OIL SUPPLY DISRUPTIONS: FREQUENCY AND IMPACT
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP83B00231R000200020001-0
Release Decision:
RIPPUB
Original Classification:
C
Document Page Count:
17
Document Creation Date:
December 20, 2016
Document Release Date:
January 24, 2007
Sequence Number:
1
Case Number:
Publication Date:
May 1, 1982
Content Type:
REPORT
File:
Attachment | Size |
---|---|
CIA-RDP83B00231R000200020001-0.pdf | 879.81 KB |
Body:
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Directorate of Confidential
Intelligence
Oil Supply Disruptions:
Frequency and Impact
An Intelligence Assessment
Confidential
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Directorate of Confidential
Intelligence
Oil Supply Disruptions:
Frequency and Impact
An Intelligence Assessment
Information available as of 26 April 1982
has been used in the preparation of this report.
This assessment was prepared b of
the Office of Global Issues of the
Office of East Asian Analysis, an
f the Office of Current Production and
Analytic Support. Comments and queries are
welcome and may be directed to the Chief, Ener
Markets Branch, OGI
The paper was coordinated with the National
Intelligence Council
Confidential
GI 82-10086
May 1982
25;
25X
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Confidential
Oil Supply Disruptions:
Frequency and Impact
0
Key Judgments The consensus forecast of a soft world oil market over the next two to three
years is causing complacency within major oil-consuming countries. Sever-
al forecasters have even predicted a collapse in oil prices, as well as the de-
mise of the OPEC cartel. While we expect considerable market softness
because of stagnant demand and adequate supplies, the frequency of
supply disruptions in the past gives little reason to place much faith in a
"no surprises" forecast.
Given present market conditions, only a major disruption of output from
Saudi Arabia would have significant impact on oil prices during the next
two to three years. After the mid-1980s, however, the production-capacity
cushion that now exists is likely to shrink as OECD economic growth
rebounds from the current recession. We are uncertain about how fast the
change would occur, but oil market vulnerability to moderate supply
disruptions could greatly increase. To avert panic buying normally associ-
ated with a significant supply disruption, consumers must be convinced
that adequate energy supplies will be available from surplus productive
capacity in other producing countries, from alternative fuels such as coal or
gas, or from stockpiles. Our analysis indicates that a release of oil stocks at
the onset of a crisis would sharply cut the runup in oil prices.
Confidential
GI 82-10086
May 1982
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Figure 1
Disruptions of World Oil Supplies, 1951-81
Iran I
Suez War
Syria
Six-Day War
Nigeria
Libya
Algeria
Lebanon l
October War
Lebanon II
Saudi Arabia
Iran II
Iran-Iraq War
Iran I
Suez War
Syria
Six-Day War
Nigeria
Libya
Algeria
Lebanon l
October War
Lebanon II
Saudi Arabia
Iran II
Iran-Iraq War
Em
0 5 10 15 20 25 30 35 40 45 50
Duration of Disruption in Months
I1 I I I I I I I I I
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Magnitude of Disruption in Million Barrels
per Day
5IM4111 1-111
1
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Confidential
Oil Supply Disruptions.
Frequency and Impact
0
Most forecasters generally agree that over the next
two to three years the oil market will remain stable,
with a significant decline in real oil prices and the
possibility of erosion in nominal prices. Given the
importance of oil and energy use in Western econo-
mies, such predictions imply that economic recovery
can safely proceed at a fairly rapid pace, uncon-
strained by concern over further shocks from oil price
hikes. These forecasts also assume that oil supplies are
not disrupted by political unrest in key exporting
countries or damage to key oilfield and export facili-
ties. A review of recent supply disruptions, however,
suggests that such events occur rather frequently.
