23

Aviation fuel demand modelling in OECD

and developing countries: impacts of
fuel efficiency
Mohammad Mazraati* and Osama M. Alyousif**
*Energy Studies Department, The Organization of the Petroleum Exporting Countries (OPEC), Obere
Donaustrasse 93, 1020 Vienna, Austria. E-mail: [email protected] and [email protected]
**Mechanical Engineering Department, College of Engineering & Petroleum, Kuwait University,
PO Box 5969 Safat, Kuwait. E-mail: [email protected]

Abstract
On the quest for reducing the fuel consumption per passenger per flight for economical and environ-
mental reasons, commercial aircraft manufacturers are implementing new strategies for optimising
aircraft performance by using new lighter and stronger materials and enhancing engines’ effici-
encies in terms of fuel consumption and maintenance requirements. With the rising and falling
of economies, whether in the Organization for Economic Cooperation and Development (OECD)
countries or other developing countries, the aviation industry has been affected by multiple factors
such as passenger traffic, freight traffic, airport capacities and oil prices. Aircraft manufacturers have
worked on improving the engine efficiency of their newly built airplanes (e.g. Airbus’s A-380 and
Boeing’s B-787), and many airports in the world have increased the number of their runways to face
the increasing demand for air traffic in the world. Aviation efficiency can also be achieved through
better load management, which in return enables airliners to cope with higher oil prices or rising
costs. Aviation fuel demand is modelled in OECD North America, Europe and Pacific regions and
some selected developing countries. Price elasticities of fuel demand in all regions are low, while
income elasticities are high. The elasticity of aviation fuel demand on passenger kilometre per-
formed (PKP) is considerably low. One per cent increase in PKP leads to less than half a per cent
increase in aviation fuel demand, confirming an ongoing fuel efficiency in aviation industry.

1. Introduction
Passenger traffic is expected to grow at a rate of 5 per cent for the next 20 years, and current
numbers of passengers should double by the year 2020 (FAA, 2008). Aircraft manufac-
turers aim to meet the demand for more airplanes to carry the expected increasing numbers
of passengers, but such goals are challenged with economical and environmental chal-
lenges (Airbus, 2007; Boeing, 2007). Increased crude oil prices have affected the aviation
industry with increased jet fuel prices, and consequently, have brought additional costs
to air carriers. Environmental regulations for carbon gases emissions have also induced

© 2009 The Authors. Journal compilation © 2009 Organization of the Petroleum Exporting Countries. Published by

Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
24 Mohammad Mazraati and Osama M. Alyousif

further restrictions in the aviation industry. In order to meet the demand for more air travel
and to meet the economical and environmental restrictions, fuel-efficient airplanes must
be used in aviation. Aircraft manufacturers are using innovative concepts for reducing fuel
consumptions. Any estimates for the future jet fuel demand must account for the increased
efficiency in fuel consumption and the retirement of ageing fuel-inefficient aircrafts. This
study aims to predict the jet fuel (aviation fuel) demand for the Organization for Economic
Cooperation and Development (OECD) countries and developing countries. The develop-
ing countries include China, India, Indonesia, Malaysia, Philippines, Singapore, Thailand
and Vietnam. These countries have shown aggressive growth in their economies. OECD
countries are modelled by grouping in North America, Europe and Pacific.
Section 2 covers the fuel efficiency improvements in commercial aircrafts; Section 3
explains the key variables in aviation industry using the historical trends. In Section 4, the
fuel demand model is specified and estimated. The forecast of fuel demand in the reference
case is also discussed in this section. The last section contains conclusions.

2. Fuel efficiency improvements in the commercial aircrafts

In order to achieve satisfactory fuel reduction rates, it is required that the following
changes be implemented: (i) reduce the weight by 15–25 per cent; (ii) reduce the drag by
15–20 per cent; and (iii) reduce fuel consumption by 20–30 per cent (Greene, 1990). In
addition, the size of the aircraft can have direct and indirect fuel consumption efficiency.
Larger aircrafts can carry more numbers of passengers and thus increase the fuel effi-
ciency for the aircraft (L/passenger km), and using larger aircrafts will ease the conges-
tions in major airports and reduce the holding time before landing or idling time before
take-offs (Lee et al., 2001; Babikian et al., 2005). Such reductions in engine running time
will have indirect improvements in reducing the extra amounts of fuel needed, which even-
tually lower the take-off airplane gross weight.
Such improvements need changes in aircraft engines, aerodynamics and structural
materials for airframe and high-temperature materials for engine components. The use
of supercomputers have provided a powerful tool for complex calculations in the field of
computational fluid mechanics for analysis of aircraft aerodynamics, and the field of stress
analysis for load analysis to reach the optimum structure and shape design.

2.1. Aircraft engines

When the first turbojet engine was introduced in 1960, it was very fuel inefficient and had
high levels of noise. The technological limitations affected the pressure ratio and the
turbine inlet temperature. The early modifications were improving the components ma-
terials and doubling the number of air compressors. These enhancements in the turbojet
engine were insufficient because of the high velocity exhaust, which was more suitable to

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 25

supersonic airplanes. To solve the high exhaust velocity, the new low-bypass turbofan
engine was introduced by Rolls-Royce. Later on, the high-bypass turbofan engine devel-
oped by General Electric have improved the aircraft engine performance and have reduced
the fuel consumption by 25–28 per cent, in addition to considerable reduction in noise
levels.
Other improvements come from the high-temperature components used in the
low-pressure compressors in the engine, which will increase the thermal efficiency of
the engine. Old airplanes’ engines used steel as the primary material for manufacturing
the components because of its high-temperature resistance and strength. Nickel-based
superalloys replaced the steel components. These alloys were coated with thermal barrier
ceramics to increase their heat resistance and are widely used because of their equal or
superior qualities over steel in strength and weight, in addition to their excellent resistance
to corrosion and oxidation.
In the future, ultra-high bypass ratio engines for subsonic airplanes may add further
fuel consumption savings of 15 per cent more than currently used turbofans.

