World Energy Model 2005

WORLD ENERGY MODEL 20051 Background Since 1993, the IEA has provided medium- to long-term energy projections using the World Energy Model (WEM). The WEM has undergone significant transformations for recent editions of the World Energy Outlook (WEO), increasing the regional and sectoral disaggregation and enhancing the supplyside modules. The WEM used to produce this Outlook is the ninth version of the model. It is designed to analyse: x

x

x

Global and regional energy prospects: These include trends in demand, supply availability and constraints, international trade and energy balances by sector and by fuel to 2030. Environmental impact of energy use: CO2 emissions from fuel combustion are derived from the detailed projections of energy consumption. Investment in the energy sector: The model evaluates investment requirements in the fuel supply chain needed to meet projected energy demand to 2030 as well as demand-side capital expenditures.

The analysis is conducted under the following scenarios: x

x

x

1

Reference Scenario: The primary objective of this scenario is to identify and quantify the key factors that are likely to affect energy demand and supply and to build projections based on these factors. Key variables such as economic growth, demographic trends and technological innovation are given exogenously to the model. The Reference Scenario takes into account only those government policies and measures that had been enacted by mid-2005. World Alternative Policy Scenario: The World Alternative Policy Scenario is developed and used to analyse the impact of stronger policy actions and technological developments on energy demand, supply, trade, investments and emissions. This scenario, first introduced in WEO-2000, has been improved and includes an expanded number of policies and measures that countries are either currently considering or might reasonably be expected to implement over the projection period. Deferred Investment Scenario: The Deferred Investment Scenario (DIS), new to this Outlook, assesses how global energy markets might evolve if upstream oil investment in each MENA country were to

This note provides an overview of the methodology used to produce the World Energy Outlook 2005. For further information, please send all queries to [email protected].

increase slower over the projection period than in the Reference Scenario.

New Features in WEO 2005 The model for this Outlook has been further expanded to include nine new country models, Algeria, Egypt, Iran, Iraq, Kuwait, Libya, Qatar, Saudi Arabia and the United Arab Emirates and two new regional aggregates, other Middle East and other North Africa, to cover the entire Middle East and North Africa (MENA) region (Figure 1).

Other important additions include: x x x x

An oil and gas field-by-field production analysis for all the key countries in the MENA region. A World Oil Equilibrium Model to support the analysis in the Deferred Investment Scenario. A water desalination module to project energy demand for desalination in Algeria, Kuwait, Libya, Qatar, Saudi Arabia and the UAE. A global refinery model to project product demand and capacity additions to 2030.

3

* North Africa includes Algeria, Egypt and Libya as well as Other North Africa (Morocco and Tunisia).

* Middle East includes Iran Iraq, Kuwait, Qatar, Saudi Arabia and the UAE as well as Other Middle East (Bahrain, Israel, Jordan, Lebanon, Oman, Syria and Yemen).

Figure 1: World Energy Model Regions

Model Structure The WEM is a mathematical model made up of the following main modules: x

Final energy demand

x

Power generation and heat plants2

x

Refinery

x

Other transformation

x

CO2 emissions

x

Fossil fuel supply

x

Investment

Figure 2 provides a simplified overview of the structure of the model.

Figure 2: World Energy Model Overview

The main exogenous assumptions concern economic growth, demographics, international fossil fuel prices3 and technological developments4. Electricity

2 3

Includes the desalinated water module. The oil price is endogenously determined in the Deferred Investment Scenario.

consumption and electricity prices dynamically link the final energy demand and power generation modules. The refinery model projects throughput and capacity requirements based on global oil demand. Primary demand for fossil fuels serves as input for the supply modules. Complete energy balances are compiled at a regional level and the CO2 emissions of each region are then calculated using derived carbon factors. Investment needs for fuel supply are evaluated through the methodology described in the World Energy Investment Outlook 2003. A new methodology for calculating demand-side capital expenditures was introduced in WEO-2004.

Technical Aspects The development and running of the WEM requires access to significant quantities of historical data on economic and energy variables. Most of the data are obtained from the IEA’s own databases of energy and economic statistics. A significant amount of additional data from a wide range of external sources is also used. The parameters of the equations of the demand-side modules are estimated econometrically, usually using data for the period 1971-20035. Shorter periods are sometimes used where data are unavailable or significant structural breaks have occurred. To take into account expected changes in structure, policy or technology, adjustments to these parameters are sometimes made over the Outlook period, using econometric and other modelling techniques. Simulations are carried out on an annual basis. Demand modules can be isolated and simulations run separately. This is particularly useful in the adjustment process and in sensitivity analyses related to specific factors. The WEM makes use of a wide range of software, including specific database management tools, econometric software and simulation programmes.