0
The Historical Record
Since 1950, oil supplies from major exporting coun-
tries have been interrupted on 13 occasions (table 1
and figure 1). The duration of these disruptions
ranged from one month (a 1977 fire at a Saudi
Arabian gas-oil separation plant) to 44 months (na-
tionalization of Iranian oilfields in 1951). Supply
losses ranged from 300,000 barrels per day (b/d) in
the case of the 1976 Lebanese civil war, to 3.7 million
b/d as a result of the Iranian revolution in 1978-79.
In most cases, the disruption to supplies had little or
no measurable impact on prices. Oil companies were
able to switch to alternate sources with relative ease
because of the ample surplus capacity that existed
worldwide and the considerable flexibility companies
maintained in their distribution systems. Indeed, the
lesson learned from early disruptions-such as the
closings of the Suez Canal in 1956 and 1967-
reemphasized the need for large supply overhangs and
redundant distribution systems.
of oil more than doubled from about 19 million b/d
in 1960 to almost 40 million b/d by 1970, while
productive capacity grew by less than 80 percent.
? Oil trade expanded in line with the growth in
consumption; inventory requirements grew faster.
? US oil production peaked in 1970 and imports,
which had doubled in the 1960s, began to grow even
more rapidly.
? Producing country governments increased control
over oil liftings through such actions as nationaliza-
tion, reducing major oil companies' flexibility in
managing production and exploration.
Libya's move to reduce the authorized production
level of foreign oil companies in 1970, coincident with
major pipeline sabotage in Syria, marked the first
instance in which a supply disruption caused a sharp
rise in oil prices. The loss of about 1.3 million b/d for
a nine-month period resulted in a 25-percent rise in oil
prices.
In the 11 years since the 1970 Libyan dispute, seven
additional disruptions have occurred. Only two had an
impact on the oil market, but the effects of both were
severe:
? The 1973 Arab oil embargo following the October
Arab-Israeli war caused a loss of 1.6 million b/d in
oil supplies, more than tripled oil prices, and con-
tributed to the abrupt curtailment of gross national
product growth in the Organization for Economic
Cooperation and Development economies from 6.1
percent in 1973 to a negative 0.4 percent by 1975.
By the early 1970s, however, several changes had
occurred to make the oil market increasingly vulner-
able to supply disruptions:
? Growth in oil consumption began to outpace in-
creases in productive capacity; non-Communist use
? Supply losses resulting from the Iranian revolution
more than doubled oil prices between late 1978 and
early 1980. Moreover, OECD/GNP growth fell
from 3.6 percent in 1979 to 1.3 percent in 1980.r
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
March 1951-October 1954 Mossadegh and the Anglo-Iranian Oil Companies.
Iranian oilfields were nationalized on May 1, following two
months of unrest and strikes in the Abadan area. Major oil
companies boycotted Iranian oil on the world oil market and
instituted court actions to deter potential buyers.
November 1956-March 1957 Suez War.
Nasser closed the Suez Canal in the wake of the Anglo-French-
Israeli incursion. At the same time, the Iraqi Petroleum
Company (IPC) pipeline was damaged in Syria, and Saudi
Arabia embargoed oil shipments to the United Kingdom and
France.
December 1966-March 1967 Syrian Transit Fee Dispute.
Syria imposed new transit taxes on the IPC pipelines to the
Mediterranean. IPC stopped the flow of oil, refusing to pay the
higher taxes. Iraq demanded full revenues from the oil companies
operating there, despite the diminished production.
June 1967-August 1967 Six Day War.
Suez Canal closed, and IPC and Tapline pipelines shut down
following the Israeli strike into the Sinai. Oil exports were
embargoed to Western Europe-particularly to the United
Kingdom and West Germany-and the United States.
July 1967-October 1968 Nigerian Civil War.
Oil terminals were blockaded by the Nigerian Federal Navy,
choking off exports. Shell and BP ceased production in the
country during most of the war.
May 1970-January 1971 Libyan Price Controversy.
Libya gradually reduced the authorized production by selected
oil companies, claiming potential damage to the oilfields. At the
same time, the Tapline was damaged in Syria, forcing a
shutdown until Syria authorized repairs. Production cuts were
restored, and the pipeline in Syria was quickly repaired once
Libya obtained higher oil prices from the companies.