2.2. Airframe and aerodynamics

Airframe improvements come from reducing the weight while maintaining the required
material’s strength. Aluminium alloys were the first materials used in manufacturing
aircraft structural components. The improvements of the mechanical properties, such as
weight and strength in aluminium alloys, kept them in use for many generations of
airplanes until newer materials were found to have better properties and lighter weights.
These materials were fibre-reinforced materials or what are commercially known as ‘com-
posites materials’. Composites materials consist of high-strength fibres that are oriented in
specific directions, and these fibres are held together by a compatible resin. The material
can have maximum strength only in one direction, which will be the direction of highest
applied loads on the structure. Composites materials were earlier introduced in spacecrafts
and missiles to benefit from their light weight and tremendous strength. Aircraft manufac-
tures began to utilise the excellent features of composites materials, and some parts of the
fuselage are made from either carbon fibre-reinforced materials or aluminium lithium
composites. New generations of passenger aircrafts are using greater amounts of compo-
sites materials than before: Boeing’s new B-787 ‘Dreamliner’ will be fabricated with 50
per cent of its weight in composites materials. The new Airbus A-350 will introduce an all-
composite wing and an airframe with 60 per cent of its materials made from composites
materials. Airbus’s super jumbo, the A-380, which was introduced in 2008, has used large
amounts of composites in its airframe. An added feature to using composites materials in
passenger aircrafts is its resistance to corrosion and fatigue, which cuts down the inspec-
tions required for cracks and corroded components. Reducing the maintenance frequency
will, in return, cut down the costs of inspection and maintenance.

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

26 Mohammad Mazraati and Osama M. Alyousif

The shape and size of any current in use or future design aircraft is governed by the aero-
dynamics of the airframe structure. Massive body sizes will increase the drag on the air-
plane. Drag is a term used to describe the resistance induced by the air friction to the shape
and size of the craft to forward movement. Continuous improvements in the field of aero-
dynamics can reduce the drag on an airplane by employing structure enhancements. One
of these enhancements is the use of ‘winglets’, small attachments to the tip of the aircraft
wing that will reduce the drag vortices produced by the wing tips. A vortex is a turbulent
flow of fluid. The motion of the fluid swirling rapidly around a centre is called a vortex. The
speed and rate of rotation of the fluid are greatest at the centre, and decrease progressively
with distance from the centre. Vortices can produce forces that are opposite to the motion of
an airplane, and hence increase the drag. Its estimated that 40 per cent of forward drag on the
airplane is due to vortices. Winglets reduce the intensity of the vortices at the wing tips,
which will reduce the drag on the airplane. Winglets can reduce the fuel consumption in an
aircraft by 3–4 per cent, which made the addition of winglets to new aircrafts a must, and
retrofitting older generations of crafts with substantial airworthiness with winglets can be
beneficial. Kroo (2005) has studied the developments in non-planar wing concepts and
their effects on aircraft efficiency. It was found that non-planar wings variations, including
winglet, might have a potential for reducing fuel consumption in passenger aircrafts.

2.3. Operations
Fuel consumption reductions can also be achieved by proper utilisation of the aircraft in
terms of available seating capacities and optimising the best routes to reach their destina-
tions. Such improvements may not be limited to:
1. Increasing the number of passengers and reducing the number of empty seats by better
selection of airplane sizes and flight timings.
2. Several airlines taxiing on with one engine when conditions permit; it could save
US$10–12 million/year.
3. Reducing the weight of the onboard equipment (excess seats, equipment and galley
equipment).
4. Improved navigation for better and faster routes. Alaska Airlines is deploying a new
flight planning system to yield more direct routings.
5. Using ground transportation, e.g. American Airlines and Delta Air Lines use super
tugs on the ground to reposition aircraft where feasible.
6. Using ground facility: American Airlines, Southwest Airlines and others are using
ground power to provide electricity and ground-conditioned air, rather than the
plane’s auxiliary power unit.
The science of aircraft building and operating is involved with very complex
mathematical calculations. To build and test a new aircraft, there are many sophisticated

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 27

calculations to be made. Wing shape calculations is an example of these calculations

where the aspect ratio (wing length to width) is an important factor in fuel consumption.
Structural stress calculations are also another example for identifying the points where
the highest loads will be on the airframe.
For operation, the calculations for the shortest and direct routes may influence the
fuel consumption, and estimating the time required for landing approach and taxiing to
the landing gate without consuming extra amount of fuel is involved with optimising the
airplane speed, flight path and altitude.
Air traffic management (ATM) can have a major part in improving the fuel consump-
tion in the aviation industry. Better management systems and accurate climb and decent
procedures will prevent delays and reduce the extra fuel needed for holding flying patterns.

2.4. Efficiency prospects

The expected growth in aviation industry is estimated to be 5 per cent. There is a need
for new aircrafts to feed the demand for travel in many regions of the world, with increased
demand in developing countries like China and India. New airplanes to enter the service
and retiring old airplanes will play an important role in reducing fuel consumption and
controlling emissions. The improvements in fuel efficiency will continue as new ideas
for improvements are tested and implemented in new airplanes. Figure 1 shows the fuel
efficiency for wide and narrow-body airplanes introduced in service since 1958. The
figure shows gradual improvement in fuel efficiency by 1 per cent, with room for improve-
ments in the future. The future generations of passenger airplanes may use new concepts
of flying to accommodate the traffic demand and satisfy the environmental regulations
(Abadie, 2007).
Based on the International Air Transport Association (IATA), new aircrafts are 70 per
cent more fuel efficient than 40 years ago, and 20 per cent better than 10 years ago (IATA,

0.08
Liter/passenger kilometer

Comet-4
0.07
0.06
0.05
B707
0.04 B747-100 A310-300

0.03 A350

0.02 B767-400ER A380-800

0.01
0.00
1950 1960 1970 1980 1990 2000 2010 2020

Figure 1 Fuel efficiency for wide and narrow body passenger airplanes.
Source: Abadie (2007).

