting of the same oil discovery data used by Aleklett
et al. (2010) yields only 113 Gb for the period 2013-203016, the possibility of a stabilized discovery rate
on account on increased exploration spending and higher oil prices, as claimed by the IEA, must be
acknowledged. However, the choice to validate the discoveries based on the USGS P95 data is
questionable. Firstly, the P95 resources in the updated USGS 2012 assessment of undiscovered
16
The typical lead time between discovery and first production is about 5 years for giant oil fields (Höök et al.,
2009b). This indicates that any oil field expected to be in production by 2035 should be discovered in 2030 at
latest.
32
resources for all regions except the U.S. only adds up to 199 Gb (Schenk et al., 2012) or 226 Gb if the
mean values for the U.S. is added (USGS, 2013). This discrepancy is possibly explained by the
incorporation of USGS data into the IEAs own database (IEA, 2013a). Secondly, the P95 data from the
USGS 2012 assessment does not indicate in any way that the oil will be found within the scenario time
frame, only that it is likely to be discovered at some point in time17. The 250 Gb stated by the IEA will
however serve as an upper limit of discoveries within the scenario in the sensitivity analysis (chapter
5.3), with 113 Gb as the lower limit.
YTF depletion rates compared to the North Sea
12
10
%
8
6
4
2
0
1
6
11
16
21
Years from scenario start
Global (WEO2008)
Global (WEO2013)
North Sea
Adjusted depletion
Figure 23: RRR depletion rates for fields yet-to-be-found compared to the depletion of the North Sea region. Projected
discoveries (URR): 114 Gb for WEO2008, 170 Gb for WEO2013. The URR of the North Sea region is approximately 75 Gb
(Aleklett et al., 2010)
Figure 23 shows the depletion rates of WEO2008 and WEO2013 based on the predicted discoveries,
and compared to that of the North Sea. There is some lead time from the scenario start to first
production from fields YTF, which also explains the lower depletion rate at the end of the scenario for
the North Sea, compared to Figure 20. Both the WEO2008 and 2013 depletion rate starts out very
reasonably compared to the North Sea, but in mid/late scenario the development pace increases. The
WEO2008 depletion rate increases far faster than both that of WEO2013 and the North Sea, and was
like the YTD scenario deemed unrealistic by Aleklett et al. (2010). The lower projected production
coupled with higher expectations on discoveries in WEO2013 lead to a more moderate depletion rate,
even though it still exceeds that of the North Sea. The YTF scenario is therefore seen the as realistic,
but somewhat optimistic. As with YTD however, the YTF production is adjusted to the North Sea trend
depletion, although the depletion rate in the first few years of the scenario is accepted (see the
“Adjusted depletion” in Figure 23).
17
The USGS 2012 updated assessment of worldwide undiscovered resources do not mention any time frame for
the oil to be discovered within (Schenk et al., 2012). However, the USGS 2000 assessment cites that the resources
should be found between 1995 and 2025 (Schindler and Zittel, 2008) . It is unclear what data the IEA are using as
their numbers do not match any of the reports, and they cite a combination of USGS data and their own database
(IEA, 2013a).
33
3.4.1.4 Enhanced oil recovery
As finding new oil fields (especially the sought after giants) is becoming increasingly difficult, it is
generally seen as necessary to increase the amount of oil that can be extracted from fields already in
production, i.e. increase the recovery factor. Simply drilling a well and utilizing natural pressure
(primary recovery) yields around 10% of the oil originally in place (OOIP) (BP, 2012). The majority of
the world’s producing oil fields also employ secondary recovery techniques, such as increasing the
pressure by injecting water, to increase the recovery factor. Primary and secondary recovery can
generally achieve recovery rates of 20-50%, depending on field characteristics. Tertiary recovery or
enhanced oil recovery (‘EOR’) is an umbrella term for techniques aimed at increasing the recovery
factor beyond that. In WEO2008, IEA defines the three main EOR techniques as chemical flooding,
miscible displacement (using for example nitrogen or CO2), and thermal recovery. Actual production
as a direct result of the application of EOR techniques is difficult to assess. Some claim that additional
recovery attributed to EOR amounted to approximately 2.5 Mb/d in 2008 (IEA, 2008; Sorrell et al.,
2009). In contrast, IEA states in WEO2013 that production contribution from EOR has remained steady
at around 1,3 Mb/d for the last decade (IEA, 2013b), while Al Adasani and Bai (2011) cite 1.8 Mb/d in
2010 based on a database on 652 EOR projects worldwide. This may be caused by differing definitions
or databases. It would seem certain that EOR so far has contributed only a small part of daily world
production.
30
140
25
120
100
Mb/d
20
80
15
60
10
40
5
20
0
0
EOR production by 2030 (left axis)
Times mentioned in report
EOR production by 2030
Word count (right axis)
Figure 24: Number of times EOR is mentioned in the WEO reports compared to production outlook for 2030. Occurrences
include a word search for “EOR”, “Enhanced oil” and “Enhanced recovery”. Production data for WEO2004 and 2008 has been
digitized and the exact numbers are slightly uncertain
As evident by Figure 24, scenarios for additional recovery by EOR is generally not included in the WEO
reports and rarely discussed in great detail, with three major exceptions: the 2004, 2008 and 2013
Outlooks. According to the 2004 Outlook, added production from utilization of EOR will amount to
almost 25 Mb/d by 2030 (see Figure 15), over 20% of the global production (118.3 Mb/d). To justify
production of such magnitude, one would expect a detailed description of prospects for EOR and the
impact it might have on the cost of production. Instead enhanced oil recovery (and any form or
abbreviation of the term) only occurs four times in the entire report, one of which is in figure 3.20 and
another one is in the appendix. The only times EOR, in any form, is mentioned in text is on page 412,
in regard to carbon capture and storage:
34
“The use of CO2 in enhanced oil- and gas-recovery techniques could in certain situations offset part of
the capture cost. //…// The financing and construction of CO2 storage demonstration projects, including
enhanced oil recovery, with CO2 storage to monitor long-term storage and possible leaks, should be
given highest priority.” (IEA, 2004).
Interestingly enough, reports adjacent to WEO2004 such as the 2005 Outlook, features some actual
examples of EOR-projects and prospects, while not providing a scenario. Of course, there is no direct
link between the number of times a phrase or word is mentioned and how correct the underlying
assumptions are. But any claim of massive production increases from a specified technique or region,
such as the case with EOR in WEO2004, should be properly justified. It cannot be said that the quotes
above from the 2004 Outlook provides such justification.
In stark contrast, WEO2008 produces a much lower scenario for additional production of 6.4 Mb/d by
EOR (a drop of over 70% from WEO2004), and uses several examples of current and planned EOR
projects to justify the production increase. The 2008 Outlook states that “EOR costs have risen sharply
in the last few years, in line with upstream costs generally” (IEA, 2008), which could be part of the
explanation for the downward revision. Since then, no WEO has featured any scenarios specifically for
EOR, although it is assumed to still be a part of the crude oil production. The significant drop in the
EOR production scenario between the 2004 and 2008 Outlooks could also possibly be explained by
looking at the production profile for the giant field Cantarell in Mexico. In 2000, Cantarell was
subjected to nitrogen injections which by 2004 had doubled the output, from about 1 to 2.1 Mb/d.
Production peaked in March 2006 and has been in rapid decline since, and by 2011 production had
fallen by 74% down to 0.45 mb/d (Rodriguez, 2012). It could be theorized that IEA analysts in 2004 saw
the rapid increase in Cantarell’s production a confirmation of the potential for EOR, although this
would imply serious shortsightedness and is not confirmed by the Outlook reports.
In WEO2013 the IEA projects only 2.2 Mb/d of EOR production by 2030, and 2.7 Mb/d by 2035, while
discussing it in greater detail than any other Outlook with the possible exception of WEO2008. They
also approximate that the current contribution from EOR projects is around 1.3 Mb/d and has been at
that level for the last decade. This is only around half of what is reported as actual production in
WEO2008 and by Sorrell et al. (2009). A possible explanation is that the WEO2013 EOR scenario
excludes fields that IEA classify as unconventional, such as Canadian oil sands and the Orinoco belt in
Venezuela, while it is not clear whether those fields were included in WEO2008. The IEA also cites the
limited data available on EOR projects outside of North America as a source of uncertainty (IEA, 2013b).
The use of EOR historically has had a rather limited effect on global production, which concerns IEA
insofar as they state that it “remains somewhat of a puzzle that these techniques have not yet made a
more substantial contribution to oil production” (IEA, 2013b). The effect and applicability of tertiary
recovery techniques varies with field geology and distribution, and lengthy and often expensive studies
will have to be done on each separate oil field before applying EOR techniques. The application of EOR
techniques can also take months or years before affecting the oil production rate, and may be
impossible to implement on more mature offshore fields due to lack of space (Muggeridge et al., 2014).
The IEA cites the large initial costs as one of the reasons why EOR sees so little use globally (IEA, 2013b).
It is also disputed as to what extent EOR techniques add to the recovery factor, and thus to the URR of
a field. For at least some fields and certain EOR techniques, only the production rate increases at the
cost of a steeper decline, while there is little effect on the fields URR (Gowdy and Juliá, 2007). The
Weyburn field in Canada, often cited in the Outlooks in regards to improved URR by EOR, can be seen
as a counter-example to this (Sorrell et al., 2009).
35
In conclusion however, IEAs predictions for EOR production in WEO2013 is seen as both properly
justified and reasonable, perhaps even a bit pessimistic. This portion of the scenario is accepted
without any alterations into the reevaluated model.
3.4.2 Unconventional oil
Unconventional oil production scenarios
16
14
12
Mb/d
10
8
6
4
2
0
1997
2003
WEO2000
WEO2005
WEO2009
WEO2013
2009
2015
WEO2002
WEO2006
WEO2010
WEO2012 (LTO)
2021
WEO2003
WEO2007
WEO2011
WEO2013 (LTO)
2027
2033
WEO2004
WEO2008
WEO2012
Figure 25: Forecasts for oil categorized as unconventional. Note that the IEA definition of unconventional oil changes
between the Outlooks (see Table 1). WEO2012 and 2013 are presented both with and without light tight oil (LTO)
Figure 25 shows the scenarios for oil categorized by IEA as unconventional for each Outlook. Since IEAs
definition of unconventional oil has changed repeatedly over the years (see Table 1), no actual
production data can be presented that would be relevant for comparative purposes. The WEO20122013 scenarios are shown both with and without LTO, seeing as its categorization as unconventional
oil in WEO2012 and rapid expansion has significantly altered the scenarios (see chapter 3.4.2.3 for
more details on LTO). Excluding LTO, the scenarios for total unconventional oil production has not
changed much between the Outlooks, even though there has been much re-categorization of different
types of oil. For example, between 2006 and 2009 large parts of the Venezuelan extra-heavy oil was
categorized as conventional oil (even though the unconventional scenarios remained much the same).
The total unconventional category does not say much about the actual scenario or IEAs beliefs
however, and thus the most prominent parts of the unconventional category has been examined in
greater detail. The following chapters analyze specific types of oil that the IEA classify as
unconventional, namely Canadian oil sands, gas- and coal-to-liquids and LTO. Venezuelan extra-heavy
oil is discussed in the conjunction with Venezuela as a whole (chapter 3.3.1.3). There are other sources
of unconventional oil, such as extra-heavy oil from the Neutral zone between Kuwait and Saudi Arabia,
but the analyzed areas provides more than 90% of all unconventional oil in IEAs scenarios. In common
for most unconventional oil is that it is generally more expensive to extract (requiring higher oil prices)
and often the subject of environmental concerns (Bolonkin et al., 2014). As these above-ground effects
can have large impacts on the production of unconventional oil, it is generally more difficult to find
empirical data relevant for future forecasting. Hence, no decline or depletion rate analysis can be
36
applied to these scenarios, and instead the review relies on other studies to investigate whether or not
the scenarios are reasonable.
3.4.2.1
Canadian oil sands
Canadian oil sands production scenarios
7
6
Mb/d
5
4
3
2
1
0
2000
2010
2020
2030
WEO2003
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
IEA actual
Figure 26: Forecasts for Canadian oil sands production, with actual production data from the WEO reports.
The Canadian oil sands is a vast resource located primarily in Alberta, Canada with some additional
resources located in Saskatchewan. The remaining established reserves of 169 billion barrels of crude
bitumen makes the Alberta oil sands the third largest deposit of oil in the world, behind Saudi Arabia
and Venezuela (Ouellette et al., 2014). Oil sands consist of sand, silt and clay along with crude bitumen,
which is a thick black tar-like substance that often contains significant amounts of sulfur and heavy
metals. The bitumen is both dense and very viscous18, and is deficient in hydrogen meaning it has to
be upgraded to synthetic crude oil (SCO) before it is acceptable for conventional refineries (Söderbergh
et al., 2007).
There are two primary ways of extracting the bitumen, mainly depending on the depth of the
resources. Bitumen close to the surface is often mined directly, while in situ extraction utilizing both
primary conventional techniques and EOR is used for resources more than 75 meters below the
ground. Mining is still the largest form of extraction, although in situ recovery is generally thought to
overtake the position in the future as around 80% of the established remaining reserves lie below 75
meters (CERI, 2013).
