Running out of and into oil:
Analyzing global oil depletion to 2050
David L. Greene
The Q Group
Spring 2005 Seminar
Key Largo, Florida
April 4, 2005
Lost our bearings?
“Oil prices have broken $50 a barrel. Soaring Chinese demand is
rocking energy markets. Climate-destabilising carbon emissions
grow apace. New energy investments will cost over $500 billion per
year.
“As the World Energy Outlook 2004 goes to press, the energy world
has lost its bearings.”
“The price is assumed to remain flat ($22/bbl) until 2010, and then to
begin to climb steadily to $29 in 2030.” p. 39
“OPEC countries, mainly in the Middle East, will meet most of the
increase in global demand. By 2030, OPEC will supply over half of the
world’s oil needs – an even larger share than in the 1970s.” p. 32
International Energy Agency, 2004.
Transportation runs on oil.
The emergence of OPEC and consequent oil price shocks in the
1970s and 80s temporarily reversed the global trend of increasing
petroleum use for increasing global mobility.
WHERE WILL THE OIL COME FROM?
The graph below was not presented by Colin Campbell or Jean Laherrere, but
Rex Tillerson, President of Exxon Mobil Corporation (3/11/04).
One solution: OPEC will provide.
Why should OPEC do this?
Because of the inelasticity of world oil
demand and supply…
“The payoffs to OPEC are relatively
insensitive to faster output growth;
aggressive output expansion yields
slightly lower payoffs than just maintaining
current market share.”
D. Gately, The Energy Journal, 2004.
The U.S. Energy Information Administration says
that OPEC will increase production 50% by 2025,
spending its patrimony for no profit.
“For OPEC members, cumulative production of almost
280 billion barrels in the high A world oil price case is
projected to bring in $9.9 trillion, as compared with
cumulative production of 343 billion barrels and
revenues of $9.7 trillion in the reference case.” AEO
2005, p. 46.
“Undiscounted cumulative revenues from OPEC
member country production in the high B world oil
price case exceed those in the reference and high A
world oil price cases, despite lower production;…” (255
billion barrels)
AEO 2005, p. 47.
Are we running out of oil?
{
“Pessimists” aka “Geologists”
z
z
z
z
{
Geology rules
Discoveries lag production
Peaking, not running out matters
Expect peak by 2010
“Optimists” aka “Economists”
z
z
z
z
Economics & technology rule
Rate of technological progress will exceed rate of
depletion
Market system will provide incentives to expand,
redefine resources
Stone age did not end for lack of stones
Take optimists’ view but quantify.
{
{
{
{
{
How much oil remains to be discovered?
How fast might technology increase
recovery rates?
How much will reserves grow?
How fast will technology reduce the cost
of unconventional sources?
How much unconventional oil is there and
where is it?
The optimists’ approach is… optimistic.
{
No Hubbert’s curves
z
z
{
RESOURCE/Production ratio limits
expansion of production
z
z
{
No geologic constraints on production rates
Costs do rise with depletion, however
Analogous to a limit based on life of capital
No explicit calculation of capital investment
No environmental/social/political
constraints on production
z
ANWAR, offshore, etc. fair game
What is oil?
{
Conventional Oil
z
z
{
Unconventional Oil
z
{
Liquid hydrocarbons of light and medium
gravity and viscosity, in porous and
permeable reservoirs.
Plus enhanced recovery and NGLs
Deposits of density > water (heavy oil),
viscosities > 10,000 cP (oil sands) and tight
formations (shale oil).
Liquid fuels can be made from coal or
natural gas (not considered here).
Do we know how much oil there is?
In 2000 the USGS published a major assessment
of world oil resources, including uncertainty and
technological progress.
(Billions of Barrels)
Speculative Resources
Estimated Additional
Reserve Growth
Proved Reserves
= Undiscovered 50th to 5th percentile
= Undiscovered 50th percentile
There is even greater uncertainty about
unconventional oil resources, but regions seem to
divide into oil sand/heavy oil or shale oil.
(1 Gtoe = 7.33 billion bbls, 20.1 mmbd)
MEA = Middle East + North Africa
Pessimists dispute the USGS estimates with the
following arguments:
{
OPEC members overstate proved reserves
{
Reserve growth methodology biased
{
Range of uncertainty exaggerated
{
Unconventional resources also much smaller than
implied by my estimates
Campbell, 2003.
A resource accounting model was constructed to simulate oil
resource depletion, expansion and transition under various
scenarios. It does not include Hubbert curves. If anything, its
rules are optimistic.
World energy scenarios were derived from existing
projections. The Reference Scenario represents
“business as usual”.
Most future growth of energy use is expected in
the developing world (2.7%/yr v. 1%/yr).
The average growth of world oil use from 2000
to 2050 is 1.9%/yr.