April 1971-August 1971 Algerian-French Nationalization Struggle.
Algeria nationalized 51 percent of the oil companies and
pipelines-all French owned-operating within its borders and
announced compensation plans. The Algerians then unilaterally
Duration Magnitude
(Months of Lost of Supply
Production) Shortfall
(Million b/d)
44
0.7
4
2.0
3
0.7
2
2.0
15
0.5
9
1.3
5
0.6
raised oil export prices to French firms, leading the companies to
seek a worldwide embargo against Algerian oil. Algeria retaliat-
ed by suspending oil deliveries to French tankers, claiming that
the companies owed back taxes. Oil flows resumed in August
after a CFP-Algerian agreement in July. All other companies
accepted the Algerian terms by December.
March 1973-May 1973 Lebanese Political Con lict. 2 0.5
Unrest in Lebanon disrupted the flow of oil from Iraq and Saudi
Arabia. Following a Lebanese takeover of IPC facilities near
Tripoli, Iraq cut off oil exports to Lebanon and was denied use of
Lebanon's transit facilities. In April, a storage tank and parts of
the Tapline facility at Sidon were destroyed by sabotage.
Approved For Release 2007/01/24 :, CIA-RDP83B00231 R000200020001 -0
Confidential
Table 1 (continued)
vcwuor 17 t.3-marcn 1 v t4
1 he October Arab-Israeli War.
and reduced oil production. Transit through the Suez Canal was
October 1973 in the first quarter of 1974. Saudi Arabia finally
5
1.6
April 1976-May 1976
Civil War in Lebanon.
2
0.3
May 1977
Damage at a Saudi OiUeld.
1
0.7
November 1978-April 1979
The Khomeini Revolution in Iran.
gradually restored to around 4 million b/d starting in May 1979.
6
3.7
October 1980-December 1981a
Iran .Iraq War.
There were intermittent air attacks and sabotage damage to
15
1.3
a For analytical purposes, we assumed that this ongoing conflict ends
with the conclusion of our sample period, December 1981.
Current Vulnerabilities to Supply Disruptions
The continued slide in oil consumption since the 1979-
80 price runup has forced OPEC producers to lower
output sharply. Total OPEC production in March
1982 was about 20 million b/d, or roughly 8 million
b/d below available capacity in these countries, even
with the continuation of the Iran-Iraq war. Some of
this surplus capacity-perhaps as much as 3-4 million
b/d-will be eroded once the ongoing inventory ad-
justment process is completed and demand for OPEC
oil rebounds. Resolution of the Iran-Iraq war, howev-
er, could result in an increase of 2-4 million b/d in
production capacity within a year. As for oil inven-
tories, stocks in non-Communist countries at the end
of 1981 totaled 4.4 billion barrels, or roughly 93 days,
of forward consumption. Some 300-500 million bar-
rels of these stocks were considered excess-an
amount equal to 6 to days of forward consumption.
The present combination of surplus productive capac-
ity, excess stocks, and declining consumption affords
OECD countries considerable protection against an
oil supply disruption. Indeed, the market may be
vulnerable only to a major disruption in Saudi oil
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
production over the next three to four years. Exclud-
ing Saudi Arabia, it would take simultaneous disrup-
tions in at least two other countries to cause a supply
loss of 3 million b/d. In the unlikely event that such
disruptions occurred, the Saudis probably would use
their own surplus capacity to offset the bulk of the lost
supplies
The cushion of surplus stocks and excess capacity will
be gradually eroded once OECD economic recovery
gets under way and oil use rebounds. The stock
surplus probably will be depleted later this year, while
excess productive capacity will still be available be-
yond 1985. The level and duration of the excess in
productive capacity will depend on several factors:
? The pace of economic growth and the effect of
structural changes in the economy on oil use.
? The magnitude of decline in oil production in the
OECD as a result of the expected drop in US
production.