As evident by Figure 26, the IEA scenarios for oil sands are divided into three distinct patterns, differing
in anticipated growth rates and production levels. The earlier Outlooks (up to and including WEO2008)
all feature evidently optimistic growth rates and high production levels by the end of the scenario.
WEO2008 features the largest anticipated production of all WEOs, reaching 5.9 Mb/d by 2030. The
scenarios from later Outlooks (2009-2013) are all rather similar, with slower growth rates and lower
production levels.
18
Bitumen is often specified as having API<10° and a viscosity greater than 10 000 centipoise (Chew, 2014).
37
Exactly why the WEO2008 scenario is so much higher than all surrounding Outlooks is somewhat
unclear, although IEA cites a high oil price and a range of ongoing projects as justification for its
assumptions (IEA, 2008). The financial crisis between the 2008 and 2009 Outlooks explains the sudden
drop in anticipated production, following lower oil prices and many cancelled oil sand projects (IEA,
2009). But even as the oil price quickly returned to the previous high levels, the IEAs scenarios for oil
sands has remained the same. The WEO2008 scenario was also criticized by Aleklett et al. (2010), and
exceeded the crash programme scenario proposed by Söderbergh et al. (2007).
The Crash programme presented by Söderbergh et al. (2007), allows for 5,0 Mb/d of production by
2030, although it is considered difficult to reach. However some of the conditions assumed in the crash
programme has changed. Söderbergh et al. (2007) assumes that strain on natural gas supply will force
the construction of nuclear power plants to provide energy for upgrading the bitumen to SCO, and also
questions if the crash programme can be achieved without breaking Canada’s commitments to the
Kyoto protocol. The shale gas revolution in the U.S. has since then made cheap natural gas readily
available (Moryadee et al., 2014), and in 2011 the Canadian government officially abandoned its
obligations to the Kyoto treaty (Hu and Monroy, 2012). Raising production to 5.0 Mb/d by 2030 should
therefore not be completely unreasonable (Chew, 2014), provided that proper investments are made
and that the projects receive public acceptance.
The more recent Outlooks end up with production levels of between 4,2 and 4,5 Mb/d by 2035, with
WEO2013 reaching 4,3 Mb/d by the end of the scenario. Aleklett et al. (2010) revised the WEO2008
scenario to 3,9 Mb/d by 2030, citing this to more reasonable In comparison, the WEO2013 scenario
reaches 3,8 Mb/d in 2030, some 1.2 Mb/d below the Crash programme scenario. CERIs reference case
for bitumen production peaks at 6,3 Mb/d in 2034 (CERI, 2013). Although bitumen production is not
directly comparable to SCO because of losses in the upgrading process, it still gives some validation to
the WEO2013 scenario.
In conclusion, the WEO2013 scenario for Canadian oil sands is seen as reasonable and accepted into
the revised model as is.
3.4.2.2
Coal- and gas-to-liquids
Global GTL production scenarios
3
2,5
Mb/d
2
1,5
1
0,5
0
2000
2010
WEO2002
WEO2006
WEO2011
2020
WEO2003
WEO2008
WEO2012
Figure 27: Global GTL production scenarios.
38
WEO2004
WEO2009
WEO2013
2030
WEO2005
WEO2010
Global CTL production scenarios
1,4
1,2
Mb/d
1
0,8
0,6
0,4
0,2
0
2000
WEO2006
2010
WEO2008
2020
WEO2010
WEO2011
2030
WEO2012
WEO2013
Figure 28: Global CTL production scenarios.
Coal-to-liquids and gas-to-liquids (abbreviated ‘CTL’ and ‘GTL’ respectively) are techniques used to
convert coal and gas to synthetic liquid fuels and are proposed as a mitigation strategy for peak oil
(Höök et al., 2014b). Several established processes for both CTL and GTL have been commercially
proven and may be used as a substitute for petroleum-derived fuels. Current world CTL and GTL
capacity stands at around 0.4 Mb/d, or approximately 0.5% of world oil production (Höök et al., 2014b).
The most common process of liquefaction of coal and natural gas is the Fischer-Tropsch process (FT),
described in more detail by Wood et al. (2012). The techniques are generally thought of as useful for
freeing up stranded resources, such as gas that has no functioning infrastructure to transport it.
However, heavy capital investment needs, oil price volatility and environmental concerns have
resulted in a slow growth of the techniques.
Figure 27 and Figure 28 show IEAs scenarios for GTL and CTL, which both lack actual production for
comparison. This is due to the lack of data on the actual production from CTL and GTL plants, although
the combined capacity seems to be as previously mentioned around 0.4 Mb/d (Höök et al., 2014b).
Most of the GTL scenarios in Figure 27 have no production at the beginning of the scenario, although
it is implied in several of the related Outlooks that actual production is and has been around 0.2 Mb/d
for at least the latest decade. This is likely due to IEAs practice of rounding the numbers, but for the
sake of consistency the numbers derived from the Outlook tables are used.
The GTL scenarios show two distinct patterns. Outlooks predating WEO2008 predict a more rapid
growth in GTL production to around 2.3-2.4 Mb/d by 2030, while later Outlooks see GTL production at
around 0.7 Mb/d in 2030 and 0.9 Mb/d by 2035. Concerns regarding large upfront investment needs
and profitability of both GTL and CTL plants are frequently mentioned in WEO2005 and subsequent
Outlooks, and it is said that the GTL scenarios should be treated with caution (IEA, 2005). The CTL
scenarios have been more consistent, with most pointing to a rise from around 0.2 to 1.1-1.3 Mb/d.
In WEO2013 the combined production from CTL and GTL in the New Policies Scenario is 2.1 Mb/d in
2035. Although large investments need and the environmental concerns may halt the development of
39
future CTL and GTL plants, there is no scientific ground to object to the WEO2013 scenarios and they
are viewed as reasonable.
3.4.2.3
Light tight oil
Light tight oil production scenarios
7
6
Mb/d
5
4
3
2
1
0
2008
2012
2016
2020
WEO2012
2024
WEO2013
2028
2032
EIA
Figure 29: Light tight oil production scenarios. WEO2012 and 2013 are the only reports to feature separate scenarios for LTO.
Actual production data from EIA (2014, 2012) features only U.S. LTO
Light tight oil (‘LTO’) is found in tight rock formations with very low permeability, commonly extracted
with a combination of multi-stage hydraulic fracturing and horizontal drilling techniques (Fic and
Pedersen, 2013). The oil itself is light and sweet19, and while the IEA classifies LTO as unconventional
oil as of WEO2012 on account of the production techniques, others argue that it should be classified
as conventional crude on the basis of its API gravity (McGlade, 2012). The so called LTO boom in the
U.S. is still very new, with production expanding rapidly. Figure 29 shows the only two LTO production
scenarios available from the Outlooks, with actual production in the year 2012 far surpassing the
predictions of WEO2012. While the WEO2013 scenario represents a large upward revision from
WEO2012, the production is still expected to peak and start declining within the scenario time frame.
Compared to actual LTO production from the U.S., which currently constitutes almost all of the global
LTO production, the IEA have underestimated the rate of growth in LTO production.
WEO2013 features an investigation into the natural decline rates of LTO wells, i.e. the decline rate that
would occur if no investments were put in to expand the life cycle of the well. IEA finds that the
production from LTO wells increases rapidly the first year but then enters a sharp decline. It therefore
becomes necessary to continuously drill new wells to compensate for the decline in others, hence
maintaining a larger production requires massive continuous drilling. The IEA argues that the U.S. and
to some extent Canada are the only countries currently capable of maintaining large production levels,
due to existing infrastructure and a large fleet of drilling rigs. As such the WEO2013 LTO outlook sees
only a minor contribution from countries outside North America even by 2035 (IEA, 2013b). This is
supported by, for instance, the findings of Alquist and Guénette (2014). In comparison, the WEO2013
LTO outlook lies somewhere between the predictions of the EIA of around 5 Mb/d by 2030 and BP of
approximately 7.5 Mb/d by 2035 (BP, 2014; Sieminski, 2013).
19
Light meaning API > 10° (often much greater). Sweet means a low sulfur content.
40
The IEAs opinion is that the current LTO resource base (349 Gb) and the natural decline rates will limit
the effect to a regional scale, and that LTO will only put an eventual oil peak off by a few years. There
is no ground to object to this analysis, and find the IEAs reasoning to be sound. The emergence of LTO
in the U.S. and North America is also one of the main differences between the analysis of WEO2008 by
Aleklett et al. (2010) and this report. It is rather interesting however that despite the 5.8 Mb/d LTO
production by 2030, the WEO2013 global oil projection is still lower than that of WEO2008.
3.4.3 Natural gas liquids
NGL production scenarios (adjusted for energy content)
16
14
Mboe/d
12
10
8
6
4
2
0
1980
1990
WEO2006
WEO2010
WEO2013
2000
2010
WEO2008
WEO2011
EIA NGPL production
2020
2030
WEO2009
WEO2012
Figure 30: Production scenarios for NGL compared to actual production data from EIA (2013c). Note that the IEA NGL scenarios
have been adjusted for energy content.
Natural gas liquids, or NGL for short, are light hydrocarbons that are in either liquid form at normal
pressure and temperature (condensate) or can be relatively easily separated from natural gas and
turned to liquid under moderate pressures (Sorrell et al., 2009). As seen in Figure 30, NGL production
has increased steadily for the last 30 years and every Outlook that reports NGL scenarios assume a
continued growth, albeit at different rates. It is worth noting that the IEA prefers to report NGL
production in terms of volume instead of in oil equivalents. However the energy content of a barrel of
NGL is somewhere around 75% of one barrel of crude oil (Aleklett et al., 2010). Some researchers, such
as Aleklett et al. (2010), argue that the energy content of one barrel of NGL is more important than the
volume to replace one barrel of crude oil. Others claim that for the petrochemical industries, which in
the U.S. uses 55% of all NGLs (Troner, 2013), one barrel of NGL can replace one barrel of crude oil,
regardless of the energy content (McGlade, 2013). However, as the EIA reports historical NGL
production in barrels of oil equivalents, the WEO scenarios are adjusted to mboe/d, using 75% energy
content as a benchmark, the result being that the starting points of the scenarios match the actual
production (see Figure 30). In subsequent chapters NGL is presented in volume terms in line with IEA
practice, to simplify comparisons.
Figure 30 shows that WEO2008 predicts the largest increase in NGL production, up to 18.9 Mb/d (or
around 15 mboe/d) by 2030, followed by the 2009 Outlook. Outlooks from 2010 and onwards predict
a slower development and reaching lower levels even by 2035. The 2006 Outlooks, which is the only
report predating WEO2008 to feature a separate NGL scenario, stands as the low case. NGL production
is heavily coupled to the extraction of natural gas, which leads the IEA to base the growth of NGLs on
41
its natural gas production scenarios, rather than on a separate analysis (IEA, 2013b). Figure 31 shows
the historical production of natural gas and NGL, and also demonstrates that the fraction of NGL to
natural gas has remained around 14-16% for the last 40 years, with a slightly upwards trend.
Production [Mboe/d]
20
50
15
40
30
10
20
5
10
0
0
1973
1982
1991
Natural gas
NGL
2000
Fraction
Share of NGL to Natural Gas [%]
World NGL and Gas production 1973-2012
60
2009
Figure 31: Historical NGL production compared to natural gas production. Source: (BP, 2013; EIA, 2013c, 2008).
In The Peak of The Oil Age it is noted that the IEA in WEO2008 confirms the view that the NGL content
of natural gas is more or less constant. However, Aleklett et al. (2010) also observed that the 2008
scenario predicted a much larger increase in NGL production than in natural gas, and hence the IEA
contradicts their own statements. Figure 32 illustrates the forecasted increases in production of both
NGL and natural gas for several Outlooks. If the NGL content of the natural gas should remain constant,
as both the IEA and historical production suggests, the increases in percentage should be about the
same. It is seen that while the 2006 Outlook features a rather similar growth between NGL and natural
gas, this is heavily disrupted in WEO2008, where NGL increases by almost 90%, compared to the 50%
increase in natural gas production. Similar discrepancies can be seen for both the 2009 and 2010
Outlooks, even if the gaps in increase becomes smaller, and by WEO2011 the growth rates match once
more. No explanation to this discrepancy has been found in any of the Outlooks, and even while the
problem seems to be fixed by now, it remains unclear why it occurred in WEO2008 and why the error
remained for another two reports.
Production increase by end of scenario period
100
80
%
60
40
20
0
WEO2006
WEO2008
WEO2009
WEO2010
Natural gas
WEO2011
WEO2012
WEO2013
NGL
Figure 32: Increase in production of NGL compared to natural gas over the scenario period. Note that the starting value for
natural gas lags by one year (i.e. WEO2013 present actual production of NGL for 2012, but natural gas for 2011).