An “ecologically driven” scenario foresees only
an 0.8% annual growth in energy use.
In this scenario, there is a demand-driven peak
in oil use.
Six depletion/transition scenarios were constructed.
{
Two IIASA/WEC scenarios
{
Three EIA Int. Energy Outlook to 2020
{
Two DOE/NRCan NA transport projections
{
{
Three sources of conventional oil resource
estimates
Three unconventional oil estimates
A risk analysis was carried out, defining key
parameters as random variables.
Reference/USGS: non-Middle East oil production
peaks by 2030 with 90% probability.
Reference/Rogner: Non-MEA peak likely
anytime 2010-2040.
If Campbell’s estimates are correct, the non-MEA
peak will occur before 2010.
The most important determinant of the date of
peaking is…how much oil there is.
From 2.7 Gtoe in 2001, non-MEA oil production is
estimated to increase substantially.
Lower R/P ratios, more oil resources, slower
growth of MEA production all raise the level of
peak ROW output.
The total world oil production peak could be
significantly later, perhaps after 2050.
The world peaking date depends strongly on
the rate of expansion of Middle East production.
Under median assumptions, unconventional oil
production must expand rapidly after 2020.
Using the upper range of values of the 5
factors that most strongly influence the world
peaking date yields a broad, flat ROW curve.
Slowing the growth of MEA production raises
prices and further delays the ROW peak.
The price estimates of my model are not predictions. They
reflect optimistic assumptions about supply technologies.
Their purpose is to regulate the market mechanisms by which
unconventional resources are introduced.
The optimism of the model is reflected in
increasing US production to 2020.
The Middle East could maintain a dominant
position through 2050.
Rapid expansion of heavy oil and oil sands is
needed to allow world oil use to continue to grow.
The ability to produce vast quantities of shale oil (or
liquids from coal) is even more uncertain.
Conclusions
{
{
{
{
{
Rate of production likely
to decrease after 2020 in
any case
Caveats
{
Transition to
unconventional oil may be
rapid: 7-9%/yr growth
{
First supplies from
Venezuela, Canada,
Russia
{
Present trends imply ROW
oil peak 2010-30
{
Vast quantities of shale oil
(or coal, NG) may be
needed before 2050
{
Model doesn’t include
geologic constraints on
production rates; relies on
target resource-toproduction ratios
Does not include
environmental or political
constraints
Does not include gas- or
coal-to-liquids
Unconventional oil
resource estimates weak
Scenario, not market
equilibrium based
modeling of oil demand
T hank you for your
interest.
For a copy of the report:
Visit http://www-cta.ornl.gov/cta/Publications/Publications_2003.html
Or contact David L. Greene at:
dlgreene@ornl.gov
EIA used a few simple assumptions to produce a
range of peak year estimates with implausible
transitions and ignoring OPEC.
The model predicts that production may peak before
proved reserves (caveat).
“Never mind that man behind the curtain.”
(Wizard of Oz)
OPEC is an imperfect, partially monopolistic
cartel. (von Stackelberg, 1952)
{
{
{
{
It is imperfect, because it is a cartel of
sovereign states with differing interests.
It faces a (mostly) competitive fringe.
Some members of the fringe collude with the
cartel at times.
Five factors determine the cartel’s market
power.
z
z
z
z
z
Price elasticity of ROW supply
Price elasticity of World Demand
Market Share
Rate of growth of World Demand
Rate of growth/decline in ROW supply
In theory, a partial monopolist’s power (in a static market)
depends on three things.
P=
C
⎛ 1
⎞
1 + ⎜⎜
S (μ ( P) + 1)⎟⎟
⎝ β ( P)
⎠
β = price elasticity of world oil demand ( β < 0 )
S = Share of world oil market ( 0 < S < 1 )
µ= Rest-Of-World supply response ( -1 < µ < 0 )
Short- v. long-run elasticities differ by an order of magnitude!
Growing demand or declining supply amplify market power.
These economic parameters define the space in which the cartel
can operate, not what it will do.
Source: BP, EIA; OPEC Core is Venezuela, Iran, Iraq, Kuwait, Qatar, Saudi Arabia, UAE, Algeria & Libya.
Growing world oil demand and peaking ROW oil supply affect the
calculus in two ways.
{
Cartel effectiveness enhanced
z
{
It is easier to not expand capacity than to cut
production.
Cartel market power magnified
z
z
P=
Growing demand increases the inverse
elasticity term
Peaking ROW production diminishes the ROW
supply response term
C
⎛ (1 + ρ )
⎞
S ((1 − δ ) μ ( P ) + 1)⎟⎟
1 + ⎜⎜
⎝ β ( P)
⎠
, 0 < ρ < 1, 0 < δ < 1