? The rate of growth in oil consumption in LDC
economies, especially within OPEC.
? The level of exports from Communist countries.
On balance, these factors will lead to an increase in
the demand for OPEC oil and cause producers to
approach the level of available capacity later in the
decade, leaving the market once again vulnerable to
even a moderate supply disruption. Throughout the
decade, the market will be more vulnerable to disrup-
tions during the upswing of the business cycle because
of changes in inventory patterns and the rapid surge
Disruption Analysis
If the historical record of major exporting countries is
indicative of future activity, there is a good chance
that some type of shortfall will occur in any given
year. In fact, disruptions may occur with even greater
frequency:
? Five of the major exporting countries are located in
the volatile Persian Gulf region.
? Five other exporters-Mexico, Algeria, Nigeria,
Indonesia, and Venezuela-could face social tur-
moil that would disrupt oil production should the
present soft market conditions persist for a year or
more.
? Libya is subject to internal and external threats to
its ability to produce oil, given the present govern-
By superimposing the historic pattern of oil supply
disruptions on oil market conditions that we think
might materialize between 1982 and 1990, we were
able to simulate possible market conditions during
disruptions.' The simulations allowed for an inventory
buildup and liquidation in response to a supply short-
fall as well as for a supply response from oil producers
or strategic reserve drawdowns by consumers.' The
baseline scenario used for oil market conditions in the
absence of supply shortfalls enabled us to measure the
impact of supply disruptions.' (See a endix for de-
tails.)
Given the record, consuming countries should be
prepared to deal with disruptions of various sizes and
duration. We used the historic incidence of disrup-
tions to calculate the frequency, size, and duration of
potential supply disruptions between 1982 and 1990.
The. results of our simulation analysis are as follows:
? Forty percent of the simulated disruptions entailed
losses of less than 500,000 b/d.'
? Twenty percent entailed losses between 500,000 b/d
and I million b/d.
? About 15 percent resulted in disruptions of 1-1.5
million b/d.
? About 10 percent resulted in disruptions of 1.5-3.0
million b/d.
? Only 15 percent of the potential disruptions involved
losses of more than 3 million b/d.
'The possibility of supply disruptions was addressed for the 10
LDC oil exporters that we ex ect will !tnucing at least 1
million b/d by 1990.
' Inventory building in e simulations occurs at the rate of 50
percent of the oil lost in the shortfall in the first year and 10 percent
of the shortfall in the second year, to a maximum of I million b/d.
These shortfall-induced additions to inventories are liquidated in
the year following the end to the disruption.
'Assumptions used in the baseline scenario include:
? OECD economic growth averaging 2.5 percent annually
between 1982 and 1990.
? Constant nominal oil prices through 1984, an increase in the
nominal price of less than $3 per barrel by 1985, and market-
clearing prices from then onward.
? Nonoil su plies increase an average 1.3 million b/d
annually.
`Our simulations express oil supply responses in
annual rates] I
25
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Confidential
Supply Offsets
Assessing the size of potential disruptions was only
part of the analytical process. Another key element
was the extent to which these disruptions can be' offset
by production increases in nonaffected countries. The
historic record shows that most supply disruptions
have been associated with substantial offsets. Indeed,
these offsets largely account for the absence of a price
runup during most supply disruptions in the past.
Production offsets could reduce or eliminate the im-
pact of most disruptions likely to be encountered over
the next few years (figure 2). The magnitude of these
offsets could be substantial:
? Over the next two to three years, the supply re-
sponse to a disruption could be perhaps as much as 6
million b/d.
? The Saudi production response alone could be the
difference between 10 million b/d and the level of
Saudi production before a disruption. Over the next
two to four years, this difference could exceed 3
million b/d. Beyond 1986, however, the difference
probably will equal 1.5 million b/d or less.
? Non-Saudi production responses are likely to be
limited to million b/d throughout
the decade
Our analysis does not identify the specific source of
the disruption. As a result, the supply response during
a real disruption is not likely to be as automatic as we
have assumed and could be somewhat less. For exam-
ple, a disruption that involves Saudi Arabia would
obviously reduce the amount of the available supply
response.