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

28 Mohammad Mazraati and Osama M. Alyousif

2008). Airlines are planning to gain a further 25 per cent fuel efficiency improvement
by 2020. Modern aircrafts achieve fuel efficiencies of 3.5 L per 100 passenger km. IATA
airlines have adopted a voluntary fuel efficiency goal to reduce fuel consumption and CO2
emissions (per revenue tonne kilometre) by at least 25 per cent by 2020, compared to 2005
levels (IATA, 2008). A portion of this target might realise, based on IATA; ATM enhance-
ments could improve fuel efficiency and CO2 emissions reductions by up to 12 per cent,
and operational improvements could provide a 6 per cent overall fuel saving (IATA, 2008).
The fuel efficiency in US airliners improved highly over the recent decades. It has
improved more than 25 per cent in 2006 when compared to 1990. It is expected that
this will reach to over 30.8 per cent in 2010 when compared to the 1990 efficiency level
(Heimlich, 2008).
The efficiency improvement is considered endogenously in the model.

3. Statistical data and analysis

OECD countries were divided into three regions: (i) OECD North America region, which
includes North American countries; (ii) OECD Europe region, which includes the Euro-
pean members in the OECD; and (iii) OECD Pacific region, which includes the Pacific
members in the OECD. Developing countries chosen for this study were the Pacific Rim
countries that showed aggressive development and promising progress in their economical
futures. China, India, Indonesia, Malaysia, Philippines, Singapore, Thailand and Vietnam
are the countries that constitute the ‘developing countries’ group in the statistical data and
analysis and modelling.
The air traffic and financial data in this report were obtained from the International
Civil Aviation Organization (ICAO); the data include the passengers and freight traffic
as well as the number of passengers carried and load factors (LFs). The revenues from
passengers and freight and fleet details of every airline in the OECD countries and the
developing countries were also obtained from the ICAO database. Jet fuel prices and
consumptions for the countries used in this study were obtained from the International
Energy Agency. The economical and population data for this study were obtained from
the International Monetary Fund.
Reviewing the past trends of aviation sector by different regions reveals the facts on
development of aviation sector. There are many questions that can be raised on the level
of market saturation, LF, fuel efficiency, PKP and cargo kilometre performed. Knowing
the impacts of fuel price on development of aviation transportation by different regions,
among other issues, could provide guidance for a better model specification. There are
many other variables that affect air travel growth, besides the socio-economic growth,
price and cost factors. This might include the network of infrastructure and its evolution,
improvement in service offering by airliners, such as routs, frequencies, promotions and

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 29

modal choices, and the effect of globalisation on increasing trade and business. Consider-
ing the mentioned issues brings a deeper analysis but requires more resources, which are
out of the scope of this study.

3.1. Passenger and freight traffic

In order to understand the passenger traffic pattern for each region considered in this study,
it was necessary to obtain the PKP and Freight Kilometre Performed (FKP) for the airlines
operating in the considered regions. The PKP is a term used for passenger traffic, and it is
defined as the number of paying passengers carried for 1 km. The FKP is a term used for
freight traffic, and it is considered as the number of tonnes carried for 1 km. This might
also be referred as tonne kilometre performed.1
Another factor used in the aviation industry for the ratio between available seats in an
airplane and the number of passengers in the same airplane is called the LF. Flights with
high LF correspond to less empty seats. LF is also a measurement of air carrier efficiency
to fill the airplanes with passengers to make optimum consumption of fuel designated
for the flight journey. Figure 2 shows the passenger and freight traffic for OECD North

1600 100
Billions

PKP FKP PLF WLF

90
1400
80
1200
Kilometre performed

70

1000
60
Load factor %

800 50

40
600
30
400
20

200
10

0 0
1980 1985 1990 1995 2000 2005
Year

Figure 2 Passenger and freight traffic in OECD North America.

Source: ICAO (2008).
PKP, passenger kilometre performed; FKP, freight kilometre performed; PLF, passenger load factor;
WLF, weight load factor.

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

30 Mohammad Mazraati and Osama M. Alyousif

America; it can be seen that the North American region is a large market, with some signs
of approaching the area of saturation. The region has tripled its passenger traffic since
1985, but declined in 2000. The dip in traffic in the years of 2001–2003 was because of the
9/11 terrorist attacks, which affected the travel pattern of North American people. On the
average, the PKP has increased 5 per cent per annum from 1980 to 2007, while it is only
2 per cent per annum for the period 2000–2007. The same trend can be seen for the FKP.
The FKP has grown 5.8 per cent per annum during 1980–2007, while its average annual
growth rate was 2.4 per cent during 2000–2007.
The passenger LF trend shown in Fig. 2 depicts that it is changing between 62 and 67
per cent till 2000, but in the recent years, this value has increased to 75 per cent, which
reflects an improved efficiency in aviation load management. This might be mainly due to
the higher fuel prices. This has helped the aviation sector to be a bit resilient to the recent
high fuel prices.
Figure 3 shows the passenger and freight traffic and LFs for OECD Europe region.
The European market is a growing market, and the introduction of low-cost travelling has
pushed this market to higher levels. However, the OECD Europe market is half the size of

700 100
Billions

PKP FKP PLF WLF

90
600
80
Kilometre performed

500 70
Load factor %
60
400
50
300
40

200 30

20
100
10

0 0
1980 1985 1990 1995 2000 2005
Year

Figure 3 Passenger and freight traffic in OECD Europe.