Description of the Modules Final Energy Demand The OECD regions and the major non-OECD regions are modelled in considerable sectoral and end-use detail. Specifically: x

4

Industry is separated into six sub-sectors, allowing a detailed analysis of trends and drivers in the industrial sector.

Parts of the technological developments are endogenously determined through appropriate learning curves. 5 Data for 2004 is used when available.

x x x

Residential energy demand is separated into five end-uses by fuel. Services demand is modelled as three end-uses by fuel. Transport demand is modelled in detail by mode and fuel.

Final energy demand is modelled at the sectoral level for each of the MENA countries and regions, but not at such a disaggregated end-use level. Total final energy demand is the sum of energy consumption in each final demand sector. In each sub-sector or end-use, at least six types of energy are shown: coal, oil, gas, electricity, heat and renewables. However, this level of aggregation conceals more detail. For example, the different oil products are modelled separately as an input to the refinery model. Within each sub-sector or end-use, energy demand is estimated as the product of an energy intensity and an activity variable. In most of the equations, energy demand is a function of the following variables: x

Activity variables: This is often a GDP or GDP-per-capita variable. In many cases, however, a specific activity variable, which is usually driven by GDP, is used.

x

Prices: End-user prices are calculated from assumed international energy prices. They take into account both variable and fixed taxes, as well as transformation and distribution costs. For each sector, a representative price (usually a weighted average) is derived. This takes account of the product mix in final consumption and differences between countries. This representative price is then used as an explanatory variable directly, with a lag, or as a moving average.

x

Other variables: Other variables are used to take into account structural and technological changes, saturation effects or other important drivers.

Industry Sector The industrial sector in the OECD regions is split into six sub-sectors: iron and steel, chemicals, paper and pulp, food and beverages, non-metallic minerals and other industry. For the non-OECD regions, the breakdown is typically based on four instead of six sub-sectors. The intensity of fuel consumption per unit of each sub-sector’s output is projected on an econometric basis. The output level of each sub-sector is modelled separately and is combined with projections of its fuel intensity to derive the consumption of each fuel by sub-sector. This allows more detailed analysis of the drivers of demand and of the impact of structural change on fuel consumption trends.

The increased disaggregation also facilitates the modelling of alternative scenarios, where end-use shares and technology descriptions are applied in conjunction with capital stock turnover models to analyse in detail the impact of alternative policies or different choices of technology.

Transport Sector The WEM fully incorporates a detailed bottom-up approach for the transport sector in all OECD and major non-OECD regions. Transport energy demand is split between passenger and freight and is broken down among light duty vehicles, buses, trucks, rail, aviation and navigation (Figure 3). Passenger cars and light trucks are subdivided by fuel used – gasoline, diesel, alternative fuels or hybrids of these. Freight trucks are divided between gasoline- and dieseldriven. The gap between test and on-road fuel efficiency is also projected. For each region, activity levels for each mode of transport are estimated econometrically as a function of population, GDP and price. Additional assumptions to reflect saturation of passenger vehicle ownership are also made. Transport activity is linked to price through elasticity of per km fuel cost, which is estimated for all modes except passenger buses and trains and inland navigation. This elasticity variable accounts for the “rebound” effect of increased car use that follows improved fuel intensity. Energy intensity is projected by transport mode, taking into account changes in energy efficiency and fuel prices. Stock turnover is explicitly modelled in order to allow for the effects of fuel efficiency regulations for new cars on the energy intensity of the whole fleet.

8

Figure 3: Structure of the Transport Demand Module

Residential and Services Sectors In WEO-2004, detail in the energy demand model for the residential and services sectors was significantly increased for the major non-OECD regions (Figure 4). For the other non-OECD regions, energy consumption in these sectors is calculated econometrically for each fuel as a function of GDP, the related fuel price and the lag of energy consumption. The residential sector module splits consumption by fuel into five end-uses: space heating, water heating, cooking, lighting and appliances use. The total fuel demand is projected per household consumption. The services sector module disaggregates demand into the following components: heating, hot water and cooking uses (HHC) and other electricity end-uses, including ventilation, space cooling and lighting. The total fuel demand for HHC is projected per square metre of floor area. Floor area in services is estimated as a function of valueadded in the sector, which in turn is projected from GDP assumptions. The total demand for HHC is then allocated to two components: an “existing stock” model determines energy consumption based on the historical shares of each fuel, while a portion of demand is allocated to ”new stock”, where fuel shares are a function of both relative prices and existing shares of each fuel.