42
Part of the explanation to the discrepancies shown in Figure 32 may lie in the difference in reliance on
OPEC of the Outlook scenarios and the way that NGL is defined in different countries. Condensate, or
field condensate as it is also called, is the part of the natural gas that is liquid at normal temperatures
and pressure, and is thus extracted directly at the wellhead (Sorrell et al., 2009). On page 423 of
WEO2013, IEA notes that some countries, especially members of OPEC, report field condensate as
NGL, while others (like the OECD members) report this as part of crude oil20. This, along with the fact
that some OPEC Middle East countries have large condensate fields in production, lead the NGL
content of natural gas for OPEC to be significantly higher than for non-OPEC countries, illustrated in
Figure 33. It is however unclear if this is would explain the discrepancies in WEO2008-2010, and will
require further investigation. It is also noted by McGlade (2013) that the IEA has indicated that natural
gas production is shifting towards wetter sources, producing relatively more NGL.
NGL fraction of natural gas for 2012
50%
40%
30%
20%
10%
0%
OPEC
Non-OPEC
World
Figure 33: NGL production compared in percent to natural gas production for different regions. NGL data is from WEO2013,
natural gas production from BP (BP, 2013).
Another part of the reasoning behind the increase in the NGL content of natural gas in WEO2008-2010
may be found in the IEA report Natural Gas Liquids – Supply Outlook 2008-2015 (IEA, 2010b). Released
as a supplement to the monthly Oil Market Report in April 2010 and by a different team of analysts
than the WEOs, the report features a medium term view of NGL supply and also discusses the coupling
between NGL and natural gas production. While different IEA reports have a history of disagreeing
with each other (Miller, 2011), and the data is not supposed to be used interchangeably (IEA, 2010b),
some points raised within the NGL report are likely of relevance for the WEO NGL scenarios. These
include improved NGL extraction techniques, higher liquid content of new gas sources and that the
liquids are produced earlier in the lifetime of a gas field (IEA, 2010b). While the outlook for the growth
in the NGL to natural gas fraction has changed from WEO2008-2010 to WEO2013, this may in part
explain why the NGL content grew in the aforementioned report. However, it does not explain why
the IEA explicitly stated in WEO2008 that the NGL content of natural gas should remain constant
(Aleklett et al., 2010).
In the WEO2013 scenario the relative increase in NGL production is lower than that for natural gas.
Since the difference is not great at around 40% increase in NGL to 47% increase in natural gas, and the
natural gas scenario lies outside the scope of this project, the NGL scenario is accepted into the
Alternative scenario model.
20
Often referred to as crude + lease condensate
43
4 Track record analysis
Table 4 presents the track record of the main WEO scenarios for world oil supply, in terms relative of
actual production stated in each Outlook report, as described by equation 3.
𝑥=
𝑆𝑐𝑒𝑛𝑎𝑟𝑖𝑜 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛𝑛 −𝐴𝑐𝑡𝑢𝑎𝑙 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛𝑛
𝐴𝑐𝑡𝑢𝑎𝑙 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛𝑛
(3)
IEA actual production data has been used to make sure that the categorization and definitions of oil
types coincide as much as possible, though there are still some discrepancies (e.g. the
inclusion/exclusion of Venezuelan extra heavy oil for the crude and conventional oil categories). Some
of the earlier reports, such as WEO2000, only presents actual production for 1997, and then a scenario
prediction for 2010 and 2020. In such cases, actual production for the starting year of the scenario has
been taken either from other Outlook reports, or in a few cases the BP 2013 statistical review (BP,
2012). Since most of the values in Table 4 are linearly interpolated between the fixed points presented
by the IEA, there exists some uncertainties as to the accuracy of the track record analysis. However,
the linearized values corresponds well with the digitized scenarios in Figure 15, and should hence not
be so far off as to not be useful for analysis. The characters in bold print notes for which years actual
values are available, and the rest represents years which have been interpolated. Appendix I features
tables for the complete track record analysis for categories where such analysis has been applicable,
namely conventional oil, OPEC, Brazil, Russia, Saudi Arabia, Iraq, NGL and conventional crude oil.
Table 4: Track record of IEA world oil supply scenarios, relative to IEA actual production, and the average error by numbers
of years from scenario start
World oil supply scenarios track record, relative terms [%]
WEO2000
1999
2000
2001
2002
2003
2004
2006
2007
2008
2009
0.0
0.0
0.0
7.1
7.1
0.0
7.3
7.3
0.0
0.0
8.0
8.0
0.6
0.6
0.0
4.1 5.3 5.9
4.1 5.3 5.9
-3.2 -2.1 -1.7
-3.2 -2.1 -1.7
-3.3 -1.9 -1.1
0.0 1.5 2.3
0.0 0.6
0.0
8.3
8.3
0.5
0.5
1.5
5.0
3.1
2.6
0.0
9.0
9.0
1.0
1.0
2.4
5.9
3.8
3.4
0.4
0.0
13.7 12.0 13.0 11.7
13.7 12.0 13.0 11.7
5.3 3.6 4.2 2.8
5.3 3.6 4.2 2.8
7.1 5.7 6.5 5.1
10.7 9.3 9.6 7.7
8.4 6.8 7.5 6.1
8.0 6.6 7.1 5.6
4.5 2.8 3.1 1.5
3.2 0.6 0.1 -2.3
0.0 -1.9 -1.8 -3.6
0.0 0.1 -1.8
0.0 -1.6
0.0
1
1.6
2.0
1.5
2
3.1
3.6
2.2
3
4.4
4.8
2.2
4
4.4
4.8
2.7
5
4.4
4.8
1.5
6
5.3
9
7.1
10
8.2
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
2005
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Years
Average
Pre 2008
Post 2008
7
6.0
8
6.7
2010
2011
2012
Table 4 also presents the average absolute error, based on the number of years since the start of the
scenario. Hence year 1 represents the average error of all scenarios one year after the scenario start.
The Outlooks are also divided into categories of ‘Pre 2008’, representing WEO2000-2007, and ‘Post
2008’ (WEO2008-2013). This is to make a distinction between Outlooks utilizing decline rate analysis
and those that do not, and to analyze whether the more recent reports are more accurate. Since actual
44
production data is currently limited to year 2012, the ‘Post 2008’ Outlooks can only be compared for
a few years ahead. The 5 year average for Post 2008 Outlooks weighs solely on the track record of
WEO2008, since not enough time has passed for the other reports. In the same way the 10 year
average for all Outlooks weighs only on WEO2000-2003. The average error for all categories that have
been analyzed are presented in Table 5.
Table 5: Average error by years from scenario start for all track record analyzed categories. The color coding of the cells is
unique to each category. The number of data points vary between categories (see Appendix I)
Average absolute error [%]
Year
1
2
3
4
5
6
7
8
9
10
World oil supply
3.1
2.1
6.2
6.7
6.8
8.5
17.0
10.8
2.0
4.4
3.4
8.9
9.7
9.9
8.7
21.5
15.2
2.8
4.4
3.5
8.5
11.2
13.7
11.5
28.7
17.2
5.0
4.4
3.9
9.7
13.0
15.6
13.2
25.9
14.2
6.5
5.3
4.7
7.0
11.5
14.2
23.7
30.8
10.9
11.9
6.0
5.3
6.1
13.6
16.2
22.7
20.0
7.5
11.7
6.7
6.4
6.8
16.0
16.6
37.8
22.8
7.4
7.1
7.2
10.6
19.4
16.3
44.1
18.3
8.2
9.1
15.6
25.3
12.4
36.0
Conventional crude
1.6
1.4
3.4
4.5
5.1
7.6
12.3
5.5
1.7
NGL
2.8
3.3
3.1
2.2
4.4
8.4
10.6
Conventional oil
OPEC
Russia
Brazil
Iraq
Canadian oil sands
Saudi Arabia
Despite aforementioned uncertainties, some conclusions can indeed be draw from the analysis. Firstly,
the scenarios become increasingly inaccurate as more and more years pass by, as seen in Table 5. This
is as expected as uncertainty generally increases with increasingly longer outlooks. Secondly, the track
record analysis points towards an improvement in the accuracy of the Outlooks scenarios, with Post
2008 reports constantly outperforming Pre 2008 WEOs in 1-5 years averages, with only a few
exceptions (such as Brazil and to some extent the Canadian oil sands, mainly due to the WEO2008
scenario). Most of the categories featured in all or almost all Outlooks display the same behavior,
where accuracy increases in the Post 2008 WEOs. The same trend can be observed in categories less
frequently reported in the Outlooks.
By individual categories, the IEA seem to lean more towards overestimating production that
underestimating. In eight out of the ten categories analyzed in this chapter, the production was
overestimated more than 50% of the time (most ranging between 70-90%). Only the scenarios for
Russia and for NGL have underestimated production more frequently than overestimated. This is
mitigated to some extent in the Post 2008 Outlooks, but that may be due to a smaller data sample.
In conclusion, the track record analysis is subject to uncertainties such as lack of sufficient data points
for some scenarios and the effects of interpolation. However, it is still useful as a tool for quantifying
some of the findings in this report. It is observed that in general the IEA scenarios get more inaccurate
the more years pass by and that the IEA, especially for earlier Outlooks, has overestimated scenarios
to a much greater extent than underestimated them. It is also to some extent quantified that the Post
2008 Outlooks are more accurate than their earlier counterparts.
5 Remodeling and extrapolation
The New Policies Scenario in WEO2013 predicts that global oil supply will continue to rise steadily until
the end of the scenario period, reaching 98.1 Mb/d by 2035 (excluding processing gains and biofuels).
45
In chapter 3.4 and related subchapters the scenarios for WEO2013 are compared mainly to the
criticized WEO2008 Reference scenario, applying said criticism to the WEO2013 outlook. In reviewing
the WEO2013 main scenario in detail, no assumptions that could be viewed as unrealistic have been
found, although it should be argued that they are in some cases somewhat optimistic. This perceived
optimism notwithstanding, the WEO2013 New Policies Scenario can be accepted as a plausible future
for global oil supply.
There are however still some issues to raise in regards to said optimism, and to the sensitive nature of
some of the areas that have been analyzed. Chapter 5.1 features a remodeling of the WEO2013
scenario, with some rather modest changes to what is viewed as IEAs most optimistic assumptions. In
chapter 5.2 a long-term extrapolation to 2100 of both the WEO2013 scenario and the reevaluated
model is presented, using resource data found in WEO2013. Chapter 5.3 features a sensitivity analysis
on the scenarios, viewing the effects of, for example, changes to the URR in both the Alternative
scenario and the WEO2013 scenario.
5.1 The Alternative scenario
The changes to the WEO2013 New Policies Scenario are based on data gathered from the report and
the findings in chapter 3.4 and associated subchapters. Reviewing the changes in order by type of oil:
Crude Oil
-
-
-
-
Fields currently producing – the applied decline rates on fields currently in production is well
in line with other sources and empirical studies (Höök et al., 2014a; Jackson and Smith, 2014).
This section is therefore accepted as is.
YTD – The depletions rates for YTD used in WEO2013 exceeds those found from the fastest
developed region in the world, the North Sea. While the North Sea depletion is by no means a
theoretical maximum, it is arguably optimistic enough to assume that the development of
fields YTD will go as fast as the North Sea, and not faster. The YTD category is therefore limited
to follow the North Sea trend depletion (see chapter 3.4.1.2).
YTF – The same arguments used for YTD are applied here, although as shown in chapter 3.4.1.3,
URR for this category is more uncertain. For the main models URR the 170 Gb of discoveries
that the IEA state is necessary for the New Policies Scenario is used, and again the development
is limited to the North Sea depletion trend (see chapter 3.4.1.3).
EOR – As stated in chapter 3.4.1.4, the EOR scenario is viewed as both reasonable and properly
justified by the IEA, and is accepted without any revisions.
Unconventional Oil
-
-
Canadian oil sands – As discussed in chapter 3.4.2.1, there are no objections to the oil sands
scenario, and it is accepted.
Venezuelan Extra Heavy – Makes a rather small contribution of 2.1 Mb/d by 2035, and no
objections are found (see chapter 3.3.1.3 for details).
CTL and GTL – In WEO2013 CTL and GTL provides a combined total production of 2 Mb/d by
2035. The contribution is quite small and no objections to this part of the scenario are found
(see chapter 3.4.2.2).
LTO – In the WEO2013 scenario, LTO is predicted to provide more than one third of all
unconventional oil by 2035. While indeed a large contribution, there is no scientific ground to
object to the scenario (see chapter 3.4.2.3 for a more detailed discussion), and it is accepted.
46
The individual categories studied here makes up 13.9 of the 15 Mb/d of unconventional oil
forecasted for 2035. As all categories are perceived to be well enough justified, the IEAs scenario
for unconventional oil is accepted as is.
NGL
NGL is strongly coupled to the production of natural gas, as described in chapter 3.4.3. The
scenarios for natural gas are outside the scope of this project, and are thus assumed to be correct.
The objections raised by Aleklett et al. (2010) on the WEO2008 NGL scenario have been rectified
by the IEA, and the outlook for NGL can therefore be accepted.
Summarizing the different fractions, it is found that global oil production by 2035 will be 91.1 Mb/d,
some 7.1 Mb/d lower than the WEO2013 scenario, and as shown in Table 6.