Source: ICAO (2008).
PKP, passenger kilometre performed; FKP, freight kilometre performed; PLF, passenger load factor;
WLF, weight load factor.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 31

the North American market, which leaves some room for growth and expansion. The small
dip in traffic in the year 2002 is similar to the North American region, but the effect was
quickly recovered in the following year. LFs in OECD Europe were similar to the North
American LFs, where the average value was hovering around 65 per cent till 2000, but this
value increased to 74 per cent in 2007.
The average annual growth rate of passenger traffic was 5.5 per cent within 1980–
2007, a bit less than that of North America, but it has come down to 3.3 per cent per annum
during 2000–2007, still higher than the growth rate in North America, notifying an
expanding market in the Europe.
OECD Pacific region aviation traffic is shown in Fig. 4. This region, at its current
levels, is smaller than the OECD Europe market, and the growth has slowed down from
the second half of the 1990s. The 9/11 effect is apparent in the traffic pattern in the years
2002 and 2003. The passenger LF values were maintained at an average of 70 per cent with
lower values in the years 1997–2003, and this was due to the economical difficulties in that
region. The passenger LF has recently reached 73 per cent. It is seen that in all OECD
regions, the LF has been improving recently along with higher fuel prices. It is worthy to
mention that in all OECD regions, the aviation traffic similarly shows a behaviour of a not

500 100
Billions

PKP FKP PLF WLF

450 90

400 80
Kilometre performed

350 70 Load Factor %

300 60

250 50

200 40

150 30

100 20

50 10

0 0
1980 1985 1990 1995 2000 2005
Year

Figure 4 Passenger and freight traffic in OECD Pacific.

Source: ICAO (2008).
PKP, passenger kilometre performed; FKP, freight kilometre performed; PLF, passenger load factor;
WLF, weight load factor.

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

32 Mohammad Mazraati and Osama M. Alyousif

700 100.0
Billions

PKP FKP PLF WLF

90.0
600
80.0
Kilometer performed

500 70.0

Load Factor
60.0
400
50.0
300
40.0

200 30.0

20.0
100
10.0

0 0.0
1980 1985 1990 1995 2000 2005
Year

Figure 5 Passenger and freight traffic in selected developing countries.

Source: ICAO (2008).
PKP, passenger kilometre performed; FKP, freight kilometre performed; PLF, passenger load factor;
WLF, weight load factor.

sharp consecutive increase from the second half of the 1990s. In all regions, the passenger
LF has experienced improvement in recent years along with higher oil prices.
The developing countries’ passenger and freight traffic is shown in Figure 5. This
market, at first glance, looks different to that of OECD in passenger traffic pattern. The
PKP draws first phase of an S-shaped curve, meaning that the aviation market in these
countries has a potential for growth. There are many factors that affect aviation transpor-
tation: the fast economic growth rates, urbanisation, population growth, expansion of the
aviation sector and the ‘open market’ policy, which allowed airliners to boost the aviation
industry in these countries. The PKP was grown at the rate of 9.1 per cent per annum
within 1980–2007, which is 3–4 per cent higher than that of OECD countries. The passen-
ger traffic grew at an even higher rate during recent years. The annual average growth rate
of PKP was 9.3 per cent within 2000–2007. This was not the case for all regions of OECD,
where the aviation traffic growth rate was very low when compared to long-term average.
In the selected developing countries, the LFs were variable from 60 to 70 per cent, but
in recent years, like other OECD regions, were showing signs of improvements.
More precisely, the developing countries are in the stage of growth, and the markets,
despite high oil prices, are expanding fast. In developed countries, although the market is
developed, the passenger traffic is still growing relatively moderate.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 33

3.2. Jet fuel prices and consumption

Jet fuel price is considered as a key variable affecting the jet fuel demand. Figure 6 draws
trend of jet fuel prices along with the jet fuel consumption trends in different regions. The
simple correlation relation between prices and jet fuel consumption suggests that the
consumption due to growing passenger traffic is not affected negatively. But this is not
the case for all periods. For example, high oil prices due to the Iranian revolution affected
the aviation fuel consumption negatively, and later on, a sharp decrease in prices led
to higher aviation fuel consumption. As shown in Table 1, the price and consumption
are highly negatively correlated during 1980–1985. For the most recent period, i.e. 2000–
2007, again the correlation is positive, which reflects the idea that the aviation sector has

2.5 OECD Pacific

80
OECD North America
OECD Europe 70
Developing Countries
2 Jet Fuel prices
60

50
1.5

US $ / Bbl
mb/d

40

1
30

20
0.5
10

0 0
1970 1975 1980 1985 1990 1995 2000 2005
Year

Figure 6 Jet fuel consumption versus jet fuel price.

Source: IEA (2008).

Table 1 Jet fuel price and aviation oil consumption correlation

Region 1972–2007 1980–1985

OECD Pacific 52.3 -97.5
OECD North America 42.4 -74.0
OECD Europe 62.4 -65.8
Selected developing countries 68.4 -95.8
OECD, Organization for Economic Cooperation and Development.

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

34 Mohammad Mazraati and Osama M. Alyousif

managed to bypass the impact of high prices through better LF and cost reduction activi-
ties. One should not ignore the effects of technical fuel efficiency of the new aircrafts that
came into the navigation system in the course of time (see Fig. 1).

3.3. Airline ticket prices

The average ticket price has an indication about people’s sentiment about travelling. When
ticket prices are low, it will encourage people to travel, and when they are high, it will dis-
courage people from travelling. Table 2 shows the average ticket price per passenger.
These data were obtained by dividing the total ticket sale revenues earned by the air
carriers over the total number of passengers carried over that year. OECD North America
region showed fluctuating average ticket prices, which is a sign of market maturity. The
average price has declined, on the average, by 1 per cent per annum from 2000 to 2005. The
OECD Europe and OECD Pacific regions, however, have different average ticket prices,
and the increase in the average ticket price for 2005 reflects the fact that the industry was
able to transfer a portion of the high jet fuel price on the customers’ shoulders. Developing
countries have the same pattern like the OECD Europe and OECD Pacific, including the
sharp increase in average ticket prices in recent years due to increase in jet fuel prices. It
can be shown that ticket prices are affected by jet fuel prices, and air carriers tend to trans-
fer a portion of the increase in fuel prices to travelling passengers to make up for some
of the increased expenses. The increase in average ticket prices, which is due to increased
fuel prices, may affect the demand for travel, and consequently will affect the demand for
jet fuel. This point should be taken in consideration when predicting the future demand
of jet fuel. Brons et al. (2001) have studied the effect of air ticket prices on travel demand
and concluded that tourist travellers tend to be more sensitive to air ticket prices than
business travellers, and increasing the air ticket prices would negatively affect the travel