World Energy Model 2005

10

Figure 4: Structure of the Residential and Services Sectors Demand Modules

Refinery Module For each WEO region, the module estimates a base case refinery output, given past domestic demand in the Reference Scenario and the region’s share in global trade. Demand in the medium term is based on existing projects in all WEO regions, including MENA, China, India and the OECD. On the basis of information gathered about these projects and regional capacity utilisation rates, the model determines the additional capacity needed by region. Historical capacity figures for the module are based on data from the Oil and Gas Journal and British Petroleum (2005). The 2004 installed capacity figures for MENA countries, however, are based on IEA analysis and on industry journals and national statistics from countries in the region, the Arab Oil and Gas Directory and the Arab Oil and Gas Journal. Current capacity figures for non-MENA countries are from the IEA’s Oil Market Report and company sources. Throughput and capacity projections are based on the Reference Scenario oil demand projections by world region/country. The model adjusts global refinery demand among the major regions: OECD, transition economies, China, India, Middle East, North Africa and other developing regions. Capacities are then disaggregated in each MENA country according to production, demand, exports of crude and refined products, and costs. Figure 5 shows the structure of the refinery module. After determining individual refinery output and capacity, the module calculates the global oil products balance. Total demand for oil products (excluding direct use of crude) is matched with the total supply of oil products, including products from NGLs and GTLs. Thus, at the global level:

S ref where

D   NGLs  GTLs  Dref

Sref = total refinery output D = Total oil product demand NGLs= natural gas liquid products GTLs= gas-to-liquid products Dref =refinery own use

The refinery model balances supply and demand through an optimisation process. Excess demand is split according to an optimisation matrix which takes into account unit costs, environmental and political constraints and capacity constraints. Figure 5: Structure of the Refinery Module

There are three types of distillation capacity additions: new refinery (highest cost); added capacity at an existing refinery; and capacity creep (lowest cost). Distillation capacity refers to calendar day capacity. In OECD countries, no new refineries are assumed to be built. Over the projection period, production increases are all projected to be derived from capacity creep. Investment requirements are separated between additions investment and conversion investment. Additions investments are based on current costs which vary among regions/countries. For additions investments, the model projects the share of distillation capacity additions for each region/country and allocates costs accordingly. Conversion investments are based on the estimated costs of modifying existing capacity to cope with new demand (lighter products) or new environmental restrictions on products (sulphur content). Demand is divided into three products: light, middle and heavy. The model uses the sectoral breakdown and the region/country specification to project product demand.

The refinery module projects the necessary capacity conversions, based on the demand projections for light, middle and heavy products and on anticipated environmental regulations. The projections are used to calculate investment in cost per mb/d converted. A weighted average technology cost was calculated, taking into account the cost of each different technology such as catalytic crackers and hydro-skimmers, from industry sources. The refinery investments do not include maintenance costs. Power Generation and Heat Plants The power generation module calculates the following: x

Amount of electricity generated by each type of plant to meet electricity demand

x

Amount of new generating capacity needed

x

Type of new plants to be built

x

Fuel consumption of the power generation sector

x

Electricity prices

The structure of the power generation module is described in Figure 6. Electricity generation is calculated using the demand for electricity, taking into account electricity used by power plants themselves and system losses. The need for baseload, medium and peaking capacity is based on an assumed load curve. New generating capacity is the difference between total capacity requirements and plant retirements using assumed plant lives. When a new plant is needed, the model makes its choice on the basis of electricity generating costs, which combine capital, operating and fuel costs over the whole operating life of a plant, using a given discount rate, plant efficiency and plant utilisation rate. The model considers the following types of plant: x x x x x x x x x x x x