Later in the decade, growth in the demand for OPEC
oil, coupled with erosion of productive capacity in
some OPEC countries will reduce the size of the
potential production response during a disruption.
Most analysts agree that oil prices remain stable if
there are 2-3 million b/d of spare capacity available.
By 1988, however, the production response could be
limited to less than 2 million b/d, leaving the market
vulnerable to even a moderate supply disruption=
Figure 2
Potential Production Offsets, 1982-90a
Million b/d
aBased on Saudi Arabian capacity at 10 million b/d and combined capacity
in Iran and Iraq reaching 5.5 million b/d in 1984 and remaining at this level
for the remainder of the decade.
Nonoil supplies, especially of coal and natural gas,
can also make a significant contribution to offsetting
oil losses during a disruption. In the United States, for
example, gas consumption surged 4 percent immedi-
ately following the Iranian revolution after several
years of stable demand. The impact these supplies
have on lessening pressures on oil prices probably will
increase over time:
? More oil using industries and power stations are
developing a dual-fuel capability.
? New pipeline systems are increasing gas supply
flexibility and further development of gas reserves
could provide added surge capacity.
? Expansion of coal-handling facilities, especially in
the United States, are eliminating constraints to the
expansion of coal supplies.
? Growing nuclear power capacity will rovide some
surge capacity.
Some Policy Issues
One factor in dampening the impact of any disruption
is the timely drawdown of consumer nations' stocks.
Indeed, stock levels and stock behavior have been key
factors in shaping market reactions during the last
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Nonoil Supply Disruptions
Natural Gas
Rapid growth in natural gas trade and growing
OECD reliance on gas imports from non-OECD
sources will make gas supplies increasingly vulner-
able to disruptions. International shipments
of natural gas have more than doubled since 1973,
with a growing volume coming from Algeria and the
USSR. Both countries have curtailed gas shipments
in the recent past:
? Moscow has interrupted supplies to Western Eu-
rope on several occasions during the peak winter
demand period for technical reasons and to meet
domestic needs.
? Algeria suspended gas shipments to France and the
United States and has delayed startup of deliveries
to Italy overpricing disputes.
Western Europe will be particularly vulnerable to gas
cutoffs because reliance on imports from these two
suppliers and other non-OECD sources is expected to
increase to 35 percent of total gas needs by 1990.
Coal
Coal supplies are also susceptible to disruption, as
recent events in Poland have indicated. A more likely
cause for a disruption of coal supplies in the future,
however, is a labor dispute such as occurred in the
United States in 1977. Japan could be particularly
vulnerable to a coal supply disruption because Tokyo
relies heavily on imported coal.
three major supply disruptions (table 2). Price runups
following the Arab oil embargo and the Iranian
revolution were due in large part to demand pressures
resulting from stockholder efforts to rebuild and add
to inventories. In both cases, consumption was in-
creasing and stocks were at about normal levels prior
to the disruption:
? During the Arab embargo, stocks were initially
depleted at an above-normal rate in late 1973. Oil
stocks were then drawn down at only half the
normal rate in the first quarter of 1974, and com-
panies continued to build inventories at above-
Major Supply Disruptions:
Impact on Oil Stocks and Oil Prices
Primary Oil Stocks Oil Prices a
Rate of Change (US $ per barrel)
(million b/d)
a OPEC: Average crude oil official sales prices.
b Government-owned stocks increased by 600,000 b/d.
normal rates through the third quarter. From late
1973 through the third quarter of 1974, average
OPEC official sales prices rose from $5.18 per
barrel to $11.34 per barrel.