Table 2 Average ticket prices per passenger in US$

OECD North OECD OECD Developing

Regions America Europe Pacific countries
1985 112.47 143.57 192.78 168.20
1990 139.96 212.16 198.68 186.94
1995 133.64 217.20 215.15 188.39
2000 151.94 173.33 186.60 171.04
2005 144.44 205.92 195.37 195.63
GR 2000–2005 (%) -1.01 3.45 0.92 2.69
Source: ICAO (2008).
OECD, Organization for Economic Cooperation and Development; GR, growth rate.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 35

demand (Brons et al., 2001). Hayashi and Trapani (1987) have studied the impact of
energy cost on passenger domestic air travel in the United States, and concluded that rising
fuel prices will negatively affect the travel demand and the quality of services provided by
the air carriers if the ticket prices were kept unaffected, which indirectly affect the aviation
industry.

4. Jet fuel demand model

4.1. Model specification
Gately (1987) has used a direct econometric method to estimate jet fuel demand. He used
a recursive system of equations. US jet fuel demand in a log-linear functional form was
considered as a function of seat-miles, fuel price and lagged seat-miles/gallon. The seat-
miles/gallon is a criterion of aviation efficiency. He considered the fuel efficiency as an
endogenous variable that is a function of fuel price and time trend variable. The seat-miles
variable is a function of per capita income and ticket price. Modelling jet fuel demand
could also be indirectly estimated through demand for aviation transportation demand
(i.e. business passenger, personal passenger and civil freight). The changes in aviation
transportation demand could be formulated as a logistics functional from. The aviation
transportation demand is a function of an intrinsic growth rate and a time-variant capacity
variable. In order to estimate demand for jet fuel, an efficiency index is being used. Using
the available fuel efficiency indices, the jet fuel demand is forecasted (Vedantham and
Oppenheimer, 1998):

dD ⎛ D⎞
= rD ⎜1 − ⎟ (1)
dt ⎝ K⎠

Jf = η ∗ D (2)
where D, r and K stand for aviation traffic variable, intrinsic rate of expansion and capa-
city, respectively.The term Jf stands for jet fuel demand and h is the average exogenous fuel
efficiency of the aviation system that is calculated as litre per tonne kilometre of homo-
genised freight and passenger travelled.
In this paper, jet fuel demand is modelled directly as a function of key variables,
i.e. Gross Domestic Product (GDP), jet fuel price, urban population, PKP and a dummy
variable.2 Since the data is aggregate for regions, this type of model could generate better
results. The coefficient of PKP could be interpreted as the implicit fuel efficiency criterion.
In the log-linear functional form, it is known as efficiency elasticity of fuel demand.
It reflects the percentage increase in jet fuel demand due to 1 per cent increase in PKP.
ln ( jf jt ) = α + β ∗ ln ( y jt up jt ) + θ ∗ ln ( p jt ) + φ ∗ ln ( PKPjt ) + δ ∗ d jt + ε jt (3)
where:

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

36 Mohammad Mazraati and Osama M. Alyousif

j = OECD North America, OECD Europe, OECD Pacific and selected developing
countries;
jfjt = jet fuel demand for region j at time t (million barrel/annum);
yjt = real GDPs for region j at time t (million US$ at 2000 prices);
upjt = urban population for region j at time t (million inhabitant);
pjt = real jet fuel prices for region j at time t (US$ per barrel). The fuel prices are
adjusted by the Consumer Price Index (CPI) of a representative country in each
region;
PKPjt = PKP for region j at time t;
djt = dummy variable that captures the effects of the 9/11 terrorist attacks on aviation
transportation. It takes value 1 for year 2003 and 0 otherwise.

4.2. Model estimation

In order to avoid estimation of a spurious equation, the unit root and cointegration tests are
used. Table 3 summarises the unit root test for variables used in the model for all regions
by the Augmented Dickey-Fuller (ADF) test. In addition, the results of the cointegration
test are presented at the bottom of Table 3 to show that there is a cointegrating or long-term
relationship.
All variables used for OCED North America equation, except the PKP, are non-
stationary, and the ADF test does not reject the null hypothesis of unit root existence. As it
can be seen from Table 3, the same can be said for all other variables for other regions. But

Table 3 The ADF unit root test for variables used in the jet fuel demand model

OECD North OECD OECD Developing

Variables America Europe Pacific countries
Jet fuel demand -1.2^ 0.34^ -2.52^ 0.06^
Real jet fuel price -2.4^ -1.40^ -2.3^ -2.29^
GDP per urban population 0.24^ -0.91^ -2.5^ 0.18^
Passenger kilometre performed -4.3* 0.64^ -2.0^ -0.19
Cointegration test: stationary of -4.9* -4.8* -6.3* -3.8*
residuals by ADF
Note: Null hypothesis, i.e. existence of the unit root is rejected at * 1% critical levels.
Maximum lag 5 selected by Schwarz Information Criterion.
^ The time series is I(1).
ADF, Augmented Dickey-Fuller test; OECD, Organization for Economic Cooperation and Develop-
ment; GDP, gross domestic product.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 37

Table 4 Estimation of the coefficients

a b q f d R2
Passenger
Intercept Income* Fuel price kilometre Dummy
OECD North America -2.6 0.55 -0.08 0.14 -0.05 0.97
t-statistics -2.9 2.3 -2.9 2.1 -1.6
OECD Europe -13.4 0.95 -0.03 0.36 -0.02 0.99
t-statistics -11.6 3.5 -2.2 5.7 -0.86
OECD Pacific -5.2 1.1 -0.07 0.04 0.01Q 0.99
t-statistics -3.5 4.9 -2.3 0.54m 3.5
Developing Countries -11.3 0.64¡ -0.02 0.26 -0.1 0.99
t-statistics -10.5 3.9 -1.1 2.01 -2.7
* Real GDP per urban population.
Q Trend variable.
m Null hypothesis rejects at 0.58 critical value.
¡ GDP.
OECD, Organization for Economic Cooperation and Development; GDP, Gross Domestic Product.