Coal, oil and gas steam boilers Combined-cycle gas turbine (CCGT) Open-cycle gas turbine (OCGT) Integrated gasification combined cycle (IGCC) Oil and gas internal combustion Fuel cell Nuclear Biomass Geothermal Wind (onshore) Wind (offshore) Hydro (conventional)

x x x x

Hydro (pumped storage) Solar (photovoltaics) Solar (thermal) Tidal/wave

Capacities for nuclear power plants are based on assumptions, which are in turn based on an assessment of government plans. Where market conditions prevail, the assumptions are influenced by international fossil fuel prices. Fossil fuel prices and efficiencies are used to rank plants in ascending order of their short-run marginal operating costs, allowing for assumed plant availability. Once the mix of generation plants has been determined, the fuel requirements are then deduced by plant type, using an assumed efficiency. The marginal generation cost of the system is calculated, and this cost is then fed back to the demand model to determine the final electricity price. The combined heat and power (CHP) option is considered for fossil fuel and biomass plants. CHP, renewables and distributed generation are sub-modules of the power generation module. The CHP sub-module uses the potential for heat production in industry and buildings together with heat demand projections, which are estimated econometrically in the demand modules. The distributed generation (DG) sub-module is based on assumptions about market penetration of DG technologies. The projections for renewable electricity generation are derived in a separate module. In WEO-2004, the future deployment of renewable energies for electricity generation and the investment needed for such deployment were assessed on the basis of the renewable energy potential in each country/region. The development of renewables is based on an assessment of the potential and costs for each source (biomass, hydro, photovoltaics, solar thermal electricity, geothermal electricity, on- and offshore wind, tidal and wave) in each of twenty world regions. By including financial incentives for the use of renewables and non-financial barriers in each market, as well as technical and social constraints, the model calculates deployment as well as the resulting investment needs on a yearly base for each renewable source in each region.

World Energy Model 2005

15

Figure 6: Structure of the Power Generation Module

Desalinated Water Module The desalinated water module is closely linked to the power generation module through the interrelationship between water and electricity production in combined water and power (CWP) plant production. The desalinated water module generates projections for Algeria, Kuwait, Libya, Qatar, Saudi Arabia and the UAE, for the following variables: x x x x x x

Agricultural, residential and industrial water consumption by country. Water production from desalination plants. Desalination capacity. Fuel consumption for desalinated water production. Electricity capacity and electricity production in combined water and power plants. Investment needs for desalination plants.

Water demand projections are calculated on the basis of historical data, using assumptions for population growth (the main driver for residential water sector demand) and per capita income. Losses in the distribution systems are assumed to decline over the projection period. Water supply includes surface water, renewable groundwater resources, nonrenewable groundwater resources, re-treated waste-water and current production of desalinated water. Surface water is assumed to remain constant unless specific plans are in place to increase supply, such as in Libya. Non-renewable groundwater resources are assumed to decrease at different depletion rates. The average lifetime of desalination plants is assumed to be 30 years. Over the projection period, if water consumption is higher than supply from existing sources, the difference is assumed to be covered by production at new desalination plants. New plants are split between technologies (reverse osmosis and distillation units). Water production, desalination capacity and electricity consumption are calculated for reverse osmosis plants. All new distillation plants are assumed to be associated with electricity plants in combined water and power (CWP) plants. Demand projections for distillation units are based on water demand, capacity factors, water-to-power ratios and performance ratios. CWP electrical capacity and production are outputs of this sub-module and are used in the power generation module. Energy consumption in CWP plants for water production is the sum of the electricity consumed and the fuel requirements for the steam consumed. Figure 5 shows the structure of the desalinated water module. Investment needs are based on the desalination capacity additions, unitary costs and the type of technology.

World Energy Model 2005

17

Figure 7: Structure of the Desalinated Water Module

Allocation of Fuel Requirements Desalination plants using reverse osmosis (RO) consume electricity to drive the pump. The fuel requirements for this type of plant are therefore linked to the efficiency of the country’s power plants. For distillation units, the calculation of the primary energy needed to produce the desalinated water is more complex and is heavily dependent on the methodology adopted for the allocation of the fuel needed to produce the steam. Distillation plants use both electricity and steam to produce water. The electricity consumption (A) is directly derived from the water production and the primary energy needs (as for the RO plants) are calculated applying the power plants’ electrical efficiency. The calculation of the fuel needs to produce steam depends on the methodology adopted for the allocation of the fuel used to produce the electricity and the fuel used to produce the steam. The following methodology is adopted in the WEM: 1. 2. 3.

A fixed efficiency of 90% is applied to the steam production in order to estimate the related input needs (B). The rest of the fuel input is assumed to be used for electricity generation. The total energy requirements (A+B) for water production in a CWP plant are the sum of the primary energy needs for the electricity consumption (A) and for the steam production (B).

CO2 Emissions For each region, sector and fuel, CO2 emissions are calculated by multiplying energy demand by an implied carbon emission factor. Implied emission factors for coal, oil and gas differ between sectors and regions, reflecting the product mix. They have been calculated from year-2003 IEA emissions data for all regions.