Table 6: Summary of reported production numbers in WEO2013 and results from the analysis in this report, compared to the
results of the analysis in The Peak of the Oil Age (Aleklett et al., 2010). All numbers in Mb/d
Production in 2030
Source
Currently producing
YTD
YTF
EOR
Crude oil - total
NGL
LTO
Other Unconventional
WEO2008
27.1
22.5
19.2
6.4
75.2
19.8
0
8.8
Aleklett et al.
(2010)
27.1
13.6
8.7
6.4
55.8
15.3
0
6.5
Global oil production
103.8
77.6
Production in 2035
Alternative
scenario
WEO2013
(this study)
27.1
27.1
19.8
16.0
15.9
12.6
2.7
2.7
65.5
58.4
17.7
17.7
5.6
5.6
9.4
9.4
98.2
91.1
Table 6 also shows the results of the analysis performed by Aleklett et al. (2010), with NGL in volumetric
terms, for comparative purposes. It is apparent that the revisions to the IEA scenario in this project is
much more modest than those done in The Peak of the Oil Age. There exists two main reasons for this.
The first is the lack of transparency in WEO2013 concerning the crude oil categories YTD and YTF.
WEO2013 contains almost no discussion about fields YTD or YTF, whereas WEO2008 discusses at least
the YTD category in some detail. The lack of detailed URRs for YTD and YTF in the 2013 Outlook mean
that this study has to settle for analysis on a global scale, which in this case may result in a more positive
result (see chapter 3.4.1.2 for a more detailed discussion).
The second and most prominent reason for the more modest revision is simply the fact that the
WEO2013 scenario is better substantiated than WEO2008s. The crude oil category for fields currently
producing relies on established decline rates, while the YTD and YTF categories are within some reason,
at least on the assumption of a correct URR presented in the report. No problems are found with the
EOR scenario, and the problems regarding NGL highlighted by Aleklett et al. (2010) (i.e. the mismatch
between NGL and natural gas projections) have been rectified. The scenario for Canadian oil sands is
also more realistic, and with the Venezuelan extra-heavy oil once again categorized as unconventional
oil, no objections to the Unconventional category are found. The emergence of LTO in North America
further distances the projections in this report from that presented by Aleklett et al. (2010). It has a
47
large effect on the scenario, but as discussed in chapter 3.4.2.3, there are no current grounds to object
to these projections.
The Alternative scenario compared to WEO2013
120
100
Mb/d
80
60
40
20
0
2012
2018
Currently producing
EOR
Light Tight
2024
YTD
NGL
WEO2013
2030
YTF
Other Unconventional
Figure 34: The Alternative scenario compared to the WEO2013 New Policies Scenario, excluding processing gains and biofuels
In conclusion, the IEAs methods and practices have seemingly improved in more recent years, and the
WEO2013 oil scenario appears more substantiated than its 2008 counterpart (of course, it remains to
be seen if it is more correct). However, while the WEO2013 is viewed as realistic, it may also be said to
be somewhat optimistic, as it hinges on a lot of things going the right way, such as timely investments
and the absence of large scale geopolitical events, especially such concerning oil producing countries.
Table 6 presents the end result of the analysis performed in this project, while Figure 34 shows the
Alternative scenario model for the entire length of the scenario. In contrast to the WEO2013 New
Policies Scenario, which predicts a steady increase in global oil production, the revised model presents
future oil production in more of a plateau. However, even the Alternative scenario should be viewed
as rather optimistic, as it depends on a lot of production coming on stream from sources that have
either not yet been developed for various reasons (such as ‘fallow fields’, see chapter 3.4.1.2) or that
have not yet been found. The effect of slower development of YTD or slower discoveries of new crude
oil, which translates to a lower URR for the YTD and YTF categories, is further analyzed in the sensitivity
analysis in chapter 5.3. There should also be some concern as to the implications for oil supply in the
longer term, if the WEO2013 scenario comes true, especially in terms of conventional crude
production. In chapter 5.2 the scenario is extrapolated to the year 2100, based on decline rates and
URR limitations presented in WEO2013.
5.2 Long term extrapolation of the scenarios
In the WEO2013 New Policies Scenario, the IEA foresees no peak in oil production within the scenario
time frame, although there is ample mention of the risks of under-investments. In a spotlight called
“Has the rise of LTO resolved the debate about peak oil?” on pages 447-448 IEA claims that the LTO
resources are not enough to move the oil peak more than a few years, but concludes that
48
“Taking into account the large amount of unconventional resources that becomes available as oil prices
increase, in addition to the significant remaining conventional resources and the sizable potential for
EOR in conventional fields, no peak occurs before the end of the projection period.” (IEA, 2013b).
The notion that LTO will not move an eventual peak in oil production is in line with the statements of
the German Federal Institute of Geosciences and Natural Resources that non-conventional oil
production will not prevent a peak in oil production, but rather modify the subsequent decline (BGR,
2008). This chapter examines the implications of the New Policies Scenario on the longer term (up until
the year 2100), extrapolating from the implied URRs and production trends. The assumptions used are
described below.
For crude oil from fields currently producing a continuing decline rate of 4% per year up until 2050 is
assumed, and then reduced to 3.5% until 2100. This allows for some continued long-term production
from the super-giants, while phasing out the smaller fields that dries up. Total production by 2100
lands at 626 Gb, roughly equaling the RRR that digitization of figure 13.5 in WEO2013 gives. EOR is
assumed to remain constant at 2.7 Mb/d for the period 2035-2100.
For fields YTD and YTF a continuing depletion rate trend is assumed, based on the predicted production
up until 2035 and the YTD URR given by digitization of figure 13.5 in WEO2013 (236.8 Gb). For YTF IEA
states that it expects 170 Gb of new conventional oil to be discovered within the scenario time frame,
while total undiscovered oil amounts to 649 Gb (IEA, 2013a). This amount is said to be based upon a
combination of the IEA database, the USGS 2000 assessment and the USGS 2012 updated assessment
of undiscovered oil, although the IEA directly cite the USGS 2012 assessment (IEA, 2013a). The 649 Gb
of undiscovered resources do however exactly match the mean values for undiscovered resources
from the USGS 2000 assessment (Schindler and Zittel, 2008). By comparison, the USGS 2012
assessment leaves the mean value of undiscovered conventional oil at 592 Gb (McGlade, 2012), falling
some 57 Gb short of the IEA estimates. Despite these inconsistencies, the 649 Gb is accepted as URR
for undiscovered resources, in line with the practice of this study.
It is also assumed that all 649 Gb of remaining conventional oil will be discovered before 2100 and put
in production. Thus the depletion rate of IEAs YTF production scenario has been recalculated using 649
Gb as URR, and the depletion rate trend is then extrapolated to2100. The extrapolation implies that
by 2100, 235 Gb of oil YTD and 515 Gb of oil YTF would have been produced, leaving around 140 Gb of
URR left. The URRs of these categories are subject to great uncertainties, and in the sensitivity analysis
in chapter 5.3 the effect of revisions to these numbers are examined.
NGL is assumed to remain at the production levels of 2035 (17.7 Mb/d), due to lack of data suggesting
otherwise. There is the possibility of natural gas production continuing to rise, in turn increasing NGL
supply. However, there likely exists limits to how much crude oil NGLs can substitute as it is mainly
used in the petrochemical industries (Miller and Sorrell, 2014). Also, on page 423 of WEO2013 the IEA
claims that the RTRR of conventional NGL (including U.S. unconventional NGL) to be 465 Gb. It is
unclear if this is presented in terms of volume or oil equivalents. However, under the assumption of a
steady NGL production in the period of 2035-2100, 555 Gb of NGL would be produced in volumetric
terms, and 416 Gb in oil-equivalents. With both numbers roughly equaling the RTRR of NGL, it is
assumed NGL production to stay at the level IEA predicts for 2035.
The production of unconventional oil grows according to a curve-fit to the New Polices scenario, but is
limited to 16.5 Mb/d, a number derived from the Rystad Energy UCube database (Rystad Energy AS,
2014). This allows for the possibility of some continuing growth in Canadian oil sands, Venezuelan
extra-heavy oil and CTL/GTL, while also taking into account the possible decline in LTO production.
49
However, this category remains very uncertain. Finally, processing gains is set to add 2.6% of all oil
production to the total.
Extrapolation of WEO2013 New Policies Scenario
120
100
Mb/d
80
60
40
20
0
2012
2023
Currently producing
2034
YTD
YTF
2045
EOR
2056
NGL
2067
2078
Total Unconventional
2089
2100
Processing gains
Figure 35: Long term extrapolation of the WEO2013 New Policies Scenario
The result of the long-term extrapolation is shown in Figure 35. It is found that under aforementioned
assumptions on URR and growth rates of NGLs and unconventional oil, world oil production would
peak in 2035 and enter a steady decline, dropping an average of 0.8 Mb/d per year, reaching around
53.4 Mb/d by 2100. However, given the huge uncertainties surrounding this extrapolation, it should
not be viewed as an actual scenario. Rather, it should be regarded as a conceptualization of the effect
that a continued expansion of oil production would have on the long-term world supply. The 2035
peak in Figure 35 appears rather sudden and abrupt, and is caused mainly by the depletion of the URR
in fields YTD. A perhaps more likely scenario would be a smoother peak or even a plateau, albeit the
constrains on spare capacity in recent years (Benes et al., 2012) may suggest a sharper peak. Utilizing
the framework for peak oil mitigation planning developed by Hirsch (2008), this would likely constitute
a Middling Case Scenario, wherein “World oil production grows to a relatively sharp break maximum,
after which it goes into a monotonic decline” (Hirsch, 2008).
Extrapolation of the Alternative scenario
120
100
Mb/d
80
60
40
20
0
2012
2023
Currently producing
2034
YTD
YTF
2045
EOR
2056
NGL
Figure 36: Long term extrapolation of the reevaluated model
50
2067
2078
Total Unconventional
2089
2100
Processing gains
Figure 36 provides a similar long-term extrapolation of the Alternative scenario. The same assumptions
are used for this extrapolation, the only difference being slower depletion rates for fields YTD and YTF
(which continues along the North Sea depletion trend), and how they are applied to the YTF category.
For fields YTF, the depletion trend has been applied to an URR of 170 Gb, which is the IEAs assumed
discoveries for 2012-2035. The resulting production levels for 2012-2035 are then used to recalculate
a new depletion rate which is in turn linearly extrapolated and applied on the total URR of 649 Gb, as
to not assume that all undiscovered oil can be found within the period of 2012-2035. This is further
discussed in the sensitivity analysis of the Alternative scenario (chapter 5.3). The result is a smoother
curve with a peak a little earlier than the WEO2013 extrapolation (in the year 2032), and at a lower
level of production. Production then plateaus for a few years and later starts to decline by
approximately 0.6 Mb/d per year. By 2100 production is at 55.7 Mb/d, slightly higher than the
WEO2013 extrapolation owing mainly to the slower depletion of the YTF resources. Within the
framework described by Hirsch (2008), this scenario constitutes a Best Case Scenario. In both
extrapolations, oil supply peaks right by the end of the original scenario time frame. Similar behavior
can be found in multiple official prognoses (Sorrell et al., 2010, 2009).
5.3 Sensitivity analysis
As with any scenario or model of the future, the Alternative scenario and the long-term extrapolations
comes with great uncertainties. The aim of this chapter is to analyze the sensitivity of these scenarios
to changes in some of the base assumptions. It is previously stated that the Alternative scenarios relies
on a set depletion rate of the resource base for both fields YTD and YTF, however the URR of these
categories remains highly uncertain. The amount of oil found in fields YTD can be argued to be
somewhat certain, at least to the extent that the oil likely exists. As discussed in previous chapters oil
fields are often the subject of reserve growth, increasing the URR. The different components of the
total URR are illustrated in Figure 37. There also exists the possibility that some of the oil is in ‘fallow
fields’, some of which may be unsuitable for development, thus limiting the URR. As for fields YTF, the
URR corresponds to oil that is assumed to be discovered within the scenario time frame, an estimation
that of course comes with great uncertainty. As the conventional crude oil peaks and declines in both
extrapolation scenarios, it becomes likely that producers will want to expand the unconventional
production. Due to aforementioned limitations, the likelihood of a further increased production of
unconventional oil seems low, although high enough oil prices may encourage a slight expansion.
Based on these possibilities, some cases to analyze the effects of such changes on the Alternative
scenario and the long term extrapolations have been designed.
51
Figure 37: Illustration of the different components of the URR. Adapted from Sorrell et al. (2009)
The main purpose of these cases are to provide a solid interval for the sensitivity analysis. For the
WEO2013 long term extrapolation three different cases, the Low/High URR cases and the High
Unconventional case, as well two combinations of the URR and unconventional cases are presented.
The Low/High URR cases are based on changes to the URR of fields YTD and YTF. For YTD a simple
interval of 236.8 ±100 Gb is applied to the URR, accommodating for possible reserve growth or ‘fallow
fields’ not suited for development. Thus the YTD URR is 136.8 Gb in the Low URR case, and 336.8 Gb
in the High URR case. For the URR of fields YTF, i.e. oil presumed to be discovered within the scenario,
the P95 and P5 estimations21 from USGSs latest assessment of undiscovered resources worldwide are
used (Schenk et al., 2012). Since this assessment excludes the U.S. resources, the 27 Gb of
undiscovered resources that is the USGS P50 estimation for the U.S. (USGS, 2013) are incorporated.