? During the Iranian revolution, oil inventories were
drawn down at an above-normal rate during the
first quarter of 1979 and then accumulated at rapid
rates during the remainder of the year. The OPEC
average official sales price rose from $12.91 per
barrel to $23.55 per barrel during this period. I
In contrast, the oil market remained fairly stable
during the supply disruption caused by the Iran-Iraq
war in September 1980. The market calmness was
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Confidential
due in large part to the existence of about 400 million
barrels of excess stocks that stockholders were willing
to deplete. In addition, consumption was falling and
the International Energy Agency (IEA) was urging
members to refrain from spot market purchases. Oil
stocks fell at about the normal rate throughout the
winter period (October-March), and the inventory
buildup during the second quarter remained at about
the normal seasonal rate of 2 million b/d. The OPEC
average official sales price rose only from $31.74 to
$34.64 per barrel during the period despite a loss of
4 million b/d in supplies from Iran and Iraq.
If oil users anticipate a fairly quick release of oil from
stockpiles or can readily switch to alternate fuels, the
initial scramble to build and hoard inventories may be
toned down or averted. As a result, the sharp escala-
tion in spot prices and the ensuing rise in official
prices may be dampened considerably. Our simula-
tions indicate that if oil stocks or available nonoil
supplies were used at the onset of a crisis to offset a
1-million-b/d supply shortfall, the disruption-induced
runup in real oil prices would be sharply curtailed.
Holding down disruption-induced real oil prices by $5
per barrel (1980 dollars) would, on average, avert a
loss of some $350-400 billion in GNP for the OECD
over the 1982-90 period.
Commercial stocks represent the bulk of oil inven-
tories held in consuming countries. The majority of
commercial inventories are operating stocks needed to
ensure the smooth operation of the distribution sys-
tem. Because costs are involved in holding inventories,
companies will not maintain stocks at above-normal
levels unless appropriate economic incentives are pro-
vided or governments mandate such action. Most
West European governments have adjusted for this
normal market behavior by requiring companies to
maintain stocks at specified levels. West Germany,
through the Oil Stockpiling Corporation, EBV, en-
ables companies to share the financial burden of
holding excess commercial inventories.
Sizable strategic stockpiles-oil purchased and owned
commercial firms-are located only in the United
by governments as opposed to inventories held by
States, Japan, and West Germany. The Strategic
Petroleum Reserve in the United States now contains
about 250 million barrels of oil stored in underground
salt caverns. Japan has a strategic reserve of about 70
million barrels, mostly held in floating tankers off-
shore. The West German stockpile contains about 50
million barrels of oil.
At present, these foreign governments have not indi-
cated clear plans regarding the disposition of these
strategic stockpiles during an oil supply disruption,
although the stocks are viewed as a last resort. In both
Japan and West Germany, the level of stocks is still 25
below the announced stockpile goal. Even so, the
potential for sizable releases of stockpiled oil could
have a significant price-dampening effect during a
disruption, if buyers refrain from panic purchases on
the spot market in anticipation of the availability of
oil from strategic stocks.
Some other factors that consuming nations have
considered or implemented to cope with a supply
disruption include:
? Removal of controls on crude and product prices to
allow market forces to allocate supplies.
? Use of the IEA oil-sharing scheme.
? Standby legislation to allocate oil supplies.
? Maintenance of dual-fuel capacity in industry and 25
electric power facilities; the Dutch, for example, can
reduce oil use by 10 percent by switching to gas.
? Surge capacity for domestic production of oil and
natural gas.
? Stockpiles of alternative fuels such as coal and
natural gas.
Altogether these measures provide consuming coun-
tries with some protection against small oil and gas
disruptions through the mid-1980s. Beyond this peri-
od, the measures now in place will provide less
protection because the production-capacity cushion
will diminish and leave the market more vulnerable to
any size disruption. By that time-the late 1980s-
some consuming countries, especially in Western Eu-
rope, will also become increasingly vulnerable to gas
supply disruptions.
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Appendix
Disruption Methodology
Probability of Disruptions
To estimate the likelihood, duration, and magnitude
of possible future oil disruptions, we examined the
historical instances of disruptions to oil production
between 1951 and 1981. Oil research focused on those
cases where oil production declined for nonmarket
reasons such as damage to oil facilities, nationaliza-
tion of com anies or political decisions about produc-
tion levels.