all variables are first difference stationary. The long term or cointegrated equation
may exist, should the error term of each equation reject the null hypothesis of unit root
existence. As shown in Table 3, all equations represent a long-term cointegrating relation
among the variables.
The equations were estimated using the data sample of 1980–2005 using ordinary least
squares technique. Table 4 provides the estimated coefficients along with their t-statistics.
In all regions, the income elasticity of jet fuel demand is higher than 0.5, indicating that
the most dominant variable in aviation transportation demand is economic growth. This
elasticity is close to unity in Europe and is higher than unity in OECD Pacific. Therefore, a
1 per cent increase in income led to more than 1 per cent increase in jet fuel demand in the
long term in OECD Pacific ceteris paribus. This would be about 1 per cent in OECD
Europe and about half a per cent in OECD North America and developing countries. The
price elasticity of jet fuel demand is very low for all regions, confirming a better manage-
ment of the aviation industry to cope with higher fuel prices through better LF, and trans-
ferring a portion of fuel costs to customers. Better technical fuel efficiency and LF, along
with low-price elasticity of business passenger demand, has led the aviation industry to be
resilient to higher oil prices.
Fuel demand elasticity of PKP is low for all regions. One per cent increase in PKP led
to about 0.3, 0.2, 0.1 and 0.04 per cent increase in jet fuel demand in OECD Europe, devel-
oping countries, OECD North America and OECD Pacific, respectively.

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

38 Mohammad Mazraati and Osama M. Alyousif

4.3. Forecast
The reference case is forecast based on assumptions on exogenous variables in each
region. The assumptions are based on the historical trend and the most recent information
that allows us to assume the future values on our best guess. Assumption on jet fuel price is
identical for all regions. Due to the fact that jet fuel prices have experienced high increase
in recent years, we assume that a moderate half a per cent increase will settle around its
long-term value. It is worth mentioning that jet fuel demand responses to jet fuel prices are
insignificant, as was seen in estimated price elasticities. The GDP assumptions are in line
with those of the Organization of the Petroleum Exporting Countries (OPEC) OPEC
World Energy Model (OWEM) reference case in World Oil Outlook 2008 (OPEC, 2008).

OECD North America

Assumptions for OECD North America on exogenous variables in the reference case are
presented in Table 5. We assume that the real GDP of OECD North America would grow
around 2.4 per cent, on the average, per annum. This rate locates between the average long
term and the most recent growth rates. Despite high oil prices, it is assumed that the CPI
would grow not more than 3.2 per cent per annum. But its effects on jet fuel demand would
be moderate due to the fact that price elasticity of jet fuel demand is very low and the
market is already well developed.
Urban population has increased by 1.6 and 1.4 per cent per annum during 1980–2005
and 2000–2005, respectively. It is assumed that urbanisation is continued to expand at the
rate of 1 per cent per annum. The PKP has experienced 5.3 and 1.9 per cent increase during
the mentioned periods, respectively, and we assume that it would grow at 1.5 per cent per
annum till 2030.
Figure 7 draws the fit and forecast of the model. It is obvious that the jet fuel demand
growth rate is moderate. Jet fuel demand grew at 1.81 per cent per annum during 1980–
2005 and 2.97 per cent during 2000–2005, but it is expected to grow at an average lower
rate in the future. Jet fuel demand would increase by 1.7 per cent per annum during 2006–

Table 5 Growth rate of exogenous variables for OECD North America (%)

Regions Variables 1980–2005 2000–2005 2006–2030

OECD North America Real GDP 2.9 2.3 2.4
CPI 5.3 3.2 3.2
Urban population 1.6 1.4 1.0
PKP 5.3 1.9 1.5
OECD, Organization for Economic Cooperation and Development; GDP, gross domestic product;
CPI, Consumer Price Index; PKP, passenger kilometre performed.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 39

2.5

1.5
mb/d

0.5

0
1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Fit and forecast Actual

Figure 7 Fit and forecast of aviation fuel demand in OECD North America.

Table 6 Growth rate of exogenous variables for OECD Europe (%)

Regions Variables 1980–2005 2000–2005 2006–2030

OECD Europe Real GDP 2.2 1.8 1.8
CPI 5.3 4.3 4.0
Urban population 0.78 0.76 0.75
PKP 6.1 5.5 5.0
OECD, Organization for Economic Cooperation and Development; GDP, gross domestic product;
CPI, Consumer Price Index; PKP, passenger kilometre performed.

2030. Demand level will reach to 2.4 million barrels per day (mb/d) in 2030 from 1.8 mb/d
in 2005, showing an increment of half a million barrels per day.

OECD Europe
In the OECD Europe region, GDP has experienced an average annual rate of 2.2 per cent
within 1980–2005 and 1.8 per cent during the most recent years (see Table 6). It is
assumed that GDP would grow 1.8 per cent per annum in the future. The long-term CPI
growth rate was 5.3 per cent, but declined to 4.3 per cent during the most resent years. A 4
per cent annual growth rate for the future is assumed. Urban population grows very steady

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

40 Mohammad Mazraati and Osama M. Alyousif

2.5

1.5
mb/d

0.5

0
1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Fit and forecast Actual

Figure 8 Fit and forecast of aviation fuel demand in OECD Europe.

in the OECD Europe region. Urban population growth rates in different periods were low.
The rate is set to be 0.75 per cent per annum for the future. The PKP grew at 6.1 and 5.5 per
cent per annum within 1980–2005 and 2000–2005, respectively. In line with the historical
data and the level of market saturation, it is assumed that PKP to grow 5 per cent per annum
up to 2030.
Figure 8 shows the fitness and forecast of aviation demand in OECD Europe. The
aviation fuel demand was less than half a million barrels per day in 1980 that reached to
more than 1 mb/d in 2005. It experienced about 2.3 per cent increase per annum during this
period. It is expected to grow at 2.1 per cent per annum till 2030. Aviation demand will
reach to 2.2 mb/d in 2030. This region would contribute an increment of 1.1 mb/d by 2030.