Fossil Fuel Supply Oil Module Two different oil modules are used for the Reference and Deferred Investment Scenarios: x

The Reference Scenario oil module uses the methodology described in Annex C in WEO-2004 (summarised below) with the addition of a field-by-field analysis for MENA countries. The field-by field-analysis is a bottom-up approach that serves to support our top-down results for the call on MENA oil.

x

The Deferred Investment Scenario oil module uses the field-by-field analysis to calculate the MENA production (as described in Chapter 7 of WEO-2005), which is an assumption behind the World Oil Equilibrium Model (described below).

Reference Scenario Oil Module The purpose of this module is to determine the level of oil production in each region. Production is split into three categories:6 x

Non-MENA

x

MENA

x

Non-conventional oil production

Total oil demand is the sum of regional oil demand, international marine bunkers and stock changes. MENA conventional oil production is assumed to fill the gap between non-MENA production and non-conventional and total world oil demand (Figure 8).

6

MENA and non-MENA production includes crude oil, NGLs and condensates.

Figure 8: Structure of the Oil Supply Module

The derivation of non-MENA production of conventional oil (crude and natural gas liquids) uses a long-term approach. This approach involves the determination of production according to the level of ultimately recoverable resources and a depletion rate estimated by using historical data and industry sources. Ultimately recoverable resources depend on a recovery factor. This recovery factor reflects reserves growth, which results from, among other things, improvements in drilling, exploration and production technologies. The trend in the recovery rate is, in turn, a function of the oil price and of a technological improvement factor. Non-conventional oil supply is determined mainly by the oil price. Higher oil prices bring forth greater non-conventional oil supply over time.

Field-by-Field Analysis for MENA countries In the field-by-field analysis, some 200 fields in the MENA region are analysed according to a two-step methodology: i) a supply curve analysis and ii) assessment and modifications based on existing or planned projects for a specific field. This method is applied to both oil and gas fields Projected output from each field is a function of the field discovery year, yearly crude oil, condensates and gas production, recoverable reserves and hydrocarbons initially in place. Data has been checked and verified to ensure internal consistency. The primary source is the IHS Energy database. Additional information stems from a number of other sources, including international

oilfield service companies, national and international oil companies, consultants and the IEA’s own databases.7 For each field, supply projections are based on a time series of production from that field and are adjusted after an assessment of current and planned field development projects. Production decline-curve analysis is used to regress the historical data and to forecast production over the next 25 years (2005-2030). The following exponential decline equation is used: Q t0

where



n

Q

t0 ˜ e

k ˜n

Q = production volume n = number of years following the initial year, t0, k = coefficient calculated as: k=

ln [Q t / Q t  T ] T

where T is the number of past production years which are used in the calculation of the k coefficient. The coefficient k is the slope of the production curve. Figure 9 shows historical production data and the projection for Saudi Arabia’s super-giant Abqaiq field using the above exponential function.

7

See Box 4.4 in Chapter 4 of WEO-2005 for a more in-depth explanation of the field production analysis methodology.

Figure 9: Abqaiq Production – Historical and Forecast

billion barrels per year

1

Production forecast

0.1 Production data

0.01 1970

1980

1990

2000

2010

2020

2030

Output from the field peaked at more than 1 mb/d in 1973 and then maintained an average of 0.8 mb/d in the 1970s. Production declined dramatically in the 1980s, due to the oil price collapse and to the introduction of OPEC quotas. Saudi Arabia acted as a swing producer, significantly reducing the production from its fields. In the 1990s, production recovered to around 0.6 mb/d. Output in 2003 was slightly less than 0.4 mb/d. In the regression analysis, only data from 1970 to 1980 and from 1990 to 2003 are used. The lower production rates in the 1980s are the result of political/commercial upheavals and do not reflect reservoir performance. Based on the historical trend, production would decline at an effective rate of 2.25% per year (annual decline rate = 1 - ek), falling to 0.24 mb/d in 2030. However, given the field’s huge reserves and improved oil recovery technologies, Abqaiq can easily increase its production over the next 5 to 10 years. Thus, the WEO-2005 projects that production will be higher than 0.24 mb/d in 2030, largely because improved technology implies an upward shift of the production forecast curve. All of the oil and gas production projections in this Outlook were subject to the same scrutiny as the Abqaiq example above. Energy experts, within and outside of the MENA region, were consulted in order to establish if the results of the regression analyses were sound and defendable.