Hence, the URR of YTF for the sensitivity analysis is 649 Gb as basis, 226 Gb for the Low URR case and
1235 Gb for the High URR case. The YTD and YTF production scenarios from WEO2013 are still used
and limits the production up until 2035, but the depletion rate is recalculated for the new URR in each
case and continues in a linear fashion for the remainder of the scenario. The High unconventional case
uses the same curve-fit of the WEO2013 unconventional oil scenario as the long-term extrapolation,
but removes the 16.5 Mb/d production cap, allowing continued growth of unconventional oil
production throughout the scenario, for example from LTO outside of North America. Table 7 presents
all the cases in the sensitivity analysis. Note that the last two cases, the Early low URR and Early high
URR, are only used in the analysis of the Alternative scenario.
21
According to the USGS definitions, P95 undiscovered resources have a probability of 95% at least the amount
of oil existing. Likewise, P5 corresponds to a probability of 5%, and P50 to 50%.
52
Table 7: Variations to the URR of YTD and YTF and the unconventional oil production limit in the sensitivity cases
Parameters for the sensitivity analysis cases
YTD URR [Gb]
YTF URR [Gb]
YTF URR by 2035
[Gb]
Unconv. Limit
[mb/d]
Low URR,
High
unconv.
High URR,
high
unconv.
Early low
URR
Early high
URR
Base case
Low URR
High URR
High
Unconv.
236.8
649
136.8
227
336.8
1235
236.8
649
136.8
227
336.8
1235
136.8
227
336.8
1235
170
170
170
170
170
170
113
250
16.5
16.5
16.5
-
-
-
16.5
16.5
Figure 38 presents the production in each of the sensitivity cases. As this is an extrapolation of the
WEO2013 New Policies Scenario, production remains the same until 2035 regardless of changes to the
URR. It should also be emphasized that these cases are not likely scenarios, but rather provides an
interval for a possible future supply. The URR cases affect both the shape and size of the supply curve,
while the additional unconventional cases mainly raises the supply some. In the High URR cases the
2035 supply peak turns into a plateau that lasts for at least another decade, before again beginning
the decline. The Low URR cases emphasizes the peak, owing mainly to the depletion of the YTD URR,
of which only 26 Gb remains after 2035. By 2100, all cases fall within the range of 40-80 Mb/d of world
oil supply, well below the 87 Mb/d supply of 2012, and indicating that if the WEO2013 New Policies
Scenario is correct the peak in oil supply would occur right around the year 2035.
Sensitivity analysis of the WEO2013 long term extrapolation
120
100
Mb/d
80
60
40
20
0
2012
2023
2034
2045
2056
Base case
High URR case
Low URR, high unconventional
2067
2078
2089
2100
Low URR case
High unconventional case
High URR, high unconventional
Figure 38: Sensitivity analysis of the long term extrapolation of WEO2013. Production up until 2035 is the same as in the
WEO2013 New Policies Scenario
A similar analysis has been performed on the Alternative scenario and the associated extrapolation.
The same basic cases of URR and Unconventional oil as presented in Table 7 are used for the
Alternative scenario, with the addition of the ‘Early’ cases. This is because the Alternative scenario
53
model is based on the depletion rate trend of the North Sea region, as discussed in chapter 5.1. Thus
any changes to the URR of the fields YTD and YTF will affect the scenario right from the start, providing
a sensitivity analysis for the entire scenario length. For fields YTD, this simply means that the Low/High
URR cases will change production levels accordingly.
The YTF category becomes more complex however, as it would depend not only on the amount of oil
that will be discovered within the time frame, but also on the rate of discovery. For example, if the
North Sea depletion trend were to be applied to the YTF URR in the High URR case (1235 Gb),
production from fields YTF would skyrocket. This would imply that all oil resources could be discovered
within the first 23 years of the scenario (2012-2035). In the extrapolation of the Alternative scenario
this is counteracted by applying the North Sea depletion trend to the 170 Gb of presumed discoveries
for 2012-2035, and using the resulting production to calculate a new depletion rate based on the total
undiscovered URR. The new depletion rate is then linearly extrapolated for 2036-2100 and applied to
the total YTF URR. To accommodate the possibility of lower or higher discoveries within the Alternative
scenario time frame (2012-2035), two more cases have been devised for the sensitivity analysis; the
Early Low URR and Early High URR cases (see Table 7). These are based on the same amount of total
URR as their Low/High URR counterparts, but in the Early Low URR case only 113 Gb of oil is discovered
between 2012 and 2035, based on the curve-fit of historical discoveries (see chapter 3.4.1.3). The
corresponding 2012-2035 discoveries in the Early High URR case is 250 Gb.
Sensitivity analysis of the Alternative scenario and the
long term extrapolation
120
100
Mb/d
80
60
40
20
0
2012
2023
2034
2045
2056
2067
2078
2089
Base Case
High URR case
Low URR case
Early high URR
Early low URR
High unconv.
High URR, high unconv.
Low URR, high unconv.
2100
Figure 39: Sensitivity analysis of the Alternative scenario. Early high/low cases allow for more/less oil to be discovered before
2035
Figure 39 show the results of the different cases in the sensitivity analysis. The spread is similar to that
seen in Figure 38, with production in 2100 landing between 44 and 81 Mb/d. The different variants of
the High URR case all see a continued increase in supply in the Alternative scenario time frame, but
peaks and plateaus within a decade after. The Early High URR case sees the largest production, even
surpassing the WEO2013 scenario on account of a rapidly expanding YTF production, but declines
faster than the High URR case by the end of the extrapolation. The Low URR cases all exhibit very
similar behavior, the Early Low URR case declining somewhat faster early on in the scenario. Table 8
54
presents oil supply for all cases by the end of the Alternative scenario and the long term extrapolation,
as well as peak supply and peak year.
Table 8: Supply in 2035 and 2100, with peak supply and peak year for the cases in the sensitivity analysis of the Alternative
scenario.
Alternative cases
Supply in 2035 [Mb/d]
Supply in 2100 [Mb/d]
Peak supply [Mb/d]
Peak year
Base Case
Low URR
Early Low
URR
High URR
Early High
URR
High
unconv.
Low
URR,
high
unconv.
93.1
55.7
93.5
2032
86.2
44.4
89.3
2012
81.9
45.4
89.3
2012
100.0
71.6
100.2
2033
106.1
74.9
106.2
2036
93.1
64.6
93.5
2032
86.2
50.4
89.3
2012
High
URR,
high
unconv.
100.0
80.5
100.2
2033
Table 8 shows that despite large differences in all supply categories, all scenarios have peaked and
entered a state of decline or plateau by 2036. This is in line with the findings of other studies (Jakobsson
et al., 2009; Sorrell et al., 2009). The High URR cases all reach supply levels equal or surpassing the
WEO2013 New Policies Scenario, although only by a few Mb/d at best and with resources far
surpassing the assumptions the IEA uses. It should also be emphasized once more that the cases are
to be considered rather extreme, and are only featured to create a plausible interval for a possible
future oil supply outlook.
There are of course other uncertainties to both the WEO2013 New Policies Scenario and the
Alternative scenario, such as unrest in Iraq, continued embargos on Iran or failures at Brazil’s pre-salt
fields, the effects of which are much more difficult to quantify. If Iraq were to erupt in civil war, caused
by for instance an al Qaeda resurgence (Riedel, 2014; Weston, 2014), production may stay at current
level or even drop. This would mean a loss of 6 Mb/d of expected growth in the New Policies Scenario.
This would not necessarily translate to a 6 Mb/d lower production in 2035, as the demand may cause
another rise in oil price leading to other producers stepping in. However, production may already be
strained in most countries, and there may not exist enough capacity to replace the loss. The point of
this sensitivity analysis has therefore been to create an interval for possible oil supply outlooks with as
few assumptions on above-ground factors as possible.
6 Discussion
6.1 IEA track record
The track record analysis presented in chapter 4 indicates that the accuracy of IEA scenarios has
apparently improved in recent years, and the detailed analysis of WEO2013 shows that current
assumptions are more or less in line with contemporary literature. However, the statistical basis for
the track record analysis of the most recent WEOs is limited to only a few years. The analysis shows
decreasing accuracy the longer a scenario runs, indicating that the recent scenarios may also continue
to err in the longer term. While previous Outlooks has shown a clear tendency towards overestimating
oil supply, the trends in more recent reports are not as clear, which could indicate that the current
scenarios may even underestimate oil supply just as much as overestimate. However, this is based
solely on the IEAs track record, and not on any geological, policy or economic analysis. Of course, some
of the errors stem from large and unforeseeable above-ground events, such as the 2003 U.S. invasion
of Iraq and to a greater extent the global financial crisis of 2008. Still, large revision to the IEA scenarios
55
was made well before the crisis, and as seen in Figure 4 and Figure 6, both the oil price and world oil
supply rebounded quickly after 2008.
While difficult to quantify, the IEA has seemingly often had high expectations on new technology and
new supply sources. When the IEA abandoned its early reliance on the ‘call-on-OPEC’, more and more
attention turned to unconventional resources. Contributions from EOR techniques, NGL and Canadian
oil sands, and to some extent CTL and GTL techniques22 have all been revised downwards from previous
scenario highs. Compared to a tremendous growth in contribution from conventional crude oil in
earlier Outlooks23, the WEO2008 reference scenario with a special focus on oil and an extensive decline
rate analysis saw a steep revision of some 15 Mb/d by 2030 from the WEO2006 scenario. By
comparison, the difference in total world oil supply between the scenarios by 2030 is only 10 Mb/d.
Interestingly, the outlooks for both Canadian oil sands and NGL were highest in the WEO2008, both of
which came under criticism from Aleklett et al. (2010)24 and have subsequently been revised
downwards again. Currently the largest and most untested resources (at least in large scale
production) would be the Brazilian pre-salt finds and the North American LTO. If viewed purely from
the IEAs track record of new resources and untested extraction techniques, one could expect both
categories to be revised downwards within a few years. However, pointing to the opposite, recent
WEO scenarios have underestimated the rapid LTO production growth.
Since the special focus on oil in WEO2008 and the inclusion of a decline rate analysis, the IEA have
developed a seemingly more empirical approach to the future of supply, and combine geological and
economic factors to a greater extent. In contrast, the earlier Outlooks seemingly relied heavily on
presumed demand and OPECs ability and willingness to increase its supply. Indeed, while the
dependence on OPEC and the Middle East has somewhat lessened in more recent Outlooks (See
chapter 3.3.1), its importance for future oil supply cannot be neglected. Any unforeseen shortcomings
in other parts of the world regarding petroleum production would likely have to be compensated by
OPEC, as IEAs scenarios put most non-OPEC countries at their maximum production capacities. OPECs
willingness and often debated capability of filling this void would then be of utmost importance. The
IEAs scenarios for OPEC supply in turn depend heavily on growing contributions from countries
associated with severe political issues such as Iraq and Iran. As mentioned in the sensitivity analysis,
uncertainties regarding such countries is difficult to quantify and almost impossible to foresee, but any
shortcomings in Iraqi production could easily mean a much lower oil supply growth than the WEO2013
New Policies Scenario indicates. The IEAs history of underestimating domestic oil consumption in OPEC
and Middle East countries may also lead to an overestimation of their export capabilities (Gately et al.,
2013).
6.2 The Alternative scenario and long term extrapolations
The Alternative scenario represents what is believed to be the most realistic scenario for future oil
supply, based solely on the figures presented in the WEO2013 report. In contrast to WEO2013, in which
oil supply grows throughout the scenario, the Alternative scenario sees world supply in a plateau phase
for the entire time frame, with only a slight increase of some 4-5 Mb/d. The differences between the
22
Although both expected contributions from and revisions to the CTL and GTL categories have been
comparatively small in absolute terms, the reductions in the GTL scenarios amount to 70% in 2030 from the
highest scenario to WEO2013, as seen in chapter 3.4.2.2.
23
See Figure 16. Of the Outlooks predating WEO2008, only WEO2006 features a separate scenario for
conventional crude oil. The conventional oil scenarios (crude + NGL) of earlier Outlooks are however directly
comparable in size to WEO2006, indicating that expectations on crude oil were just as high in those reports.
24
Note that the criticism of the WEO oil sands scenario by Aleklett et al. (2010) was based purely on a crash
programme assessment by Söderbergh et al. (2007).
56
scenarios comes entirely from the assumptions on the development pace of fields YTD and YTF, i.e.
the depletion rate. An important concept such as this is not discussed within WEO2013, which instead
focuses solely on decline rate analysis of both conventional and unconventional oil. While decline rates
are of importance for future oil supply, such an analysis focuses more on how much capacity that will
have to be replaced, and not on how it will be replaced. While the decline rate analysis in WEO2013 is
sound, the assumption in the New Policies Scenario is that almost all conventional crude oil capacity
lost to decline will be replaced in the scenario time frame. Based on the resources presumed to be
developed, this implies the application of depletion rates greater than any than any regional rates seen
historically. The IEA has not motivated this deviation in any of the Outlooks.