Thirteen cases constitute the historical data base.
These disruptions ranged in length from one month to
44 months and the magnitude ranged from 300,000 to
3.7 million b/d. In calculating the magnitude for our
simulations, the amount of oil lost was expressed as a
percentage of oil production in significant exporting
countries over an average of three months prior to the
disruption. The largest percentage loss-32 percent-
resulted from the 1956-57 closure of the Suez Canal.
We assumed that a disruption was equally likely to
occur in any future year as in the past, and we
calculated the estimate of the yearly probability of,i
disruption of any size to the oil production of signifi-
cant oil-exporting countries to be 0.42 (13 divided by
31).5 Probability distributions for the duration and
magnitude of oil disruptions were also constructed,
with equal weight given to each historical case.
Generating Disruptions
Scenarios of the occurrence of market disruptions
between 1982 and 1990 were created by combining
random selections of the likelihood, length, and mag-
nitude of a disruption in each year. To generate each
? A random number between 0 and 1 was drawn for
each year and matched against the estimate of the
likelihood of a disruption. Where a random number
' Disruptions to the combined production of the 10 OPEC and non-
OPEC LDC oil exporters that we expect to be producing at least
1 million b/d in 1990 were considered in our baseline estimates-
Indonesia, Iran, Iraq, Kuwait, Libya, Mexico, Nigeria Saudi
Arabia, the United Arab Emirates, and Venezuela.
.was less than or equal to the estimate of the
probability of a disruption-0.42-a shortfall in oil
production was recorded. Should the random num-
ber exceed 0.42, oil production follows a baseline oil
scenario.
? Once a disruption was recorded for a particular
year, two additional random numbers were drawn to
pick from the probability distributions calculated
for the duration and magnitude of past disruptions.
? As a third step, duration times magnitude times the
baseline estimate of oil production was calculated to
obtain the oil lost in each disruption. For example, a
hypothetical disruption of six months (duration
= 0.5), shutting down 3 percent of the oil produced
(magnitude = 0.03) in a year in which our baseline
production projection for the group of 10 countries
would otherwise be 24 million b/d, implies a loss of
360,000 b/d in oil on the world market.'
The method of randomly selecting from the individual
probability distributions constructed from the data
base allows for combined results that exceed any of
the historical disruptions. A random pairing of the
longest disruption with the largest percentage of oil
lost over any period would yield a disruption that
exceeds any recorded case. Conversely, smaller than
historically observed disruptions can also result from
the random selection.)
The Baseline Scenario
A "no surprises" scenario in the CIA energy model
was used to calculate benchmarks for energy demand,
GNP growth, and oil prices for the 1982-90 period.
The underlying characteristics of the model included:
? Oil and nonoil supplies are exogenous with the
exception of Saudi Arabia's oil output; Saudi pro-
duction is endogenous up to the limit of the Saudis'
preferred ceiling of 8.5 million b/d.
Because disruptions can extend more than one year, the longer
disturbances were treated by reducing the amount of production in
one year and a fraction of the amount in the following year(s). For
example, a 3-percent disruption that lasts for 18 months would
result in a loss of 720,000 b/d in the first d a loss of 360,000
b/d in the second year
Confidential 25X1
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
? Income elasticity of energy demand is 1.0 with no
lags.
? Price elasticity of energy demand is 0.12 in the first
year, with a cumulative impact of 0.5 through the
ninth year.
Energy prices in the model are determined in the
following manner:
? The cost of crude oil is determined by the OPEC
price, subject to price controls on domestic produc-
tion, where applicable.
? Prices for petroleum products are determined from
costs of crude oil, refining, and transportation.
? Natural gas prices move with petroleum product
prices, subject to national controls on end-use prices.
? Coal prices move with the GDP deflator.