OECD Pacific
Growth rates of exogenous variables are provided in Table 7. Real GDP grew at 2.7 per
cent within 1980–2005 but it was 1.8 per cent during 2000 to 2005. It is assumed that GDP
would grow at the moderate rate of 1.7 per cent per annum. Inflation rate is expected to be
2.2 per cent per annum. Urban population grows at a very low level of 0.09 per cent per
annum, as it was the case historically. The PKP grew at high rate of 7.3 per cent per annum
during 1980–2005, but declined to only 0.2 per cent during 2000–2005, mainly due to a
sharp decline in 2002. The PKP declined 13 per cent in that year, followed by a spiky

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 41

Table 7 Growth rate of exogenous variables for OECD Pacific (%)

Regions Variables 1980–2005 2000–2005 2006–2030

OECD Pacific Real GDP 2.7 1.8 1.7
CPI 3.7 2.1 2.2
Urban population 0.12 0.09 0.09
PKP 7.3 0.2 5.0
OECD, Organization for Economic Cooperation and Development; GDP, gross domestic product;
CPI, Consumer Price Index; PKP, passenger kilometre performed.

1.20

1.00

0.80
mb/d

0.60

0.40

0.20

0.00
1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Fit and forecast Actual

Figure 9 Fit and forecast of aviation fuel demand in OECD Pacific.

increase in the following year but stabilised around 2 per cent afterward. It is assumed that
the passenger traffic would grow at its historical average rate of 5 per cent per annum.
Figure 9 depicts the fitness and forecast for aviation fuel demand model for this
region. Aviation fuel demand was very low in 1980 at the level of 0.13 mb/d. It grew at
2.5 per cent per annum and reached to 0.4 mb/d in 2005. Based on model’s result, it is
expected to grow at 2.2 per cent per annum to reach to around 1 mb/d in 2030. The incre-
mental oil demand would be about half a million barrels per day by 2030.

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

42 Mohammad Mazraati and Osama M. Alyousif

Table 8 Growth rate of exogenous variables for selected developing countries (%)

Regions Variables 1980–2005 2000–2005 2006–2030

Developing countries Real GDP 7.2 7.6 7.0
CPI 6.7 4.1 3.5
Urban population 1.0 1.0 0.8
PKP 9.2 9.6 8
GDP, gross domestic product; CPI, Consumer Price Index; PKP, passenger kilometre performed.

Developing countries
In the selected developing countries, GDP grew at 7 per cent rate per annum during 1980–
2005 (Table 8). It grew at an even higher rate during the most recent years. It assumed that
GDP would grow at 7.0 per cent per annum over the period from 2006 to 2030. The urban
population grew at 1 per cent during both periods. It is expected that the urbanisation grow
at a lower rate of 0.8 per cent per annum over the forecast period. Since the aviation in-
dustry is in it early expansion, the demand for aviation services is expanding. The PKP
grew at 9.3 per cent per annum during 1980–2005. The annual growth rate was even higher
during the most recent years. It was 9.6 per cent per annum within 2000–2005. The
industry expansion is expected for the future. It is plausible to consider 8 per cent annual
growth rate over the period to 2030.
Figure 10 draws the actual, fitted and forecast trend of aviation fuel demand in
selected developing countries. It is obvious that the behaviour is distinguished from those
of OECD countries. Demand for aviation fuel was only 0.09 mb/d in 1980. It grew annu-
ally by more that 3 per cent over the period to 2005. The demand reached to half a million
barrels per day in 2005. The model forecast shows that demand for these countries would
grow at about 3 per cent per annum to reach 2.5 mb/d in 2030. The incremental aviation
fuel demand would be 2 mb/d in 2030.
Countries of this study would add more than 4 mb/d to the aviation fuel demand in
2030. About 50 per cent of this incremental oil demand will be contributed by developing
countries.

4.4. Efficiency evolution

The fuel efficiency has been improving in the course of time and it is expected to continue
in the future. The realised efficiency during the last two decades has saved huge amount
of fuel demand. Figure 11 shows fuel efficiency criteria through fuel consumption per
1000 PKP by each region. It is obvious that the current fuel per passenger kilometre is
far below that of the early 1980s, justifying the realised better current fuel efficiency.
The model forecasts the continuation of efficiency improvement at a lower rate, showing
a convergence by regions.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 43

2.5

2
mb/d

1.5

0.5

0
1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Fit and forecast Actual

Figure 10 Fit and forecast of aviation fuel demand in selected developing countries.

1.60

1.40

1.20
Liter per 1000 PKP

1.00

0.80

0.60

0.40

0.20

0.00
1980

1982

1984
1986

1988
1990

1992
1994
1996

1998
2000
2002
2004

2006
2008

2010
2012

2014

2016
2018

2020
2022

2024
2026

2028

2030

OECD N.A. OECD Europe OECD Pacific Developing countries

Figure 11 Aviation fuel consumption per 1000 passenger kilometre performed (PKP).

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

44 Mohammad Mazraati and Osama M. Alyousif

In the absence of current ongoing efficiency improvement, the demand for aviation
fuel would have been 4.7 mb/d more in 2030 for the countries of this study. The amount of
fuel saving due to ongoing efficiency improvement would have been 2.42 mb/d in devel-
oping countries, 1.4 mb/d in OECD Europe, 0.44 mb/d in OECD Pacific and 0.37 mb/d in
OECD North America in 2030. There is no doubt that the forecast faces different uncer-
tainties. Mismanagement of load traffic, congestion in airports, low LF and many other
factors may lead to less energy saving. But a big portion of efficiency comes from techni-
cal efficiency, which is granted by replacing the ageing and retired, less efficient aircrafts
with new, efficient aircrafts.