In the Alternative scenario the development pace of new resources is limited to that of the historical
trend of the North Sea, a region with the highest observed depletion rate in the world. It has previously
been emphasized that this is in no way a theoretical maximum, and some authors criticize comparison
of a single region developed from scratch to a vast number of fields in separate regions worldwide, as
is the case with the YTD and YTF fields (McGlade, 2013). An argument can still be made however, that
using the North Sea as an analogy and not a theoretical maximum provides a good estimation of the
feasible development pace (Höök, 2014). Any scenario requiring a depletion rate greater than that of
the North Sea should therefore have to be well substantiated, something that cannot be said of the
WEO2013 New Policies Scenario. As the analysis of WEO2013 has shown however, the depletion rates
do not exceed the North Sea by far, as they did in WEO2008, and the scenario can be viewed as realistic
but optimistic.
The depletion rate analysis of WEO2013 also rests on some assumptions of the URR used by the IEA.
For fields YTF the URR used is stated by the IEA in the text, but the use of the YTD URR used for analysis
in this report is at best implied by the IEA. However, seeing as the YTD URR closely match the data
provided in WEO2008, it is assumed to be correct. As assumptions and resource data used in WEO2013
is less readily available than in for instance WEO2008, especially in regards to fields YTD and YTF, both
the analysis of WEO2013 and the Alternative scenario suffers from some uncertainties. These
problems are quantified in the sensitivity analysis, which shows that under the models assumptions on
the depletion of resources, changes to the URR can have great consequences. In turn, the sensitivity
analysis of the long term extrapolations indicate that while changes to the URR would have great
consequences on future supply levels, even extreme amounts of new discoveries25 are not enough to
offset a peak for more than a few decades at best. This is in line with the findings of other studies
(Jakobsson et al., 2009; Sorrell et al., 2009)
The general discovery trend is towards smaller and smaller fields being discovered, which tend to
decline at a much faster rate than the old giants (Sorrell et al., 2009). This suggests that future crude
oil production will lean increasingly on smaller fields, and more and more fields will have to come on
stream continually to offset the increasing decline rate. Also, while technical advancements and
increasing oil prices makes smaller and smaller fields more viable for development, there should exist
a lower limit due to the ‘energy return on investment’ (EROI) (Sorrell et al., 2012). At some level, the
energy that is extractable from small fields may be lower than that spent on exploration and
development of that particular field. As such, some small fields may never become viable for
development (Sorrell et al., 2012). Hence, maintaining high levels of crude oil production is likely to
become increasingly difficult, suggesting that future oil supply may well be closer to the lower cases
presented in the sensitivity analysis.
25
The USGS P5 values are likely to be considered a maximum URR available for discoveries.
57
The sensitivity analysis of the Alternative scenario features a total of five cases and three combinations
of those cases, designed to evaluate the scenario and its long term extrapolation. Combined, the cases
should provide a decent and probable interval for the future of oil supply, which interestingly
encompasses the WEO2013 New Policies Scenario. Some cases or combinations of cases have been
omitted from the study, such as a ‘Low unconventional case’ or an ‘Early High URR, high
unconventional case’. The Low unconventional case is left out on the assumption that any decline in
conventional crude oil production will incite larger investments into unconventional production to
offset said decline. Likewise, an ample early supply of crude oil would at more likely lead to lesser
investments in unconventional production as oil prices drop.
7 Conclusions
Compared to historical production, IEA central oil supply scenarios published in the World Energy
Outlooks of 2000 to 2013 have predominately been overestimated and past scenario revisions have
mainly been downward adjustments of future production rates. A track record analysis of ten scenario
categories, consisting of global oil supply, individual regions and different types of oil, shows that
accuracy decreases with scenario time frame. For the IEA category ‘World oil supply’ the average
absolute error on a ten year horizon was 8.2%. According to the analysis, the accuracy of scenarios has
improved in recent years. The average absolute error on a five year horizon for ‘World oil supply’ was
4.8% for pre 2008 scenarios, and 1.5% for post 2008 scenarios. A possible contributing factor to this
improvement is the extended modeling methodology employed in WEO 2008 and onwards which
include both geologic and economic factors to a higher extent. It should be noted that the apparent
increase of accuracy in post 2008 scenarios is subject to higher uncertainty due fewer available data
points. The accuracy of post 2008 scenarios for longer time-horizons will need further evaluation, as
actual production data is not yet available.
From the first report to project oil supply to 2030, WEO2002, to the latest report, WEO2013, the
outlook for global oil supply has seen considerable downward revisions, amounting to 20 Mb/d for
2030. Revisions to the scenarios for OPEC, owing mainly to high expectations of the organizations
willingness and capability to act as a swing producer in earlier Outlooks, closely match the reductions
in global supply scenarios. While the OPEC revisions stands for the bulk of the drop, scenarios for many
countries and types of oil have been revised both up and down interchangeably. New types of
unconventional oil, requiring additional and advanced extraction techniques or upgrading facilities,
have generally been associated with high expectations in the early stages of development but later
often been revised downwards after a few years, as in the case for Canadian oil sands and gas to liquids.
However, the opposite seems to apply for light tight oil where initial outlooks have underestimated
actual production and scenarios have been revised upwards. Individual downward revisions in
categories are seldom commented at length in the reports, even in cases where the size of the revision
is considerable.
The IEA sees oil becoming more and more expensive as production scenarios drop and while oil price
and investment assumptions and scenarios increase. The two are of course interconnected, with
increasing investment demands leading to higher oil prices, and higher prices offering more incentive
for added investments. These changes to the scenarios are in line with the expressed view of IEA chief
economist Dr. Fatih Birol that “the age of cheap oil is over” (Kumar, 2011).
A few cases of inconsistencies in referring to correct data sources has been found, where the IEA
reports for instance resource estimates for undiscovered oil that do not match the data from the
reported source. Some scenario estimates lacks published motivation (such as enhanced oil recovery
production in WEO2004), while in other cases IEA makes statements which their own scenarios
58
disagree on, as is the case with natural gas liquids (‘NGL’). In the NGL case in particular, the IEA clearly
states in WEO2008 that the NGL fraction of natural gas production is assumed to be consistent
throughout the scenario, although the actual scenario suggest that the NGL fraction will rise. This
inconsistency was pointed out by Aleklett et al. (2010), and although never commented on by the IEA
in their WEO reports, it seems to have been amended over the course of four Outlooks. No such
apparent inconsistencies have been found in the latest report.
The WEO2013 New Policies Scenario has been evaluated using decline and depletion rate analysis, and
found to be a reasonable outlook for future oil supply, on the assumption that the reported resource
data is accurate. The depletion rates of fields yet to be developed (‘YTD’) and yet to be found (‘YTF’)
were found to be slightly higher than that of the North Sea, the fastest developed region in history.
While the depletion rate of the North Sea is not a theoretical maximum, such a claim should require
proper justification, which is absent in the 2013 Outlook. The other parts of the New Policies Scenario
appears sound, and as the depletion rates can still be considered to be within reason the scenario is
not objected to, although it can be perceived as optimistic. An extrapolation of the WEO2013 scenario
to the year 2100, based on reported resource data and continuing decline and depletion rate trends,
shows that oil supply would peak in 2035. Assuming that additional resources can be discovered or
extracted mainly serves to slow the rate of decline after a mid-2030 peak.
In the Alternative Scenario the depletion rates of fields YTF and YTD are limited to the North Sea
depletion trend. This results in a scenario where oil supply goes into a plateau phase for the period
2012-2035, with world supply by 2035 providing 7.1 Mb/d less than the WEO2013 New Policies
Scenario. A sensitivity analysis shows that increasing the amount of ultimately recoverable resources
(‘URR’), by for example assuming reserve growth in fields YTD or a higher discovery rate results in a
scenario more comparable with the New Policies Scenario. In the most extreme case oil supply
increases to 106 Mb/d at the peak, although this case is designed to provide the maximum in a
sensitivity interval, and depends on highly unlikely values for URR. An extrapolation of the Alternative
scenario to 2100 results in a peak in 2032, with additional oil discoveries at best postponing the peak
by a few years.
The extrapolations of both the WEO2013 New Policies Scenario and the Alternative scenario in this
report indicate that global oil supply peaks before or around 2035. At best, the peak may occur a few
years after 2035, but this relies on highly unlikely values for URR. The tendency of oil production
peaking soon after the end of the scenario time frame has been noted in previous assessments of
several other official projections for conventional oil (Bentley et al., 2007; Sorrell et al., 2009).
59
8 References
Al Adasani, A., Bai, B., 2011. Analysis of EOR projects and updated screening criteria. J. Pet. Sci. Eng. 79, 10–24.
doi:10.1016/j.petrol.2011.07.005
Aleklett, K., Höök, M., Jakobsson, K., Lardelli, M., Snowden, S., Söderbergh, B., 2010. The Peak of the Oil Age –
Analyzing the world oil production Reference Scenario in World Energy Outlook 2008. Energy Policy
38, 1398–1414. doi:10.1016/j.enpol.2009.11.021
Alquist, R., Guénette, J.-D., 2014. A blessing in disguise: The implications of high global oil prices for the North
American market. Energy Policy 64, 49–57. doi:10.1016/j.enpol.2013.07.002
Alquist, R., Kilian, L., Vigfusson, R.J., 2013. Chapter 8 - Forecasting the Price of Oil, in: Graham Elliott and Allan
Timmermann (Ed.), Handbook of Economic Forecasting. Elsevier, pp. 427–507.
Bardi, U., 2009. Peak oil: The four stages of a new idea. Energy 34, 323–326. doi:10.1016/j.energy.2008.08.015
Benes, J., Chauvet, M., Kamenik, O., Kumhof, M., Laxton, D., Mursula, S., Selody, J., 2012. The Future of Oil:
Geology versus Technology. IMF Working Paper.
Bentley, R.W., Mannan, S.A., Wheeler, S.J., 2007. Assessing the date of the global oil peak: The need to use 2P
reserves. Energy Policy 35, 6364–6382. doi:10.1016/j.enpol.2007.08.001
BGR, 2008. Reserves, Resources and Availability of Energy Resources 2007. German Federal Institute of
Geosciences and Natural Resources.
Bloomberg, 2013. Petrobras said to struggle pumping oil at Brazil’s largest prospect [WWW Document]. Fuel
Fix. URL http://fuelfix.com/blog/2013/05/20/petrobras-said-to-struggle-pumping-oil-at-brazils-largestprospect/ (accessed 11.11.13).
Bolonkin, A., Friedlander, J., Neumann, S., 2014. Innovative unconventional oil extraction technologies. Fuel
Process. Technol. 124, 228–242. doi:10.1016/j.fuproc.2014.01.024
BP, 2012. The recovery factor. BP Mag.
BP, 2013. BP Statistical Review 2013.
BP, 2014. Energy Outlook 2035.
Büyükşahin, B., Lee, T.K., Moser, J.T., Robe, M.A., 2013. Physical markets, paper markets and the WTI-Brent
spread. Energy J. 34. doi:http://dx.doi.org/10.5547/ 01956574.34.3
Campbell, C.J., 2013a. Recognising the Second Half of the Oil Age. Environ. Innov. Soc. Transit., Energy,
materials and growth: A homage to Robert Ayres 9, 13–17. doi:10.1016/j.eist.2013.08.004
Campbell, C.J., 2013b. Campbell’s Atlas of Oil and Gas Depletion. Springer New York.
Campbell, C.J., Laherrère, J.H., 2008. The End of Cheap Oil. Sci. Am. 278, 78–83.
CERI, 2013. Canadian Oil Sands Supply Costs and Development Projects (2012-2046). Canadian Energy Research
Institute.
Chapman, I., 2014. The end of Peak Oil? Why this topic is still relevant despite recent denials. Energy Policy 64,
93–101. doi:10.1016/j.enpol.2013.05.010
Chew, K.J., 2014. The future of oil: unconventional fossil fuels. Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 372,
20120324. doi:10.1098/rsta.2012.0324
Chu, H.-P., 2012. Oil consumption and output: What causes what? Bootstrap panel causality for 49 countries.
Energy Policy, Renewable Energy in China 51, 907–915. doi:10.1016/j.enpol.2012.09.050
De Almeida, P., Silva, P.D., 2009. The peak of oil production—Timings and market recognition. Energy Policy 37,
1267–1276. doi:10.1016/j.enpol.2008.11.016
De Castro, C., Miguel, L.J., Mediavilla, M., 2009. The role of non conventional oil in the attenuation of peak oil.
Energy Policy 37, 1825–1833. doi:10.1016/j.enpol.2009.01.022
De Oliveira, R., 2011. Dealing with plenty: Brazil in the era of surplus oil. University of Illinois, Urbana, Illinois.
EIA, 2008. Annual Energy Review 2007 (No. EIA-0384(2007)). U.S. Energy Information Administration,
Washington D.C.
EIA, 2012. Annual Energy Review 2012 (No. EIA-0383(2012)). U.S. Energy Information Administration,
Washington D.C.
EIA, 2013a. Country Analysis - Venezuela.
EIA, 2013b. Country Analysis - Brazil.