? Electricity prices are obtained from prices on fossil
fuels, weighted by shares of the fuel used to produce
final electricity.
? The end-use price for total energy is built up from
prices for individual fuels, weighted by the demand
for each fuel, and based to a common year.
Given these relationships, the model is solved for a
market-clearing OPEC price.
The energy model summarizes the macroeconomic
impacts of higher energy prices through reaction
functions derived from simulations with the CIA
linked econometric model. The model is an ISLM
model with a small supply side and is composed of
medium-size models for major industrialized coun-
tries. It estimates total demand for energy and ties
energy and macroeconomic sectors together through
prices.
Simulation of Oil Disruptions
As a first step in assessing the impact of disruptions
on the oil market and OECD economies, 100 scenar-
ios of oil-production disruptions for each year between
1982 and 1990 were generated. Of a total of 900
possible disruption-years, 451 cases were obtained in
the first simulations; 63 of these, however, were the
result of multiyear disruptions. The average disrup-
tion was 1.3 million b/d, and the largest exceeded
10.9 million b/d:
? Forty percent of these disruptions involved a loss of
less than 500,000 b/d in a single year.
? Another 20 percent involved a loss of less than 1
million b/d.
? Nearly 75 percent of the disruptions resulted in a
loss of less than 1.5 million b/d in oil production.1
Stock Behavior and Supply Responses
Historical experience since 1970 indicates that oil
consumers typically build inventories in the months
following a disruption in oil production. To capture
this behavior pattern, a three-year inventory cycle was
developed in which stocks increase by 50 percent of
the disruption in the first year and 10 percent of the
disruption in the second year, to a cumulative total of
1 million b/d. Stocks are liquidated by an amount
equal to the buildup in the third year, or in the case of
a multiyear disruption, in the year after it ends.I
In past disruptions, a number of OPEC countries have
responded with increased production. To approximate
the amount of additional production forthcoming, the
productive capacity of each OPEC member was ex-
amined, as well as their recent responses during the
1978/79 Iranian revolution. Production from OPEC
states (excluding Iran) rose by 3.3 million b/d on a
year-over-year basis in that period; 90 percent came
from Saudi Arabia, Iraq, Kuwait, and Nigeria.I I
Three alternative rules were employed to simulate the
supply response from unaffected producers and con-
sumers. One set of simulations provided as much as
1.5 million b/d in additional supplies from countries
other than Saudi Arabia. Given the continuing ero-
sion of productive capacity among some of the OPEC
states, this amount represented an upper limit (figure
3). The Saudis were assigned a maximum production
of 10 million b/d, with a production response equal to
the difference between this amount and Saudi produc-
tion in our baseline scenario. Because of growing
requirements for Saudi production in our baseline
scenario, the combined OPEC supply response was
limited to 3 million b/d after 1986 (figure 4).I
A more optimistic supply response allowed for an
additional I million b/d to be added in the event of a
disruption. Such supplies could be forthcoming from
strategic oil stocks in the consuming countries. The
third supply response assumption allowed for addi-
tional supplies only to the extent that inventory
demand was countered (figure 5).
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Confidential
Figure 3
Comparision of Market-Clearing Prices for Oil:
Assuming Optimistic Supply Response and
Stock Behavior
Median Price
0 Lowest Price
Real Oil Prices, 1980 $ per Barrel
100
- Baseline
- Highest Price
Figure 4
Comparison of Market-Clearing Prices for Oil:
Assuming Pessimistic Supply Response and
Stock Behavior
Baseline
- Highest Price
Median Price
Lowest Price
Real Oil Prices, 1980 S per Barrel
4-82
Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Figure 5
Comparision of Market-Clearing Prices for Oil:
Assuming Most Likely Supply Response and
Stock Behavior
Baseline
Highest Price
Median Price
Lowest Price
Real Oil Prices, 1980 $ per Barrel
Confidential 12
Confidential Approved For Release 2007/01/24: CIA-RDP83B00231 R000200020001-0
Confidential