5. Conclusions
High oil prices impose financial burdens on the aviation sector. Fuel cost as per cent of
passenger revenue was less than 15 per cent in 2000 but reached to over 40 per cent in the
first quarter of 2008 (Heimlich, 2008). But price elasticities show an insignificant impact
of fuel price on aviation fuel demand. That is due to a better ongoing efficiency initiated
from better LF and aircraft fuel efficiency, labour restructuring (Heimlich, 2008) and a
moderate increase in fare prices. For example, fuel efficiency of US airlines has been up
30.8 per cent since 1990. Forecast in the reference case explains an increase of 4 mb/d in
aviation fuel demand over the period to 2030. The gradual improvement of aircraft fuel
efficiency and retiring ageing fleets (1 per cent reduction in fuel consumption) and load
management could relax a portion of the future demand expansion, while a better eco-
nomic growth may offset that. The model estimates that in the absence of the ongoing effi-
ciency improvement in the aviation sector, the aviation fuel demand for the countries of
this study would have been 4.7 mb/d more in 2030.
Aviation fuel in all OECD regions grows steady, but consumption of developing coun-
tries is expected to increase faster. The aviation fuel consumption in selected developing
countries would surpass OECD Pacific soon. The consumption would surpass OECD
Europe and OECD North America around 2025 and 2030, respectively.

Acknowledgements
The authors gratefully would like to acknowledge the opportunity provided by the
Organization of the Petroleum Exporting Countries (OPEC) through the summer
fellowship program to conduct this study. The assistance and support from the Statistical
Department Studies in the office of the OPEC Secretariat is also acknowledged. The
authors also would like to acknowledge the support provided by Kuwait Petroleum
Corporation.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

 Impacts of fuel efficiency 45

Notes
1. An example of this is when a carrier like United Airlines flies a passenger plane carrying 200
revenue passengers from New York to London; the distance from New York to London is
5533 kilometres, then the passenger kilometre performed (PKP) for this flight is
200 ¥ 5533 = 1,106,600 passengers per kilometre flown.
2. Aviation traffic is defined as the sum of PKP and freight kilometre performed. In this study,
the PKP is a proxy variable for aviation traffic, although a homogenised passenger and freight
travelled could also be used as aviation traffic.

References
Abadie, O., 2007. Jet Fuel: How High a Flyer? Demand, Supply, and the Endless Quest for
Efficiency. Cambridge Energy Research Associates, Cambridge, Massachusetts, USA.
Airbus, 2007. Global Market Forecast 2007–2026. URL
http://www.airbus.com/fileadmin/documents/gmf/PDF_dl/00-all-gmf_2007.pdf
[accessed July 2008].
Babikian, R., Lukachko, S.P., Waitz, I.A., 2005. The Historical Fuel Efficiency Characteristics
of Regional Aircraft from Technological, Operational, and Cost Perspectives. URL
http://web.mit.edu/aeroastro/people/waitz/publications/Babikian.pdf [accessed July 2008].
Boeing, 2007. Current Market Outlook 2008–2027. URL
http://www.boeing.com/commercial/cmo/pdf/Boeing_Current_Market_Outlook_2008_to_
2027.pdf [accessed July 2008].
Brons, M., Pels, E., Nijkamp, P., Rietveld, P., 2001. Price elasticities of demand for passenger air
travel: a meta-analysis. Tinbergen Institute Discussion Paper.
FAA (Federal Aviation Agency), 2008. FAA Aerospace Forecast. URL
http://www.faa.gov/data_research/aviation/aerospace_forecasts/2008-2025/media/
2008%20Forecast%20Document.pdf [accessed July 2008].
Gately, D., 1987. Taking off: the U.S. demand for air travel and jet fuel. The Energy Journal 9, 4,
63–91.
Greene, D.L., 1990. Commercial aircraft fuel efficiency potential through 2010. Proceedings of the
25th Intersociety Energy Conversion Engineering Conference 4, 106–111.
Hayashi, P. and Trapani, J., 1987. The impact of energy costs on domestic airline passenger travel.
Journal of Transport Economics and Policy 21, 1, 73–86.
Heimlich, J., 2008. Coping with Sky-High Jet Fuel Prices. URL
http://www.airlines.org/airlines_fuel.pdf [accessed 22 July 2008].
IATA, 2008. Fuel Efficiency. URL http://www.iata.org/whatwedo/environment/fuel_efficiency.
htm [accessed July 2008].
ICAO, 2008. ICAO data website. International Civil Aviation Organization. URL
http://icaodata.com [accessed July 2008].
IEA, 2008. Energy Balances of OECD Countries. International Energy Agency, Paris.

© 2009 The Authors. OPEC Energy Review March 2009

Journal compilation © 2009 Organization of the Petroleum Exporting Countries

46 Mohammad Mazraati and Osama M. Alyousif

Kroo, I., 2005. Nonplanar wing concepts for increased aircraft efficiency. VKI Lecture Series on
Innovative Configurations and Advanced Concepts for Future Civil Aircraft, 6–10 June,
Stanford University, USA.
Lee, J.J., Lukachko, S.P., Waitz, I.A., Schafer, A., 2001. Historical and future trends in aircraft
performance, cost, and emissions. Annual Review of Energy Environment 26, 167–200.
OPEC, 2008. World Oil Outlook (WOO) 2008. OPEC Secretariat, Vienna, Austria. Also available
on http://www.opec.org.
Vedantham, A. and Oppenheimer, M., 1998. Long-term scenarios for aviation: demand and
emissions of CO2 and NOx. Energy Policy 26, 8, 625–641.

OPEC Energy Review March 2009 © 2009 The Authors.

Journal compilation © 2009 Organization of the Petroleum Exporting Countries