EIA, 2013c. International Energy Statistics - Natural Gas Plant Liquids Production [WWW Document]. URL
http://www.eia.gov/cfapps/ipdbproject/iedindex3.cfm?tid=5&pid=58&aid=1&cid=ww,&syid=1980&ey
id=2012&unit=TBPD (accessed 12.10.13).
EIA, 2014. Annual Energy Review 2014 (No. EIA-0383(2014)). U.S. Energy Information Administration,
Washington D.C.
60
Fic, J., Pedersen, P.K., 2013. Reservoir characterization of a “tight” oil reservoir, the middle Jurassic Upper
Shaunavon Member in the Whitemud and Eastbrook pools, SW Saskatchewan. Mar. Pet. Geol. 44, 41–
59. doi:10.1016/j.marpetgeo.2013.03.013
Gately, D., Al-Yousef, N., Al-Sheikh, H.M.H., 2012. The rapid growth of domestic oil consumption in Saudi Arabia
and the opportunity cost of oil exports foregone. Energy Policy 47, 57–68.
doi:10.1016/j.enpol.2012.04.011
Gately, D., Al-Yousef, N., Al-Sheikh, H.M.H., 2013. The rapid growth of OPEC′s domestic oil consumption. Energy
Policy 62, 844–859. doi:10.1016/j.enpol.2013.07.044
Gowdy, J., Juliá, R., 2007. Technology and petroleum exhaustion: Evidence from two mega-oilfields. Energy 32,
1448–1454. doi:10.1016/j.energy.2006.10.019
Grant, J., Chazan, G., 2013. Petronas pulls out of Venezuela heavy oil project Carabobo-1. Financ. Times.
Hirsch, R.L., 2008. Mitigation of maximum world oil production: Shortage scenarios. Energy Policy 36, 881–889.
doi:10.1016/j.enpol.2007.11.009
Holland, S.P., 2013. Economics of Peak Oil, in: Shogren, J.F. (Ed.), Encyclopedia of Energy, Natural Resource, and
Environmental Economics. Elsevier, Waltham, pp. 146–150.
Höök, M., 2009. Depletion and decline curve analysis in crude oil production. Uppsala University, Global Energy
Systems.
Höök, M., 2014. Depletion rate analysis of fields and regions: A methodological foundation. Fuel 121, 95–108.
doi:10.1016/j.fuel.2013.12.024
Höök, M., Davidsson, S., Johansson, S., Tang, X., 2014a. Decline and depletion rates of oil production: a
comprehensive investigation. Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 372, 20120448.
doi:10.1098/rsta.2012.0448
Höök, M., Fantazzini, D., Angelantoni, A., Snowden, S., 2014b. Hydrocarbon liquefaction: viability as a peak oil
mitigation strategy. Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 372, 20120319.
doi:10.1098/rsta.2012.0319
Höök, M., Hirsch, R., Aleklett, K., 2009a. Giant oil field decline rates and their influence on world oil production.
Energy Policy 37, 2262–2272. doi:10.1016/j.enpol.2009.02.020
Höök, M., Söderbergh, B., Jakobsson, K., Aleklett, K., 2009b. The Evolution of Giant Oil Field Production
Behavior. Nat. Resour. Res. 18, 39–56. doi:10.1007/s11053-009-9087-z
Hu, Y., Monroy, C.R., 2012. Chinese energy and climate policies after Durban: Save the Kyoto Protocol. Renew.
Sustain. Energy Rev. 16, 3243–3250. doi:10.1016/j.rser.2012.02.048
Hubbert, M.K., 1956. Nuclear Energy and the Fossil Fuels.
Hughes, L., Rudolph, J., 2011. Future world oil production: growth, plateau, or peak? Curr. Opin. Environ.
Sustain. 3, 225–234. doi:10.1016/j.cosust.2011.05.001
IEA, 2004. World Energy Outlook 2004. Organisation for Economic Co-operation and Development, Paris.
IEA, 2005. World Energy Outlook 2005. Organisation for Economic Co-operation and Development, Paris.
IEA, 2006. World Energy Outlook 2006. Organisation for Economic Co-operation and Development, Paris.
IEA, 2008. World Energy Outlook 2008. Organisation for Economic Co-operation and Development, Paris.
IEA, 2009. World Energy Outlook 2009. Organisation for Economic Co-operation and Development, Paris.
IEA, 2010a. World Energy Outlook 2010. Organisation for Economic Co-operation and Development, Paris.
IEA, 2010b. Natural Gas Liquids - Supply Outlook 2008-2015, Oil Market Report. International Energy Agency Oil Industry and Market Division, Paris.
IEA, 2011. World Energy Outlook 2011. Organisation for Economic Co-operation and Development, Paris.
IEA, 2012a. Key World Energy Statistics 2012. International Energy Agency, Paris.
IEA, 2012b. World Energy Outlook 2012. Organisation for Economic Co-operation and Development, Paris.
IEA, 2013a. World Energy Model Documentation - 2013 Version.
IEA, 2013b. World Energy Outlook 2013. Organisation for Economic Co-operation and Development, Paris.
Jackson, P.M., Smith, L.K., 2014. Exploring the undulating plateau: the future of global oil supply. Philos. Trans.
R. Soc. Math. Phys. Eng. Sci. 372, 20120491. doi:10.1098/rsta.2012.0491
Jakobsson, K., Bentley, R., Söderbergh, B., Aleklett, K., 2012. The end of cheap oil: Bottom-up economic and
geologic modeling of aggregate oil production curves. Energy Policy 41, 860–870.
doi:10.1016/j.enpol.2011.11.073
Jakobsson, K., Söderbergh, B., Höök, M., Aleklett, K., 2009. How reasonable are oil production scenarios from
public agencies? Energy Policy 37, 4809–4818. doi:10.1016/j.enpol.2009.06.042
Jakobsson, K., Söderbergh, B., Snowden, S., Aleklett, K., 2014. Bottom-up modeling of oil production: A review
of approaches. Energy Policy 64, 113–123. doi:10.1016/j.enpol.2013.09.043
Kato, H., 2003. World Energy Model 2002.
61
Kent, S., 2013. Iraq close to agreeing cuts in oil output targets [WWW Document]. Wall Str. J. URL
http://online.wsj.com/news/articles/SB10001424052702304854804579236050689378182 (accessed
12.9.13).
Kerschner, C., Prell, C., Feng, K., Hubacek, K., 2013. Economic vulnerability to Peak Oil. Glob. Environ. Change
23, 1424–1433. doi:10.1016/j.gloenvcha.2013.08.015
Kohl, W.L., 2010. Consumer Country Energy Cooperation: The International Energy Agency and the Global
Energy Order, in: Goldthau, A., Witte, J.M. (Eds.), Global Energy Governance: The New Rules of the
Game. Brookings Institution Press, Washington D.C, pp. 195–220.
Kumar, N., 2011. Market turmoil as IEA warns “age of cheap oil is over” [WWW Document]. The Independent.
URL http://www.independent.co.uk/news/business/news/market-turmoil-as-iea-warns-age-of-cheapoil-is-over-2230581.html (accessed 5.19.14).
MacCracken, M., 2001. Prediction versus Projection - Forecast versus Possibility. WeatherZine.
McGlade, C.E., 2012. A review of the uncertainties in estimates of global oil resources. Energy 47, 262–270.
doi:10.1016/j.energy.2012.07.048
McGlade, C.E., 2013. Comment on the “Uppsala critique”. Energy Policy. doi:10.1016/j.enpol.2013.11.036
McKelvey, V.E., 1972. Mineral Resource Estimates and Public Policy: Better methods for estimating the
magnitude of potential mineral resources are needed to provide the knowledge that should guide the
design of many key public policies. Am. Sci. 60, 32–40.
Miller, R.G., 2011. Future oil supply: The changing stance of the International Energy Agency. Energy Policy 39,
1569–1574. doi:10.1016/j.enpol.2010.12.032
Miller, R.G., Sorrell, S.R., 2014. The future of oil supply. Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 372,
20130179. doi:10.1098/rsta.2013.0179
Monbiot, G., 2008. When will the oil run out? The Guardian.
Moryadee, S., Gabriel, S.A., Avetisyan, H.G., 2014. Investigating the potential effects of U.S. LNG exports on
global natural gas markets. Energy Strategy Rev., Sustainable Energy System Changes 2, 273–288.
doi:10.1016/j.esr.2013.12.004
Muggeridge, A., Cockin, A., Webb, K., Frampton, H., Collins, I., Moulds, T., Salino, P., 2014. Recovery rates,
enhanced oil recovery and technological limits. Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 372,
20120320. doi:10.1098/rsta.2012.0320
OPEC, 2007. World Oil Outlook 2007. Organization of the Petroleum Exporting Countries, Vienna.
OPEC, 2013. World Oil Outlook 2013. Organization of the Petroleum Exporting Countries, Vienna.
Ouellette, A., Rowe, A., Sopinka, A., Wild, P., 2014. Achieving emissions reduction through oil sands
cogeneration in Alberta’s deregulated electricity market. Energy Policy.
doi:10.1016/j.enpol.2014.04.020
Riedel, B., 2014. Al Qaeda’s Next Comeback Could Be Afghanistan And Pakistan [WWW Document]. Dly. Beast.
URL http://www.thedailybeast.com/articles/2014/01/13/al-qaeda-s-next-comeback-could-beafghanistan-and-pakistan.html (accessed 1.14.14).
Riveras, I., Luna, D., 2010. Huge Brazil oil field will bear Lula’s name. Reuters.
Robelius, F., 2007. Giant Oil Fields - The Highway to Oil : Giant Oil Fields and their Importance for Future Oil
Production. Uppsala University, Department of Nuclear and Particle Physics.
Rodriguez, C.M., 2012. Pemex Oil Production Falls as Cantarell Hits Record Low [WWW Document]. Bloomberg.
URL http://www.bloomberg.com/news/2012-04-30/pemex-oil-production-falls-as-cantarell-hitsrecord-low.html (accessed 4.23.14).
Rystad Energy AS, 2014. Rystad Energy UCube Database. Rystad Energy AS.
Salameh, M.G., 2004. Oil Crises, Historical Perspective, in: Cleveland, C.J. (Ed.), Encyclopedia of Energy. Elsevier,
New York, pp. 633–648.
Saraiva, T.A., Szklo, A., Lucena, A.F.P., Chavez-Rodriguez, M.F., 2014. Forecasting Brazil’s crude oil production
using a multi-Hubbert model variant. Fuel 115, 24–31. doi:10.1016/j.fuel.2013.07.006
Schenk, C.J., Brownfield, M.E., Charpentier, R.R., Cook, T.A., Gautier, D.L., Higley, D.K., Kirschbaum, M.A., Klett,
T.R., Pitman, J.K., Pollastro, R.M., Tennyson, M.E., Wandrey, C.J., Whidden, K.J., 2012. An Estimate of
Undiscovered Conventional Oil and Gas Resources of the World, 2012. USGS Fact Sheet 2012-3028.
Schindler, J., Zittel, W., 2008. Crude oil: the supply outlook, EWG-Series. Energy Watch Group, Berlin, Germany.
Sieminski, A., 2013. Outlook for shale gas and tight oil development in the U.S.
Simmons, M.R., 2002. The World’s Giant Oilfields.
Söderbergh, B., Robelius, F., Aleklett, K., 2007. A crash programme scenario for the Canadian oil sands industry.
Energy Policy 35, 1931–1947. doi:10.1016/j.enpol.2006.06.007
62
Sorrell, S., Miller, R., Bentley, R., Speirs, J., 2010. Oil futures: A comparison of global supply forecasts. Energy
Policy, Special Section on Carbon Emissions and Carbon Management in Cities with Regular Papers 38,
4990–5003. doi:10.1016/j.enpol.2010.04.020
Sorrell, S., Speirs, J., 2014. Using growth curves to forecast regional resource recovery: approaches, analytics
and consistency tests. Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 372, 20120317.
doi:10.1098/rsta.2012.0317
Sorrell, S., Speirs, J., Bentley, R., Brandt, A., Miller, R., 2009. Global oil depletion - an assessment of the
evidence for a near-term peak in global oil production.
Sorrell, S., Speirs, J., Bentley, R., Miller, R., Thompson, E., 2012. Shaping the global oil peak: A review of the
evidence on field sizes, reserve growth, decline rates and depletion rates. Energy 37, 709–724.
doi:10.1016/j.energy.2011.10.010
Troner, A., 2013. Natural Gas Liquids in the Shale Revolution. James A. Baker III Institute for Public Policy.
Tsoskounoglou, M., Ayerides, G., Tritopoulou, E., 2008. The end of cheap oil: Current status and prospects.
Energy Policy 36, 3797–3806. doi:10.1016/j.enpol.2008.05.011
USGS, 2013. National Oil and Gas Assessment 2013 Assessment Updates. United States Geological Survey.
Van de Graaf, T., 2012. Obsolete or resurgent? The International Energy Agency in a changing global landscape.
Energy Policy, Special Section: Frontiers of Sustainability 48, 233–241.
doi:10.1016/j.enpol.2012.05.012
Weston, J.K., 2014. Their Fight…But Our Legacy: The New Battle for Fallujah [WWW Document]. Dly. Beast. URL
http://www.thedailybeast.com/articles/2014/01/12/their-fight-but-our-legacy-the-new-battle-forfallujah.html (accessed 1.14.14).