Deferred Investment Scenario Oil Module: World Oil Equilibrium Model 8 The World Oil Equilibrium Model used in the Deferred Investment Scenario is a simulation model which solves for the equilibrium oil price that equates world oil supply and demand.9 The model is not intended to optimise MENA or nonMENA production. In the World Oil Equilibrium Model , MENA oil supply is exogenous and is calculated through the field production analysis, according to the methodology and assumptions underlying the Deferred Investment Scenario (Chapter 7). Non-MENA supply is a function of the oil price and a technology parameter , which reflects the future role oilfield technology is expected to play. This parameter is based on historic data, especially in the North Sea, and it ranges from 0.01 to 0.05.10 World oil demand is a function of price and GDP growth which in turn is endogenously determined as a function of the oil price. The rate of GDP growth is influenced by both relative and absolute change in price. The model incorporates short and long-run price elasticises of demand which were estimated econometrically. The World Oil Equilibrium Model is used to analyse the impact that reduced investment in the upstream oil sector in MENA countries would have on international oil prices and on global oil demand. It draws on the existing World Energy Model to calculate world oil demand, keeping the same sectoral and regional breakdown as in Reference Scenario. The World Oil Equilibrium Model is based on the following equation for the global oil balance: S MENA  S Non MENA P, D DWorld ( P, GDP( P ))

where

SMENA = oil supply in MENA countries (fixed) Snon-MENA = oil supply in non-MENA countries. P = average IEA oil import price = technology parameter. D = world oil demand.

8 A number of experts, including Dermot Gately of New York University, assisted with the specification of the model and estimation of coefficients. Experts from the Statistics Norway and from the Research Department of the International Monetary Fund also provided input to the analysis. 9 For a detailed simulation model, see Gately, Dermot (2004), “OPEC’s Incentives for Faster Output Growth”, Energy Journal, Vol. 25, No. 2, IAEE, Toronto. 10 In the World Oil Equilibrium Model, the value of the parameter was at the conservative (lower) end of the range.

The equation below describes the non-MENA supply curve. The parameter measures the responsiveness of oil production in non-MENA countries to price changes. The higher the value of this parameter, the greater the ability of nonMENA countries to increase oil production when supply in MENA countries is reduced. St

where

§ S t , RS ˜ ¨ ¨ ©

· ¸ S t 1, RS ¸¹

S t 1

1D

§ Pt ˜¨ ¨P © t , RS

· ¸ ¸ ¹

D

S = oil supply in non-MENA countries. RS = Reference Scenario. = technology parameter. P = average IEA oil import price

Gas Module In the Reference Scenario, gas fields in MENA countries are also analysed according to the field by field analysis described above for oil. Non-MENA gas output projections are based on the level of ultimately recoverable resources and on a depletion rate. There are some important differences from the oil module. In particular, three regional gas markets are considered — America, Europe and Asia — whereas oil is modelled as a single international market. Two country types are modelled: net importers and net exporters. Once gas production from each netimporting region is estimated, taking into account ultimately recoverable resources and depletion rates, the remaining regional demand is derived and then allocated to the net-exporting regions, again according to recoverable resources and depletion rates. Production in the net-exporting regions is subsequently calculated from their own demand projections and export needs. Trade is split between LNG and pipelines according to: x x x

The terms of existing long-term contracts and the pattern of LNG and pipeline projects, under construction or planned. The less costly option. Minimisation of transportation distances.

In the Deferred Investment Scenario, the new level of gas prices and GDP affects global gas demand, which in turn has an impact on MENA and non-MENA production. Moreover, MENA associated gas production is affected by the change in oil production relative to the Reference Scenario.

Coal Module The coal module is a combination of a resources approach and an assessment of the development of domestic and international markets, based on the international coal price. Production, imports and exports are based on coal demand projections and historical data, on a country basis. Three markets are considered: coking coal, steam coal and brown coal. World coal trade, principally constituted of coking coal and steam coal, is separately modelled for the two markets and balanced on an annual basis.

Investment in the Fuel Supply Chains Projections of investment needs in the fuel supply chains are based on the methodology reported in the World Energy Investment Outlook 2003. This involved, for each fuel and region, the following steps: x

New-build capacity needs for production, transportation and (where appropriate) transformation were calculated on the basis of projected supply trends, estimated rates of retirement of the existing supply infrastructure and decline rates for oil and gas production.

x

Unit cost estimates were compiled for each component in the supply chain. These costs were then adjusted for each year of the projection period using projected rates of change based on a detailed analysis of the potential for technology-driven cost reductions and on country-specific factors.

Incremental capacity needs were multiplied by unit costs to yield the amount of investment needed.