Wood, D.A., Nwaoha, C., Towler, B.F., 2012. Gas-to-liquids (GTL): A review of an industry offering several routes
for monetizing natural gas. J. Nat. Gas Sci. Eng. 9, 196–208. doi:10.1016/j.jngse.2012.07.001
63
9 Appendix I
The resulting tables of the complete track record analysis.
Table 9: Track record of IEA world oil supply scenarios, relative to IEA actual production, and the average error by numbers
of years from scenario start
World oil supply scenarios track record, relative terms [%]
WEO2000
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
0.0
0.0
0.0
7.1
7.1
0.0
7.3
7.3
0.0
0.0
8.0
8.0
0.6
0.6
0.0
4.1 5.3
4.1 5.3
-3.2 -2.1
-3.2 -2.1
-3.3 -1.9
0.0 1.5
0.0
5.9
5.9
-1.7
-1.7
-1.1
2.3
0.6
0.0
8.3
8.3
0.5
0.5
1.5
5.0
3.1
2.6
0.0
9.0
9.0
1.0
1.0
2.4
5.9
3.8
3.4
0.4
0.0
13.7 12.0 13.0 11.7
13.7 12.0 13.0 11.7
5.3 3.6 4.2 2.8
5.3 3.6 4.2 2.8
7.1 5.7 6.5 5.1
10.7 9.3 9.6 7.7
8.4 6.8 7.5 6.1
8.0 6.6 7.1 5.6
4.5 2.8 3.1 1.5
3.2 0.6 0.1 -2.3
0.0 -1.9 -1.8 -3.6
0.0 0.1 -1.8
0.0 -1.6
0.0
1
1.6
2.0
1.5
2
3.1
3.6
2.2
3
4.4
4.8
2.2
4
4.4
4.8
2.7
5
4.4
4.8
1.5
6
5.3
8
6.7
9
7.1
10
8.2
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
7
6.0
2010
2011
2012
Table 10: Track record of IEA conventional oil production scenarios, relative to IEA actual production, and the average error
by numbers of years from scenario start
Conventional oil production scenarios track record, relative terms [%]
WEO2000
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
0.0
0.9
0.0
2.2
1.4
0.0
4.5
3.7
1.5
0.0
3.6
3.0
-0.1
-1.4
0.0
1.8
1.2
-2.6
-3.6
-2.3
0.0
2.5
2.1
-2.5
-3.3
-2.1
0.6
0.0
3.7
3.4
-2.0
-2.7
-1.6
1.5
0.6
0.0
6.1
5.9
-0.3
-0.8
0.2
3.7
2.4
2.0
0.0
7.5
7.3
0.4
0.0
1.0
4.8
3.2
3.0
0.3
0.0
13.3 12.4 14.9 14.9
13.2 12.4 14.9 14.9
5.1 3.7 5.6 5.1
5.0 3.7 5.6 5.1
5.9 4.6 6.6 6.2
10.2 9.1 10.6 9.6
8.3 6.9 8.1 6.8
8.2 7.0 8.8 8.2
4.7 2.9 4.3 3.4
3.6 1.1 1.8 0.3
0.0 -2.0 -1.0 -2.1
0.0 1.1 0.0
0.0 -1.3
0.0
1
1.4
1.3
1.7
2
2.1
2.3
1.7
3
3.4
3.7
2.3
4
3.5
3.8
2.3
5
3.9
4.2
3.4
6
4.7
7
5.3
8
6.4
9
7.2
10
9.1
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
i
2010
2011
2012
Table 11: Track record of OPEC oil production scenarios, relative to IEA actual production, and the average error by numbers
of years from scenario start
OPEC oil production scenarios track record, relative terms [%]
WEO2000
1999
2000
2001
2002
2010
2011
2012
0.0
-0.5
0.0
5.2
5.6
0.0
14.8 11.8 12.2 11.7 8.5 11.9 14.3 28.1 26.7
15.2 12.1 12.4 11.9 8.7 12.0 14.3 28.2 26.7
6.8 2.0 0.4 -1.8 -6.2 -4.8 -4.3 5.8 3.2
0.0 -3.6 -4.3 -5.5 -9.0 -6.9 -5.7 5.0 3.2
0.0 -2.4 -5.3 -10.3 -9.7 -9.9 -1.2 -4.3
0.0 -1.6 -5.5 -3.6 -2.6 8.2 6.0
0.0 -4.9 -3.8 -3.6 6.1 3.2
0.0
3.1 5.2 18.0 16.7
0.0 1.8 13.8 12.3
0.0 10.4 7.6
0.0 -1.6
0.0
28.5
28.5
4.6
4.6
-2.1
6.3
4.0
16.8
12.5
6.5
-1.7
-0.9
0.0
26.7
26.7
3.1
3.1
-2.7
3.7
2.0
13.7
9.6
2.6
-4.4
-4.4
-4.0
0.0
1
3.4
3.6
3.7
2
6.2
5.8
6.9
3
8.9
8.5
7.7
4
8.5
8.8
7.5
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
2003
2004
2005
2006
2007
2008
2009
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
5
9.7
9.4
9.6
6
7.0
7
6.1
8
6.8
9
10
10.6 15.6
Table 12: Track record of Russia oil production scenarios, relative to IEA actual production, and the average error by
numbers of years from scenario start. Note that WEO2009 excludes NGL, affecting the analysis.
Russia oil production scenarios track record, relative terms [%]
1999 2000
WEO2000
WEO2001
WEO2002
WEO2003
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
0.0 -5.7 -11.4 -17.1
0.0 -6.6 -13.2
0.0 -5.6
0.0
-24.7
-21.7
-13.5
-9.2
0.0
-28.6
-26.2
-17.3
-13.9
-3.7
0.0
-27.7
-25.6
-15.5
-12.6
-0.9
2.7
0.0
-30.5
-28.9
-18.2
-16.0
-3.4
0.0
-2.5
0.0
-32.4
-31.2
-19.8
-18.2
-4.7
-1.5
-3.8
-1.7
0.0
-27.1
-26.4
-13.0
-11.8
4.1
7.4
5.1
6.8
6.7
0.0
-31.3
-30.9
-17.3
-16.8
-0.6
2.5
0.4
1.7
-0.2
-7.3
0.0
-32.4
-32.4
-18.1
-18.1
-1.0
1.9
0.0
1.0
-2.7
-10.3
-2.9
0.0
-32.3
-32.3
-18.5
-18.5
-1.7
1.1
-0.8
0.4
-3.3
-11.6
-3.8
-1.1
0.0
-32.1
-32.1
-18.9
-18.9
-2.4
0.4
-1.5
-0.2
-3.9
-12.8
-4.7
-2.2
-1.2
0.0
WEO2004
WEO2005
WEO2006
2001
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
1
4.5
4.7
3.8
2
3
4
5
6
7
8
9
10
6.7 9.7 11.2 13.0 11.5 13.6 16.0 19.4 25.3
7.9 10.0 12.0 12.2
4.2 6.3 8.1 3.9
ii
Table 13: Track record of Brazil oil production scenarios, relative to IEA actual production, and the average error by numbers
of years from scenario start
Brazil oil production scenarios track record, relative terms [%]
WEO2000
1999
2000
2001
2002
2003
2004
2005
2006
2007
0.0
-1.9
0.0
2.0
3.5
0.0
-1.2
0.0
-3.5
2.9
4.0
0.2
13.9
14.9
10.5
7.3
8.0
3.8
7.7
8.3
4.0
14.1 20.5 14.2 14.3 12.7 16.4
14.5 20.8 14.4 14.3 12.7 16.4
9.9 15.9 9.7 9.5 8.6 12.7
0.0
12.3 7.1
0.0 -2.0
0.0
8.7
1.9
5.6
0.0
16.2
11.1
13.0
8.3
0.0
WEO2001
WEO2002
2008
2009
2010
2011
2012
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
23.8
20.4
20.4
16.7
11.8
0.0
WEO2010
18.2
16.7
15.0
12.5
11.3
-0.7
0.0
WEO2011
19.0
19.0
15.9
14.3
16.1
3.4
4.0
0.0
WEO2012
17.3
17.3
16.7
14.5
20.5
7.1
7.6
3.6
0.0
WEO2013
Year
Average
Pre 2008
Post 2008
1
5.1
5.3
5.4
2
6.8
5.8
8.5
20.9
20.9
22.7
20.0
30.1
15.6
15.9
11.8
6.8
0.0
3
4
5
6
7
8
9
10
9.9 13.7 15.6 14.2 16.2 16.6 16.3 12.4
9.8 12.5 13.8
13.0 18.0 30.1
Table 14: Track record of Iraq oil production scenarios, relative to IEA actual production, and the average error by numbers
of years from scenario start
Iraq oil production scenarios track record, relative terms [%]
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
54.2 44.1
36.0
WEO2000
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
0.0 22.2 20.1 23.8 16.7 20.0 33.3 26.7 21.3
0.0 -6.0 -6.7 -15.0 -15.2 -8.3 -14.1 -18.7
WEO2007
0.0
WEO2008
WEO2009
-7.8
0.0
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
1
2
3
4
5
6
7
8
9
10
7.6 8.5 8.7 11.5 13.2 23.7 22.7 37.8 44.1 36.0
14.1 13.4 19.4 15.9 14.2
5.0 6.1 1.6 7.1 11.3
iii
-7.0
-0.6
0.0
1.6
7.1
11.8
0.0
-5.6
-1.6
6.2
2.2
0.0
-11.3
-8.6
1.7
4.0
2.5
0.0
Table 15: Track record of Canadian oil sands production scenarios, relative to IEA actual production, and the average error
by numbers of years from scenario start
Canadian oil sands production scenarios track record, relative terms [%]
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
0.0
5.5
10.0 13.8
5.8
16.9 18.1 28.3 20.0 22.8 18.3
0.0
0.0
0.0
WEO2000
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
16.7 23.1 38.5 33.3 37.5 33.3
12.5 15.4 26.9 20.0 25.0 22.2
0.0 12.5 32.7 32.5 40.6 39.6
0.0 8.8 1.9 2.7 -2.4
0.0 -1.1 4.2 2.8
0.0 6.2 5.6
0.0 -1.4
0.0
1
2
3
4
5
6
7
8
9
10
12.3 17.0 21.5 28.7 25.9 30.8 20.0 22.8 18.3
13.4 18.4 22.1 25.1 27.8
11.9 16.2 21.0 32.3 22.2
Table 16: Track record of Saudi Arabia oil production scenarios, relative to IEA actual production, and the average error by
numbers of years from scenario start
Saudi Arabia oil production scenarios track record, relative terms [%]
1999
2000
2001
2002
2003
2004
2005 2006 2007
2008
2009
2010
2011
2012
0.0
0.0
0.0
3.4
2.9
0.0
8.8
7.8
6.6
0.0
23.4
21.7
22.3
16.6
0.0
20.8
18.8
21.1
17.2
-1.6
0.0
18.5 10.0 7.4
16.0 7.6
5.0
20.0 10.3 6.7
17.8 10.8 9.6
-3.1 -10.6 -13.1
-1.3 -8.7 -11.1
0.0 -9.0 -12.8
0.0
-5.6
0.0
1
2
3
4
5
6
7
5.5 10.8 15.2 17.2 14.2 10.9 7.5
3.2 11.2 17.2 20.7 15.7
6.8 10.5 13.1 11.9 9.6
8
7.4
9
10
WEO2000
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
iv
Table 17: Track record of Conventional crude oil production scenarios, relative to IEA actual production, and the average
error by numbers of years from scenario start
Conventional crude oil production scenarios track record, relative terms [%]
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
0.0
2.3
3.6
4.6
10.0
9.2
11.9 11.7
0.0
0.1
0.0
4.4
3.6
0.0
2.8
1.2
-1.9
0.0
4.5
2.1
-0.6
1.3
0.0
9
10
WEO2000
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
1
1.7
2.3
1.6
2
2.0
3.6
1.6
3
2.8
4.6
2.2
4
5.0
10.0
2.5
5
6.5
9.2
3.7
6
7
11.9 11.7
8
3.7
0.5
-1.8
0.1
-1.3
0.0
Table 18: Track record of NGL production scenarios, relative to IEA actual production, and the average error by numbers of
years from scenario start
NGL production scenarios track record, relative terms [%]
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
0.0
-6.7 -5.7 -5.6 -2.8 -6.9 -8.4
-10.6
WEO2000
WEO2001
WEO2002
WEO2003
WEO2004
WEO2005
WEO2006
WEO2007
WEO2008
0.0
1.7
0.0
9
10
WEO2009
WEO2010
WEO2011
WEO2012
WEO2013
Year
Average
Pre 2008
Post 2008
1
2.8
6.7
2.0
2
3.3
5.7
2.7
3
3.1
5.6
2.3
4
2.2
2.8
2.0
5
4.4
6.9
1.9
6
8.4
7
10.6
v
8
6.3
4.1
0.0
3.1 2.9
0.7 0.2
-2.7 -2.8
0.0 -0.2
0.0
1.9
-1.0
-3.5
-1.1
-1.4
0.0
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertising